Murthy 1991.

Methods	Randomised controlled trial, 6 weeks
Participants	208 outpatients with moderate to severe depression meeting Feighner’s criteria for primary depression and DSM‐III major depression criteria
Interventions	Alprazolam versus imipramine (105/103 patients) Placebo 'washout' period: 3 to 7 days Maximum dosages: alprazolam 4.5 mg per day, imipramine 225 mg per day Mean dosages: alprazolam 2.5 mg per day, imipramine 125 mg per day, and 80% required less than 6 capsules of 0.5 mg per day
Outcomes	Primary outcome: Mean HDRS after 6 weeks: alprazolam 9.75 (initial score 23.81), SD 4.63; imipramine 9.20 (initial score of 23.44); SD 4.72 Secondary outcome: All‐cause withdrawals: alprazolam 29 (28%), imipramine 34 (33%) "The frequency of side effects was higher for imipramine compared to alprazolam. Significantly higher number of patients reported insomnia (P < 0.01) and tremor as side effects (P < 0.01). None of the side effects reported was significantly more in the alprazolam group."
Notes	HDRS 21 items
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Assigned in double blind random fashion." In each consecutive group, there were 3 persons who received alprazolam and 3 imipramine; but there is no further information
Allocation concealment (selection bias)	Unclear risk	There is no information, except that "The drug code of each patient was kept in a sealed envelope which could be opened in case of emergency."
Blinding (performance bias and detection bias) Patients	Low risk	Yes, identical capsules (there could have been a specific and rapid tranquillising effect of alprazolam)
Blinding (performance bias and detection bias) Physicians	Low risk	Yes, identical capsules and the code was kept in a sealed envelope: "The drugs were dispensed in identical capsules, each capsule containing alprazolam 0.5 mg or imipramine 25 mg. Patients were started with one capsule twice daily. The drug code for each patient was kept in a sealed envelope which could be opened in case of emergency."
Blinding (performance bias and detection bias) Investigators	Unclear risk	No further information is provided
Incomplete outcome data (attrition bias) HDRS	Unclear risk	No information is provided
Incomplete outcome data (attrition bias) Withdrawals/side effects	Unclear risk	Unclear, although dropouts are documented
Selective reporting (reporting bias)	Unclear risk	No information is provided
Other bias	Unclear risk	It seems that Upjohn sponsored the study
Placebo washout	High risk	3 to 7 days