{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T21:57:51Z","timestamp":1791583071963,"version":"4.3.4"},"reference-count":23,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Form Res"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Type 2 diabetes is a widespread chronic condition in which blood glucose and body weight management constitute essential therapeutic targets. Emerging technologies have the potential to aid complex therapeutic pharmacotherapy choices that are optimally tailored to individual needs.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n In this study, we developed and evaluated an AI model combining guidelines with clinical features and continuous glucose monitoring (CGM) to optimize therapeutic decision-making.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n \n Therapeutic guidelines were first encoded using a rule-based model and trained on a feed-forward neural network to predict the probability of therapeutic success for individual treatment recommendations. This approach relied on real-world evidence from a specialist diabetes outpatient clinic, using historical clinical data generated between 2009 and 2023. We used data from 533 patients with a diagnosis of type 2 diabetes and complete baseline data for weight and hemoglobin A\n 1c<\/jats:sub>\n within relevant therapy windows, resulting in a total of 853 treatment regimens. Transfer learning was used to optimize for glucose-lowering therapies that led to successful treatment outcomes, defined as an absolute 0.3% reduction in hemoglobin A\n 1c<\/jats:sub>\n (when it is over 6.5%) without weight gain in patients with a BMI over 28 kg\/m\n 2<\/jats:sup>\n . Recommendations that deviated from the guidelines were described using Shapley values and tested in digital twins for statistical significance. Four CGM-derived glucose-insulin response dynamic factors served as additional biomarkers.\n <\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n Dual glycemic and weight targets were achieved in actual clinical practice in 51.2% (131\/256) of cases, increasing to 54% (20\/37) when clinical guidelines were followed. After selecting outcomes in the test set that followed individualized recommendations, this increased further to 58% (21\/36) when using only phenotypic markers and to 65% (22\/34) when adding CGM-derived dynamic factors.<\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n Tested on the limited number of patients available, our findings show that our AI model was associated with improved retrospective outcomes compared to the guidelines in complex type 2 diabetes cases by integrating multiple data sources, drawing on experiential clinical insights, and selecting treatments most likely to meet each patient\u2019s clinical targets for glucose and weight control. Future research is needed with a larger dataset.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/92877","type":"journal-article","created":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T21:21:11Z","timestamp":1791580871000},"page":"e92877-e92877","source":"Crossref","is-referenced-by-count":0,"title":["Individualized Therapy Optimization for Type 2 Diabetes: Retrospective Study of Model Development, Internal Validation, and Physician Evaluation"],"prefix":"10.2196","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-5553-4790","authenticated-orcid":false,"given":"Arina","family":"Lozhkina","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-5553-3224","authenticated-orcid":false,"given":"Camillo Dario","family":"Piazza","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-8083-4343","authenticated-orcid":false,"given":"Zeina","family":"Gabr","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-2012-0372","authenticated-orcid":false,"given":"Maurice","family":"Rupp","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1028-9445","authenticated-orcid":false,"given":"David","family":"Herzig","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1993-7672","authenticated-orcid":false,"given":"Lia","family":"Bally","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2987-3932","authenticated-orcid":false,"given":"Andre","family":"Jaun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,10,9]]},"reference":[{"issue":"1 Suppl 1","key":"R1","doi-asserted-by":"publisher","first-page":"S1","DOI":"10.2337\/dc25-SINT","article-title":"Introduction and methodology: Standards of Care in Diabetes-2025","volume":"48","author":"American Diabetes Association Professional Practice Committee","journal-title":"Diabetes Care"},{"issue":"Suppl 1","key":"R2","doi-asserted-by":"publisher","first-page":"S145","DOI":"10.2337\/dc24-S008","article-title":"8. Obesity and weight management for the prevention and treatment of type 2 diabetes: Standards of Care in Diabetes-2024","volume":"47","author":"American Diabetes Association Professional Practice Committee","journal-title":"Diabetes Care"},{"issue":"7","key":"R3","doi-asserted-by":"publisher","first-page":"1617","DOI":"10.2337\/dci20-0022","article-title":"Precision medicine in diabetes: a consensus report from the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD)","volume":"43","author":"Chung","journal-title":"Diabetes Care"},{"issue":"1","key":"R4","doi-asserted-by":"publisher","DOI":"10.1136\/bmjmed-2022-000419","article-title":"Comparative effectiveness of second line glucose lowering drug treatments using real world data: emulation of a target trial","volume":"2","author":"Deng","journal-title":"BMJ Med"},{"issue":"1","key":"R5","doi-asserted-by":"publisher","DOI":"10.1038\/s41746-021-00394-8","article-title":"Predicting adverse outcomes due to diabetes complications with machine learning using administrative health data","volume":"4","author":"Ravaut","journal-title":"NPJ Digit Med"},{"key":"R6","doi-asserted-by":"publisher","DOI":"10.1016\/j.jbi.2022.104001","article-title":"Predicting pharmacotherapeutic outcomes for type 2 diabetes: an evaluation of three approaches to leveraging electronic health record data from multiple sources","volume":"129","author":"Tarumi","journal-title":"J Biomed Inform"},{"issue":"6","key":"R7","doi-asserted-by":"publisher","first-page":"415","DOI":"10.1016\/S2213-8587(19)30118-4","article-title":"Predicting individual treatment response in diabetes","volume":"7","author":"Heerspink","journal-title":"Lancet Diabetes Endocrinol"},{"issue":"10480","key":"R8","doi-asserted-by":"publisher","first-page":"701","DOI":"10.1016\/S0140-6736(24)02617-5","article-title":"A five-drug class model using routinely available clinical features to optimise prescribing in type 2 diabetes: a prediction model development and validation study","volume":"405","author":"Dennis","journal-title":"Lancet"},{"issue":"2","key":"R9","doi-asserted-by":"publisher","first-page":"376","DOI":"10.1038\/s41591-022-02120-7","article-title":"Patient stratification for determining optimal second-line and third-line therapy for type 2 diabetes: the TriMaster study","volume":"29","author":"Shields","journal-title":"Nat Med"},{"issue":"1","key":"R10","doi-asserted-by":"publisher","DOI":"10.1038\/s43856-026-01442-8","article-title":"A machine learning model for optimizing treatment of patients with poorly controlled type 2 diabetes","volume":"6","author":"Shi","journal-title":"Commun Med (Lond)"},{"issue":"7","key":"R11","doi-asserted-by":"publisher","first-page":"103086","DOI":"10.1016\/j.dsx.2024.103086","article-title":"Computational approaches for clinical, genomic and proteomic markers of response to glucagon-like peptide-1 therapy in type-2 diabetes mellitus: an exploratory analysis with machine learning algorithms","volume":"18","author":"Villikudathil","journal-title":"Diabetes Metab Syndr"},{"issue":"8","key":"R12","doi-asserted-by":"publisher","first-page":"3439","DOI":"10.1111\/dom.15689","article-title":"Integrated biomarker profiling for predicting the response of type 2 diabetes to metformin","volume":"26","author":"Long","journal-title":"Diabetes Obes Metab"},{"key":"R13","doi-asserted-by":"publisher","DOI":"10.2196\/67748","article-title":"Enhancing antidiabetic drug selection using transformers: machine-learning model development","volume":"13","author":"Kurasawa","journal-title":"JMIR Med Inform"},{"issue":"10","key":"R14","doi-asserted-by":"publisher","first-page":"5761","DOI":"10.1111\/dom.16630","article-title":"Exploring the value of ChatGPT in selecting antidiabetic agents for type 2 diabetes","volume":"27","author":"Jang","journal-title":"Diabetes Obes Metab"},{"issue":"7","key":"R15","doi-asserted-by":"publisher","first-page":"1666","DOI":"10.1111\/dom.14381","article-title":"Prototype of an evidence-based tool to aid individualized treatment for type 2 diabetes","volume":"23","author":"Buse","journal-title":"Diabetes Obes Metab"},{"issue":"8","key":"R16","doi-asserted-by":"publisher","first-page":"1241","DOI":"10.1007\/s13300-023-01431-3","article-title":"Expanding the role of continuous glucose monitoring in modern diabetes care beyond type 1 disease","volume":"14","author":"Klupa","journal-title":"Diabetes Ther"},{"issue":"10","key":"R17","doi-asserted-by":"publisher","first-page":"e3990","DOI":"10.1210\/clinem\/dgac459","article-title":"The importance of continuous glucose monitoring-derived metrics beyond HbA1c for optimal individualized glycemic control","volume":"107","author":"Yoshii","journal-title":"J Clin Endocrinol Metab"},{"issue":"6","key":"R18","doi-asserted-by":"publisher","first-page":"2082","DOI":"10.1111\/dom.15512","article-title":"Clinically relevant stratification of patients with type 2 diabetes by using continuous glucose monitoring data","volume":"26","author":"Shao","journal-title":"Diabetes Obes Metab"},{"key":"R19","doi-asserted-by":"crossref","unstructured":"Metwally AA Perelman D Park H et al. Predicting type 2 diabetes metabolic phenotypes using continuous glucose monitoring and a machine learning framework. medRxiv. Sep 9, 2024:2024.07.20.24310737. [doi: 10.1101\/2024.07.20.24310737] [Medline: 39108516]","DOI":"10.1101\/2024.07.20.24310737"},{"issue":"2_suppl","key":"R20","volume":"27","author":"Lozhkina","journal-title":"Diabetes Technol Ther"},{"key":"R21","unstructured":"Abadi M Agarwal A Barham P et al. TensorFlow: large-scale machine learning on heterogeneous distributed systems. TensorFlow. 2015. URL: https:\/\/www.tensorflow.org\/extras\/tensorflow-whitepaper2015.pdf [Accessed 10-09-2026]"},{"key":"R22","unstructured":"Shapley LS . Notes on the n-person game \u2014 II: the value of an n-person game. RAND Corporation; 1951. URL: https:\/\/www.rand.org\/pubs\/research_memoranda\/RM0670.html [Accessed 10-09-2026]"},{"key":"R23","unstructured":"Lundberg SM Lee SI . A unified approach to interpreting model predictions. In: NIPS\u201917: Proceedings of the 31st International Conference on Neural Information Processing Systems. Curran Associates Inc; 2017:4768-4777. [doi: 10.5555\/3295222.3295230]"}],"container-title":["JMIR Formative Research"],"original-title":[],"language":"en","deposited":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T21:21:18Z","timestamp":1791580878000},"score":1,"resource":{"primary":{"URL":"https:\/\/formative.jmir.org\/2026\/1\/e92877"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,9]]},"references-count":23,"URL":"https:\/\/doi.org\/10.2196\/92877","relation":{},"ISSN":["2561-326X"],"issn-type":[{"value":"2561-326X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,9]]},"article-number":"v10i1 Suppl 1e92877"}}