人工流产后静脉血栓栓塞:加拿大基于人群的倾向评分匹配队列研究
该研究利用加拿大安大略省全民医疗数据,将2003年1月1日至2015年12月31日期间首次妊娠以人工流产结束的初孕妇,按倾向评分与活产初孕妇(1:1)及非妊娠未孕女性(5:1)匹配。主要结局为索引日期后42天内静脉血栓栓塞风险。人工流产后42天内静脉血栓栓塞发生率为每10万女性30.1例(95%CI 22.0-38.2),非妊娠组为13.5例(95%CI 11.1-16.0),HR 2.23(95%CI 1.61-3.08);与活产组(产后42天内发生率184.7例/10万,95%CI 164.6-204.7)相比HR 0.16(95%CI 0.12-0.22)。
要点
- 人工流产后42天内静脉血栓栓塞风险约为匹配非妊娠女性的2倍(HR 2.23,95%CI 1.61-3.08)。
- 人工流产后42天内静脉血栓栓塞风险显著低于活产后(HR 0.16,95%CI 0.12-0.22)。
- 作者认为临床医生可考虑对人工流产后女性采用比非妊娠女性更低的诊断检查门槛以排除静脉血栓栓塞。
核心结论
- 人工流产后42天内静脉血栓栓塞风险是匹配非妊娠女性的两倍,且显著低于活产后。
- 该新信息可用于估算人工流产后围操作期静脉血栓栓塞风险。
- 作者认为临床医生可考虑对人工流产后女性采用比非妊娠女性更低的诊断检查门槛以排除静脉血栓栓塞。
研究设计与样本
- 基于加拿大安大略省全民医疗系统的倾向评分匹配队列研究。
- 纳入2003年1月1日至2015年12月31日期间首次妊娠以人工流产结束的初孕妇。
- 按倾向评分将人工流产初孕妇与活产初孕妇(1:1)及非妊娠未孕女性(5:1)匹配,非妊娠女性在匹配对象手术日期未妊娠且此后1年内未受孕。
- 排除小于15岁或大于49岁女性,以及性别、居住区域、居住收入或世界来源地区信息缺失或无效者。
- 主要结局为索引日期后42天内任何静脉血栓栓塞风险,索引日期定义为人工流产日期、活产分娩日期或非妊娠女性匹配对象的人工流产日期。
主要发现
- 识别194 086例首次妊娠以人工流产结束的合格女性,其中176 001例(90.7%)可与首次妊娠以活产结束的女性匹配。
- 176 001例人工流产女性同时与880 005例非妊娠女性匹配。
- 人工流产后42天内静脉血栓栓塞发生率为每10万女性30.1例(95%CI 22.0-38.2),非妊娠组为每10万女性13.5例(95%CI 11.1-16.0),HR 2.23(95%CI 1.61-3.08)。
- 与活产队列相比HR为0.16(95%CI 0.12-0.22),活产队列产后42天内静脉血栓栓塞发生率为每10万女性184.7例(95%CI 164.6-204.7)。
关键数值与推荐
- 人工流产后42天内静脉血栓栓塞发生率每10万女性30.1例(95%CI 22.0-38.2)。
- 非妊娠组42天内静脉血栓栓塞发生率每10万女性13.5例(95%CI 11.1-16.0)。
- 人工流产组与非妊娠组比较HR 2.23(95%CI 1.61-3.08)。
- 活产组产后42天内静脉血栓栓塞发生率每10万女性184.7例(95%CI 164.6-204.7)。
- 人工流产组与活产组比较HR 0.16(95%CI 0.12-0.22)。
临床怎么用
- 作者认为临床医生可考虑对人工流产后女性采用比非妊娠女性更低的诊断检查门槛以排除静脉血栓栓塞。
局限性
- 原文未报告局限性
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: A woman's risk of venous thromboembolism during pregnancy is estimated to be two-to-six times higher than her risk when she is not pregnant. Such risk estimates are largely based on pregnancies that result in delivery of a newborn baby; no estimates exist for the risk of venous thromboembolism after induced abortion, another common pregnancy outcome. To fill this knowledge gap, we aimed to assess the risk of venous thromboembolism in women whose first pregnancy ended with induced abortion. METHODS: We did this propensity score-matched cohort study using data from the universal health-care system of Ontario, Canada. We included primigravid women who had an induced abortion between Jan 1, 2003, and Dec 31, 2015, and used a propensity score to match them to primigravid women who had a livebirth (1:1) or nulligravid women who were not pregnant on the procedure date of their matched counterpart and who did not conceive within 1 year afterwards (5:1). We excluded from our analysis women younger than 15 years or older than 49 years and individuals who had missing or invalid information about their sex, area of residence, residential income, or world region of origin. The primary outcome was risk of any venous thromboembolism within 42 days of the index date (defined as the date of an induced abortion, delivery date for livebirth, or for non-pregnant women the induced abortion date of their matched counterpart). We compared the rate of venous thromboembolism in primigravid women who had an induced abortion with the rate of venous thromboembolism in propensity-score-matched non-pregnant women and propensity-score-matched primigravid women whose pregnancy ended with a livebirth. We generated hazard ratios (HRs) of 42-day risk of venous thromboembolism after induced abortion using Cox proportional hazard models. FINDINGS: We identified 194 086 eligible women whose first pregnancy ended with induced abortion, of whom 176 001 (90·7%) could be matched with women whose first pregnancy ended in delivery of a newborn. These 176 001 women were also matched to 880 005 non-pregnant women. The rate of venous thromboembolism within 42 days of an induced abortion was 30·1 (95% CI 22·0-38·2) per 100 000 women compared with 13·5 (11·1-16·0) per 100 000 women in the non-pregnant group (HR 2·23, 95% CI 1·61-3·08). The HR was 0·16 (95% CI 0·12-0·22) when compared with the women in the livebirth cohort, whose venous thromboembolism rate within 42 days postpartum was 184·7 (95% CI 164·6-204·7) per 100 000 women. INTERPRETATION: The 42-day risk of venous thromboembolism after an induced abortion is double that of a matched non-pregnant woman, and is significantly lower than after a livebirth. This novel information can inform estimates of peri-procedural risk of venous thromboembolism after induced abortion. Clinicians could consider a lower threshold for ordering a diagnostic test to rule out venous thromboembolism after induced abortion than they would in a non-pregnant woman. FUNDING: Institute for Clinical Evaluative Sciences.
引用
- Liu N, Vigod SN, Farrugia MM, et al. Venous thromboembolism after induced abortion: a population-based, propensity-score-matched cohort study in Canada. Lancet Haematol. 2018. PMID: 29891177. doi:10.1016/S2352-3026(18)30069-3
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。