Review on knee and hip arthroplasty outcomes in sub-Saharan Africa

Dear Editor,

We read with interest the article by Davies et al., ¹ describing their findings on the systematic review of total joint replacement surgery in sub-Saharan Africa (SSA). However, we note significant methodological and reporting flaws that may affect the utilisation of these findings for enhanced care arthroplasty programs across this region.

It is noteworthy that the authors reported adherence to the PRISMA recommendations ² for their review. Nonetheless, the search strategy described was inadequate in achieving high sensitivity. First, it did not include relevant keywords and selective outcomes such as HIV as reported in the results and each of the 47 countries in SSA. We replicated the published search in PubMed and retrieved 686 citations. Including all countries in the region in the search yielded an additional 13 citations. This is a sharp contrast to the 191 citations reported by the authors for the same database. Second, limited access to funding for research conducted in SSA reduces the likelihood that studies related to this topic would be published in the databases listed. ³ Additional methods to examine other sources such as regional-specific databases such as African Journal Online, grey literature or unpublished data are recommended to minimise bias. ⁴ Third, it is unclear whether one or both authors were involved in the search and screening processes and how discrepancies were addressed. It is imperative that the roles of the authors are clearly delineated so that end-users can make informed assessments on the sources of bias and its impact on the validity of the findings. ²

Further, the authors made a poor justification for excluding South Africa from the list of SSA nations. They opined that despite its high poverty rate, South Africa is an outlier in terms of development and its health system is comparable to that of developed nations. However, according to the United Nations Development Programme, ⁵ South Africa has high inequalities in health, income and education which mirrors socioeconomic indices found in other countries in the region and may have similar population characteristics. In addition, the review question was not predicated on the economic status of individual countries and therefore rendered this justification inconsequential. Excluding studies from South Africa may affect the generalisability of the findings.

Notwithstanding the efforts by Davies et al. to contribute to the growing body of knowledge in arthroplasty in SSA, we conclude that the accuracy of the conclusions drawn from the review is questionable and should be interpreted with caution.