TB Notes Newsletter

No. 4, 2008

Highlights From State and Local Programs

New England TB Heroes Awards Go to Two Outstanding Nurses

The New England TB Consortium has launched a new award that recognizes persons who have made an exceptional contribution to TB control and the well-being of patients and the community. By creating the TB Heroes Award, the New England TB programs pay homage to those individuals who by some act or consistent performance embody the highest qualities of public health caring and service. The Award also demonstrates our appreciation for the recipients’ dedication to patients and the example they set for all of us.

The 2008 recipients for the New England TB Heroes Award are Susan Nutini, RN, clinical nurse at the RISE TB Clinic at Miriam Hospital in Providence, Rhode Island, and Suzanne Gunston, RN, MS, former coordinator of the Maine TB control program.

On September 19, 2008, before a full auditorium at the Northeast TB Controllers meeting in New York City, Kathleen Gensheimer, MD, Maine State Epidemiologist, and Jill Fournier, RN, New Hampshire TB control program coordinator, introduced the award recipients.

The audience listened in rapt attention as the recipients and award presenters stood before the group. Susan Nutini was nominated for the award for her 23 years of service at the RISE TB Clinic where her tenacity, compassion, sense of humor, and dedication to patient care is legendary. In the years prior to the implementation of directly observed therapy in Rhode Island, Susan was known to drop off medications to patients at their homes after clinic hours to ensure that they stayed with the program. Over the years, Susan has trained countless physicians and nurses who have rotated through the clinic. Physicians around the state call the clinic to discuss cases with her and to gain her advice on care, a true measure of the medical community’s respect for Susan.

Photo caption: New England TB Heroes Suzanne Gunston and Susan Nutini, in center, holding awards, are congratulated by Mark Lobato, DTBE regional consultant and Kathleen Gensheimer, Maine State Epidemiologist, on left, and Jane Carter, former medical director of the RISE TB Clinic, and Jill Fournier, New Hampshire TB control program coordinator, on right.

Through her dedication, she has saved lives by finding the exposed child who did not look quite right, by astutely interviewing patients to identify community exposures, by engaging difficult patients in care through a combination of tenacity and kindness, and by educating the community about TB. In the words of her nomination, “She can retire when TB is eliminated from Rhode Island or when cloning is possible — so we can be blessed with many Susan Nutinis!  Susan Nutini is a hero in TB!”

Suzanne Gunston, RN, MS, is no novice to public health. After many years as a public health nurse, she found a passion for TB as the Maine TB program coordinator. Her nomination was for exemplary leadership, initiative, creativity, and quality-of-care improvements. Cutting her teeth on a large TB outbreak among the homeless that started in 2002, Suzanne became a TB leader respected throughout the state. Over time, Suzanne developed mutual respect, collaborative relationships, and the ability to solve problems with provider partners throughout the state through clear communication and by working toward excellence in service delivery and design. She also collaborated with other public health nurses to enhance their capacity to work with high-risk groups, as well as with primary care medical providers on the implementation of the federal CDC guidelines on prevention of TB in health care facilities. She was able to understand the needs and concerns of medical providers, and to align these with the needs of patients. This meant not being dogmatic, but rather working with people to meet their needs and still accomplish good TB control. She also recognized the times when she could not compromise on regulatory issues because of the insidiousness of TB, and in these situations her patience and graceful persistence were recognized as a means to accomplishing the necessary goals.

During her tenure as TB program coordinator, Suzanne spearheaded the design of a calendar with animal figures to improve adherence in children being treated for latent TB infection, led the development of a toolkit for conducting screening for TB in homeless shelters, and demonstrated creativity and cultural competence in finding solutions for TB management among the Somali immigrant population. Retiring on May 1, 2008, “Suzanne has indeed left her mark on TB control in Maine,” the nomination noted, “and for that, we are all grateful to her. It would truly be fitting to award her with the 2008 New England TB Heroes award.”

—Submitted by the New England TB Consortium Nominating Committee:
Sue Etkind, RN, MS, Director,
Div of TB Prevention and Control,
Massachusetts Dept of Public Health;
Kathleen Gensheimer, MD
Maine State Epidemiologist;
Jill Fournier, RN, BSN, Coordinator, New Hampshire TB Program; and Mark Lobato, MD, New England TB Consultant, Div of TB Elimination

 

TB Death Assessment Tool in California: Development and Pilot Test

Background

In 2006, a total of 268 (9.7%) of California’s 2,778 tuberculosis (TB) patients died with TB. Of these, 23.0% were diagnosed after death, 4.4% were alive at TB diagnosis but died before starting TB therapy, and 72.7% died during TB therapy.  In 1995, CDC recommended that “TB control programs should evaluate … each death caused by TB to determine whether … the death could have been prevented.  Based on such a review, new policies should be developed and implemented to reduce the number of preventable … deaths.”¹

More than 10 years after these recommendations were published, there is no systematic method for determining which of these deaths are caused by TB, nor is there a method to assess preventability. CDC’s revised Report of Verified Case of Tuberculosis (RVCT),² to be implemented in 2009, will capture information about whether a death is related to TB disease or to TB medication.  A systematic approach is needed to accurately capture these new RVCT variables.

In order to respond to California’s local TB control programs’ requests for assistance in evaluating preventability of TB deaths, the California Department of Public Health (CDPH) Tuberculosis Control Branch (TBCB), with input from a working group of TB controllers, developed a tool to fill this important gap.  The TB Death Assessment Tool systematically examines 1) whether or not TB contributed to death for each case patient who dies with TB, and 2) missed prevention opportunities for TB-related deaths.  The goal of the tool is to assess TB-relatedness of deaths and gather information from missed opportunities, which will be used to guide public health action at the local and state levels to prevent future TB deaths.

Description of the Tool

To determine whether a death is TB-related, data are abstracted from the public health record, hospital and out-patient records, laboratory reports, imaging studies, autopsy, and death certificate.  The tool classifies deaths into one of five categories: definitely, possibly, not likely, definitely not, and unknown TB-related.  This classification is performed by using an algorithm that considers the extent, severity, and consequences of TB disease; adverse effects associated with TB medications; and adverse outcomes associated with medical procedures within 30 days prior to death.

In this classification, TB-related deaths are defined as either definitely or possibly related to TB.  For TB-related deaths, missed prevention opportunities are assessed.  Deaths that are classified as not likely, definitely not, or unknown TB-related are not evaluated further.  There are 35 missed opportunities in three categories: case detection (9), medical treatment (13), and case management (13).  Two of the missed opportunities are defined based on expert opinion; the remainder are based on state regulations or state or national guidelines.

Methods

From September to December 2006, the tool was pilot tested by using it to assess TB deaths in a sample of 20 M. tuberculosis culture-positive TB cases.  These cases were reported to four California local TB control programs from January 1, 2005, to June 30, 2006, with final treatment outcomes submitted to the State TB Registry by September 30, 2006.  The sample, drawn from a cohort of 54 eligible cases, consisted of five (25%) case patients who were diagnosed at death or alive at diagnosis but died before starting therapy; three (15%) who received less than 2 weeks of TB therapy; and 12 (60%) who received 2 or more weeks of TB therapy.  Cases were randomly selected from these groups in approximately the same proportion as each was present in the 2000 to 2004 statewide cohort of TB deaths. 

Local TB program staff provided copies of the public health record to state TB control staff, who reviewed the record for information gaps. If gaps were present, local staff attempted to obtain additional information, using all reasonable means, until the record was as complete as possible. 

Two CDPH investigators (a TB clinician and an epidemiologist) abstracted information from each patient record into an electronic version of the tool and used the tool’s algorithm to assess the contribution of TB to death for each case and evaluate missed opportunities for TB-related deaths. Missed opportunities were categorized as “potential” when information in the medical record was inadequate to assess the presence of a missed opportunity, but available information was suggestive.  Microsoft Excel 2003 was used for data analysis. 

To assess feasibility of the tool, the availability of essential patient information was determined and the time to complete the examination of case records using the tool was calculated.  To examine utility, the tool’s ability to determine the TB-relatedness of death and assess missed prevention opportunities was evaluated for each case.  Inter-rater reliability was examined by assessing the concordance of two investigators’ classification of the TB-relatedness of death for each case.

A survey instrument was developed to evaluate local TB control staff practices and needs in assessing TB deaths.  One staff member from each program completed the survey prior to the pilot, either by interview or in writing.

Results

The final pilot sample comprised 20 cases: 15 (75%) men and five (25%) women, with median age at death of 77 (range 46–93) years.  Fourteen (70%) patients were foreign-born and had been living in the United States for a median of 20 (range 3–30) years.

Eleven (55%) patients had pulmonary TB only, five (25%) had extrapulmonary TB only, and four (20%) had both pulmonary and extrapulmonary TB.  Of all cases, seven (35%) had acid-fast bacillus (AFB) smear-positive TB disease. Eighteen (90%) had pansensitive TB, one (5%) had multidrug-resistant TB, and one (5%) had isoniazid monodrug-resistant disease.  The following co-morbidities were observed in these 20 patients: acquired immunodeficiency syndrome (AIDS) and lung cancer (1), lung cancer and diabetes mellitus (DM) (1), end-stage renal disease (ESRD) and DM (1), AIDS (1), lung cancer (1), ESRD (1), and DM (2).  Other co-morbidities were also noted.

In the 20-case sample, six (30%) deaths were categorized as definitely TB-related, 11 (55%) as possibly TB-related, and three (15%) as unlikely TB-related.  The cases that were definitely TB-related included one patient with TB meningoencephalitis complicated by hydrocephalus, one with tubercular empyema, one with miliary TB and AIDS, one with pulmonary and lymphatic TB who died of a pulmonary hemorrhage, and two with extensive pulmonary TB.  No deaths were classified as definitely not or unknown TB-related.  Overall, 17 (85%) deaths were determined to be TB-related.

All 17 patients with TB-related deaths were hospitalized during diagnostic work-up and/or treatment.  For these patients, the median time from initiation of the diagnostic evaluation for TB to death was 38 (range 5–390) days, and, on average, patients spent half of this time hospitalized.  For the 10 patients with TB-related deaths who received at least 2 weeks of TB treatment, the median duration of treatment was 45 (range 23–346) days. 

Among the 17 TB-related deaths, the death certificate for nine (53%) listed TB as a cause of death. On the death certificates for the three deaths determined not to be TB-related, none listed TB.  Thus, the overall concordance between the tool and the death certificate in assessing TB as a cause of death was 60% (12 of 20). The death certificate is already known to have suboptimal sensitivity for TB.  Three of four (75%) patients with TB-related deaths who were dead at TB diagnosis did not have TB listed as a cause of death on the death certificate.  For these patients, confirmation of TB diagnosis was received after the death was certified.

Of the 17 TB-related deaths, eight (47%) patients had missed opportunities in case detection only; two (12%) patients had missed opportunities in case detection and medical treatment; and two (12%) patients had missed opportunities in case detection, medical treatment, and case management. No missed opportunities were found for five (29%) patients.

For the 12 (71%) patients for whom opportunities were missed, a total of 28 definite and five possible missed opportunities were documented. Specific missed opportunities are detailed below.

Definite Missed Opportunities
Type	n
Case detection	19
Provider delays in reporting cases to the local TB program	7
Provider diagnostic delay	6
Patient diagnostic delay	3
Lab delay in processing and/or reporting positive AFB smears or pathology consistent with TB to TB program	3
Medical treatment	5
Delay in TB treatment initiation	2
Inappropriate TB treatment	1
Inadequate management of adverse reactions to TB medications	1
Lab delay in processing and/or reporting initial drug susceptibility results to TB program	1
Case management	3
Inadequate assessment of indications for directly observed therapy (DOT)	1
Failure to initiate DOT	1
Inadequate transfers of TB patients	1
Other missed opportunities	1
Pharmacy transcription error	1
Total	28*

*Some cases had no missed opportunities, and some had >1

 

Possible Missed Opportunities
Type	n
Case detection	2
Provider diagnostic delays	2
Medical treatment	3
Delays in treatment initiation	2
Inadequate monitoring of adverse effects to TB medications	1
Total	5**

**All cases with possible missed opportunities also had >1 definite missed opportunity

The tool was deemed feasible in the pilot test.  Specifically, sufficient patient information was available to assess TB deaths, and the time required for all aspects of data collection and examination was approximately 4 hours per case. Inter-rater reliability in determining the TB contribution-to-death category was 95% (19 of 20). The tool appeared to accomplish what it was designed to do. Specifically, it was able to assess the contribution of TB to death and evaluate missed opportunities for TB-related deaths. However, there is no “gold standard” against which to compare the results and assess the accuracy of the tool.

In the survey of local program practices and needs to assess TB deaths, two of four programs reported conducting limited reviews of TB deaths.  These programs noted a number of challenges, including lack of a standardized method to assess the TB contribution to death and prevention opportunities; consequent difficulty using findings to inform interventions; insufficient time to conduct reviews of TB deaths; and information gaps in the patient record.  All four respondents considered examining prevention opportunities of TB-related deaths a program priority and noted the need for a structured, systematic, and timely review process that would minimize clinician time and that could readily be used to inform interventions.

Some limitations of this pilot test were noted.  First, the sample was relatively small, limiting generalizations beyond the study population.  Second, despite our attempts to obtain all pertinent medical records, we did not have complete records for all cases; for some, handwritten notes by health care providers were not legible. Inadequate medical documentation could result in inaccurate conclusions.  This limitation was addressed, in part, by categorizing missed opportunities as “possible” if information to fully assess a missed opportunity was lacking but available information was suggestive.  Third, California’s experience in assessing TB deaths may not be representative of other program experience.  Fourth, although the tool is an electronic form that allows the user to enter data, the lack of an electronic database linked to the entered data tended to limit data analysis.  

Conclusions

Use of this death assessment tool provided a systematic method for determining TB-relatedness of death and detected missed prevention opportunities.  In this sample, a high proportion of deaths were found to be TB-related, and a high proportion of TB-related deaths had at least one missed opportunity to prevent death. 

These findings may be used to inform interventions to prevent future deaths. The most common missed opportunities were provider diagnostic delays and provider delays in reporting suspected or confirmed cases to the local program (which could result in delay in treatment initiation or initiation of an inappropriate TB regimen). Interventions suggested by these findings include feedback to private providers regarding diagnostic standards for TB, including the importance of a high index of suspicion for patients with certain risk factors, use of empiric TB therapy, and special treatment regimens for patients with pre-existing liver disease; reporting requirements, including the reporting of TB suspects; and the need to update the death certificate when a TB diagnosis is made after death and TB contributes to death. Other suggested interventions include intensification of private provider oversight during treatment and hospitalization of TB patients with complex co-morbidities, expanding DOT to patients with co-morbid conditions, and strengthening of local program transfer of care during moves.

The pilot highlighted the substantial toll of TB on both the patients and society.  Interventions to prevent TB deaths may give patients the best chance of cure and survival and also result in substantial cost savings to society.

During 2008–2010, the CDC Tuberculosis Epidemiologic Studies Consortium will undertake a multisite mortality investigation of approximately 1,400 TB deaths which will extend beyond California.  Study objectives include assessing the proportion of TB-related deaths, patient predictors of mortality, and missed prevention opportunities.  Findings of this study, which will include a case-control study design, will be used to further test, revise, and finalize a systematic approach to assess TB deaths.

—Reported by Joan Sprinson, CA Dept of Public Health, TB Control Branch; Manisha Bahl, Univ of California–San Francisco School of Medicine; Robert Benjamin, Alameda Co. Public Health Dept, TB Control Program; Charles M. Crane, Contra Costa Health Services, TB Program; Lisa Gooze, San Mateo County Health Dept, TB Control; Connie Haley, Vanderbilt Univ; Masae Kawamura, San Francisco Dept of Public Health, TB Control; Marisa  Moore, DTBE, CDC and San Diego Co. TB Control Branch; Karen Redwine, Alameda Co. Public Health Dept, TB Control Program; Karen Smith, Public Health Div, Napa Co. Health and Human Services Agency; Francie Wise, Contra Costa Health Services, TB Program; Jennifer Flood, CA Dept of Public Health, TB Control Branch

References

1. CDC Essential components of tuberculosis prevention and control program. Recommendations of the Advisory Group for the Elimination of Tuberculosis. MMWR 1995; 44(No. RR-11).

2. U.S. Department of Health and Human Services Centers of Disease Control and Prevention. Form OMB 0920-0096.

 

Getting TB Suspect Information from Hospitals When You Need It–Quickly!

Sending a cover letter to hospitals with an information request form can foster rapid reporting of TB suspects and cases to the local health department. As soon as new TB suspects and cases are found by hospital staff or private physicians, they should be reported to local TB programs so initial interviews can be scheduled. There are a number of factors contributing to reporting delays.  Nurses, attending physicians, and resident physicians are often changing jobs and relocating to other cities or states. New personnel may not be acquainted with the local reporting requirements. Private physicians may see TB patients infrequently in their practice and may not know when to report or whom to contact at the local health department. Hospital staff may delay reporting TB suspects until culture results have been obtained.

Delays in reporting TB suspects and cases cause a number of concerns.  Patients may be discharged from the hospital before an interview can be scheduled; this could be problematic, especially if young children could be exposed at the returning patient’s home. Patients may be homeless and be discharged to the streets at the classic 4 pm on Friday, and may try to gain entry to a shelter while still infectious. Patients may be discharged without an adequate supply of TB medications. For these and other reasons, it is important for local TB programs to have an ongoing policy to ensure prompt notification of TB suspects and cases and to address notification problems when they occur.

To address these concerns, TB program staff members in some states have developed a cover letter for hospitals and a request-for-information form. In developing these documents, the TB controller, state nurse consultants, local TB program nurses, and physicians should all have an opportunity to review and tailor them as needed. For instance, in Houston, the TB control staff added the state reporting law in the first sentence of the hospital cover letter. They also included information on the Health Insurance Portability and Accountability Act (HIPAA) on the information request form to address concerns of reluctance to provide medical record data. The value of the request-for-information form is that it can expedite the collection of patient information and permit the review of primary information on the TB suspect by the TB program staff. This helps in setting priorities for interviews and contact investigations. To provide added weight to the importance of these documents, the Houston TB program included on the hospital cover letter the signatures of the TB bureau chief, the director of the health department, and the physician in charge of the Houston Health Authority.

In Houston, once the cover letter was completed, the clinical nurse case manager supervisor scheduled appointments and visits to various hospitals in the area. Accompanied by two TB program nurse case managers, she visited the infection control nurse and staff nurses at the hospital. This was a positive step because it allowed everyone to see and meet the colleagues they had been speaking to when reviewing patient cases by telephone. The nurses also brought TB literature to the hospital as well as a small Mr. TB Germ and a bag of jellybeans.  The visits began in early November 2007 to hospitals where delayed reporting had been noticed. Letters and request-for-information forms have now been sent, and visits made, to the infectious control nurses and staff at most of the hospitals that report TB suspects and cases. The TB control staff also fax the cover letter and form to other hospitals that rarely report TB. The fruits of these efforts have been a thank-you letter from one of the infection control nurses, stating she appreciated the time taken in coming to the hospital. In addition, there has been more prompt reporting of TB suspects at her hospital.

Lessons Learned. TB control staff should not assume that all hospitals and private physicians will report TB suspects promptly every time. As mentioned before, personnel come and go. It may take time for new staff to learn the local reporting requirements. In addition, stays at hospitals are getting shorter. TB programs need to know about these TB suspects and cases as soon as possible in order to interview them before discharge, or to discuss the need to keep the infectious TB patient longer in the hospital until suitable housing can be located.

Future Plans. Future plans include establishing a regular schedule of these visits to all hospitals in the Houston area that routinely report TB suspects and cases. In addition, the interviewers who visit these hospitals will discuss and provide the hospital cover letter and request-for-information form to nurses in the units housing the TB patients. Finally, efforts will also be made to provide this information to private physicians through office visits, telephone calls, or local conferences, or by working with the Houston Medical Society.

For more information, or for sample copies of either the cover letter or the information request form developed in Houston, please contact Ted Misselbeck.

—Reported by Ted Misselbeck
Div of TB Elimination