Abstract Typing Instructions

Outcome and resource use in critical care (electronic supplementary material)

Variability in outcome and resource use in intensive care units

Hans U. Rothen, Kay Stricker, Johanna Einfalt, Peter Bauer, Philip G.H. Metnitz,

Rui P. Moreno, Jukka Takala

Electronic supplementary material

--- 1 --- Outcome and resource use in critical care (electronic supplementary material)

Table 2E (electronic material): Key variables for structure and process

The total number of intensive care units included is 275; the total number of patients is 16,560.

Table 2E-a: Structure and staffing

mean ± SD median (25th – 75th percentile)

Number of staffed beds 10.94 ± 6.97 9.00 (7.00 - 12.00)

Physicians per bed 0.93 ± 0.77 0.81 (0.50 - 1.13)

Intensive care medicine specialists per bed 0.53 ± 0.67 0.37 (0.18 - 0.74)

Number of medical specialties in hospital 11.53 ± 4.11 13.00 (10.00 - 15.00)

Nurses per bed 3.44 ± 2.75 3.00 (2.33 - 3.88)

Physicians per nurse 0.33 ± 0.30 0.28 (0.17 - 0.39)

Data are mean ± SD and median (25th –75th percentile).

Table 2E-b: Hospital characteristics and clinical processes in the ICU

yes no yes no

n n % %

University hospital 122 141 46 54

Presence of emergency department in hospital 239 24 91 9

Multidisciplinary meetings in the ICU 121 136 47 53

Clinical rounds (physicians + nursing staff) in the ICU 158 99 61 39

Availability of physicians (weekdays) in the ICU 210 51 80 20

Availability of physicians (nights/weekends) in the ICU 204 57 78 22

Note: Due to missing data, the numbers do not sum up to n = 275.

Table 4E (electronic material): Estimated efficiency: Univariate logistic regression analysis of

explanatory variables. Stratification of groups by tertiles.

Variable OR 95% CI P

Number of staffed beds 1.05 0.97 – 1.13 0.231

Physicians per bed 0.36 0.15 – 0.88 0.026

Number of intensive care medicine specialists 0.95 0.84 – 1.06 0.362

Intensive care medicine specialists per bed 0.28 0.08 – 0.98 0.047

Number of medical specialties in hospital 1.01 0.92 – 1.11 0.852

Nurses per bed 2.02 1.30 – 3.15 0.002

Physicians per nurse 0.01 0.00 – 0.14 0.001

Type of hospital 1.13 0.46 – 2.78 0.783

Emergency department in hospital 7.16 0.59 – 86.62 0.122

Multidisciplinary meetings 1.24 0.49 – 3.17 0.648

Clinical rounds 4.07 1.54 – 10.75 0.005

Physicians (weekdays) 0.22 0.06 – 0.80 0.022

Physicians (nights/weekends) 0.34 0.10 – 1.23 0.101

Region

1 1.94 0.11 – 34.92 0.655

2 0.02 0.00 – 0.12 <0.001

3 1.00

4 0.11 0.01 – 1.43 0.09

5 0.01 0.02 – 0.33 < 0.001

Stratification of groups by tertiles. “Most efficient” units are units with both SMR and SRU in the lowest

tertile. These are compared to “least efficient” units (both SMR and SRU in the highest tertile). See

also figure 1 and table 2.

Type of hospital: Y = university hospital; N = non-university hospital

Emergency department (ED): Y = hospital has an ED; N = hospital has no ED.

Clinical rounds: Y = clinical rounds (physicians and nursing staff together) in the ICU; N = no common

clinical rounds in ICU.

Physicians (weekdays): Y = physicians are available in ICU on weekdays; N = physicians are not

available.

Physicians (nights/weekends): Y = physicians are available in ICU during the night and/or on

weekends; N = physicians are not available.

Region: see legend to table 3.

OR: Odds Ratio for being in the lowest SMR and lowest SRU tertile as compared to the highest SMR and highest SRU tertile.

For region: OR is calculated as compared to region 3 (= reference group).

Table 5E (electronic material): Estimated efficiency: Multiple regression analysis of explanatory

variables. Stratification of groups by tertiles.

Variable OR 95% CI P

Physicians per bed na

Intensive care medicine specialists per bed na

Nurses per bed na

Physicians per nurse na

Clinical rounds na

Physicians (weekdays) na

Region

1 1.88 0.10 – 35.45 0.673

2 0.03 0.00 – 0.20 < 0.001

3 1.00

4 0.01 0.00 – 2.06 0.093

5 0.11 0.03 – 0.46 0.002

Stratification of groups by tertiles. “Most efficient” units are units with both SMR and SRU in the lowest

tertile. These are compared to “least efficient” units (both SMR and SRU in the highest tertile). See

also figure 1 and table 2.

Stepwise logistic regression analysis. Initially, all significant variables of univariate analysis are

entered into the model. For region, OR is calculated as compared to region 3 (= reference group)

R2 = 0.28 (Cox and Snell), R2 = 0.40 (Nagelkerke), area under the receiver operating

characteristics (ROC) curve = 0.72 na: variable did not enter the model.

For region: OR is calculated as compared to region 3 (= reference group).

For further legends see table 4.

Figure 2E (electronic material): SMR and SRU vs. geographical region

SMR vs. geographical region

0 1 3 4 5 2 on n on on on gi gio gi gi gi re re re re re

SRU vs. geographical region

10 U R S

1 3 4 5 2 on on on on on gi gi gi gi gi re re re re re

Geographical regions (x-axis) are ordered according to median SRU.

Region: see legend to table 3

Figure 3E (electronic material): SMR and SRU vs. structural properties of the ICU

SMR vs. type of hospital, presence of emergency department, and presence of clinical rounds

2 R M S 1

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

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

SRU vs. type of hospital, presence of emergency department, and presence of clinical rounds

10 U R S

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

Bottom panel: SRU vs. type of hospital, presence of emergency department, and presence of common clinical rounds (physicians and nurses).

--- 7 ---