Prevalence and risks factors of caregiving-related low back pain among caregivers of stroke survivors: a systematic review and meta-analysis.

All 5 included studies were cross-sectional and all were hospital based, and the review could not get extra information from the authors of those studies, so several details of who took part and how they were assessed are missing. The studies disagreed substantially with one another on how common low back pain was (I2=96%), and the authors point out that with fewer than 10 studies that estimate of disagreement is itself unreliable. Each risk-factor comparison came from only two studies, and both showed significant heterogeneity (I2=86% and I2=81%), so the null results for sex and for partial versus complete caregiving are weak. Every included study fell down on the same three points: no repeat checking of the main measure, no explanation of why individual participants were left out of analyses, and no description of how confounders were assessed or controlled. One of the 5 studies was rated high risk of bias and the other 4 moderate, and only 2 studies reported disability levels, so how low back pain relates to disability could not be pooled.

The authors declare no conflict of interest with any financial organisation regarding the material discussed. The work was funded by the research funding of the Research Centre for Chinese Medicine Innovation of the Hong Kong Polytechnic University (ref. N. P0041139), awarded to Prof Shamay Ng and her team, and by a PolyU Distinguished Postdoctoral Fellowship Scheme (N. P0035217).

Do not conclude that men and women who care for stroke survivors are at equal risk of low back pain. The pooled comparison found no significant difference, but it rested on only two studies with wide confidence intervals and significant heterogeneity, and the individual studies plus previous observational evidence point to higher risk in women.