Table 1

Parameters used in our decision analytic models.

Parameter

Base case

Source

Univariate Sensitivity analyses


Effectiveness measure

Life-years gained

Based on long-term survival

Quality-adjusted life years (QALYs)

Groups

All patients

APACHE II ≥ 25

-

-

Mean entry age

60 years

PROWESS [13] and ADDRESS [14]

-

Life-expectancy for a 60-year old

Males: 20 years

Females: 24 years

Statistics Canada Life-tables[59]

-

Time horizon

20 years

Statistics Canada Life-tables [59]

30 months to 30 years

Discount rate

3%

-

0, 5%

Perspective of the analyses

Public health care provider

-

-

Survival rates (see Tables 4-6 for details)

Short-term (28-days)

PROWESS [13] and ADDRESS [14] RCTs

Probabilistic sensitivity analysis using a beta distribution defined by the point estimate and variance from RCTs for each group.

Mid-term (hospital discharge – 30 months)

Long-term PROWESS [8] RCT

Long-term beyond 30 months

Observational study in severe sepsis patients [40] and Canadian life table (2000–2002) [59] adjusted for a higher mortality in severe sepsis patients*

Complication rates

28-day bleeding rates

PROWESS [13] and ADDRESS [14]

-

Resources included in the cost analyses**

Drug acquisition

Pharmacy department MUHC

-

Hospitalization for the sepsis episode

Canadian long-term observational study in severe sepsis patients [40]

Treatment complications Years 1–3 follow-up healthcare treatment costs


* The lifetime annual survival rates in the general population were adjusted for a higher severity in severe sepsis patients according to the absolute difference in mortality between the age-specific survival in the general population (1.1% for a 63-year old) [59] and that of a 3-year Canadian long-term observation study in severe sepsis patients (4.2% and 6.2% in all patients and those with APACHE II ≥ 25 respectively at 3 years, mean age at cohort entry: 61.1 years [40]).

** Costs associated with the severe sepsis episode incurred after three years were not available and were considered identical for the two groups.

Our cost analysis included direct costs such as hospitalization, emergency room visits, day-surgery, and physician charges [40].

Costa and Brophy BMC Anesthesiology 2007 7:5   doi:10.1186/1471-2253-7-5

Open Data