Table 2. Characteristics of the Cohort With Hospital-Onset Sepsis Before and After Being Rebalanced for Treatment and Observable Characteristicsa.
| Characteristic | Before rebalancing | After rebalancingb | ||||
|---|---|---|---|---|---|---|
| SEP-1 | Standardized difference | SEP-1 | Standardized difference | |||
| Received (n = 281) | Did not receive (n = 2015) | Received (n = 274) | Did not receive (n = 1790) | |||
| Time 0 of sepsis, mean (SD), hc | 51.0 (105.2) | 73.3 (137.2) | −0.18d | 52.0 (106.4) | 51.2 (91.4) | 0.01 |
| Age, mean (SD), y | 61.9 (17.4) | 61.7 (18.1) | 0.01 | 62.1 (17.2) | 62.7 (16.7) | −0.03 |
| Male sex, No. (%) | 163 (58.0) | 1116 (55.4) | −0.05 | 159 (58.0) | 1035 (57.8) | −0.004 |
| Elixhauser comorbidities, mean (SD), No. | 5.0 (2.9) | 5.0 (3.1) | −0.01 | 5.0 (2.8) | 4.8 (2.5) | 0.04 |
| Postoperative status, No. (%) | 30 (10.6) | 363 (18.1) | −0.21d | 30 (10.9) | 231 (12.9) | −0.06 |
| Present at admission, No. (%) | ||||||
| Pneumoniae | 74 (26.3) | 394 (19.5) | 0.16d | 72 (26.2) | 470 (26.3) | 0 |
| Urinary tract infection | 55 (19.6) | 330 (16.3) | 0.08 | 54 (19.7) | 353 (19.7) | 0 |
| Skin or soft-tissue infection | 23 (8.2) | 109 (5.3) | 0.12d | 22 (8.0) | 110 (6.1) | 0.08 |
| New bacteremia, No. (%)f | 26 (9.3) | 233 (11.7) | −0.08 | 26 (9.5) | 200 (11.2) | −0.06 |
| Immunosuppression, No. (%) | 66 (23.5) | 618 (30.7) | −0.16d | 66 (24.1) | 618 (27.0) | −0.07 |
| Fever, No. (%) | 95 (33.8) | 351 (17.6) | 0.38d | 90 (32.8) | 588 (32.8) | 0 |
| Septic shock, No. (%) | 94 (33.5) | 759 (37.9) | −0.09 | 91 (33.2) | 584 (32.6) | 0.01 |
| Hypotension, No. (%) | 106 (37.7) | 1169 (57.9) | −0.41d | 102 (37.2) | 703 (39.3) | −0.04 |
| Elevated serum lactate level, No. (%)g | 63 (22.4) | 135 (6.7) | 0.45d | 63 (23.0) | 167 (9.3) | 0.40d |
| Hospital, No. (%) | ||||||
| 1 | 27 (9.6) | 251 (12.6) | −0.09 | 24 (8.8) | 157 (8.8) | 0 |
| 2 | 67 (23.8) | 412 (20.4) | 0.08 | 67 (24.5) | 372 (20.8) | 0.09 |
| 3 | 84 (29.9) | 848 (42.0) | −0.25d | 83 (30.3) | 608 (33.9) | −0.08 |
| 4 | 103 (36.7) | 504 (25.1) | 0.25d | 100 (36.5) | 653 (36.5) | 0 |
| Outcomes | ||||||
| In-hospital mortality, No. (%) | 54 (19.2) | 411 (20.6) | −0.04 | 53 (19.7) | 339 (19.5) | 0.01 |
| Vasopressor days, mean (SD), d | 2.3 (3.3) | 2.4 (3.3) | −0.03 | 2.3 (3.3) | 2.1 (3.1) | 0.06 |
| Length of stay, mean (SD), dh | 15.7 (14.4) | 18.9 (18.9) | −0.19d | 15.7 (14.8) | 17.0 (17.3) | −0.08 |
Abbreviation: SEP-1, Early Management Bundle for Severe Sepsis/Septic Shock.
SI conversion factor: To convert plasma lactate level to mg/dL, divide by 0.111.
Community-onset sepsis was defined by time 0 in the emergency department.
In the rebalanced cohort, SEP-1 recipients and nonrecipients were matched and reweighted on observable characteristics using propensity scores and Mahalanobis distance between covariates.
Hours from arrival to the hospital.
Difference was significant at α = .05.
Source of infection present at admission was determined by International Classification of Diseases, Ninth Revision and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision diagnosis codes as appropriate per time period.
New bacteremia was defined by positive blood culture results after time 0 of sepsis in a patient who did not have a positive blood culture result in the preceding week. Possible skin contaminants were excluded.
Elevated serum lactate level was defined as a value of more than 20 mmol/L at onset of sepsis or in the preceding 6 hours. To meet the definition of shock, the level serum lactate was 4 mmol/L or greater.
Length of stay was measured from time of onset of sepsis until hospital discharge.