LAUSR

The utilization of quality metrics to determine the value of patient care is one of the most important means by which regulatory agencies such as the Center for Medicare & Medicaid Services (CMS) determine which institutions are reimbursed or penalized. Readmission rates are often utilized for this purpose as it is easily tracked, modifiable, and allows for comparison over time. In this editionof the Journal of SurgicalOncology,Dr.Bagante,Dr.Beal, and colleagues examined 90-day readmissions after hepatopancreatic (HP) procedures utilizing the Nationwide Readmissions Database.1 The studyexaminedover41 000patientsand found that22%ofpatientshada 90-day readmission with a slightly higher rate in patients with malignancy compared to thosewithbenigndisease.Wewould like to congratulate the authors for tackling this complicated topic and it is an important study that demonstrates thehigh rateof readmission inpatients undergoing complex HP operations. The study further illustrates that readmission is a complicated metric to assess as the authors found that patients with malignancy were more likely to have comorbidities and higher severity of disease versus patients with benign disease. Thus, readmission was not only dependent on the type of procedure but also disease characteristics, and patient characteristics. Our group has previously examined readmission rates in a multi-institutional study in patients undergoing pancreaticoduodenectomy for benign and malignant disease and similarly found that diagnosis is an important variable influencing recidivism.2 In the present study, patients with malignancy had higher readmission rates than those with benign disease. It needs to be emphasized that these two groupswere not equally matched as patients withmalignancywere older, had greater medical co-morbidities, and had different insurance types; factors which alone are associatedwith higher readmission rates. Utilization of propensity-matching analysis would also have allowed for better comparisons between these groups when consideringdifferent pathologies.Another limitationof thepresent study was the inability to perform an in-depth analysis of the reasons for readmission.Withmoregranular data, not onlywoulddirect comparisons be possible but also identification of areas in need of improvement. Although a national patient data set was used, there are questions about the accuracy of the database. In this study, 19% of patients were admitted to the non-index hospital, a value significantly less than has been previously reported.Yermilovetal3performedapopulation-basedappraisal of readmissions following pancreaticoduodenectomy for pancreas cancer using the California Cancer Registry. They found that 47% of readmitted patients were admitted to a non-index hospital, which is likely more reflective of true readmission rates. With any institutional database, it is challenging to capture data points outside of the primary institution and similarly, the current study is likely not capturing all non-index readmissions. In conclusion, this study shows that patient readmission should not be used as a metric by governing agencies, such as the CMS, to measure the quality of patient care. This is a complex variable influenced by multiple patient, pathologic, and operative-based factors and cannot be used as an isolated variable for all procedures. This study highlights not only that readmission rates are high after HP surgery but also that it is a complexmetric. The authorsneed tobecommendedon investigating this challenging topic and their data should provide impetus for centers to evaluate current processes and identify opportunities for improvement.