Model May Predict Which UC Patients Risk Rehospitalization

Four variables easily accessible at hospital discharge could predict the risk for rehospitalization at 90 days among patients with ulcerative colitis (UC), a preliminary modeling study suggests.

“Absence of a gastroenterologist consultation within the year prior to admission, male sex, shorter length of hospital stay, and narcotic prescription at the time of discharge were independently associated with the risk for 90-day rehospitalization for a UC-related indication,” study author Sanjay Murthy, MD, associate professor of gastroenterology at the University of Ottawa, Ottawa, Ontario, Canada, and staff gastroenterologist at the Inflammatory Bowel Disease Centre at The Ottawa Hospital, Ottawa, Ontario, Canada, told Medscape Medical News.

“While some hospital readmissions are likely unavoidable, a subset of them, particularly readmissions that occur soon after discharge, may be preventable with early and intensive postdischarge outpatient management,” he said. “Identifying those who are at high risk for early readmission is a rational first step toward applying targeted outpatient interventions that reduce this risk.”

The study was published on January 24 in The Journal of the Canadian Association of Gastroenterology.

Major Predictor Variables 

The researchers conducted a retrospective study in adults with UC who were admitted to The Ottawa Hospital between 2009 and 2016 for a UC flare or UC-related complication, excluding bowel cancer. Using medical records and administrative health databases, they derived and validated a multivariable logistic regression model of 90-day UC-related rehospitalization risk.

Participants’ mean age at UC diagnosis was 35.3 years and 50.4% were men. In the year before the index hospitalization, 138 (55.6%) participants had a gastroenterologist visit, whereas 41 (16.5%) were hospitalized.

During the index hospitalization, 42 (16.9%) patients were newly diagnosed with UC, and 25 (10.1%) underwent intra-abdominal surgery. At discharge, 34 (13.7%) patients were prescribed an outpatient narcotic. The mean length of hospital stay was 9.97 days. Twenty-seven individuals (10.9%) were rehospitalized within 90 days of discharge.

Out of 35 variables, the model identified the following four as significant predictors of 90-day rehospitalization: Gastroenterologist consultation within the prior year (adjusted odds ratio [aOR], 0.09), male sex (aOR, 3.77), length of hospital stay (aOR, 0.93), and discharge with narcotics prescription (aOR, 5.94).

The model had 77.8% sensitivity, 80.9% specificity, 33% positive predictive value, and 96.7% negative predictive value for predicting high vs low risk for 90-day hospital readmission.

The researchers noted several study limitations. The cohort was relatively small, which limited the statistical power for model building and identifying variable associations with the outcome. In addition, the study was conducted in a single tertiary care center, which limits its generalizability. Retrospective data may have affected the accuracy of the measurements, and information on some relevant variables was not available.

Nevertheless, Murthy said, “Optimally applying our prediction model at the point of hospital discharge would have classified only about a quarter of individuals in our cohort as being at high-risk for 90-day readmission and potentially needing targeted early outpatient intervention, and this would have captured close to 80% of individuals who were destined for early readmission.”

“However, our research is still preliminary and requires considerably more work to ensure that the findings are suitable for application to clinical practice,” he added. “In the meantime, practitioners may reflect on the potential importance of the major predictor variables identified in our study within their practices.”

Careful Follow-Up Key 

Rajiv Bhuta, MD, assistant professor of clinical gastroenterology and hepatology at Temple University and a gastroenterologist at Temple University Hospital, both in Philadelphia, commented on the study for Medscape Medical News.

“The model performed fairly well (c-statistic of 0.78) using four variables: Gastroenterologist consultation within the prior year (protective), male sex (higher risk), length of stay (marginally protective), and narcotic prescription at discharge (higher risk). These are intuitive predictors that align with prior literature on UC hospitalizations,” said Bhuta, who was not involved in the study.

“From a clinical perspective, this type of tool could be useful for targeting high-risk patients for early outpatient interventions (eg, close gastroenterology follow-up and pain management strategies). The negative predictive value (96.7%) suggests that it is particularly good at identifying patients at low risk for rehospitalization, which may help prioritize resource allocation more efficiently. However, practical implementation will require external validation and integration into electronic medical records to automatically flag high-risk patients at discharge.”

In addition, Bhuta noted, “the study only examines patient data through 2016. Why have the last 8 years been excluded? Given the small sample size and the sea change in available inflammatory bowel disease therapies since 2016, there could be significantly different findings with more current data.”

Furthermore, there is a lack of specific data supporting the protective effect of a gastroenterology visit in the previous year, and the readmission rate was lower than that reported by others (10% vs 20%), which, he said “may skew their findings.”

“The strong protective effect of prior gastroenterologist visits underscores the importance of specialty proactive disease management in these complex patients,” Bhuta continued. “Narcotic prescriptions at discharge may indicate inadequate disease activity control, thus making these patients important targets for close follow-up. Narcotics are generally not required once successful disease control has been achieved with steroids or biologics.

“While promising, this tool should not yet replace clinical judgment until it undergoes external validation,” he concluded. “In the meantime, clinicians should focus on structured outpatient follow-up and careful discharge planning to minimize UC-related rehospitalizations.”

This study was funded by a grant provided to Murthy by the Department of Medicine at the University of Ottawa. Murthy and Bhuta declared having no relevant financial relationships.

Marilynn Larkin, MA, is an award-winning medical writer and editor whose work has appeared in numerous publications, including Medscape Medical News and its sister publication MDedge, The Lancet (where she was a contributing editor), and Reuters Health.