Longer-term mortality seen as more reliable metric
Thirty-day mortality paints an incomplete picture of hospital performance in aortic valve replacement, whereas 90-day mortality is the better quality metric, a study showed.
Hospitals ranked as top-performers (lowest 10% mortality), middle-performers (middle 80%), and bottom-performers (highest 10% mortality) by Medicare records of 30-day mortality averaged over 4 years often changed ranks when they were judged by 90-day mortality.
For transcatheter aortic valve replacement (TAVR), 21.7% of centers shifted to a different performance category between the two metrics, reported Tsuyoshi Kaneko, MD, and colleagues of Brigham and Women's Hospital and Harvard Medical School in Boston, and colleagues in a study published online in JAMA Cardiology.
Nearly half of the "best" declined and 29.6% of the "worst" improved rank when considered by the 90-day measure. By 1 year, 30.4% of hospitals had a change in ranking compared with the 30-day metric.
Findings were similar for surgical aortic valve replacement (SAVR), for which 17.3% of hospitals changed rankings from the 30- to 90-day measures and 30.3% from the 30-day to 1-year metrics. Results persisted after adjusting for patient risk.
"An analysis of the instantaneous hazard found that, although the hazard (risk of mortality) diminished rapidly during the first 30-day period, it remained elevated through 90 days for the overall cohort and then plateaued to fairly constant levels thereafter during the first year for TAVR or SAVR. Thus, capturing 90- day events was more robustly informative regarding expected 1-year outcomes," the authors reported.
Evaluating hospital performance by 30-day mortality therefore runs the risk of underestimating outcomes and misrepresenting institutional performance after TAVR and SAVR, they concluded. "Although 30-day mortality has been validated, 90-day mortality may be a more reliable outcome metric for measuring hospital performance and capturing procedure-related mortality."
Their cohort study included Medicare beneficiaries who had valve procedures from 2012 to 2015 without concomitant coronary artery bypass grafting or other surgeries. Hospitals were also excluded if they performed fewer than 50 TAVRs or 70 SAVRs per year.
Investigators ended up with more than 30,000 TAVR admissions at 184 hospitals and more than 26,000 SAVR admissions at 191 hospitals.
Thirty-day mortality has been challenged before as a quality indicator of hospital outcomes in non-cardiac surgery.
"[I]t should not be surprising that 90-day mortality rates more closely reflects 1-year results than 30-day outcomes. The extended period yields more information to inform what is likely to happen at 1 year," wrote a group led by Michael Mack, MD, of Houston Methodist DeBakey Heart & Vascular Institute, in an accompanying editorial.
"However, outcomes at some point become more reflective of underlying patient intrinsic factors and comorbidities extended out until the risk from the procedure itself is negligible. When is that point reached? At 30 days? At 90 days? At 180 days?" they posited.
Kaneko and colleagues cautioned that the hospital claims database used for the study was subject to possible coding errors and that patients who switched hospitals for their care could not be accounted for. Moreover, the data they had in hand may not reflect current practice.
"For example, the current study was performed between 2012 and 2015 when TAVR was performed mainly in patients at high surgical risk," agreed Mack's team. Since then, they noted, low-risk TAVR has come out victorious in trials, with 30-day and 1-year mortality at 0.4% and 1.0%, respectively, in PARTNER 3 and 0.5% and 2.3% in Evolut Low Risk.
"Clearly, the addition of collecting 90-day outcomes in this population would be highly unlikely to discriminate hospital-performance differences. Of course, current clinical practice will be somewhere between these 2 extremes and not reflective of either," the editorialists said.
Ninety-day mortality as a hospital performance metric would fit in with the Centers for Medicare and Medicare Services (CMS) bundled payment models in which hospitals are responsible for all costs incurred from the time of the index procedure through 90 days, Mack and colleagues noted.
"There will clearly become an added burden of data collection and reporting to clinical sites if this does become a standard metric for hospital performance. It is therefore mandatory that we continue to examine whether the juice is worth the squeeze, as the saying goes," they urged.
In the meantime, demand for TAVR procedures is expected to spike what with the recent FDA approval for low-risk patients and the reduced site volume requirements in the latest National Coverage Determination from CMS.
The study was supported by institutional Levinger gift funds.
Kaneko disclosed being a speaker for Edwards Life Sciences and Medtronic
Mack reported nonfinancial support from Edwards Lifesciences, Medtronic, and Abbott.
Source Reference: Hirji S, et al "Utility of 90-day mortality vs 30-day mortality as a quality metric for transcatheter and surgical aortic valve replacement outcomes" JAMA Cardiol 2019; DOI: 10.1001/jamacardio.2019.4657.
Source Reference: Mack M, et al "Ninety-day outcome assessment after transcatheter and surgical aortic valve replacement -- is the juice worth the squeeze?" JAMA Cardiol 2019; DOI: 10.1001/jamacardio.2019.4821.
Read the original article on Medpage Today: Study: 30 Days Not Enough to Track TAVR, SAVR Quality