Page 179 - Read Online

Page 10 of 13 Kolba et al. Vessel Plus 2023;7:12 https://dx.doi.org/10.20517/2574-1209.2022.61

DISCUSSION
Multivariable analysis of the SPARCs database records showed no difference in short-term outcomes of
MEI New York State adult residents. Given the increased baseline comorbidities and healthcare spending
reported in the literature in this population, this result was unexpected .
[22]

In this SPARCS New York State database analysis, only 9.95% of AVR and 14.05% of r-AVR patients had
preoperative MEI diagnoses. This is below the rate of New Yorkers in the general population with MEI
reported by the Department of Health. As MEI patients do not appear to receive AVR or r-AVR procedures
at the same rates as non-MEI patients, there may be multiple potential explanations for these differential
procedural use rates.

For MEI patients, studies have reported lower medication compliance rates and follow-up rates as
compared to non-MEI patients [23,24] . Follow-up difficulties with MEI patients may be related to the
“revolving door phenomenon,” where MEI patients experience a temporary improvement of symptoms
following a facility-based treatment, but then stop their facility-based encounters until a relapse episode or a
[25]
re-hospitalization occurs . Additionally, a MEI diagnosis might preclude patients from being selected or
referred for an invasive treatment, such as a surgical procedure. Importantly, MEI patients that are well
controlled might present otherwise similarly to patients without a MEI diagnosis. For this SPARCS analysis,
the level of detailed diagnostic information available to clinicians for their pre-procedural AVR referral or
AVR treatment selection decisions was not documented. Although possibly either a referral for AVR
treatment or an AVR treatment selection bias of MEI patients may have occurred, sensitivity analyses run
for multivariable analyses that included a MEI treatment propensity variable had no substantive changes.

For all three of the multivariable AVR logistic regressions, however, SAVR procedures had statistically
significantly higher odds ratios associated with POAF, MM, and READMIT adverse outcomes as compared
to TAVR procedures. Independent of a patient’s pre-existing MEI status, therefore, treatment type appears
to be an important predictor of adverse AVR outcomes. Given these SPARCS study records were extracted
from 2005 to 2018, however, this preliminary finding should be re-confirmed by updated longitudinal
analyses.

Specifically, TAVR procedures were approved for Medicare payment in 2012 for high-risk patients. In 2016,
TAVR procedures received expanded Medicare approval for application to intermediate risk patients. Thus,
it was not until after this study’s time-period had already ended - in August 2019 - that TAVR procedures
[26]
were approved for widespread use in low-risk patients . Given these period-specific policy changes, it
would not be appropriate to directly report a TAVR vs. SAVR procedural comparison. Based on these
preliminary findings, additional research appears needed to evaluate differential TAVR vs. SAVR utilization
in MEI vs. non-MEI patients while adjusting for other patient risk factors.

Surprisingly, there were no risk adjusted POAF rate differences between the MEI and non-MEI patients.
Given that MEI patients are commonly prescribed anti-psychotics, anti-depressants, anti-anxiolytics,
sedatives, opioids, stimulants, and pain management medications that may result in postcardiac surgery
arrhythmias, this finding was unexpected.

Consistent with already published literature, this study also found that smokers were less likely to be
readmitted to the hospital. This has been documented as the “smoker’s paradox,” a phenomenon in which
the varying age and comorbidities between smokers lead to this unexpected outcome [27,28] .