Children's and adolescent's self - assessment of metabolic control versus professional judgment: a cross-sectional retrospective and prospective cohort study

Abstract

Background: Morbidity and mortality in T1DM depend on metabolic control, which is assessed by HbA1c measurements every 3-4 months. Patients' self-perception of glycemic control depends on daily blood glucose monitoring. Little is known about the congruence of patients' and professionals' perception of metabolic control in T1DM.

Objective: To assess the actual patients' self-perception and objective assessment (HbA1c) of metabolic control in T1DM children and adolescents and to investigate the possible factors involved in any difference.

Methods: Patients with T1DM aged 8 - 18 years were recruited in a cross-sectional, retrospective and prospective cohort study. Data collection consisted of clinical details, measured HbA1c, self-monitored blood glucose values and questionnaires assessing self and professionals' judgment of metabolic control.

Results: 91 patients participated. Mean HbA1c was 8.03%. HbA1c was higher in patients with a diabetes duration > 2 years (p = 0.025) and in patients of lower socioeconomic level (p = 0.032). No significant correlation was found for self-perception of metabolic control in well and poorly controlled patients. We found a trend towards false-positive memory of the last HbA1c in patients with a HbA1c > 8.5% (p = 0.069) but no difference in patients' knowledge on target HbA1c between well and poorly controlled patients.

Conclusions: T1DM patients are aware of a target HbA1c representing good metabolic control. Ill controlled patients appear to have a poorer recollection of their HbA1c. Self-perception of actual metabolic control is similar in well and poorly controlled T1DM children and adolescents. Therefore, professionals should pay special attention that ill controlled T1DM patients perceive their HbA1c correctly.

Figure 1

HbA1c in relation to (A) age, (B) duration of diabetes, (C) glucose self-monitoring and (D) socioeconomic level. There is a tendency towards higher HbA1c values with age (p = 0.065). HbA1c values correlate with the duration of diabetes (p = 0.025). HbA1c does not correlate with the number of blood glucose self-measurements (p = 0.173) but correlates with the socioeconomic level (p = 0.032). Data are given as boxplots and were statistically analyzed by Kruskal-Wallis tests with a significance level of p ≤ 0.05.

Figure 2

Memory of last HbA1c. The recollection of the last measured HbA1c values was assessed by comparing the objective HbA1c values 3 months ago with the patient’s recollection of this HbA1c. The gap between the last measured and remembered HbA1c value was then compared to the actual HbA1c. Data are shown as boxplots with the actual HbA1c in categorized form on the x-axis. Note that there is a tendency towards wrong positive memory of the last HbA1c in patients having an HbA1c > 8.5% (p = 0.069). Data were analyzed by the Kruskal-Wallis test with a significance level of p ≤ 0.05.

Figure 3

Knowledge of target HbA1c. All patients were asked for the currently recommended HbA1c level for good glycemic control (y-axis). A) These data were then compared to the actual HbA1c of each patient (x-axis). No significant difference was found (p = 0.154). B) Data were also correlated with the age finding significantly higher target HbA1c levels in older patients (p = 0.017). Data were analyzed by the Kruskal-Wallis test with a significance level of p ≤ 0.05.

Figure 4

Self-perception of metabolic control in T1DM. HbA1c levels were put in relation to a self assessment score (SAS). Patients were asked to predict their HbA1c qualitatively. Data were collected with questionnaires and categorized from -2 to +2. A SAS 0 meant that patient’s perception overlapped with the objective result. A SAS of +1 or +2 meant that the measured HbA1c value was better than the last one but this improvement was not perceived by the patient. A SAS -1 or -2 meant that the actual HbA1c value was worse than the last one but predicted otherwise by the patient. No significant correlation was found between the SAS and the actual HbA1c level (p = 0.99). Data are shown as bar graphs and were analyzed by the Kruskal-Wallis test.

Figure 5

Comparison between professionals’ and patients’ suggestions to improve metabolic control in T1DM. Professionals and patients were invited to make suggestions to improve or maintain metabolic control. A list of items was given. Professionals and patients could choose one or more of the listed items. Only professionals had the possibility to choose the item Change of treatment regimen while only patients could choose the answer Don’t know. Data were analyzed descriptively and are shown as % of all.