{"doi":"10.1111/dom.70170","title":"Postoperative insulin requirements in surgical patients receiving fully automated insulin delivery in the hospital","abstract":"Surgery-induced metabolic stress, immobilisation and adjunctive therapies (e.g., glucocorticoids, nutritional support) cause dynamic changes in insulin requirements.1, 2 A mismatch between prescribed and actual insulin needs predisposes patients to dysglycaemia, which is associated with prolonged hospitalisation and adverse outcomes.3-5 Fully automated insulin delivery (AID) systems continuously titrate insulin to maintain target glucose levels, offering both effective perioperative support for complex patients6, 7 and insights into insulin requirements. This study examined perioperative insulin delivery in patients with fully AID and identified determinants of insulin needs to inform postoperative glycaemic strategies. Data from two randomised controlled trials (NCT05392452 and NCT04361799) investigating the glycaemic efficacy of fully AID in hospitalised surgical patients with insulin-requiring non-type 1 diabetes were pooled and retrospectively analysed. Both trials had ethics approval (2020-01024, 2022-D0034) and all participants provided written informed consent. Eligible adults underwent elective surgery, were expected to require insulin, and remained hospitalised ≥72 h postoperatively; type 1 diabetes was excluded. The AID system comprised the Dexcom G6 subcutaneous continuous glucose monitoring (Dexcom, USA), an android smartphone hosting the CamAPS HX application with a model predictive control algorithm (University of Cambridge, Cambridge, UK), and a subcutaneous insulin pump—either YpsoPump (Ypsomed AG, Burgdorf, Switzerland) or the DANA RS insulin pump (Diabecare, Sooil, Seoul, South Korea), delivering fast-acting insulin aspart (Fiasp, Novo Nordisk). AID was initialised using participants' body weight and estimated total daily insulin dose (TDD), with a glucose target of 5.8 mmol/L. Continuous insulin delivery and glucose data were retrieved from a cloud-based diabetes management platform (Glooko, Inc., Palo Alto, California, USA). Clinical data, including nutrition support type (parenteral, enteral or both) and glucocorticoid use (documented daily as a binary variable), were extracted from electronic health records. For a subset of patients, daily carbohydrate (CHO) intake (oral plus nutritional support, calculated over 24-h periods [00:00 to 23:59]) was obtained from the hospital's electronic meal management software. TDD and weight-normalised TDD (TDD/kg) were calculated for each patient, from the first postoperative day (00:00) until the 10th postoperative day or discharge, excluding the day of surgery and discharge. Day-to-day variability was assessed by the coefficient of variation (CV) of TDD, calculated for each patient. Changes in TDD were analysed with a linear mixed-effects model. Predictor significance was tested with Type III Wald chi-square tests, and model assumptions were verified graphically. Linear mixed-effects models assessed effects of postoperative day, nutrition support (parenteral vs. enteral vs. concomitantly administered parenteral and enteral) and glucocorticoid therapy on TDD/kg, adjusted for surgery type, glycated haemoglobin (HbA1c), sex, age and body mass index (BMI). Postoperative day was included as a fixed effect and participant as a random intercept to account for within-subject correlation. Non-significant terms were removed to reduce model complexity and improve interpretability. For patients with CHO data, a quadratic term accounted for non-linear effects. Results from the fixed-effects estimates are presented with 95% confidence intervals [95% CI], p-values, and using Forest plots, while non-linear effects of CHO supply on TDD/kg are visualised graphically. Descriptive statistics are presented as mean ± standard deviation (SD) for normally distributed variables or median [25th; 75th percentile] for non-normally distributed variables unless specified otherwise. Analyses were performed using R (4.4.1). Thirty-six patients (277 patient-days, mean age 67.8 ± 11 years, 39% female,","journal":"Diabetes Obesity and Metabolism","year":2025,"id":576986,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":0,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9517,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2025-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1485765,"name":"Clara Escorihuela‐Altaba","orcid":"0009-0003-1680-920X","position":1,"is_corresponding":false},{"id":1041503,"name":"Michael S. Hughes","orcid":"0000-0002-3729-0350","position":2,"is_corresponding":false},{"id":766734,"name":"Christos T. Nakas","orcid":"0000-0003-4155-722X","position":3,"is_corresponding":false},{"id":1485766,"name":"David Herzig","orcid":"0000-0003-1028-9445","position":4,"is_corresponding":false},{"id":49193,"name":"Lia Bally","orcid":"0000-0003-1993-7672","position":5,"is_corresponding":false},{"id":1485764,"name":"Gabija Krutkyte","orcid":"0009-0004-5987-4615","position":0,"is_corresponding":true}],"reference_count":10,"raw_metadata":null,"created_at":"2026-07-19T02:58:00.620755Z","pmid":"41048189","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}