Introduction
Materials and methods
Case Mix Programme Database
Selection of cases
Data
Case mix
Outcome
Activity
Analyses
Descriptive statistics
Modelling ultimate hospital mortality
Comparison of existing mortality prediction models
Results
Data
Category | Parameter | All patients (n = 17,326) | Oligurica (n = 5,687) | Nonoligurica (n = 10,133) |
---|---|---|---|---|
Case mix | Age (years; mean ± SD) | 63.2 ± 15.6 | 63.5 ± 15.0 | 63.1 ± 15.7 |
Sex (male; n [%]) | 11,511 (66.4) | 3,617 (63.6) | 6,925 (68.3) | |
Surgical status (n [%]) | ||||
Nonsurgical | 14,479 (83.7) | 5,005 (88.1) | 8,125 (80.3) | |
Elective surgery | 961 (5.6) | 187 (3.3) | 717 (7.1) | |
Emergency surgery | 1,868 (10.8) | 487 (8.6) | 1,280 (12.6) | |
Serum parameters (mean ± SD) | ||||
Highest serum creatinine (μmol/l) | 459 ± 232 | 502 ± 244 | 431 ± 216 | |
Highest serum urea (mmol/l) | 32.2 ± 27.3 | 29.5 ± 15.2 | 33.9 ± 32.3 | |
Lowest serum albumin (g/l) | 21.3 ± 7.2 | 20.5 ± 6.9 | 21.6 ± 7.2 | |
Lowest haematocrit (%) | 28.6 ± 6.4 | 28.1 ± 6.3 | 28.8 ± 6.3 | |
APACHE II (mean ± SD) | ||||
APSb | 20.2 ± 7.0 | 24.6 ± 6.5 | 17.9 ± 6.1 | |
Total scoreb | 24.8 ± 7.5 | 29.3 ± 7.0 | 22.4 ± 6.5 | |
Outcome | Mortality (n [%]) | |||
CMP unit | 7,508 (43.3) | 3,176 (55.8) | 3,387 (33.4) | |
Any hospitalc | 9,725 (58.6) | 3,850 (70.3) | 4,770 (49.3) | |
Activity | Length of stay (days; median [IQR]) | |||
CMP unit: survivors | 4.1 (1.7–10.3) | 5.7 (2.1–13.2) | 4.0 (1.8–9.8) | |
CMP unit: nonsurvivors | 2.0 (0.8–6.1) | 1.7 (0.9–5.0) | 3.4 (1.5–8.9) | |
Any hospitalc: survivors | 31 (17–54) | 38 (24–62) | 29 (16–51) | |
Any hospitalc: nonsurvivors | 8 (3–20) | 7 (2–18) | 11 (4–23) | |
Transfers from another ICU (n [%]) | 1,355 (7.8) | 561 (9.9) | 730 (7.2) | |
Readmissions within hospital stay (n [%]) | 388 (2.2) | 116 (2.0) | 221 (2.2) |
Case mix
Outcome
Activity
Relationship of case mix factors with ultimate hospital mortality
Case mix factor | Deaths (n) | n | % | Adjusted OR (95% CI) | P value | |
---|---|---|---|---|---|---|
Age (years) | < 45 | 817 | 2,128 | 38.4 | 1.48 (1.44–1.52) per 10 year increase | < 0.001 |
45–54 | 846 | 1,776 | 47.6 | |||
55–64 | 1,671 | 3,086 | 54.1 | |||
65–74 | 2,991 | 4,866 | 61.5 | |||
75+ | 2,553 | 3,720 | 68.6 | |||
Sex | Female | 2,905 | 5,217 | 55.7 | Reference | 0.004 |
Male | 5,973 | 10,359 | 57.7 | 1.13 (1.04–1.22) | ||
Past medical history | Absent | 7,355 | 13,298 | 55.3 | Reference | < 0.001 |
Present | 1,523 | 2,278 | 66.9 | 1.82 (1.63–2.03) | ||
CPR before admission | No | 7,679 | 14,062 | 54.6 | Reference | < 0.001 |
Yes | 1,199 | 1,514 | 79.2 | 1.64 (1.41–1.90) | ||
Mechanical ventilation | No | 2,086 | 5,250 | 39.7 | Reference | < 0.001 |
Yes | 6,792 | 10,326 | 65.8 | 1.93 (1.76–2.12) | ||
Surgery within 1 week of admission | No | 6,324 | 10,751 | 58.8 | Reference | < 0.001 |
Yes | 2,554 | 4,825 | 52.9 | 0.80 (0.73–0.87) | ||
Sepsis | No | 4,362 | 8,207 | 53.1 | Reference | < 0.001 |
Yes | 4,516 | 7,369 | 61.3 | 1.14 (1.06–1.21) | ||
Oliguria | No | 4,766 | 9,660 | 49.3 | Reference | < 0.001 |
Yes | 3,844 | 5,466 | 70.3 | 1.90 (1.75–2.07) | ||
Length of stay before unit admission | 0 days | 2,451 | 4,261 | 57.5 | Reference | < 0.001 |
1 day | 1,952 | 3,666 | 53.2 | 0.90 (0.81–1.01) | ||
2–6 days | 2,269 | 4,133 | 54.9 | 1.00 (0.90–1.12) | ||
7+ days | 2,199 | 3,508 | 62.7 | 1.54 (1.38–1.73) | ||
Temperature (°C)a | < 32 | 68 | 89 | 76.4 | 1.96 (1.12–3.40) | < 0.001 |
32–34 | 399 | 557 | 71.6 | 1.32 (1.05–1.65) | ||
34–36 | 2,779 | 4,595 | 60.5 | 1.11 (1.01–1.22) | ||
36–38.5 | 2,918 | 5,789 | 50.4 | Reference | ||
38.5–39 | 882 | 1,680 | 52.5 | 0.96 (0.85–1.10) | ||
39–41 | 1,614 | 2,595 | 62.2 | 1.16 (1.03–1.30) | ||
≥ 41 | 185 | 217 | 85.3 | 3.35 (2.18–5.15) | ||
Mean arterial pressure (mmHg)a | < 50 | 2,923 | 3,803 | 76.9 | 2.08 (1.80–2.41) | < 0.001 |
50–70 | 4,164 | 7,013 | 59.4 | 1.45 (1.28–1.65) | ||
70–110 | 634 | 1,682 | 37.7 | Reference | ||
110–130 | 724 | 1,988 | 36.4 | 0.89 (0.77–1.04) | ||
130–160 | 366 | 955 | 38.3 | 0.94 (0.78–1.13) | ||
≥ 160 | 48 | 96 | 50.0 | 1.33 (0.83–2.12) | ||
Heart rate (beats/minute)b | < 40 | 134 | 187 | 71.7 | 1.42 (0.97–2.08) | < 0.001 |
40–50 | 179 | 324 | 55.2 | 1.14 (0.87–1.50) | ||
50–100 | 1,229 | 2,775 | 44.3 | Reference | ||
100–110 | 1,044 | 2,085 | 50.1 | 1.18 (1.03–1.35) | ||
110–120 | 1,287 | 2,408 | 53.4 | 1.26 (1.11–1.44) | ||
120–140 | 2,543 | 4,251 | 59.8 | 1.46 (1.30–1.64) | ||
140–155 | 1,262 | 1,878 | 67.2 | 1.66 (1.43–1.93) | ||
≥ 155 | 1,173 | 1,623 | 72.3 | 1.94 (1.66–2.28) | ||
Respiratory rate (breaths/minute)b | < 6 | 403 | 628 | 64.2 | 1.05 (0.85–1.30) | 0.001 |
6–12 | 1,836 | 3,509 | 52.3 | 1.00 (0.88–1.13) | ||
12–14 | 1,771 | 2,867 | 61.8 | 1.19 (1.05–1.36) | ||
14–25 | 1,395 | 2,397 | 58.2 | Reference | ||
25–35 | 2,007 | 3,784 | 53.0 | 1.04 (0.92–1.18) | ||
35–40 | 689 | 1,109 | 62.1 | 1.33 (1.12–1.58) | ||
40–50 | 608 | 1,015 | 59.9 | 1.16 (0.97–1.39) | ||
≥ 50 | 135 | 203 | 66.5 | 1.49 (1.05–2.13) | ||
Oxygenation (mmHg; A-aDO2 [FiO2 ≥0.5])b | A-aDO2 (FIO2 ≥ 0.5) | < 0.001 | ||||
< 250 | 1,238 | 2,338 | 53.0 | Reference | ||
250–350 | 1,576 | 2,540 | 62.0 | 1.25 (1.12–1.40) | ||
350–500 | 1,034 | 1,453 | 71.2 | 1.68 (1.45–1.94) | ||
≥ 500 | 1,480 | 1,951 | 75.9 | 1.94 (1.69–2.23) | ||
PaO2 (FIO2 < 0.5) | ||||||
< 50 | 114 | 206 | 55.3 | 1.10 (0.80–1.53) | ||
50–70 | 918 | 1,786 | 51.4 | 1.11 (0.98–1.26) | ||
70–80 | 786 | 1,637 | 48.0 | 1.02 (0.90–1.16) | ||
≥ 80 | 1,732 | 3,665 | 47.3 | Reference | ||
Arterial pHa | < 7.15 | 1,380 | 1,676 | 82.3 | 2.13 (1.82–2.49) | < 0.001 |
7.15–7.25 | 1,871 | 2,655 | 70.5 | 1.63 (1.46–1.83) | ||
7.25–7.33 | 2,177 | 3,744 | 58.1 | 1.20 (1.09–1.32) | ||
7.33–7.5 | 3,290 | 7,156 | 46.0 | Reference | ||
≥ 7.5 | 160 | 345 | 46.4 | 0.98 (0.76–1.26) | ||
Serum sodium (mmol/l)a | < 120 | 187 | 326 | 57.4 | 1.49 (1.14–1.94) | < 0.001 |
120–130 | 1,506 | 2,654 | 56.7 | 1.19 (1.07–1.32) | ||
130–150 | 6,590 | 11,687 | 56.4 | Reference | ||
150–155 | 358 | 537 | 66.7 | 1.45 (1.17–1.80) | ||
155–160 | 153 | 221 | 69.2 | 1.48 (1.06–2.07) | ||
≥ 160 | 84 | 151 | 55.6 | 0.92 (0.63–1.36) | ||
Serum potassium (mmol/l)a | < 2.5 | 56 | 110 | 50.9 | 0.67 (0.43–1.05) | < 0.001 |
2.5–3 | 302 | 535 | 56.4 | 0.93 (0.75–1.14) | ||
3–3.5 | 1,106 | 2,008 | 55.1 | 0.91 (0.81–1.02) | ||
3.5–5.5 | 4,639 | 8,297 | 55.9 | Reference | ||
5.5–6 | 1,137 | 1,858 | 61.2 | 1.17 (1.04–1.32) | ||
6–7 | 1,244 | 2,028 | 61.3 | 1.13 (1.00–1.27) | ||
≥ 7 | 394 | 740 | 53.2 | 0.79 (0.66–0.95) | ||
Serum albumin (g/l)b | < 20 | 3,544 | 5,499 | 64.4 | 1.21 (1.11–1.33) | < 0.001 |
20–25 | 1,834 | 3,356 | 54.6 | 1.02 (0.93–1.13) | ||
25–45 | 3,469 | 6,635 | 52.3 | Reference | ||
≥ 45 | 31 | 86 | 36.0 | 0.55 (0.33–0.93) | ||
Haematocrit (%)a | < 20 | 612 | 1,008 | 60.7 | 0.99 (0.84–1.17) | 0.011 |
20–30 | 4,748 | 8,536 | 55.6 | 0.89 (0.82–0.97) | ||
30–46 | 3,168 | 5,473 | 57.9 | Reference | ||
46–50 | 218 | 335 | 65.1 | 1.28 (0.98–1.68) | ||
≥ 50 | 132 | 224 | 58.9 | 0.99 (0.71–1.38) | ||
White blood count (×1,000/mm3)a | < 1 | 176 | 223 | 78.9 | 2.52 (1.73–3.67) | < 0.001 |
1–3 | 506 | 680 | 74.4 | 1.77 (1.44–2.18) | ||
3–15 | 3,862 | 7,468 | 51.7 | Reference | ||
15–20 | 1,756 | 3,071 | 57.2 | 1.11 (1.00–1.23) | ||
20–40 | 2,278 | 3,681 | 61.9 | 1.21 (1.09–1.33) | ||
≥ 40 | 300 | 453 | 66.2 | 1.16 (0.91–1.47) | ||
Glasgow Coma Scale scored | 3–4 | 1,596 | 2,019 | 79.0 | 1.07 (1.06–1.08) per decrease of 1 point | < 0.001 |
5–12 | 1,178 | 1,924 | 61.2 | |||
13–14 | 1,099 | 1,880 | 58.5 | |||
15 | 5,005 | 9,753 | 51.3 |
Comparison of existing mortality prediction models
SHARF T0 | SHARF II0 | Mehta | UK APACHE II | |
---|---|---|---|---|
Discrimination (AUC [95% CI]) | 0.632 (0.624–0.640) | 0.668 (0.660–0.676) | 0.693 (0.686–0.701) | 0.74 (0.733–0.748) |
Calibration (mortality ratio [95% CI]) | 0.759 (0.747–0.771) | 0.950 (0.936–0.965) | 1.223 (1.205–1.242) | 1.147 (1.129–1.165) |
Hosmer-Lemeshow (C-statistic [P value]) | 20,894 (< 0.001) | 2,575 (< 0.001) | 2,170 (< 0.001) | 575 (< 0.001) |
Cox's regression calibration: | ||||
Slope (95% CI) | 0.247 (0.229–0.266) | 0.415 (0.390–0.439) | 0.444 (0.417–0.470) | 0.753 (0.720–0.787) |
Intercept (95% CI) | -0.195 (-0.243 to -0.153) | 0.005 (-0.031 to -0.040) | 0.314 (0.278–0.350) | 0.306 (0.269–0.343) |
χ2 (1) (P value)a | 11,598 (< 0.001) | 2,303 (< 0.001) | 2,602 (< 0.001) | 596 (< 0.001) |
Fit (Shapiro's Q [R statistic]) | 0.392 | 0.491 | 0.487 | 0.539 |
Discussion
Conclusion
Key messages
-
Severe AKI is responsible for about 10% of all ICU bed-days, and the presence of AKI doubles ICU and overall hospital lengths of stay.
-
In adult, general critical care units in England, Wales and Northern Ireland, the mean age of patients with AKI in ICU is increasing and nonsurgical admissions now predominate, representing a change from the mid-1980s, when surgical conditions dominated.
-
Although all severity of illness models tested have problems that lead to reduced performance, the UK APACHE II model when used in UK ICUs performs best.
-
Traditional risk factors such as oliguria, CPR, mechanical ventilation, sepsis, length of stay more than 7 days before admission to ICU and extremes of physiology continue to be associated with the greatest increased risks for mortality.