Geographical variations in patient-reported outcomes after total hip arthroplasty between 2008 - 2012

• Main Authors: Linnea Oldsberg, Göran Garellick, Ingrid Osika Friberg, Anke Samulowitz, Ola Rolfson, Szilárd Nemes
• Format: Article
• Language: English
• Published: BMC 2019-05-01
• Series: BMC Health Services Research
• Subjects: Patient-reported outcomes; Total hip arthroplasty; Geographical variations; Equity in health care
• Online Access: http://link.springer.com/article/10.1186/s12913-019-4171-5

Abstract Background Health care on equal terms is a cornerstone of the Swedish health care system. Total hip arthroplasty (THA) is considered a success story in Sweden with low frequency of reoperations and restored health-related quality of life (HRQoL). Administratively, health care in Sweden is locally self-governed by 21 counties. In this longitudinal nation-wide observational study we assessed the possible geographical variations in 1-year follow-up patient-reported outcomes (PROs): EQ-5D index, EQ VAS, Pain VAS and Satisfaction VAS. Methods Study population consisted of 36,235 Swedish THA patients, operated during 2008 to 2012 due to hip osteoarthritis. Individual data came from Swedish Hip Arthroplasty Register, Statistics Sweden and National Board of Health and Welfare. We used descriptive statistics together with multivariable regression analysis to analyse the data. Results We observed county level differences in both preoperative and postoperative PROs. The results showed that the differences observed in preoperative PROs could not fully explain the differences observed in postoperative PROs, even after adjustment for patient demographics (age, sex, BMI, Elixhauser comorbidity index, marital status, educational level and disposable income). This indicates that other factors might influence the outcome after THA. Conclusion Likely, structural and process differences such as indication for surgery have an influence on PROs after surgery. Standardization of care at hospital levels may decrease geographical variations in postoperative HRQoL. Remaining differences will then possibly be associated to patient demographics.