How real are patients in placebo-controlled studies of acute manic episode?

OBJECTIVE: To determine whether the results from placebo-controlled studies conducted in patients with manic episode can be generalised to a routine population of hospitalised acute manic patients.
METHODS: A list of four most prevalent inclusion and the nine most prevalent exclusion criteria was constructed for participation in previous randomised-controlled trials (RCTs). On the basis of this list, a consecutive series of 68 patients with 74 episodes of acute mania who had been referred for routine treatment were retrospectively assessed to determine their eligibility for a hypothetical but representative randomised controlled trial.
RESULTS: Only 16% of the manic episodes would qualify for the hypothetical trial (male episodes 28%, female episodes 10%), whereas 37%, 20% and 27% of the manic episodes would have to be excluded because they did no fulfil one, two or at least three of the inclusion or exclusion criteria. The most common exclusion criterion was "no use of contraceptives". If this criterion was not taken into account, 28% of the male episodes and 33% of the female episodes would qualify for inclusion in the hypothetical study. Apart from the use of contraceptives, no significant differences between male and female episodes were observed in the reasons for exclusion: 11% suicidal ideation, 29% prior mood stabilising medication, 1% depot medication, 22% other axis I diagnosis, 27% internal disease somatic disease, 5% neurological disorder, 15% alcohol use disorder and 10% drug use disorder.
CONCLUSION: Only a small percentage acute manic episodes in a routine mental hospital seem to qualify for a standard placebo-controlled RCT. It could be argued, however, that certain exclusion criteria (e.g. no use of contraceptives) are not very likely to reduce the external validity of a standard RCT. In contrast, some other exclusion criteria (e.g. comorbid alcohol and drug use disorders) may have resulted in an overestimation of the efficacy of anti-manic medications. These notions should be taken into account when evaluating the results of RCTs in bipolar patients with an acute manic episode.