FACTORS AFFECTING HEALTH-RELATED QUALITY OF LIFE IN TYPE 1 DIABETES PATIENTS IN IRELAND

Purpose:

   Generic measures of quality of life (QoL) are often used an inputs in economic evaluation and thus play a key role in reimbursement of health interventions. It is important to understand factors that might influence patients’ scores on these measures. 

Method:

   The Irish DAFNE (Dose Adjustment for Normal Eating) trial was a cluster-randomized trial of Type 1 diabetes patients to determine the effectiveness of group versus individual follow-up for Type 1 diabetes management. Using baseline data from this study, we identified potential covariates that might influence patients QoL, as measured by the EuroQol 5 Dimension (3 Level version) (EQ-5D-3L). The primary outcome was EQ-5D-3L scores and a secondary analysis was conducted using the visual analogue scale (VAS) component of the EQ-5D. We carried out a series of univariate regression analyses to determine potential covariates in the multivariate model (using p<0.2 for inclusion). The final multivariate model was chosen by stepwise addition comparing Akaike Information Criterion (AIC) values.

Result:

  The Irish DAFNE study enrolled 494 Type 1 diabetes patients from 6 centers in Ireland between 2012-2013. The median (interquartile range, IQR) age was 40 (IQR: 31-49) years. Of the trial patients, 52.6% were female. Thirteen per cent of trial patients had a body mass index (BMI) of greater than 30 kg/m². The median number of years since disease diagnosis was 14 (IQR: 7-23) years.  The mean glycated hemoglobin value in the patient population was 8.30% (SD: 1.34). There was low correlation between the EQ-5D (0.33) and Diabetes Specific Quality of Life Scale (0.24). The mean EQ-5D score and VAS score and were 0.871 (SD: 0.219) and 71.8 (SD: 16.8), respectively. In multivariate regression, insulin injections, self-reported mental illness, HbA1c, and obesity (BMI>=30 kg/m²) all had statistically significant effects on EQ-5D score (p-value<0.05). Mental illness had the greatest impact (-0.18, p=0.002) on EQ-5D score. In a secondary analysis, the VAS component of the EQ-5D questionnaire was used as the outcome. Similarly, HbA1c, obesity, and mental illness were all statistically significant (p<0.05) and in negatively associated with the outcome. 

Conclusion:

   Obesity, HbA1c, and mental illness appear to have significant effects on EQ-5D scores in patients with Type 1 diabetes. Knowledge that these factors affect quality of life may have implications for patient management.