A Cochrane review included 5 RCTs with a total of 213 stroke patients. Stroke had taken place between 6 days and 13 years prior to randomization. Patients suffered from emotionalism: lability or pathological crying and laughing. No standard criteria were used to define emotionalism at entry across the trials. The treatments included nortriptyline, sertraline (2 studies) and fluoxetine (2 studies). Duration of treatment ranged from 10 to 182 days. Methods of assessment of emotionalism varied across different trials. Large effects of treatment were seen: 50% reduction in emotionalism, diminished tearfulness, improvements (reduction) in lability, tearfulness and scores on the Pathological Laughter and Crying Scale. However, confidence intervals were wide indicating that treatment may have had only a small positive effect, or even a small negative effect (in one trial). Only two studies systematically reported adverse events; no discernible differences were seen between groups.
Comment: The quality of evidence is downgraded by study quality (unclear allocation concealment, inadequate follow up), inconsistency (heterogeneity in patients and evaluation methods) and indirectness (differences in studied patients).