Escobar 1985.

Methods	Randomisation: “randomly assigned”. Blinding: double‐blind. Duration: 4 weeks. Design: parallel. Location: two‐centre study (Brentwood VA Medical Center, Los Angeles; Payne Whitney Clinic, New York Hospital‐Cornell Medical Center, New York). Setting: “Typically, patients remain in these acute wards for only a few days.” Hospitalisation at least during the injectable phase of the trial. No further details.
Participants	Diagnosis: schizophrenia (DSM‐III). N = 35. Gender: 32M, 3F. Age: mean 37 years. History: duration stable: not indicated, duration of illness: not indicated, number of previous hospitalisations: mean 6, age at onset: not indicated, severity of illness: not indicated, baseline antipsychotic dose: not indicated.
Interventions	1. Haloperidol: flexible dose, dose range: maximum mean oral dose 32.4 mg/day, mean dose: not indicated. N = 15 (participants completing at least 1 week of the oral phase of the trial).  2. Molindone: flexible dose, dose range: maximum mean oral dose 160 mg/day, mean dose: not indicated. N = 15 (participants completing at least 1 week of the oral phase of the trial).    “Dosages were clinically determined.” “Injections were administered only during the first 12‐72 hours.”
Outcomes	Examined: Global state general: Clinical Global Impressions (CGI) ‐ Symptom Severity (no SDs available). Adverse effects: akathisia, dystonia, rigor, tremor, weight gain. Unable to use: Mental state general: Brief Psychiatric Rating Scale (BPRS) (no total score available). Adverse effects: Treatment Emergent Symptoms Scale (TESS) (no total score available).
Notes	“ongoing study”. The study drugs were “given for the first 2‐3 days of hospitalisation and then continued orally for up to 4 weeks.”
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	“Randomly assigned”. No further details.
Allocation concealment (selection bias)	Unclear risk	No further details.
Blinding (performance bias and detection bias) (performance bias)	Low risk	“Double‐blind”. “identical appearing tablets”.
Blinding (performance bias and detection bias) (detection bias)	Low risk	“Double‐blind”. “The rater was blind to the type of medication the subject received.”
Incomplete outcome data (attrition bias) All outcomes	High risk	“Only 30 subjects completed 1 week of treatment and 25 completed 2 weeks.” Thus, at least 10 of 35 participants (28.6%) left the trial early and therefore the overall attrition can be considered as being high. It is not explicitly mentioned how many patients were randomised to each study group (haloperidol or molindone group). “Analyses [were] based on the 30 subjects who completed at least the first week of the study.”
Selective reporting (reporting bias)	High risk	The outcome data were not fully addressed (no total score for the BPRS and TESS; no SDs for the CGI).
Other bias	High risk	“This study is limited by…..the relatively short evaluation period.” “Because of attrition, analyses for the oral portion of the study [were] limited to ratings at baseline, days 2‐3, and weeks 1 and 2.”