Does Study Design Affect Redislocation Rates After Primary Shoulder Dislocations? A Systematic Review Comparing Prospective and Retrospective Studies
- Additional Document Info
- View All
PURPOSE: To compare recurrence rates between prospectively collected and retrospectively collected data on primary anterior shoulder dislocations, as this could influence the timing of surgical decision making. METHODS: A comprehensive literature search of Medline, Embase, CINAHL, and hand searches was performed. Recurrence rates of anterior shoulder dislocations were collected from relevant articles, along with follow-up length, age, and gender. An independent sample t test was conducted to evaluate our hypothesis. A multiple linear regression model was used to examine the variance in recurrence rates while controlling for covariates. RESULTS: A total of 1,379 articles were identified, of which 25 were relevant to our study-16 prospective and 9 retrospective. The average rate of recurrence of anterior shoulder dislocations in retrospective studies (mean [M] = 45.2, standard deviation [SD] = 31.67) was not significantly different from that in prospective studies (M = 56.7, SD = 22.55). The 95% confidence interval for the difference of the means ranged from -34.05 to 10.91. After controlling for covariates with the multiple linear regression, only 1.9% of the variance in recurrence rates was due to study type and was not significant (P = .42). The t test performed to evaluate our hypothesis was also not significant t(23) = -1.07, P = .298. CONCLUSIONS: When comparing prospective and retrospective studies, there was no significant difference in recurrence rates of primary anterior shoulder dislocations treated nonoperatively. The average redislocation rate was 56.7% in prospective studies and 45.2% in retrospective studies. Furthermore, the majority of this difference was accounted for by varying rates between age groups. Further research is needed to determine the risk of redislocation in specific age groups, to guide treatment decisions based on varying risk. LEVEL OF EVIDENCE: Systematic review of Level II and III studies.
has subject area