Table II.

Meta-analysis-based effect estimates as input for the cost-effectiveness model.

Event probabilities are shown for patients in the liberal and restrictive groups. The probabilities of complications in the restrictive group were calculated with a pooled relative risk (RR) compared to the liberal group. The lower and upper bounds of the 95% confidence interval (CI) are given.

Probabilities	Pooled effect liberal group	Pooled relative risk	Effect estimate restrictive group in % (95% CI)	p-value
Non-cardiac surgery
Sepsis (total)	19.32%	0.30	5.72% (5.41–6.05)	<0.01
a) Sepsis with pneumonia	11.08%	0.14	1.56% (1.41–1.74)	<0.01
b) Sepsis without pneumonia	8.24%	0.50	4.16% (4.00–4.31)	<0.01
Acute renal failure	4.02%	0.49	1.98% (1.65–2.38)	0.01
Acute myocardial infarction	5.96%	0.76	4.50% (3.58–5.66)	0.02
Acute stroke	1.58%	0.63	1.00% (0.64–1.55)	<0.01
Total (any complication)	30.88%	0.43	13.20% (11.28–15.64)
Cardiac surgery
Sepsis (total)	11.31%	0.41	4.61% (3.39–6.29)	0.04
a) Sepsis with pneumonia	6.48%	0.41	2.65% (1.94–3.60)	<0.01
b) Sepsis without pneumonia	4.82%	0.41	1.97% (1.45–2.68)	<0.01
Acute renal failure	2.50%	0.38	0.94% (0.78–1.13)	<0.01
Acute myocardial infarction	3.03%	0.02	0.07% (0.02–0.18)	<0.01
Acute Stroke	1.90%	0.35	0.66% (0.54–0.81)	<0.01
Total (any complication)	18.73%	0.34	6.28% (4.73–8.4)