Table. Description of the Underlying Population-Weighted Cohort and the Cohort Following Inverse Probability of Treatment Weighting for CABG vs PCIa.
| Cohort after population weighting | Inverse probability-weighted cohort | |||||
|---|---|---|---|---|---|---|
| No. (%) | Difference, % (95% CI) | % | Difference, % (95% CI) | |||
| CABG | PCI | CABG | PCI | |||
| Sample No. | 665 | 1015 | 665 | 1015 | ||
| Sample No. (complete case)b | 652 | 1001 | NA | 649 | 997 | NA |
| Demographics | ||||||
| Age at procedure, mean (SD), y | 74.8 (6.6) | 75.0 (7.2) | −0.2 (−1.0 to 0.5) | 74.8 (5.3) | 74.7 (6.9) | 0.1 (−0.7 to 1.0) |
| Sex | ||||||
| Male | 446 (65.8) | 556 (54.4) | 11.4 (6.2 to 16.7) | 58.7 | 58.3 | 0.4 (−6.1 to 7.0) |
| Female | 219 (34.2) | 459 (45.6) | 41.3 | 41.7 | ||
| Race/ethnicity | ||||||
| Non-Hispanic | ||||||
| White | 572 (89.4) | 829 (85.5) | 3.9 (−0.4 to 8.3) | 86.5 | 86.8 | −0.3 (−5.9 to 5.4) |
| Black | 41 (3.9) | 98 (6.2) | −2.3 (−4.3 to 0.3) | 5.8 | 5.5 | 0.2 (−3.1 to 3.6) |
| Hispanic | 45 (5.5) | 61 (5.7) | −0.2 (−3.5 to 3.1) | 5.8 | 5.7 | 0.1 (−3.4 to 3.6) |
| Otherc | 7 (1.2) | 27 (2.7) | −1.5 (−3.1 to 0.1) | 1.9 | 1.9 | −0.1 (−2.4 to 2.2) |
| Education | ||||||
| <High school/GED | 220 (30.8) | 329 (30.3) | 0.5 (−5.5 to 6.4) | 28.4 | 28.3 | 0.2 (−7.2 to 7.6) |
| High school | 200 (30.8) | 343 (34.3) | −3.5 (−8.5 to 1.5) | 32.8 | 32.7 | 0.1 (−6.5 to 6.7) |
| Some college | 113 (18.5) | 192 (19.8) | −1.3 (5.6 to 3.1) | 20.1 | 20.9 | −0.8 (5.7 to 4.2) |
| >College | 132 (19.9) | 151 (15.5) | 4.3 (0 to 8.7) | 18.7 | 18.2 | 0.5 (−5.2 to 6.2) |
| Wealth greater than cohort median | 298 (56.5) | 507 (49.7) | 6.9 (1.2 to 12.5) | 46.1 | 46.3 | 0.2 (−6.2 to 6.7) |
| Married/partnered | 479 (69.4) | 647 (59.3) | 10.1 (4.7 to 15.5) | 63.6 | 63.0 | 0.6 (−6.3 to 7.4) |
| Health covariates | ||||||
| BMI | ||||||
| <18.5 | 4 (0.8) | 9 (0.9) | −0.1 (−1.0 to 0.8) | 0.8 | 0.8 | 0.04 (−1.3 to 1.4) |
| 18.5-24.9 | 187 (27.3) | 277 (26.9) | 0.4 (−4.8 to 5.6) | 27.2 | 26.5 | 0.8 (−4.6 to 6.1) |
| 25-29.9 | 297 (47.0) | 452 (45.5) | 1.5 (−4.4 to 7.4) | 46.4 | 46.8 | −0.4 (−6.4 to 5.6) |
| ≥30 | 172 (24.9) | 272 (26.7) | −1.8 (−6.8 to 3.2) | 25.6 | 26.0 | −0.4 (−5.9 to 5.1) |
| Currently smoking | 89 (13.0) | 137 (15.7) | −2.7 (−5.6 to 0.2) | 12.4 | 12.9 | −0.6 (−5.4 to 4.3) |
| Hypertension | 457 (68.3) | 676 (66.1) | 2.2 (−2.9 to 7.4) | 64.2 | 64.7 | −0.5 (−6.9 to 5.9) |
| Prior diagnosis of heart problems | 300 (45.7) | 476 (46.5) | −0.8 (−7.1 to 5.6) | 44.1 | 43.7 | 0.3 (−7.1 to 7.9) |
| Diabetes | 211 (31.4) | 314 (30.6) | 0.8 (−4.1 to 5.7) | 28.2 | 28.6 | −0.5 (−7.0 to 6.1) |
| Moderate-severe paind | 169 (24.3) | 299 (30.0) | −5.7 (−11.0 to −0.5) | 27.0 | 27.5 | −0.5 (−6.8 to 5.7) |
| Depressive symptomse | 129 (21.0) | 243 (24.9) | −3.9 (−8.8 to 1.0) | 20.7 | 21.1 | −0.3 (−5.6 to 4.9) |
| Prior stroke | 88 (13.3) | 146 (13.8) | −0.6 (−4.0 to 2.9) | 13.2 | 13.4 | −0.2 (−4.6 to 4.2) |
| Lung disease | 54 (7.8) | 135 (14.8) | −7.0 (−10.7 to −3.2) | 8.4 | 10.4 | −2.0 (−5.4 to 1.4) |
| Difficulty with at least 1 activity of daily living | 126 (19.1) | 196 (19.2) | −0.1 (−5.1 to 4.9) | 16.9 | 17.5 | −0.6 (−6.4 to 5.2) |
| Interview data, mean (SD) | ||||||
| No. of interviews | 5.0 (1.9) | 5.0 (1.9) | 0 (−0.2 to 0.2) | 5.4 (1.3) | 5.5 (1.7) | −0.1 (−0.2 to 0.1) |
| No. of preprocedure interviews | 2.2 (0.7) | 2.2 (0.7) | −0.01 (−0.08 to 0.07) | 2.3 (0.5) | 2.3 (0.7) | −0.01 (−0.08 to 0.06) |
| No. of postprocedure interviews | 2.8 (1.8) | 2.8 (1.9) | −0.01 (−0.19 to 0.17) | 3.2 (1.3) | 3.2 (1.6) | −0.06 (−0.25 to 0.14) |
| Mean person-years of preprocedure data | 2.4 (1.4) | 2.3 (1.4) | −0.02 (−0.07 to 0.02) | 2.4 (1.2) | 2.4 (1.5) | 0.003 (−0.05 to 0.06) |
| Mean person-years of postprocedure data | 3.9 (2.7) | 3.9 (2.7) | 0.07 (−0.18 to 0.11) | 3.9 (2.4) | 4.0 (2.9) | −0.06 (−0.22 to 0.11) |
| Preprocedure memory score, mean (SD) | ||||||
| Preprocedure memory score, memory unitsf | 0.84 (0.49) | 0.82 (0.54) | 0.01 (−0.04 to 0.07) | 0.90 (0.35) | 0.89 (0.46) | 0.01 (−0.04 to 0.07) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CABG, coronary artery bypass grafting; GED, General Educational Development test; NA, not applicable; PCI, percutaneous coronary intervention.
The actual number of participants prior to weighting, and proportion of the sample after application of population-level survey weights, are reported in the cells.
Complete case No. includes those who were completed via multiple imputation, and best reflects the No. included in the final analytic models. The inverse probability of treatment weighted models required more variables (ie, those used in the CABG vs PCI propensity score, as well as the other main adjustment variables), hence the complete case No. is lower.
Other race is a composite of Asian, Pacific Islander, American Indian, and Alaska Native, which are masked as a single category to protect participant confidentiality.
Moderate-severe pain was determined at the preprocedure interview (as were all health covariates listed here) and was ascertained as a 2-part question: “Are you often troubled with pain?” and “Is the pain mild, moderate, or severe?” Participants who reported they were often troubled by moderate or severe pain met the definition for this health covariate.
Depressive symptoms were considered a score of 3 or greater on the revised 8-item Center for Epidemiologic Studies Depression scale (range, 0-8).
Memory score is a summary measure that uses direct response to immediate and delayed word recall or proxy cognition assessments for those who do not or cannot directly participate in cognitive assessments. The score is roughly analogous to a population cognitive z score and is normally distributed. A total of 95% of the study population memory score values were between −0.93 and 1.5, to offer an idea of range. Higher memory scores indicate better composite memory function.