Frequency of Blood Glucose Monitoring in Relation to Glycemic Control in Patients With Type 2 Diabetes

Data were analyzed from the third National Health and Nutrition Examination Survey (NHANES III), in which questionnaire, clinical, and laboratory data were obtained for a representative sample of adults with type 2 diabetes. NHANES III was conducted from September 1988 to October 1994 and included a stratified probability sample of the civilian noninstitutionalized U.S. population (2). Participants were interviewed in their homes and were given a standardized set of examinations and laboratory measurements in a mobile examination center. There were 16,993 participants aged ≥25 years, of whom 1,608 had been diagnosed with diabetes by a physician before the survey. Women with diabetes diagnosed only during pregnancy (n = 105) and subjects with type 1 diabetes, defined as those with age at diagnosis <30 years who had continuous insulin use since diagnosis of diabetes (n = 23), were excluded from analysis. The remaining 1,480 subjects were considered to have type 2 diabetes.

Information was obtained by structured questionnaires on diabetes therapy and blood glucose self-monitoring. Measurement of HbA_1c was made during a separate clinical examination in which 88.2% of the interviewed subjects participated (3). HbA_1c was measured by a high-performance liquid chromatographic assay as used in the Diabetes Control and Complications Trial. The upper limit of normal for HbA_1c in the assay system is 6.1%, defined as the mean + 2 SDs (5.27 + 0.86%) for the group of people with fasting plasma glucose <110 mg/dl and 2-h postchallenge glucose <140 mg/dl. This value (6.1%) is virtually identical to the upper limit of normal (6.0%) recommended by the ADA using the same assay system (1). Statistical analyses were performed using SAS (Statistical Analysis System; Cary, NC) with appropriate survey sampling weights. Logistic regression was performed using SUDAAN (Research Triangle Institute, Research Triangle Park, NC) to assess the relationship between HbA_1c and self-monitoring in those treated with insulin, oral agents, or diet alone.

The median age of the patients was 62.5 years. About 44% were male, 26% were of minority race or ethnicity, 55% had a high school education or more, 93% had health insurance, 96% had a regular source of primary medical care, and 85% had at least two physician visits in the 12 months before the survey.

Table 1 shows HbA_1c values and the frequency of blood glucose self-monitoring, according to diabetes therapy. One-fourth of patients treated with insulin and most patients treated with oral agents or diet alone had never monitored their blood glucose or monitored it less than once per month. Self-monitoring at least once per day was practiced by 39% of those taking insulin and was infrequent for those treated with oral agents or diet alone.

In Fig. 1, the patients are distributed according to deciles of HbA_1c. The percentage of patients in each decile who were treated with insulin rises with increasing HbA_1c value. The percentage treated with oral agents is relatively constant across deciles 3–10, and the percentage treated with diet alone declines markedly with increasing HbA_1c value.

The frequency of self-monitoring by patients in each decile of the HbA_1c distribution is shown in Fig. 1. The proportion of patients who tested at least once per week or who tested at least once per day increased with increasing HbA_1c value. Testing at least twice per day was uncommon and showed no relation to HbA_1c value.

Figure 2 illustrates the frequency of self-monitoring by patients for those treated with insulin, oral agents, or diet alone in each quintile of the HbA_1c distribution. There was little relationship between HbA_1c value and the proportion testing at least once per week for those treated with insulin or oral agents, although, for those treated with diet alone, the test frequency was higher in the highest two quintiles of HbA_1c (comprising 14% of diet-treated patients). The proportion testing at least once per day was lower in the middle quintile for those taking insulin, but there was no relationship to HbA_1c for those treated with oral agents or diet alone. The proportion testing at least twice per day was highest in the lowest quintile for those taking insulin (8% of insulin-treated patients) and was low throughout the spectrum of HbA_1c for those taking oral agents or treated with diet alone.

Logistic regression models were constructed to further evaluate the association of self-monitoring and HbA_1c. In these models, there was no relationship between the frequency of self-monitoring and HbA_1c level for those treated with insulin, oral agents, or diet alone (P > 0.5).

The data for all patients combined indicate that self-monitoring of blood glucose is more common as HbA_1c increases, suggesting that patients with poorer glycemic control have a greater tendency to self-monitor. However, the increase in frequency of self-monitoring with increasing HbA_1c value is related to the higher proportion of insulin-treated patients, who are more likely to self-monitor, in higher HbA_1c categories. Within each diabetes therapy category there was little relationship between the frequency of testing and HbA_1c value. Indeed, half of patients with HbA_1c >8%, the ADA value at which intensification of glucose control is recommended (1), monitored less than once per week.

Prior studies have shown that self-monitoring is more common in Caucasian patients than in African-American or Mexican-American patients (3,4). Those with more education, those who have had a diabetes patient education class, and those who have frequent physician visits are more likely to practice self-monitoring (4). However, self-monitoring was not related to higher income or to having health insurance (4).

The relationship between HbA_1c value and self-monitoring frequency was investigated in a study in England of 290 type 2 diabetic patients treated with insulin; no association was found (5). A study in Missouri of 61 type 2 diabetic patients who self-monitored found no difference in mean GHb values based on the frequency of testing (6). Among 115 patients treated with oral agents at a Veterans Administration Medical Center in Arizona, glucose control was independent of the number of blood glucose test strips dispensed (7). Some reports have questioned the efficacy of self-monitoring in patients with type 2 diabetes (8,9). In contrast, the importance of self-monitoring in patients with type 1 diabetes is more certain, and an increased frequency of self-monitoring is generally associated with decreased HbA_1c values for these patients (5,10).

Self-monitoring is considered to be a tool to guide patient and physician action with respect to changes in diet, physical activity, and use of antihyperglycemic medication. Both patient and physician action are needed to change diabetes management when hyperglycemia is evident. Patients with differing levels of insulin resistance and β-cell reserve may respond differently to self-monitoring and changes in diabetes control regimens. The true impact of self-monitoring could be assessed in a randomized clinical trial with pre-established guidelines for how glucose monitoring results would be used to facilitate achievement of glycemic targets.

Adequate co-involvement of the medical care team, the patient, and the family appears to be important in effective management of diabetes. Clinical trials have demonstrated that health care systems that utilize nurses and other nonphysician personnel as case managers and those with telephone-based contact are efficacious in improving glycemic control, blood pressure, and lipid levels (11–14). Key elements in the success of the Diabetes Control and Complications Trial were the frequent clinic visits involving nurses and dietitians and the extensive telephone access to these health care practitioners. These elements were afforded to patients in the intensive intervention group, and HbA_1c remained low throughout the study (15). Availability of social support, such as that provided by nurse case-managers, appears to be a major factor that contributes to adherence by diabetic patients to such behaviors as following a diabetic diet, weight loss, taking prescribed medications, checking blood glucose, and checking feet (16). In a logistic regression analysis of the NHANES III type 2 diabetic patients, the HbA_1c level was not significantly associated with having a primary source of ambulatory medical care, the number of physician visits per year, having any type of health insurance, or having private insurance (17).

The NHANES III is a cross-sectional study, and it cannot be determined from this study whether self-monitoring influenced blood glucose control. Furthermore, the study was conducted during the period when Medicare reimbursed the cost of blood glucose monitors and test strips for monitors only for patients treated with insulin. Although self-testing of blood glucose does not require monitors, it is likely that the lower frequency of testing by patients treated with oral agents and diet alone was a result, in part, of the lack of reimbursement. In 1998, this policy was changed to reimburse patients for these items, regardless of whether the patient is treated with insulin. In a survey conducted in 1997–1999, the proportion of all diabetic patients who self-monitored at least once per day was 44% (18), which was higher than that seen in NHANES III.