Research Activities, March 2010, No. 355
Alexander, G. L. and Madsen, R. (2009). "IT sophistication and quality measures in nursing homes." (AHRQ grant HS16862). Journal of Gerontological Nursing 35(7), pp. 22-27.
The relationship between the degree of information technology (IT) sophistication and quality measures of care at nursing homes is strongest when IT is used to manage resident care and clinical support, found this study of 210 nursing homes in Missouri. The researchers reported comparable IT sophistication scores in 3 domains (resident care management, clinical support, and administrative processes) for 95 nursing homes that made quality measure data available to the researchers and 115 that did not. Among the nursing homes providing quality measure data, the most commonly occurring quality measure issue was incontinence (mean of 39.4 percent of residents), followed by a decline in activities of daily living, or ADLs, (mean of 13.4 percent), and worsening locomotion (mean of 10.6 percent). The researchers found weak but significant correlations between the use of IT for resident care management and reduction in ADL decline.
Alper, S. J. and Karsh B.-T. (2009). "A systematic review of safety violations in industry." (AHRQ grant HS13610). Accident Analysis and Prevention 41(4), pp. 739-754.
Instead of blaming workers for violating safety rules, managers should try to understand the factors in the work environment that precipitate these violations, note the authors of this review of the literature. They identified 13 studies that discussed intentional violations of safety procedures and norms without a desire to cause harm (termed "nonmalevolent violations"). Of these studies, five were concerned with health care; the other fields reported on were commercial driving, aviation, aviation maintenance, mining, railroad transportation, and construction. The studies tested 57 different variables to determine their association with violations. The researchers grouped these variables into six categories (individual characteristics, information/education/training, design to support worker needs, safety climate, competing goals, and problems with rules). In some cases, violation of poorly designed rules was found to be necessary to increase safety.
Butt, A. A., Tsevat, J., Leonard, A., and others. (2009). "Effect of race and HIV co-infection upon treatment prescription for hepatitis C virus." (AHRQ grant HS13220). International Journal of Infectious Diseases 13, pp. 449-455.
These researchers compared treatment of 241 patients with hepatitis C virus (HCV) infection and 158 patients with HCV and HIV co-infection from 3 academic medical centers. They found that HIV co-infection was an independent predictor of not receiving treatment for HCV. This result persisted even after researchers controlled for such factors as patient age, minority race, and coexisting medical conditions. The researchers offer several explanations. First, HCV/HIV co-infected patients have lower response rates to HCV medications compared with patients with just HCV infection. Co-infected patients receiving treatment for HIV are also at risk for a number of drug interactions with ribavirin, a common therapy for HCV. Blacks were less likely to receive treatment for HCV compared with whites. The researchers attribute the treatment disparities for blacks with HCV to their reduced likelihood of achieving good treatment responses to HCV medications compared with whites.
Castle, N. G., Hanlon, J. T., and Handler, S. M. (2009). "Results of a longitudinal analysis of national data to examine relationships between organizational and market characteristics and changes in antipsychotic prescribing in U.S. nursing homes from 1996-2006." (AHRQ grant HS16547). The American Journal of Geriatric Pharmacotherapy 7, pp. 143-150.
In order to help patients who are agitated or behave aggressively, nursing homes use antipsychotic medications. Currently, between 24.8 percent and 27.6 percent of patients in nursing homes receive these drugs. In some cases, nursing home caregivers may use these drugs inappropriately in an attempt to keep patients quiet and less troublesome to staff. The researchers tracked the use of these drugs between 1996 and 2006 with data from the Medicare and Medicaid programs. Factors such as nursing home for-profit status, chain membership, competition levels, and Medicaid reimbursement were included in the analysis. Overall, the use of antipsychotics in the facilities studied increased by almost 60 percent during the time period. Theresearchers also found that for-profit nursing homes used more antipsychotic drugs than others. Factors resulting in a lower rate of antipsychotic use included chain membership, increased Medicaid reimbursement, and increased market competition.
Chung, J. W. and Meltzer, D. O. (2009, November). "Estimate of the carbon footprint of the US health care sector." (AHRQ grant HS16967). Journal of the American Medical Association 302(18), pp. 1970-1972.
To quantify the environmental impact of health care—its carbon footprint, these researchers estimated total greenhouse gas (GHG) emissions for the U.S. health care sector. They accounted for carbon dioxide, methane, nitrous oxide, and chlorofluorocarbons. They estimated GHG emissions using 2007 data on health expenditures by the National Health Accounts Team and the Environmental Input-Output Life-Cycle Assessment Model developed by Carnegie Mellon University. Both the direct effects of health care activities and indirect effects that included upstream supply-chain effects were measured. In 2007, the health care sector accounted for 16 percent of the U.S. gross domestic product. The total effects of health care activities contributed 8 percent of total U.S. GHG and 7 percent of U.S. total carbon dioxide emissions as measured in millions of metric tons of carbon dioxide equivalent.
Clancy, C. M. (2009, November.) "The canary's warning: Why infections matter." American Journal of Medical Quality 24(6), pp. 462-464. Reprints (AHRQ Publication No. 10-R018) are available from the AHRQ Publications Clearinghouse.
Health care organizations that desire a brief but accurate view of their patient safety performance can use the Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicator (PSI) #7 "Selected Infections Due to Medical Care" as a simple barometer, suggests the Director of AHRQ in this paper. Constructing national trends of safety outcomes is difficult because of data limitations. The AHRQ PSI #7 is useful as a canary measure because the data that can be collected for it are significant, and its results strongly correlate with health care organization performance on other PSIs. Also, the PSI #7 is useful because of the nature of health care-related infections, which have become the most common complication of hospital care, notes the author.
Clancy, C. M. (2009, November). "Ten years after To Err is Human." American Journal of Medical Quality 24(6), pp. 525-528. Reprints (AHRQ Publication No. 10-R019) are available from the AHRQ Publications Clearinghouse.
A decade ago, the discovery of the large numbers of people who die each year from medical errors in U.S. hospitals came as a surprise, even to experts in health care quality and patient safety. Since that time, the Agency for Healthcare Research and Quality (AHRQ), as the lead Federal agency in supporting Federal research on efforts to reduce patient harms, has built a wider and deeper evidence base about the root causes of errors. AHRQ has also created patient safety tools that providers have begun using to address medical errors in a systematic way. For example, AHRQ's suite of patient safety culture surveys help hospitals, nursing homes, and medical offices to assess, improve, and monitor their patient safety performance. Other areas of AHRQ involvement include reducing health care-associated infections, reducing hospital readmission rates, and understanding the relationship between medical resident fatigue and medical errors.
DeFrank, J. T., Rimer, B. K., Gierisch, J. M., and others. (2009). "Impact of mailed and automated telephone reminders on receipt of repeat mammograms." (AHRQ grant T32 HS00079). American Journal of Preventive Medicine 36(6), pp. 459-467.
Only two-thirds of U.S. women over 40 years and under 74 years old report having had mammograms within the past 1-2 years. To see if they could increase the proportion of women with recent mammograms, the researchers tested the effectiveness of automated telephone reminders (ATRs), enhanced reminder letters, and standard letters on the likelihood of repeat mammograms. The study included 3,547 adult women who were randomly assigned to 3 different groups. The researchers found that ATRs (76 percent) were more effective in ensuring repeat mammography than either standard (74 percent) or enhanced reminder letters (72 percent), even though the information contained in the ATR and the standard letter was the same. The telephone reminders were also the least costly of the three interventions. Overall, 74 percent of the women in the study had a repeat mammogram within 10-14 months after the previous mammogram compared with 57 percent before the reminders.
Gallagher, P., Ding, L., Ham, H. P., and others. (2009, November). "Development of a new patient-based measure of pediatric ambulatory care." (AHRQ grant HS09205). Pediatrics 124(5), pp. 1348-1354.
There are validated versions of the Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey for assessing pediatric care. However, many in the field thought that the pediatric CAHPS would be improved by including questions about developmental and preventive care. In this study, the researchers developed and tested two new pediatric CAHPS multi-item scales (composites) of developmental surveillance and preventive care that met high standards of reliability and validity. As part of this process, they conducted two focus groups and nine cognitive interviews. The multi-item scales were then tested on a group of 670 parents. The researchers found that reliable care assessments at the physician level can be obtained for both composites with as few as 50 patients. The CAHPS Consortium approved the new instrument, CAHPS-CG Child Primary Care, which incorporates the two composites, for assessing ambulatory pediatric care by clinicians and groups.
Henderson, M. J. (2009). "Synthesis of 'statistical innovations for cost effectiveness analysis.' Translating Research into Policy and Practice (TRIPP)," in: Hofmann, B. R., ed. Health Care Costs: Causes, Effects and Control. Hauppauge, NY: Nova Science Publishers, 2009, pp. 121-151. (Reprints (AHRQ Publication No. 10-RG001) are available from the AHRQ Publications Clearinghouse.
This article pulls together the findings of a 6-year study of new statistical models and methods for improving cost-effectiveness analysis (CEA) in health services research. CEA is seen as a useful tool for setting priorities in expenditures for health care programs, allowing comparison in costs and benefits between alternative interventions to improve health. The methods developed by the study allow analysts to express and quantify the degree of uncertainty in estimates of cost and benefit. Applications of the new models and methods include calculating the cost-effectiveness of the implantable cardioverter defibrillator and technologically improved versions of the device. The author of the article suggests that these methods may prove useful in evaluating the Medicare Part D prescription drug benefit.
Henriksen, K., Joseph, A., and Zayas-Cab�n, T. (2009, December). "The human factors model of home health care: A conceptual model for examining safety and quality concerns." Journal of Patient Safety 5(4), pp. 229-236. Reprints (AHRQ Publication No. 10-R020) are available from the AHRQ Publications Clearinghouse.
As the U.S. population ages, the rise in multiple chronic and acute care needs among individual patients is placing new demands on home health care. Patients can be discharged from the hospital while not fully recovered. At the same time, within the past decade, a growing number of sophisticated medical devices and equipment originally designed for use by trained personnel in hospitals and clinics are now used at home. The authors examine the human factors challenges associated with these converging trends in the home health care sector. Based on an analysis of literature related to home health care, they present a sociotechnical systems conceptual model in order to explore safety and quality concerns. The model consists of five tiers: the external environment, the physical and social/community environments together with medical devices and technology, the nature of home health care tasks, provider characteristics, and patient characteristics.
Lopez, L., Weissman, J.S., Schneider, E.C., and others. (2009, November). "Disclosure of hospital adverse events and its association with patients' ratings of the quality of care." (AHRQ grant HS11928). Archives of Internal Medicine 169(20), pp. 1888-1894.
To discover the relationship between the disclosure of hospital adverse events and patients' ratings of the quality of care, the researchers surveyed a random sample of 603 medical and surgical acute care adults patients in Massachusetts hospitals in 2003. Adverse events (AEs) are injuries caused by medical management rather than an underlying condition of the patient. Overall, the 603 patients reported 845 AEs. Forty percent of the AEs were disclosed to the patients by someone from the hospital, who explained why the negative effects occurred. Patients who reported an AE that had been disclosed to them gave a higher rating of the quality of care than patients who reported an AE that was not disclosed. Clinicians were less likely to disclose an AE that was associated with a more prolonged impact on the patient. The researchers suggest that a disclosure gap remains, even though patients rate their care favorably when AEs are disclosed.
McAuley, W. J., Spector, W., and Van Nostrand, J. (2009). "Formal home care utilization patterns by rural-urban community residence." Journal of Gerontology: Social Sciences 64(2), pp. 258-268. Reprints (AHRQ Publication No. 10-R003) are available from the AHRQ Publications Clearinghouse.
The researchers used data taken from the Medical Expenditure Panel Survey for 2002 and 2003 to look at home care utilization patterns in various residential areas. These areas included large metropolitan counties, micropolitan counties with towns of 10,000 to 50,000 residents, and remote counties with towns of less than 2,500 residents. Residents of micro counties were nearly twice as likely as metro residents to use home care and residents of remote counties had a three times greater likelihood of using home care. More home care provider days were received by users residing in counties adjacent to micro counties compared with those living in nonmetro counties next to metro counties and those residing in remote counties. Factors associated with a greater use of formal home care included increasing age, being white, having Medicaid, and being a nonsenior with Medicare.
Malat, J., van Ryan, M., and Purcell, D. (2009). "Blacks' and whites' attitudes toward race and nativity concordance with doctors." (AHRQ grant HS13280). Journal of the National Medical Association 101(8), pp. 800-807.
Very little is known about patients' general attitudes toward doctors' race and nativity. In addition, there is little research on how American-born patients evaluate care from foreign-born doctors. In order to explore these questions, the researchers decided to focus on how patients felt about the two dimensions of technical competence and interpersonal skills. Using a telephone survey conducted in the spring of 2004, the researchers queried 695 whites and 510 blacks in Hamilton County, Ohio. They found that black patients were more likely to believe that same-race doctors better understand their health problems (27 percent vs. 12 percent) and expected to be more at ease with same-race doctors than white patients (27 percent vs. 20 percent). Blacks were also more likely than whites to believe that U.S.-born doctors better understood their health problems and expected to be more at ease with U.S.-born doctors.
Manwell, L. B., Williams, E. S., Babbott, S., and others. (2009, May). "Physician perspectives on quality and error in the outpatient setting." (AHRQ grant HS11955). Wisconsin Medical Journal 108(3), pp. 139-144.
In this study, 32 primary care physicians participated in 9 focus groups in 5 cities to identify workplace factors that were related to greater or reduced safety and error. Participant comments were sorted into three major areas (factors affecting quality, factors affecting errors, and cross-cutting factors that affected both quality and errors). Many participants attributed limitations on quality of care to differences in basic values between staff physicians and practice leadership, to the availability of resources (supplies, medications, and referrals), and to lack of job control and inability to participate in decisionmaking. Factors influencing errors included the lack of formal error reporting systems and a focus on individuals rather than systems in preventing errors. Patient advocates and on-site pharmacists were noted as resources to help reduce mistakes by patients and clinicians. Cross-cutting factors included increased complexity of needs for aging patients, and lack of attention to special issues for women and minority physicians and patients.
Margolis, P. and Halfon, N. (2009). "Innovation networks: A strategy to transform primary health care." (AHRQ grant HS16957). Journal of the American Medical Association 302(13), pp. 1461-1462.
Collaborative networks of primary care practices can enhance the capability of individual practices, as well as the capacity of local primary care systems, to improve health outcomes. However, most primary care practices are outside of large health care systems. Therefore, providing policy and practice supports to them poses an enormous challenge. One way to meet this challenge is through user-led innovation networks, suggest the authors. These networks, linked by technology and common purpose, are becoming a widespread means of sharing the work of innovation while improving large complex systems. Operational components of the networks include ongoing dialogue to achieve common understanding and interchange of widely varying perspectives, a platform for experimenting and testing ideas, and a means to execute and implement ideas. One innovation network, called Improving Performance in Practice, currently operates in seven States, and serves as a prototype of a primary health care innovation network.
Olfson, M., Crystal, S., Gerhard, T., and others. (2009). "Mental health treatment received by youths in the year before and after a new diagnosis of bipolar disorder." (AHRQ grant HS16097). Psychiatric Services 60, pp. 1098-1106.
A recent sharp increase in the number of youths diagnosed as having bipolar disorder has focused attention on community practice patterns. Concern exists that this diagnosis may be excessively used by health care professionals who treat children and adolescents, but little is known about the circumstances under which clinicians diagnose young people as having bipolar disorders. The researchers examined service patterns and pharmacy claims surrounding new clinical diagnoses of bipolar disorder among 1,274,726 privately insured youths (17 and younger) during a 1-year period. The rate of new diagnoses was 0.23 percent (2,900 youths). Most of these youths had already been diagnosed as having other mental disorders, usually depressive (46.5 percent) or disruptive (36.7 percent) behavior disorders and they were already being treated with psychotropic medications. In the year after the new diagnosis, more than half had three or fewer additional claims for bipolar disorder. This service pattern suggests that the diagnostic label often does not persist as new symptom patterns emerge or resolve.
Rodriguez, M., Shoultz, J., and Richardson, E. (2009). "Intimate partner violence screening and pregnant Latinas." ( AHRQ grant HS11104). Violence and Victims 24(4), pp. 520-532.
The American College of Obstetrics and Gynecology recommends that pregnant women be routinely screened for intimate partner violence (IPV). IPV has a high prevalence and is associated with adverse health consequences. The researchers decided to explore whether pregnant Latina women were being screened in accordance with the recommendation. In their survey of 210 pregnant Latina women in the Los Angeles area, they found that almost two-thirds had never been asked about being abused. Nearly 83 percent of women who were asked about abuse received care at provider practices that had systems in place to prompt routine screening for abuse. Health care providers who possess good listening skills and are able to explain information in a way women can understand were also likely to ask women about abuse. The researchers suggest that training care providers to ask women about potential abuse may increase clinicians' confidence and boost the number of women whose abuse is detected.
Schiff, G.D., Hasan, O., Kim, S., and others. (2009, November). "Diagnostic error in medicine." (AHRQ grant HS11552). Archives of Internal Medicine 169(20), pp. 1881-1887.
Errors related to delayed or missed diagnoses are a frequent and underappreciated cause of patient injury. Despite the fact that such errors are the leading cause of malpractice litigation, few studies have examined diagnostic errors in detail. The researchers surveyed 310 clinicians from 22 institutions about diagnostic errors which they personally committed or observed. Of the 583 cases examined, 162 (28 percent) were considered major, 241(41 percent) moderate, and 180 (31 percent) minor. The largest proportion of errors (44 percent) took place during laboratory and radiology testing (including test ordering, performance, and clinician processing), followed by clinician assessment (32 percent) (including hypothesis generation, weighing or prioritizing, and recognizing urgency or complications). Pulmonary embolism and drug reactions (including overdose and poisoning) were the two most commonly missed diagnoses (4.5 percent each), followed closely by lung cancer (3.9 percent) and colorectal cancer (3.3 percent).
Sharma, M., Ansari, M.T., Abou-Setta, A.M., and others. (2009). "Systematic review: Comparative effectiveness and harms of combination therapy and monotherapy for dsylipidemia." (AHRQ Contract No. 290-02-0021). Annals of Internal Medicine 151, pp.622-630.
Patients who need an intensive lowering of cholesterol levels may take an increased dose of a statin alone or use a statin in combination with other lipid-lowering agents of another class, such as ezetimibe, niacin, or bile-acid sequestrants. The researchers undertook a comparative effectiveness review to evaluate which of these strategies is superior with respect to clinical outcomes. In their review of 102 studies, they found no evidence of combination therapy being superior to high-dose statin monotherapy in terms of mortality, heart attack, stroke, and revascularization procedures in patients requiring intensive lipid-lowering therapy. There was limited evidence for a greater reduction in LDL cholesterol levels with any combination compared with high-dose statin monotherapy in patients requiring intensive lipid-lowering therapy. However, few studies examined treatment combinations other than statin-ezetimibe.
West, S. L., Blake, C., Liu, Z., and others (2009). "Reflections on the use of electronic health record data for clinical research." (AHRQ Contract No. 290-05-0040). Health Informatics Journal 15(2), pp. 108-121.
To gain insight into the usefulness and limitations of electronic health records (EHRs) for clinical research, the authors of this paper focused on a Web-based EHR system at a major academic medical center. The goal of the clinical research case study was to assess medication patterns in patients with newly diagnosed type 2 diabetes. The researchers discuss the problems in extracting deidentified data from the EHR databases. From 12,424 records of patients with HbA1c measurements (a measure of blood sugar control), 1,664 met their criteria for "newly diagnosed diabetics." The researchers found they needed to use a variety of methods to screen out patients, including text mining of transcribed notes of the patient visits. They found that fragmentation of care at the practice or provider level made it difficult to determine when a patient was newly diagnosed with diabetes and how well they were being treated for this condition.
White, R. H., Sadeghi, B., Tancredi, D. J., and others. (2009, December). "How valid is the ICD-9-CM based AHRQ patient safety indicator for postoperative venous thromboembolism?" (AHRQ Contract No. 290-04-0020). Medical Care 47(12), pp. 1237-1243.
Patient Safety Indicators (PSIs) are a set of indicators providing information on potential in-hospital complications and adverse events following surgeries and procedures. One PSI, PSI-12, focuses on postoperative venous thromboembolism (VTE), either pulmonary embolism (PE) or deep-vein thrombosis (DVT). However, current PSI 12 criteria do not accurately identify patients with acute postoperative lower extremity DVT or PE, found this study. Like all PSIs, it relies on ICD-9-CM diagnostic coding of hospital records. However, the coding process for VTE is particularly prone to inaccuracy. Patients from 80 hospitals were divided into two groups. In the combined group, 451 of 573 VTE flag-positive cases identified using PSI-12 had any acute documented VTE at any time during hospitalization. However, the positive predictive value for acute lower extremity DVT or PE diagnosed after an operation was 44 percent in one group and 48 percent in the second group. Modification of the ICD-9-CM codes and implementation of "present on admission" flags should improve the predictive value for clinically important VTE events.