Research Briefs

Alexander, G.C. (2008, Spring). "Commentary on the case of Mr. A.B.: Dilemmas for a reason." (AHRQ grant HS15699). The Journal of Clinical Ethics, pp. 70-71.

The author comments on the following ethical dilemma: should the staff not tell Mr. A.B. about the tragic loss of his daughter prior to a planned bypass surgery, as his family asks? The medical team must grapple with a classic tension between beneficence and patient autonomy. A complicating factor is the difficulty of knowing how the timing of the message will affect Mr. A.B., because the physiologic consequences of grief in the context of his critical illness are not clear. Given his comorbid conditions and clinical instability, Mr. A.B. is more likely than most patients to be at increased risk of a poor outcome triggered by acute grief. As with many cases of delivering bad news, the crux of the issue is how, not whether, it should be told. The clinician shares with the family a responsibility for the situation. In this case, the team reached a decision through thoughtful deliberation and careful consultation with the affected parties. Reasonable people, of course, may differ about this or any such decision.

Arora, V. and Meltzer, D. (2008, June). "Effects of ACGME duty hours on attending physician teaching and satisfaction." (AHRQ grant HS10597). Archives of Internal Medicine 168(11), pp. 1226-1227.

The implementation of restricted duty hours by the Accreditation Council for Graduate Medical Education (ACGME) has raised concerns regarding possible negative effects not only on resident education and patient care, but also on teaching faculty. The researchers studied the effects of restricted duty hours on inpatient attending physician teaching and satisfaction using data collected for 5 years before and after implementation at a single institution. After implementation of restricted duty hours, attending physicians reported fewer hours of teaching more times per week when residents missed conferences due to rounds, and a lower percentage of patients seen on the day of admissions. They were also less satisfied with time for teaching, ability to determine patient length of stay, and influence on hospital policy. The researchers caution that since the survey was originally designed to test the effects of hospitalists, it may have missed other effects of duty hours.

Brady, J., Ho, K., and Clancy, C.M. (2008). "Slowed progress in improving quality and minimizing disparities." AORN Journal 87(5), pp. 1007-1009.

In its fifth annual National Healthcare Quality Report and the National Healthcare Disparities Report, the Agency for Healthcare Research and Quality (AHRQ) finds that the rate of quality improvement appears to be slowing. Overall quality improved by an average of just 1.5 percent per year between the years 2000 and 2005. This represents a decline when compared with the 2.3 percent average annual rate over the longer reporting period of 1994 to 2005. Some areas have improved, such as counseling to quit smoking and reduced disparities in childhood vaccinations, according to the 2007 reports. However, measures of patient safety showed an average annual improvement of only 1 percent. Part of improving health care quality is reducing the variation in health care delivery across the country, so that patients in all States receive the same level of high-quality, appropriate care. On the average, since the year 2000, variation has decreased across the measures for which State data is tracked, but this progress is not uniform.

Bravata, D.M., Shojania, K.G., Olin, I., and Rave, A. (2008). "CoPlot: A tool for visualizing multivariate data in medicine." (AHRQ Contract No. 290-02-0017). Statistics in Medicine 27, pp. 2234-2247.

Many research questions in medicine require the analysis of complex multivariate data. Multidimensional scaling (MDS) facilitates the analysis of multivariate data by reducing the multidimensional data into a two-dimensional structure that attempts to uncover the hidden structure in a data set by creating a pictorial representation of the data. CoPlot, an adaptation of MDS, addresses two key limitations of MDS—the inability to simultaneously map the variables and observations and the lack of orientation of the MDS map. CoPlot maps the observations and variables in a manner that preserves their relationships, allowing richer interpretation of the data. The authors describe CoPlot and its methodology and present the results of the application of CoPlot to multivariate data describing clinical presentations and treatment responses of children infected with anthrax. They also provide recommendations for the use of CoPlot for evaluating and interpreting other health care data sets.

Clancy, C.M. (2008, May/June). "Evidence shows cost and patient safety benefits of emergency pharmacists." American Journal of Medical Quality 23(3), pp. 231-233.

Adverse drug events (ADEs) account for nearly 20 percent of adverse events overall and hospital emergency departments (EDs) have the highest rates of reportable errors, notes the author of this commentary. There is some evidence that use of ED pharmacists results in substantially lower rates of ADEs. One hospital found that within 6 months of employing an emergency pharmacist, medication errors in the ED had dropped by 50 percent. A recent survey of pharmacy directors found that 10.5 percent of level 1 trauma centers employed an emergency pharmacist (EPh), while only 2.7 percent of hospitals without a level 1 trauma center employed an EPh.

Clancy, C. (2008, June). "AHRQ's Effective Health Care Program: Accelerated access to decisive clinical information." The Journal of Family Practice S20-S23.

To mine the potential of practice-based research networks and existing literature, the Agency for Healthcare Research and Quality (AHRQ) has formed partnerships with major universities, health maintenance organizations, and independent office practices. According to Carolyn M. Clancy, M.D., director of AHRQ, the goal of these efforts is to develop better, more accessible clinical information and to disseminate it faster. One such partnership is Developing Evidence to Inform Decisions about Effectiveness (DEcIDE), a consortium of health care organizations that conduct rapid practical research in different topic areas with existing patient databases. AHRQ has also established 14 Evidence-based Practice Centers (EPCs) to compile results of scientific literature, promote evidence-based practice on various clinical topics, and conduct comparative effectiveness reviews on medications, devices, and other interventions. In addition, AHRQ has established the Eisenberg Clinical Decisions and Communications Science Center to translate the complex knowledge gleaned by DEcIDE and the EPCs into practical tools for clinicians, patients, and policymakers. Readers are invited to suggest topics for research in progress (visit http://effectivehealthcare.ahrq.gov).

Clancy, C. (2008, June). "Improving care quality and reducing disparities: Physicians' roles." Archives of Internal Medicine 168(11), pp. 1135-1136.

As the U.S. population becomes increasingly diverse, there is growing urgency to identify solutions to the challenges of "unequal treatment." Most studies confirm that disparities remain pervasive, even though these differences vary by specific racial or ethnic group, specific disease condition, and specific region, health plan, or hospital. The author of this paper, Carolyn M. Clancy, M.D., director of the Agency for Healthcare Research and Quality, discusses a research article focusing on the role of variation among individual physicians in contributing to observed disparities in quality of care. That article found that rates of achieving target blood sugar and cholesterol levels were significantly lower for black patients than for white patients. Within-physician effects rather than patient sociodemographic or clinical factors explained most of these observed differences. Since hypotheses differ as to why this is the case, Dr. Clancy calls for more research to better understand how to close such gaps in disparities of care and for physician leadership to assure that the care provided is evidence-based, patient-centered, effective, consistent, and equitable.

Dalby, D.M. and Hirdes, J.P. (2008). "The relationship between agency characteristics and quality of home care." (AHRQ grant HS09455). Home Health Care Services Quarterly 27(1), pp. 59-74.

To assess the relationship between Canadian home health care agency characteristics and the quality of care, the researchers examined data from 12 agencies in Ontario and the Winnipeg region of Manitoba and 11,767 of their adult home care clients. These agencies used a set of home care quality indicators (HCQIs) based on the Resident Assessment Instrument for Home Care (RAI-HC), a questionnaire to which their clients responded. The agencies also responded to a mailed survey about their characteristics. The researchers found that agencies that served a larger population had lower quality of care with respect to 11 of the HCQIs and the overall summary measure of quality. Home care case managers had between 90 and 130 clients and agencies with fewer clients per case manager had better performance.

Dougherty, D. and Conway, P.H. (2008, May). "The '3T's' road map to transform US health care." Journal of the American Medical Association 299(19), pp. 2319-2321.

The authors outline a model or "road map" to transform the U.S. health care system by accelerating the pace at which innovations are implemented in clinical settings and addressing the "how" of health care delivery. Translation 1 (T1) is basic science and its translation into clinical research; Translation 2 (T2) focuses on practice guidelines and tools for patients; Translation 3 (T3) activities address the "how" of health care delivery so that evidence-based treatment, prevention, and other interventions are delivered reliably to all patients in all settings of care. The 3T's model of transformation has four main activities: 1) measurement and accountability, 2) implementation and system redesign, 3) scaling and spread, and 4) research. For these activities to truly transform the health system to achieve the goal of high-quality care and better health outcomes, the key facilitators of leadership, teamwork, tools, and resources must be established and integrated.

Edwards, J.C., Kang, J., and Silenas, R. (2008, Summer). "Promoting regional disaster preparedness." (AHRQ grant HS13715). Journal of Rural Health 24(3), pp. 321-325.

Rural as well as urban areas need to be prepared for natural disasters. Rural hospitals face multiple obstacles to preparedness. The authors created and implemented a simple and effective planning and training exercise to assist rural hospitals to improve disaster preparedness and enhance regional collaboration among these hospitals. The intervention selected for use was a tabletop exercise, one of the training methods from the discipline of emergency management. Tabletop exercises are scripted scenarios depicting a public health emergency that are led by a facilitator who asks a group of participants to respond to a series of incidents in the scenario. The 3-hour exercise emphasizing regional issues in an avian flu pandemic was followed by a 1-hour debriefing. The exercise was successful in identifying problems such as insufficient staff for incident command, facility constraints, and the need to develop more regional cooperation.

Ensign, J. and Ammerman, S. (2008). "Ethical issues in research with homeless youths." (AHRQ grant HS11414). Journal of Advanced Nursing 62(3), 365-372.

As a vulnerable group, homeless youths are the focus of an increasing body of research documenting their unique health and social needs. Unlike research with other adolescents, there are no specific guidelines for the ethical conduct of research with homeless youth. Specific ethical issues concern the ability of minor homeless youths to consent to participation in research and what constitutes appropriate research payments for homeless youths. The authors used a 10-15 minute questionnaire to document researcher, health care provider, and program administrators' experiences with ethical issues of research with homeless youths in the U.S. and Canada. Of the 72 respondents to the questionnaire, 37 respondents obtained written consent and 14 oral consent from the youths. Overall, 27 of the researchers used money as an incentive, while 26 used vouchers or gift cards, 16 used food, and 10 used no incentive. Mental health and/or substance abuse researchers tended to use money as an incentive, while health care providers and program administrators tended to use nonmonetary incentives.

Fiscella, K. and Meldrum, K. (2008). "Race and ethnicity coding agreement between hospitals and between hospital and death data." (AHRQ grant HS10910). Medical Science Monitor 14(3), pp. SR9-13.

Reliable coding of race and ethnicity by hospitals represents a critical step toward assessing and addressing racial and ethnic disparities in acute inpatient care. The authors of this study used California State Inpatient Data and matching vital data to examine the reliability of race and ethnicity by hospitals and death certificates for 1998-2000. The analysis focused on the rates of agreement for race and ethnicity coding among patients admitted to different hospitals, the rates of agreement of coding between hospital data and death certificate data among 548,006 persons, and the rates of agreement for coding Asian and Hispanic ethnicity between hospital and death data by country of origin coded on death certificate data. The rates of agreement were high for most major racial and ethnic groups both among hospitals and between hospitals and death certificates but were too low for American Indians, persons born in India, and selected Hispanic subgroups. Among patients admitted to different hospitals, the rates of agreement were 98 percent for Asians, 91 percent for Whites, 88 percent for Blacks, 19 percent for American Indian, 26 percent for "Other" race, and 4 percent for unknown race.

Fuhlbrigge, A., Carey, V.J., Finkelstein, J.A., and others (2008). "Are performance measures based on automated medical records valid for physician/practice profiling of asthma care?" (AHRQ grant HS08368). Medical Care 46, pp. 620-626.

The researchers sought to determine if physician practices in treating children with asthma could be reliably ranked using the Health Employers Data Information System (HEDIS) performance measure plus three other measures. The research design was based on a simulation describing the relationship between practice size and precision of practice measures to estimate performance. The study group consisted of 39 practices with a total of 1,457 children meeting the criteria for persistent asthma. The main outcome was reproducibility of the HEDIS measure and three other measures (proportion of children with asthma-related hospitalization, emergency department visits, and oral steroid dispensings for asthma). Of the four measures evaluated, none achieved a reproducibility of greater than 85 percent for a practice size of 50 or less. Only with a practice size larger than 100 children with persistent asthma was reproducibility greater than 85 percent for all measures. The researchers concluded that only at the level of the health care organization can the asthma measures available within claims data be used to reliably rank physician/practice performance.

Hazlehurst, B., Gorman, P.N., and McMullen, C.K. (2008). "Distributed cognition: An alternative model of cognition for medical informatics." (AHRQ grant HS12003). International Journal of Medical Informatics 77, pp. 226-234.

Cognitive science has provided medical informatics with theory, methods, and findings for understanding clinical knowledge, problemsolving, decisionmaking, and other cognitive phenomena in health care. Classical cognitive theory takes the individual person as the relevant unit of cognitive analysis, and medical informatics has largely inherited this approach in its teaching, research, and design of workplace technologies. However, health care activities include resources internal to individuals as well as resources provided by workplace tools and technologies and the organization of individuals into task-oriented and role-based groups and collaborative teams. The authors review developments in cognitive science that have generated a theory of distributed cognition, where the unit of analysis is the activity system, which includes individual agents, their technologies and tools, and their understandings, roles, and relationships defined by their history of interaction. They argue that the theory of distributed cognition is relevant to medical informatics in its efforts to understand and to improve information processing in health care.

Henderson, G., Garret, J., Bussey-Jones, J., and others (2008, March). "Great expectations: Views of genetic research participants regarding current and future genetic studies." (AHRQ grant T32 HS00032). Genetics in Medicine 10(3), pp. 193-200.

People who have already participated in genetic research represent an important resource for future genetic research. However, few studies have investigated their attitudes about their willingness to participate in future genetic research. The researchers interviewed 801 black and white individuals who had participated in a genetic epidemiology study of colon cancer risk factors. Overall, 63 percent felt "very positive" and 32 percent felt "positive" about research looking at whether genes put people at risk for disease or illness. Most reported being "very likely" (49 percent) or "somewhat likely" (40 percent) to participate in a genetic research study in the future. When asked to list the good things about such research, respondents cited the potential for discovering the causes and cures of various types of cancer, and increased awareness and preventive health activities that might reduce or eliminate the risk of disease. When asked to list the "bad things," over half of the subsample of 194 said "none." The most common expressed concern was about the implications of genetic information.

Kuzniewicz, M.W., Vasilevskis, E.E., Lane, R., and others (2008). "Variations in ICU risk-adjusted mortality. Impact of methods of assessment and potential confounders." (AHRQ grant HS13919). Chest Journal 133, pp. 1319-1327.

The intensive care unit (ICU) has become a focus of efforts to improve the quality of care because of the growing number of severely ill hospital patients and the increased complexity of care. In order for Federal and State agencies to assess ICU performance, an accurate method of measuring performance must be selected. The objectives of this study were to determine whether substantial variation in ICU mortality performance still exists in modern ICUs and to compare the updated ICU risk-prediction models for predictive accuracy, reliability, and data burden. The three models compared in the study were the mortality probability model (MPM) III, the simplified acute physiology score (SAPS) III, and the acute physiology and chronic health evaluation (APACHE) IV. A total of 11,300 patients from 35 California hospitals were used to compare the models. The researchers found substantial variation in ICU risk-adjusted mortality rates, regardless of the risk adjustment model used. Since there is no "gold standard" against which to judge the available models, there is no way to tell if variations in outcomes represent true differences in performance or merely the inability of the models to account for unmeasured differences in case mix.

Lanoue, E. and Still, C.J. (2008, May/June). "Patient identification: Producing a better barcoded wristband." (AHRQ grant HS15270). Patient Safety & Quality Healthcare.

Patient identification is the cornerstone of patient safety. The use of barcode technology for patient identification is a growing trend at many health care organizations. After deciding that a radio-frequency identification system was too new and costly, a small Vermont-based health system explored barcode symbologies and print options in light of the needs of patients, clinicians, and the hospital mission. The system switched from using laser-generated labels, which were easily damaged, to using thermal printers specifically designed for barcoding that produce durable barcodes easily scanned at the point of care. These wristbands can survive a week of wear and temperatures up to 130 degrees. In addition, there is not a significant cost difference between thermal and laser printers.

Lee, I., Thompson, S., Lautenbach, E., and others (2008). "Effect of accessibility of influenza vaccination on the rate of childcare staff vaccination." (AHRQ grant HS10399). Infection Control and Hospital Epidemiology 29(5). pp. 465-467.

In 2002, the Centers for Disease Control and Prevention expanded its influenza vaccination recommendations to include children aged 6-23 months and their caregivers, and again in 2006 to include children aged 6-59 months and their caregivers. The researchers conducted a study to evaluate the impact of free on-site vaccination on childcare staff vaccination rates. Included in the study were four influenza seasons: 2002-2003, 2003-2004, 2005-2006, and 2006-2007. Free on-site vaccinations were offered in the 2003-2004 and the 2006-2007 seasons. Vaccination rates were markedly higher in the two intervention seasons (51 percent in 2003-2004, 45 percent in 2006-2007) than in the two nonintervention seasons (28 percent in 2002-2003, 26 percent in 2005-2006). Of those vaccinated, two-thirds said they would not have been vaccinated without the intervention, with one-third stating that they would not have been vaccinated if required to pay for it.

Leonard, C.E., Haynes, K., Localio, A.R., and others (2008). "Diagnostic E-codes for commonly used, narrow therapeutic index medications poorly predict adverse drug events." (AHRQ grant HS11530). Journal of Clinical Epidemiology 61, pp. 561-571.

Adverse drug events (ADEs) are a common, major clinical problem resulting in patient morbidity and mortality and increased cost of care. The researchers sought to determine the validity of hospital discharge E-codes (External-Cause-of-Injury codes) in identifying drug toxicity precipitating hospitalization among elderly users of warfarin, digoxin, or phenytoin. Included in the study were 4,803 patients with 11,409 person-years of experience with at least 1 of the 3 drug groups and with 8,756 hospitalizations of which 304 were considered to be ADEs related to the use of warfarin, digoxin, or phenytoin. The researchers found that the positive predictive values of E-codes indicating toxicity were too low to confirm hospitalizations due to ADEs for any of the three drugs. They concluded that such rare events identified in studies based solely on discharge coding should be confirmed by medical record review.

Lieberthal, R.D. (2008, June). "Hospital quality: A PRIDIT approach." (AHRQ grant T32 HS00009). HSR: Health Services Research 43(3), pp. 988-1005.

Measuring hospital care quality can be done by measuring inputs to care (process measures) or measuring outputs from care (patient outcomes). This study used the PRIDIT method to retro-spectively analyze Medicare hospital data to determine a relative measure of hospital care quality for 4,217 hospitals. The PRIDIT approach uses 20 care process measures in 4 categories: heart attack care, heart failure care, pneumonia care, and surgical infection prevention and five structural measures of hospital type. The best indicators of hospital quality were heart failure patients given an assessment of left ventricular function and heart attack patients given a beta-blocker at hospital arrival and at discharge. Hospital teaching status was also an important indicator of higher quality of care. The author concludes that the PRIDIT method is an alternative to the use of clinical outcome measures in measuring hospital quality.

Ness, R. and Grainger, D.A. (2008). "Male reproductive proteins and reproductive outcomes." (AHRQ grant HS08358). American Journal of Obstetrics and Gynecology 198, pp. 620.e1-620.e4.

Male reproductive proteins (MRPs) are among the most rapidly evolving functional genes known. They play a central role in embryo implantation and placentation by inducing inflammation. No studies to date have directly examined whether and how MRPs have an impact on the efficacy and outcomes of human reproduction. There is some evidence to suggest that novel exposure to MRPs may elevate preeclampsia risk since first pregnancies, teenage pregnancies, out-of-wedlock pregnancies, paternity change, and donor sperm insemination are markers that increase the risk for preeclampsia. MRPs may also contribute to spontaneous preterm birth. If novel MRPs influence reproductive outcomes, donor insemination might provide a useful framework to demonstrate this. The proposed link between MRPs and adverse pregnancy outcomes in humans is speculative. Direct measures relating specific sperm and semen proteins to conception and birth outcomes will be needed to test the import of MRPs.

Osler, T., Glance, L., Buzas, J.S., and others (2008). "A trauma mortality prediction model based on the anatomic injury scale." (AHRQ grant HS16737). Annals of Surgery 247, pp. 1041-1048.

Improving outcomes, whether through private comparisons among trauma centers or public report cards, depends on the ability to accurately compare performance across trauma centers. This, in turn, depends on accurate statistical adjustment, especially for the severity of injury. Measuring this factor is difficult, because there are hundreds of possible individual injuries, and patients often have more than one injury. The researchers developed three new models based on empiric estimates of injury severity in the abbreviated injury scale (AIS) and tested them against the injury severity score (ISS). For each of the 1,322 AIS coded injuries, the researchers derived an empiric severity model-averaged regression coefficient (MARC) value ranging from -1.01 for a trivial injury to a value of 4.03 for an unsurvivable injury. Applied to a National Trauma Data Bank containing information on 702,229 patients, these continuous values provided much greater granularity than the 6 integer values available to the AIS. All three of the models discriminated survivors from nonsurvivors better than the ISS, but the trauma mortality prediction model had both better discrimination and better calibration than the ISS.

Quigley, D.D., Elliott, M.N., Hays, R.D., and others (2008, July). "Bridging from the Picker hospital survey to the CAHPS® hospital survey." (AHRQ grants HS09204 and HS16980). Medical Care 46, pp. 654-661.

The Consumer Assessment of Healthcare Providers and System (CAHPS®) Hospital Survey compares performance information on hospitals to inform consumer choice and provide incentives to hospitals to improve the care they provide. With the introduction of the CAHPS® Hospital Survey in early 2007, hospitals that previously used other surveys are faced with bridging and trending issues as they adopt the new survey. The authors of this paper illustrate an accessible method of bridging data from earlier surveys to the CAHPS® Hospital Survey to support hospitals' internal quality improvement efforts. They administered 6 pairs of parallel items from the CAHPS® and Picker Hospital Surveys to the same 734 patients. Differences in wording, response options, and cut points for "problem scores" yielded large differences in problem score rates between the Picker and CAHPS® Hospital Surveys, which required bridging formulas that they detail.

Return to Contents
Proceed to Next Article