Diabetic Ketoacidosis: Systematic Review and Meta-Analysis

Diabetic ketoacidosis (DKA) remains a major cause of diabetes-related morbidity and mortality worldwide, yet evidence on its management and clinical outcomes across healthcare settings remains limited. We conducted a systematic review and meta-analysis to evaluate the precipitants, management and outcomes of adult DKA.

We searched PubMed, Embase, Web of Science, Scopus, MEDLINE, CINAHL and ERIC from 2000 to April 2026 for studies reporting precipitating factors, management and outcomes among adults (≥ 16 years) with DKA. Articles that did not report all three were excluded. Two reviewers independently screened studies, extracted data and assessed risk of bias. Descriptive synthesis was performed for demographic, biochemical, management and outcome variables. A random-effects meta-analysis estimated summary in-hospital all-cause mortality. This review was registered with PROSPERO (ID: CRD42025641665). https://www.crd.york.ac.uk/PROSPERO/view/CRD42025641665.

Nineteen studies met the inclusion criteria, of which 15 studies comprising 1278 DKA episodes contributed to the descriptive synthesis and nine studies including 724 participants to the mortality meta-analysis. Most episodes were classified as moderate or severe DKA. Preadmission glycaemic control was frequently suboptimal. Missed or inadequate diabetes medication was the commonest precipitant (42.2%), followed by infection (28.6%) and newly diagnosed diabetes (11.2%). Substantial regional variation was observed in guideline use, insulin and fluid administration, intensive care admission and treatment-related complications. The in-hospital mortality summary estimate was 2.17% (95% CI, 0.93%-4.99%).

Medication omission and infection remain the dominant precipitants of adult DKA worldwide. Marked variation in management practices and outcomes persists across healthcare systems, suggesting inconsistent implementation of DKA care pathways.