JMIR Publications

ABSTRACT

Background:

Hypoglycemia prognostic models contingent on prospective, self-reported survey data offer a powerful avenue for determining real-world event susceptibility and enhanced interventional targets.

Objective:

This protocol describes the design and implementation of iNPHORM, a 12-wave internet-based panel survey in the United States (US). The iNPHORM study has two primary objectives: 1) to measure the incidence of real-world severe and non-severe daytime and nocturnal hypoglycemia (SDH, SNH, NSDH, NSNH) in American adults with type 1 or 2 diabetes mellitus (T1DM, T2DM) taking insulin and/or secretagogues; and 2) to develop and validate three hypoglycemia prognostic models for total severe hypoglycemia (SH), NSDH, and NSNH. The secondary objective is to assess the impact of different antihyperglycemic regimens on hypoglycemia rates.

Methods:

Americans (≥18 years old) with T1DM or T2DM taking insulin and/or secretagogues were conveniently sampled from a large, probability-based internet panel. A predetermined sample size of N = 1,250 baseline responders was specified. We obtained anonymized and deidentified, self-reported data via online questionnaires at baseline and over successive monthly interwaves for one year. Prognostic information pertained to participants’ clinical (i.e., diabetes, general health status, and health-related quality of life) and non-clinical (i.e., anthropometric, demographic, situational/environmental, and lifestyle/behavioural) characteristics. Outcome data were collected on SDH, SNH, NSDH, and NSNH event frequencies, identification/treatment modes, management behaviours/supports, as well as impaired awareness and fear of hypoglycemia. Summary and hypoglycemia frequency statistics for the incidence of hypoglycemia (incidence densities and proportions) will be calculated. Predictor and outcome responses will be analyzed to develop and internally validate three real-world prognostic models for SH (combined daytime and nocturnal), NSDH, and NSNH. Severe hypoglycemia will be modelled over one-year using Cox regression; NSDH and NSNH frequencies will be modelled over 30-days using Poisson or negative binomial regression. Discrimination and calibration will be evaluated to assess model performance. We will also examine the causal impact of different antihyperglycemic regimens on hypoglycemia rates using advanced analytic methods.

Results:

Recruitment and data collection commenced February 2020 and concluded March 2021. Our target sample size was achieved; 1,694 individuals completed the baseline questionnaire and 1,206 participated in the iNPHORM longitudinal panel. Non-response and dropout was minimal.

Conclusions:

iNPHORM is the first US-based prognostic study on hypoglycemia to leverage prospective, longitudinal data. The results of this investigation will help clarify hypoglycemia incidence and risk estimation, helping guide the development of more effective interventions in the real world. Clinical Trial: ClinicalTrials.gov NCT04219514; https://clinicaltrials.gov/ct2/show/NCT04219514