Skip to main content
. 2019 Jun 17;179(9):1176–1185. doi: 10.1001/jamainternmed.2019.1051

Table 3. Outcomes by Risk-Standardized Quartiles of Echocardiography Use.

Variable Unadjusted Outcomes Fully Adjusted Outcome Effect Size Estimate, OR (95% CI)a
Quartile 1 Quartile 2 Quartile 3 Quartile 4 P Value (Quartile 4 vs Quartile 1) Quartile 1 Quartile 2 Quartile 3 Quartile 4 P Value (Quartile 4 vs Quartile 1) Overall P Value
Length of stay, db,c 4.3 (4.0) 4.6 (4.3) 4.6 (4.3) 4.7 (4.3) .03 1 [Reference] 0.12 (−0.06 to 0.30) 0.17 (−0.01 to 0.36) 0.23 (0.04 to 0.41) .01 .09
Total cost, $c,d 15 747 (14 282) 17 105 (15 817) 19 633 (17 807) 18 737 (16 734) <.001 1 [Reference] 842 (−467 to 2150) 3232 (1912 to 4551) 3164 (1843 to 4485) <.001 <.001
In-hospital mortalitye,f 919 (4.6) 1513 (5.1) 1019 (4.8) 1162 (4.6) .79 1 [Reference] 1.15 (0.99 to 1.34) 1.09 (0.94 to 1.27) 1.02 (0.88 to 1.19) .80 .22
3-mo Readmission among survivorsf 3599 (18.2) 5191 (18.0) 4201 (19.9) 4405 (17.9) .40 1 [Reference] 1.01 (0.93 to 1.10) 1.16 (1.06 to 1.27) 1.00 (0.92 to 1.09) .80 .001

Abbreviation: OR, odds ratio.

a

Hierarchical generalized linear models accounted for patient clustering within hospitals and adjusted for patient demographics, comorbidities, acute organ failures, hospital characteristics, and hospital interventional capabilities.

b

Winsorized at the 99th percentile.

c

Expressed as mean (SD) in unadjusted outcomes.

d

Winsorized at the first and 99th percentiles.

e

One admission was randomly selected.

f

Expressed as No. (%) of patients in unadjusted outcomes.