Low back pain is a leading cause of disability, more prevalent in high-income countries than in low- and middle-income countries.1 However, these differences may be partially explained by the lack of high-quality epidemiologic and clinical research in low- and middle-income countries.2 Nevertheless, low back pain dramatically impacts people's lives and health systems regardless of a country's income level.3–5

Despite the high burden across countries, most low back pain research has been conducted in high-income countries.6–9 Research in these countries has systematically excluded people from linguistically diverse backgrounds, with one in five physical therapy trials excluding participants based on language proficiency.10 These two pieces of information raise questions about whether evidence generated in high-income countries is applicable and transferrable to low- and middle-income countries. Among the reasons that explain the underrepresentation of research conducted in low- and middle-income countries are research infrastructure and funding mechanisms.3

Emerging evidence suggests clinicians and researchers have biases toward research conducted in high-income countries.11 Observational studies reveal clinicians and researchers implicitly associate high-quality research with rich countries compared to developing countries.12,13 For example, a recent randomized experiment showed that changing the affiliation of the lead author of a study from a low- to a high-income country positively affected the perception of the relevance of the research and the likelihood of recommending that research to a peer.12,13 Some studies have shown that trials conducted in low-income countries are more likely to have lower methodological quality14 and more commonly report favorable treatment effects15 compared to trials conducted in high-income countries. Furthermore, a few studies have focused on specific differences across methodological quality domains between trials from countries with different income levels.14 According to some authors worldwide, this difference between income levels could be similar in low back pain trials. However, there is a lack of data to support this assumption, and the magnitude of the difference is unknown.16–19

There is an international concern that most interventions reported in guidelines have been tested in high-income countries; some cannot be applicable in contexts with fewer resources because of the different work policies, education access, and health systems.8,16–19 Also, most of these published trials have a low methodological quality,20 so the applicability of these treatments in low- and lower-middle-income countries is questionable. In addition, it is important to know the quality of trials from low- and lower-middle-income countries because low-quality trials represent a waste of scarce resources for research because this information cannot be used in practice or included in systematic reviews.14

The main objective of this meta-epidemiological study was to compare the methodological quality of randomized controlled trials of physical therapy interventions for low back pain conducted in countries with different income levels. Secondary aims were to describe which quality domains were more frequently achieved and the characteristics of the trials (trial design, intervention type used, purpose of the trial, treatment type evaluated) and the participants (age, duration of low back pain, nonspecific or specific diagnosis) in trials from countries with different income levels.

We conducted a meta-epidemiological study21 and registered the protocol a priori in the Open Science Framework.22 All data will be publicly available (see Supplementary material).

We hypothesize that the PEDro score of trials of physical therapy interventions for low back pain performed in countries with a high-income level will be higher than that performed in low- and lower-middle-income countries.

Most (approximately 70%) randomized controlled trials evaluating physical therapy interventions for people with low back pain were conducted in high-income countries, this raised the concern that most interventions for low back pain could not be suitable to the context and needs of people from countries with less resources.7,8,16,18,19 Slight differences in methodological quality were observed between income level groups, with trials from low- or lower-middle-income countries having lower total PEDro scores compared to trials from upper-middle-income countries and high-income countries. A more considerable proportion of trials from high-income countries achieved the eligibility criteria and source, concealed allocation, blinding of assessors, intention-to-treat analysis, and point and variability measures items of the PEDro scale compared to upper-middle-income and low- or lower-middle-income countries. Completeness follow-up, intention-to-treat analysis, and point and variability measures were more common in trials from upper-middle-income than low- or lower-middle-income countries. An explanation is that having more economic resources could promote the accomplishment of several characteristics required for a trial with high methodological quality, like research assistants, adequate infrastructure, a multidisciplinary team, access to adequate technology, and enough support from public and academic institutions.8

Our study has the following strengths: We included a large sample size of trials published in 15 languages. A rigorous process was used to evaluate the methodological quality of these trials as we downloaded PEDro scores from PEDro, which means that pairs of trained raters generated scores. Also, other trial data were extracted by pairs of independent reviewers using a pilot-tested spreadsheet and clear written instructions.

On the other hand, the study had some limitations that can underestimate or overestimate the results like the unidimensional assessment of income level based on GNI, so we did not analyze other factors related to economic resources that could influence the quality of trials, like external funding or industry sponsorship, because it is complex to measure these variables in all trials included.

Another limitation is that we used the World Bank income classification of 2020 independently of the year of publication of each trial; this could represent a measurement bias because the income classification of some countries may have changed recently.

Also, the PEDro scale was designed to appraise the methodological quality of randomized controlled trials, so it may underestimate those with the economic evaluations in our study. Nevertheless, less than 2% of trials were classified as economic evaluations, so the impact of this issue might be non-significant in our overall results. In addition, we did not extract other variables related to the methodological quality of trials, like research expertise, use of technology, or institution support, which could explain in different ways the differences found in the quality of trials acting as confounding variables that could be controlled in future studies. Moreover, we only searched articles in three databases using a search strategy that considered three languages, but it is important to note that we did not restrict the trials retrieved by language. One of the databases consulted was PEDro, the most comprehensive database of trials in physical therapy.23,31

Lastly, we recognize that we may have omitted some articles, mostly from low- and lower-middle-income countries, if we consider that most research produced there is written in languages different than English and published in gray literature, local journals, or predatory journals; besides we did not consider the work that is not published, but may have good quality.8,14,34

Our finding that the majority of trials were conducted in high-income countries was similar to trials included in Cochrane reviews that evaluated interventions for hypertension, diabetes, stroke, or heart disease (78%),14 smoking cessation (96%),35 or reported mortality outcomes (76%),15 and bibliometric studies of low back pain research that indicates high-income countries were the most productive.6,36 Factors that may explain the low scientific productivity of low- and middle-income countries include the pressure placed on clinician-researchers to manage a sizeable clinical caseload in the presence of financial constraints,37 a few researchers with enough academic formation, lack of research awareness from the society, failure to publish in scientific journals due to English-language proficiency,8,38 prohibitive publication costs,39–42 and publication bias.8,12,38,43 The low prevalence of low back pain trials from low- and middle-income countries prompted a key recommendation in the Lancet series on low back pain, which recommended that more trials were needed in low-resource settings to identify and design context-specific strategies for managing back pain.7

Trials evaluating physical therapy interventions for low back pain have an average total PEDro score of 5.5 out of 10, consistent with other meta-epidemiological studies that report mean scores of 5.3 for musculoskeletal populations44 and 4.7 for cardiorespiratory populations.45 These studies also report a similar percentage achievement of the PEDro scale items. Gonzalez et al.44 investigated predictors of methodological quality in trials of physical therapy interventions for musculoskeletal conditions, concluding that adherence to reporting guidelines, sample size calculation, fewer primary outcomes, and publication in English were associated with higher methodological quality. However, they did not investigate whether the country's income level was related.

Our finding that there are minor differences in methodological quality for trials conducted in countries with different income levels is consistent with a meta-epidemiological study that investigated associations between country income level and risk of bias for trials included in Cochrane reviews.14 Compared to trials conducted in middle-income countries, trials in high-income countries were more likely to have a low risk of bias, including a lower risk for sequence generation, concealed allocation, and blinding.14 In contrast, the country's income level had little impact on the quality of trials included in Cochrane reviews.35

Regardless of the country's income level where the trial was conducted, we observed that methodological quality was low for low back pain trials. Some barriers are likely to impact the quality of randomized controlled trials,8,46 particularly in low- and middle-income countries. Improving financial and institutional support as well as research training and mentoring, could reduce bias in the trial design and encourage better reporting of outcomes in research articles.8 However, we must be cautious with these recommendations as we need more information about what causes the observed differences.

As a recommendation, researchers could consider using established, freely available resources to improve the quality of their trials (i.e., trial registration,47 Standard Protocol Items Recommendations for Interventional Trials [SPIRIT],48 Consolidated Standards of Reporting Trials [CONSORT]49). Funders should invest in multi-country trials to evaluate cost-effective interventions to treat or prevent low back pain and journals could implement strategies to encourage publishing and adequate reporting of trials from low- and lower-middle-income countries like ensure blinded peer-review processes, include editors from diverse contexts, and fair article processing charges to authors from countries with less resources.8 This action could improve the methodological quality of trials of physical therapy interventions in patients with low back pain in countries with any income level.

We recognize that there is a wide range of factors, different from income level, that can influence methodological quality; future research should describe which variables could explain a low quality of trials of low back pain interventions. Furthermore, it is important to know if strategies to improve research training and facilitate access to funding could increase the quality of future trials.

There is a minor difference in the methodological quality of trials of physical therapy intervention for low back pain between the income levels of the countries where these studies were conducted. Moreover, seven out of every 10 trials of physical therapy interventions for low back pain are conducted in high-income countries. Therefore, more trials of physical therapy interventions for low back pain from low- and middle-income countries are needed. Finally, implementing strategies to improve the methodological rigor of trials in patients with low back pain is necessary in all countries, regardless of income level.

Carlos Maximiliano Sánchez Medina: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Project administration. Claudia Gutiérrez Camacho: Conceptualization, Methodology, Investigation, Formal analysis, Writing – original draft, Writing – review & editing, Supervision. Anne M. Moseley: Methodology, Software, Investigation, Validation, Writing – original draft, Writing – review & editing. Xochiquetzalli Tejeda Castellanos: Formal analysis, Investigation, Writing – review & editing, Data curation. Qiuzhe Chen: Methodology, Validation, Investigation, Data curation. Edgar Denova-Gutiérrez: Formal analysis. Aidan G Cashin: Methodology, Validation, Writing – review & editing. Viridiana Valderrama Godínez: Investigation. Akari Fuentes Gómez: Investigation. Ana Elisabeth Olivares Hernández: Investigation. Giovanni E. Ferreira: Methodology, Validation, Investigation, Writing – original draft, Writing – review & editing.