Schatz P E
California State University, Los Angeles.
J Am Diet Assoc. 1988 Jun;88(6):708-12.
In the past, research directed at understanding patient compliance behavior has generally resulted in findings that were not predictive of compliance or that were at best contradictory. A study of 98 adult patients with diabetes was conducted to investigate some of the variables affecting compliance in that patient population. Compliance and knowledge scores were computed for all 98 patients interviewed. Twelve patients at each extreme of the compliance distribution (no. = 24) were selected for further interviews. Nonparametric analyses were used to test for group differences. Significant differences were found in the mean values for the following variables (mean +/- standard error of the mean for low and high compliers): years diabetic, 5.7 +/- 1.5 and 12.1 +/- 2.4, respectively (p = .05); knowledge scores, 12.2 +/- 1.4 and 20.9 +/- 2.6, respectively (p = .001); and Health Belief Model score, 35.5 +/- 2.6 and 47.2 +/- 1.2, respectively (p = .002). In addition, significant differences were found in the use of home glucose monitoring, which was practiced by 33% of low compliers and 100% of high compliers (p = .002). No significant differences were found between the groups for sex, marital status, employment status, or ethnicity. The patients' perception of health care delivery and the knowledge acquired about the disease condition appeared to be important factors that had a bearing upon the control the patients chose to exert upon their care. There also appears to be some indication that the Health Belief Model can be used for predicting compliance.(ABSTRACT TRUNCATED AT 250 WORDS)