Introduction: Neuroleptics are commonly used drugs to treat different conditions (e.g. psychosis, migraines) in the acute care setting and the emergency department. Their side effects can be disabling or, worse, fatal. The use of diphenhydramine to prevent those side-effects is widespread, but remains controversial. We performed a systematic review to determine if prophylactic administration of diphenhydramine (PAD) reduces the incidence of neuroleptic side-effects. Methods: Data sources: Medline, Embase, Cochrane Library, PsycInfo and Web of Science were searched. References from reviews that were identified in the search and from included studied were also reviewed for inclusion. Study selection: Randomized controlled trials evaluating any neuroleptic with PAD versus the same neuroleptic alone or with any inactive agent. Primary outcome was incidence of any extra-pyramidal side-effect. Secondary outcomes were akathisia, usage of rescue medication, subjective restlessness, neuroleptic malignant syndrome, sedation and sedation intensity. Data extraction: Independent reviewers scanned identified citations, extracted data and assessed for risk of bias. Data analysis: Meta-analysis was performed using random effect models. Heterogeneity and quality of evidence were assessed using, respectively, I2 and the GRADE approach. Results: Results: Of 1566 identified citations, nine studies (n=1436) met all eligibility criteria. Four studies were specifically designed to assess for neuroleptic side-effects. Four studies were at high risk of bias. In primary analysis, PAD had no effect on the incidence of extra-pyramidal symptoms (7 studies, n=1393 patients, RR 0.70 [0.40-1.22]), akathisia (5 studies, n=1094 patients, RR 0.81 [0.36-1.82]) and sedation (5 studies; n=1079, RR 1,48 [0.90-2.42]). Higher dosage of diphenhydramine was not associated with a greater reduction of extra-pyramidal side-effects. In a sensitivity analysis excluding an outlier study (n=120, RR 6.63 [1.55-28,35]), PAD was associated with a significant decrease in extra-pyramidal side-effects (6 studies, n=1273, RR 0.56 [0,38-0.82]), but not with any of the secondary outcome measures. Conclusion: Conclusion: When excluding an outlier study, PAD was associated with a significant reduction of extra-pyramidal side-effects. However, PAD did not significantly influence the incidence of akathisia. Overall quality of evidence is low. Further studies are warranted. PAD represents an interesting treatment option against neuroleptic side-effects, but its widespread usage whitout strong evidence to support it raises concerns.
