Radiologists develop calculator for predicting imaging no-shows
Radiologists have developed a calculator that can predict whether patients will fail to show up for their imaging appointment—a problem that costs their institution an estimated $58 million annually.
Outpatient radiology no-shows can delay timely diagnoses and exacerbate healthcare disparities, resulting in significant operational and financial burden, experts wrote Aug. 27 in JACR. Machine learning models have previously shown promise in helping to predict these circumstances, with socioeconomic data potentially making them even stronger.
Researchers with the University Hospitals Cleveland recently sought to test this theory, building their own Radiology No-Show Calculator enriched by social determinants of health data. Their work appears to be paying off, with the potential to duplicate their AI program at other institutions.
“The calculator demonstrates moderate discrimination for predicting no-shows and enables prospective targeted outreach,” corresponding author Wyatt Anderson, MD, with University Hospitals’ Department of Radiology, and colleagues concluded. “Future deployment of this model holds promise for improving health outcomes among socioeconomically vulnerable patients while lowering institutional cost burden.”
For the study, researchers analyzed data from nearly 33,000 outpatient radiology appointments corresponding to 10,000 patients—half of whom had at least one no-show between 2023 and 2024. They built an AI model utilizing 16 predictors including area deprivation index, a tool that measures socioeconomic disadvantage for a geographic region.
The AI model achieved an area under the receiver operating characteristic of 0.76—indicating acceptable predictive power with a 76% chance of correct ranking. No-shows were strongly associated with socioeconomic disadvantage, with rates rising from 15% in the least deprived neighborhoods to 32% in the most deprived. Targeting the top 10% of highest-risk appointments would capture about 26% of all no-shows, with a “number needed to intervene” of 1.8, according to the calculator.
“Because the calculator uses routinely available EHR data, it can potentially be integrated into scheduling workflows to prioritize outreach. However, prospective and external validation are needed before clinical deployment,” the authors advised.
Read much more, including potential study limitations, in the Journal of the American College of Radiology.
Marty Stempniak has covered healthcare since 2012, with his byline appearing in the American Hospital Association's member magazine, Modern Healthcare and McKnight's. Prior to that, he wrote about village government and local business for his hometown newspaper in Oak Park, Illinois. He won a Peter Lisagor and Gold EXCEL awards in 2017 for his coverage of the opioid epidemic.