Introduction: The CLEAR III trial found that removal of intraventricular hemorrhage was not associated with a dichotomized improvement in 6-month outcome measured by the modified Rankin Scale score (p=0.554). Despite this, shifts towards improved outcome were evident for the alteplase (n=249) versus saline (n=251) group measured by the extended Glasgow Outcome Score (GOS-E). Shared patient/physician decision-making would be aided by characterizing the magnitude of this benefit, using the common clinical metrics of number needed to treat (NNTB), number needed to harm (NNH), and benefit per hundred (BPH) treated.
Methods: Analyses were performed on CLEAR III findings using a 6-level version of the GOS-E (GOS-E6), in which the 3 poorest outcome levels of the original GOS-E (extremely severe disability, vegetative state, and death) are combined into a single worst outcome category. In this way, shifts between these categories are not counted as improved outcome. For all possible dichotomizations of the GOS-E6 net NNTB values were derived by taking the inverse of absolute risk difference, and net BPHs by multiplying absolute risk difference by 100. For benefits accruing across all disability state transitions on the GOS-E6 (shift analysis), net NNTB, and net BPH, values were derived using automated, algorithmic min-max joint outcome table derivation technique. In addition, NNTB and NNH values for dichotomous and shift outcomes were derived using expert’s joint outcome tables, and are presented as the geometric mean.
Results: For the 6 level GOS-E, automated algorithmic analysis indicated that the net NNTB for 1 additional patient to have a better outcome by ≥ 1 grade at 6 months, with thrombolytic rather than saline irrigation, was 7.4. Expert joint outcome table analyses indicated that the NNTB for improved final outcome at 6 months was 6.3 (range 5.8-6.7) and NNH 39.0 (32.3-47.6). For every 100 patients treated, 15.9 had a better outcome and 2.6 a worse outcome. The likelihood of help to harm ratio was 6.2.
Conclusions: Thrombolytic removal of intraventricular hemorrhage confers net benefit on 9% of treated patients. Approximately 1 in 11 patients has a better outcome, while 1 in 39 has a worse outcome.