Page 100 - Read Online

Page 4 of 11 Hawkins et al. Vessel Plus 2022;6:42 https://dx.doi.org/10.20517/2574-1209.2021.116

Patients who developed POAF were older (70 median years of age vs. 64 median years of age, P < 0.001), less
likely to be women (25.0% vs. 27.7%, P < 0.001), and, in general, had a higher burden of comorbid disease
including hypertension (89.1% vs. 86.4%, P < 0.001), diabetes (44.3% vs. 47.1%, P < 0.001), moderate to
severe chronic lung disease (12.2 % vs. 9.92%, P < 0.001), and heart failure (31.7% vs. 28.2%, P < 0.001).
Patients who developed POAF were less likely to receive a preoperative beta-blocker, relative to those who
did not develop POAF (86.7% vs. 88.2%, P = 0.002). STS predicted risk of mortality was significantly greater
for patients who developed POAF (1.46% vs. 1.02%, P < 0.001).

Patients developing POAF were less likely to receive surgery on an emergent or urgent basis, relative to
those not developing POAF (52.9% vs. 57.8%, P < 0.001). A significantly greater proportion of POAF
patients had an IABP placed during their hospitalization, relative to those without POAF (7.68% vs. 6.44%, P
< 0.001). Rates of POAF did vary significantly by procedure, and the number of disease coronary arteries
present (P < 0.001 for both). Patients with POAF underwent longer cross-clamp (75.0 min vs. 70.0 min, P <
0.001) and cardiopulmonary bypass times (101 min vs. 94.0 min, P < 0.001), relative to those without POAF.

Postoperative outcomes and resource utilization
By univariate analysis POAF was associated with significantly increased rates of STS operative mortality
(3.04% vs. 1.43%, P < 0.001), and major morbidity (13.7% vs. 6.07%, P < 0.001). Incidence of cardiac arrest
(2.95% vs. 1.02%, P < 0.001), pneumonia (3.71 vs. 1.15, P < 0.001), and transfusion requirement (34.4% vs.
21.5%, P < 0.001) was elevated among patients who developed POAF relative to those who did not
[Supplementary Table 3].

POAF was also associated with significantly increased resource utilization. On univariate analysis
postoperative length of stay (7.0 days vs. 5.0 days, P < 0.001), intensive care unit (ICU) length of stay (67.0 h
vs. 46.0 h, P < 0.001), discharge to facility (29.2% vs. 16.6%, P < 0.001), and readmission (11.7% vs. 8.15%, P <
0.001) were all significantly increased among patients with POAF vs. those without POAF. Total hospital
costs were significantly increased for patients who developed POAF, relative to those who did not ($42,324
vs. $36,682, P < 0.001).

Propensity matched baseline and operative characteristics
A total of 5926 pairs of patients were matched between POAF and no POAF patients. These were well
matched with all baseline covariates having a standardized mean difference of < 10% [Table 1,
Supplementary Figures 1 and 2]. Median predicted risk of mortality was similar between both groups of
matched patients (1.32% vs. 1.29%, P = 0.505).

Risk-adjusted outcomes and resource utilization
After risk adjustment, there were no significant differences between groups for postoperative morbidity or
mortality [Table 2]. However, postoperative atrial fibrillation continued to be associated with increased
resource utilization, including transfusion, increased postoperative length of stay, ICU length of stay,
discharge to facilities, and readmission.

The unpaired cost comparisons are seen in [Table 2], with visual representation of mean total and sub-
group costs in [Figure 1]. On paired univariate analysis the mean additional cost (95% confidence interval)
associated with postoperative atrial fibrillation was $6705 (5568-7842) in total hospital cost, $3159 (2720-
3598) for total stay cost, $699 (545-854) for diagnostic cost, $799 (391-1207) for intervention cost, $1971
(1599-2342) for general care cost, and $77 (2-152) for other costs. All paired differences were statistically
significant at P < 0.0001 except for other cost (P = 0.047). Cost estimates over times did not demonstrate
either an upward or downward trend [Figure 2].