Methodological approaches to previously reported adjusted analyses
| Concept | Approach used for clinical trial report* | Most informative model† |
| Handling of patients who had incomplete data for baseline NIHSS score | Exclusion‡ | Inclusion using imputed scores§ |
| Approach to baseline NIHSS score variable for adjustment | Categorical with five categories (0–5, 6–10, 11–15, 16–20, >20) | Continuous |
| Approach to time from symptom onset variable for adjustment | Categorical with seven categories (3–4.5 hours in 15 min increments, >4.5 hours) | Continuous |
*As described in clinical trial report (dated 15 December 2008).
†As considered in alternative models applied to replication of previously reported analyses (see table 3).
‡According to correspondence with the study sponsor.
§Imputed values for baseline NIHSS scores were very close to the likely actual values for 10 of 11 patients requiring imputation. These 10 patients had values for 15 or 16 of the 17 component scores used to calculate the baseline NIHSS scores.
NIHSS, National Institutes of Health Stroke Scale.