Methods

Top Methods Results Discussion Author & Article Info References

Information about medical care of diabetic persons in the U.S. population is derived from the 1989 National Health Interview Survey (NHIS) [11, 12]. The NHIS is an annual personal interview survey done in the households of a nationwide sample of the civilian, noninstitutionalized U.S. population. The survey consists of questions about various sociodemographic and health-related factors asked of all household members. In 1989, the eligible sample of persons aged 18 years or older was 84 572, and the response rate was 96%. All persons who were known to have diabetes were identified by self-report or by a knowledgeable household respondent (n = 2829). In several studies, diabetes has been shown to be reported accurately by subjects in an interview setting [13-17]. The diabetic individuals were eligible for a special diabetes questionnaire. Nonresponse to the diabetes questionnaire was 4.6%, and 10.4% of the subjects indicated that they did not have diabetes or had only potential, borderline, or gestational diabetes. The final sample size was 2405 persons aged 18 years or older who were identified and self-confirmed as having been diagnosed by a physician as having diabetes.

Questions about duration of diabetes, duration of insulin use, height, and weight permitted differentiation of diabetic subjects into insulin-dependent diabetes mellitus (IDDM) and non–insulin-dependent diabetes mellitus (NIDDM) [12]. Persons were classified as having IDDM if they were younger than 30 years when they were diagnosed as having diabetes, had been taking insulin since the time of diagnosis, and were less than 120% of desirable weight (body mass index less than 27 kg/m² for men and less than 25 kg/m² for women, based on self-reported height and weight; n = 124). Thirteen subjects were missing data on these variables and could not be classified by diabetes type. The remaining 2268 persons were considered to have NIDDM, and these were differentiated into those who did (n = 922) and did not (n = 1346) use insulin. Some insulin-treated NIDDM patients with a diagnosis of diabetes at age 30 years or older may have IDDM but are estimated to comprise only about 7% of all adults with diagnosed diabetes [18].

All analyses were done using the computer program SAS (SAS Institute Inc., Cary, North Carolina). Because the NHIS population was selected by stratified sampling, the data were weighted by age, sex, race, income, and geographic location to yield estimates representative of the U.S. population [11]. Standard errors of means and proportions were estimated using the Taylor series linearization method [19] and were calculated by the SESUDAAN computer program developed for complex sample surveys [20]. Two-tailed, large-sample z tests were used to test for significant differences in means and proportions.

Results

Top Methods Results Discussion Author & Article Info References

Demographic and medical care characteristics of the diabetic subjects are shown in Table 1. About 10% of diabetic adults indicated that they did not have a physician for regular care of their diabetes; 32% of subjects made fewer than four visits to their physician each year, 33% made four to six visits per year, and 26% made more than six visits per year (Figure 1). There were no significant differences in these rates by age throughout the adult age range. The proportions were similar for persons with NIDDM treated with insulin and not treated with insulin, but adults with IDDM made fewer visits per year to their regular diabetes physician, with only 36% reporting four or more visits. The lower frequency of medical care for subjects with IDDM is likely caused by the much younger age of the IDDM patients and the accompanying lower frequency of comorbid conditions.

View larger version (62K): [in this window] [in a new window]   Figure 1. Frequency of visits to a regular physician by adults with insulin-dependent (IDDM) and non–insulin-dependent diabetes mellitus (NIDDM) for diabetes care. Sample sizes: patients with IDDM equals 124; patients with insulin-treated NIDDM equals 922; patients with NIDDM not treated with insulin equals 1346.

Diabetic persons in the NHIS were asked about visits to certain physician specialists: 23% had seen a cardiologist in the past year, and 15% of women had seen a gynecologist (Table 1). The proportion of subjects who had seen an ophthalmologist in the past year (45%) was slightly less than the proportion who had had a dilated eye examination in the past year (49%). About 8% to 23% had been treated by a podiatrist in the past year, and 24% to 39% reported that their feet had been checked by a health professional at least twice in the past 6 months (Table 1). Of at least equal importance is the high frequency with which diabetic patients inspect their own feet: 82% of insulin-treated subjects and 65% of those not treated with insulin reported that they check their feet at least once per week.

Diabetic adults were hospitalized substantially more frequently than nondiabetic persons (Figure 2). Among those with insulin-treated NIDDM, the proportion hospitalized in the past year ranged from 6 times that of adults without diabetes at age 18 to 44 years to 2.3 times at age 75 or older (P < 0.001). The frequency of hospitalization for persons with NIDDM not treated with insulin was also considerably higher than for nondiabetic adults. Multiple hospitalizations in the past year were reported by 8.3% of diabetic adults and 1.3% of nondiabetic adults. Mean length of stay per hospitalization was 8.3 days for diabetic patients and 6.5 days for nondiabetic adults.

View larger version (41K): [in this window] [in a new window]   Figure 2. Proportion of adults with non–insulin-dependent diabetes mellitus (NIDDM) and without diabetes who were hospitalized in the past year (excluding hospitalizations for childbirth). Sample sizes: patients with insulin-treated NIDDM equals 922; patients with NIDDM not treated by insulin equals 1346; nondiabetic patients equals 20 181.

Among all diabetic subjects who were 18 years or older, 43% were treated with insulin, 49% were treated with oral agents, and 64% reported that they were following prescribed diet therapy. Ten percent of the insulin-treated subjects with NIDDM were using oral agents in addition to insulin. Two or more insulin injections daily were taken by 61% of IDDM and 48% of insulin-treated NIDDM patients; insulin pump use was rare for both insulin-treated groups. The proportion of subjects with NIDDM who were treated with insulin was relatively constant across the adult age range but increased with longer duration of diabetes, from 22% at 0 to 4 years duration to 58% at 20 or more years (Figure 3). Concomitantly, the proportion treated with oral agents declined from 64% at 0 to 4 years duration to 37% at 20 or more years. The proportion following a diabetes diet all or most of the time was relatively constant across the range of diabetes duration. Although only 64% of diabetic individuals stated they were following their diet, 90% answered "yes" to a question about whether they believed that diet is important in diabetes control.

View larger version (14K): [in this window] [in a new window]   Figure 3. Percentage of adults with non–insulin-dependent diabetes mellitus (NIDDM) who are treated with insulin and oral agents and who are following a diet for their diabetes, by years since diagnosis of diabetes. Sample size: 2268 subjects with NIDDM.

For those who reported that they were trying to follow a diabetes diet but had difficulty adhering to it, questions were asked about particular situations that they found difficult (Table 2). The subjects reported that various situations were problematic, most notably the desire to eat foods not on the diabetes diet. Of importance, two situations were not issues for these patients: lack of support from family and friends and being unsure about what foods they should eat. In general, difficulties with following a diet for diabetes were expressed less frequently as age increased.

About 40% of patients with IDDM and 26% of those with insulin-treated NIDDM reported that they test their blood glucose level at least once per day, but this proportion was substantially lower for patients with NIDDM not treated with insulin (5.3%) (Table 1). The proportion who self-monitored their glucose level at least once per day declined markedly with increasing age (Figure 4). About 20% to 30% of diabetic subjects also tested their urine glucose at least once per week (Table 1). Frequent hyperglycemia and glycosuria were reported by about one fourth of diabetic subjects who self-tested or knew the results of tests done by their physicians (Table 1).

View larger version (17K): [in this window] [in a new window]   Figure 4. Percentage of adults with diabetes who self-monitor their blood glucose level at least once per day. Sample sizes: patients with insulin-dependent diabetes mellitus (IDDM) equals 124; patients with insulin-treated non–insulin-dependent diabetes mellitus (NIDDM) equals 922; patients with NIDDM not treated with insulin equals 1346. (Adapted from reference 12.).

Diabetic subjects in the 1989 NHIS were asked whether they had attended any education program or class about their diabetes, including a diabetes management course. Insulin use was associated with having had diabetes education. Subjects with IDDM had the highest proportion (59%) who had received patient education compared with 49% of those with insulin-treated NIDDM and 24% of those with NIDDM not treated with insulin. Subjects were also questioned about where they had obtained information about diabetes. Almost all had obtained information from some source, with a physician being by far the most likely source (for 86% of subjects).

Discussion

Top Methods Results Discussion Author & Article Info References

The nature of appropriate diabetes care has been defined by many individuals and organizations. Interest in this subject has been heightened by the DCCT, which found an approximately 50% to 70% reduction in the risk for retinopathy, nephropathy, and neuropathy in subjects with intensively treated IDDM compared with that in conventionally treated patients, although full normalization of blood glucose level was not achieved in either group [6]. This risk reduction was found for both the delay in onset and the progression of complications and was found regardless of age, sex, or duration of diabetes.

The DCCT did not study patients with NIDDM. However, an extensive body of scientific data, including data from epidemiologic studies, controlled clinical trials, and studies in animals with experimental diabetes, provides evidence that hyperglycemia is the proximate cause of diabetic microvascular complications in both IDDM and NIDDM [21, 22]. Studies of representative patient populations in southern Wisconsin found a direct relation between the degree of hyperglycemia and the incidence and progression of diabetic retinopathy in both IDDM and NIDDM (23; see also "Relation of Glycemic Control to Diabetic Microvascular Complications in Diabetes Mellitus"). It is important to note that when patients were stratified by level of glycosylated hemoglobin, there were virtually no differences in risk between the two types of diabetes. Because the pathophysiology of microvascular disease in NIDDM thus appears to be the same as in IDDM, intensive metabolic control has been recommended for both diabetic groups [6-9].

Achievement of glucose control in the DCCT required the coordinated efforts of physicians, diabetes educators, nutritionists, and behavioral specialists. However, in practice, this set of health practitioners is often not available to the patient, probably for logistic and financial reasons. Indeed, the representative sample of diabetic subjects in the NHIS used few diabetes specialists. A minority of diabetic subjects received care from health professionals such as ophthalmologists, diabetes educators, or dietitian-nutritionists. The proportion of diabetic patients who may currently be in comprehensive care plans is unknown. In the 1989 NHIS, approximately 10% of diabetic subjects had health insurance coverage through health maintenance organizations and approximately 5% had coverage through individual practice associations [24].

The frequency of visits by diabetic patients to their physicians may be less than that required for intensive management of diabetes. Approximately 10% of the NHIS diabetic subjects did not have a physician for regular care of their diabetes. Of those with a regular physician, one third had fewer than four visits each year. Other studies have shown that the physicians who care for diabetic patients are usually not specialists in diabetes: Less than 10% of all visits for diabetes to office-based physicians are made to diabetologists or endocrinologists [25, 26]. A nation-wide survey of physicians found that most patients with diabetes receive their primary care from internists and general and family practitioners [27, 28].

Almost all NHIS subjects were treated with either oral agents or insulin. A substantial proportion of insulin-treated subjects were not using multiple daily insulin injections. Only a small proportion of those treated with insulin, and an even smaller proportion of those treated with oral agents, were monitoring their blood glucose level. These practices probably will not lead to a level of metabolic control that will prevent diabetic microvascular complications [6, 23]. Indeed, in community-based studies of patients with NIDDM, the mean fasting plasma glucose level ranged from 7.8 mmol/L among a group of upper-class whites in California to 12.9 mmol/L among the Pima Indians in Arizona (Figure 5). These values are far above the level of 5.2 mmol/L for U.S. adults who do not have diabetes [29].

View larger version (36K): [in this window] [in a new window]   Figure 5. The mean fasting plasma glucose level in subjects without diabetes and in representative patient populations in U.S. community-based studies. DM equals diabetes; Japn Amer equals Japanese American; Hisp equals Hispanic; Native Amer equals Native American. Adapted from reference 29.

Blood glucose self-monitoring can result in improved glycemic control, a lower incidence of diabetic retinopathy, and reduction of patient expenses [6, 12, 30-33]. The NHIS data show that blood glucose measurements were obtained during physician visits, but these were too infrequent to achieve a level of glycemic control that would lead to reduction in the diabetic microvascular complication rate. Although cost has been considered a barrier to self-monitoring, one study did not find that economic factors, including health insurance and income, were statistically significant determinants of self-testing [12]. For patients with IDDM in another study, no correlation was found between self-testing and the presence of health insurance [34]. Only a small minority of diabetic persons (14% of those aged 18 to 65 years and 2% of those more than equals 65 years) do not have health insurance [24], and blood glucometers and strips can be covered by Medicare and commercial health insurance [4].

The critical role of diabetes education in quality diabetes care is defined in the standards of care adopted by the American Diabetes Association [35]. Patient education can result in increased rates of blood glucose self-monitoring, compliance with overall management, improved glycemia, and reduced incidence of diabetic complications [12, 36-39]. In studies based on diabetic subjects in the NHIS, two outcomes occurred with substantially greater frequency for patients with NIDDM who had had patient education compared with those who had not: blood glucose self-monitoring at least once per day and having a dilated eye examination at least once in the past year [12, 38]. Despite these favorable effects, the NHIS data indicate that only 35% of diabetic adults in the United States have ever attended a diabetes education class or program.

In summary, achieving and maintaining near-normal glycemia in both IDDM and NIDDM may require close monitoring and ongoing support from a health care team, ample financial resources, and advanced patient knowledge and motivation. Providing this level of diabetes management to all diabetic persons may require major changes in the health care system for diabetes and in patient self-care practices.