Price 1982.

Methods	Allocation: randomised, no details. Blindness: double, described and adequate. Duration: 9‐11 days. Design: parallel. Raters: one blinded rater. Setting: inpatients, USA.
Participants	Diagnosis: schizophrenia (69%), OBS (29%), bipolar (2%) + TD (diagnosed by criteria), thorough evaluation to rule out differential diagnostic categories. History: mean duration of antipsychotic treatment 17 (SD 8.5) years (range 2‐26 years). TD for at least 3 months. N = 45. Sex: all male. Age: mean 56 years, range 26‐77 years.
Interventions	1. Lecithin: dose 60 g/day containing PC dose of 33 g/day. N = 15. 2. Placebo. N = 15. 3. No‐treatment control group. N = 15. Antipsychotics stable, anticholinergics used by 7 participants.
Outcomes	TD symptoms. Adverse effects. Leaving study early. Unable to use ‐ TD symptom scores: Simpson TDRS, SRTDRS (reported in ANCOVA tables, unable to extract data).
Notes	Sponsorship source: Sponsorship source not reported. Review uses data only from lecithin and placebo groups for whom blinding adequate and reporting consistent. (N = 15 + 15 = 30). Author contacted to confirm lack of additional data. 60% of participants overlapped with Beckham 1981 study. Extensive neuropsychological and motor tests performed.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"randomly selected". Details not reported
Allocation concealment (selection bias)	Unclear risk	Allocation concealment not reported.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"double‐blind". "Although the subjects assigned to one of the treatment groups were informed that they would receive either the lecithin treatment or a placebo control treatment, neither the patients nor the researcher knew to which group any individual had been assigned." "The placebo substance resembled the lecithin mixture in taste, .appearance, and thickness" The no treatment group's participants could not have been blinded
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Rater and self‐report were used as outcome measures. Research personnel and participants seem to have been blinded to the assignment. Self‐report ratings were not recorded for the "no treatment" group.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All participants completed the trial.
Selective reporting (reporting bias)	High risk	Dissertation. All outcomes seem to have been reported. However, adverse effects expected to be reported in such trials have not been reported. Data for TD scores are not extractable.
Other bias	Low risk	"A one‐way analysis of variance was performed on subject variables to determine if there were initial differences among the groups. These analyses show no significant differences for age... , duration of antipsychotic treatment..., or initial symptom severity. A chi‐square analysis of diagnostic categories demonstrates no significant difference among the groups...". The study seems to have been free of other sources of bias.