Fletcher 2014.

Study characteristics
Methods	Randomised, double‐blind, placebo‐controlled non‐inferiority crossover trial.
Participants	Adults with active cancer and self‐reported nausea at least 4 out of 10 (n=22 enrolled, 20 of whom completed the study). Many participants (unspecified) were attending the bone marrow transplant clinic, some were enrolled from the palliative care clinic or palliative care in‐patient setting. For full inclusion and exclusion criteria see Fletcher 2014.
Interventions	ABH gel (contains lorazepam, diphenhydramine and haloperidol) or placebo gel rubbed on the wrist. ABH gel contains lorazepam 20mg, diphenhydramine 250mg, haloperidol 20mg, lecithin organogel 2mL, ethoxydiglycol 0.83mL, water 0.2mL, pluronic gel 20% (sufficient to make 10mL). These 10mL were divided into ten doses of 1mL. Placebo was 1mL pluronic lecithin organogel alone. If no effect (up to 1 point on 0 to 10 scale) was seen after one hour, the participant was offered other medication for nausea and was considered to have completed the trial. Otherwise the participant crossed over to the other treatment after four hours.
Outcomes	Primary outcome: difference in nausea score from baseline to 60 minutes after intervention. A change of 2 points on a 0 to 10 scale was thought to be significant. Secondary outcomes: difference in vomiting episodes and side effects determined by the Memorial Symptom Assessment Scale ‐ Condensed. Assessments were made at baseline (before administration of gel), 30, 60, 90, 120, 180, 240 minutes. Participants were also asked questions about the perceived effectiveness of each gel.
Notes	There was no significant difference in the change in nausea scores for each group (1.7 +/‐ 2.05 for the ABH gel group and 0.9 +/‐ 2.45 for the placebo group on a 0 to10 scale; P = 0.42). Authors concluded that placebo was non‐inferior to ABH gel. A previous study had shown that there was no significant absorption of haloperidol via this route (Smith 2012a).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The study biostatistician generated the randomisation list."
Allocation concealment (selection bias)	Low risk	"Randomisation was done in the investigational drug pharmacy with allocation concealed by the use of sealed opaque envelopes not accessible to investigators."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Described as double‐blind. Participants received ABH gel or placebo gel in randomised order.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Participants and investigators not aware of which treatment was being used.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Two of the 22 participants did not complete both arms of the study, because they "did not feel like waiting the length of the study but denied side effects from the medications". It is not clear which treatments were completed for the two who did not complete the study.
Selective reporting (reporting bias)	Low risk	Placebo is reported to be non‐inferior to ABH gel in this study.
Other bias	Unclear risk	This is a relatively small study. Participants were enrolled from the bone marrow transplant clinic, inpatient palliative care unit or seen as an inpatient by the palliative care team. The study authors state that, "Many patients were bone marrow transplant patients" but it is not clear what proportion. The study participants may not be fully representative of patients receiving palliative care in other settings. However, this is unlikely to have been a significant cause of bias for the study findings.