JMIR Publications

ABSTRACT

Background:

Ulcerative colitis (UC) is a chronic inflammatory disease characterized by considerable variability in symptoms and severity. Prognostic prediction models are integral to precision medicine, facilitating personalized treatment strategies tailored to individual patient characteristics. Despite advancements in model development, their quality and clinical applicability remain insufficiently evaluated.

Objective:

This study systematically reviewed the development, performance, and clinical utility of prognostic prediction models for UC.

Methods:

To identify prognostic models for UC, a comprehensive search was conducted in PubMed, Embase, the Cochrane Library, Web of Science, SinoMed, China National Knowledge Infrastructure, Wanfang, and VIP Database up to November 2, 2024. Extracted data included study characteristics, model development methods, validation metrics (e.g., area under the curve [AUC], C-index). The risk of bias and applicability was evaluated using the Prediction Model Risk of Bias Assessment Tool (PROBAST). A meta-analysis was conducted to assess model performance.

Results:

A total of 30 studies involving 7,816 UC patients were included, with the largest numbers conducted in China (n=11) and Japan (n=4). Most studies were retrospective (n=22), with 67% being multi-center studies. The primary objectives of the UC prognostic models included predicting therapeutic effect and response to treatment, particularly to TNF-α inhibitors (e.g., infliximab, adalimumab), and assessing risks of surgery, disease progression, or relapse. Logistic regression was the most frequently used method for both predictor selection (n=6) and model construction (n=12). Common predictors included age, C-reactive protein, albumin, hemoglobin, disease extent, and Mayo scores. The meta-analysis yielded a pooled AUC of 0.86 (95% confidence intervals [CI]: 0.80-0.92). Most studies exhibited a high risk of bias (n=29), particularly in participant selection and statistical analysis. Applicability concerns were identified in 18 studies, primarily due to subgroup-specific designs that limited the generalizability of the findings. External validation data (n=16) were limited, and a limited number of studies (n=14) included calibration curves or decision curve analyses.

Conclusions:

This study demonstrates that prognostic models for UC have some potential in predictive performance and clinical application. However, most models are constrained by high bias risk, insufficient external validation, and limited generalizability due to small sample sizes and subgroup-specific designs. Future research should prioritize multi-center validations, refine model development approaches, and enhance model applicability to support broader clinical implementation. Clinical Trial: This study was registered in the International Prospective Register of Systematic Reviews under registration number CRD42024609424.