Page 178 - Read Online

Kolba et al. Vessel Plus 2023;7:12 https://dx.doi.org/10.20517/2574-1209.2022.61 Page 9 of 13

Table 4. Preoperative mental illness diagnoses among redo-SAVR and ViV-TAVR patients
Percentage of patients with specific mental illness Percentage of redo-SAVR and ViV-TAVR
Mental illness
diagnoses (total = 41; %) patients (%)
PTSD 0.05 0.83
PTSD-related 0.05 0.83
Depression 0.63 10.74
Anxiety 0.24 4.13
Obsessive-compulsive 0.02 0.41
disorder
Bipolar disorder 0.02 0.41
Schizophrenia 0.27 4.55
Mild cognitive 0.02 0.41
impairment
Dementia 0.02 0.41
Definitions: PTSD (post-traumatic stress disorder).

Table 5. First-time AVR and outcomes in patients with mental illness
Variable OR (95%CI) P-value
POAF 0.982 (0.920-1.047) 0.5746
30-day readmission 1.059 (0.977-1.149) 0.1632
30-day STS composite rates 1.010 (0.914-1.116) 0.8478
Listed respectively:
New-onset POAF model’s statistically significant variables in the model included age, surgery type, type of admission, gender, race, ethnicity,
insurance type, history of stroke, bicuspid aortic valve, non-rheumatic aortic stenosis, obstructive sleep apnea, pulmonary hypertension, and
Elixhauser Mortality Index. The c-index of this model was 0.658 and the Hosmer and Lemeshow Goodness-of-Fit Test p-value was 0.0001.
30-Day Readmission Model’s statistically significant variables in the model included age, surgery type, gender, insurance type, alcohol abuse,
tobacco/smoking, stroke, myocardial infarction, bicuspid aortic valve, abdominal aortic aneurysm, non-rheumatic aortic stenosis, leukemia, and
Elixhauser Readmission Index. The c-index of this model was 0.633 and the Hosmer and Lemeshow Goodness-of-Fit Test p-value was 0.7041.
30-Day Mortality/Morbidity (MM) composite model’s statistically significant variables in the model included age, surgery type, type of
admission, insurance type, tobacco/smoking, stroke, bicuspid aortic valve, peripheral vascular disease, abdominal aortic aneurysm, non-
rheumatic aortic stenosis, obstructive sleep apnea, leukemia, hypothyroidism, intra-aortic balloon pump, and Elixhauser Mortality Index. The c-
index of this model was 0.733 and the Hosmer and Lemeshow Goodness-of-Fit Test p-value was 0.0117.

POAF: Postoperative atrial fibrillation.

Table 6. Redo AVR and outcomes in patients with mental illness
Variable OR (95%CI) P-value
POAF 1.335 (0.580-3.072) 0.4973
30-day readmission 0.660 (0.206-2.122) 0.4860
30-day STS composite rates 0.591 (0.221-1.581) 0.2950
Listed respectively:
For new-onset POAF model, there were no statistically significant risk factors. The c-index of this model was 0.729 and the Hosmer and
Lemeshow Goodness-of-Fit Test p-value was 0.0282.
The only 30-day readmission model’s statistically significant variable was cerebral vascular disease. The c-index of this model was 0.645 and
the Hosmer and Lemeshow Goodness-of-Fit Test p-value was 0.1790.
The Mortality/Morbidity Composite model’s statistically significant variables included race and cerebral vascular disease. The c-index of this
model was 0.706 and the Hosmer and Lemeshow Goodness-of-Fit Test p-value was 0.7659.
POAF: Postoperative atrial fibrillation.

Sensitivity analyses: no impact of new mental illness propensity variable
To assure no findings substantively changed, a new sensitivity analyses was run by including a new
propensity variable (i.e., the likelihood of mentally ill patients to receive AVR treatment). Although these
reports have not been provided herein, the current multivariable models were found to be robust.