Summary of findings 1. Antipsychotics compared with placebo for people with borderline personality disorder.

Antipsychotics compared with placebo for people with borderline personality disorder
Patient or population: people with borderline personality disorder Settings: inpatient and outpatient Intervention: antipsychotics Comparison: placebo
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	No of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Placebo	Antipsychotics
BPD symptom severity Timing of outcome assessment: end of treatment (5 to 12 weeks treatment duration)	‐	The mean score in the intervention groups was 0.18 SD lower (0.45 lower to 0.08 higher, I2 = 70%) than in the placebo group	‐	951 (8 trials)	⊕⊝⊝⊝ Very Lowa,b	An SMD of 0.18 represents a small effect. **TSA‐adjusted CI is −3.27 to 0.86 on the Zanarini BPD scale. Z value in futility area TSA DARIS = 951 TSA in futility area
Self‐harm Timing of outcome assessment: end of treatment (8 to 24 weeks treatment duration)	310 per 1000	208 per 1000 (211 less to 127 more self‐harm incidents than in the placebo group)	RR 0.66 (95% CI 0.15 to 2.84, I2 = 67%)	76 (2 trials)	⊕⊝⊝⊝ Very Lowc,d	‐
Suicide‐related outcomes Timing of outcome assessment: end of treatment (6 to 12 weeks treatment duration)	‐	The mean score in the intervention groups was 0.05 SD higher (0.18 lower to 0.29 higher, I2 = 55%) than in the placebo group	‐	854 (7 trials)	⊕⊝⊝⊝ Very Lowa,e	A SMD of 0.05 represents a marginal effect. RR 0.73 (95% CI 0.31 to 1.73; 2 trials, 61 participants), low‐certainty evidence
Psychosocial functioning Timing of outcome assessment: end of treatment (5 to 12 weeks treatment duration)	‐	The mean score in the intervention groups was 0.16 SD lower (0.33 lower to 0.00 lower, I2 = 75%) than in the placebo group	‐	904 (7 trials)	⊕⊝⊝⊝ Very Lowa,f,g	A SMD of 0.16 represents a small effect.
Interpersonal problems Timing of outcome assessment: end of treatment (5 to 12 weeks treatment duration)	‐	The mean score in the intervention groups was 0.21 SD lower (0.34 lower to 0.08 lower, I2 = 0%) than in the placebo group	‐	907 (8 trials)	⊕⊕⊝⊝ Lowa	A SMD of 0.21 represents a small effect. TSA‐adjusted CI is −0.60 to 0.08 on Zanarini BPD Scale ‐Interpersonal Problem Index TSA DARIS = 386
Attrition Timing of outcome assessment: end of treatment (5 weeks to 6 months treatment duration)	325 per 1000	361 per 1000 (36 less to 123 more than in the placebo group dropped out)	RR 1.11 (0.89 to 1.38; I2 = 35%)	1216 (13 trials)	⊕⊝⊝⊝ Very Lowa,h	TSA‐adjusted CI is 0.74 to 2.13 TSA DARIS = 2008
Non‐serious adverse events Timing of outcome assessment: end of treatment (8 to 12 weeks treatment duration)	560 per 1000	599 per 1000 (56 less to 162 more than in the placebo group dropped out)	RR 1.07 (0.90 to 1.29; I2 = 57%)	814 (5 trials)	⊕⊝⊝⊝ Very Lowa,i	TSA‐adjusted CI is 0.79 to 1.47 TSA DARIS = 1250 TSA in futility area
*The basis for the assumed risk (e.g. the median control group risk across trials) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). **TSA did not include cross‐over trials. CI: Confidence interval; RR: Risk Ratio; BPD: Borderline Personality Disorder; SD: standard deviation; SMD: Standardised mean difference TSA Trial Sequential Analysis, DARIS: Diversity‐adjusted required information size
GRADE Working Group grades of evidence High certainty: Further research is very unlikely to change our confidence in the estimate of effect. Moderate certainty: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low certainty: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low certainty: We are very uncertain about the estimate.

^aWe downgraded the evidence two levels due to risk of bias of the included studies (indication of selective outcome reporting, indication of incomplete outcome data presented in the included studies) (rated by HEC and OJS (E))

^bWe downgraded the evidence one level due to inconsistency (I²= 70%) (rated by HEC and OJS)

^c We downgraded the evidence two levels due to imprecision (wide CI around the pooled effect estimate suggest both an appreciable effect and no effect; few participants included in the studies) (rated by HEC and OJS)

^dWe downgraded the evidence one level due to inconsistency (I² = 67%) (rated by HEC and OJS (D))

^eWe downgraded the evidence one level due to inconsistency (I² = 55%) (rated by HEC and OJS)

^fWe downgraded the evidence one level due to inconsistency (I² = 75%) (rated by HEC and OJS)

^gWe downgraded the evidence one level due to imprecision (wide CI around the pooled effect estimate suggest both an appreciable effect and no effect) (rated by HEC and OJS)

^h We downgraded the evidence one level due to imprecision (wide CI around the pooled effect estimate suggest both an appreciable effect and no effect; the TSA that did not reach the required information size (RIS)) (rated by HEC and OJS)

ⁱ We downgraded the evidence one level due to inconsistency (I²= 57%) (rated by HEC and OJS)