Outcomes of total hip and knee arthroplasty in special populations: a synopsis and critical appraisal of systematic reviews

Introduction This study aimed to present and critically appraise the best available evidence investigating associations between some pre-defined patient-related characteristics and perioperative complications or other outcomes in THA and TKA. Methods Electronic databases were searched (Medline, EMBASE, Scopus, CENTRAL) for systematic reviews assessing the following pre-defined patient-related characteristics as possible risk factors for worse peri-operative outcomes in THA and TKA: smoking, alcohol excess, rheumatoid arthritis, human immunodeficiency virus infection, hepatitis C virus infection, mental health conditions, and solid organ transplantation. Our primary outcome was periprosthetic joint infection. Results were analysed separately for THA, TKA and THA/TKA (mixed data). Results Based on at least two systematic reviews being in agreement, the following patient-related characteristics were associated with increased incidence of complications as follows: a) Smoking for all-cause revision in THA, for periprosthetic joint infection in TKA and THA/TKA; b) alcohol excess for periprosthetic joint infection in THA/TKA; c) human immunodeficiency virus for periprosthetic joint infection in THA/TKA; d) hepatitis C virus for overall complications, periprosthetic joint infection and all-cause revision in THA and THA/TKA, and for overall complications in TKA. Our study found conflicting evidence for a) smoking as a risk factor for periprosthetic joint infection and aseptic loosening in THA; b) human immunodeficiency virus as a risk factor for all-cause revision for THA/TKA; c) hepatitis C virus as a risk factor for periprosthetic joint infection and all-cause revision in TKA. No certainty of evidence was assigned to these results as this was not assessed by the authors of the majority of the included systematic reviews. Conclusion We found that smoking, excess alcohol consumption, RA, and infection with HIV and HCV were associated with a higher incidence of periprosthetic joint infection in one or both of THA and TKA or mixed THA/TKA data. All our results should be interpreted and communicated to patients with caution as the quality of the included systematic reviews was generally poor.