<p>Outcome and resource use in critical care (electronic supplementary material) </p><p>Variability in outcome and resource use in intensive care units</p><p>Hans U. Rothen, Kay Stricker, Johanna Einfalt, Peter Bauer, Philip G.H. Metnitz, </p><p>Rui P. Moreno, Jukka Takala</p><p>Electronic supplementary material</p><p>--- 1 --- Outcome and resource use in critical care (electronic supplementary material) </p><p>Table 2E (electronic material): Key variables for structure and process </p><p>The total number of intensive care units included is 275; the total number of patients is 16,560. </p><p>Table 2E-a: Structure and staffing </p><p> mean ± SD median (25th – 75th percentile)</p><p>Number of staffed beds 10.94 ± 6.97 9.00 (7.00 - 12.00) </p><p>Physicians per bed 0.93 ± 0.77 0.81 (0.50 - 1.13) </p><p>Intensive care medicine specialists per bed 0.53 ± 0.67 0.37 (0.18 - 0.74) </p><p>Number of medical specialties in hospital 11.53 ± 4.11 13.00 (10.00 - 15.00) </p><p>Nurses per bed 3.44 ± 2.75 3.00 (2.33 - 3.88) </p><p>Physicians per nurse 0.33 ± 0.30 0.28 (0.17 - 0.39) </p><p>Data are mean ± SD and median (25th –75th percentile). </p><p>Table 2E-b: Hospital characteristics and clinical processes in the ICU </p><p> yes no yes no </p><p> n n % % </p><p>University hospital 122 141 46 54 </p><p>Presence of emergency department in hospital 239 24 91 9 </p><p>Multidisciplinary meetings in the ICU 121 136 47 53 </p><p>Clinical rounds (physicians + nursing staff) in the ICU 158 99 61 39 </p><p>Availability of physicians (weekdays) in the ICU 210 51 80 20 </p><p>Availability of physicians (nights/weekends) in the ICU 204 57 78 22 </p><p>Note: Due to missing data, the numbers do not sum up to n = 275. </p><p>--- 2 --- Outcome and resource use in critical care (electronic supplementary material) </p><p>Table 4E (electronic material): Estimated efficiency: Univariate logistic regression analysis of</p><p> explanatory variables. Stratification of groups by tertiles. </p><p>Variable OR 95% CI P </p><p>Number of staffed beds 1.05 0.97 – 1.13 0.231 </p><p>Physicians per bed 0.36 0.15 – 0.88 0.026 </p><p>Number of intensive care medicine specialists 0.95 0.84 – 1.06 0.362 </p><p>Intensive care medicine specialists per bed 0.28 0.08 – 0.98 0.047 </p><p>Number of medical specialties in hospital 1.01 0.92 – 1.11 0.852 </p><p>Nurses per bed 2.02 1.30 – 3.15 0.002 </p><p>Physicians per nurse 0.01 0.00 – 0.14 0.001 </p><p>Type of hospital 1.13 0.46 – 2.78 0.783 </p><p>Emergency department in hospital 7.16 0.59 – 86.62 0.122 </p><p>Multidisciplinary meetings 1.24 0.49 – 3.17 0.648 </p><p>Clinical rounds 4.07 1.54 – 10.75 0.005 </p><p>Physicians (weekdays) 0.22 0.06 – 0.80 0.022 </p><p>Physicians (nights/weekends) 0.34 0.10 – 1.23 0.101 </p><p>Region </p><p>1 1.94 0.11 – 34.92 0.655 </p><p>2 0.02 0.00 – 0.12 <0.001 </p><p>3 1.00 </p><p>4 0.11 0.01 – 1.43 0.09 </p><p>5 0.01 0.02 – 0.33 < 0.001 </p><p>Stratification of groups by tertiles. “Most efficient” units are units with both SMR and SRU in the lowest </p><p> tertile. These are compared to “least efficient” units (both SMR and SRU in the highest tertile). See </p><p> also figure 1 and table 2. </p><p>--- 3 --- Outcome and resource use in critical care (electronic supplementary material) </p><p>Type of hospital: Y = university hospital; N = non-university hospital </p><p>Emergency department (ED): Y = hospital has an ED; N = hospital has no ED. </p><p>Clinical rounds: Y = clinical rounds (physicians and nursing staff together) in the ICU; N = no common </p><p> clinical rounds in ICU. </p><p>Physicians (weekdays): Y = physicians are available in ICU on weekdays; N = physicians are not </p><p> available. </p><p>Physicians (nights/weekends): Y = physicians are available in ICU during the night and/or on </p><p> weekends; N = physicians are not available. </p><p>Region: see legend to table 3. </p><p>OR: Odds Ratio for being in the lowest SMR and lowest SRU tertile as compared to the highest SMR and highest SRU tertile.</p><p>For region: OR is calculated as compared to region 3 (= reference group). </p><p>--- 4 --- Outcome and resource use in critical care (electronic supplementary material) </p><p>Table 5E (electronic material): Estimated efficiency: Multiple regression analysis of explanatory</p><p> variables. Stratification of groups by tertiles. </p><p>Variable OR 95% CI P </p><p>Physicians per bed na </p><p>Intensive care medicine specialists per bed na </p><p>Nurses per bed na </p><p>Physicians per nurse na </p><p>Clinical rounds na </p><p>Physicians (weekdays) na </p><p>Region</p><p>1 1.88 0.10 – 35.45 0.673 </p><p>2 0.03 0.00 – 0.20 < 0.001 </p><p>3 1.00 </p><p>4 0.01 0.00 – 2.06 0.093 </p><p>5 0.11 0.03 – 0.46 0.002 </p><p>Stratification of groups by tertiles. “Most efficient” units are units with both SMR and SRU in the lowest </p><p> tertile. These are compared to “least efficient” units (both SMR and SRU in the highest tertile). See </p><p> also figure 1 and table 2. </p><p>Stepwise logistic regression analysis. Initially, all significant variables of univariate analysis are </p><p> entered into the model. For region, OR is calculated as compared to region 3 (= reference group) </p><p>R2 = 0.28 (Cox and Snell), R2 = 0.40 (Nagelkerke), area under the receiver operating </p><p> characteristics (ROC) curve = 0.72 na: variable did not enter the model. </p><p>For region: OR is calculated as compared to region 3 (= reference group). </p><p>For further legends see table 4. </p><p>--- 5 --- Outcome and resource use in critical care (electronic supplementary material) </p><p>Figure 2E (electronic material): SMR and SRU vs. geographical region </p><p>SMR vs. geographical region</p><p>3</p><p>2</p><p>SMR</p><p>1</p><p>0 1 3 4 5 2 on n on on on gi gio gi gi gi re re re re re</p><p>SRU vs. geographical region</p><p>10 U R S</p><p>1</p><p>1 3 4 5 2 on on on on on gi gi gi gi gi re re re re re</p><p>Geographical regions (x-axis) are ordered according to median SRU. </p><p>Region: see legend to table 3 </p><p>--- 6 --- Outcome and resource use in critical care (electronic supplementary material) </p><p>Figure 3E (electronic material): SMR and SRU vs. structural properties of the ICU </p><p>SMR vs. type of hospital, presence of emergency department, and presence of clinical rounds</p><p>2 R M S 1</p><p>0 l nt ita e ds sp rtm un ho pa ro y e al sit d ic er cy in iv n cl n ge u er em</p><p>Top panel: SMR vs. type of hospital, presence of emergency department, and presence of common clinical rounds (physicians and nurses). Yes = open boxes, no = striped boxes</p><p>SRU vs. type of hospital, presence of emergency department, and presence of clinical rounds</p><p>10 U R S</p><p>1</p><p> l nt ita e ds sp rtm un ho a ro y ep l it d ica rs y in ve nc cl ni ge u er em</p><p>Bottom panel: SRU vs. type of hospital, presence of emergency department, and presence of common clinical rounds (physicians and nurses). </p><p>--- 7 --- </p>