Laboratory-based parameters on day of ICU admission analysed were: neutrophil count, platelet count, serum bilirubin and urea.
As the deaths were evenly distributed across the hospital volume groups in terms of the proportions of patients in the groups, and the censoring of cases in the analyses of length of stay occurred in only about 0.1% of the operations studied, we do not consider that censoring was a problem in performing the LOS, LUIC, or short-term outcome (2- and 6-week re-admission) analyses.
Similarly, due to the small number of daily cause-specific admissions, analyses were only stratified by gender (all ages combined) or age (i.e. younger/older than 15 years of age).
According to a study carried out by French pharmacovigilance centres, haemorrhage subsequent to oral anticoagulant treatment (OAT) is the most common drug-related side effect resulting in hospitalisation in public hospitals in France (13% of such admissions, and 0.41% of the 3,137 admissions analysed).
For the subgroup of patients who died during admission, all analyses were replicated.
In contrast to admission rate analyses, adjustment for covariates was undertaken in only half of studies (n = 7), five (36%%) adjusted for age, sex and case-mix while one other (7 %) additionally controlled for other factors (i.e. alcohol consumption, blood culture results).
Only the first admission was analysed.
Crude 30-day all-cause mortality (from admission) was analysed for both NHS Tayside cohorts.
Associations between thrombus location and NIHSS score on admission were analysed with the Kruskal Wallis test.
Carbachol concentrations in serum and urine on day 1 and 2 of hospital admission were analysed by HPLC-mass spectrometry.
The association between survival and categorical potential risk variables (gender, age-group, where admitted from, year of admission) were analysed using Kaplan-Meier analysis (log-rank test).
