Applying Hospital Readmissions to Oncology: A Square Peg in a Round Hole?

A large volume of recent research and policy work has sought to understand and reduce readmission within 30 days of an initial, or index, hospitalization.Since 2012, Medicare's Hospital Readmissions Reduction Program (HRRP) has profiled hospitals by riskadjusted readmission rate, penalizing institutions for higher than expected readmission rates among a narrow set of common and relatively homogenous conditions (acute myocardial infarction, heart failure, pneumonia, and chronic obstructive pulmonary disease) and procedures (coronary artery bypass graft surgery, elective total hip, and total knee arthroplasty). 1 Although some research on multicondition predictive modeling has included patients with cancer, few studies have exclusively focused on hospital readmissions in cancer.Against this backdrop, Manzano et al 2 developed a multivariable regression model and linked Cancer READMIT risk score to predict 30-day hospital readmissions among adult patients with solid tumor using data from 2011 to 2015 at the University of Texas MD Anderson Cancer Center.They reported a 16% readmission rate, which was comparable with rates for other conditions.The modeling included many domains associated with readmissions in other conditions, including disease severity (cancer type and distant metastases), social determinants of health (non-English speaking and nonmarried), prior emergency department utilization, laboratory abnormalities on discharge (hemoglobin , 10 and WBC .11), and length of stay (. 4 days).