The problem
“Despite research showing how social determinants of health (SDOH) affect orthopedic outcomes, little is known about how many THA and TKA randomized controlled trials (RCTs) actually include SDOH variables such as race, ethnicity, insurance, income, and education.”
The approach
“My co-authors and I conducted a PubMed literature review of RCTs published from 2017-2019 in four orthopedic journals: The Journal of Bone and Joint Surgery (JBJS), Journal of Arthroplasty, Clinical Orthopaedics and Related Research (CORR), and Osteoarthritis and Cartilage.”
“Using our inclusion criteria that the study must pertain to either TKAs or THAs and include a patient demographic table, we ended up with 72 publications. We collected data on the publication year, type of surgery, and the inclusion of race, ethnicity, insurance, income, and education in the demographic table or discussion.”
“In addition, we performed an analysis of counts and percentages of SDOH variables by journal year, journal name, and surgery type; Fisher’s exact tests were used for comparison.”
The results
“We found that race was discussed in 4 of the manuscripts, and only included in 3 of the demographic tables. Ethnicity was discussed in 3 manuscripts and included in 1 demographic table, while insurance, income, and education were not included in any of the discussions or demographic tables. None of the SDOH variables were included in any of the multivariable regressions. Overall, only 5 of the manuscripts included any 1 of the SDOH variables, and only 3 included any of these variables within the demographic table. Therefore, there is a less than 10% reporting rate of any of the SDOH variables.”
“Looking at the publications broken down by year, journal, and surgery type, we found that in 2017 there were only 2 publications that discussed any one of the SDOH variables, only 1 in 2018, and 2 in 2019. We found an insignificant difference within the 3 years analyzed, showing that there was no significant improvement and there was not an occurrence of just one bad year.”
“When broken down by journal type, no publications met the inclusion criteria in Osteoarthritis and Cartilage. Only 4 publications from the Journal of Arthroplasty discussed any of the SDOH variables; 1 publication in CORR addressed these variables, but no articles in JBJS looked at SDOH. We found an insignificant difference when broken down by journal type. They were also divided by the type of surgery, and we found that there was an insignificant difference.”
“Thus, when looking at 72 publications, we found that SDOH variables are vastly underreported in the TKA/TKA literature. There is growing evidence that SDOH variables impact outcomes and should be considered as consequential as gender and body mass index.”
“Inclusion of these SDOH variables allows researchers to determine the applicability of RCT findings to various patient populations outside of this study. We found an insignificant difference over the 3 years analyzed. It is also important to note that we only looked at RCTs, meaning that investigators had the choice of considering SDOH variables.”
“In conclusion, there is a significant shortcoming in the inclusion of SDOH variables in the TKA/TKA literature. Exclusion of these variables when establishing national standards and policies may be indirectly perpetuating disparities.”
“Research that does not rely on representative patient samples should be used with caution. We are advocating for the standardization of the inclusion of SDOH variables. Our call to action is that journals and their audiences should demand greater inclusivity and diversity in research.”