primary-care physicians; only 9% received care in a specialized diabetes clinic. Despite frequent primary- care visits, most diabetic patients in this county health-.
SHORT REPORTS
Preventive Care in Diabetes Mellitus Current Practice in Urban Health-Care System
Early identification and treatment of complications of diabetes mellitus may reduce the severity of the complications. As part of a program to reduce these complications in the Denver Department of Health and Hospitals patient population, our study determined how frequently preventive care, e.g., fundoscopic examinations, referral to an ophthalmologist, foot examinations, and assessment of cardiovascular risk factors, was provided to diabetic patients. With the use of billing records to identify a large sample of diabetic patients, a chart review of 544 patients was conducted. During the study year, the mean ± SE number of visits to primary-care clinics was 5.7 ± 0.22, with 86.4% having at least one visit. Most diabetic patients were seen by primary-care physicians; only 9% received care in a specialized diabetes clinic. Despite frequent primarycare visits, most diabetic patients in this county healthcare system did not have documentation of care to detect complications of diabetes mellitus, and referral services for detection and treatment of these complications were infrequently used. Moreover, among patients seen on >10 occasions in a primary-care setting, preventive care was not provided to 30% of the patients. Preventive care does not appear to be a regular part of a primary-care visit for most of the diabetic patients in this study. Diabetes Care 12:745-47, 1989
E
arly identification of complications of diabetes mellitus, such as diabetic ophthalmopathy and lower-extremity ulcerations, may reduce the severity of these complications. Because diabetic patients are at higher risk for cardiovascular and renal disease, reducing reversible cardiovascular risk factors, including treatment of hypertension, is also important. For these reasons, primary-care practitioners have been encouraged to ensure that their diabetic patients receive regular ophthalmological examinations, foot care, and have their cardiovascular risk factors controlled (1,2). Moreover, diabetic patients with reduced access to care of this type are thought to be at higher risk for the development of ophthalmological, lower-extremity, and cardiovascular complications (3). As part of a combined state and county program to reduce the incidence of diabetic complications, we determined how frequently such preventive care was given to diabetic patients receiving treatment in the Denver Department of Health and Hospitals (DDHH). We found that despite frequent visits to community and hospital
DIABETES CARE, V O L . 12, N O . 10, NOVEMBER/DECEMBER
1989
Thomas H. Payne, MD Barbara A. Gabella, BA Sharon L. Michael, RN, MS Walter F. Young, MA Jeffrey Pickard, MD Fred D. Hofeldt, MD Frances Fan, MD Jerome S. Stromberg, PhD Richard F. Hamman, MD, DrPH
continuity-care clinics, most patients did not receive recommended preventive care during the study year.
RESEARCH DESIGN AND METHODS
The DDHH system includes eight neighborhood health clinics, medical and surgical clinics, a walk-in clinic, emergency room, and Denver General Hospital inpatient medical and surgical services. There are -690,000 outpatient visits and 78,000 hospitalizations in the system annually. Interviews with diabetic patients identified in this study indicate that of those receiving care in the DDHH system, 5% received additional care outside the system. The actual number of patients receiving care outside the system may be higher Magnetic tapes from the DDHH billing system for all 1984 outpatient visits and hospitalizations were searched for the presence of ICD 9 codes for diabetes mellitus and its complications (250.0-250.9). Patients whose billing records included such a diagnosis code for any clinic or hospital visit at any time during the study year were identified. Because no more than two ICD 9 codes could appear in each visit or hospitalization on billing tapes, and because not all diabetic patients were seen during the study year, some diabetic patients receiving care in the DDHH were not identified by this method. The medical records of a randomly selected 50% sample of these patients were abstracted by trained reviewers to determine demographic and clinical characteristics, the number of foot examinations, measurements of visual acuity, funduscopic examinations, ophthalmology clinic visits, podiatry clinic visits, and total outpatient visits and hospitalizations during the previous year. An interrater reliability study revealed chart review reliability to be >95% in all variables, except family income. Prevalence of renal, ophthalmological, and lower-extremity complications and hypertension were determined by chart review.
RESULTS A search of DDHH billing tapes for 1984 revealed 1196 patients whose records included an ICD 9 code for diabetes mellitus on at least one occasion. A randomly selected sample of 602 of these patients was selected
745
SHORT REPORTS
for manual review of the patient record. Seventeen (2.8%) of the records were unavailable for review and 41 (6.8%) were found not to have diabetes mellitus. The remaining 544 patients had 6153 outpatient visits and 293 hospitalizations during 1984. The mean age of the 544 patients was 55 yr; 64.5% were female. Hispanics comprised 39.6% of the sample, and Blacks and Whites represented 34.4 and 23.4%, respectively. The median family income was $4685/yr (based on chart information); 33% of the sample was covered by Medicaid, 33% by Medicare, and 38% had no insurance coverage. Of the 544 study patients, 425 (78.1%) were characterized by their physician as having type II (non-insulin-dependent) diabetes, 62 (11.4%) with type I (insulin-dependent) diabetes, 1 (0.2%) with gestational diabetes, and in 56 (10.3%), no clinical type was recorded. Over half (53.7%) of the patients were treated with insulin, 30.7% were treated with oral hypoglycemic agents, 2.4% were treated with both insulin and an oral agent, and 13.2% were treated with diet alone. Services used by these patients are presented in Table 1. The mean ± SE number of primary-care visits during the study year was 5.7 ± 0.22. During the study year, the mean number of hospitalizations was 0.5 ± 0.4; the mean number of emergency room and walk-in clinic visits was low (1.0 ± 0.08). The diagnosis of hypertension was noted in 64% of patients. Ophthalmological complications (retinopathy, cataract, glaucoma) occurred in 34.4%, renal disease (proteinuria, diagnosis of nephropathy, BUN >40 mg/dl, or serum creatinine >2 mg/dl) in 30.1%, and lower-extremity complications (peripheral vascular disease, lower-extremity lesions on examination, or amputation) in 16% of patients. During the study year, 13.1% of the 544 study patients were referred to the ophthalmology clinic and 24.4% were seen (11.3% had been referred earlier). Only 29.8% of patients had a record of their visual acu-
ity in their medical record in the study year, and only 46.6% had a recorded funduscopic examination. If the 24.4% of patients seen in the ophthalmology clinic are excluded, the number of patients with recorded visual acuity drops to 7.4%, and the number with a recorded fundoscopic examination decreases to 34.1%. Fifty-five percent of the patients had a recorded foot examination and 2.6% were referred to the podiatry clinic. When the patients seen in the podiatry clinic (Table 1) are excluded, the number of patients with a recorded foot examination decreases to 48.8%. In 45.2% of patients the chart reflected tnat a smoking history had been obtained. A serum cholesterol value had been obtained in the preceding year in 67.8% of patients. Few patients were seen in the diabetes clinic, but patients seen on at least one occasion during the study year were more likely to have had visual acuity checked (49 vs. 27.7%, P = .003), had a fundoscopic examination (87.8 vs. 46%, P < .0C1), been seen in the ophthalmology clinic (46.9 vs. 22.2%, P < .01), and have had their feet examined (71.4 vs. 54.1%, P = .02) than patients not seen there. The percentage of patients with foot examinations, visual acuity measurements, fundoscopic examinations, and referrals to ophthalmologist increased significantly as the number of primary-care visits increased (KruskalWaHis exact test, P < .05 for each); however, the rate of increase was low. For example, among patients with five visits to a primary-care clinic during the study year, 60% had foot examinations, 30% had measurements of visual acuity, 44% had fundoscopic examinations, and 26% were referred to an ophthalmologist; for patients seen on >10 occasions in primary-care clinics, the percentages were 64.6, 38.8, 34.5, and 33.3%, respectively. Only 35.1% of patients had more than five primary-care visits during the study year.
TABLE 1 Services used by diabetic patients in Denver Department of Health and Hospitals system in 1984
espite frequent visits to primary-care clinics (mean 5.7 visits), most diabetic patients in this county health-care system did not have a record of care designed to detect complications of diabetes, and referral services for detection and treatment of these complications were infrequently used. Preventive care did not appear to be a regular part of a primary-care visit for most of the diabetic patients in this study. Bailey et al. (4) found in a study of 65 patients in a diabetes clinic that only 49.2% had their feet examined in the previous year. Deeb et al. (5) found similar percentages of patients who had retinal examinations, ophthalmology referral, and lower-extremity examinations, and noted a high clinic dropout rate. Our study shows that even diabetic patients who have frequent primarycare visits usually lack preventive care that could be provided by primary-care practitioners. Why do so few diabetic patients have documentation
General medical clinics* Inpatient services Emergency room Walk-in clinic Ophthalmology clinic Podiatry clinic Nutritionist Diabetes clinic Diabetes education classes Vascular surgery clinic
n
%
470 180 158 143 133 79 53 49 42 20
86.4 33.1 29.0 26.3 24.4 14.5 9.7 9.0 7.7 3.7
Percentage of patients (n = 544) that used each service at least once in 1984. includes 8 neighborhood health clinics and hospital internal medicine clinic.
746
DISCUSSION
D
DIABETES CARE, VOL. 12, NO. 10, NOVEMBER/DECEMBER 1989
SHORT REPORTS
of preventive care? Cost is probably not the major problem. Although these patients are poor, they are seen frequently, and most preventive care can be provided by primary-care practitioners. Availability of timely ophthalmology referral services was found to be a problem. It is possible that preventive care was given more frequently than the charts reflected, but it is more likely that it was overlooked. There was no uniform method of documenting such care that would allow another provider to rapidly ascertain the status of preventive care in a particular patient. Because more patients in this study were seen in primary-care clinics than any other site, these clinics are the logical setting for the identification of diabetic patients in need of preventive care. However, this study suggests that simply increasing the frequency of visits is not an efficient method of increasing preventive care. Instead, programs to increase practitioner awareness of appropriate preventive care, chart or computer-based reminders, and improved accessibility of referral care are needed (6).
Creatinine 88.4 p,M = 0.011 mg/dl
Urea nitrogen 0.357 mM = 2.8 mg/dl
From the Denver Health and Hospitals Diabetes Study Group; the Denver Department of Health and Hospitals; the Diabetes Control Program of the Colorado Department of Health; and the Departments of Medicine, Preventive Medicine, and Biometrics, University of Colorado School of Medicine, Denver, Colorado. Address correspondence and reprint requests to Thomas H. Payne, MD, Center for Health Studies, Group Health Cooperative of Puget Sound, 521 Wall Street, Seattle, WA 98121.
DIABETES CARE, VOL. 12, NO. 10, NOVEMBER/DECEMBER 1989
ACKNOWLEDGMENTS
This study was supported in part by the Division of Diabetes Control, Centers for Disease Control under cooperative agreement U-32-CCU-800342-05. Portions of this paper were presented in abstract form at the 5th annual Epidemiological Research Exchange of the University of Colorado School of Medicine, 24 October 1985. The views expressed herein are those of the authors and do not necessarily represent the position of the Denver Department of Health and Hospitals. REFERENCES
1. Browner WS: Preventable complications of diabetes mellitus. West I Med 145:701-703, 1986 2. National Diabetes Advisory Board: The Prevention and Treatment of Five Complications of Diabetes: A Guide for Primary Care Practitioners. Washington, DC, U.S. Govt. Printing Office, 1983 (Dept. of Health and Human Services publ. no. 83-8392) 3. Herman WH, Teutsch SM, Geiss LS: Closing the gap: the problem of diabetes mellitus in the United States. Diabetes Care 8:391-406, 1985 4. Bailey TS, Yu HM, Rayfield EJ: Patterns of foot examination in a diabetes clinic. Am / Med 78:371-74, 1985 5. Deeb LC, Pettijohn FP, Shirah JK, Freeman GF: Interventions among primary-care practitioners to improve care for preventable complications of diabetes. Diabetes Care 3:275-80, 1988 6. Barnett GO, Winickoff RN, Morgan MM, Zielstorff RD: A computer-based monitoring system for follow-up of elevated blood pressure. Med Care 21:400-409, 1983
747
