Objective. To determine if there is a relationship between diabetes outcome as measured by HbA1C and the number of multidisciplinary clinic visits per year in children and youth with diabetes.
Research Design and Methods. The number of clinic visits per year, the mean HbA1C (by DCA 2000, Miles, Tarrytown, NY), type of insurance, parental marital status, parental age, maximal parental grade level achieved, family income, self-identified race, and scores on adherence and knowledge tests were compared for 1995 in 360 patients with a mean age of 11.6 ± 4.8 years (1–2 visits, 85 patients; 3–4 visits, 275 patients), for 1996 in 412 patients with a mean age of 11.6 ± 4.7 years (1–2 visits, 115 patients; 3–4 visits, 297 patients), and for 1997 in 442 patients with a mean age of 11.8 ± 4.9 years (1–2 visits, 126 patients; 3–4 visits, 332 patients).
Results. There was a significant difference in the mean HbA1C levels between subjects with 1 to 2 visits versus 3 to 4 visits during the 3 years of this study. In 1995, the mean HbA1C was 9.0 ± 2.0% for subjects with 1 to 2 visits and 8.3 ± 1.6% for subjects with 3 to 4 visits. In 1996, the mean HbA1C was 9.3 ± 2.0% for subjects with 1 to 2 visits and 8.4 ± 1.6% in those with 3 to 4 visits, whereas in 1997, the mean HbA1C was 9.1 ± 1.9% with 1 to 2 visits and 8.3 ± 1.5% with 3 to 4 visits. There was a significant difference in the number of visits by the age of the subject. The mean age of patients with 1 to 2 visits was 13.6 ± 4.5 years; it was 10.8 ± 4.6 years with 3 to 4 visits. However, for age groups <13 years versus ≥13 years, there was still a difference between HbA1C levels for subjects with 1 to 2 visits compared with 3 to 4 visits (8.9 ± 1.7% versus 8.1 ± 1.3%, respectively). The only patient/family characteristic that had an association with number of visits was the marital status of the parents. Children from single-parent households had fewer visits. There was no association between health insurance status and number of visits and there was no difference between the number of visits and the mean scores on tests of adherence or knowledge. Multivariate analysis showed that the number of visits was a significant predictor of HbA1C even after controlling for age, duration of diabetes, and scores on adherence and knowledge tests.
Conclusions. The finding that subjects with more frequent visits to a multidisciplinary diabetes clinic had lower HbA1C levels during the 3 years of this study suggests that strategies should be developed to promote adherence with quarterly visits, particularly targeted to children from single-parent households and to teens.
The Diabetes Control and Complication Trial1showed the long-term benefits of maintaining optimum blood glucose levels over years for patients with type 1 diabetes. However, the process of health care delivery that was used to achieve this improved outcome has not become the basis of routine clinical practice. The intensive management cohort had monthly clinic visits with a multidisciplinary team comprised of physicians, nurse educators, nutritionists, and psychologists. This visit regimen has not been adopted or advocated by most diabetes specialists or the American Diabetes Association as the standard of care for people with type 1 diabetes.2
Although most centers, particularly those caring for children and teens with diabetes,3 ,4 have suggested that quarterly visits are adequate to promote optimal outcome, few previous studies have been done to determine the effect of routine clinic visits. Jacobson et al5 evaluated pediatric patients with diabetes over a 4-year time span and compared those who had continuous follow-up with those who had irregular clinic attendance. Those with irregular and interrupted follow-up had increased episodes of diabetic ketoacidosis, poor glycemic control, and an increased incidence of retinopathy. This group had lower socioeconomic status and a parental history of separation or divorce. Baumer et al6 showed that children followed for diabetes care by designated diabetes clinics compared with children followed by nonspecialists had lower glycohemoglobin levels, spent less time in the hospital at diagnosis, and had higher rates of screening for microalbuminuria and retinopathy. These studies suggested that subjects' routine care given by multidisciplinary teams had improved diabetes outcomes compared with those with irregular care given by specialists and those with care given by nonspecialists. However, the exact frequency of routine visits was not explored.
The present study was done to determine if pediatric subjects evaluated at yearly or 6-month intervals had a similar outcome to those subjects seen every 3 to 4 months. The association between diabetes outcome as measured by HbA1C levels and the number of multidisciplinary visits in children and teens with this chronic disease was explored.
We tracked patients with diabetes followed in our center and its satellite clinics in Southern California on a computerized database. Number of visits per year to the clinic, HbA1C by DCA 2000 (Miles, Tarrytown, NY), socioeconomic data such as type of insurance including private indemnity, health maintenance organization, or MediCal/CCS (California's entitlement programs), parental marital status, parental age, maximal parental grade level achieved, family income, and self-identified race were included in the database. In addition, the results of the physician and nurse examinations and evaluations were recorded for each patient visit and data were entered after each clinic visit. A measure of adherence to the diabetes regimen, a modified version of the Israel Self Report Measure of Adherence,7was administered to subjects during the study. The 16-item questionnaire that assesses a variety of behaviors linked to the diabetes regimen was modified by Nadeau8 to a 5-point Likert scale (reliability, 0.71). Knowledge was evaluated with a 30-item Diabetes Knowledge Questionnaire developed at Childrens Hospital Los Angeles (reliability, 0.89). The present study was approved by the Committee on Clinical Investigations at Childrens Hospital Los Angeles and informed consent was obtained from the patient and/or parents before administering the measures.
Patients followed in our center were told that it was expected that they would be seen in the clinic setting on a quarterly basis. At clinic visits, patients and their parents were seen by their pediatric endocrinologist for a physical examination, assessment of blood glucose control, and evaluation of their insulin regimen. Patients and their parents also met with the diabetes nurse educator during which time the written or computerized blood glucose records were reviewed. In addition, the checklist outlined in Table 1 was covered to promote diabetes self-care by increasing family competency from basic to advanced in diabetes management. The nutritionist evaluation consisted of assessing the medical nutritional plan, teaching carbohydrate management, and changing advised caloric intake as indicated by growth. A psychosocial assessment was performed by a licensed clinical social worker in patients if suggested by the endocrinologist or nurse.
Statistical analyses were done to compute correlation, relationships between groups, and multiple regressions. Pearson correlation was used to determine associations between HbA1C and number of visits and other patient variables. Paired t tests were computed to determine relationships between mean values of the 1 to 2 visit groups versus the 3 to 4 visit groups for HbA1C, age, and other patient characteristics. χ2 was used to test for differences in frequencies for data categories such as insurance groups and marital status. Multiple regression analyses were performed to estimate combined effect of multiple variables on HbA1C.
Variables that were statistically associated with HbA1C over the 3-year time period were number of patient visits per year, age, duration of diabetes, adherence, and knowledge. HbA1C and number of visits were statistically significant with a negative correlation ofr = −.32 (P = .3 −10E). Age was positively associated with HbA1C (r = .46;P < .002). Duration of diabetes was also positively associated with HbA1C (r =.42; P = .002). The association between adherence and HbA1C was −.34 (P = .0008) and the association between knowledge and HbA1C was −.47 (P = .0012). These variables were subsequently analyzed for differences in clinic visit attendance.
Visits and HbA1C
As shown in Table 2, in 1995, 85 patients (24%) were seen 1 to 2 times and 275 (76%) of the subjects were seen 3 to 4 times. The mean HbA1C was 9.0 ± 2.0% for the group with 1 to 2 visits and 8.3 ± 1.6% for the group seen 3 to 4 times (P = .006). In 1996, 115 (28%) patients were seen 1 to 2 times and 297 (72%) subjects were seen 3 to 4 times in the clinic setting by the multidisciplinary team. For subjects seen 1 to 2 times, the mean HbA1C was 9.3 ± 2.0% and significantly higher than the mean HbA1C of 8.4 ± 1.6% observed in subjects with 3 or 4 visits (P < .0001). Similar results were obtained for 1997; for the 126 (28%) patients seen 1 to 2 times, the HbA1C was 9.1 ± 1.9% compared with the mean HbA1C of 8.3 ± 1.5% observed in 322 (72%) subjects seen 3 to 4 times (P < .0001).
Visits and Age
A significant effect of the age of subjects was seen with regard to number of visits. The mean age of subjects seen 1 to 2 times was 13.6 ± 4.5 years; the mean age of patients seen 3 to 4 times was 10.8 ± 4.6 (P < .0001). However, for age groups <13 years versus ≥13 years, there was still a difference between the HbA1C levels for subjects who came for 1 to 2 visits compared with 3 to 4 visits. The mean HbA1C for subjects <13 years with 1 to 2 visits was 8.9 ± 1.7% (n = 46) versus 8.1 ± 1.3% (n = 185) for those with 3 to 4 visits (P < .003); for patients ≥13 years, it was 9.7 ± 2.2% (n = 68) for those with 1 to 2 visits and 8.9 ± 1.9% (n = 108) for those with 3 to 4 visits (P = .017).
Visits and Duration
The mean duration of diabetes for patients with 1 to 2 visits was 5.3 ± 3.7 years (n = 73), and for 3 to 4 visits, 4.0 ± 3.7 years (n = 243) and statistically different (P = .009).
Visits and Adherence Measure
There was no difference in the mean scores on the measure of adherence with diabetes regimen for subjects that had 1 to 2 vs 3 to 4 clinic visits. The mean score on adherence for patients seen 1 to 2 times was 77 ± 8% (n = 35) and 78 ± 9 for those with 3 to 4 visits (n = 156) (P = .156).
Visits and Knowledge Measure
There was no difference in mean knowledge scores for patients who had 1 to 2 visits or 3 to 4 visits. The mean knowledge score for patients with 1 to 2 visits per year was 79.0 ± 16.5% (n = 140) and 79.1 ± 12.9% for patients with 3 to 4 visits (n = 48) (P = .14).
Visits and Socioeconomic Variables
Table 3 shows the mean parental age, parental education level by grade and family income, marital status of parents, self-described race, and insurance status for patients who had 1 to 2 versus 3 to 4 clinic visits. As shown, there was no difference in these characteristics for the 2 groups except with regard to the marital status of parents. A significantly greater number of subjects with 1 to 2 visits came from single-parent homes compared with those with 3 to 4 visits (P = .02).
Number of visits, age, and duration were examined by using multiple regression to evaluate the combined effect on diabetes control (HbA1C). As shown by using the formula in Table 4, HbA1C was impacted most by patient visits (P = .0003,) then duration of diabetes (P = .02), and then by age (P= .03). The overall formula was statistically significant (P = 2.8 E-10) comprising 14% of variance.
Adding adherence and knowledge to the analyses increased the percentage of variance explained to 22%, with visits being significant (P = .04) as well as adherence (P = .006) and knowledge (P = .03). Patient age and duration of diabetes were independently correlated with HbA1C but were counteracted with the addition of knowledge and adherence scores. This could be because children who were able to perform these questionnaires were closer in age and duration of diabetes.
These 3-year data show that subjects who had 3 to 4 visits per year to the diabetes clinic, during which they were assessed and treated by a multidisciplinary team, had significantly lower mean HbA1C levels compared with subjects seen only 1 to 2 times per year. The number of visits was a significant predictor of HbA1C even after controlling for age, duration, adherence, and knowledge. Therefore, frequent multidisciplinary visits addressing diabetes self-care issues and assessing adherence to diabetes-specific guidelines for the patient and family appeared to be beneficial and predictive toward improved outcome over a 3-year time period. In the changing climate of health care delivery, it would appear reasonable to assume that quarterly visits with a multidisciplinary team should be a covered benefit, particularly for the pediatric population.
Because all patients were encouraged to return for clinic visits quarterly, an attempt was made to determine if parental age, education level, income, race, or insurance status influenced the frequency of clinic visits. These variables did not appear to be associated with the difference in the number of clinic visits. However, parental marital status did have a relationship; a higher percentage of children whose parents were married had 3 to 4 clinic visits compared with those whose parents were separated or divorced. This finding is similar to that of Jacobson et al,5 who showed that children whose parents were separated or divorced had irregular clinic visits over a 4-year time period and that this was associated with poor glycemic control and an increased risk of long-term diabetes complications such as retinopathy. Therefore, specific strategies targeted to children from single-parent households might need to be developed to promote adherence with scheduled diabetes health care visits.
It is well-known that adolescence is associated with a deterioration in glycemic control. In the present study, increasing age was not only associated with increasing HbA1C, but also with decreasing number of clinic visits. Although it has been hypothesized that several barriers to adherence in adolescents, such as loss of parental involvement with blood glucose monitoring,9 might promote this deterioration in glycemic control, it is possible that failure to keep diabetes-related health care visits might also contribute to poor glucose control. Encouraging quarterly clinic visits in teens should be considered as a potential method to reverse the negative effect of increasing age on diabetes control.
It is possible that the higher HbA1C level in patients with fewer clinic visits relates to poorer adherence with the diabetes regimen and that the number of visits is a surrogate marker for adherence. The results of the present study would suggest that this is not the case, because there was no association between the scores for overall adherence and the number of visits.
The exact component or components of the health care visit that promote improved diabetes control include knowledge and adherence. Adjustments in the regimen (ie, switches in insulin type, timing of injections, or insulin delivery system), and improvement in competencies and motivational strategies, which occur with increased frequency of visits, are also possible factors. Future studies must be conducted to determine the mechanism for improvement in HbA1C level that is associated with a greater number of multidisciplinary visits.
The present study supports the importance of multidisciplinary visits aimed at increasing diabetes self-management for children and teens and their families. In particular, visits should be strongly encouraged for hard-to-reach targeted groups such as children from single-parent households and adolescents. As the practice of medicine enters a new era and as health care delivery systems struggle with cost and emerging definitions of quality,10 the importance of health care visits with a multidisciplinary team must be considered.
- Received March 12, 1998.
- Accepted November 30, 1998.
Reprint requests to (F.R.K.) Childrens Hospital Los Angeles, Box 61, 4650 Sunset Blvd, Los Angeles, CA 90027. E-mail:
- ↵Wilson D. Diabetes mellitus in children and adolescents. In: Behrman RE, Kliegman RM, Arvin AM, eds. Nelson Textbook of Pediatics: Update 8. 15th ed. Philadelphia, PA: WB Saunders; 1995:1–12
- Kaufman FR
- Baumer JH,
- Hunt LP,
- Shield JP
- ↵Nadeau M. Living with diabetes: the relationship between quality of life, coping style, adherence and metabolic control in adolescence. Doctoral dissertation. 1993
- Anderson B,
- Ho J,
- Brackett J,
- Finkelstein D,
- Laffel L
- Sidorov A,
- Harris R
- Copyright © 1999 American Academy of Pediatrics