Comment
Unmatched analysis in this study demonstrates an association between the use of MAC for TAVR and reductions in OR time, ICU LOS, total LOS, hospital cost, and total cost. However, MAC was also associated with lower VASQIP estimated mortality (Table 1 ). This difference reflects temporal trends in TAVR use. TAVR was initially applied only to patients not considered surgical candidates as a salvage alternative to medical therapy, then to high risk patients, and eventually to intermediate and low risk patients in later years.2, 27, 30-32 MAC was not used for TAVR at this institution until 2015, and as it became increasingly implemented, the use of GA declined (Figure 2 ). This trend is consistent with national trends.14 Given the asymmetrical temporal distribution of surgical risk and anesthesia modalities with time, propensity matching was necessary to reduce confounding bias and for meaningful comparisons between anesthesia modalities for TAVR.
The incidence of adverse outcomes, including atrial fibrillation, cardiac arrest, CVA, MI, and renal failure requiring renal replacement therapy, were not significantly different between propensity-matched groups of TAVR patients. The incidence of these events after TAVR is known to be low, and this study was insufficiently powered to detect differences in the occurrence of these uncommon events.14 25, 29, 33 No significant differences were found in 30-day mortality or the percentage of patients discharged elsewhere besides home, but MAC was associated with lower 180-day mortality. It is difficult to ascertain whether this is a true effect of the anesthesia modality or a reflection of underlying bias not eliminated by propensity matching. Studies have suggested 30-day mortality may underestimate outcomes, and longer-term outcomes may be better metrics for assessing hospital performance and patient outcomes.34
Several studies have compared costs between MAC and GA in TAVR.14, 24, 25, 35-37 This study demonstrated significant differences between anesthesia modalities in OR time, ICU LOS, and total LOS after propensity-matching. It was hypothesized that these differences would translate to reductions in patient charges, but this was not the case. The absence of significant cost savings among MAC TAVR has been reported in a prior unmatched single-center study, which similarly found decreased ICU and hospital LOS with MAC but no cost benefit.36 The authors found similarity in costs was mostly driven by increased “miscellaneous” costs in the MAC group, which primarily included the costs of treating unrelated comorbidities during the hospital stay. A 2014 retrospective unmatched comparison between TAVR via “minimalist approach,” involving MAC and the use of catheterization lab, and a “standard approach,” which involved GA and the use of a hybrid OR, found the minimalist approach was associated with an average cost reduction of $9,892.24 The high cost of a hybrid OR likely explains these findings, and the use of MAC may not have driven savings.38, 39 A 2017 propensity-matched analysis found a 28% decrease in direct cost when MAC was used for TAVR.25 However, this latter study was of high-risk patients, with a large percentage of cases being done by transapical approach. Average hospital stay for both arms was nearly double that observed in this series, highlighting the limited generalizability of the study to more recent trends in the use of TAVR. Other comparisons of high-risk patients, confounded by baseline differences between groups, have demonstrated reduced OR, ICU, and total hospital LOS times with the use of MAC for TAVR.35, 37With the inclusion of intermediate and low risk patients, the data presented herein offers more generalizability to the population currently treated with TAVR.
Hospital and operator volume have been demonstrated to correlate with outcomes in TAVR.40-42 To minimize the impact of physician and institutional experience on OR time, we compared MAC and GA TAVR within the year 2016. This year was chosen as it had the most even distribution of TAVR between MAC and GA and followed over 60 cases and two years of institutional experience with TAVR. In this analysis, MAC was still associated with significantly shorter OR time, suggesting the difference is due to the anesthesia modality itself, not confounders such as increased institutional experience or new deployment devices.
Physician experience and case volume vary considerably between studies, which poses a challenge when comparing our findings to previous analyses. Among the TAVR programs developed at 13 national VAMC’s, our center is the highest volume VAMC TAVR program based on internal data. While prior studies were performed at non-VA hospitals, this is the first study comparing costs between MAC and GA TAVR at a VAMC, which has a very standardized cost structure. Differences in administrative, clinical, and billing operations between non-VA and VA centers may limit extrapolation of costs to non-VA centers. For instance, the relative value units (RVU) used for reimbursement purposes are calculated differently between VAMCs and other hospitals. RVUs defined by the Centers for Medicare and Medicaid Services (CMS) for non-VA facilities are based on factors like technical skill, stress, and judgement required for a treatment associated with a current procedural terminology (CPT) code. The VA instead uses a unique RVU formula that is based on the resources used for a service (e.g., OR time, ICU LOS) and further specifies RVUs for several different resource categories (labor, supplies, overhead, contractors, etc.).43 It can be difficult to extrapolate costs between VA and non-VA hospitals, as studies of these differences are limited. Healthcare costs for end-stage renal disease patients have been shown to be more expensive at VA hospitals compared to at non-VA hospitals, but studies like this are confounded by comorbidity distribution and different approaches to medical management.44 More recent comparisons have demonstrated increased costs at VA hospitals compared to community hospitals for coronary artery bypass grafting, but lower costs for percutaneous coronary intervention.45 Cost data for other hospitals would include indirect costs, such as administrative expenses.
While our study was performed at the highest volume VA TAVR program, it remains limited by the propensity matched sample size and its retrospective nature. Additionally, the VA patient population is almost exclusively male, limiting the generalizability of our results. Despite these limitations, our study demonstrates strong advantages to the use of MAC for TAVR. In a high volume VAMC, TAVR performed under MAC is safe, leads to an improvement in OR efficiency, and reduces ICU and hospital length of stay, but does not translate into cost savings.