Tool for predicting early readmission risk

27 Mar 2017. NUS clinician scientists have developed a web-based tool for predicting hospital readmission risk in Singapore.

Some patients may require readmission soon after they leave the hospital. The researchers have developed an online tool to help doctors identify them. With this tool, readmission may be prevented among patients.

In Singapore, approximately 15% of patients discharged from hospitals today will succumb to a readmission in the following 30 days. This places an immense strain on the healthcare system. Efforts to identify and intervene on high-risk patients may help reduce readmission rates. However, only one-third of all readmissions are believed to be potentially preventable. Risk factors that independently predict early readmission risk remain poorly investigated. This hinders efficient allocation of interventional healthcare resources to reduce readmission rates.

A research team involving Sreemanee Raaj DORAJOO, a Ph.D. student and his supervisor, Prof Alexandre CHAN, collaborated with pharmacists from Khoo Teck Puat Hospital and Singapore General Hospital to study this issue. As part of the study, they have developed a prediction tool that uses patients’ current state of health to predict the risk of early readmission. They found that for each additional medication prescribed to the patient, the 15-day readmission risk rises by approximately 6%.The tool was tested on more than 600 local patients from two cohorts, showing relatively high accuracy.

With this prediction tool, risk stratification based on the 15-day readmission risk can be performed on discharging patients. This allows healthcare teams to identify high-risk patients who can be assessed separately and administered interventions tailored to their needs. These include specialised discharge planning, medication counselling, caregiver training or placing patients on a home visit programme. By selectively administering these interventions to patients who are most likely to benefit from them, readmission risk can be reduced. This could potentially raise the overall effectiveness of the interventions and reduce healthcare costs.

The research team plans to integrate the tool with the electronic medical records system in healthcare institutions in Singapore. Further work should include a comprehensive evaluation of the clinical and economic impact of their interventional programmes by preventing readmission using this tool. Additional validation of the tool in other tertiary healthcare settings could be pursued to further refine prediction accuracy.

Figure shows the screenshot of the prediction tool implemented as a web application to facilitate 15-day readmission risk assessment by healthcare providers.

Reference

SR Dorajoo; V See; CT Chan; JZ Tan; DSY Tan;, SMBA Razak; TT Ong; N Koomanan; CW Yap; A Chan*, “Identifying potentially avoidable readmissions: A medication-based 15-day readmission risk stratification algorithm” PHARMACOTHERAPY DOI: 10.1002/phar.1896 Published: 2017 [Epub ahead of print]