Have interventions aimed at assisting general practitioners in facilitating earlier diagnosis of type 1 diabetes in children been successful in preventing acute complications? A systematic review

Background: Diabetic ketoacidosis (DKA) is a life-threatening emergency that can result from delayed diagnosis of type 1 diabetes mellitus (T1DM). Three-quarters of Australian children with a new diagnosis of T1DM visit their general practitioner (GP) the week prior to developing DKA, with similar trends observed internationally.

Objective: To summarise interventions in general practice to reduce diagnostic delay in paediatric T1DM and to evaluate their effectiveness.

Methods: Six databases (Ovid, Web of Science, CINAHL, Evidence-Based Medicine Reviews, Google Scholar and EMBASE) were searched. Any English language, less than 20 years study involving interventions targeting GPs specifically in the prevention of paediatric DKA, was included. Primary outcomes were (a) the number of children presenting to the hospital in DKA following diagnostic delay after a GP visit and (b) DKA rate. The secondary outcome was changes in GPs' behaviour regarding timeliness of referrals. Two reviewers completed title, abstract and full-text review, with conflicts resolved by a third reviewer. ROBINS-I risk of bias was used for appraisal. High heterogeneity among studies rendered meta-analysis unsuitable. Structured tabulation of results was completed for analysis. The date of last search was 2 July 2023.

Results: Eight studies were included (three conference abstracts and five peer-reviewed publications.) We identified six intervention types attempting to facilitate timely diagnosis of type 1 diabetes in the general practice setting: direct communication, indirect communication, education sessions, electronic clinical decision support tools, updated referral pathways and provision of glucose and/or ketone monitors. Due to the limited number of peer-reviewed studies identified by this review, we were not able to identify the extent to which these interventions were successful.

Conclusion: Paucity of information regarding study methodology and high heterogeneity among study design and outcome measures limited our conclusions regarding acceptability, effectiveness and reach. Future studies should include GPs in their design and consider the sustainability of interventions in the long term.

Prospero registration number: CRD42023412504.