Baine, W.B. (2003). "Systematic screening of secondary diagnoses in Medicare administrative data to identify candidate risk factors for the principal diagnosis." Annals of Epidemiology 13,
pp. 443-449.
This researcher used the Medicare Provider Analysis and Review Files on hospital stays, which include a patient's principal diagnosis and up to nine secondary diagnoses, to calculate hospital discharge rates for lung cancer, heart attack, and acute renal failure as proxies for incidence rates. He then ranked common secondary diagnoses by the magnitude of their odds difference (overrepresentation in a group with a certain principal diagnosis). He concluded that ranking common secondary diagnoses by the magnitude of their odds difference between groups with disparate discharge rates for a given principal diagnosis may disclose secondary diagnoses that merit evaluation as candidate direct or indirect risk factors. For example, tobacco use and abnormally high blood lipids ranked high in association with the principal diagnosis of heart attack.
Reprints (AHRQ Publication No. 03-R051) are available from the AHRQ Publications and Clearinghouse.
Bazzoli, G.J., Manheim, L.M., and Waters, T.M. (2003, Spring). "U.S. hospital industry restructuring and the hospital safety net." (AHRQ grant HS10040). Inquiry 40, pp. 6-24.
During the 1990s, many hospitals became members of health systems and networks. These investigators examined whether safety net hospitals (SNHs), hospitals with a special commitment to serving the uninsured, were included or excluded from these arrangements and the factors associated with their involvement. They constructed measures for hospital market conditions, management, and mission and examined network and system affiliation patterns between 1994 and 1998. Larger and more technically advanced hospitals joined systems in the 1990s. SNH participation in networks and systems was more common when hospitals faced less market pressure and where only a limited number of unaffiliated hospitals remained. If networks and systems are key parties in negotiating with private payers, SNHs may be going it alone in these negotiations in highly competitive markets.
Bennett, S.J., Oldridge, N.B., Eckert, G.J., and others (2003, July). "Comparison of quality of life measures in heart failure." (AHRQ grant HS09822). Nursing Research 52(4), pp. 207-216.
These researchers compared three instruments for measuring health-related quality of life in (HRQL) in patients with heart failure: the Chronic Heart Failure Questionnaire (CHQ), the Minnesota Living with Heart Failure Questionnaire (LHFQ), and the General Health Survey Short-form-12 (SF-12), in a convenience sample of 211 patients with heart failure (165 patients completed the entire study). Patients completed interviews at baseline and at 4, 8, and 26 weeks after baseline. Overall, patients reported low to moderate HRQL. Reliability of the three instruments was satisfactory. However, the CHQ and LHFQ were more responsive than the SF-12 to clinically important changes over time. The LHFQ and SF-12 were easier and took less time to administer than the CHQ.
Bland, D.R., Dugan, E., Cohen, S.J. (2003, July). "The effects of implementation of the Agency for Healthcare Policy and Research urinary incontinence guidelines in primary care practices." (AHRQ grant HS08716). Journal of the American Geriatrics Society 51, pp. 979-984.
This study found that attempts at increasing screening and management of urinary incontinence (UI) by primary care physicians using a guideline developed in the mid-1990s by the Agency for Healthcare Research and Quality were not successful. The authors conclude that these guidelines may not be the best approach to treating UI in the primary care setting. They randomized 20 of 41 primary care practices in North Carolina to an intervention that included a 3-hour continuing medical education course, training in UI management, patient educational materials, and on-site physician and office support. The remaining 21 practices served as usual care controls. They conducted telephone surveys of UI status and quality of life of 1,145 patients before and after the intervention. Rates of UI assessment and management of existing UI were low in both groups.
Clauser, S.B., and Bierman, A.S. (2003, Spring). "Significance of functional status data for payment and quality." Health Care Financing Review 24(3), pp. 1-12.
To date, the Medicare Program has used functional status information (FSI) in patient assessment tools, performance assessment, payment mechanisms, and most recently in measures of quality of care to inform consumer choice. These researchers explored the rationale for the collection of functional status data to promote innovative models of care. They also examined issues related to data collection for quality improvement, performance measurement, and payment. For example, most Medicare programs including functional assessment have targeted the most seriously disabled patients in need of nursing home care. More research is needed on how to use functional status data and geriatric models of care in integrated systems that target Medicare beneficiaries with less severe functional impairments and those who are at risk for functional decline. There is also much to be learned about functional assessment for children.
Reprints (AHRQ Publication No. 03-R049) are available from the AHRQ Publications and Clearinghouse.
Cohen-Mansfield, J., Libin, A., and Lipson, S. (2003). "Differences in presenting advance directives in the chart, in the minimum data set, and through the staff's perceptions." (AHRQ grant HS09833). Gerontologist 43(3), pp. 302-308.
Decisions concerning end-of-life care may depend on information contained in advance directives that are documented in residents' charts in the nursing home. These investigators examined how advance directives are summarized in patients' records and how physicians perceive the same advance directives and formal orders. They collected data regarding advance directives of 122 elderly nursing home residents from three sources: the Minimum Data Set (MDS), the front cover of the resident's chart, and from inside the chart. They found higher rates of agreement between information inside the chart and on the chart cover than between the MDS and the other two sources. The reasons for these discrepancies may lie in the different functions and procedures pertaining to these source documents.
Glueck, D.H., and Muller, K.E. (2003). "Adjusting power for a baseline covariate in linear models." (AHRQ National Research Service Award training grant T32 HS00058). Statistics in Medicine 22, pp. 2535-2551.
Medical studies often use a random baseline covariate to increase precision and statistical power. Although of no consequence in data analysis, including any random predictors substantially complicates the theory and computation of power. However, failing to account for randomness of predictors may distort power analysis, and many data analysts fail to fully account for this complication in planning a study, assert these authors. In this study, they provide new exact and approximate methods for power analysis of a range of multivariate models with a Gaussian baseline covariate for both small and large samples. The techniques allow rapid calculation and an interactive, graphical approach to sample size choice, which the authors illustrate by calculating power for a clinical trial of a treatment for increasing bone density.
Jennings, R.M., Thompson, L.A., Townsend, C.L., and others (2003, July). "The relationship between pediatric residency program size and inpatient illness severity and diversity." (AHRQ National Research Service Award training grant T32 HS00070). Archives of Pediatric and Adolescent Medicine 157, pp. 676-680.
This study challenges the assumption that better learning opportunities for residents in pediatrics are offered at larger, urban hospitals. The researchers found that pediatric inpatient illness severity and diagnostic diversity varied among hospitals with residency programs. However, these differences were poorly related to the size of the residency programs. The five common diagnoses were almost identical at small, medium, and large programs. On average, smaller programs had similar levels of inpatient illness severity for medical discharges compared with larger programs. In contrast, larger programs had a higher proportion of discharges concentrated in their top five diagnoses, particularly, asthma, thereby decreasing diversity. The findings are based on retrospective analysis of hospital discharges among children in a sample of pediatric residency programs within the University HealthSystems Consortium.
Kaissi, A., Kralewski, J., and Dowd, B. (2003, July). "Financial and organizational factors affecting the employment of nurse practitioners and physician assistants in medical group practices." (National Research Service Award training grant T32 HS00036). Journal of Ambulatory Care Management 26(3), pp. 209-216.
These researchers used data from a survey of 128 medical group practices in Minnesota to examine the financial and organizational factors that are associated with the employment of nurse practitioners (NPs) and physician assistants (PAs) in medical group practices. The findings suggest that the employment of NPs and PAs and their ratios to primary care physicians in practices that employ them are influenced by the organizational characteristics of the group practice but not by the degree of financial risk-sharing for patient care. Large practices, those located in rural areas, not-for-profit practices, and those that scored low on cohesive cultural traits were more likely to employ midlevel practitioners (MLPs). As medical group practices become larger and have more organizational capacity, they can be expected to increase the employment of MLPs and integrate them into their organizations.
Kaplan, R.M. (2003). "The significance of quality of life in health care." (AHRQ grant HS09170). Quality of Life Research 12(Suppl. 1), pp. 3-16.
This author compared a traditional biomedical model with an outcomes model for evaluating health care. The traditional model emphasizes diagnosis and disease-specific outcomes. In contrast, the outcomes model emphasizes life expectancy and health-related quality of life. Although the models are similar, they lead to different conclusions with regard to some interventions. For some conditions, diagnosis and treatment may reduce the impact of a particular disease without extending life expectancy or improving quality of life. Older individuals with multiple medical problems may not benefit from treatments for a particular disease if competing health problems threaten life or reduce quality of life. In some circumstances, successful diagnosis and treatment may actually reduce life expectancy or overall life quality.
Lo, B., and Froman, M. (2003, June). "Oversight of quality improvement: Focusing on benefits and risks." (AHRQ grant HS10961). Archives of Internal Medicine 163, pp. 1481-1486.
Quality improvement (QI) programs may substantially improve patient outcomes while posing little risk to subjects. However, some may pose risks to participants or raise ethical concerns. A monolithic approach to oversight of QI is inappropriate in light of variation in benefits and risks of QI projects and their overlap with research, conclude these authors. They suggest that an explicit protocol for the ethical review of QI would benefit both patients and leaders of QI projects. If a project is considered research rather than QI, review by an institutional review board and informed consent from subjects may be required. When a project poses only minimal risk, stringent oversight beyond what is in place for clinical care is counterproductive. The key ethical issue is not the classification of a project as QI or research, but rather the balance of expected benefits and harms.
Moeller, J.F., Cohen, S.B., Mathiowetz, N.A., and Wun, L-M. (2003). "Regression-based sampling for persons with high health expenditures: Evaluating accuracy and yield with the 1997 MEPS." Medical Care, 41(7 Suppl.), III-44-52.
Given the highly skewed nature of the distribution of health care expenditures in the population, it is critical to capture a sufficient number of individuals with high expenditures in the tail of this distribution to measure national health expenditures accurately. The results presented in this paper validate the accuracy of using model-based sampling to identify households likely to contain working-age individuals with high health expenditures. Estimates made from the logistic 1997 National Medical Expenditure Survey (NMES) selection model prior to drawing the Medical Expenditure Panel Survey Panel 2 sample showed an expected yield of 60.6 percent for NMES households selected to be over-sampled. This means that 60.6 percent of NMES households selected for over-sampling by the model were expected to in fact contain at least one person aged 18 to 64 with health expenditures in the top 15 percent of the distribution in 1997. This over-sampling approach increased the actual number of high expenditure individuals in the sample by 20 percent compared to the expected number cases had there been no over-sampling.
Reprints (AHRQ Publication No. 03-R059) are available from AHRQ Publications and Clearinghouse.
AHRQ Publication No. 04-0007
Current as of October 2003
Internet Citation:
Research Activities newsletter. October 2003, No. 278. AHRQ Publication No. 04-0007. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/research/oct03/
540 Gaither Road Rockville, MD 20850