Differences between original research and big database publications in hip arthroscopy
Author: Jessica Boh
Quick Summary: A systematic review found rates of major complications, reoperations, and conversion to arthroplasty following hip arthroscopy are higher in database research than original research. The disparity is attributed to differences in surgeon experience, demographics and length of follow-up, as individual research papers tend to report outcomes for surgeons with a higher level of specific expertise and lower risk patients with shorter follow-up.
Hip arthroscopy is an increasingly common technique used for numerous conditions, in part due to low rates of complication, reoperation and conversion to total hip arthroplasty reported in the literature. However, there are discrepancies in the complication and reoperation rates reported by original research and database publications.
Sochacki and colleagues (2018) have conducted a systematic review and meta-analysis to investigate the differences in hip arthroscopy outcomes reported by individual research and database publications. They hypothesised that database studies would have higher rates of complications, reoperations and conversions to total hip arthroplasty compared to original research publications.
The systematic review followed PRISMA guidelines to identify hip arthroscopy clinical outcome studies published in English. Papers reporting on open or mini-open hip surgery and surgical technique papers were excluded. Studies were divided into two groups for comparison and demographics, complications, reoperations, and conversions to THA were extracted.
Two hundred and seven papers were included in the final analysis, with data for 16,260 hips from individual studies (201 papers) and 21,044 hips from database studies (six papers).
Chi squared comparisons found differences between original research and database publications. Database articles reported a higher percentage of cases of reoperation, conversion to hip arthroplasty (total hip arthroplasty, hemi, or resurfacing), major complications, femoral neck fracture, and hip dislocation. Most significantly, the study found the relative risk of femoral neck fracture was eight times greater in database studies than individual studies.
The authors suggest that surgeon experience may explain a large amount of variation in outcomes. Most database studies capture the long learning curve of hip arthroscopy, whereas individual studies tend to report data of high-volume experienced orthopaedic surgeons who are often hip arthroscopy experts. Patients in the database studies were also more likely to be female, older and have higher BMI than those in the original research publications, and these are all risk factors for poorer outcomes following hip surgery. Additionally, rates of reoperation and conversion increase with length of follow-up, and the database studies had a significantly higher mean follow-up of 3.1 years compared to 2.7 years in original research.
While it is difficult to identify its precise cause, the difference in outcomes between original research and database publications highlights the importance of context when interpreting research. Both surgeons and researchers alike must consider not only the quality of methodology and data analysis but also the generalisability of a publications findings to their own situation, in their effort to apply orthopaedic research for best practices.
For more detail, access the full article here.
Sochacki, K. R., Jack, R. A., II, Safran, M. R., Nho, S. J., & Harris, J. D. (2018). There Is a Significant Discrepancy Between “Big Data” Database and Original Research Publications on Hip Arthroscopy Outcomes: A Systematic Review. Arthroscopy: The Journal of Arthroscopic & Related Surgery: Official Publication of the Arthroscopy Association of North America and the International Arthroscopy Association. https://doi.org/10.1016/j.arthro.2018.01.018