Invited commentary

Abstract

The authors provide an exhaustive review of patients captured in the Nationwide Readmissions Database who had thoracic endovascular aortic repair performed in 2013. They conclude that early readmission was more likely due to cardiac factors, with later readmissions increasingly related to the aneurysm repair. Not surprisingly, thoracic endovascular aortic repair for rupture was associated with a high likelihood of readmission at all intervals. This review highlights both the strengths and inherent weaknesses of the many database reviews that currently populate our literature. The large sample size, capturing a more realistic population than Medicare-based reviews, adds weight to their observations. They raise intriguing questions about increased length of stay as protective of early readmission. Similarly, they demonstrate the high rate of readmission at all intervals from non-aneurysm-related causes, which has particular significance for 30-day and 90-day readmissions linked to reimbursement. How, then, do we use these results to effect changes to improve our patient care? The limitations of the database do not permit a finer analysis of factors that may help. How many patients were discharged to home or to an extended care facility? What were the discharge medications, and was medical care at discharge optimized? Is there a role for targeting certain factors related to cardiac and infectious complications in the postdischarge setting that could provide for early outpatient intervention and ward off readmission? Of the non-aneurysm-related readmissions, how many and which ones were due to complications directly related to the procedures and how many to the patient's underlying comorbidities, particularly at the later intervals? The choice of database is also important and should be directed by the hypotheses to be tested. As the authors state, the use of the Nationwide Readmissions Database data allows them to investigate long-term readmissions in a realistic population, a study that could not be conducted in the Medicare, Vascular Quality Initiative, or National Surgical Quality Improvement Program databases because of their inherent weaknesses. Those databases, however, provided finer granularity for investigating pertinent clinical factors. I would recommend that all authors contemplating a database review re-read the excellent article “Analysis of large databases in vascular surgery” by Nguyen and Barshes1Nguyen L.L. Barshes N.R. Analysis of large databases in vascular surgery.J Vasc Surg. 2010; 52: 768-774Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar as well as the overview of available databases with their strengths and weaknesses summarized in the Comparative Effectiveness Research Resources Guide published by the Society for Vascular Surgery.2Comparative effectiveness research resources guide.https://vascular.org/sites/default/files/ComparativeEffecResearchGuide_FINAL.pdfGoogle Scholar As we are presented with an increasing number of database analyses, it is important to acknowledge the limitations as well as the value of large population studies. They often do not settle a question but raise issues that benefit from a prospective institutional or multi-institutional study or perhaps validation from another database or registry with different limitations. Finally, many of these studies can serve as a starting point for internal quality initiative projects to more completely analyze and to improve factors related to poor patient outcomes. Readmissions after thoracic endovascular aortic repairJournal of Vascular SurgeryVol. 68Issue 2PreviewThe care of patients undergoing thoracic endovascular aortic repair (TEVAR) can be resource intensive, which can be driven by readmissions. Our objective was to characterize index readmissions at 30, 90, and 180 days after TEVAR. Full-Text PDF Open Archive