Table 2. Characteristics and Predictors of LTAC vs SNF Transfer Among Hospitalized Older Medicare Beneficiaries.
| Characteristica | LTAC (n = 3093) | SNF (n = 62 432) | Univariate ORb (95% CI) | Adjusted ORc (95% CI) |
|---|---|---|---|---|
| Patient Factors Prior to Hospitalization | ||||
| Age, y | ||||
| 65-69 | 567 (18.3) | 5435 (8.7) | 1 [Reference] | 1 [Reference] |
| 70-74 | 694 (22.4) | 7881 (12.6) | 0.83 (0.73-0.94) | 0.87 (0.74-1.02) |
| 75-79 | 622 (20.1) | 10 418 (16.7) | 0.56 (0.49-0.64) | 0.76 (0.65-0.90) |
| 80-84 | 577 (18.7) | 13 775 (22.1) | 0.39 (0.34-0.44) | 0.62 (0.53-0.73) |
| ≥85 | 633 (20.5) | 24 923 (39.9) | 0.23 (0.20-0.26) | 0.51 (0.43-0.59) |
| Female | 1621 (52.4) | 40 840 (65.4) | 0.56 (0.52-0.61) | 0.79 (0.72-0.87) |
| Race/ethnicity | ||||
| White | 2337 (75.6) | 54 765 (87.7) | 1 [Reference] | 1 [Reference] |
| Black | 516 (16.7) | 5284 (8.5) | 1.87 (1.66-2.10) | 1.33 (1.15-1.54) |
| Other | 240 (7.8) | 2383 (3.8) | 1.46 (1.24-1.73) | 1.29 (1.05-1.57) |
| Prior hospitalizations | ||||
| 0 | 1264 (40.8) | 35 522 (56.9) | 1 [Reference] | 1 [Reference] |
| 1 | 785 (25.4) | 14 736 (23.6) | 1.47 (1.34-1.62) | 1.31 (1.17-1.48) |
| ≥2 | 1044 (33.8) | 12 174 (19.5) | 2.27 (2.07-2.49) | 1.59 (1.40-1.80) |
| Prior LTAC use | 222 (7.2) | 378 (0.6) | 5.78 (4.73-7.06) | 3.97 (3.12-5.06) |
| Prior SNF use | 530 (17.1) | 12 233 (19.6) | 0.77 (0.70-0.85) | 0.51 (0.44-0.58) |
| Wheelchair | 477 (15.4) | 6332 (10.1) | 1.41 (1.26-1.57) | 1.23 (1.07-1.42) |
| Home hospital bed | 313 (10.1) | 3201 (5.1) | 1.77 (1.54-2.02) | 1.19 (1.00-1.42) |
| Home oxygen | ||||
| None | 2420 (78.2) | 54 828 (87.8) | 1 [Reference] | 1 [Reference] |
| Transient (≥1 but <12 mo of use) | 447 (14.5) | 4846 (7.8) | 2.19 (1.95-2.46) | 1.49 (1.29-1.73) |
| Chronic (use for prior 12 mo) | 226 (7.3) | 2758 (4.4) | 2.02 (1.73-2.36) | 1.34 (1.11-1.62) |
| Patient Factors of Index Hospitalization | ||||
| Length of hospital stay,d days | 12 (7-19) | 6 (4-9) | 1.35 (1.33-1.36) | 1.09 (1.07-1.10) |
| Prolonged ICU stay (≥3 d) | 1855 (60.0) | 17 769 (28.5) | 4.64 (4.25-5.06) | 1.15 (1.02-1.29) |
| Primary diagnosis | ||||
| DRG resource intensity weight, per unit | 2.1 (1.5-5.5) | 1.5 (1.0-2.1) | 1.56 (1.53-1.58) | 1.07 (1.04-1.11) |
| DRG with an MCC designation | 1629 (52.7) | 20 285 (32.5) | 2.30 (2.13-2.49) | 1.65 (1.48-1.85) |
| Respiratory MDC | 481 (15.6) | 7464 (12.0) | 1.30 (1.16-1.45) | 1.64 (1.43-1.88) |
| Circulatory MDC | 367 (11.9) | 9189 (14.7) | 0.77 (0.70-0.86) | 0.70 (0.60-0.81) |
| Musculoskeletal MDC | 230 (7.4) | 17 633 (28.2) | 0.22 (0.19-0.25) | 0.59 (0.50-0.70) |
| Secondary diagnosese | ||||
| Respiratory failure | 1294 (41.8) | 7068 (11.3) | 7.14 (6.54-7.79) | 1.38 (1.21-1.58) |
| Sepsis | 986 (31.9) | 5370 (8.6) | 5.49 (5.01-6.03) | 1.77 (1.56-2.00) |
| Skin, soft-tissue, or joint infection | 454 (14.7) | 2897 (4.6) | 3.93 (3.48-4.44) | 2.35 (2.02-2.74) |
| Chronic skin ulcer | 655 (21.2) | 5859 (9.4) | 2.77 (2.51-3.07) | 1.88 (1.65-2.14) |
| Device, graft, implant complication | 230 (7.4) | 1424 (2.3) | 3.92 (3.33-4.62) | 1.41 (1.14-1.73) |
| Complication of care | 495 (16.0) | 3718 (6.0) | 3.67 (3.27-4.11) | 1.20 (1.02-1.40) |
| Delirium or dementia | 599 (19.4) | 18 099 (29.0) | 0.48 (0.43-0.52) | 0.71 (0.63-0.80) |
| Selected intensive treatments and procedures | ||||
| Mechanical ventilation | ||||
| None | 2163 (69.9) | 60 394 (96.7) | 1 [Reference] | 1 [Reference] |
| Transient (<96 h) | 234 (7.6) | 1357 (2.2) | 5.89 (4.92-6.99) | 1.41 (1.15-1.74) |
| Prolonged (≥96 h) | 696 (22.5) | 681 (1.1) | 51.0 (43.9-59.3) | 2.50 (1.98-3.17) |
| Tracheostomy | 520 (16.8) | 140 (0.2) | 248.5 (194-319) | 23.8 (15.8-35.9) |
| Total parenteral nutrition | 195 (6.3) | 686 (1.1) | 8.01 (6.60- 9.71) | 2.20 (1.73-2.80) |
| Central venous line | 991 (32.0) | 4307 (6.9) | 7.63 (6.93-8.41) | 2.01 (1.78-2.28) |
| Dialysis | 454 (14.7) | 2277 (3.6) | 4.56 (4.03-5.17) | 1.59 (1.36-1.87) |
| Excisional debridement | 101 (3.3) | 376 (0.6) | 7.89 (6.05-10.27) | 3.49 (2.55-4.78) |
| Hospital Factors | ||||
| For profit ownership | 708 (22.9) | 8350 (13.4) | 1.30 (1.11-1.52) | 1.36 (1.14-1.62) |
| Bed size (ORs per 50 beds) | 326 (201-526) | 313 (177-493) | 1.04 (1.03-1.05) | 0.96 (0.95-0.98) |
| Disproportionate share index (ORs per 0.10)f | 0.29 (0.20-0.39) | 0.23 (0.17-0.32) | 1.21 (1.16-1.26) | 1.12 (1.07-1.18) |
| Regional Factors | ||||
| Linear arc distance to nearest LTAC, miles | ||||
| >42 | 100 (3.2) | 11 302 (18.1) | 1 [Reference] | 1 [Reference] |
| 18-42 | 231 (7.5) | 12 316 (19.7) | 2.00 (1.49-2.68) | 1.86 (1.34-2.58) |
| 7-18 | 502 (16.2) | 12 538 (20.1) | 4.11(3.11-5.44) | 4.03 (2.93-5.54) |
| 2-7 | 1006 (32.5) | 14 444 (23.1) | 6.74 (5.13-8.87) | 6.10 (4.45-8.36) |
| 0-2 | 1254 (40.5) | 11 832 (19.0) | 9.04 (6.88-11.9) | 8.40 (6.12-11.50) |
| State without Certificate of Need law | 1817 (58.7) | 22 065 (35.3) | 1.80 (1.38-2.36) | 1.54 (1.21-1.95) |
| LTAC supply, beds/100K persons (ORs per 5) | 10.8 (7.3-21.5) | 6.9 (2.6-11.4) | 1.40 (1.32-1.49) | 1.17 (1.10-1.23) |
| Median household income, per $10 000g | 5.2 (4.5-6.1) | 5.4 (4.8-6.2) | 0.76 (0.67-0.86) | 0.64 (0.57-0.71) |
| Medicare EOL PAC spending, per $1000 | 14.9 (11.5-19.2) | 12.3 (10.1-15.0) | 1.21 (1.16-1.26) | 1.10 (1.06-1.15) |
Abbreviations: DRG, diagnosis-related group; EOL, end of life; HRR, hospital referral region; ICU, intensive care unit; LTAC, long-term acute care hospital; MCC, major complication or comorbidity; MDC, Major Diagnostic Category; OR, odds ratio; PAC, postacute care; SNF, skilled nursing facility.
a
Categorical data are shown as number (percentage), and numerical data are shown as median (interquartile range).
b
Unadjusted ORs are from univariable multilevel mixed-effects logistic regression models including random effects for hospital and region.
c
Adjusted ORs are from a multilevel mixed-effects logistic regression model including all predictors listed in the Table.
d
Length of stay was Winsorized at 4 and 14 days based on functional form analysis using restricted cubic splines.
e
Categorized using the Agency for Healthcare Research and Quality Clinical Classification Software.
f
Disproportionate share index is a measure of the share of low-income patients served by the hospital.
g
Median household income was obtained from the US Census Bureau and aggregated at the HRR level.