Clinica chimica acta; international journal of clinical chemistry · 2026
RefInterval studio: development of a guided R shiny workflow for establishing discrete and age-continuous reference intervals.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Chaochao Ma, Haoyi Zhou, Xinqi Cheng, Shi Chen, Haiyu Pang, Tengda Xu, Danchen Wang, Lian Hou, Ling Qiu
Аннотация
BACKGROUND: Establishing reference intervals (RIs) using a direct approach requires multiple analytical procedures, including participant-level data preparation, reference population screening, partition assessment, and statistical modelling. Existing software tools often focus on individual analytical components and may require substantial programming expertise. We developed RefInterval Studio, an interactive R Shiny application primarily designed to provide a standardized and reproducible workflow for guideline-based direct RI studies, including both discrete and age-continuous RI estimation.
METHODS: RefInterval Studio was developed as a modular web-based application integrating data input and cleaning, reference-population screening, baseline characterization, partition analysis, discrete RI estimation, and age-continuous modelling using generalized additive models for location, scale and shape (GAMLSS). The application supports parametric, non-parametric, and robust RI estimation methods and incorporates automated reporting and visualization functions. Its functionality, operational stability, and workflow feasibility were evaluated using simulated datasets containing common data-quality issues, sensitivity analyses under varying data conditions, and real-world health-examination datasets from Peking Union Medical College Hospital.
RESULTS: RefInterval Studio completed the tested analytical workflows without runtime errors under the evaluated data conditions. Simulated-data testing demonstrated successful handling of duplicated records, missing values, invalid entries, censoring symbols, and extreme observations. Sensitivity analyses showed that the application remained operational under varying levels of missingness, duplication, class imbalance, outlier frequency, sample size, and distributional skewness. Evaluation using real-world health-examination datasets for thyroid-stimulating hormone, urea, and total protein demonstrated compatibility with realistic laboratory-data structures and successful implementation of participant-level screening, descriptive characterization, partition analysis, discrete RI estimation, and age-continuous percentile modelling.
CONCLUSIONS: RefInterval Studio provides an integrated, transparent, and user-friendly analytical environment primarily designed to support guideline-based direct RI workflows. By combining participant-level data preprocessing, reference-population.
METHODS: RefInterval Studio was developed as a modular web-based application integrating data input and cleaning, reference-population screening, baseline characterization, partition analysis, discrete RI estimation, and age-continuous modelling using generalized additive models for location, scale and shape (GAMLSS). The application supports parametric, non-parametric, and robust RI estimation methods and incorporates automated reporting and visualization functions. Its functionality, operational stability, and workflow feasibility were evaluated using simulated datasets containing common data-quality issues, sensitivity analyses under varying data conditions, and real-world health-examination datasets from Peking Union Medical College Hospital.
RESULTS: RefInterval Studio completed the tested analytical workflows without runtime errors under the evaluated data conditions. Simulated-data testing demonstrated successful handling of duplicated records, missing values, invalid entries, censoring symbols, and extreme observations. Sensitivity analyses showed that the application remained operational under varying levels of missingness, duplication, class imbalance, outlier frequency, sample size, and distributional skewness. Evaluation using real-world health-examination datasets for thyroid-stimulating hormone, urea, and total protein demonstrated compatibility with realistic laboratory-data structures and successful implementation of participant-level screening, descriptive characterization, partition analysis, discrete RI estimation, and age-continuous percentile modelling.
CONCLUSIONS: RefInterval Studio provides an integrated, transparent, and user-friendly analytical environment primarily designed to support guideline-based direct RI workflows. By combining participant-level data preprocessing, reference-population.