Table 1 Observational studies

Fogerty et al. (2010) [32]

Retrospective cohort study

223 patients, aged ≥55 years, with thermal burns, admitted to a regional burns centre

Pre-injury statin use (n = 70), duration not specified, continued after hospitalisation in 77%

No pre-injury statin use (n = 153)

In-hospital mortality

Infection

Septic shock

OR 0.17 (95% CI 0.05-0.57)

OR 0.90 (95% CI 0.48-1.7)

OR 0.50 (95% CI 0.20-1.30)

No change in odds ratio when stratified by cardiovascular comorbidities

Statin therapy continued after hospitalisation in 77%

Multivariate regression analysis determined odds ratios of death and sepsis by statin use, adjusting for cardiovascular comorbidities

Efron et al. (2008) [28]

Retrospective cohort study

1,224 patients, aged 65-84 years with moderate-severe traumatic injury (AIS ≥3), survival >24 h, participating in NSCOT study

Pre-injury statin use (21.1%), duration not specified, continuation after admission not known

No pre-injury statin use (78.9%)

In-hospital mortality

OR 0.33 (95% CI 0.12-0.92)

Subgroup with cardiovascular comorbidity (n = 414): OR 1.41 (95% CI 0.72-2.72)

Subgroup without cardiovascular comorbidity (n = 775): OR 0.30 (95% CI 0.10-0.91)

Multivariate logistic regression analysis

NSCOT study captured pre-injury medication by class only. No data on compliance, duration, dose, or whether continued after admission to hospital

NSCOT study used very complex statistical sampling model

Neal et al. (2009) [29]

Retrospective cohort study

295 patients, aged 55-90 years, blunt mechanism of injury, hypotension (systolic blood pressure <90 mmHg) or biochemical evidence of hypoperfusion (base deficit >5 meq/L) on admission, blood transfusion requirement, at least one AIS ≥2 other than head, survival >24 h, participating in Host Response to Injury Large Scale Collaborative Program

Pre-injury statin use (n = 71), as verified by patient or relative

No pre-injury statin use (n = 224)

In-hospital mortality

Nosocomial infection (microbiologically confirmed pneumonia, catheter-related bloodstream infection or urinary tract infection)

Multi-organ failure (defined as Marshall score >5)

HR 1.98 (95% CI 0.9-4.0)

HR 0.78 (95% CI 0.5-1.4)

HR 1.81 (95% CI 1.1-2.9)

Propensity score adjusted regression analysis to control for differences in baseline characteristics

No data on whether statin therapy was continued after hospital admission

Schneider et al. (2011) [30]

Retrospective cohort study

523 patients, aged 65 years and older, with head AIS ≥3, survival >24 h, participating in NSCOT study

Pre-injury statin use (22.3%), duration not specified, continuation after admission not known

No pre-injury statin use (77.7%)

In-hospital mortality

Extended Glasgow Outcome Scale insurvivors at 3 and 12 months after injury

RR 0.24 (95% Cl 0.08-0.69)

RR at 3 months 0.77 (95% Cl 0.42-1.41)

RR at 12 months 1.13 (95% Cl 1.01-1.26)

Mortality subgroup with cardiovascular comorbidity: RR 0.87 (95% Cl 0.50-1.50)

Subgroup without cardiovascular comorbidity: RR 0.17 (95% Cl 0.05-0.63)

Multivariate logistic regression analysis. NSCOT study captured pre-injury medication by class only. No data on compliance, duration, dose, or whether continued after admission to hospital. NSCOT study used very complex statistical sampling model