Published online by Cambridge University Press: 12 January 2018
Postoperative delirium has been associated with poorer long term survival in Transcatheter aortic valve replacement (TAVR) and Surgival aortic valve replacement (SAVR) patients. However, its effect on hospitalization costs and length of stay in these populations has not been formally assessed.
Using the Medicare Provider Analysis and Review File, we retrospectively analyzed elderly (80 years of age and older) Medicare patients receiving TAVR and SAVR in the United States during the 2015 fiscal year. ICD-9-CM codes were used to identify postoperative delirium diagnoses. The incremental hospital resource consumption, measured as hospital cost and length of stay, was estimated for patients with postoperative delirium during their TAVR or SAVR index hospitalization. Multivariate regression models were used for the adjusted cost estimates controlling for patient demographics, comorbidities, and complications.
A total of 21,088 claims were available for analysis (12,114 TAVR and 8,974 SAVR). The mean age of the TAVR group was older compared to the SAVR group (87 versus 84; p < .001) and TAVR patients presented with a higher comorbidity burden (Charlson Index score 3.0 versus 2.1; p < .0001). TAVR patients experiencing postoperative delirium during the index hospitalization was 1.6 percent compared to 3.6 percent of surgical patients (p < .0001). For the overall cohort, the regression adjusted incremental cost of postoperative delirium was (USD15,592; p < .0001). Patients experiencing delirium also had significantly longer hospital length of stay (4.16 days; p < .0001). When stratified by treatment approach, the adjusted incremental cost was USD13,862 for TAVR (p < .0001) and USD16,656 for SAVR (p < .0001).
While infrequent, postoperative delirium significantly increased hospital cost and length of stay following transcatheter or surgical aortic valve replacement (AVR). Despite a significantly higher comorbidity burden, TAVR was associated with lower postoperative delirium rates compared to SAVR. Moreover, post-TAVR delirium may be associated with less resource consumption than post-SAVR delirium. Future studies should seek to determine whether general anesthesia avoidance in appropriately selected transfemoral TAVR patients can further decrease rates of delirium.