From the study of adherence to treatments for diabetes, it is apparently important to assess the level of adherence to each component of the treatment regimen independently (i.e. self-monitoring of blood glucose, administration of insulin or oral hypoglycaemic agents, diet, physical activity, foot care and other self-care practices) instead of using a single measure to assess adherence to the overall treatment. This is because there appears to be little correlation between adherence to the separate self-care behaviours, suggesting that adherence is not a unidimensional construct (21,30). This finding has been reported for both type 1 and type 2 diabetes (31). Furthermore, there appear to be different relationships between adherence and metabolic control for persons with different types of diabetes (32). Consequently, the following section on adherence rates has been organized to reflect these two issues. First there is a discussion of adherence to each element of the regimen; this is followed by an analysis of adherence by diabetes type.
A. Adherence to treatment for type 1 diabetes
Self-monitoring of glucose. The extent of adherence to prescribed self-monitoring of glucose levels in blood varies widely, depending on the frequency or aspect assessed in the study. For example, in a sample of children and adolescents with type 1 diabetes (33), only 26% of study participants reported monitoring glucose levels as recommended (3 - 4 times daily), compared to approximately 40% of the adults with type 1 diabetes (34). Similar findings were reported in a Finnish study (n = 213; patients aged 17 - 65 years), in which 20% of the study participants monitored their blood glucose as recommended, and 21% of respondents made daily or almost daily adjustments to their insulin dosage according to the results of self-monitoring of blood glucose. Only 6% reported never performing the prescribed blood glucose tests (35). A study conducted in the United States replicated the latter result in patients with type 1 diabetes (mean age = 30 years), of whom 7% reported never testing their glucose levels (21).
Other studies have assessed adherence based on incorrect performance (intentional or unintentional) of the component behaviours involved in glucose monitoring in urine or blood. One study reported that up to 80% of adolescents made significant mistakes when estimating glucose concentrations in urine (36). Between 30% and 60% made errors in the timing procedures involved in self-monitoring of blood glucose (37). Others inaccurately reported concentrations; up to 75% may under-report actual mean concentrations of blood glucose, and up to 40% have been found to over-report or to invent phantom values (38).Other studies have found that between 40% and 60% of patients fabricated results (39,40) and 18% failed to record their results (40). In recent years, the development of blood glucose meters with electronic memory has made it more difficult, though not impossible, for patients to fabricate the results of blood-glucose monitoring.
Administration of insulin. The prevalence of adherence to insulin administration varies widely. In a study conducted in Finland (35) most of the respondents reported adhering to insulin injections as scheduled either daily (84%) or almost daily (15%).Other studies have framed the adherence question differently. Rates for "never missing a shot" varied from 92% in a sample of young adults (21) to 53% in a sample of children (41); while 25% of adolescents reported "missing insulin shots within 10 days before a clinic visit" (42).
A study conducted by Wing et al. (37) assessed the quality of performance of insulin administration (intentional or unintentional errors).The use of unhygienic injections was noted in 80% of patients and the administration of incorrect doses of insulin in 58%. In studies assessing the intentional omission of insulin to control weight, Polonsky et al. (43) reported that 31% of study participants (n = 341; female patients aged 13 - 60 years) admitted to intentional omission of insulin, but only 9% reported frequent omission to control weight. More recently, Bryden et al. (44) reported that 30% of female adolescents (but none of the males in the sample) admitted under-using insulin to control weight.
Diet. The results of research on adherence to prescribed dietary recommendations have been inconsistent. In studies by Carvajal et al. (45) in Cuba, and Wing et al. (37) in the United States, 70 - 75% of study participants reported not adhering to dietary recommendations, but in a study in Finland by Toljamo et al. (35), adherence to dietary recommendations was high: 70% of participants reported always or often having a regular main meal, while only 8% reported always having irregular mealtimes. In answer to questions regarding the foods prescribed, over half of the participants reported assessing both the content and amount of food that they ate daily (48%) while 14% of the respondents did not evaluate their food at all. Christensen et al. reported similar findings (46): 60% of study participants (n = 97) adhered to the number and timing of planned meals, while only 10% of patients adhered to planned exchanges, 90% of the time.
Physical activity and other self-care measures. Literature on the extent of adherence to prescribed recommendations for physical activity among patients with type 1 diabetes is scarce. One study conducted in Finland indicated that two-thirds of study participants (n = 213) took regular daily exercise (35%) or almost daily exercise (30%), while 10% took no exercise at all (35). In the same study, only 25% of study participants reported taking care of their feet daily or almost daily, while 16% reported never taking care of their feet as recommended (35).
B. Adherence rates for type 2 diabetes
Glucose monitoring. In a study conducted to assess patterns of self-monitoring of blood glucose in northern California, United States, 67% of patients with type 2 diabetes reported not performing self-monitoring of blood glucose as frequently as recommended (i.e. once daily for type 2 diabetes treated pharmacologically) (34). Similar findings were reported in a study conducted in India, in which only 23% of study participants reported performing glucose monitoring at home (47).
Administration of medication. Among patients receiving their medication from community pharmacies (n = 91), adherence to oral hypoglycaemic agents was 75%. Dose omissions represented the most prevalent form of nonadherence; however, more than one-third of the patients took more doses than prescribed. This over-medication was observed more frequently in those patients prescribed a once-daily dose (48). Similar adherence rates of between 70 and 80% were reported from the United States in a study of oral hypoglycaemic agents in a sample of patients whose health insurance paid for prescribed drugs (49). Dailey et al. (50) studied 37 431 Medicaid-funded patients in the United States, and used pharmacy records to show that patients with type 2 diabetes averaged about 130 days per year of continuous drug therapy, and that at the end of 1 year, only 15% of the patients who had been prescribed a single oral medication were still taking it regularly.
Diet. In a study conducted in India, dietary prescriptions were followed regularly by only 37% of patients (47), while in a study in the United States about half (52%) followed a meal plan (51). Anderson & Gustafson (52) reported good-to-excellent adherence in 70% of patients who had been prescribed a high-carbohydrate, high-fibre diet. Wing et al. (53) showed that patients with type 2 diabetes lost less weight than their nondiabetic spouses and that the difference was mainly due to poor adherence to the prescribed diet by the diabetic patients. Adherence to dietary protocols may depend upon the nature of the treatment objective (e.g. weight loss, reduction of dietary fat or increased fibre intake).
Physical activity. Several studies have reported on adherence to prescribed physical activity. For example, in a study in Canada of a sample of patients with type 2 diabetes randomly selected from provincial health records, few respondents participated in informal (37%) or organized (7.7%) physical activity programmes (54). A survey in the United States found that only 26% of respondents followed a physical activity plan (51). A study assessing the attitudes and adherence of patients who had completed outpatient diabetes counselling observed that only 52% exercised on three or more days per week after the counselling programme was completed (55).
C. Adherence to treatment for gestational diabetes
One study was found that had assessed adherence to treatment for gestational diabetes. Forty-nine pregnant women with pre-existing (overt) diabetes (68% with type 1 and 32% with type 2 diabetes) were assessed, using self-report, on their adherence to a number of self-care tasks on three occasions during pregnancy (mid-second, early third and late third trimester) (56). In general, the participants reported being adherent. However, there was considerable variation across different regimen components: 74-79% of women reported always following dietary recommendations, compared to 86-88% who followed the recommendations for insulin administration, 85-89% who followed the recommendations for managing insulin reactions and 94-96% who followed those for glucose testing.