Background and Aims

The DFRQ is a widely-used measure of parent involvement in diabetes management, created >30 years ago. With increased use of diabetes technologies, it is timely to update the DFRQ regarding CGM use. We evaluated the psychometric properties of an updated DFRQ.

Methods

Youth with T1D participating in a CGM study and their parents completed the DFRQ survey with four additional CGM-specific items (i.e., responding to CGM alerts/alarms). Higher scores indicate more parent involvement. Item-to-total correlations and Cronbach’s α assessed internal consistency; correlations determined concurrent and predictive validity.

Results

Participants (N=119, 49% female) were (mean±SD) aged 13.2±2.7yrs, with T1D for 6.6±3.5yrs.

The revised survey had 14 items after removal of items with little variability, low item-to-total correlation, or outdated features.

Cronbach’s α for the updated DFRQ was 0.88 for youth and 0.93 for parents. Youth and parent DFRQ scores were highly correlated (r=0.81, p<0.0001). Youth and parent scores were inversely correlated with youth age (r=-0.76, r=-0.81, respectively, both p<0.0001) and T1D duration (r=-0.23, p=0.01; r=-0.33, p=0.0004).

Higher youth and parent scores were also correlated with higher parent reports of youth diabetes treatment adherence (r=0.43, r=0.49, respectively, both p<0.0001).

For youth aged <13yrs, higher parent DFRQ scores were positively associated with youth CGM percent time-in-range (TIR) (70-180mg/dL) 3 months later (r=0.36, p=0.02).

Conclusions

The updated DFRQ youth and parent surveys demonstrated strong psychometric properties and predictive validity for percent TIR for youth aged <13yrs. These surveys can help assess family responsibility-sharing for CGM use in clinical and research settings.