Attention bias modification (ABM) , a newly emerging focus in emotion modification research, is a treatment to reduce individuals' attention bias toward negative stimuli to improve their abnormal cognition via repeated computer-based attention trainings toward neutral or positive stimuli. Recent studies about depression treated using ABM are increasing, but based on insufficient evidence with various efficacy assessment indicators.

To systematically assess the effect of ABM in patients with depression.

Databases including PubMed, The Cochrane Library, EMBase, CBM, CNKI, WanFang Data and VIP were searched from inception to December 31st, 2021 for randomized controlled trials (RCTs) regarding depressive patients treated by ABM (experimental group with computer-based ABM trainings) compared with placebo trainings under the same conditions (control group with no interventions or sham ABM trainings) . Primary outcome measures included Hamilton Rating Scale for Depression (HAMD/HRSD) , Beck Depression Inventory (BDI) and Center for Epidemiologic Studies Depression Scale (CES-D) . Secondary outcome measures included the State-Trait Anxiety Inventory (STAI-T) and Ruminative Responses Scale (RRS) . Two researchers independently performed literature enrollment, and extracted data, and evaluated the risk of bias using Risk of Bias 2 (RoB 2.0) tool. RevMan 5.4 and Stata 12.0 were used for Meta-analysis.

Thirteen RCTs involving 968 patients were included, among which one had two trials. Risk of bias was low in seven RCTs, unclear in four, and high in two. Meta-analysis showed that the overall improvement of depression, anxiety and rumination in the experimental group was better than that in the control group (P<0.05) . Further analysis indicated that two groups had no significant differences in mean scores of BDI and HAMD within a follow-up period of less than two months (P>0.05) . When the follow-up time was prolonged to at least two months, the mean BDI score was still similar in both groups (P>0.05) .

ABM treatment could improve depressive, anxiety symptom and rumination in patients with depression, but its long-term effect needs to be studied further.

注意偏向矫正（ABM）是近年来情绪调节领域研究的热点之一，指通过计算机程序反复训练个体对中性或正性刺激的关注以达到纠正其对负性刺激的注意偏向、改善异常认知的过程。近年来，国内外公开发表的ABM治疗抑郁的文献报道逐年增多，但评价指标较为分散，相关循证证据不足。

系统评价ABM疗法对抑郁患者的干预效果。

计算机检索PubMed、The Cochrane Library、EMBase、中国生物医学文献数据库（CBMdisc）、中国知网（CNKI）、万方数据知识服务平台和维普网（VIP），搜集有关ABM治疗抑郁患者的随机对照试验（RCT），检索时限均为建库至2021-12-31。干预措施：试验组采用基于计算机的ABM治疗；对照组实施同等条件下的安慰训练，包括空白对照、假ABM刺激等。主要结局指标包括汉密尔顿抑郁评定量表（HAMD/HRSD）、贝克抑郁量表（BDI）和流行病学研究中心抑郁量表（CES-D）评分；次要结局指标包括状态-特质焦虑量表（STAI）、冗思反应量表（RRS）评分。由两位研究者独立筛选文献、提取资料，并采用Cochrane提供的偏倚风险评估工具2.0版本（RoB 2.0）进行偏倚风险评价，采用RevMan 5.4和Stata 12.0软件进行Meta分析。

最终纳入13篇文献，其中1篇文献提取为2个试验报告，共包含968例患者。13篇文献均为RCT，其中7篇文献评价为低偏倚风险，4篇文献评价为中偏倚风险，2篇文献评价为高偏倚风险。Meta分析结果显示，试验组抑郁水平、焦虑情绪及冗思反应的改善优于对照组（P≤0.05）；亚组分析：干预后随访时间<2个月时，试验组与对照组BDI-Ⅱ、HAMD评分比较，差异无统计学意义（P>0.05）；干预后随访时间≥2个月时，试验组与对照组BDI-Ⅱ评分比较，差异无统计学意义（P>0.05）。

ABM疗法能明显改善抑郁患者的抑郁、焦虑及冗思反应，但该疗法的长期疗效仍需进一步研究。