Deadline for receiving credit for this article:
October 31, 2006 |
JHQ 153 - Diabetes Preventive Care and Non-traumatic Lower Extremity Amputation Rates
Michael E. Moreland, MSW; Amy M. Kilbourne, PhD MPH; Joseph B. Engelhardt, PhD; Rajiv Jain MD; Jian Gao, PhD; David S. Macpherson, MD MPH; Ali F. Sonel, MD FACC FACP; Guibo Xing, PhD
Keywords:
Diabetes mellitus, Health services research, Prevention, Quality of healthcare
September/October 2004
| Clinical performance monitoring data on processes of care from a 3-year period were used to assess whether preventive foot care was associated with improved health outcomes in diabetes mellitus patients. Preventive foot care as well as sensory and pedal-pulse examinations were associated with reduced rates of lower extremity amputation. It is believed that an administrative focus, resource direction, and improvement in process monitoring will lead to better patient outcomes. External review measures can be used by administrators and clinicians to determine trends in quality of care and patient outcomes and to provide feedback on prevention efforts. |
In 1995 the primary care model of the Veterans Health Administration (VHA) was changed from episodic, inpatient subspecialty-oriented care delivery to comprehensive outpatient primary care (Kizer, 1999). The new VHA primary care model emphasizes performance improvement in delivering preventive services (Rubenstein et al., 1996). The model has increased local control through regional networks known as Veterans Integrated Service Networks (VISNs). At the national level, the VHA designated 22 VISNs, each of which comprises 5–10 individual VHA medical centers (VAMCs).
Clinical performance measures, especially clinical process measures and actions to demonstrate improvement, have received a significant emphasis as well as financial resources and administrative attention. One focus of prevention activity is care improvement for patients with chronic diseases, notably diabetes mellitus.
The prevalence of diabetes mellitus among veterans served by the VHA is higher than the prevalence in the general population (Pogach et al., 1998). The rate of non-traumatic lower extremity amputation (LEA), a serious and costly complication of diabetes (Mayfield, Reiber, Nelson, & Greene, 2000), is higher in persons with diabetes compared with the general population because (a) diabetic neuropathy impairs the ability of persons with diabetes to sense minor skin trauma that can lead to skin ulcers (Armstrong & Lavery, 1998) and (b) micro- and macrovascular disease in persons with diabetes leads to poor blood flow and impaired skin-ulcer healing (Muggeo, 1998). As a result, persons with diabetes are more likely to develop nonhealing ulcers that progress to deep-tissue or bone infection and eventually the need for amputation (Muggeo). Persons with diabetes who have evidence of peripheral vascular disease, as revealed by diminished pedal pulse or inability to sense a standardized filament, are at particularly high risk for skin injury and ulcers. Hence, the care provider should routinely examine the patient’s feet to detect sensory loss or diminished pedal pulse. The examination should be followed by continuing patient education in foot care to prevent initial skin injury.
The VHA instituted a national quality improvement initiative to address a potential gap in quality of care for persons with diabetes and other chronic conditions, using the external peer review program (EPRP), an ongoing quality improvement initiative that collects national data on key healthcare process measures each fiscal year. Since 1996, EPRP participants have reviewed hundreds of thousands of charts to determine quality of medical care and provide feedback to clinicians and administrators. An EPRP was formed to collect data on several performance measures derived from evidence-based practice guidelines (Doebbeling et al., 2002). These data were used in benchmarking VHA facilities that had performed consistently better or worse than expected. Measures pertaining to diabetes care included foot care (i.e., the proportion of persons with diabetes for whom there were recorded results of a foot sensory examination with a standardized monofilament or a pedal-pulse examination).
Recent evidence suggests that the greatest practice variation in quality of care measures occurs at the VAMC level rather than the provider level (Krein, Hofer, Kerr, & Hayward, 2002). This finding mirrors that of an Institute of Medicine report (Committee on Quality of Health Care in America, 2001) concluding that system-level changes are needed to improve the overall quality of healthcare. Hence, profiling the quality of care at the VHA network and VAMC levels rather than at the individual-provider level may be a useful and relatively unobtrusive method of monitoring continuous quality improvement.
In light of the prevalence of diabetes among VHA patients and the VHA’s emphasis on diabetes prevention, this study attempted to determine whether improved diabetes care is associated with reduced LEA rates. The study was reviewed and approved by local VHA institutional review boards in Pittsburgh, PA, and Albany, NY.
Methods
Data Sources
Data on patients with diabetes and LEAs were obtained from the VHA National Patient Care Database, which includes all visits (both inpatient and outpatient) and diagnoses. Patients with diabetes mellitus were identified using International Classification of Diseases, 9th ed. (ICD-9) codes 250–250.92 inclusive. The number of VHA patients with diabetes mellitus was 535,220 in fiscal year (FY) 1999, 621,320 in FY 2000, and 743,120 in FY 2001. Patients undergoing a non-traumatic LEA attributed to diabetes were identified using surgical ICD-9 codes 8410, 8411, 8412, 8414, 8415, 8417, 8419, and 8491.
EPRP reviewers outside the VHA collected data on preventive care from a random sample of medical chart reviews. Patients included in the chart review had made at least two visits during the 3-year period 1999–2001 to a physician, physician assistant, or nurse practitioner at an outpatient clinic (i.e., a primary care, general medicine, cardiology, endocrinology/ metabolism, diabetes, hypertension, pulmonary/chest, or women’s clinic). For the purpose of this review, performance data on diabetic foot care were collected in aggregate from the VHA Web site (http://klfmenu.med.va. gov/pm/eprp.htm). Approximately 200,000 charts from all VHA medical facilities were reviewed for each year (1999–2001). EPRP calculated the number of patients who should have received a pedal-pulse or foot sensory examination and the number who actually did receive each intervention. The rates for each facility were posted on the Web site.
Study Period
The study period (i.e., FY 1999–FY 2001) was selected because it began several years after implementation of the VHA quality improvement initiatives and because complete data on preventive care performance are available for these fiscal years.
Units of Observation
The key outcome was the rate of non-traumatic LEA. LEA rates were analyzed at the two operational levels of the VHA system (excluding the VHA headquarters in Washington, DC): the 22 VISN regions and 135 VAMCs. Therefore, the units of observation are grouped data, not individual patients.
The independent preventive care variables (i.e., foot sensory and pedal-pulse examination rates at each VAMC) were determined based on the EPRP external audits for FY 1999 and FY 2000. The chart review question for pedal-pulse examination was, “Within the past year, does the record document pulses were checked in the patient’s feet?” Acceptable forms of chart documentation included a reference to palpable pulses, or capillary filling, or a body outline marked with pulse points. The chart review question for foot sensory examination was, “Within the past year, does the record document examination of sensation in the patient’s feet?” Acceptable documentation included whether sensation, Babinski neuro checks, or documentation of “intact to touch” sensation in feet was tested. Documentation of a visual foot examination or referral to a podiatrist were not accepted as positive responses.
For 135 VAMCs, data were available on all study variables (both preventive care and LEA outcome). Not all VAMCs perform LEAs, so an adjustment was made to allocate LEA patients among all VAMCs on the basis of proportional use of primary care. Proportional assignment of patients among all VAMCs minimized overestimation of LEA rates in the VAMCs where LEAs were performed.
Statistical Analyses
Descriptive statistics were used to describe LEA rates and number of LEAs. The relationship between preventive care (i.e., foot sensory and pedal-pulse examinations) and LEA rates was assessed using linear regression analyses. Regression models were run to determine the association between preventive care rates in FY 2000 and LEA rates in FY 2001 as well as preventive care rates in FY 1999 and LEA rates in FY 2000. The unit of analysis for the regression model was the VAMC. For the regression results, LEA rates were transformed using the natural log to approximate normal distribution. The criterion for statistical significance for analyses was p < .05.
Results
The nationwide LEA rate within the VHA declined 38% during the study period, from 6.59 per thousand in FY 1999 to 4.79 per thousand in FY 2001. For the entire study period, there was a decline in every VISN. In 18 of the 22 VISNs, a decline occurred every year. A similar pattern of decline in LEA rates was evident across all VISNs (seeFigure 1). There were no statistically significant differences in LEA rates among any of the VISNs for any year of the study.
The number of patients treated in the VHA increased 11.8% from FY 1999 to FY 2001 (from 1,464,911 unique patients in FY 1999 to 1,638,306 unique patients in FY 2001). The number of VHA patients with diabetes mellitus increased 14% from FY 1999 (535,220) to FY 2000 (621,320) and 16% from FY 2000 to FY 2001 (743,120). The number of LEAs essentially remained constant, despite the large increase in the number of diabetic patients. Because the number of LEAs remained constant and the number of persons with diabetes increased during the study period, the rate of LEAs declined.
In the context of these declining rates, the purpose of this study was to determine whether preventive foot care processes were associated with fewer LEAs. A negative relationship at the VAMC level between higher performance on diabetic foot care measures and lower LEA rates would suggest that primary-care foot examinations may prevent amputations. It was reasoned that the beneficial effect of foot examinations would not be immediate and therefore would be correlated with reduced LEA rates in later years. The regression results demonstrated the expected negative relationship, in which a VAMC’s higher preventive care rates in one year were associated with lower LEA rates in the following year (see Table 1).
Statistically significant associations were found between higher rates of foot sensory and pedal-pulse examinations in FY 2000 and lower LEA rates in FY 2001, as well as higher rates of FY 1999 pulse-pedal examinations and lower FY 2000 LEA rates. FY 1999 foot sensory examinations displayed a nonsignificant trend toward lower FY 2000 LEA rates.
Discussion
The results of this study indicate that in the VHA system the number of persons with diabetes is increasing and rate of LEAs in persons with diabetes is declining. The LEA rates are consistent with earlier findings (Collins et al., 2001; Mayfield, Reiber, Maynard et al., 2000) and are within the normative values VAMC-level practitioners and administrators might expect. The rate reduction is relatively uniform. It could be attributed to the implementation of the VHA national quality improvement initiative. The results lend partial support to the idea that preventive foot care may contribute to lower LEA rates. They are consistent with earlier studies linking diabetes foot care to lower LEA rates (Rith-Najarian et al., 1998). Although in FY 2000 there was only a nonsignificant trend toward lower LEA rates in VAMCs that had provided more foot sensory examinations in FY 1999, in FY 2000 there were statistically significant lower LEA rates in VAMCs that had provided more pedal-pulse examinations in FY 1999. VAMCs that in FY 2000 provided more pedal-pulse and foot sensory examinations displayed in FY 2001 a statistically significant reduction in LEAs.
The postulated mechanism through which foot examinations prevent LEA is improved foot care. When an abnormal foot examination result is discovered during an office visit, the healthcare provider is likely to emphasize simple precautions through which the patient can minimize the chance of foot injury (e.g., avoidance of barefoot walking, daily foot examination). Education may continue when subsequent annual foot examinations are performed. Most VHA facilities also offer a more intense foot evaluation and education program, to which at-risk patients may be referred.
Patient self-care of feet may improve after discovery that a foot is at risk. Self-care can prevent or lead to early treatment of ulcer formation, subsequent deeper infection, and the need for amputation. Future investigation could assemble data to support this mechanism of improved outcome.
These findings could have a significant impact on the quality of healthcare for persons with diabetes. Earlier studies suggested that diabetes complications are a major contributor to the need for LEA (Armstrong & Lavery, 1998). In the VHA system between 1989 and 1998, 63% of LEAs were attributed to diabetes complications (Mayfield, Reiber, Maynard, et al., 2000). To our knowledge, this is one of the first studies to demonstrate a link between diabetes performance measures and LEA rates in a large, nationally representative sample. Mayfield, Reiber, Maynard, et al. (2000) reported a decline in VHA LEA rates from 1989 to 1998, in a period before implementation of the VHA quality improvement initiative, but they found that LEA rates specific to diabetes complications did not change. Another regional study (Mayfield, Reiber, Nelson, & Greene, 2000) did not find a link between foot examinations and LEA rates, in part because of a limited sample size.
This study had some limitations, especially with regard to linking preventive care and LEA rates. The study did not control for other structural factors that could explain reduced LEA rates, such as academic affiliations or surgical interventions. Because data were only available in aggregate, we could not control for individual patient factors (e.g., age). Additional research is warranted into other possible patient, provider, and system-level causes for the VHA-wide reduction in the LEA rate. In addition, factors related to the development of diabetic neuropathy or vascular disease, such as glycemic control, lipid control, and blood pressure control, were not measured as part of this investigation. However, it is likely that improvements in these clinical factors would lead to a reduction in morbid events only after several years—not the rapid decline in LEA rates seen in this investigation. Finally, the data were examined at the group rather than individual level, and therefore are vulnerable to ecologic fallacy (Epidemiologic Research and Information Center, 2000).
This study demonstrates the way in which external review of care measures and outcomes data can be applied to determine trends in patients’ quality of care and outcomes and provide feedback about prevention efforts. The application of process and outcome measures to determine trends in diabetes quality of care may be generalizable to practices outside of the VHA. Although our findings as to the link between preventive care and LEA rates are not definitive, they tend to confirm the effectiveness of the VHA’s effort to provide preventive foot care and systemwide monitoring of such care.
The implication for healthcare administrators, clinicians, and healthcare quality professionals—whether or not they are part of the VHA system—is that the implementation of clinical process measures can have a significant impact on patient health outcomes. The VHA implementation of these process measures is intended to improve veterans’ health outcomes, and this study has found evidence that such clinical quality efforts do have a significant impact. Profiling the quality of care at the VAMC level rather than the provider level may be a feasible and less obtrusive method for monitoring and benchmarking continuous quality improvement in the VHA. An additional implication of this study is that documenting and researching the efficacy of these process measures is required to further support clinical quality efforts.
Take a test on the article you just read for continuing education credit!
Author's Biography
Michael E. Moreland is the director of the VA Pittsburgh Healthcare System, Pittsburgh, PA.
Amy M. Kilbourne is an assistant professor of medicine and psychiatry at the Center for Health Equity Research and Promotion, VA Pittsburgh Healthcare System and University of Pittsburgh School of Medicine.
Joseph B. Engelhardt is a health services researcher at the VA Healthcare Network, Upstate New York.
Rajiv Jain is chief of staff at the VA Pittsburgh Healthcare System and professor of medicine at the University of Pittsburgh.
Jian Gao is a health economist at the VA Healthcare Network, Upstate New York.
David S. Macpherson is vice president of the VA Pittsburgh Healthcare System Primary Care Service Line and professor of medicine, University of Pittsburgh.
Ali F. Sonel is chair of the Institutional Review Board Committee, VA Pittsburgh Healthcare System and assistant professor of medicine, University of Pittsburgh.
Guibo Xing is a health services researcher at the VA Healthcare Network, Upstate New York; University at Albany; and State University of New York.
For more information on this article, contact Amy M. Kilbourne by phone at 412/688-6000 ext 5977 or by e-mail at amy.kil bourne@med.va.gov.
Acknowledgements
This research was supported by the Veterans Administration Pittsburgh Health Care System and the Office of Research and Development, Health Services Research and Development Service, Department of Veterans Affairs. Dr. Kilbourne is funded by a Career Development Award from the Veterans Administration Health Services Research and Development program.
References
Armstrong, D.G., & Lavery, L.A. (1998). Diabetic foot ulcers: Prevention, diagnosis and classification. American Family Physician, 57, 1325–1332.
Collins, T.C., Johnson, M., Daley, J., Henderson, W.G., Khuri, S.F., & Gordon, H.S. (2001). Preoperative risk factors for 30-day mortality after elective surgery for vascular disease in Department of Veterans Affairs hospitals: Is race important? Journal of Vascular Surgery, 34, 634–640.
Committee on Quality of Health Care in America, Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academies Press.
Doebbeling, B.N., Vaughn, T.E., Woolson, R.F., Peloso, P.M., Ward, M.M., Letuchy, E., Boots-Miller, B.J., Tripp-Reimer, T., & Branch, L.G. (2002). Benchmarking Veterans Affairs Medical Centers in the delivery of preventive health services: Comparison of methods. Medical Care, 40, 540–554.
Epidemiologic Research and Information Center. (2000). Ecologic studies (ERIC Notebook No. 12). Durham, NC: Author.
Kizer, K.W. (1999). The “New VA”: A national laboratory for health care quality management. American Journal of Medical Quality, 14(1), 3–20.
Krein, S.A., Hofer, T.P., Kerr, E.A., & Hayward, R.A. (2002). Whom should we profile? Examining diabetes care practice variation among primary care providers, provider groups, and health care facilities. Health Services Research, 37, 1159–1180.
Mayfield, J.A., Reiber, G.E., Maynard, C., Czerniecki, J.M., Caps, M.T., & Sangeorzan, B.J. (2000). Trends in lower limb amputation in the Veterans Health Administration, 1989–1998. Journal of Rehabilitation Research and Development, 37(1), 23–30.
Mayfield, J.A., Reiber, G.E., Nelson, R.G., & Greene, T. (2000). Do foot examinations reduce the risk of diabetic amputation? Journal of Family Practice, 49, 499–504.
Muggeo, M. (1998). Accelerated complications in Type 2 diabetes mellitus: The need for greater awareness and earlier detection. Diabetic Medicine, 15(Suppl. 4), 60–62.
Pogach, L.M., Hawley, G., Weinstock, R., Sawin, C., Schiebe, H., Cutler, F., Zieve, F., Bates, M., & Repke, D. (1998). Diabetes prevalence and hospital and pharmacy use in the Veterans Health Administration (1994): Use of an ambulatory care pharmacy-derived database. Diabetes Care, 21, 368–373.
Rith-Najarian, S., Branchaud, C., Beaulieu, O., Gohdes, D., Simonson, G., & Mazze, R. (1998). Reducing lower-extremity amputations due to diabetes: Application of the staged diabetes management approach in a primary care setting. Journal of Family Practice, 47(2), 127–132.
Rubenstein, L.V., Yano, E.M., Fink, A., Lanto, A.B., Simon, B., Graham, M., & Robbins, A.S. (1996). Evaluation of the VA’s pilot program in institutional reorganization toward primary and ambulatory care: Part I—Changes in process and outcomes of care. Academic Medicine, 71, 772–783.
Objectives
Journal for Healthcare Quality is pleased to offer the opportunity to earn continuing education (CE) credit to those who read this article, take the posttest online and complete the hard copy or online form. This continuing education offering, JHQ153, will provide one contact hour to those who complete it appropriately.
Core CPHQ Examination Content Areas
II. Information Management
|
