JMIR Medical Informatics

Refining Our Focus by Bridging Theory and Practice for Real-World Impact: An Updated Scope for JMIR Medical Informatics

2026-09-04T16:45:11-04:00

To align with contemporary trends and place greater focus on persistent and novel challenges in the field, JMIR Medical Informatics updated its focus and scope. Each submitted work will be evaluated against a stronger criterion for translational impact. Additionally, submissions on artificial intelligence (AI) methods and applications have more explicit standards. In this editorial, we explain further what changed, why, and what it means for prospective authors, reviewers, and readers.

Developing Country-Specific Charlson Comorbidity Index Mappings for Use With German Administrative Data: Methodological Comparative Study

2026-09-03T16:00:19-04:00

Background: The Charlson Comorbidity Index (CCI) is widely used to quantify comorbidity burden in observational research. Applying it to real-world data requires accurate International Classification of Diseases–based mappings. The commonly used mapping does not reflect current German coding standards. Objective: This study aimed to develop year-specific mappings (2004 to 2026) based on the German modification of the International Classification of Diseases, 10th Revision (ICD-10-GM), for the CCI with an open-source implementation and compare them with the commonly used mapping. Methods: For each year, mappings were curated via independent dual review with consensus. In 515,827 inpatient cases from a German tertiary center (2010-2024), year-appropriate mappings and the commonly used mapping were applied, and the results were pooled into 2 corresponding datasets that were subsequently compared. Agreement was assessed using the intraclass correlation coefficient (2-way mixed-effects model for absolute agreement based on single measurement), and the significance of the discordances was measured using the McNemar test. Results: Agreement was very high; 3.1% (16,136/515,827) of cases differed. A single code was present in 48.9% (7896/16,136) of the discordant cases and was intentionally excluded in the new mappings as it was not diagnostic for the category it was assigned to. Discordant results were found in 58.8% (10/17) of the categories (“peripheral vascular disease,” “mild liver disease,” and “severe liver disease”) being present in, respectively, 51.7% (8339/16,136), 17.4% (2813/16,136), and 16.4% (2657/16,136) of the discordant cases. After filtering the dataset to only include observations with any diagnosis of a malignant disease, 6.9% (9377/136,607) of cases differed, and 79.3% (7434/9377) of those differences were caused by the “peripheral vascular disease” category, suggesting significant discrepancies for specific research questions. Conclusions: We provide transparent, year-specific ICD-10-GM (2004-2026) mappings and an open-source tool for CCI calculation. The mappings align with current German coding practice and enable reproducible, year-appropriate comorbidity adjustment using administrative data.

Markov Decision Process–Based Personalized Follow-Up Planning for Type 2 Diabetes: Retrospective Cohort Study

2026-09-02T17:45:17-04:00

Background: Personalized follow-up for type 2 diabetes may improve the alignment between monitoring intensity and patient needs, but operational approaches that jointly consider follow-up timing, modality, expected health benefits, and resource use remain limited. Objective: This study develops a Markov decision process (MDP) framework for personalized follow-up planning, externally validates complementary risk-prediction models, and estimates the projected 12-month cost-effectiveness of model-generated follow-up strategies. Methods: We retrospectively analyzed longitudinal data from 41,398 patients with type 2 diabetes managed in 10 community health centers in Nanjing, China, over the 2015-2024 calendar period. An independent Shanghai cohort included 25,506 patients from 10 communities. We developed least absolute shrinkage and selection operator (LASSO)-Cox models to predict 1-year incident complication and mortality risk and evaluated discrimination in Shanghai. Separately, a 12-cycle finite-horizon MDP used observed action-conditional state transitions with prespecified utility, cost, access, and willingness-to-pay parameters to generate personalized follow-up policies. Results: For an example patient initially without recorded complications (S0), the personalized policy increased annual effectiveness by 0.48 quality-adjusted life days (QALDs; 0.0013 quality-adjusted life years [QALYs]) and cost by 33.46 CNY (1 CNY=US $0.15), yielding an incremental cost-effectiveness ratio (ICER) of 25,499 CNY/QALY. In a heterogeneous simulated cohort of 1000 patients initialized in S0, mean annual effectiveness increased from 313.52 to 314.05 QALDs (0.85896 to 0.86041 QALYs), and mean annual cost increased by 24.67 CNY, yielding an ICER of 17,016 CNY/QALY. External C-indices were 0.778 for incident complications and 0.812 for mortality. Deterministic sensitivity analyses did not materially alter the cost-effectiveness conclusion. Conclusions: The MDP framework generated individualized 12-month follow-up policies with small projected QALY gains at modest incremental program cost, while the complementary risk models demonstrated external discrimination in an independent cohort. As action-conditional transitions were estimated from observational records and several economic parameters were prespecified, the modeled differences represent projections rather than causal treatment effects and require prospective implementation and economic validation before clinical adoption.

Automatic Kidney Image Segmentation During Robot-Assisted Partial Nephrectomy Using a Deep Learning Model Based on a Multiannotator Dataset: Model Development and Validation Study

2026-09-02T13:00:03-04:00

Background: Augmented reality (AR) has emerged as a promising tool to enhance surgical precision during robot-assisted partial nephrectomy (RAPN), particularly by enabling the intraoperative overlay of 3D anatomical models. However, real-time AR implementation requires robust image segmentation of anatomical structures, such as the kidney, which remains technically challenging in dynamic laparoscopic environments. Objective: This study aimed to develop and validate a large, annotated image dataset to train deep learning models for automated segmentation of the renal parenchyma during RAPN, as a prerequisite for real-time AR guidance. Methods: We conducted a single-center, observational image annotation study using prospectively collected surgical videos from 131 RAPN procedures performed between 2022 and 2024. Patients had localized renal tumors, with 11 presenting with multifocal disease (160 tumors in total). A total of 48,000 frames were extracted based on image sharpness, diversity, and the exclusion of artifacts. A subset of 454 images was annotated by 9 contributors (surgeons, engineers, and nonexperts) after structured training. Interannotator agreement was assessed using the Dice similarity coefficient (DSC) and sensitivity against an expert reference. A convolutional neural network (AlbuNet-34) was trained using 12,546 annotated images and evaluated on a validation set of 3137 images. Model performance was analyzed overall and across surgical phases. Results: Annotators achieved high agreement, with median DSC values ranging from 0.91 to 0.95 and sensitivity consistently more than 0.89. The deep learning model achieved a mean DSC of 0.75 (SD 0.23) and a sensitivity of 0.71 (SD 0.24) on the validation set. Segmentation accuracy varied significantly across surgical phases, with lower performance observed during tumor resection and tumor bed reconstruction due to increased visual complexity. Conclusions: This study demonstrates the feasibility of automated renal parenchyma segmentation using deep learning in real-world intraoperative settings. Although current performance remains below that of expert-level annotations, the creation of a large, annotated dataset and the implementation of a structured multiannotator workflow represent key milestones toward reliable AR-assisted surgery. Ongoing refinements in annotation quality, dataset diversity, and neural network optimization are expected to enhance future real-time AR applications in urology.

A Health Informatics System in the South Australian Public Health Network: Implementation Report

2026-09-01T14:00:17-04:00

Background: Following 2 decades of electronic medical record (EMR) adoption, most large public health systems hold comprehensive digital clinical data but lack the complementary informatics capability to return those data to clinicians, coders, and operational teams in a usable form. In South Australia (SA), the statewide Sunrise (Altera Digital Health) EMR system has digitized documentation and ordering since 2017; however, tools for back-end data extraction and enriched clinical information displays were not prioritized, and clinical, operational, and research users have reported ongoing difficulties accessing timely data. Objective: The aim of this study is to describe the development, governance, and deployment of a cloud-native health informatics system (HIS) in the Central Adelaide Local Health Network (CALHN), which serves approximately 40% of SA public patients, and to report implementation outcomes in accordance with the iCHECK-DH (Guidelines and Checklist for the Reporting on Digital Health Implementations) guidelines. Methods: The HIS comprises 6 architectural layers deployed on Microsoft Azure with Red Hat OpenShift (IBM), extracting Sunrise EMR data in near real time under a read-only model in which all patient data remain within the SA Health network at all times. Implementation proceeded through 5 overlapping phases (2021-2026), governed across 4 domains: technical (security impact assessment, information asset classification, and independent cybersecurity review), clinical (a clinical governance committee that has met quarterly since May 2022), corporate (incorporation of HeartAI Pty Ltd in 2022, with conflicts of interest declared and managed under SA Health policy), and ethical (human research ethics committee [HREC], reference 18079 with subsequent amendments). Development was funded through approximately Aus $1.84 million (Aus $1=US $0.72 as of August 24, 2026) in competitive and institutional grants under a public-private model with CALHN and AusHealth. Implementation (Results): Three applications have reached deliberately different stages of maturity. The CALHN Critical Care Informatics System (CCCIS) has been implemented and evaluated: it has been deployed across the 2 CALHN intensive care units (ICUs; 54 beds, approximately 5000 admissions per year) since 2022, is in daily clinical use, and automates the submission of 115 variables to the Australian and New Zealand Intensive Care Society Centre for Outcome and Resource Evaluation (ANZICS CORE) registry. In a formal evaluation by 8 senior intensivists, its mean System Usability Scale score was 76 of 100, and the overall workload was low (NASA Task Load Index: 21/100). CODEXA, an AI-assisted clinical coding application, is in operational validation: models trained on 500,000 episodes and tested on 50,000 held-out episodes achieved a pooled -score of 71% across the full case mix and 57.86% (precision 64.72%; recall 55.53%) on complex acute episodes, in line with published benchmarks of 58% to 61% obtained from curated research datasets. The Patient Flow application is in co-design with CALHN’s Network Operations Centre, with interface prototype evaluations completed in July 2026. In May 2026, the platform received unconditional endorsement from Digital Health SA’s Technical Design Review Committee, the highest technical governance approval in SA Health, concluding a 5-year governance pathway. Conclusions: A locally developed informatics platform can be governed, deployed, and validated in a public health system; demonstrating clinical outcomes and transitioning to operational status are the next priorities.

Predicting Call Abandonment in a Health Care Call Center Using Nonpersonal Operational Data: Machine Learning Study

2026-08-31T13:00:03-04:00

Background: Call abandonment is a critical barrier to patient access in health care call centers; yet, predictive modeling efforts are limited by strict privacy regulations that restrict the use of personal or behavioral data. Whether abandonment can be accurately predicted using only anonymized operational metrics remains unclear. Objective: This study evaluated the feasibility, performance, and operational use of machine learning models trained exclusively on nonpersonal, routinely collected call center metrics to predict call abandonment across distinct organizational phases. Methods: We analyzed 1,037,363 call records from a large academic health care system spanning 4 operational periods marked by workflow changes and skill consolidation. Features included temporal variables, skill identifiers, and rolling operational metrics (in-queue time, occupancy, handle time, after-call work time, and active agents). Random forest and CatBoost models were trained on 3 phases defined as: (T1 [January-April 2023; original workflows], T2 [May-August 2023; post skill consolidation, cross-training, and new workflows], and T3a [September-December 2023; optimized processes]) using 5-fold cross-validation with 3 imbalance-handling strategies (none, class weighting, and synthetic minority oversampling technique). Temporal generalizability was assessed by evaluating all models on all 4 phases, with T3b serving as an unseen holdout set. Performance was evaluated using area under the curve (AUC), precision-recall area under the curve (PR-AUC), Brier score, and calibration error. Shapley additive explanations (SHAP) values quantified feature contributions. Results: Across all training phases and algorithms, adding operational metrics improved the area under the receiver operating characteristic curve by 0.03-0.13 vs models using only temporal and skill features. The best configuration was a CatBoost model trained on T2 with operational metrics and no imbalance correction (T3b AUC 0.767, PR-AUC 0.068, expected calibration error 0.006, and Brier 0.027). Models trained solely on preintervention data (T1) generalized poorly to postintervention periods when restricted to temporal and skill features (AUC 0.32-0.38) but achieved AUC 0.715 on T3b when operational metrics were included. SHAP analysis consistently identified in-queue time as the dominant predictor, with the number of logged-in agents, hour-of-day, and skill identifiers comprising the remaining top features. Abandonment declined from 8.7% (20,809/238,722) in T1 to 2.8% (8293/300,060) in T3b; model-based analyses of temporal features (day of week and hour of day) showed the highest risk on Mondays and between 11:00 and 16:00. Skill-level analyses showed marked improvement in high-volume imaging teams with high abandonment. Conclusions: Models trained on nonpersonal operational data predicted call abandonment on the T3b temporal holdout, with real-time queue and staffing metrics providing the dominant signal. These findings support the feasibility of interpretable, operationally grounded abandonment prediction in health care call centers and indicate that models should be retrained after major operational changes. Trial Registration:

Point-of-Care Ultrasound Integrated With Teleconsultation for Rural Home-Based Medical Care: Pilot Implementation and Financial Analysis

2026-08-28T16:01:08-04:00

Background: Population aging and geographic health disparities challenge health care delivery in rural settings worldwide. Point-of-care ultrasound (PoCUS) combined with teleconsultation may enhance home-based medical care by extending specialist expertise to underserved communities, yet evidence from real-world implementation in Asian settings remains scarce. Taiwan, with its high digital literacy and universal National Health Insurance (NHI) system, provides a unique context for evaluating integrated PoCUS-teleconsultation service delivery. Objective: This study aimed to describe the implementation, clinical applications, and financial sustainability of an integrated PoCUS-teleconsultation service within a dedicated home-based medical care practice in rural eastern Taiwan. Methods: We conducted a retrospective analysis of prospectively recorded routine care data at D Clinic, Taitung County, from April 2020 to May 2022. PoCUS examinations conducted in the clinic, during home visits, and through mobile outreach were recorded. A business-to-business-to-consumer (B2B2C) teleconsultation model linking on-site physicians with remote specialists via real-time ultrasound streaming was implemented from March 2021. Seven-year net present value (NPV) analyses evaluated financial viability across 24 equipment reimbursement scenarios. Results: Among the 26 patients who received teleconsultation, the mean age was 79.4 (SD 9.2) years, 15 (58%) were male, and 20 (77%) were in the intensive home health care tier. A total of 676 PoCUS examinations (defined as single-region imaging studies) were performed by 3 trained physicians in the following settings: 375 (55.5%) in the clinic, 174 (25.7%) during home visits, and 127 (18.8%) as mobile outreach. Twenty-six teleconsultation sessions were conducted (n=15, 58% during home visits; n=7, 27% during mobile outreach; and n=4, 15% in the clinic), with 13 (50%) involving multitarget scanning (cardiac, pulmonary, and abdominal). Of the 26 sessions, 24 (92%) were completed without technical interruption, while 2 (8%) experienced transient 4G buffering but were completed. Monthly PoCUS volume reached a sustained plateau of ≥28 examinations from month 17 of implementation. Four clinical service domains were identified: hospital-at-home acute management, home-based hospice multiorgan support, green channel surgical referral, and chronic disease serial monitoring. NPV analysis demonstrated that wireless PoCUS equipment with the proposed, but not yet implemented, dual-specialist teleconsultation reimbursement achieved the most favorable projected return (NT $882,000; NT $1=US $0.036 in 2021 on average), requiring only 2.72 monthly cases to break even under the modeled assumptions. Conclusions: Integrating PoCUS with teleconsultation is operationally feasible in rural home-based care when wireless equipment is adopted. Financial scenario modeling suggests favorable projected returns under proposed reimbursement structures, although these remain contingent on NHI policy adoption. These findings are hypothesis generating and require validation through multisite studies with patient-level outcome measurement before informing national reimbursement policy.

A Conceptual Model for Ambient AI Adoption: Perspectives From Academia and Industry

2026-08-28T14:30:13-04:00

Ambient AI technologies are increasingly marketed as solutions to reduce clinician burden and improve care efficiency; however, real-world performance varies widely across clinical settings. Health care provider organizations face challenges in determining which aspects of ambient AI performance matter most and how to obtain meaningful information about those aspects from vendors or through internal evaluation. This article presents a shared mental model to guide health system leaders in conceptualizing ambient AI performance across 3 interdependent dimensions: technical, interface, and system level. For each dimension, we outline the types of information relevant to assessment; what vendors should reasonably be expected to provide; and how health care provider organizations can conduct their own evaluations to contextualize, verify, or supplement vendor claims. By integrating both vendor and health system perspectives, this work offers a grounded, practical structure to support organizations of all sizes in understanding and making informed decisions about ambient AI technologies.

Relevance of the uMap Collaborative Platform as Support for Choropleth Mapping of a Traffic-Light Statistical Signal Atlas of All-Cause Mortality During the First French Lockdown: Geospatial Analysis

2026-08-27T16:45:11-04:00

Background: The growing need for and interest in geomatics in the medical sector, as well as the pandemic crisis, led us to create a France-wide geomatics project aimed at producing several atlases of all-cause mortality at the municipal and submunicipal district levels via uMap France, a free and open-source collaborative map-sharing platform. In 2020, we decided to circumvent the obstacle of accessing detailed COVID-19 data by adopting a mortality-based approach to map the consequences of the crisis. Objective: The aim of the uMap study was to provide a webmapping platform with original visualization and knowledge as well as decision-making aids that complement existing information and are relevant to public and health care professionals. Our main hypotheses were as follows: (1) The medical sector could develop a private uMap platform dedicated to health; (2) interest in a municipal mortality atlas for France linked to the pandemic crisis will increase, even if it is produced after the pandemic; and (3) sharing the atlases with the uMap community will enhance their appeal and inspire the creation of similar atlases, owing to the new “experimental choropleth layer” recently developed by the uMap team. Methods: This approach focused on 3 main parts—data management (data collection, cleansing, and scheduling) and geomatic engineering through a 2-step geomatic action plan to create atlases of the first lockdown period in France displayed on the uMap platform. A logarithmically transformed variable allowed us to obtain an immediate statistical signal of excess mortality or submortality via the Traffic-Light Atlas. Results: The uMap Traffic-Light Atlas display provides instant statistical signals at a glance, owing to the semantic interplay of colors. The atlas’s double legends make it easy to compare specific regions (northeast, northwest, southeast, and southwest) with all of France. The atlas revealed excess mortality in 41.7% (14,503/34,833) of the municipalities. Of the municipalities, 35% (12,198/34,833) were in the green class (close to average to 2 times the average), 5% (1724/34,833) were in the orange class (2‐4 times the average), and 1.7% (581/34,833) were in the red class (4‐11 times higher than average). Conclusions: We innovated, enriched, and reinforced the value of uMap for visual rendering by instantiating colored choropleth map atlases and double legends and showed its relevance to the health care sector. We focused on the Traffic-Light Atlas, which is the most relevant aspect because of the instant message it conveys and its interpretability for all audiences. The uMap community can share our all-cause mortality atlases. A second version of the atlas encompassing 4 periods in 2020 and containing a minor error will be updated using either the “experimental choropleth layer” feature recently developed by the uMap team or, if this feature proves insufficient, the geomatic optimization process via the R project.

Bayesian Analysis of AI-Driven Cost Savings in UK and Australian Health Care Systems: Cross-Sector Implementation Study

2026-08-26T14:00:20-04:00

Background: Health care systems face growing fiscal pressure while AI reaches clinical parity in several domains. UK National Health Service expenditure rose by 52%, while Australia's health expenditure grew by 29% between 2019 and 2023. Yet large-scale cost savings from AI remain limited, largely because implementation constraints continue to outweigh technical capability. Objective: This study develops a Bayesian budget-impact framework to estimate AI-driven gross cost savings in radiology, workflow optimization, and workforce optimization in the United Kingdom and Australia, explicitly accounting for adoption uncertainty, effectiveness, and implementation risk. Methods: We used a sequential Monte Carlo simulation with 1000 particles to estimate annual gross cost savings. The core savings function combined expenditure base, sector weight, adoption, effectiveness, and implementation risk. Priors were informed by a structured review of multiple studies per domain. The savings likelihood used a heteroscedastic noise specification in which the SD followed an exponential prior, with the mean set to 15% of each sector’s observed savings estimate, ranging from US $12.0 million for Australian radiology to US $87.2 million for UK workforce optimization. The likelihood was also augmented with sector-specific observations that anchored effectiveness to published cost-reduction estimates and implementation risk to observed deployment failure rates. Scenario analysis applied multipliers to 1000 bootstrap posterior draws across optimistic, conservative, and pessimistic settings. Sensitivity analyses varied the σ scaling factor from 0.10 to 0.20 and perturbed prior means for adoption, effectiveness, and implementation risk by ±20%. Results: Posterior annual savings were US $949 million (95% credible interval [CrI] US $720.6-US $1173.5 million) for the United Kingdom and US $737 million (95% CrI US $526.1-US $953.6 million) for Australia. Workforce optimization generated the largest share of savings in both countries, contributing 62.4% of UK savings (US $591.9 million, 95% CrI US $398.8-US $854.8 million) and 76.2% of Australian savings (US $561.4 million, 95% CrI US $331.9-US $815.3 million). Posterior implementation risk estimates ranged from 35.5% to 47.4%, below prior means of 49.1% to 57.2%, reflecting the empirical anchoring introduced through deployment-failure data. Across scenarios, projected savings ranged from US $357 million to US $1.845 billion in the UK and from US $267.2 million to US $1.454 billion in Australia. Baseline cumulative projections for 2024-2030 were US $10.1 billion for the United Kingdom and US $8.0 billion for Australia. The sensitivity analyses confirmed the robustness of the posterior savings estimates. Conclusions: AI could generate substantial expenditure reductions in both health systems, but implementation risk remains the main constraint on realizing those gains. Workforce savings should be interpreted primarily as capacity gains that can be redeployed to higher-value care, not as automatic cash savings. The Bayesian framework offers probabilistic planning ranges rather than point forecasts and provides a practical basis for policy planning under uncertainty.