Abstract
Objective
The Problem Areas In Diabetes (PAID) questionnaire is widely used to assess emotional distress related to living with diabetes, although it is lengthy for routine clinical use. Our aim was to determine whether the original 20-item PAID questionnaire can be abbreviated, whilst maintaining its reliability, validity and utility.
Methods
We analysed data from the UK DAFNE (Dose Adjustment For Normal Eating) education programme for adults with Type 1 diabetes. Data were analysed at baseline (n = 1547) and 1-year post intervention (n = 846). Exploratory factor analysis (EFA) with principal axis factoring method was used to examine PAID responses within a random half of the baseline data (n = 746). Then, two confirmatory factor analyses (CFA) were conducted using the remaining baseline (n = 801) and 1-year data. Reliability, predictive validity, convergent validity and responsiveness were also examined.
Results
Based on the EFA results, which were corroborated by CFA, an 11-item PAID questionnaire was identified with a cut-off score of 18 indicating severe diabetes distress. In the current sample, this brief version had high internal consistency (Cronbach’s α = 0.93). Predictive validity was demonstrated with the PAID-11 identifying severe diabetes distress from the original 20-item measure, with 95% sensitivity and 96% specificity. Convergent validity was demonstrated by strong positive correlations with HADS anxiety and depressive symptoms (average r = 0.65 and r = 0.55, respectively), while divergent validity was shown with weaker correlations with EQ5D health status (average r = 0.37).
Conclusions
Based on present results, PAID-11 appears to be a valid and reliable measure, which seems suitable for use as a brief tool for the detection of diabetes distress in adults with type 1 diabetes. Importantly, this tool may reduce participant burden in multi-measure studies. However, further studies are urgently needed to determine the validity and utility of PAID-11 beyond the UK DAFNE population.
The Problem Areas In Diabetes (PAID) questionnaire is widely used to assess emotional distress related to living with diabetes, although it is lengthy for routine clinical use. Our aim was to determine whether the original 20-item PAID questionnaire can be abbreviated, whilst maintaining its reliability, validity and utility.
Methods
We analysed data from the UK DAFNE (Dose Adjustment For Normal Eating) education programme for adults with Type 1 diabetes. Data were analysed at baseline (n = 1547) and 1-year post intervention (n = 846). Exploratory factor analysis (EFA) with principal axis factoring method was used to examine PAID responses within a random half of the baseline data (n = 746). Then, two confirmatory factor analyses (CFA) were conducted using the remaining baseline (n = 801) and 1-year data. Reliability, predictive validity, convergent validity and responsiveness were also examined.
Results
Based on the EFA results, which were corroborated by CFA, an 11-item PAID questionnaire was identified with a cut-off score of 18 indicating severe diabetes distress. In the current sample, this brief version had high internal consistency (Cronbach’s α = 0.93). Predictive validity was demonstrated with the PAID-11 identifying severe diabetes distress from the original 20-item measure, with 95% sensitivity and 96% specificity. Convergent validity was demonstrated by strong positive correlations with HADS anxiety and depressive symptoms (average r = 0.65 and r = 0.55, respectively), while divergent validity was shown with weaker correlations with EQ5D health status (average r = 0.37).
Conclusions
Based on present results, PAID-11 appears to be a valid and reliable measure, which seems suitable for use as a brief tool for the detection of diabetes distress in adults with type 1 diabetes. Importantly, this tool may reduce participant burden in multi-measure studies. However, further studies are urgently needed to determine the validity and utility of PAID-11 beyond the UK DAFNE population.
| Original language | English |
|---|---|
| Pages (from-to) | 27-38 |
| Journal | Diabetes Research and Clinical Practice |
| Volume | 149 |
| Early online date | 30 Jan 2019 |
| DOIs | |
| Publication status | Published - Mar 2019 |