Psychometric properties of an innovative smartphone application to investigate the daily impact of hypoglycemia in people with type 1 or type 2 diabetes

Uffe Søholm, Melanie Broadley, Natalie Zaremba, Patrick Divilly, Giesje Nefs, Jill Carlton, Julia K Mader, Petra Martina Baumann, Mikel Gomes, Gilberte Martine-Edith, Daniel J Pollard, Dajana Rath, Simon Heller, Ulrik Pedersen-Bjergaard, Rory J McCrimmon, Eric Renard, Mark Evans, Bastiaan de Galan, Thomas Forkmann, Stephanie A Amiel, Christel Hendrieckx, Jane Speight, Pratik Choudhary, Frans Pouwer

OAI: oai:www.repository.cam.ac.uk:1810/348789 • DOI: 10.17863/CAM.96215

Introduction

The aim of this study was to determine the acceptability and psychometric properties of the Hypo-METRICS (Hypoglycemia MEasurement, ThResholds and ImpaCtS) application (app): a novel tool designed to assess the direct impact of symptomatic and asymptomatic hypoglycemia on daily functioning in people with insulin-treated diabetes.

Materials and methods

100 adults with type 1 diabetes mellitus (T1DM, n = 64) or insulin-treated type 2 diabetes mellitus (T2DM, n = 36) completed three daily 'check-ins' (morning, afternoon and evening) via the Hypo-METRICs app across 10 weeks, to respond to 29 unique questions about their subjective daily functioning. Questions addressed sleep quality, energy level, mood, affect, cognitive functioning, fear of hypoglycemia and hyperglycemia, social functioning, and work/productivity. Completion rates, structural validity, internal consistency, and test-retest reliability were explored. App responses were correlated with validated person-reported outcome measures to investigate convergent (rs>±0.3) and divergent (rs<±0.3) validity.

Results

Participants' mean±SD age was 54±16 years, diabetes duration was 23±13 years, and most recent HbA1c was 56.6±9.8 mmol/mol. Participants submitted mean±SD 191±16 out of 210 possible 'check-ins' (91%). Structural validity was confirmed with multi-level confirmatory factor analysis showing good model fit on the adjusted model (Comparative Fit Index >0.95, Root-Mean-Square Error of Approximation <0.06, Standardized Root-Mean-square Residual<0.08). Scales had satisfactory internal consistency (all ω≥0.5), and high test-retest reliability (rs≥0.7). Convergent and divergent validity were demonstrated for most scales.

Conclusion

High completion rates and satisfactory psychometric properties demonstrated that the Hypo-METRICS app is acceptable to adults with T1DM and T2DM, and a reliable and valid tool to explore the daily impact of hypoglycemia.

Humans Diabetes mellitus Type 1 Diabetes mellitus Type 2 Hypoglycemia Insulin Reproducibility of Results Psychometrics Adult Aged Middle Aged Benchmarking Mobile Applications Smartphone Surveys and Questionnaires