原作资料
追加：搬运姿势与腰痛 · 核实记录（2026-10-04）
任务来源：issue #76 读者建议加一条「搬重物要屈膝、让重物贴身、别弯腰猛提」，附 HSE 指南和 Saraceni 2020（DOI 10.2519/jospt.2020.9218）。用户定：不按原稿收。「腿使劲别弯腰」是常识，而且手上的证据指向反面。改成第 6 节（反面清单）节末第 29 条，否掉「学会正确搬运姿势、上搬运培训就能防腰痛」这个说法。
原有覆盖：全书搜「腰痛」「腰背痛」「搬重物」「搬运」，没有讲搬东西和腰痛的条目，只有第 27 节第 8 条无痛分娩备注里提到「长期腰痛」，和本条无关。
靶子
读者真信的是两句：「只要姿势对（屈膝、不弯腰）就不会伤腰」和「腰弯着搬一定伤腰」。前一句由 Cochrane 培训试验驳（培训改不了腰痛结局），后一句由 Saraceni 驳（腰弯多少和腰痛无关）。读者接着会问「那到底怎么搬」，答案（减重量和次数、规律锻炼）按 CLAUDE.md「否定式条目」的规矩写进收益栏，说人话也只用收益栏里的数字。
数字与原文
左右滑动查看整张表格
| 数字 | URL | 原文引句（逐字，取自 Europe PMC 摘要或官网正文） |
|---|---|---|
| 9 项随机试验、20101 名员工，另有 9 项队列 1280 人 | https://doi.org/10.1002/14651858.CD005958.pub3（Europe PMC PMID 21678349） | "We included nine RCTs (20,101 employees) and nine cohort studies (1280 employees) on the prevention of back pain in this updated review." |
| 7 项试验 19317 人，OR 1.17（0.68–2.02），视频对照 RR 0.93（0.69–1.25），中等质量 | 同上 | "There was moderate quality evidence from seven RCTs (19,317 employees) that those who received training reported levels of back pain similar to those who received no intervention, with an odds ratio of 1.17 (95% confidence intervals (CI) 0.68 to 2.02) or minor advice (video), with a relative risk of 0.93 (95% CI 0.69 to 1.25)." |
| 作者结论：配不配工具都防不住 | 同上 | "There is moderate quality evidence that MMH advice and training with or without assistive devices does not prevent back pain or back pain-related disability when compared to no intervention or alternative interventions. There is no evidence available from RCTs for the effectiveness of MMH advice and training or MMH assistive devices for treating back pain." |
| 可信范围很宽 | 同上 | "Confidence intervals around the effect estimates were still wide due to the adjustment for the design effect of clustered studies." |
| 跟踪研究差 1.5 度（−0.7 到 3.7） | https://doi.org/10.2519/jospt.2020.9218（PMID 31775556） | "found no difference in peak lumbar spine flexion when lifting (1.5°; 95% confidence interval [CI]: -0.7°, 3.7°; P = .19 for the longitudinal study and -0.9°; 95% CI: -2.5°, 0.7°; P = .29 for the cross-sectional studies)" |
| 7 项研究腰痛者少弯 6.0 度（−11.2 到 −0.9） | 同上 | "Seven cross-sectional studies measured lumbar flexion with thoracopelvic angles and found that people with LBP lifted with 6.0° less lumbar flexion than people without LBP (95% CI: -11.2°, -0.9°; P = .02)." |
| 11 项里 9 项无差别，证据质量低 | 同上 | "Most (9/11) studies reported no significant between-group differences in lumbar flexion during lifting. The included studies were of low quality." 结论："There was low-quality evidence that greater lumbar spine flexion during lifting was not a risk factor for LBP onset/persistence or a differentiator of people with and without LBP." |
| 8 项跟踪研究；每 10 公斤 OR 1.11（1.05–1.18）；每天每 10 次 OR 1.09（1.03–1.15） | https://doi.org/10.1136/oemed-2014-102346（PMID 25165395） | "Eight longitudinal studies were included. Pooled estimates resulted in ORs of 1.11 (1.05 to 1.18) per 10 kg lifted and 1.09 (1.03 to 1.15) per 10 lifts/day." |
| 21 项试验、30850 人；锻炼加讲解 RR 0.55（0.41–0.74）中等质量；对病假 0.74（0.44–1.26） | https://doi.org/10.1001/jamainternmed.2015.7431（PMID 26752509） | "23 published reports (on 21 different randomized clinical trials including 30,850 unique participants) met the inclusion criteria. … there was moderate-quality evidence that exercise combined with education reduces the risk of an episode of LBP (0.55 [0.41-0.74]) and low-quality evidence of no effect on sick leave (0.74 [0.44-1.26])." |
| 单靠锻炼 RR 0.65（0.50–0.86），低到很低质量 | 同上 | "Low- to very low-quality evidence suggested that exercise alone may reduce the risk of both an LBP episode (0.65 [0.50-0.86]) and use of sick leave (0.22 [0.06-0.76])." |
| 单靠讲解 RR 1.03（0.83–1.27）；护腰带 RR 1.01（0.71–1.44） | 同上 | "For education alone, there was moderate- to very low-quality evidence of no effect on LBP (1.03 [0.83-1.27]) … There was low- to very low-quality evidence that back belts do not reduce the risk of LBP episodes (1.01 [0.71-1.44])" |
| HSE：培训单靠自己不够 | https://www.hse.gov.uk/msd/manual-handling/training.htm | "Training can be important in raising awareness and reducing risk, but it won't ensure safe manual handling on its own. You should first design the manual handling operations to be as safe as reasonably practicable." |
| HSE：手法替代不了其他措施；微弯优于完全弯腰或深蹲 | https://www.hse.gov.uk/msd/manual-handling/good-handling-technique.htm | "Good handling technique is not a substitute for other steps you should take to reduce risk, like providing lifting aids, or making improvements to the task, load or working environment, but it is helpful as an addition to those measures." "Slight bending of your back, hips and knees at the start of the lift is preferable to either fully flexing your back (stooping) or fully flexing your hips and knees (full/deep squatting)." 另有 "Keep the load close to your waist"、"Avoid twisting your back"、"Do not jerk or snatch the load" |
| 反方：4 项有关联、5 项没有，不太可能单独致病 | https://doi.org/10.1016/j.spinee.2010.03.033（PMID 20494816） | "35 studies met eligibility criteria and 9 were considered high methodological quality studies … there was conflicting evidence for association with four studies reporting significant associations and five studies reporting nonsignificant results." "Based on these results, it is unlikely that occupational lifting is independently causative of LBP in the populations of workers studied." |
说人话里的换算：「风险高约一成」对应 OR 1.11 和 1.09；「少三成多到四成多」对应 RR 0.65（低约 35%）和 0.55（低约 45%）。
版本核查
Cochrane CD005958 最新版是 pub3（2011-06-15）。Crossref 查 10.1002/14651858.CD005958.pub4、.pub5 均 404，Europe PMC 也只有 pub2（2007）和 pub3（2011）。2012 年 Occup Environ Med 69:79-80 那篇是同一综述的期刊摘要版，不另引。
定级
- 证据等级 A：Cochrane 综述（中等质量）和 JAMA Intern Med 的随机试验荟萃都给了具体数字。标「争议」，理由写在备注开头：负重和腰痛的关系有 Wai 2010 的反方，Saraceni 的研究质量低，Cochrane 的可信范围宽。
- 口径：死亡率（健康终点，这里是腰痛发作）。
- 收益量级「中」：按阈值，锻炼加讲解低约 45%、单靠锻炼低约 35%，机械套是「大」。降一档的理由：单靠锻炼的证据质量低到很低；锻炼加讲解对请病假没看出效果；腰痛不致命。条目本身否掉的培训一项效果为零，换回的主要是不再为姿势和护腰产品花心力和钱。
- 成本标签：钱=0；时间=中（每周锻炼，小时级）；毅力=些（要养成锻炼习惯）。
没收的内容
- 读者原稿的「屈膝、贴身、别弯腰猛提」：用户定不按原稿收。HSE 指南本身也没要求深蹲，反而写微弯背、髋、膝优于完全深蹲。贴身、别扭腰、别猛拽这几条放进备注，写明「照做没坏处，只是别指望它单独防住腰痛」。
- Coenen 2014 的「搬 25 公斤以上，年发病率增加 4.32%」：原文 "increase the annual incidence of LBP by 4.32%" 没说清是百分点还是相对增幅，读者容易读错，不写。
- Richarz 2023 护理人员移位辅助器械荟萃（Journal of Occupational Health 65:e12423，PMID 37712457）：摘要说所有结局都显著，但腰痛那项 g = 1.54 的可信范围是 −0.016 到 3.088，跨过 0，前后矛盾；只针对护理人员搬病人。没收，推车、找人这类出路只按 Coenen 的剂量关系写，不写成有试验证明。
- 护腰带单独成条：Steffens 里护腰带 RR 1.01 已写进收益栏，并入本条，不另开。
- 已有腰痛怎么治：Cochrane 说培训和工具治腰痛没有随机试验，只在备注提一句；治疗不属于本条。
- 受益人是你自己（第 ① 档）。