Analytical approach:
The cost-effectiveness analysis was based on a published clinical trial, and new epidemiological data to inform life expectancy. The authors stated that they took a French National Health System perspective, with a five-year time horizon.
Effectiveness data:
The key effectiveness outcome was the reduction in mortality with albumin. This was from a subset analysis of the published Saline vs Albumin Fluid Evaluation (SAFE) study, a large prospective, double-blind, multicentre, randomised study of albumin in ICUs. The patient life expectancy was estimated using the Declining Exponential Approximation of Life Expectancy (DEALE). This was adjusted for age, gender, Simplified Acute Physiology Score (SAPS II), and the McCabe score, which were from a database of ICUs in a network of hospitals in Paris and its suburbs (CUB-Rea). For this study, all patients treated for severe sepsis, in the 35 participating CUB-Rea ICUs, between January 1998 and December 2002, were selected if they met the stated inclusion criteria.
Monetary benefit and utility valuations:
Not relevant.
Measure of benefit:
The benefit measures were the number of lives saved and the number of years of life gained.
Cost data:
The cost categories were albumin, cardiac failure and shock, septicaemia, and the ICU stay. These costs were from diagnosis-related group (DRG) data, and fixed daily prices for French ICUs. The cost of albumin was estimated for the quantities administered in the SAFE study. All costs were presented in 2005 Euros (EUR).
Analysis of uncertainty:
Several one-way sensitivity analyses were undertaken. The mortality with albumin was varied from -1% to +10%; the volume of infused albumin was varied from 1.5L to 6L; the McCabe score was varied from 35% to 65% of natural life expectancy; and mortality was varied based on the SAPS II.