Shoulder dislocation in Ontario, Canada from 1994 To 2011: The incidence, rate and risk factors for recurrence

Abstract

Objectives: Recurrent shoulder dislocation is influenced by age, activity level and bone loss. Original estimates of recurrence risk approached 90% among persons under the age of 20, but declined with increasing age. Recent literature, however, suggests that the rate of recurrent dislocation is lower. The goals of this study were to: (1) define the incidence of primary shoulder dislocation in Ontario, Canada, and (2) identify the rate of and risk factors for recurrent dislocation among demographic variables. Methods: Administrative databases (OHIP) were used to build the cohort of patients aged 15 to 70 that underwent a primary closed shoulder reduction by a physician in Ontario between July 1994 and October 2009. Exclusions included: associated humeral neck fracture, posterior dislocation, previous shoulder dislocation, prior shoulder arthroplasty, and non-Ontario residents. After cohort entry, subsequent shoulder relocations by a physician were sought. The yearly incidence (per 100,000 person-years) was calculated among all eligible Ontario residents. Kaplan-Meier survival curves to subsequent dislocation were generated. A Prentice, Williams and Peterson conditional proportional hazards survivorship model of time-to-recurrence was applied examining the influence of age, gender, income quintile, physician specialty and concurrent tuberosity fracture at index dislocation on the risk of recurrence (alpha set at 0.05). Hazard Ratios (HR) with confidence intervals were calculated. Results: The primary dislocation cohort consisted of 37,356 patients. Median age was 34 years (IQR 22-50) and 74% were male. The average yearly incidence of primary shoulder dislocation was 23.1/100,000 person-years overall, but 45.2/100,000 person-years for patients younger than 20. Recurrent dislocation events were identified in 23% of the cohort (8573 patients), most of whom were younger (median age 24 years) and male (80%). In fact, patients younger than 20 had a 37.8% rate of recurrence (HR 1.9 (1.7-2.1), p<0.0001; compared to patient aged 36-40). Kaplan-Meier survival curves showed most recurrent dislocations took place in the first year: 93.4% at 6-months, 89.4% at 1-year, 85.2% at 2-years and 79.4% at 5-years. Protective factors against recurrence included primary relocation performed by an Orthopaedic Surgeon (HR 0.87 (0.79-0.94), p=0.001), age over 50 years (5.5% rate; HR 0.70 (0.61-0.80, p<0.0001) and an associated tuberosity fracture (HR 0.51 (0.42-0.63), p<0.0001), while lowest income quintile was a risk factor for recurrence (HR 1.1 (1.04-1.16), p=0.0007). Interestingly, at 2.5 years from the primary dislocation, only 14.7% of the cohort had undergone surgical shoulder stabilization. Conclusion: Patients under 20 had twice the incidence of primary dislocation and twice the risk of recurrent shoulder dislocation compared to the median cohort age. A recurrence rate of 38% in patients under 20 is high, but less than previous reports.