Hubert H. Humphrey Building
Room 705A
200
Independence Avenue, S.W.
Washington, D.C.
Public Health Surveillance Overview
Standardizing Surveillance Data for Immunizations
Reports from Subcommittees and Work Group Action Items
DR. DETMER: I would like to call us to order. We will start with the usual introductions around the room, and then we will review the Agenda and get into the day's work.
I'm Don Detmer, University of Virginia, and in my last day as Chair of the National Committee on Vital and Health Statistics.
MR. SCANLON: I'm Jim Scanlon from HHS, ASPE, and I'm the Executive Staff Director for the full Committee.
DR. LUMPKIN: I'm John Lumpkin, from the Illinois Department of Public Health and member of the Committee.
DR. STARFIELD: I'm Barbara Starfield from Johns Hopkins University, and member of the Committee.
DR. IEZZONI: I'm Lisa Iezzoni from Beth Israel-Deaconess Medical Center in Boston.
DR. GELLMAN: I'm Bob Gellman, a Privacy and Information Policy Consultant from Washington.
MS. WARD: Elizabeth Ward, member of the Committee from the Washington State Department of Health.
DR. AMARO: I'm Hortensia Amaro from Boston University School of Public Health and a member of the Committee.
DR. DAVIS: I'm Tish Davis, I'm an epidemiologist and I'm the Director of the Occupational Health Surveillance Program with the Massachusetts Department of Public Health.
DR. BIRKHEAD: I'm Gus Birkhead. I'm an epidemiologist and also Director of AIDS Programs at the New York State Health Department, and the Past President of the Council of State and Territorial Epidemiologists, and professional organization of epidemiologists working in state and local health departments.
DR. KOO: I'm Denise Koo, the Director of the Division of Public Health Surveillance and Informatix in the Epidemiology Program, Office of the CDC.
DR. FRIEDMAN: I'm Dan Friedman with the Massachusetts Department of Public Health, and a member of the Committee. And I'm not an epidemiologist, and it's not my birthday, either.
DR. TAKEUCHI: I'm David Takeuchi, Indiana University, member of the Committee.
MS. FYFFE: I'm Kathleen Fyffe, member of the Committee. I work at the Health Insurance Association of America.
MR. STREIMER: I'm Stewart Streimer, Health Care Financing Administration, liaison to the Committee.
DR. COHN: I'm Simon Cohn, Kaiser Permanent and a member of the Committee.
MR. BLAIR: This is Jeff Blair, Medical Records Institute, member of the Committee, and I'm not going to tell you whether it's my birthday or not.
DR. HARDING: Richard Harding, child psychiatrist from South Carolina, member of the Committee.
MS. FRAWLEY: Kathleen Frawley, American Health Information Management Association, member of the Committee.
MS. COLTIN: Kathy Coltin, Harvard Pilgrim Health Care, member of the Committee.
MS. GREENBERG: Marjorie Greenberg, National Center for Health Statistics, CDC, and the Executive Secretary to the Committee.
DR. DETMER: Dale?
XX: Dale (?), Data Policy Staff of the Committee.
DR. BALL: Judy Ball from the Substance Abuse and Mental Health Services Administration.
XX: Annetta (?), Staff of the Committee.
XX: (?), CDC Atlanta, Staff of the Committee.
DR. ABERNATHY: Susan Abernathy, CDC, Atlanta.
DR. YASNOUGH: Bill Yasnough, CDC, Atlanta, Staff to the Committee.
DR. HAYDOCK: Sandy Haydock, Health Care Financing Administration and staff to workers.
DR. SEINDEL: Steve Seindel, CDC, Staff to the Committee.
XX: Sue McLelland(?), CDC, Staff to the Committee.
XX: Paul (?), Health and Medical Foundation.
XX: Michael Beeby, American Medical Association.
XX: Tom (?), American Medical Association.
XX: (?) HBOC.
XX: (?)
XX: Mike Soto,
XX: Louis Gordon, (?).
XX: Jack McDonald, I'm with (?)
(MANY MORE NOT AUDIBLE).
DR. DETMER: Okay, thank you. Before we do move to our full Agenda, a number of you that weren't with us last night missed quite a wonderful dinner and evening.
We had really an exceptional dinner, or Lynette and myself, and also David Jones. It turns out that Dr. Jones, who is visiting us from the U.K., it was his birthday, too, so we even had birthday parties and songs and such.
My wife, Mary Helen, joined us and I want to express appreciation for, really, a wonderful evening. I don't think I'll have a better birthday, actually, it was wonderful.
And Lynette authorized me to say her -- appreciation as well. So, it was an exceptionally fine evening.
Today, we have a couple of panels, and then after lunch, we'll have our usual effort to try to reach closure on some of our pronouncements, and then some items discussing the 50th anniversary of implanting and so forth, and dates, before we adjourn.
I have asked, actually, Dr. Friedman to put together, with the help of staff, a group of players to give us an overview of public health surveillance, and I think what I will do is just turn the floor over to you, Dan, and let you introduce the group and get them underway.
DR. FRIEDMAN: Thank you, Don. Let me just start by making a couple of very, very quick remarks, and start by saying that the functions of public health surveillance will undoubtedly continue to be crucial in any emerging national health information infrastructure.
Having said that, the nature of the data and the nature of the data collection will probably change. What we're going to be focusing on this morning, is trying to provide an overview of the range of public health surveillance activities, and of current health information issues, related to public health surveillance, especially as they relate to standards and integration.
You will be hearing that public health surveillance in this Country possesses a rich history; it's multifaceted; and it's largely effective in its primary mission, in directing public health intervention activities.
Having said that, my guess is you are also going to be hearing that public health surveillance in the U.S. is disease-specific, it's event-specific, it is minimally integrated at the individual person level, minimally integrated over time, at the person level, and that we have not a unitary public health surveillance system at the national level, or at the state level, but essentially, a collection of systems that mostly fit together, but sometimes don't fit together particularly well.
We are going to be starting with my friend, Denise Koo, who is the Director of the Division of Public Health Surveillance and Informatix at CDC, and Denise will provide an overview of surveillance in the U.S.
And then we will be moving to Gus Birkhead, who is the Director of the AIDS Institute at the New York Department of Public Health and as he mentioned, the immediate Past President of the Council of State and Territorial Epidemiologists.
And then, to my colleague and friend, Tish Davis, and Dr. Davis directs our Occupational Health Surveillance in the Department of Public Health?
With that, let me turn it over to Denise.
DR. DENISE KOO: Good morning, and thank you for the opportunity. I just want to start with a comment about why we are here.
I think, obviously, NCVHS, as you all know, is a very important group, thinking about the future of health information, and you know, I'm sorry I missed the Australian presentation yesterday, but I've also followed your activities with regard to the National Health Information Infrastructure Vision for Health Statistics.
I think that, in a way, we want your help. I think that too often, people leave this public health surveillance function out when they are thinking about the future of health information systems, and they think only about research, and not about the public health practice function that this supports.
So, I want to make sure that NCVHS takes this essential function into account in your vision. And I think we also want your help in sort of advocacy, because I think that people don't understand. They hear the word, surveillance, and they think of CIA, and they don't want to share information, and so, we thought that you-all would understand and help us to sort of articulate this well, because public health surveillance is a cornerstone of public health practice, and the data to provide information, that's really essential for making public health decisions and taking public health action. So we are going to give you a quick overview of this.
I am going to start with the definition of epidemiology, to remind you, it's the study of the distribution and determinants of health-related states in specified populations, and the application of this study to the control of health problems.
And why did I start here? Because epidemiology, as we consider it, a basic science of public health. And because of the distinction again between clinicians and public health practitioners, the clinician tries to determine what kind of disease a given person has, while we try to determine what kind of people have a given disease, to determine risk factors, opportunities for intervention, to decrease the prevalence of disease.
So, what's the purpose of public health surveillance? It's to assess public health status; to define public health priorities; to evaluate programs; and to stimulate research; again, an essential element of public health practice.
And see here, where it fits into public health practice. We are doing surveillance on an ongoing basis to try to determine what are the problems. Once we identify problems, we do etiologic studies to try to identify, what are the risk factors, what are the cause of these problems?
Once we have done that, we try to figure out what sorts of interventions would work; we evaluate them.
And after we have found interventions that might work, we need to of course implement them. How do we do that?
Once we have implemented the programs, of course, we have to circle back down and continue to do ongoing surveillance to determine, did the programs have an effect? Did they work?
So, as you might guess, some of the characteristics of surveillance, in order to do this, are that it needs to be systematic; it's not random, it's generally meant to cover or represent entire populations. It is ongoing. It is not a one-time occurrence or a study. We have to collect the data, analyze it, interpret it, and disseminate it. And this is actually a very important point; it has to be linked to public health practice, or action.
So an example of medical practice changing due to new surveillance data was the fact that, a few years ago, based on looking at data at microbacterium tuberculosis isolates, there was a recommendation to change to an initial four-drug therapy in areas where the prevalence of isonide-resistant was at least 1 percent.
So, I am not really going to over this again. It's a longwinded sentence. We scientists always want to put everything into one sentence. Ongoing systematic collection, analysis, and interpretation of data.
And the main point here is to show you that it's all different kinds of data, regarding agents or hazards, risk factors, exposures, health events. I think you will hear a little bit more about that from Gus and Tish. And again, there is that link to public health practice. Surveillance is really information for action. It's got to be regularly disseminated and used.
Next, I am going to point out and provide some examples of some of the traditional uses of public health surveillance. It is traditionally used to estimate the magnitude of a public health problem -- and I'm going to provide some graphic examples of this -- and to follow trends, to determine the geographic distribution of an illness. To portray the natural history of a disease. To detect epidemics, or define a problem. To generate hypotheses and stimulate research.
And really, I just want to emphasize that again. Analysis of surveillance data often generates questions for further pursuit by researchers, but surveillance is not intended to serve as etiologic research.
It is also used to evaluate control measures, to monitor changes in infectious agents; to detect changes in health practices, and to facilitate planning.
So, we just picked a few examples in each of these areas. And really, for the first one, to estimate the magnitude of a public health problem, you could pick just about anything.
Here is an example from our Notifiable Disease System, an underestimate undoubtedly, but just an estimate of cases of infant botulism.
Here is an estimate of ectopic pregnancy, actually using a different data source.
And here is yet another estimate of incidents and mortality for lung and bronchus cancer.
There you can see, the geographic distribution of hepatitis-A cases. The people in the states need to know what's the problem in their own states, and we don't need just national data. All surveillance, as Gus I'm sure will mention, is local.
There you can see the natural history of chickenpox, which is actually not a nationally notifiable disease, but we do get data from 22 states. And it's obviously very important right now, with the licensure of varicella vaccine, but because we don't have the data, we are actually going to be using varicella deaths, temporarily, although actually in combination with the data that we do have.
There you can see the surveillance data detecting an epidemic here, an outbreak caused by contaminated pasteurized milk.
You can use surveillance data, also, to generate hypotheses and stimulate research. You will see here that there were measles outbreaks in the late eighties, early 1990s, and through investigation, it was determined that there needed to be an addition of a second measles vaccination for children.
Here is another example that we are using to sort of stimulate research. These are pertussis national notifiable disease data, and you can see the age distribution. There are still a fair number of cases occurring among people who should be fully vaccinated for pertussis. So it's raising a lot of questions.
Are these people not vaccinated? Is the vaccine not effective? Should there be another dose of pertussis? What's going on here? Has pertussis changed, has the bacteria changed?
You can see the surveillance data being used to evaluate control measures.
Here, we are evaluating, or trying to determine changes in infectious agents.
I put this in here, actually, just to remind me to mention that, of course, surveillance data more at the local and state level is used to identify cases, individual cases, for treatment of a communicable disease, or to track contacts for prophylactic treatment.
Here we can detect changes in health practices. This slides shows the prevalence of mammograms among women, age 40 and over, and this is from a behavioral risk factor surveillance survey.
There's a hole in the slide projector here, so -- and here is another one where you're seeing the changes -- this is an old slide, actually, it's since gone down, but this is the changes in Cesarean deliveries as a percentage of all deliveries, but actually, this is very old, it's gone back down. This is from the National Hospital Discharge Survey.
And again, surveillance can be used to facilitate planning. As you can see here, in the late eighties, early 1990s, there was an increase in the incidence of tuberculosis, which lead to advocacy for, and obtaining more resources to control tuberculosis.
So, just to reiterate and to go over the fact that you can probably tell that we need lots of types of data to do surveillance. We need data about hazardous agents, whether chemical or biological; data about behaviors, exposures, health events or conditions, vital statistics. We also need information, of course, on the healthcare system and its effects on health service delivery. Information on public health programs, including coverage of preventive services.
As you might expect, the sources for these data are extensive, and include medical records, birth and death certificates, laboratories, environmental monitoring systems, administrative records. I know you're talking about hospital discharge, just in the previous subcommittee session, workers' compensation. Police records are important, individuals, animals in vectors, and the list goes on and on.
And actually, Public Health doesn't always collect the data ourselves; in fact, police collect information about homicides and the relationship of the perpetrator to the victim.
The National Highway Traffic Safety Administration collects information about motor vehicle crashes and blood alcohol levels. And it goes on and on.
This is a slide from a different talk, and it really is less about data sources, more about methods for obtaining the data, and we often, for a given disease, will probably use more than one method, and try to combine them for a more complete picture of the problem. And it may be that some problems are best approached using one or more of these sources or methods.
I just wanted to explain a little bit about the Notifiable Disease System. I think because people sort of, we get questions a lot all the time, why doesn't CDC mandate this? Why doesn't CDC say, collect data about this
Well, reporting is actually mandated by state law or regulation, not at the CDC or the federal level. And the healthcare providers and laboratories report to the local health department, and this is often on paper, or by a report card.
The county then submits the reports to the state. And again, frequently, this is still by paper. The state reports voluntarily to the CDC, and these reports are generally reported through the national electronic telecommunications system for surveillance, although not all of them are.
So, again, back to the salmonella data. I just wanted to show another example, not just of the Notifiable Disease System, but also of the role of the laboratory, and that's why I mentioned both Notifiable Diseases and laboratory as a separate data source, because of course we can get sera type information, we can get strain information, and we can look at whether, for example, with roto virus, whether it's covered by the current vaccine.
I'll talk a little bit about how we usually obtained information for surveillance, and traditionally, we have built our own freestanding information systems for surveillance, and Dan alluded to. And the reason this has happened is that, public health is categorically funded.
We are funded for AIDS, tuberculosis, sexually transmitted diseases, and each of these areas has their own program-specific partners, with the specific programmatic needs. For example, case management, directly observed therapy for tuberculosis.
There are often distinct data sources. Not always, but you've got your STD clinics located one place, your TB clinics in another place, and there's been limited resources for infrastructure, for something that has just the glue to linking better, doesn't sell very to politicians, although bioterrorism is certainly selling right now, and we really are using that as an opportunity to build the infrastructure.
A couple of other factors are of course the lure of technology, and the fact, that working independently leads to quicker results. If you only work with your own partners, it's a lot faster.
So, this approach has led to a multiplicity of information systems for surveillance, with variable data input methods, data sent by the same person to different programs at CDC.
Data that is useful for a given disease may often end up at distinct locations at the CDC; for example, laboratory data and reports of notifiable diseases; morbidity data and mortality data. And it is difficult to analyze them because of different coding standards and software.
There are variable methods for accessing data and there is no single point of contact at CDC. So it becomes difficult to evaluate problems across the programs. Our state and local partners are very well aware of this, because then they have to access different, multiple and different information systems to assess the overall health in their community.
So, briefly, some of the activities that we're undertaking to try to get over this issue: health insurance portability and accountability act as motivators, and we have started to work more with standards development organizations.
We know that we are going to want to access clinical data, and not just financially driven data, so we have particularly started working with the clinically relevant standards development organizations, and fortunately, for this audience, we don't have to sort of explain all the acronyms.
I usually got to surveillance or public health audiences and I have to go through every one of these, but we are working with these standards development organizations, and I know you will hear more about the Immunization Registry activity in the next hour.
We are very interested in computerized medical records, because we are currently fragmented, and our systems are separate from medical systems, and we know that they can't be in the future.
So, again, briefly, some more activities. We are looking to try to capture electronic laboratory data as it is generated, and we are concerned about some issues, again, like you have talked about, like our inability to get hold of demographic data with the laboratory data.
We have various projects going on in the states, actually, for integrated reporting for managed care and Medicaid, some of which we support with funds at CDC.
Data elements for emergency department systems. The Injury Center has worked with the standards development organizations to try to standardize data that might be collected from emergency departments, because we have not traditionally collected data from emergency departments, successfully.
Immunizations, you will hear about. Our Cancer Registry folks are working with HL7, and all of these activities really challenge us to be very specific about our data needs, because we are not just building our own information systems now; we need to be able to say them, and specify them in a way that other information systems might be able to meet them.
These slides are actually, a couple of them, from a different talk. We have a Public Health Informatix Fellowship, then we have a Windows-based epidemiologic software tool that we have often used for surveillance in the states.
Another major activity is of course our surveillance systems integration project, which is meant to link, integrate, into a system, or a system of systems, some of various infectious disease surveillance systems.
And components to that are core data for electronic reporting -- and again, this is meant to have broad applicability to our other surveillance systems. Unified user interface, so you don't have to hit F10 in one system to save a record, F10 in another system to delete the record.
And of course, a secure data communications pipeline. We want to take advantage of the Internet, and we have not traditionally done that in our information systems for surveillance. And of course, NCHS is working with the other agencies on the state and federal level to integrate some of the surveys. And Dan, I think was going to talk about electronic death records.
Now, these again are stolen from another talk I give, and obviously, you all know about these two things, your Health Statistics Vision for the 21st Century, and the talk that you have had about the National Health Information Infrastructure.
So, again, I was just ending with them, because of the fact that we recognize the interdependence of public health surveillance and health information systems, and I really feel that NCVHS can help us by building surveillance into your vision for the future.
As I mentioned, too many people are afraid of this surveillance function and of our monitoring. It's not a research function, it's a daily sort of public health practice function. We need data to our jobs, and we need sort of your help, also, and I know that you have been encouraging this with regards to privacy and confidentiality issues, that it is important to share data with public health, for us to do our job.
So, we are just trying to give you a brief exposure sort of to the tip of the iceberg of public health surveillance today, and certainly, we'd be happy to spend more time with NCVHS in-depth or in various different areas at CDC.
DR. GUS BIRHEAD: Well, good morning. I'm Gus Birkhead from the New York State Health Department, and I appreciate the opportunity to come and talk and a little bit about a state perspective on public health surveillance, to kind of balance and expand on what Denise has been talking about.
I do have a handout which people should have, which have most of my slides on it, in case you want to look at, refer to this material, again.
Let me start out, though, with sort of a graphical or a picture description of what our national public health surveillance system looks like, and this is a series of concentric circles, focusing in on the federal government level, the Centers for Disease Control, the primary public health agency, but I think, depending on where you sit -- And then the state health departments and local health departments sort of branching off from that.
In New York, we have a slightly different view of the system and that focuses more closely on the New York State Health Department, where we get a lot of information in from counties and send information off to CDC, but the effort is really one that is focused on ourselves, but I think what I would like to argue, and the impression I would like to leave you with from this talk is on the next slide, and that is, surveillance system as it really appears, is one that is based at the local level.
This slides shows the county health department level as the focus with the state health department and CDC off to the side. And this is really, really, I think the message that I want to bring to you today that, from a public health perspective, for surveillance, all surveillance is really local, to paraphrase Tip O'Neill.
And let me talk a little bit, then, about what I mean, what public health surveillance is at the local level, to try and give you an idea of how that applies.
Sorry if this doesn't show well, but it's on the handout. This slide looks at features distinguishing state and local surveillance efforts from national surveillance efforts, and I wanted to highlight a couple of things.
First of all, because health really isn't mentioned specifically in the U.S. Constitution, historically, states have developed statutory codes around health, and the responsibility has really been left to the state health department level.
And for that reason, surveillance and control activities are really authorized in this Country at the state level, many of them in state statutes, and I will show you an example in a minute.
The other features that make it different at the state and local level from the federal level is at the state and local level, there is a direct link, often, between surveillance and prevention and control activities, and for that reason, timeliness of data, immediate access to data in some cases, is very important, and I think a key difference, as well.
Now, the differences are also highlighted in the different purposes and different methods that are used at the state and local level. At the state and local level, the purpose often of doing surveillance is to link to immediate control efforts, or to program evaluation, and therefore, the methods are real time access to data, in many cases, in all cases that are occurring, or to a sample of cases.
At the federal level, the purpose is more, I would argue, to monitor national trends, detect emerging problems, demonstrate the need for resources, and the methods are different. There is often, as Denise mentioned with the National Notifiable Disease Surveillance System, an aggregation of state and local data at the federal level, which is used to assess national trends and differences by region of the Country, and there is also a much heavier reliance on national sample surveys, which provide good data on a national level, but are often not helpful at the state or local level, to try and guide prevention efforts.
Let me just give some examples of how, at the state and local level, we link the surveillance activities to control activities, and how really you cannot distinguish, I would argue, surveillance from the control activities.
And let's take it at several different levels, at the individual case level, we need reporting, for example, of botulism cases, because there is a public health role to ensure correct diagnosis and treatment of those cases.
We have a system in this Country of limited access to botulism antitoxin from the federal government, that requires clinicians to go through a state health department, to be sure that the diagnosis has been correctly made, or at least being pursued, and antitoxin is then released, only at the request of the local public health authority.
So, that's an example where individual case management, there's a public health role in carrying that out and assuring the correct diagnosis. Of course, with botulism, you then also want to link to the efforts to figure out where this came from, a contaminated food source or whatever; that has to happen on a case-by-case basis as the local level.
In terms of contacts of cases of disease, there is also a public health role, and one of the reasons to have diseases reporting is to assure that contacts of cases are treated appropriately, and I think a good example is a case of meningicoccal meningitis in a daycare center. There is often need for other children in the daycare center to receive antibiotic prophylaxis quickly, within a matter of hours or a day or two, in order to prevent further cases from occurring.
So, this is where surveillance reporting from the clinical community to the public health community results in immediate community control efforts.
At the community level, there is also a public health role for removing the source of outbreaks, and so for example, reporting of e.coli ON57 cases from the laboratory and from clinical settings to the local and state health departments, can result in determination that an outbreak may exist, and determining the source and removal of that source of an outbreak, for example, a contaminated meat product.
And finally, at the program level, we use surveillance data to, as Denise indicated, evaluate program efforts, and a good example is the Behavioral Risk Factor Surveillance System, which is a telephone survey supported by CDC that each state conducts to look at things like smoking prevalence within their states, to give states immediate hands-on access to their data on what is happening in their population, to be able to respond to changes or just to see that the programs are working effectively.
So, I think this slide summarizes the types of, the reasons why we need, in public health, immediate surveillance data at an individual and community level to be able to perform our public health functions of preventing disease.
Let me just comment briefly on the National Notifiable Disease Surveillance System, because that is one of a number of surveillance systems at the national level which is built from data that comes from surveillance at the local and state level.
Basically, the way the system works is that there is a recommended list of conditions under surveillance at the national level, which is agreed to by CDC and the Council of State and Territorial Epidemiologists. CSTE.
That list, however, is only advisory and a recommendation. What then happens is, each state with statutory authority has to go back and go through their statutory or regulatory process to add conditions to their own list of required reportable conditions.
States attempt to collect standard data elements and again, with collaboration between states and CDC, we have been able to establish a standard set of data elements, a standard mechanism for gathering and reporting data, and on each disease, some specific questions that need to be answered for that, specific to that particular disease.
For example, with botulism, what the clinical findings were; what the treatment was, and what the source was. So, each disease has a slightly different set of supplemental questions that have been agreed upon in a standard format, and that we then report that to CDC.
Now, the reporting to CDC of individual level case data does occur, though the data are reported each week in the MMWR, but I want to just highlight the fact that those case level data go to CDC without identifying information and are sent, really, on a voluntary basis, by state health departments. There is no statutory basis under which those data are sent to CDC.
The next slide gets at some issues in state and local surveillance, and Denise I think hit upon a couple of these. One is integration of national surveillance systems, and in your handout, if you flip back through, you will come to a diagram that, unfortunately, got inverted somehow in the printing process, but gives you a sense of -- and I afraid, it's hard to even read, but at the top, you'll see a circular of four little computers in it at the state level.
This is indicating four different surveillance systems data sets -- Dr. Lumpkin has it right. You need to look through the back of the page. The one thing I learned from this is that CDC is the same frontwards or backwards. I'm not sure what that is saying.
The point is that the current system that we have is not an integrated system. At the state level, CDC has provided us four different categorical software packages for STD, for TB, for HIV, that the states then have to handle separately and report the separate sections at CDC, in addition to the National Notifiable Disease Reporting scheme, and what the lower portion of the diagram tries to indicate is that, what we are moving towards is a single, integrated software system, where data can be moved between these various systems for related diseases, sent to a single point at CDC where the CDC programs can access them. It's sort of at its most critical nature.
If you go down to the local health department level, you may have a single public health nurse who is doing the STD, tuberculosis, and HIV work, to have to function with three or four different software systems just doesn't work when you bring the surveillance down to the local level, so for a long time, there has been a push and It think now a lot of movement at the CDC level, trying to integrate these systems so they look and feel the same, so they can interchange data and you can bring data on a single individual together, across several diseases, which I think will be very helpful.
At the end of the packet, just to highlight, I've got a couple of tables which show a variety of different surveillance systems, and a variety of different disease categories; infectious, chronic, injury, occupational, just to make the point that a single individual data system often contains that apply across the spectrum of public health disciplines, and to the extent that we can standardize, those data will become much more useful for everyone in public health to use.
Let me just flip back to this slide for a second and just say that, the inclusion of standard public health surveillance data elements in medical databases is also critical, as we move towards managed care and reporting of laboratory-based data, we need to have standards and again, at the state level, there has been a lot of work with the HL7 and other standards-setting organizations to try to be sure that's happening.
The third bullet on the slide relates to adequate noncategorical resources for training and for surveillance and for local surveillance infrastructure. Denise mentioned the new bioterrorism money that is going to be made available to states. The hope is, and every indication is, that those will be not tied so categorically that states don't have the flexibility to help support their general surveillance infrastructure with those funds and I think that is important.
Also, in the handouts is a set of tables from a publication on a survey that was done of state and local health departments, in terms of resources for surveillance, and I think I may actually have this on the next slide.
But this is a survey that was done several years ago by CSTE indicating that nationwide, there are only about 1,000, or 1,600 FTEs devoted to public health surveillance for all infectious diseases and a bulk of those occur in the categorical programs of HIV, STD, and TB. A total of 64 individuals nationwide, FTEs in state and local health departments, working on food- and water-borne disease surveillance.
I think highlighting the need to have more resources devoted in this area, and the next slide just shows the dollar comparisons that go along with this.
These data are a little outdated,. Things have improved, but I think the message still is that the resources for infrastructure at the local level are not as adequate as they should be.
Let me just quickly mention a proposal that has come from the Council of State and Territorial Epidemiologists a couple of years ago for really a conceptual or virtual, national public health surveillance system.
This is a grouping of -- going to be a grouping or an umbrella for all surveillance activities that are occurring. This was suggested by CSTE at its 1994 National Surveillance Conference in Atlanta. It really speaks to an overarching framework to classify all public health surveillance across the disciplines of public health.
Our hope is that it would provide visibility for public health surveillance and get people out of the categorical mindset, each program arguing for its own surveillance resources to present really a united front, in looking for surveillance funding.
It is proposed to be a collaborative process between federal, state, and local levels, to agree on what conditions should be under national surveillance and define the federal, state, and local roles, and how that should occur.
Very much, we see it as encouraging collaboration and integration of surveillance efforts for efficiency and access to new data sources, improving data quality through coordinated evaluation, training, and development of new surveillance methods, and providing a single point of access for surveillance data on the worldwide Web, for example, across public health disciplines, where if you had a question about public health surveillance data, you could go to one point and find data on any condition that you were interested in.
And I wanted to just finish up with sort of a cautionary detail about an encounter between public health surveillance and administrative data sets, which we had in New York a couple of years ago, where we thought, gee, let's look at some administrative data sets and see if we can use these data for the kinds of public health surveillance activities which I had been talking about.
What we did was look at some data obtained from a fiscal intermediary for almost a million New York State municipal employees, retirees, and their dependents, and we looked specifically at enteritis codes, gastroenteritis codes, and attempted to match specific codes with our surveillance file.
What we found, looking, like the front of the ICD Code Book, you find the enteric diseases, cholera, typhoid, salmonella, and chigella, we of course found claims for these various diseases in the data set, and looking at the rates of disease, they made sense for most of the diseases, typhoid, salmonella, chigella, was about where we thought it would be in terms of a statewide rate, but we were a little puzzled by cholera, where we had not had any reports of cholera during this year, and had 35 different patients highlighted in the data set as having cholera disease.
We looked -- the cholera codes actually break out, and the bulk of these were due to el torre strain of cholera which has not been seen in the North Americas, I don't believe, in recent years. It's the epidemic strain in other parts of the world.
We were a little puzzled by this, and looking further, we discovered that a lot of these claims had billing codes for a specific procedures, which was destruction by any method, including laser on benign facial lesions, and this added to sort of the perplexion of what was going on here.
I think what was going on is sort of a frame shift in the codes, because if you look at the code for cholera l. tor 001.1, it's very similar to the code for dermatomycosis, including infection, which might be a condition that would need laser treatment, which is 110.10, so simply shifting the decimal point I think created a cholera outbreak in New York, where none existed.
I think this really has been a theme and concern of ours in using any administrative data sets, first of all, the false-positive results that would have come through such a system would have us spinning our wheels in public health, tracking down phantom epidemics that really didn't exist, and the question which is really unanswered, is to what extent the true positive diseases show up in these data, and I think we would not at this point have confidence that we would be catching the whole picture by shifting over to administrative data sets for the kind of public health surveillance activities that we have been talking.
I think there will remain a role for the kind of public health surveillance reporting, direct reporting from providers to public health organizations for some time to come. The hope is that we can tack on to laboratory, electronic laboratory systems and other systems, collecting more clinical data, and make those useful for public health purposes.
I think I will stop there and move to the next speaker. Thank you.
DR. LETITIA DAVIS: Thank you. I appreciate the opportunity to be here this morning. When one thinks of public health surveillance, and we are certainly hoping you do think of public health surveillance, one is most likely, as you have heard this morning, to think of surveillance of infectious disease.
In recent years, surveillance activities have expanded dramatically, to include surveillance of a whole range of health outcomes and health determinants.
We do surveillance of chronic illnesses, we do surveillance of injuries, and we do surveillance of risky behaviors. And I've been invited here today to talk, to really provide an example of surveillance that extends beyond communicable disease, and I am going to be talking about occupational health surveillance, or --
I am going to be talking about occupational health surveillance, or our efforts to monitor the impact of work on health.
You have seen or heard this morning the definition of surveillance that involves a systematic, ongoing collection and analysis and dissemination of health data. The health endpoints that we are interested in, in occupational health, are here.
We are interested in occupational injuries, occupational diseases, and occupational hazards. My comments today are going to focus on occupational illnesses and occupational injuries.
I have to admit that I think that surveillance of occupational hazards would perhaps be optimal. It would enable us to identify and address problems in the workplace, before we see the adverse health consequences, but frankly, it's in very early stages of development so, again, I will be talking about illness and injury surveillance.
You have also heard this morning somewhat about the goals of surveillance. In occupational health, we are seeking to identify the magnitude and the distribution of health and safety problems; specifically, we are trying to identify high risk industries and a high risk occupations where intervention is warranted, and at the state and local levels, we are actually interested in identifying high risk work places, where work site intervention should be carried out.
We are also interested in looking at illnesses and injuries in populations, in population patterns. One dramatic finding which we have now seen consistently in a number of states, for example, is that Latino workers are at high risk of overexposure to lead in the workplace. And that has been translated appropriately into targeted outreach programs that are culturally and linguistically appropriate.
As you heard from Dan in other arenas, there is no national occupational health surveillance system. What we have are a whole variety of systems, and I am not going to go through these, but I am just going to show you these to let you know that we do have our own acronyms, and there's no time to explain them all.
At the national level, national statistics on occupational injuries are derived from data supplied from employers, who are obligated to maintain logs of occupational illnesses and injuries under the OSHA Act.
Each year, the Bureau of Labor Statistics collects the logs from a sample of employers, to generate national estimates of illness and injury rates. And in some, but not in all states, the sample is large enough to generate state-specific information.
This system provides reasonably good estimates of occupational injuries, by industry and occupation; however, it is well-recognized that the system does not capture the great majority of occupational diseases, which are much more difficult to document, and this is particularly true of the chronic diseases, which may not manifest until years after exposure.
There has also been longstanding concern about underreporting of injuries on the OSHA logs. The estimate, recent estimate, of an OMB investigation is underreporting by about 25%. And the surveillance dilemma is that the more these data are used by OSHA to target interventions, the greater incentive for under-reporting. So we have one of our classic surveillance dilemmas.
The other thing that is important to realize is that these data are anonymous; clearly, they don't have individual identifiers, but there are also no corporate identifiers. There are of limited utility in targeting work site intervention efforts at the state and local level.
There has been increasing recognition that state health agencies have a vital role to play in occupational health surveillance. There has been long recognition that the states are the lead agencies in the surveillance of infectious disease, but that recognition has been a long time coming in the occupational health role.
States are in a unique position to access our health data systems at the state level, to generate data on occupational disease, and they can also provide critically needed injury data that we can use to assess or evaluate the BLS data system.
And third and perhaps most compelling from the perspective of the states is that, state-based surveillance enables us to link surveillance actively with intervention at the local level.
In recent years, a number of state agencies, largely with funding provided by the National Institute for Occupational Safety and Health, NIOSH, have been working to build surveillance systems for a variety of different occupational health outcomes.
On the next overhead, you can see that we are using a wide variety of state data sources to track different outcomes, and I have divided these data sources into two different categories; case reporting sources, which are useful in identifying individual, or sentinel cases, for sentinel case follow-up, and data systems, some of which actually enable us to track trends over time and space.
In the surveillance literature, case-based surveillance is often contrasted with population-based surveillance that involves the use of population-based data sets to generate representative summary data; however, it is important to realize that these systems are not mutually exclusive, and I think the optimal system has elements of both, enables us to identify individual cases, and at the same time, generate meaningful summary data, that enables us to target broader-based intervention efforts.
You should know that NIOSH has funded multiple states to conduct surveillance for some of these conditions here, and we have worked to develop standardized core variables, standardized data collection instruments, and software packages to facilitate aggregation of data across the states.
I would like now to really turn to the Massachusetts experience, and to provide an example that I think underscores the importance of linking surveillance with intervention.
Massachusetts has been working since 1992 to build a surveillance system for work-related injuries to minors, and I don't mean coal miners in Massachusetts, I mean persons who are less than 18 years old.
The majority of high school students in this Country work; they work substantial hours, more than children in other industrialized nations. We know they are at risk, it appears they are at high risk, and they have been a population of workers that until recently has in large part been ignored.
In 1992, the Massachusetts Department of Public Health passed a regulation that require hospitals to report work-related injuries to the health department on at least a biannual basis, and received funding from NIOSH that year to develop a model system for surveillance.
One objective of the system was to determine whether or not we could collect, whether it was feasible to collect outpatient data on occupational injuries from hospitals. We found out that it was.
We have ten hospitals participating in the surveillance system at this time. It's a convenient sample. They submit a computer-generated hard copy report to us of all work-related injuries in persons less than 18 years old on a monthly basis.
We have combined these with data from our wooers' compensation system claims file for work-related injuries in teens that come into the health department. These come to us on a weekly basis,.
We have several other ancillary sources, but our main sources are the workers' compensation data and the hospital discharge records.
We interview serious cases, We conduct follow-up interviews with the teenagers, first getting permission from their parents. If we determine that there was a egregious safety violation, or an egregious violation of child labor laws, we talk with the family about it and we make a decision as to whether or not these cases should be referred to a regulatory agency for follow-up.
We may refer them to OSHA< we may refer them to the Department of Labor, to investigate child labor law violations -- the employer, not the case. We're referring to the work site at this point.
If it is a new or an undocumented hazard, we wills end out our own research-oriented investigators to document the hazards and provide technical assistance to the employer. And we send educational materials to all parents and teens, the parents of injured teens, and teens themselves, and I brought samples of these if you want to pass them around.
We also use summary data to target broader-based intervention efforts, and I've just brought a few overheads to give you an idea of the kind of things that we're seeing in Massachusetts. This actually was not intended as part of my talk, but I couldn't resist providing it because I think it really illustrates an important surveillance lesson.
What you see here is a blue arrow, workers compensation cases, the red are our emergency departments, and remember, this is just a sample, it's ten out of over 80 emergency departments in the State, so these are not all people coming into the system.
What is compelling about this overhead, there is only 3 percent overlap between the cases that are seen in emergency departments, and the cases that enter the workers compensation system.
So the surveillance lesson here is that when we want to document the full extent of the problem, we really need to use multiple data sources. I think we have learned that over and over again.
Let's take a look at what we are seeing. This is the distribution of injuries by industry. You see that 60 percent of the injuries that we see are in five major industries: eating and drinking places, otherwise known as restaurants, grocery stores, nursing homes, retail bakeries, and department stores.
We have used these data, we have worked, we have an ongoing education program for restaurant workers and we are just developing one for nursing home workers, as well.
We have also been able to compute rates by geographic areas in the State. And what we have consistently found is high risks, high rates in the southeastern part of the State.
We were able to use these data to generate funding from the national level, to develop a model, community-based intervention program in the City of Brockton, which is right down in the red area, on the right.
We worked intensively in that community with the schools, with employers, with physicians, with teens themselves. We have a peer leadership program in place. We have developed curricula for the schools. We have a Mayor's Health and Safety Day. There are a number of things going on in the City of Brockton, to really engage the entire community in an effort to enhance the health and safety of young workers.
The data have also been used for policy purposes. They have informed a recent effort to amend the child labor legislation in Massachusetts, and they were cited extensively in a recent report, just released by the Institute of Medicine, calling for policy and research to address the issues of young workers of young workers in the United States.
In concluding, I want to return to the issue and the themes of integration and coordination, which I think you have heard about today. It is widely recognized within the occupational health community that there is a tremendous need for surveillance planning, and to integrate our isolated systems into a comprehensive whole, and clearly rapidly changing technology -- I don't need to tell this group, I am sure -- provides challenges, but also really unique and valuable opportunities.
NIOSH has recently initiated a comprehensive surveillance planning process. I am working on that, as a representative of CSTE, to provide state input.
What I think this meeting for me today really reinforces, is that, as we seek to plan within our world of occupational health -- and that's a significant step forward, in and of itself -- this planning needs to take place in a much larger context, which is public health surveillance planning as a whole, and then we need to build a comprehensive, integrated, public health surveillance system, and again, as what I have learned today, is this is imbedded in an even larger task, which is the comprehensive and coordinated health information systems.
So, thank you.
DR. DETMER: Thank you. Thanks very much, Dan, I think these have been really very clear and useful presentations. What I would like to do at this point is just open this up for the Committee to dialogue with all of you folks, and who would like to start? Barbara.
DR. STARFIELD: Thanks very much. This was very informative of you. Just as epidemiology has spread now beyond the study of diseases and conditions; it is now a method that is applied to a lot of things, I wonder whether there is any thought about surveillance going beyond specific diseases?
Is there any thinking about in the future trying to look for clusters of diseases, just on an ecological level, and even beyond that, in trying to link conditions so that you get a person, so that it doesn't become surveillance of diseases, but it becomes a surveillance of poor health?
DR. BIRKHEAD: I think there have been efforts in that regard, particularly looking for clusters or patterns of diseases within surveillance systems, and CDC has actually taken a leadership role in trying to develop new methodologies, new statistical and epidemiological methodologies, to look at surveillance data, to detect trends where you might not have seen them.
In terms of using surveillance to look at, globally at an individual, and the various factors that impact them, we probably are able to do that on a community level, but not on an individual level. But I would argue, we still have some way to go even there, because of the categorical nature of how these systems have developed, and that's one of the challenges for us, I think, is to be able to have all -- be able to look at all these various types of data, and apply it back down at the community level.
In New York, we put together data at the community level and make it available for planning purposes, and each county does a municipal health services plan, that is their public health plan, based on their aggregate data, but it's not easy to pull it altogether, and it can't be easily applied to low, to small geographic areas, and we're often missing information like race, ethnicity, or other information, particularly on risk, other risks that would be very helpful.
So, there is still a long way to go to achieve the kind of integrated system you are talking about. I think we would all like to get there. It's a slow process.
DR. KOO: I just want to comment that we do think about those things in public health surveillance. It's been a little bit more challenging work, outside the health arena.
I think the examples that would fit what you are asking about might be that we are interested in, for example, environmental data, about air pollution. And we look at that and we look at rates of asthma admissions or discharges.
And so we are often trying to combine all these different sources to figure out, what is the true picture that contributes to disease, or -- and we also look at the chain of events that leads to a disease. So we are looking --
And that is why I mentioned look for agents. We looked at behaviors, we look at exposures. And it varies by disease, what is going to be the most effective way to look at a problem. How early in the chain of events can we get? And we of course then are still looking at outcomes, like health outcomes, to evaluate, if we intervene earlier, does that have an effect?
DR. BIRKHEAD: Maybe if I could just make one further comment of a cautionary nature, again. In New York we have embarked on mapping cancer cases, and overlaying that map with the hazardous waste sites, as a result of advocates requesting that the Department do that.
I think the real questions, in terms of the methodologies, that we need to aggregate data from different sources and what that is going to be telling us. I am not sure that mapping where cancer cases occur and where hazardous waste sites are is going to provide necessarily useful information, or will not be able to really determine whether it is useful without embarking on a study around every apparent cluster that occurs on that map.
I think we need to be a little cautious in trying to pull data together when we don't have necessarily the method to assess what it is telling us
DR. STARFIELD: I wasn't thinking so much of that, but I was thinking of clustering of diseases within individual, which could lead to elucidation of a whole variety of kinds of risk factors, and how they impact on health.
DR. BIRKHEAD: One of the other things that sort of is a break on trying to do that are the confidentiality concerns, which are different. In New York, our AIDS confidentiality is very strong. We can't even, within the Department, look at see whether that same individual is appearing in the Tuberculosis Registry or another registry, because of the strength with which -- so that's another consideration that I think we need to take into account when talking about doing the sort of thing you are doing.
DR. DETMER: Kathleen and then John.
MS. FRAWLEY: Thank you. I have to commend the Panel. That was wonderful. I could probably wind up asking lots of things of lots of people, but to try to stay focused, just on the confidentiality piece, and Gus, having working in New York for many years, I know exactly, so I am going to really address this to Letitia.
At our Subcommittee meeting the other day, one of the areas that we realized we cannot get our hands around, is workers comp. And it's an area that we have not made any recommendations to the Secretary on, because we realize that we just don't know how to get our hands around it.
So, I just found your one slide very helpful, so we are correct in assuming that we have to go, really, from the workers comp. We just can't rely on the ER records and the hospital records, that we have got to kind of backtrack through the system, through the first report of injury, where the employer is asking for information from the provider, or -- we're trying to get a sense on how we approach this, and we are going to hold a meeting on this one issue.
DR. DAVIS: Well, first of all, compensation systems are different in every state, and the confidentiality restrictions are different in every single state, so there is no one answer for every state in the Country.
What I can speak to is in Massachusetts, I mean, the point I think, what the slide demonstrated is that you will have many cases who are injured, or who have work-related injuries, who are not in the comp system, and I think the opposite is true. You will have cases in the comp system that will not appear in, for example, your outpatient emergency department data. They just won't have sought treatment there.
The slide was illustrating the need to combine multiple sources of data. How you access data, really, again varies by state. In Massachusetts, the diagnostic information, coupled with any personal identifiers, is confidential, although you can get the personal -- you can get the fact that an individual filed a claim, that's public information, but what that claim was for, is not accessible outside of --
MS. FRAWLEY: Okay, so in terms of some of the work that we need to do very quickly, NIOSH would be the appropriate group to kind of bring in and help us, or?
DR. DAVIS: NIOSH might, but since they work at the federal level, there are actually -- there are people in NIOSH who have worked with states and used the state workers compensation data. That's who I would suggest.
OSHA has also worked with workers compensation data in several states, mainly Maine, New Hampshire, and I think, Michigan.
MS. FRAWLEY: Great. Thank you.
DR. DETMER: John?
DR. LUMPKIN: First, I would like to thank the Panel for discussing an important issue. I'm particularly appreciative of Dr. Koo's presentation of salmonella and the highlight of our event in Illinois in 1985, particularly since I probably wouldn't be in my job if it wasn't for that episode, so we're appreciative of that.
I had a quick comment that, one of the things that we also are concerned about in surveillance is the fact that, I have heard another definition of epidemiology, which is, it's the science of searching for a denominator. And in that regard, when you start getting into small communities, particularly intercensul years, the rates become very unstable, because you just can't estimate really good denominators in small communities.
And we have had a lot of trouble in Illinois trying to present our data for our communities, just because we have known that to be very inaccurate in trying to estimate populations.
My question, though, has to do -- and I guess this is for Gus -- to what extent does the data that you forward to the CDC, record-level data, versus aggregate data?
DR. BIRKHEAD: Well, probably differs with different surveillance systems. In the National Notifiable Disease System, it is record-level, but it does not contain identifiers. In other systems, it may be aggregate -- it may be aggregate data.
DR. GELLMAN: Let me just follow up a little bit more on the same line, because that was where I was headed. In all of the programs that fall under the public health category, to what extent is individually identifiable data collected, and you know, I mean, and how --
And then, where does it go? I mean, is it collected and then reported in aggregate form elsewhere? And I know that the answer is going to be different, for different kinds of activities, but I am trying to get a better sense of where there are more direct conflicts with confidentiality concerns.
DR. BIRKHEAD: I think if you look in my handout, there's a table that lists a series of surveillance systems, and we could probably go through that and each one may be a little different, and I think probably each state a little different.
The first one, National NNDSS, is the one I just mentioned where individual states generally have reporting requirements that collect identifiable information at either the county and/or state level.
Those statutes also generally, to some degree, address confidentiality and I don't believe any state sends identifying information on to the federal level. It stays at the state health department level.
Now, there is an issue, and I think it is an important issue. We haven't talked about it, but to the extent that confidentiality is a concern and providers are reluctant to report, it compromises the goal of the surveillance system.
Recently, CSTE and CSDE have hosted a process with Larry Gosden at Georgetown, to develop a model state statute around confidentiality of surveillance data, and that information is -- there's a draft on the Web, actually, for comment, which I can't tell you the site at the moment, but that process has been going on for the last year or so. So, it is a very a important process,.
DR. GELLMAN: Well, I actually know about that, because I have seen some of the drafts, and I am actually troubled by the whole effort, because if you start to look at the different approaches to dealing with confidentiality, you have got disease-specific confidentiality laws.
We've got a bill from the National Associate of Insurance Commissioners saying, the insurance commissioners ought to have their own confidentiality law.
Now, there is a proposal working its way to say public health authorities, and all that's happening is, we are getting more confidentiality laws, not necessarily more confidentiality, and more overlap in conflict between them all, and everyone is in the middle and the patchwork quilt is getting more quilts.
DR. BIRKHEAD: Let me just say at the moment, at least in the public health arena, I think each state does have statutes now that address it some degree. The effort around a model public health surveillance statute was to find and bring all states to a single public health standard.
I agree that there may be confidentiality issues in other areas, which this doesn't address, but I think, you know, it's a serious issue.
The other thing that we have done at CSTE is adopted a data release policy, which we have given to CDC, that does not permit release of record-level data beyond CDC, when small cell sizes are an issue.
I'm not sure that that has ever been tested in court at this point, but it is an attempt to recognize that this is an important concern. There is a need to have these data for public health purposes. Confidentiality is an important consideration.
As I say, without it, we would probably jeopardize our ability to gather the data, because providers wouldn't cooperate, so it's -- it's a very important issue.
DR. GELLMAN: How serious is the problem of providers not cooperating?
DR. BIRKHEAD: Well, even though we have statutes that require supporting, I would say provider cooperation is critical in everything that we do.
DR. GELLMAN: But, is there any measure of the extent to which --
DR. DAVIS: No. Physician case reporting?
DR. BIRKHEAD: You mean, how often providers spontaneously report now?
DR. GELLMAN: How often do they not report?
DR. BIRKHEAD: Frequently, they don't report. I think probably, for some of the communicable diseases, there has been -- there is an analysis from a number of years ago on chigella surveillance, where only about 10 percent or fewer of the cases actually occurring made it into the public health surveillance system, so it varies.
With AIDS reporting, it is probably in the 90 to 95 percent range in terms of reporting. That's not all failure of cooperation by providers, but that's one factor.
DR. DAVIS: I just want to speak to that. It's very low, the individual case reporting on the part, for the occupational conditions. It depends --
Confidentiality is an issue, but I think the other issue is the burden of reporting, because when we have been able to utilize computer-generated billing systems, for example, in clinicians' offices, we have been able to get greater cooperation.
So, I think it is wrong to attribute the lack of cooperation entirely to the confidentially issues. There is a tremendous issue about just practicality, as well.
DR. GELLMAN: Well, I suspect that is probably a perfectly fair statement, and when you say -- if you are going to say to people that the way we are going to avoid this is by linking everybody's computer system together so we can pull all the data off, it's only going to heighten the confidentiality concerns of the public, rather than diminish them.
DR. DAVIS: I agree with that.
DR. DETMER: Jim, and then Simon.
MR. SCANLON: Two questions. Just a little bit of a follow-up. Given the under-reporting that varies, conceivably, even by disease or condition, what does the surveillance data actually mean? Are you assuming, for example, do you have to make assumptions in terms of what the level of the disease is at any time, and what exactly is happening in terms of increases or decreases? So you have to make some heroic assumptions in terms of what you know, depending on what this rate of under-reporting is.
Does it assume that the under-reporting is sort of a constant bias and that you can detect trends up or down, or are there other dynamics as well?
DR. BIRKHEAD: I think there are probably two situations. With rare diseases that are severe, like meningitis and other things, I think we feel we have pretty good reporting, because there is something that happens with each individual case in terms of the public health response.
And the outcome may be death if nothing happens, so those often do come to attention, and I think we feel we get pretty good reporting of those sorts of things, where diseases that are less severe, it is incomplete reporting, and I think what we can really gauge is simply large, long term trends.
For example, in the salmonella area, looking at the change in seer type with salmonella related to unpasteurized eggs, with salmonella enduriditis, is a trend that you can follow over years, by region of the country, and it correlates roughly with the problem in the food supply.
So, I think at a crude level, for many of these things, that's where it's helpful.
DR. KOO: And I think that is the ultimate challenge of surveillance, it's very applied epidemiology, and we do our best with the data. Periodically, we will do more in-depth assessments, where we will take a state where there is active surveillance, where there is a lot more resources, and try to compare them, but of course, there's dangers in trying to extrapolate that to the entire system, but we do sort of periodically do checks, and AIDS reporting is so complete because they have traditionally had a lot of resources.
DR. DAVIS: I just want to speak to Dr. Gellman's question, and that was, I don't think that all public health surveillance requires personal identifiers. We have systems and you only require personal identifiers if you have a case-based follow-up system that you actually -- as with infectious disease.
We have a firearm injury surveillance system in Massachusetts, where we have emergency department reporting of all firearm-related injuries, and there are no personal identifiers in that systems, because all we are trying to do is document the magnitude of distribution and the problem. So we shouldn't assume that we always need personal identifiers.
DR. DETMER: Yes, but I want to check on that. What percentage would you say, of all the things you track, do you really need know person-specific information? I know that's a tough question, but just as a cut.
DR. FRIEDMAN: My sense, Don, would be 90 percent of surveillance systems, because again, they are intervention-driven. Now, just to reiterate, most of those go to a local or state health department, and stop there.
DR. DETMER: Okay, but you answered my question.
DR. LUMPKIN: And if I can just add an additional dimension, that there are a number of systems where, for instance, reporting by providers may be low. There is also a laboratory-reporting requirement, and if you don't have a way to adjudicate those reports, you get false readings. And so there is some requirement to do that, but whether or not they need to be maintained is a different issue.
Also, there is an attempt to link databases, for example, in AIDS reporting and death certificates to determine living cases of AIDS within a jurisdiction. There is a need to be able to match for that sort of a purpose.
DR. DETMER: And are there any hard data on how many security lapses actually occur in your jurisdictions, or around the Country? Because this is one of the toughest things in all this is, trying to get some, you know, firm data on how bad this, quote, disease is, by sharing information, because clearly, at some level, obviously, the public health is threatened in a variety of ways.
On the other hand, how leaky is the system, as far as the handling of this data?
DR. BIRKHEAD: From my perspective, I don't think there is much information about how often breaches occur. Breaches for some diseases would not perhaps rise to the level of seriousness. In the world HIV-AIDS, there has been one recent, well-publicized event in Florida, where a diskette with the names of AIDS cases was released, but then I think recovered before it was widely disseminated, but a single event like that in the AIDS arena, obviously raises a huge concern --
DR. DETMER: Well, I would think this would be a piece of public health surveillance information that clearly the Country could use. Certainly, this Committee, I think, could find that quite helpful.
DR. COHN: I had a question on a different topic, so is that okay?
DR. DETMER: Yes, that's --
DR. KOO: Just to make one quick comment. One of the problems for us, I mean, as you are asking an excellent question. We don't have a lot of document security lapses, which I think is good, and so that's -- part of the problem, though, is that, you know, what's surveillance if you don't have the problem, then it's hard to sort of -- we're not documenting, we are usually documenting problems.
And so, and I think that as Gus mentioned, it's going to vary by disease. AIDS is the big example, and of course, it will be widely publicized, but I think that we have a good track record, because there are not a lot of documented problems, but that's a little bit of the other way.
DR. DETMER: Well, that's sort of the dog that doesn't bark, and then you --
DR. KOO: And right.
DR. GELLMAN: Well, could I just add a word on that? That in terms of finding, you know, abuses of confidentiality in various contexts, unless you spend a lot of resources and time doing investigative work, you don't find them, because they are often by insiders who have legitimate reasons for access.
This is probably true in any medical care facility, that it is the insiders that are the people who abuse the records, and it is very hard to find, so.
DR. BIRKHEAD: I would echo that. I think the impression is that most breaches of confidentiality occur in the clinical or care setting, where many people have access to information.
Once it is within the public health system, there are also --
DR. DETMER: I'm just trying to get an idea of the incidence of prevalence. I know it happens, it's the real world.
DR. BIRKHEAD: I don't think -- what you are hearing is, we can't answer that question.
DR. DETMER: Simon?
DR. COHN: Yes, actually, I had two questions on slightly different topics. The first one is for you, Gus, and maybe others will want to comment, and it had to do with use of claims data to do some of the work that you are doing.
Now, in my non-federal life, I am in charge of data warehousing, as well as practice clinical medicine, and so, you know, I like to see data being used for more than one purpose, and certainly, there is very little that I like less than having to create a medical documentation, fill in an encounter sheet. And oh, yes, by the way, send something off to the public health department in addition, on top of everything else. So, I like to see if there are ways to leverage.
Now, you had given us some examples of claims-based data, and then you threw in the cholera as well, of course, we can't use claims-based data.
I looked at that and I sort of, perhaps because of some epidemiological background, so I said, jeez, this actually could be useful for quite a lot of things. And maybe you could explain to me a little further about whether you are using claims-based data for some of your surveillance.
I mean, do you think it's useless? How could you better leverage it?
DR. BIRKHEAD: Yes. I'm sorry, I didn't mean to imply that it was useless. I think it probably does have uses. I think for individual rare case events, it's not helpful, partly because of timeliness, and partly because coding errors begin to become -- the predictive value of a report through that system for a rare event, may not be good, if coding is not good.
I think it is useful, and we do use it for tracking larger trends in New York in the AIDS arena. We use Medicaid billing dtaa, for example, to look at how many people are having ambulatory visits versus inpatient visits? What medications are being prescribed?
There are a variety of things on a global level, where I think claims data of that sort are very, can be very helpful.
DR. DETMER: Dan, would you like to make any comments to close this -- oh, did you have a --
MS. FRAWLEY: I'm sorry, I just have one comment, because I know our schedules, we are already over time. We are not revisiting all our work for the last two days, but I just wanted to make sure, because some of the panelists may not be here this afternoon, is that, the Privacy and Confidentiality Subcommittee is going to be devoting two days -- don't ask me exactly when, but we figure sometime in April or May, and there are three problem areas that we -- besides workers comp -- that we are struggling with.
I'm just saying this so that all of you know when I call, you can help. It's following up Don's point. We have a problem in terms of public health surveillance, the issue of health services research, and the use of IRBs, and the whole issue of preemption.
And anything any of you can help us, or at least tell me who to call, because you know, this is not my area of expertise, and I can call some colleagues in New York I know, and I know I can call on Dan, and John, but I mean, this is our challenge, and we have very little time.
So, if you know that there are studies or reports that you can feed to us, because Don's question is very valid, we're going up to the Hill, and Congressional staff are asking us these questions, and I'm like, I have no idea.
I just want to make that point, and I know Don is going to pull the plug on us.
DR. BIRKHEAD: There has been an ongoing discussion between CSD and CDC about the use of surveillance data and research, and what's the difference between public health practice of surveillance, and research? And where is the line where IRB clearance would be needed?
We have a meeting in early March at CDC with a number of states coming in to try and thrash those issues out some more, so we would be happy to contribute. We don't have a report.
Actually, I guess CDC has a document which spells this out in their view, that we are using for purposes of discussion, so we can get you a copy of that.
DR. DETMER: Dan, would you like to --
DR. FRIEDMAN: I want to thank everybody, and you know, there's a lot that's been done. We could start with -- Jennifer Zilmer yesterday started with the Crimean War. We could have started with the 17th century.
My concern is not only how much has been done, but also, my lasting fear, that in fact it ends up to less than the sum of its parts. And I think, over the next few years, that is really what we are going to be needing to deal with.
DR. DETMER: I think actually probably the first epidemiologic study was when they were trying to domesticate a horse or an animal you could ride, and they had horses and zebras to choose from, and they concluded they should choose the horse, because the zebra, when it bites, it doesn't let go. At any rate.
All right. Thank you, each, very much, and let's take a break, and we will resume in about 15 minutes.
BREAK AT 11:27 A.M.
RESUME AT 11:42 a.m.
DR. DETMER: Okay. I would like to come back to order. The book Guns, Germs, and Steel, came to mind, actually, during that last Panel. I don't know if you have read that book by Jerry Diamond, but it's a fantastic book. won a Pulitzer Prize, and I think if you ever need an argument that you need some surveillance from a public health perspective, that book will do it for you, but it's a terrifically well-written book.
We have another Panel on Standardizing Surveillance Data for Immunizations, and we've gotten started slightly later than we otherwise would have, because the last one was so terrific.
We can either, if we don't have time for -- well, if you to run right after this, we will eat a little later, but if not, we may take a break, come back, chat a little more, because we had invited some folks to come after lunch that couldn't make it into their busy schedules. So, we have a little time there in case we wish, but in any event, we are delighted to have you hear, and Susan, are you going to kick this off for us?
MS. SUSAN ABERNATHY: I will, I'll start.
DR. DETMER: Thanks very much.
MS. ABERNATHY: I'm Susan Abernathy. I'm a Program Analyst in the National Immunization Program at the Centers for Disease Control and Prevention, and a member-at-large of the Board of Directors of Health Level Seven, or HL7.
Thank you for giving me this opportunity to testify. Some of you know me in my role as Co-chair of the HIPAA Implementation Team on Unique Identifiers, however today I will report to you some recent standards implementation activities and the work of CDC's National Immunization Program, or NIP.
One of our tasks is to help states and local communities overcome barriers to implementing immunization registers.
I will start by giving you some background on immunization registries, then describe where are now, and finally, discuss possibilities for the future.
Immunization registries are confidential, computerized, information systems that contain information about immunizations in children. Most people believe that registries must be tailored to local needs, therefore, the concept of one national immunization registry is not one that our program supports.
Instead, our approach to universal coverage is to encourage community and state, population-based registries that are capable of exchanging information with each other, through a nationwide network.
Typically, registries are established at the state health department level, then connect to county health departments to their systems. Today, all but four of the 64 CDC-funded immunization projects have current or planned immunization registries. Thirty-six of these are currently operational in one or more public sites, and 11 of the 36 are currently operational in all public sites in their areas.
With the progress made in public sites, many operational registries now are ready to exchange information with each other, and to exchange with private providers. To be fully successful, all immunization providers, both public and private, must participate actively in the Registry.
When all immunization providers participate, registries can combine fragmented records from various providers to maintain one complete immunization history for each uniquely identified patient.
They can determine the date when immunizations are due, and generate phone or mail reminders for upcoming appointments. They can identify and target at-risk populations within any geographic area. They can determine aggregate immunization levels for a provider's entire client base, or for any specific geographic area.
They can integrate immunizations with other preventive health programs serving at-risk populations. They can measure immunization levels over time, providing a quick and efficient method of measuring changes in these levels, and they can facilitate rapid and accurate implementation of immunization schedule changes.
Other functions can be added to compute HEDUS coverage rates, to link provider vaccine usage with a practice inventory system, and to generate records for school, camp, and daycare.
Children typically are entered into a registry either at birth, through a linkage with electronic birth records, or at their first contact with the healthcare system.
As the child is immunized at whatever source, a report of the immunization is sent to the registry. When a child goes to a different provider, and often the parent does not have the provider history with them, the provider can query the registry and can obtain a complete history.
This allows the provider to give any vaccine needed during the visit, and reduces the chance of missed opportunities to immunize, while at the same time, decreasing extra immunizations, due to an incomplete record. However, this system depends on registries having the ability to receive and process information from many different kinds of systems.
Providers are generally supportive of immunization registries, but do not want to have to purchase special equipment or devote additional staff time in order to participate.
This situation presents quite a technical challenge. In 1995, we started working with HL7, an ANSI-accredited standards development organization, to develop standardized messages, or transactions, that would transmit our core data set.
With the help of many people within HL7, including Dr. Clement MacDonald, the effort resulted in the successful inclusion of these messages in Version 2.3, which was issued in April of 1997.
We published an implementation guide to help immunization developers of registries use the standard, and made it available on our Web site. Several of the more advanced registries began development of tools to exchange records, using this standard.
In a ceremony on July 23rd, 1997, President Clinton stated his support for efforts to raise the immunization levels of children under the age of two, and announced steps to lend additional federal support for this work.
One of these steps was to direct Secretary Shalala to start working with the states on an integrated immunization registry system. Since that announcement, a great deal of work, involving our partner organizations, parent focus groups, and provider associations, has been done.
You will hear more about these efforts when the activities are announced nationally in a few months, but one area of focus is overcoming the technological challenges.
One of the efforts directed toward the technical solutions is relevant to today's discussion. In August 1998, NIP convened a meeting of Informatix experts, healthcare system vendors, immunization registry developers, and technical experts, to plan next steps that would increase participation in registries.
The group concluded that NIP's approach should be to develop standards and benchmarks, rather than attempt to develop the software itself for registries.
The vendors who were presented, including Dr. Eisenberg, who is on this Panel, encouraged the different immunization registries to work together, to develop one common exchange protocol, rather than each registry having a slightly different requirement from the others. This would make it easier for interested vendors to incorporate this functionality into the systems they sell to providers.
The registries that were represented agreed to attempt to remove optionality from the messages, so that they could approach vendors with one standard solution that would allow unambiguous record exchange among a large number of private providers and state and local registries.
The group met on September 10th last year. The registries represented were Georgia, New York State, Illinois, Michigan, California, and San Bernardino, which has its own registry within California.
These registries had described how they planned to use the HL7 standard in their own written document. These written guides allowed a field-by-field comparison of how the transactions would be formatted.
By comparing these, we could see the differences and discuss ways to eliminate them, reduce optionality in the messages, and work toward one common implementation.
HL7's broad framework provides vendors with flexible solutions for their customers, but our kind of project with so many different participants requires some agreement on how to use the options available.
Several examples of the kind of options that resulted in the need to reach agreement are attached, on the back of your handout.
The participants agreed to give up user-defined segments called Z-segments, within HL7, and to work with CDC's coordination to agree on specific codes for the user-defined code options, and to add needed codes to the HL7-defined code tables, or additional fields to the segments within the HL7 standard.
Since this cooperative effort has begun, we have also attracted the attention of the Indian Health Service, the Department of Veterans Affairs, the Department of Defense, and Kaiser Permanente's National Clinical Information System. Collaboration with Kaiser on this is natural, because they helped to develop the original messages.
We believe this project will teach some lessons that can be applied to the exchange of a more complex, longitudinal, computerized patient record. We will continue to work with HL7 to incorporate the needed fields and codes for immunization registries.
We have now sent our first draft of the new implementation guide to the participants for their review. When it is final, we will make it available to everyone on our Web site. We plan to provide this single implementation standard to vendors, and to enlist their support by building it into the systems they create for physicians to use in reporting immunization registries.
Other activities within NIP will be directed toward educating providers about these systems, so that they begin to request them when purchasing upgrades to their systems.
We appreciate the efforts of your Committee to promote standardization in the healthcare industry. We support the use of computer-based patient records as the source of immunization data, and look forward to your recommendations in this area.
Thank you.
DR. DETMER: Thank you, Susan. Brad Tait, nice to have you here from Illinois.
MR. BRAD TAIT: Thank you. Good morning. I am Brad Tait, I'm the Chief of the Division of Data Processing with the Illinois Department of Public Health.
I am pleased to be here this morning to provide some information on issues relevant to the development and standardization of data for immunization registries at the state level.
I have been with the Department about five years, but I have more than 20 years experience in data processing, most of it with large-scale, statewide systems.
This morning I would like to provide a brief overview of the status of the Registry in Illinois, drawing some contrasts between implementation in the public sector and the private sector, and describe our experience with utilization of standards and data collection and exchange.
This is sort of a case study of trying to establish a population-based registry across a broad spectrum of the state, utilizing multiple technologies to collect the information.
The Department of Public Health has been a strong proponent of, and is committed to, establishing a statewide immunization registry in Illinois.
We are currently at the midpoint of our goal with the recent completion of the public sector implementation. The Registry contains currently about 50 percent of the State's immunization events with 250,000 children under the age of two registered. There are about 1.7 million total participants and almost 8 million shot records on file.
In 1993, in response to issues surrounding non-linked maternal and child health systems that had been developed with the categorical focus we heard about earlier, a large effort to develop and deploy an integrated system called Cornerstone, was undertaken by the Department to serve the State's public health community.
This project replaced a WIC system that had been developed in 1989, and incorporated both a family case management and immunization component to it. The system was a personal computer-based system and it is tied together at each location by a Local Area Network, and currently utilizes a dial-up file transfer process on a nightly basis.
It is installed in more than 270 locations in the State, including all local health departments, as well as community health centers and community-based organizations with WIC contracts.
In this setting, with custom-developed software deployed and managed by the State, control over data definition and content is well established. Integration of the immunization component with WIC and family case management provides a useful tool for the public health community, and the system has been well received and is heavily utilized.
There are, however, some local health departments and community health centers that had preexisting immunization databases, contained either in purchased, clinic management software, or locally developed systems.
In these cases, the databases are typically linked to billing systems, and therefore the immunization data is entered into these local billing systems, rather than the State-supplied system that populates the Registry.
The existence of these localized databases undermines the goal of a true population-based registry and creates the need for electronic interfaces to these systems.
In 1996, an effort was initiated to incorporate immunizations given in the private sector into the Registry. With the existing system having been developed to support public health, entirely different strategies had to be employed to address the private sector.
The State does not have enabling legislation that mandates participation, and we utilize an informed consent approach to Registry participation. In order to reach as many providers as possible and present a variety of options for participation, an aggressive strategy had been adopted to utilize multiple technologies for data collection and dissemination.
These technologies included electronic data interchange, fax serves, voice response units, optical mark recognition, as well as PC-based software package that was developed within the Department.
Some components are currently in pilot and others will begin the second quarter of this year, 1999, an estimated 50 percent of the State's private immunization providers are currently registered to participate.
There is a handout that describes kind of an overview of the system and as I walk through the multiple technologies, we had to be able to address nonautomated as well as automated providers. And I think it is important, as we talk about surveillance, that we understand that not all of our sources of information are electronic, and that there is a large nonautomated component out there.
Various technologies were researched and acquired and implemented to enable these noncomputerized providers to receive and communicate immunization information to the Registry.
Voice interaction or interactive voice response will allow access to data via the phone, by stepping through a voice menu, a provider can have a printed immunization record, or a State-mandated school physical form for that specific client faxed back to them at their local fax machine.
Additionally, the voice system can be utilized to generate phone calls to remind parents or guardians of upcoming or missed immunizations.
Immunization information for a patient that is faxed back to the provider contains demographic data on the client that is both printed and bar-coded. The shot history, the shot forecast, and coded fields for shots given.
These coded fields are marked and the form can then be faxed back to the Department, and software automatically scans the incoming fax, reading the bar code to identify the patient, and using optical mark recognition to record the shots given. The Registry database is then updated with the new information.
For the large providers with existing systems, we began to look into the use of EDI, and we had long recognized the need for a standards-based ability to exchange data electronically with the diverse and numerous partners in the public health system. We have heard a lot about that this morning from the surveillance panel. With more than 170 different state and federal programs to administer, the Department collects and exchanges information with laboratories, hospitals, long term care facilities, local health departments, county and city governments, private physicians, and many other entities. We have heard the long list this morning.
The administration of these programs and the data collection to support them is often hindered by fragmented manual and paper collection processes, multiple transcriptions of information, and duplicate data entry.
Ideally, there would be electronic interfaces among all of these participants in the public health system. The interfaces would allow for seamless exchange of data, eliminate redundant reporting, either by paper or electronically, and lower the cost of data capture by capturing information a single time at the point of origin, and then electronically routing data to the appropriate destination.
This scenario would improve the timeliness and compliance with required reporting of public health events, and provide for a near real time, if you would, public health surveillance system.
It was readily apparent that adopting national standards for electronic data interchange, where they existed, was the only practical means of moving the State towards this ideal environment.
Customized electronic interfaces with each individual provider are expensive to construct and maintain. With the number of entities in the public health domain, anything less than a standards-based approach is unmanageable.
With these issues in mind, the Department evaluated, selected, and implemented, an EDI-interface engine to support the ANSI and HL7 standards.
The initial use and pilot of this interface engine in HL7 at the state level, focused on the immunization registry, using the HL7 transaction standard developed by the Centers for Disease Control and Prevention.
In August of 1998, the Department successfully exchanged what we believe to be the first HL7-based EDI immunization record with the State of Georgia.
Once the infrastructure was established and the capacity to exchange these standards-based records was tested, efforts were initiated to extend the pilot to target partners with existing local immunization databases.
We learned some lessons from this process. It was apparent early that the HL7 standard as it was developed did not go far enough to eliminate the problem of customization for each interface or partner.
After exchanging the first record with Georgia, a second test was conducted with the State of Arizona where a Z-segment, as Susan mentioned earlier, or user-defined segment, had been utilized to contain additional information in the HL7 transaction.
The template that we had developed in the interface engine then had to be modified and stored as a second template to accept data from Arizona. With hundreds of potential partners, the problems associated with maintaining multiple interfaces are nearly the same as those mentioned earlier for customized interfaces between systems.
The standard as it existed addressed the structure and format of the transaction, but left the contents somewhat open to interpretation, as Susan alluded to earlier.
Further work was done and continues with several states and CDC to clarify the standard and create guidelines for its use.
Discussions with local health departments and community health centers that utilize clinic management software products, in general in the State of Illinois, has revealed a lack of readiness or capability to support EDI transactions and standards.
In the case of the new HL7 transaction for immunization records, systems have to be modified at a cost and time commitment that individual partners have difficulty meeting.
To improve the opportunity for success, continued efforts such as those underway with the CDC and software vendors and public health agencies to make the standards as clear as possible, and develop good guidelines for their use, must continue. With this type of synergy, software vendors will be more inclined to include the necessary functionality within their products.
Thank you.
DR. DETMER: Floyd Eisenberg. Nice to have you here from SMS.
DR. FLOYD EISENBERG: Good morning. I am Floyd Eisenberg, a physician-consultant with Shared Medical Systems, SMS.
On behalf of SMS, I want to thank you for the opportunity to testify before you today, on the very important subject of the standardization of surveillance data for immunization.
Before I begin my comments, I would like to introduce my company to those of you who may not know us. SMS, now in its 30th year, has focused exclusively on serving the information technology needs of participants in the health industry.
We have made it our business to develop, deliver, and support the information solutions, and help our customers meet their varied and challenging business needs.
SMS provides these health information solutions to customers in 20 countries and territories across North America, Europe, and the Asia-Pacific. Our customers include integrated health networks, multi-entity health corporations, community health information networks, hospitals, physician groups, government health facilities, managed care organizations, health benefit plan administrators and payers.
Based on customer need, our solutions can include any combination of clinical, financial, and administrative applications, enabling technologies and integration, and support services.
My testimony today will consider SMS customer-based activities for immunization tracking, as well as explore potential methods to encourage private sector interaction with local and regional registries.
SMS has over 40 customers who currently perform immunization registry and tracking functions using SMS systems for hospitals and clinics.
Many of these facilities function under the auspices of county health departments. Some of these facilities, including one state health department, use immunization tracking in conjunction with practice management and billing software.
In partnership with these customers, and based no the requirements, we have worked with HL7 code sets in additional detail, to increase the specificity of the information storage. Specific granularity has included vaccine series number, vaccination site, dosage, manufacturer, route of administration, VIS state, vaccine lot number, expiration date, patient age, and individual administering the vaccine.
Immunization records are entered once at the point of care, and stored in real time in the clinical data repository, so that any provider reviewing the record anywhere in the delivery system has direct access to the most recent service delivered.
Due to the requirements of our customer base, SMS has included this type of granularity in our health information systems, and our ambulatory clinical and billing software.
We welcome the movement to further standardization of terminology, as beneficial in this area for SMS as a vendor, and for each of our customers.
Our customers have performed immunization registry functions as health departments experience similar changes as those presented earlier. Vaccinations are not included in the registry if they are provided to the commercially insured population at physician offices and health clinics not affiliated with the health department. That connectivity is of interest to our customers working in the public health sector.
Currently, many physician offices document immunizations in a paper record, using electronic transactions primarily for billing. Integrated health networks in private sector physician organizations have also identified the requirement to track immunizations and other preventive care activities for the populations they serve.
We originally developed this feature in our practice management, billing, and counter system to meet the needs of public health clinics; it quickly caught on with the primary care providers, using our system in the private sector.
The private sector customer base has not, however, identified a priority to transmit vaccination or other information to existing local or state registries.
To be effective, a process to transmit immunization information from the private sector to an immunization or regional registry, has the following requirements.
First, transmission must provide added value for the physician or business entity, which includes receipt of information that updates the records of persons actively followed in that office or clinic.
This function carries with it a requirement for security authorization, as well as the possibility that other preventive care parameters will be included in the information returned from the registry to the local information system.
The process of updating the office record is useful to improve productivity, saving the time and expense involved in record retrieval for new and existing patients.
Updated records are also beneficial for improving the accuracy of performance measurement efforts based on preventive care delivery.
Such measures are increasingly being used by health plans and practice management organizations as an indicator for performance incentive payments to the clinician.
Other potential benefits to a clinician's practice include improved patient satisfaction, and member retention, due to improved productivity, allowing more time for patient contact and decreased waiting time for service delivery.
These potential benefits, however, require that a critical mass of information is reached in the immunization registry. As the senders, clinicians bear the cost of additional information technology function, accommodating electronic reporting into routine procedures, attempting to maintain and appropriate level of compliance, and transmission of error recovery.
Until there is sufficient reliable information retrieval for the registry, the clinician will not experience significant productivity improvement, or performance improvement, with respect to patients for whom he or she is held accountable.
The receiver, the immunization registry, gains the information needed. The initial value to the physician or business entity may initially be minimal. The transmission must be cost-neutral to the physician or business entity, as well.
As noted above, unless there is perceived value, the physician office is unlikely to accept transmission charges, therefore the transmission must occur through existing equipment and connectivity; for example, a modem and telephone line currently used for other purposes, such as claim transmission to third party payers.
The transmission must also be productivity-neutral, occurring in the background, without affecting the speed of existing office practices, and it must require no additional data input for the end user.
Potential modes of transmission include electronic data interchange, EDI, direct modem, Internet or Intranet transmission. To be effective, the transmission to an immunization registry will require standardization of provider identifiers and the identifiers used to match patients, as well as appropriate security and encryption.
Patient authorization will also be required for any information returned to update the initiating private practitioner database.
Transmission to a local or state immunization registry could be performed in batch mode, which would require storage of daily activities for replication with the appropriate server at a predetermined time.
Currently, electronic transmissions are most commonly used for claim submission. One consideration for submitting immunization delivery information is simultaneous with claim transmittal. Data from billing systems, however, require management. Bills that are never issued would clearly not be transmitted to an immunization registry.
A single service may also be billed to several insurance carriers, and finally, the guarantor, until payment is collected. The transmission to an immunization registry would ideally occur on the first claim submission for the service; the receiving system, however, would need to ascertain if the transmission were a duplicate.
In addition, claims are generally submitted as X12 transactions, where the transmission of immunization data with claims, a combined solution, such as enveloping HL7 data with an X12 message, should be addressed.
EDI transactions generate transmission charges to the physician or business entity. Connectivity to an immunization registry must be simultaneous with eligibility and/or claims transmission, or preprogrammed to occur after office hours, for no additional fee. In light of the current lack of demand, it is not clear that EDI vendors would provide this function without charge.
Similarly, direct modem or Internet/Intranet transmission would need to occur through an existing connection, also after routine hours, for no additional cost.
Healthcare Data Exchange, HDX, is a subsidiary of Shared Medical Systems, having extensive experience in delivering batch and real-time transactions as a byproduct of existing workflows, as with eligibility, authorizations, and referrals between provider, payer, and utilization review organizations, and third party administrators.
HDX processes more than 5 million such transactions each month. The key to success is standardization, and SMS and HDX have contributed many resources to the development of such standards in HL7, X12, DICOM, CPRI, and others. We believe the development of a national standard is valuable, including a national implementation guide that governs the use of standards, similar to the effort for X12 transactions identified for use under HIPAA administrative simplification, and those Susan Abernathy discussed this morning.
In summary, SMS has significant experience providing functional immunization registries and tracking functions to our customers. We encourage and support standardization to increase the value of our clinical systems.
The challenges of integrated care delivery across the continuum of care are exciting. We are pleased to be included in the discussion with the National Immunization Program, and the National Center for Vital and Health Statistics.
We look forward to continued cooperation in this and similar efforts. This concludes my statement. Thank you.
DR. DETMER: Thank you very much. Paul Tang.
MR. PAUL TANG: Thanks. I appreciate the opportunity to talk to you about the standardizing immunization surveillance data, and I am especially glad to be here on Dr. Detmer's final session, because I want to personally acknowledge and thank him for the terrific job he has done as Chair, and for the tireless efforts of this Committee, bringing us closer to the goals of Healthy People 2000. I really think it's terrific and I wish you all the best, Don, in Cambridge.
Listening to the prior Panel on Public Surveillance, it really is a deja vu. I am participating in an Institute of Medicine Study on Protecting the Health of Deployed U.S. Troops, and part of the task is on medical recordkeeping and surveillance.
I am sure all of us assume that the tri-services had a uniform health system, but let me just say that the multiplicity issues that Dr. Koo mentioned are not unrecognized in the Service. And leave it at that, So, all of your recommendations I hope you will forward on to DOD.
My name is Paul Tang. I'm Vice President, Epic Research Institute, and also Medical Director of Clinical and Traumatic Health Medical Foundation.
I thought I would begin my testimony sort of with a summary vision of the future, and then derive the standards, working backwards.
I would envision that by the Year 2010 -- sorry, Don, I had to add another ten years this thing to make it more accurate - that the majority of physicians would be using CPRS, and that we would have secure, authenticated access to composite health records.
Naturally, these records are going to include health maintenance information, including immunization data, and these systems would help remind me of preventive health services, when I'm seeing a patient in the office.
At the click of a button, I would be ordering an immunization, documenting it, updating my own record, and I would imagine I would be updating the local and the national registries with this immunization data.
At the same time, then, the CDC would have instant access to the registry and be able to study and evaluate the health status and immunization status of local and national populations.
Because I imagine our CPR to have active links with this, the CDC can in turn push back any new guidelines on immunization or otherwise, and then I could, in effect, accept those, and use my computer to help remind me of these upcoming events, and improve my performance with these nationally-accepted guidelines.
Now, technical, this scenario is very, very feasible today. Among the major challenges, I would like to just emphasize three. One is the comprehensive privacy and confidentiality legislation. Second is the unique ways of identifying every individual. And third are the data standards that we are talking about today.
Now, this Committee has been very involved in the first two, and we are providing some input on this third one. For the purposes of this discussion, I would like to sort of develop phases of implementation in how a physician would use these services.
In the first phase, there would be the following phases. There would be access to national immunization registries, followed by an integration of computer-based patient records with these registries; and finally, computer-based decision support and reminders for immunization guidelines.
In the phase, the access to national registries, and when I say, national, I mean, in the sense that Susan spoke of, these nationwide networks of local registries, just to clarify that.
In the first phase, providers would have secure access, perhaps over the Internet using a Web-based interface to these national registries. Security and authentication policies would be very similar to those conducting electronic commerce.
Amazon.com, for example, claims that they processed over six million credit card transactions over the Internet already. Each provider would have an authentication certificate issued by a trusted authority, a unique provider I.D., and have a secret password.
These passwords should be issued to the provider only after they have signed stringent confidentiality requirements, which include penalties, such as exclusion from federally sponsored health programs.
Because of the migratory patterns of consumers and patients, we are forced by the constantly changing payer environment, we obviously have to have larger geographic aggregation of the data. So consequently, as Floyd mentioned, we will need to have the providers, on a regular basis, sign up and upload their data into local and national registries.
This could be done with an automatic nightly upload, as Floyd described. In return, the providers could benefit from this electronic data collection function, by getting HEDUS compliance reports, perhaps even verified by the CDC, that would be generated from these registries.
And in addition, patient's parents could get immunization records off the Net. This would not only offload the physician office's time, but it would be a great convenience to parents, speaking as a father.
Now, the technical requirements for this work, again, are primarily standards-related. We need standardized data, separate immunization transactions, and the message formats.
Now, Susan spoke of the work that CDC has done with HL7 in Version 2.3 on the message formats. We also need the content standards, as Brad mentioned, to ensure that we transfer data efficiently, and preserve the original meaning.
The standards should include a template, a standard template, for immunization data that have core data sets, and core responses to those data fields that satisfy the needs of both the practitioner, and the public health officials.
The vendors of clinical information systems recognize the need for these standards and heartily endorse the efforts of the CDC in the project that they describe.
We also unique national, unique identifiers. I cannot let this opportunity get past without a plea for adoption of nationally unique health identifiers for all individuals. I think the National Immunization Registry is an outstanding example of a public good that is critically dependent on the ability of a robust identification method for all individuals.
To maintain accurate records of people who receive immunizations from multiple providers at multiple sites, we simply need a consistent and reliable method of uniquely identifying each person.
Once comprehensive privacy legislation is passed, I really hope that agencies will quickly move to recommend and implement such an identifier.
At the conclusion of this first phase, then, all providers would have access to important health data on their patients, and the public officials would have the needed epidemiologic data they need to protect and improve the health of the population.
Even though the access would be nearly universal by the end of this first phase, I would still have to do a separate activity to go look up this information. Phase Two should eliminate the need for provider-initiated activity, and instead facilitate behind the scenes, electronic querying of the information by the provider's CPR system.
By integrating the immunization data within the CPR, then this data just becomes another piece of relevant patient data that are used.
Technically, this is going to require the collaboration between the registry operators and the vendors of CPR systems. We would need very precise standards in place so that all vendors speak the same language.
The CPR system should be able to automatically retrieve on a need-to-know basis, information about their patients, the patients that are being seen. There would have to be new functions designed that are implemented both by the CPR vendors and the registries, and added to HL7, that would include things like, messages that make sure that we have the appropriate authorizations to see that data, and messages that would support the audit functions necessary for system security.
The benefit of this further integration with CPR systems, then, would be seamless operation for the provider. In this demanding world where we have to do more and more administrative things, we can't afford to increase the burden on the provider.
So, by the end of Phase Two, we would have integrated access and update capability within the CPR system. The next step would be to have a way for the CDC and other agencies to be able to disseminate and help us get new recommendations and guidelines in a seamless way.
This is the computer-based decision support for immunization guidelines. CPR system vendors are beginning to introduce decision support capability in their products, and patient-specific reminders, you all know, are showing the increased provider adherence with the guidelines very effectively.
I will share some of my Northwestern experience in this regard. As part of an LM-sponsored research activity, we had a large general internal medicine practice that was divided into two groups, the have and the have-nots.
The haven't got to use a CPR system and the other group continued on with their traditional paper method. What we did was measure their compliance with, among other things, immunization habits. And we did that for four years, two before implementing the system, and two after.
And I am happy to say that the two years before, we were rock solid in terms of both groups having the exact same immunization rate from one year to the next.
Following implementation, the CPR system in the one group, their immunization rate went up by 78 percent, while the other group continued at the same level established by their baseline. The health and cost consequences are well known to you all.
And this is the kind of results I think we would be able to achieve in all cases, if we had national immunization reminders,.
My experience using a commercially-developed, or commercially-available CPR reconfirms the value of computer-based reminders that Dr. MacDonald established so many years ago, we also confirmed that, although doctors may be nonperfectable, to use Clem's terms, they seem to be educable.
It is interesting to note that in our study, 20 percent of the vaccines that were given, were given outside of the clinic; hence, again, the importance of having regional registries, so that we can get information and also share information for others to use.
So, the summary. These local and national immunization registries will be valuable assets, not only to public health officials, but for individual primary care providers, and linking these registries with CPR systems, would give us a seamless way for clinicians to automatically update the registries, and also to act on the latest immunization guidelines coming down from CDC and other guideline development groups.
The technology problems are largely solvable. The human processes of setting standards and setting policies are those that we need to work on. And I believe, and I believe all the other vendors very much support the work that the CDC is doing in helping us bring standards and standard templates into this arena, and I thank you for the opportunity to testify.
DR. DETMER: Thank you, Paul, and thank you, Floyd, Brad, and Susan. I think this has been a good Panel, and what we would like to do is open this up now for the Committee to interact with you on your points. Bob, why don't you start?
DR. GELLMAN: I want to take issue with one word in your statement, Susan. I don't want anyone to take my comments as questioning the value of the registry or immunization, but one of the things you said in your statement was, "Immunization registries are confidential."
I don't know what that word means. As a matter of fact, I don't think it means anything, but I doubt that the registries are exempt from subpoenas; whether they are administrative or civil or criminal or legislative.
I doubt that the records are not available to federal, state, or local program or management auditors.
I doubt they are exempt from financial audits. I doubt they are exempt from being used by criminal investigators, by fraud and abuse investigators.
They are clearly used in providing -- the information is clearly made available to lots of participants, physicians, providers -- I look on the Illinois list and I see arrows going in two directions all over the place.
I'm not saying any of these things are inappropriate, by the way. The records are shared with health departments, regional, local, or with other states. I assume records at some level are made available in some way to researchers.
I see there is something here about schools. I don't know what schools get or how they get information. I see from the Illinois thing that they are sending postcards to people. So, my point is, I don't see anything here that's the least bit confidential.
I am not saying that any of these uses are inappropriate, I'm only taking issue with the word, confidential.
MS. ABERNATHY: I agree, and it depends on your definition, and confidentiality is an issue that we are very concerned about, and we have taken the tack that parents should be informed about registries, if their children are in them. They should know what information is in the registry. They should have the opportunity to remove their child's participation from the registry, if that's what they want.
But, it's true. The confidentiality is not any more confidential than other medical records.
DR. GELLMAN: And let me follow up on the provision of how you deal with people who object to this. I know that there is at least a significant, small but significant, minority of people who object, and I wonder if you could provide more detail, perhaps, if you have estimates, of the number of people who have objections to this, and what you do to accommodate them.
MS. ABERNATHY: During the last year, we have held several parent focus groups, trying to understand more what that number might be, because the people who show up at the hearings generally are very vocal, and yet it tends to be the same people showing up at the hearings.
What we have found is that parents very much support immunization registries, and the percentage that opt out in the states that have advanced enough to have that function in place, is extremely small. I believe it is less than 2 percent. And so I think that the one thing that the focus groups did establish, though, is that parents really do want to be informed about it, and they want to have the option, even though when given the option, they generally don't opt out of the registry.
DR. GELLMAN: When somebody opts out, what does that mean?
MS. ABERNATHY: It means that they exercise their option to not have their child included in the registry.
DR. GELLMAN: So there is no record of the child in the registry?
MS. ABERNATHY: Right.
DR. GELLMAN: Okay.
DR. DETMER: John, do you want to comment on this?
DR. LUMPKIN: Yes, let me clarify that, that not every state is the same, obviously. Some states have laws that specifically govern immunization registry, when the records may be released.
For instance, in Illinois, we do not, we are not allowed to release them to the schools, but there is also a requirement that children have immunization information given to the schools, and that exists in most states, as well as the whole school health physical form.
So, the piece of the system that Brad didn't go into a lot of detail, which is the software that we give to a physician's office, allows that physician to have their own locally-held immunization registry, their own patients.
And if the parent does not sign the informed consent to enroll them into the state registry, then that record is only held at the local level, and is not transmitted to the state. So, that's kind of the hierarchy.
Now, other states would be different. In Illinois, we feel that under our current laws, we require an informed consent. Other states have implied consent laws, which are in place.
DR. FRIEDMAN: I'm going to just end up actually reiterating what John said. They are state-based system, some of them are informed consent-based, ours is informed consent-based system, and at least some of the states, including ours, have state laws that prohibit --
DR. DETMER: Dan? Could you speak into the mike?
DR. FRIEDMAN: Yes, speaking into the mike works. I was listening yesterday morning, and it was great on the Internet, and not only that, but at the break I heard one of my colleagues who shall remain nameless say, "You've got to be careful of the mike, it's being broadcast over the Internet." It really is. But, having said that, we for example, have a state law that prohibits subpoena -- use of the data in subpoenas, so.
DR. DETMER: I had question, and I would like your comments on it, because it sounded like you are fairly uniformly feeling that this would be a very good thing for the health and wellbeing of the children, particularly.
It struck me that -- and you also advised, as I heard it pretty clearly, that you would like to see the NCVHS really, really look, not only at the standards issue, but also an implementation strategy for that, as well. Is that -- did I hear you right?
DR. FRIEDMAN: Correct.
DR. DETMER: Okay. It did strike me that each of your presentations focus very much on physician provider kind of strategy, more than necessarily how the parent, or the person, might be able to have their own data, as a matter of their own copy, if you will, and so forth. And maybe I just missed it, but, you know, is this something that you would like? I just would be interested in your reactions to that, because I mean, people move around and so forth. And it's going to be awhile before we clearly would have a ubiquitous linked infrastructure and such, and at any rate, I would just like for you to chew that over for a minute.
MS. ABERNATHY: Well, there are not only differences among the states, but as many registries as there are, there are differences in those and functions that are added to them, but there are states -- there are registries that have the function that anyone who wants a legal copy of their immunization registry, can get that from the registry.
So, that service is offered by some. It's not a uniform thing, and since the development goes on independently among the different registries, they are not all alike in that area.
MR. TAIT: In our particular case, we are not providing for direct access by individuals. We are still grappling with the confidentiality and security issues of providers and input to the registry through those mechanisms, but the patient or client can get a printed copy of their immunization history from their provider, for their own purposes, and this is something that I think that we will see more pressure to do as the registry builds and becomes more a part of everyday practice in Illinois.
MR. TANG: I agree with you, Don, and when I had proposed this registry access, I included patients or parents in that. As a matter of fact, at the Medical Foundation, we were thinking about having a patient access to their own record, and one of the biggest sellers would be being able to download and print out your immunization record for school or camp.
DR. EISENBERG: Right. Some of our customers actually do print out for each parent, for school entry, a list of all the immunizations, dates and times, and clearly, a future direction of having the patients be able, or parents be able to find that themselves, is a vision. The question is, with confidentiality, are we ready for that yet?
DR. STARFIELD: Can I follow-up on that question? Do any of you have any experience with Smart Card technology to just what Don is talking about? You know, some countries do that, and I gather they are being tried out in California, somewhere.
DR. TANG: Oh, I was looking at a different Smart Card, a Smart Card way of authenticating myself to make sure that I'm the person accessing the data. You're probably talking about having it on the card itself.
MS. ABERNATHY: There are some projects that are piloting the use of Smart Cards in conjunction with WIC clinic services, and the cards as I understand it, are able to carry the benefits that the client is eligible for, and is also able to carry the immunization history on it.
The negative things that I have heard about the Smart Cards are that they are, you still have to have the registry behind that, and you still have to have the computing source. It's not a single answer.
DR. TANG: And then you need the standards to read it and so on and so forth. And once you have a widely accessible leader, then we have the same problem as that aged disk. I mean, someone else could read that card as well.
DR. STARFIELD: Well, usually they have a PIN number that the patient has to put in, as well as the provider.
DR. DETMER: Kathleen?
MS. FRAWLEY: Okay, thank you. I'm not sure that, Susan, you'll be able to answer this question, but I just want to throw it out. According to your statement, 36 of the registries are operational right now, in one or more sites, and I followed that.
I'm trying to find out, I don't know if anybody knows the answer, and if not, if somebody could tell me who could answer it, I would be very happy.
It follows upon Bob's point. Do we know how many states have statutes or regulations protecting immunization registry data? Does anybody know the answer to that, or?
MS. ABERNATHY: We actually have, we've had a project going on now for over a year in the National Immunization Program, and Gail Horlick is a vital part of it, doing a survey of what is being done in the states, what exists in the states, in the way of protections, and also, giving us more of an ability to compare across states, because the difference is, we are trying to see how large is the problem of being able to share from state to state.
But we have a state survey that Gail has largely done, and some of the results from that are on our Web site, and the Web site, the URL is in the handout, and we have spent a good deal of time trying to keep up with the legislation over the past year, and we hope that the HIPAA legislation will help to resolve some of these at a national level, but right now, it is a really fragmented situation, to try to see what the standards are from state to state.
MS. FRAWLEY: Thank you.
MS. FYFFE: Paul, you mentioned the 78 percent increase in immunizations. Could you tell us a little more about that? I mean, is there a cause and effect relationship here, or is maybe the recordkeeping better for the CPR group than for the non-CPR group?
DR. TANG: Right. What we did was, we looked at -- it's a -- now I'm going into a little bit more detail. A lot of times, you know, the science of looking for the denominator, the public health program. From the public point of view, you survey some consumers and find out who has had one recently.
Our denominator were, each time an eligible patient walked into your office, you had an opportunity to intervene. And that was the denominator we used. And intervening could also mean counseling them because a patient can decline, or they can be allergic, or they could say that they have gotten it elsewhere. All those things would quality, so we actually did look at both the paper record and the computerized record.
Now, if there were differences in documentation between the two groups, you are correct, that we would have -- that we all rely, essentially, on documentation or patient or provider memory, so that would be a limitation of the study, but we tried to be very thorough in finding out what people did when faced with an eligible patient. Does that help clarify it?
MS. FYFFE: Yes. Did you publish that?
DR. TANG: That is coming out in March. That's in JAMIA, Journal of American Medical Informatix Association.
MS. FYFFE: Thank you very much.
MS. COLTIN: Dr. Tang, you mentioned the need for content standards, as well, and I was curious, given that the data are used to evaluate how well providers are delivering immunization services, as well as to serve as a record for school entry for parents or for camp admission, most systems do very well at capturing what was done, and usually not as well at capturing what was not done, and why. And clearly, there are some children for whom the immunization may not be indicated, due to some other allergy or medical condition.
Do these systems, and the current messaging standards, allow you to put in that an immunization was not given, and why, so that that can be accounted for?
DR. TANG: And that relates to Kathleen's question. One of the things we do is we do capture that information in a codified way, and how do you incent people to enter that?
One is, let me give you another angle. Telephone encounter, the documentation for telephone encounters goes up dramatically when you implement these things. I don't think the calls are going up, but the documentation is going up, and why would you spend extra effort? Because you actually recognize very quickly that it's there for when you need it. When, on the paper case, telephone documentation is usually never there or findable when you need it.
Once people recognize there is value in having information, and they guarantee that they can get it back -- incidentally, they can get it back when they're at home or in the hospital. They take that extra effort.
Now, the documentation on why it wasn't given, either declined or received elsewhere, is in there, so you can document, but that means that the reports that we run on you, and show you in an anonymous fashion to give you feedback, is far more accurate in your mind.
So, the first report you get is, you know, but I -- you know, they're allergic, or they weren't eligible or so on and so forth. And soon you realize that, while these reports are to help you improve your own performance, and they are anonymous, and you can see yourself compared to the group. And then you know how you can code to make sure that the reports are accurate, an accurate reflection.
So, it turns out that the ability to get it in there in a codified way, and to be able to always have that accessible to either yourself or to someone running these QA reports, is a good incentive. And people like seeing their percentages go up.
It is a combination of getting the reminder, and getting the feedback that was so powerful.
MS. COLTIN: I would think the parents would also be a motivating force, if they need to, you know, prove to a school official, then in fact, they've got the entry.
DR. TANG: That is correct. And as I say, really, for pediatricians and family practitioners, there is a lot of phone calls consumed by getting immunization reports. If we give them -- well, back to hand printouts, it does everybody a favor, and it's just, the benefits of it feeds on itself. That's one of the lessons.
DR. FITZMAURICE: This has been really enjoyable, it's a mixture of vision and reality, and it is really great when the people who are responsible for making things happen, have the vision.
I want to focus more on the market demand aspects of it. I'm looking at when built into computer-based patient record systems, you can put a date that they warehoused that there, with a section for immunization. Providers have a demand for it. States have a demand for it. And event parents have a demand for it. But I want to ask Floyd Eisenberg and Paul Tang, what is the market out there for this?
Is it built into your own computer-based patient record systems that you sell? Do they ask you to create a data warehouse portion for the immunization data? What is the demand there? Who will buy it? And is it coming, or is it here?
DR. TANG: I think there is a -- I really like the work that NCQA is doing in moving HEDUS reports more and more clinical. So, clearly, immunization is up there, mammography, pap smears, and now even smoking cessation, and documenting one, who are your smokers, but two, that you did something every time, again you are faced with an eligible smoker.
CPR tools give us, one, the reminder to do this, and two, the proof that we did it. And those are positive reinforcements.
DR. EISENBERG: Just following along on that, in a prior life I did some analysis of looking at paper charts of provider, practitioner performance related to HEDUS rates. And one -- going back to another comment, one of the concerns we had is, how to accept a refusal, say, of a vaccine, and how to code that so that what is acceptable and what is not acceptable as a refusal.
But, as far as the systems you asked about, in our ambulatory system, we do have immunization tracking built in as part of the routine, and that is sold as part of the system.
In the health information system, we have worked with different health departments to actually put it in for measurement, and also tracking at the same time.
DR. FITZMAURICE: It seems clear this is a good thing. In your opinion, is the Government doing enough, or should the Government be doing something more.
DR. EISENBERG: Well, I think what would help is when we get the immunization, and when the Advisory Committee on Immunization Practices Recommendations, a change requires currently a change in coding, so that -- in each of the systems -- so that they can be updated.
I think some way to standardize that, and we talked about that at the National Immunization Program, through some type of HL7 language, like Arden Syntax, perhaps, so that these could be implemented in systems in a more standardized way and more quickly, would be helpful.
DR. TANG: I would answer like I did last month at the testimony. If the more enforceable, the more auditable standardized quality measures the government or regulatory agencies can provide us, I think that would be very good incentives.
DR. DETMER: Simon, and then John and we'll wrap it up.
DR. COHN: Okay, yes. I just a relatively quick question, just to ask you to follow up with Kathy Coltin's questions -- I think is actually for you, Susan.
Kathy had asked a question about this issue of reasons why you don't give immunization, and Paul had answered it, I think, from the viewpoint of his information system, and the fact that there is a way to say why that isn't the case.
Now, you had also talked in your presentation about the issues of implementing HL7 differently, and the issues of data, content cross information systems.
Now, is this an area where there is standardization in HL7 and everybody states why immunizations are not given the same way, and is that specified within HL7, or is that a content issue that still needs to be addressed?
MS. ABERNATHY: Within the HL7 standards, there is a place, in the same place that you would put, if you were to give the immunization, there is a place to put that it was refused, and why. And there is a code set, and you will know why it was refused.
DR. COHN: Then there is a standard code set for that.
MS. ABERNATHY: And the way the system should work is that if that is filled in, that it was refused and why, then anything tracking your compliance level should not cause an error on your part to be reported from that.
DR. COHN: Okay. So that gets reflected out in the registries, too.
MS. ABERNATHY: Right.
DR. LUMPKIN: Well, I'd like to thank all of you for your presentations, and the reason why I suggested that this be placed on our agenda is because clearly, as we move down the road to standardization, just having an HL7 standard for immunization may not be enough, and I wanted to applaud the actions of the CDC, which was to start to bring in a group of states that are already starting to implement EDI, so that it will be done in a consistent fashion.
The message that was heard was that vendors who may not be interested in doing it one state, if there is a single implementation of the HL7, they are much more interesting in participating in developing the provider side of this system, so it certainly addresses a number of issues.
I think one issue that we haven't addressed, which may be useful for some discussion in the Privacy and Confidentiality Committee is that with states having different requirements and some states having conformed consent in others, what do you do when someone moves from one state to another?
The record should be able to flow through the system, but does the consent or the less full consent that they may have given, is that able to flow? So that the other state, which may not have a restriction on giving the record to the school, will know that until the parent signs the consent or enrolls in their state system, they are still governed by the laws of the state in which the record was generated.
There are some issues that really are addressed in the differences between state operations, but to the extent that we can have a uniform system, to the extent that these can then be integrated into the computerized patient record and other devices, and really lead to much better quality control.
Finally, as a postscript, studies have indicated that when you ask and do measures of quality, when you ask clinicians what their personal immunization rate is, and then you go back to the records and chart it, there is a great disparity, and so it's not -- you know, there's -- the issue is, they think they do it, and without some sort of assistance in doing it, it doesn't happen as much as it should.
DR. DETMER: The thing that I have found strictly also very useful about this is, is that obviously the HEPA legislation was trying to move the Country where all of it worked together to try to set sensible standards in the name of both efficiency and I think health.
And I am hearing, you know, let's keep this thing moving in a sensible fashion. We've made some progress on some of them, and some not moving as much, but some things clearly have moved. And I am hearing that this is a good thing to vendors and this is a good thing to a variety of people within the system, and I think that's also a message that we have gotten today which I find really quite, quite exciting.
At any rate, thank you very much. Why don't we break for lunch and it's -- I have ten 'til 1:00. Do you want to get back at 1:30 or quarter 'til 2:00 --
SEVERAL: 1:30.
DR. DETMER: Okay. It sounded like that was a unanimous --
(Whereupon, at 1:40 p.m, a recess was taken until 1:30 that same day.)
DR. DETMER: We had, as you will recall, written letters to the new heads of the major parts of HHS, inviting them, preferably to come actually together, to share with us their own thinking and directions for their part of the Department, in the hopes that we could benefit from that and talk with one another.
Actually, one of them had actually intended to be here, whether and when we will actually see them, I don't know. But I think it's a good idea, and I think it's worth doing, and I guess the Executive Committee felt like we should give it a shot, even though these are terribly busy people and it might not come to pass. So, at any event, it did not come to pass, at least on this round, and so I think we will just move on forward to our report from subcommittees and work groups, and work on through it.
MS. GREENBERG: Let me just say that, we did when we sent them the letter, we also mentioned the June meeting as a possibility --
DR. DETMER: Right.
MS. GREENBERG: And that was basically, was not turned down.
DR. DETMER: Well, in fact we didn't --
MS. GREENBERG: It's just so far from now that --
DR. DETMER: Well, the first one that we sent was fairly close --
MS. GREENBERG: Right. Right.
DR. DETMER: -- and the next one was fairly long. So, I don't know how it will come out, but I hope we make it.
MS. GREENBERG: Right.
DR. DETMER: So, let's go ahead and start. Do you want to report out on the HIPAA?
DR. LUMPKIN: Sure. Actually, we have a report later on in the Agenda, so I will just basically star off and say that we have had the update on HPPA, so perhaps Simon may want to give an update on the work group meeting and the upcoming hearing.
DR. COHN: Yes, I think all we want to do is to announce that there will be two sets of hearings between now and our next full committee meeting.
The first set of hearings will be in March, the next ones will be in May, March 29th and 39th. And we will focus on messaging format standards and patient medical record information, as well as some issues of data quality.
And then the second one is being scheduled for May 17th and 18th, and we will concentrate more on content and medical terminology issues.
MR. BLAIR: The other thing is, we will by popular demand here be laying out the hearings for the rest of this year, as well, and be reviewing that with the group, probably by conference call and e-mail, because we probably need to be doing that before next time we get together. So, that also will be taken care of.
DR. DETMER: Questions or comments?
DR. LUMPKIN: And then finally, we had some discussion on the mechanism for collection of race and ethnicity and other demographic information.
We had quite an extensive discussion about how to get this out and came to the conclusion that this is probably a shared project between us and the Committee on Populations, and then looking at the Committee membership, we realized that just about everybody on the full Committee is on either one of those two committees, so we are going to try to address that issue with a little bit more in-depth discussion at the next meeting in June.
DR. DETMER: Okay. Questions or comments on that?
DR. COHN: Actually, I was just going to add something about the workers comp. Actually, we were going to be looking at the first report --
DR. LUMPKIN: First, yes. First report of injury. We expect that the industry is about ready to come out with a standard, so we can begin to look at that to begin to begin the process of adopting a consensus for electronic data interchange for first report of injury.
The Privacy and Confidential Committee is also going to be looking at that issue, so we expect that they will have some special interest as we conduct ultimately the hearing on that new standard.
Any questions or comments on that?
DR. DETMER: I don't know when it would weigh in, the decision on the NADO letter that I got that will go I guess your committees for discussion
DR. STARFIELD: Yes, we did discuss that.
DR. LUMPKIN: Right. And I think basically, we'll probably end up sending a response saying that we are going to be discussing that in an upcoming meeting, and this is how we are going to address the issue.
DR. DETMER: Yes. What I was planning to do, Jerry, was writing a letter, just saying that I pass this on. That the Committee is bringing to a hearing, and when they plan to have some hearings.
DR. LUMPKIN: All right.
DR. DETMER: What else do you have?
DR. LUMPKIN: That's it.
DR. DETMER: Okay. Lisa.
DR. IEZZONI: For the Subcommittee on Populations, I would like to report on three things. The first is the report on the insular areas. We reviewed a nice draft of it a couple of days ago. That draft will be revised. Once it is revised, it will be sent out to all the meeting participants, which means it will be sent very far away, to the Pacific Islands and to Puerto Rico and the Virgin Islands.
After we hear back from them, we will be finalizing our recommendations, hopefully, at a May meeting, date yet to be determined. And so we can report that report and the recommendations at June as an action item for forwarding recommendations to the Secretary.
The second thing is just to remind you, once again, that we gave you some homework on the Medicaid Managed Care, and in fact, Jackie, I think we have a copy of the principles that --
MS. GREENBERG: We have that.
DR. IEZZONI: Oh, you have a copy of the principles that Kathy Coltin presented to you all yesterday. There seemed to be a great outcry for a written summary of the principles that she mentioned in her verbal comments.
We will also be finalizing the report from that Medicaid Managed Care activity, hopefully in May. Again, we will produce recommendations that hopefully will be an action item at the June meeting.
And then the third thing that I would like to talk about is that on January 22nd, our Committee held an educational day here in Washington. January 22nd, our Committee held an educational day here in Washington.
It was actually great. I think we started out with Bob Kane, who's a pedagogue from 'way back from University of Minnesota, teaching us about how to look at care after people leave the hospital, was how he phrased it, and we spent the whole day hearing from people from a variety of different venues in what can be called postacute care.
Actually, I think that day, Barbara, it was focused on postacute care. And I think it got the Committee kind of educated, and all of us up on track in terms of understanding what the issues are, and they are big.
You know, long-term care is probably the fastest-growing single sector of the health care economy right now. And so, we will be in the future, probably starting after our June meeting, where we hope to finish the insular areas in the Medicaid Managed Care Report.
We will start another initiative that will focus on what we are calling the Continuum of Care. That will be split between our full subcommittee and the Working Group on Quality, which John, we're not sure whether you are still going to be a member of, but you need to let us know.
And this is a segue for me also to talk about Kathy Coltin and the working group's activities. Kathy had to leave for a meeting at NSQA.
MS. GREENBERG: I have the charge.
DR. IEZZONI: So, we'd like you all, while I am talking, Marjorie will pass around the charge. We would like you to look at it and maybe have a motion to approve the charge of the working group that we can vote on in a few minutes.
Basically, what we are planning to do around this Continuum of Care issue is twofold. As a full subcommittee, we are going to focus very explicitly on measuring functional status.
People might remember that in the core data elements, there was always a placeholder for functional status, and in a number of even the electronic data sets, I've heard people talk about having placeholders for functional status. But how you measure that, how you capture it, how you code it, has always been an open area for dispute.
It raises a lot of issues. For example, measuring function has to be age-specific to recognize that kids, especially infants, maybe quite different from elderly folks.
Functioning is also a multidimensional concept, thinking not only about physical functioning, but also cognitive functioning and sensory functioning. We also have to think about functioning in mental health; behavioral health raises a whole number of different functioning dimensions that we might need to consider.
So, we are going to try to help you think about how to fill in that placeholder when people talk about functioning status. So that will be what the full subcommittee will be focusing their next initiative on will be reviewing ICIDH, etceteras, and looking at the various people who have proposals for how to measure functional status.
Now, the working group will also be initiating activities on quality, looking at the Continuum of Care but with a particular emphasis on areas which we haven't really talked about very much, outside of the acute care hospital and the standard kind of office-based visit.
We will look at all of those, but we will focus primarily on looking at these other venues for giving care at the home, longer term care institutions.
The work plan for doing that is as yet a little bit open. You see that mentioned as I think Item -- one of the items of the work plan, exactly, and so Kathy will be kind of working with the rest of the working group to come up with a very, a much more specific targeted work plan for that particular initiative, but hopefully, there will be a synergy between what the full subcommittee will be doing and what the work group will be doing.
Questions, yes?
MS. GREENBERG: So that the full subcommittee will be focusing quite a bit on the measurement of functional status?
DR. IEZZONI: Functional status, exactly. Filling in that placeholder. We may not be able to come up with a specific answer, but we are going to try. That's one area that has always been left hanging.
DR. DETMER: So you do you want to make a motion to the work plan charge?
DR. KOO: Yes. I make a motion that we approve the work plan?
DR. LUMPKIN: Second.
DR. DETMER: Any discussion?
(No response.)
Hearing none, all in favor say aye.
(A chorus of ayes.)
Opposed?
(No response.)
Abstentions?
(No response.)
(Whereupon, the motion was approved unanimously.)
Okay, Ruth?
MS. GREENBERG: I have an information piece, too, from Kathy that she just wanted me to pass out.
DR. IEZZONI: Yes, I think that is off the Web site of the Joint Commission, I think, that says the Joint Commission and NCQA, I think, are going to be partnering in -- and is the AMA initiative also included in that, Marjorie?
MS. GREENBERG: AMAP, JCHO, NCQA --
DR. IEZZONI: And NCQA are going to be --
MS. GREENBERG: -- announced plans to merge performance measures development.
DR. IEZZONI: Right. This is actually a fairly cataclysmic event, for those of you who follow the politics of this. But it's a welcome event.
DR. DETMER: All right. Let see. Who's next? Are you ready?
MS. FRAWLEY: Yes. Okay, the Subcommittee on Privacy and Confidentiality met on Tuesday in all day meeting. And I am happy to report that we were able to accomplish quite a bit at that meeting.
We did have a meeting with six representatives from the pharmaceutical industry, and what our goal was at that particular meeting was to understand, from the time the consumer walks into a pharmacy, until the time the consumer walks out, exactly what the flow of information is, and we talked a lot about some of the EDI transactions, some of the consumer rights issues, privacy and confidentiality concerns.
And I think that most of the subcommittee members would join me in saying that was probably a very, very beneficial meeting, and they enjoyed it so much that they agreed to two more days, so we are going to try and find two days in April, that is our goal as a subcommittee. And just for people that are here and also listening on the Internet, we are issuing the invitation -- this will be a second invitation -- the pharmacy benefits management companies, to come and appear before the subcommittee to explain to us their role in this whole process, and really, talk a little bit about the issues of confidentiality, privacy, employer use of health information, patient rights. So, we expect to spend a half a day on that issue.
The second issue that we will spend a half a day on is this whole issue of workers compensation and occupational health, because we are trying to figure out how to put our arms around this particular issue.
The third issue that we will be addressing is employers' use of health information, and what we are talking about is people who are self-insured and administering benefit plans, and exactly what information do they have on their employees, and what privacy protections are in place? So, that will be for a day and a half in April and we will leave a half a day for the subcommittee to kind of synthesize that.
At our June meeting, we are hoping to be able to spend a session with the Medical Information Bureau, to make sure we understand some of their processes, and are not having a knee-jerk reaction to media reports.
Our plan, I am happy to report, Don, that our entire work plan will be finished in June.
DR. DETMER: Fantastic.
MS. FRAWLEY: And we revalidated that, and so that will free up a subcommittee then to provide the technical assistance to the Department, in the event the Secretary has to start going through the rule-making process, in the event that Congress fails to pass legislation by August.
So, I think the last year has been very, very rewarding, and I have to thank all of the subcommittee members and the staff for that.
The other thing that I just want to bring to your attention is that we did give you, yesterday, the letter to the NAIC. It's back at your place. I'm giving everybody a minute to find the correct piece of paper. So, if you could all get that out, so you can understand what we did, we can vote on this now, and let Don sign it.
If you remember, yesterday afternoon's presentation, we discussed the fact that we wanted to send a letter to the National Association of Insurance Commissioners, acknowledging the briefing that we received in November, and secondarily, our specific concerns regarding their Health Information Privacy Model Act.
And if you remember yesterday, there were three points raised. One was a point by Clem McDonald to take a sentence out, and the sentence is gone. And then responding to Vince Moor and Barbara Starfield specifically concerned because they thought we were being too nice in the letter, what we did --
And I will tell you where the two changes are, so that you don't have to have a heart attack, is that the first place is that we left the sentence the way it was in the second paragraph, "The NCDHS would like to call to your attention, some of our major concerns regarding the Model Act." We left it that way and decided not to add a sentence there.
What we decided to do was, in the last paragraph on the second page, it says, "The NCDHS has several additional concerns that we would like to discuss with you, and we would be happy to answer any questions that you may have."
And I will initiate the dialogue with them. So we are not waiting for them to call me, I will be calling them. And the other piece that Barbara and Vince raised was about the MIB, which is the Medical Information Bureau and not John Lumkin's "Men in Black," which is further up the page on page two. We are at the third paragraph on page two.
And it says that, "The NCDHS notes that this provision would allow that all health claims that are filed, subject to this Act, be disclosed without authorization from the Medical Information Bureau, for the purpose of detecting fraud.
"The NCVHS does not support this type of unauthorized disclosure, because this provision would vastly increase the ability of the MID to track information on people, who have never applied for underwritten life or health insurance."
So, those were three concerns that you brought over yesterday, that's the sixth, to this letter, and at this point, I would like to move consideration. I need a motion from someone. Gail, you have to come to the table please, and a mike.
MS. HORLICK: I wanted to point out that there was one other change in the second paragraph of the letter, where we had originally said that the Model Act was reviewed in light of the NCVHS recommendations to the Secretary and the Secretary's subsequent recommendations to Congress.
We also added before that, we inserted some background, and indicated "That over the past two years, that we had held a number of hearings" - and I didn't bring my good copy up here. And --
MS. FRAWLEY: Second paragraph.
MS. HORLICK: And "The Committee heard testimony from a broad spectrum of experts in the field, including representatives from industry, research, and public health, and the privacy community." And then it continues, that "The Model Act was reviewed in light of this testimony." And the rest is the same.
DR. LUMPKIN: Second.
DR. DETMER: Okay, it has been moved and seconded and is there discussion?
(No response.)
Hearing none, all in favor, say aye.
MS. FRAWLEY: Wait a second. Wait a minute, please. It's Jeff.
DR. DETMER: Jeff? I'm sorry. I was -- we were prepared to vote.
DR. HARDING: Just a technical question on the third paragraph on the second page, last sentence. "The ability of MIB to track information on people who have never applied or underwritten life or health insurance." I believe the MIB is only for life and disability at this time. And so, would there be a correction to say, "life or disability," or are you talking about in the future, if they get into health?
DR. DETMER: Kathleen?
MS. FRAWLEY: If you remember on the January conference call, John Fanning presented that information to us. So, that is why that word, health, is in there. I agree. They traditionally have done life and disability. John has now indicated that that is their new market area, and that is why we made the decision on our January conference call to invite them to come to talk to us, so that we don't believe everything we read in the newspaper, because we are getting confusing information. So, that's where we put that.
DR. HARDING: Thank you.
DR. DETMER: Okay?
DR. HARDING: Yes.
MS. FRAWLEY: Okay.
DR. DETMER: Are there other questions, comments?
(No response.)
Okay, hearing none, all in favor say, aye.
(A chorus of ayes was heard.)
Opposed?
(No response.)
Abstentions?
(No response.)
Okay, so ruled.
(Whereupon the motion was carried by a voice vote.)
MS. FRAWLEY: Thank you.
DR. DETMER: Other items, or is that it?
MS. FRAWLEY: That's it. Thank you.
DR. DETMER: The Work Group on the National Health Information Infrastructure had a, I think, a good meeting yesterday afternoon. We essentially reviewed where we had been in the past, and one of the things that I would like to see us act on is the charge to, and what the work group will be doing on that.
I would say, the summary -- we also heard a presentation on a new effort, using sort of the VISA metaphor of perhaps a way that one might be able to move health records so that consumers could have a secure, sort of lockbox on the Web, for being able to keep and maintain their own records, and that was one presentation.
Then we moved in. The staff had been charged at the last meeting to do a review of the Australian, Canadian and U.K. models and approaches to their national health information infrastructures.
Those were nicely presented, and then we had the good fortune, of course, of having people from those nations there to sort of see, had we really cut, if you will, in our view at it, cut the thing the way they in fact saw it playing out from where they were coming from?
And I think the sentiment, by the end of the discussion, was that the staff would take the deliberations of yesterday, and put them together, come back to the work group, trying to assimilate what we heard as well from our colleagues abroad.
I think there was also clear sentiment that we really would like to work hard to find a way to stay in touch with these people, as well as others. That in fact, as we try to pull together a matrix of how the U.S. ought to go at this, obviously, this is not a one-time fixed target, and being able to work iteratively with them on some common issues, struck all of us as being really a positive and important thing.
I think that is the bulk of what we conducted. We weren't able to get too much further. We did talk a little bit about the glossary. The idea is to try to pull together a set of terms, just like the presentation yesterday morning, our Canadian colleague mentioned infostructure as sort of a new kind of word that's there.
The feeling was, just having a set of terms, as these things roll forward, would be useful. To the extent that we can, we don't plan to write our own dictionary, so much as try to shift and sort and glean what other people are doing, and pull that together into a site on the Web.
There will be this desire to move forward on this matrix initiative and effort, trying to look at all the various components that were elucidated in the Concept Paper, and then try to put out a work plan that really more specifically relates to that. But I think that piece will have to wait for the next meeting.
We really, I think mostly, when it was all said and done, continued to hear what we educated ourselves on, what was happening in these other countries, as well as how the staff viewed that.
So, I would say that's a summary of our deliberations. If there are questions, I'll take those, and if not, I would like to move forward to the charge and the work plan. And I would be -- okay, hearing none, would someone like to make a motion to accept this, it's been batted about. Lisa?
DR. IEZZONI: So, move.
DR. DETMER: First and --
DR. LUMPKIN: Second.
DR. DETMER: -- and second. Okay. Is there any discussion?
(No response.)
All in favor, say, aye.
(Chorus of ayes.)
Opposed?
(No response.)
Abstentions?
(No response.)
Okay. It so has passed.
(Whereupon, the motion was unanimously approved.)
Okay, I think that's the report --
MS. FRAWLEY: Do you want to hear from Dan?
DR. DETMER: Oh, yes, sorry. Dan. Thank you very much.
DR. FRIEDMAN: This is the Work Group on Health Statistics for the 21st Century, or some such. I get the words confused, but in any case, there are four sort of subactivities to it, one of which is a series of five commission papers, all of which are under way, and the drafts are due in I think late March, early April. There is one on -- well, I won't go into the details unless people are interested.
The second activity is a couple of discussion groups dealing with what are going to be major issues in health and healthcare, 10, 15 years out, and what are their implications for health statistics? And invitations to those will be going out this week or next week, and one has been scheduled for March, and I think one for May. And then there is going to be a separate two-day session of roughly 10 to 15 people looking at a series of definitional and boundary issues around health statistics.
What is it? How is it different, is it different from surveillance? What are the criteria we can use to evaluate health statistics systems and so forth? And that's been scheduled for late March.
The third activity is the Committee on National Statistics Work Shop, and Marjorie and I and Ed Hunter and Jerry Henderschott had a conference call this week, and what we need to get back to them, and either have a meeting or a conference call to basically find out where they are, because we're not really sure at this point.
And then the fourth activity, which we have just recently added on, but I think all of us think is essential, is trying to add on a public part to this process, and the way we are envisioning it, the public part would include both some open forums, professional meetings, perhaps some feedback and individual state visits that Ed Sondik is envisioning. It's sort of the Ed Sondik Millennial Tour.
And perhaps a publics, or a public hearing sometime in the fall. Part of this activity has been -- was generated by our being very impressed with both the Canadian process, with their provincial hearings on health information needs, that Jennifer discussed yesterday, that's been very thoughtful and very well done; has resulted in province-specific reports, but also I think shows promise of generating a lot of public support and understanding for what they are trying to do.
And part of our desire to add on a public process was also generated by the NIOSH, the National Institute of Occupational Safety and Health, National Occupation Research Agenda Process, which has also been very much a public process, which they have used to develop a research agenda and really drive their allocation of research funds. So, that part, the public part, is still unfolding.
DR. DETMER: Questions or comments?
MR. SCANLON: Just generally, in terms of the pinning down the National Academy of Science's work shop would be, are we thinking more September, September or so, or --
DR. FRIEDMAN: That would be my guess, and when we had talked to them at the November meeting, we had talked about probably July, or maybe September. And we are not sure at this point.
And as I said, we really need to pin that down. And NCA -- it's -- that communication is really NCHS' to cinch that. I don't understand all of the complications --
DR. DETMER: You will.
DR. FRIEDMAN: Well, maybe I do and I don't want to. It's NCHS to cinch that, and not NCVHS, Jerry.
DR. DETMER: No, I appreciate all the work that is going in, and it looks like, obviously, it is going to be a very busy year. That looks good. You had an item, right?
MS. GREENBERG: Yes. I just wanted to update you, since I have been reporting on the Work Shop that we had on Implications of HPPA for Public Health and Health Services Research.
I think we had a wonderful demonstration this morning of the synergy here. It's really a nice case study, what Susan and the others presented, as for what I think the public health community would like to do, and support.
So, I just wanted to give you an update on what has happened since the work shop, which I reported on in November. We had just had it, and as I reported then, the work shop participates recommended the development of a consortium to organize the public health and health services research communities on data standard issues, and identified some goals, etceteras.
In December, we held two conference calls with a number of the opinion leaders and organizations that had participated in the meeting, and particularly strongly advocated for development of this consortium, and agreed to the goals which are on this Fact Sheet that I passed out, and on the back, with some of the organizations that have been working with us, or we have been working with them, really, because the really enthusiasm for this is coming in great part from the states, who see this not only as an opportunity to get some of their state health departments, state health data organizations, to get some of their data needs met, through this current standards process to standardize their data.
But even to work with other agencies at the state level, to try to come to come to some further understanding on what data needs they have, and how they can try to standardize the data that needs to be collected.
And we developed a number of priority or action items, which are on the back page of this Fact Sheet. We did have a meeting, we had actually two meetings last week, in conjunction with the Annual Meeting of the National Association of Health Data Organizations.
The day before, on the 24th, it was an open invitation for all the people who had come to the work shop, as well as everyone who was coming to the NADO meeting on the 25th and 26th.
We had about 50 people. We had an excellent discussion. Again, a real desire to see this carried forward, and in fact, decided to meet that Monday evening as well, and we set up about eight little task groups of people who were working on operationalizing different parts of this.
So, it really seems to have caught fire. Now, there is a great deal of real interest in seeing if we can make this work, and being, you know, as inclusive as possible. We have a list serve, which I think it just mentions it on here. I don't think it tells you how to sign up for it, but if you are interested, just send me an e-mail and I will --
Or, I'll just send something to the whole national committee, because I forgot to bring that piece of paper, but we have just established a list serve. We still are using the Lewin Web site, and we see most of the communication, you know, being through list serves, conference calls, possibly meeting in conjunction with other meetings, and one of the things we -- and then possibly a separate meeting if, you know, funds are available and seems worthwhile.
One of the low task groups is, the education one, to try to really inform and work with all the different organizations who have come to the table, and some who haven't, to be on their annual meeting agendas, as we were at the NADO meeting. They really gave us a great opportunity, not only in these two separate meetings, but in the actual agenda, to present this, and to get some feedback on it.
If you know of organizations that would like to learn more about HIPAA and what its relationship is to standards in public health, and for research, you know, let us know, and we will see if we can get someone out to their meeting, but there is a great deal of interest in that, and as well, in participating in the standards process, getting more public health people attending the meetings, some additional representation on the National Uniform Billing Committee and the National Uniform Claim Committee, but also, attending HL7, X12, etceteras.
Of course, all these things require resources, so this is at this point, you know, that hasn't all been settled, but I think there is enough interest that I feel very good about the fact that we will try to carry this forward.
DR. DETMER: Any questions, or? Yes.
DR. COHN: Actually, I just had a -- as I was looking at your priorities and action items, one of the things you mentioned is a user-friendly data dictionary, which I think would be a terrific addition.
My question was, is this a wish list, or is this a set of something that we can expect --
MS. GREENBERG: Okay, the --
DR. COHN: -- to be developed by this consortium in the near future, or?
MS. GREENBERG: Yes. The inventory, which is the first one, is actually something that AHCPR is funding, an update of something NAHDO did a number of years ago on, actually looking at, what are the current requirements and future requirements, actually, as well; so, looking at what states are now collecting.
The user-friendly data dictionary --
DR. COHN: Yes, my question --
MS. GREENBERG: Yes?
DR. COHN: When will that be?
MS. GREENBERG: I believe that it is supposed to be ready by June.
DR. COHN: Great,
MS. GREENBERG: And NAHDO is doing it under contract with AHCPR, and so, that's great, because we haven't had anything like this for quite awhile. And so much has happened since the last one, in the early nineties.
The user-friendly data dictionary which will try to, as it says, communicate the content of standard claims transactions to a variety of nontechnical audiences, and particularly, the elements that are relevant for public health, or that people feel are relevant.
We actually have a small, a group working on that. They are doing, you know, they don't want to reinvent the wheel, so they are looking at what's out there, and what other people have done, and they are going to sort of develop a test, or a pilot sample of, well, does this -- you know, one or two elements.
Then that will get broader review, and it will probably be reviewed on the list serve. And then we are hoping to get some funding for this. We don't have it yet, but we have a plan. And as I said, we have said up the list serve. We have a Web site.
I am going to be presenting at the NUBC-NUCC Joint Meeting on February 17th, an executive meeting that the two groups are having together, and I am a member now of the NUBC, but I will be reporting to them on what the Consortium's recommendations are for representation on those two groups. So, everything on here is pretty much, we have a plan for it.
DR. DETMER: Other questions or comments? Okay, let's move on -- yes?
DR. FRIEDMAN: I just wanted to mention that, as far as I know, this is really the first time that, you know, even within public health, these different public health data-related organizations are sitting down together to discuss these issues, and it's really -- it's a terrific endeavor, and it is really quite remarkable that it -- it really is, it's just -- it's quite remarkable that it's happening and also happening on a voluntary basis, rather than being pushed from the top down.
MS. GREENBERG: It seems to kind of be an idea whose time has come --
DR. FRIEDMAN: Absolutely.
MS. GREENBERG: Because there was a lot of that kind of sentiment coming out, and I think Bill was -- he was at the NADO meeting, and you heard that, too, didn't you, Bill? Yes, so I was really pleased.
DR. DETMER: A new era of trust and collaboration breaking out before us?
MS. GREENBERG: Why not?
DR. DETMER: No, that's great. That's great.
MS. GREENBERG: You know, it won't be simple, but just the interest and the willingness and the kind of not overly turf-oriented discussions at all, at this point. It was very encouraging.
DR. DETMER: Well, I hope it continues to make progress. That's great. Okay, I'm going to -- why don't we move on, then, to our two reports, the Committee Reports, pick those back up.
Just to lead into that, the first one that we would be talking about is the 1996-1998 Report. I think what we heard yesterday in the discussion was that it was both incomplete and deficient; that among the things that will be recommended is adding a list of all of the staff members that contributed to our work, which I think definitely needs to be there, as well as some changes in language, relative to the unique identifier, as well as security, and there may be some other items, but I think those are the two things that I particularly heard.
John and Kathleen were going to talk with Bob, really, to the unique identifier piece, and I don't know, Kathleen, if you have some thoughts on the security piece as well, but why don't we go ahead and let me just turn it over to you at this point, John and Kathleen.
DR. LUMPKIN: Well, we kind of did this independently, and basically what we suggested is that on page 13, where it talks about the unique health identifier for individuals, we would add language to the effect "That NCVHS has long identified the need for a unique health identifier."
Then it would lead into that other sentence, but because of its complexity, so there would just be an opening clause to identify that our position in fact predated HIPAA on the unique identifier, and that would be the recommendation.
DR. DETMER: Okay. Do you have more to add to that, Kathleen?
MS. FRAWLEY: No, I think that what John is pointing out is our discussion this morning at our subcommittee meeting, is that what we are focusing on? The three points, or -- I'm just confused --
DR. DETMER: No, no, we're only --
MS. FRAWLEY: What report are we looking at ?
DR. DETMER: We're looking at the 1996 --
MS. FRAWLEY: Okay? I'm sorry, I -- which report are we looking at? I'm totally confused here.
MS. GREENBERG: The 1996-1998 Report. You know, they both get called Annual Reports.
MS. FRAWLEY: I know --
DR. DETMER: Yes, it is confusing. I apologize. So, 1996-1998, right now.
MS. FRAWLEY: Right. so, the 1996-1998 Report, John, this morning, we agreed that the discussion would be, here is what we did prior to 1996. Here's what HIPAA required us to do, and here is what the Committee did from 1996 to 1998.
DR. LUMPKIN: That's a much better statement of --
MS. FRAWLEY: That's what we agreed to this morning and voted on.
MS. GREENBERG: That was for the 1996-1998.
DR. LUMPKIN: Well, no, actually, that was for the -- was that for the 1996? I thought that was for the HIPAA report.
MS. GREENBERG: Actually, the discussion that there was --
DR. LUMPKIN: Because we did not discuss --
MS. GREENBERG: Because the discussion was that, in fact, the HIPAA report was on the Agenda of the Subcommittee, but the 1996-1998 Report was not.
DR. LUMPKIN: Right. But that particular statement that particular statement that you made is a much clearer statement of how we should we change this section. Because the scope of this section is much broader than the scope of the HIPAA report.
DR. DETMER: I think actually, I mean, and Bob referenced this yesterday. Both reports need that kick, so I don't think it's -- you know, I think that's fair enough that way.
DR. LUMPKIN: Okay.
MS. FRAWLEY: Do you want to hear on the six-month security?
DR. DETMER: Well, first of all, let's go ahead and see, the other part that I think you talked about was that this would be drafted and sent to the Committee, for feedback, and then if any concerns, if there were concerns, would be adjudicated by the Executive Committee. Is that correct?
Okay, so essentially, I don't know if --
DR. LUMPKIN: For transmittal.
DR. DETMER: Pardon?
DR. LUMPKIN: I'll move that.
DR. DETMER: Okay. Is there a second
DR. STARFIELD: Second.
DR. DETMER: Okay. Is there discussion?
DR. LUMPKIN: I think we need to perhaps include the security and entertain a motion, or --
DR. DETMER: Well, we can. it will be the same process for it, too, so I don't care. Do you want to have that --
MS. FRAWLEY: Well, we didn't tell you what we were going to do on the security piece, that's what I'm trying to get to --
DR. DETMER: All right. Go ahead.
MS. FRAWLEY: -- before we started voting on something. Okay. On security, the problem that we had is the paragraph or the document, and I'm totally confused which report that is in at this point, so if I'm out of order --
MS. GREENBERG: It's in the HIPAA report.
MS. FRAWLEY: -- I give up.
MS. GREENBERG: No, it's in the HIPAA report.
MS. FRAWLEY: But to respond to the concerns that were articulated yesterday, we are going to take both reports, because it's so confusing at this point.
DR. DETMER: Well, fine. I initially aid the --
MS. FRAWLEY: Right.
DR. DETMER: Okay? Yes.
MS. FRAWLEY: So, we are going to take both reports, and do it line-by-line, and make sure we have are totally consistent all the way through.
On the security fix, and I don't know where it is, in which report, what we are doing is taking the letter that we sent to the Secretary, and that is what is going into the report, whatever page it's on, in whatever report, we were talking about.
MS. GREENBERG: Not the whole letter.
MS. FRAWLEY: Whatever report that is.
DR. LUMPKIN: It's more or a less a paraphrase of the letter.
MS. FRAWLEY: Right. And I assumed responsibility, Jim was very gracious to get the staff to get me a copy of the letter, midmorning, and I am going to rewrite that piece.
It will go back through the review process, with the motion just articulated, which is, it will come back out. The Executive Committee will adjudicate any concerns, and that's how we'll fix the security piece.
DR. LUMPKIN: Now, that is on page 27 of the Annual HIPAA Report.
MS. GREENBERG: The Annual HIPAA Report.
DR. LUMPKIN: And actually, this process is -- it's my fault, I thought there was a security piece, also, in the 1996-1998, but --
DR. DETMER: And I don't think there is --
DR. LUMPKIN: And there isn't.
MS. GREENBERG: That same language, it just references that the Committee made recommendations, but the language that was of concern in the HIPAA Report, that was not carried over, as far as I can tell, in the other --
DR. LUMPKIN: So, we probably should do it as a separate motion, as the Chair wisely suggested.
DR. DETMER: Well, that's what I would suggest. So --
DR. LUMPKIN: I apologize.
DR. DETMER: -- what is on the floor? That you --
DR. LUMPKIN: It's just the change in this on the unique health identifier on page 13.
DR. DETMER: All right. That sounds fine. And that's been seconded. Is there discussion? Yes?>
DR. GELLMAN: I mean, I don't have any problem with the motion, but when the language is revised for both reports, I reserve the right to amend the views that I read the other day, and resubmit them.
DR. DETMER: Any other comments?
(No response.)
Hearing none, all in favor say aye?
Opposed?
(No response.)
Abstentions?
(No response.)
Okay.
(Having been moved and seconded, the motion was approved by a unanimous voice vote.)
Now let's go to --
DR. LUMPKIN: Okay. The HIPAA Report, there are four changes.
DR. DETMER: The Second Annual Report to Congress on the Implementation of the Administrative Simplification Provisions of the Health Insurance Portability and Accountability Act, February 2. He'll walk us through the core places --
DR. LUMPKIN: The first change is on page three, numbering on the bottom. I'm sorry, page on page two --
MS. GREENBERG: It's not of the roster.
DR. LUMPKIN: Yes, it's not of the roster. Where it talks about progress in HIPAA data standards, we would include a short description of the Y2K issue and the overwhelming response we got on the need to not have them both happen at the same time.
The second --
DR. DETMER: They won't.
DR. LUMPKIN: What?
DR. DETMER: And that now they won't.
DR. LUMPKIN: And now, they won't. The second was on page 18 through 20, would fit somewhere in there, would be a paragraph on the activity of HIPAA and public health and health services research, that we just had a report on. There would be some paragraph summarizing that we were cosponsor, and what that conference was.
And then on page 26, we would make a similar change to that language on linkage of the individual identifier to privacy protection, as we discussed on the previous report.
And then on page 27, everything that Kathleen said, she really meant on page 27, because I didn't understand it, but it's --
MS. GREENBERG: She did.
DR. LUMPKIN: It's going to be read, like our letter, and not like it says here.
DR. DETMER: Okay.
DR. LUMPKIN: Those are the four changes.
DR. DETMER: Do you have anything to add to that?
MS. FRAWLEY: No, we finally are all in synch.
DR. DETMER: Okay -- all right?
DR. LUMPKIN: So it would be redone, right, circulated and adjudicated by the Executive Committee, then -- okay.
DR. DETMER: Is there a motion to that extent?
DR. COHN: I move.
DR. DETMER: Simon moved it. Is there a second?
DR. IEZZONI: I second it.
DR. DETMER: Okay. Any discussion?
(No response.)
If not, all in favor say, aye.
(Chorus of ayes.)
Opposed?
(No response.)
Abstentions.
(No response.)
Thank you.
(Whereupon, the motion was approved unanimously.)
DR. DETMER: Okay. Thank you very much. I think it's always tough to write a report to the approval of a committee before it goes to the committee, and I think that this has been a good process, and so from my perspective, I'm quite happy, and I appreciate it.
We have a few other items. One that's a fun one relating to the 50th anniversary event, and that's a Tab somewhere -- no, I know where it is. It's not a Tab, it's in the Executive Subcommittee Minutes.
MS. GREENBERG: Right. Which is a Tab.
DR. DETMER: Which is the M Tab. And at the back of that, at the end of that report, the Executive Committee talked about having perhaps a half-day event, and have it probably at the time of, you know, our June 2000 Committee Meeting, and essentially, try to see if maybe we -- if we could get the Great Hall of the National Academy for a reception for, it. Find somebody that might be willing, at one of the foundations or something to help support that.
But, actually, do take a little time to both look back, but mostly look forward in the sense of where we are going this, and I had said that I would be happy to help work with the Committee in that regard, and obviously, would still be willing to do that off the Committee, as the Committee so desired.
But at any rate, those are my brief recollections, but the Executive Committee, many of you are here, so why don't you weigh in and give some more thought and discussion to that, because we didn't want to get the approval of the Executive Committee, or of the full committee, so the exec may continue on this planning.
So, what others? Want to add more to that?
DR. IEZZONI: I think that's a reasonable summary of it.
MS. GREENBERG: I think we probably, staff would benefit from a few members in addition to our British office, or maybe it would just be the Executive Subcommittee, either way, helping to think through the --
DR. DETMER: Yes, I think the -- because there will be a program on that.
MS. GREENBERG: Yes, there would be a program, but I think we're talking about a half-day program, so it -- and --
DR. DETMER: Well, I don't know. John, do you want the Executive Committee to --
DR. LUMPKIN: I think so.
DR. DETMER: Yes. Okay.
MS. GREENBERG: But of course, we would welcome suggestions from anybody as to what might be involved in it. Well, Don, yes, we have to make sure that we have that June meeting --
DR. DETMER: Well, that would be helpful.
MS. GREENBERG: After your semester is over, or whatever.
DR. DETMER: Yes, that would be -- and if not, then that's where it is, but I mean --
MS. GREENBERG: Well, we haven't scheduled it yet, though actually, we do need -- these are things that Lynette always took care of, so I have to be thinking here. We do need to --
DR. DETMER: Lynette has a smile on her face.
MS. GREENBERG: Right. And I've got tears in my eyes, but we need to set up actual dates. We have dates for the rest of this year, that I hope are on people's calendars. Maybe that's the next Agenda item. But also, we should really, because June 2000 will be here sooner than we think, so we need --
DR. DETMER: Yes, it will.
MS. GREENBERG: We are going to need to set up dates for 2000, too, and I guess we'll do that through polling people.
MR. STREIMER: Marjorie, you might want to check on the availability of the Great Hall.
DR. DETMER: Yes. Well, I need to know both. I mean, we need some dates, options, and then I can check with where they are on that.
MS. GREENBERG: Right. We can --
DR. DETMER: Sure. I agree.
MS. GREENBERG: Also, we are planning to commission from Susan Kannanar, our historian, who did the 45 year anniversary, so really all she has to do -- and now she has done this three-year plan, I mean, three-year report, so it really shouldn't be a major deal to -- oh, Susan, there you are -- I thought you did a great job, actually, on both -- to update, so to just kind of update the 45th, so that we have the equivalent of a 50th.
DR. DETMER: Sure. Before I turn the gavel over to my colleague, the two chairs that are left, this actually by the terms, are the last meetings for Hortensia Amado Simon Cohn, and Kathleen Fyffe, and Vince Mor.
I certainly, personally, have really enjoyed working and benefiting from your friendship and counsel, and I am quite confident that, you know, John will want to talk with you about, you know, another tenure. As that plays out, I certainly personally would hope that you would be interested.
Obviously, I won't be party to that process. There will be some new members coming on, obviously, replacing my spot, and the Executive Committee gave a fair amount of time and attention to the categories of sort of skill sets that we felt we needed, so I do -- you know, this is also going out on the Web.
I think that there will be an announcement coming out, if it hasn't come out already. You know, when do you expect that, Jim?
MR. SCANLON: We'll be issuing, the Department will be issuing, a Federal Register notice, soliciting nominations, in probably three weeks.
DR. DETMER: So, in any event, that's I think in a way, the last piece of business, except to thank you all once again for, really, a terrific personal experience, and a wonderful professional experience, and with that, I will pass actually this gavel to John to take over what's left of the meeting that he chooses to have.
DR. LUMPKIN: Well, thank you, Don. It's truly been an honor to have had the opportunity to really get to know you well, and to work on the Committee, and to really move in the direction that this Committee has moved in.
I was a new person to the Committee, as you were when we started, and I think that, judging from what we have accomplished, and somehow I'm not sure how we've done all that, it certainly is a statement and a credit to you.
I did have a couple of closing remarks I wanted to make, and then we'll have to let everyone go. And that really is to identify the importance of the Committee and the direction that we have been on, and more importantly, the direction that we need to go in.
Health and healthcare and this system that we've, if one wants to call it that, that we're living under is, I believe, at a nexus point, or very close to one.
Managed care may be what will be the surviving structure of healthcare delivery in the future, but certainly, there is tremendous concern about what managed care is doing, and how it's performing, and the transformations that will occur in the healthcare system, we are in the midst of, and one always hates to be in a situation of the ancient Chinese curse, may you live in interesting times, and in health and healthcare, this is certainly interesting times.
I believe that if we are going to have a true transformation of this healthcare system and the health delivery system, one has to recognize that managed care has about yielded all the cost-savings that they can, with the current arrangements, and that real transformations are going to occur about actually delivering the service, which is, a clinician and a person treating another person, providing advice, providing information to keep them healthy, to allow them to get well once they become sick, or to help them live with their illness.
We have to be able to do that differently, and one has to look at the fact that the healthcare industry is somewhere under 2 percent of investment of gross revenues and information systems, and the average for most other businesses is 8 percent.
Basically, one of the major obstacles is the fact that we don't have a rational approach to health information systems, and if we are going to make a transformation, we need to rationalize that. And that's really the responsibility of this Committee.
Working with the Department, working with all the various groups and Don raised an important question yesterday. He said, if we don't do it, who else is going to do it?
And I think that's really our charge, to step into that void, to try to precipitate action. We don't have to do everything, but I think in many ways, we can precipitate action, and we have seen that with the initial stages of the HIPAA standards, where we didn't do the standards, but we recognized that, and passage of the law.
Many of the standards groups got together and did stuff that they wouldn't have done, necessarily, without that sort of impetus. And so I think we have a really, a tremendously great charge to move forth a vision, and the work of the Health Information Infrastructure Committee, I think is going to be central, because it provides the overall structure, and this may be one of the things we will want to look at for the celebration of the 50th anniversary, the new millennium, is to perhaps use that as an event to unveil this infrastructure, this vision, this model, and to begin to coalesce and bring people around it.
And I guess my last point is that, if we are going to be successful, we need to expand the ranks of the converted. We need to be evangelistic. It's too easy to fall into our lexicon of initials and ANSIs and X12s and all that stuff, and yet we need to bring it down to real terms.
A mile and a half away, just southwest of the Capitol, is a housing development, and there are a number of people who live there, and their health has been unchanged by what we have done.
There is a young mother there who has a child, who needs health services. And if we can make the transition, if we can communicate our message to people who are concerned about the health of the people of this Country, the transformation in health systems will be possible.
Our biggest challenge is not to keep this important stuff that we are doing to ourselves, but to translate it in meaningful terms so that providers will adopt it, and people who are very concerned about their health and healthcare will adopt it, and they can be the ones who will be the purchases to make the change.
And so I think we certainly have a challenge before us, but what I certainly have learned by working with this Committee is that, certainly, if not us, who is going to do it but the us that's around the table, who are certainly up to the challenge. And so it's an honor to be asked to be the Acting Chair,
Thank you. And with that, unless there is other stuff, we are adjourned.
(Whereupon, at 2:40 p.m., the meeting was concluded.)