Page 172 - Read Online

Kolba et al. Vessel Plus 2023;7:12 https://dx.doi.org/10.20517/2574-1209.2022.61 Page 3 of 13

[12]
be found online .

Ethical oversight -CORIHS exemption as “not human subjects” research
At Stony Brook University, the Office of the Vice President for Research’s Office of Research Compliance
staff coordinate research project ethical oversight and approvals. Specific to this SPARCS database study,
only de-identified summary reports were generated by the Stony Brook University Biostatistical Consulting
Core (BCC) lab staff. Thus, the Stony Brook University Committee on Research in Human Subjects’
Institutional Review Board (CORIHS IRB) office granted a “not human study” research (NHSR) written
exemption for this study [IRB 2022-00375].

Observational study design
Using the SPARCS database records from 2005-2018, this retrospective, observational cohort study
compared outcomes such as POAF and 30-day readmission rates after both AVR and r-AVR procedures for
patients with and without a preoperative mental illness diagnosis. To ensure completeness of this
observational study’s administrative data reporting, the pre-established, STROBE evidence-based medicine
[13]
report criteria were utilized . To identify opportunities to reduce this observational study’s potential risk
for bias, moreover, the Newcastle Ottowa scale for cohort studies was utilized .
[14]
Patient population: study inclusion and exclusion criteria
To date, little is known about the MEI impact upon AVR or r-AVR patients; thus, eligible SPARCS adults
(age > 18 years) undergoing AVR and r-AVR procedures were extracted based on ICD9/ICD10 hospital
billing codes [Supplementary Table 1]. To assure 30-day complete follow-up, only New York State adult
residents with either first-time aortic valve replacement (AVR) or repeat aortic valve replacement (r-AVR)
from January 2005 to November 2018 were extracted; by utilizing only New York State residents’ records,
challenges with out-of-state care provided were minimized.

To identify the subgroup of patients with new-onset POAF/AFL, all records for patients with a prior two-
year history of atrial fibrillation or flutter, Maze procedure, pacemaker implantation, or defibrillation prior
to their first-time, non-emergent AVR or subsequent r-AVR procedure were excluded from the definition
of “new onset” atrial fibrillation. Additionally, all records for pediatric patients (i.e., < 18 years of age) were
excluded.

Study definitions: major mental illness diagnoses
A preoperative major mental illness (MEI) diagnosis was used as this study's independent variable; thus, this
was the basis for study comparisons of MEI vs. non-MEI records. For this purpose, mental illness was
defined broadly [Supplementary Table 1] to include patients with a diagnosis of depression (and related
disorders such as adjustment disorder, adjustment disorder with depressed mood, etc. based on ICD codes),
post-traumatic stress disorder (PTSD) and related disorders, generalized anxiety, alcohol-induced mental
disorders, obsessive-compulsive disorder, bipolar disorder, schizophrenia, mild cognitive impairment, and
dementia were selected as patients with preoperative mental illness, based on ICD codes contained in the
SPARCS database. Recorded in SPARCS, these MEI diagnoses were based on the Diagnostic and Statistical
Manual of Mental Disorders (DSM) and audited independently by government agencies (e.g., Medicare and
Medicaid) to ensure billing accuracy .
[15]

Preprocedural risk factors
Upon admission, patients’ risk factors were assessed and recorded; these risk factors included their
demographics/socioeconomic status (i.e., age, race, ethnicity, sex, insurance status, etc.), pre-procedural
clinical diagnoses, historical cardiac-related diagnoses, historical cardiac-related procedures performed, and