The influence of preoperative determinants on quality of life, functioning and pain after total knee and hip replacement: a pooled analysis of Dutch cohorts

BACKGROUND Previous research has identified preoperative determinants that predict health related quality of life (HRQoL), functioning and pain after total knee or hip arthroplasty (TKA/THA), but these differed between studies and had opposite directions. This may be due to lack of power and not adjusting for confounders. The present study aims to identify the preoperative determinants that influence health related quality of life (HRQoL), functioning and pain after total knee or hip arthroplasty (TKA/THA). METHODS We pooled individual patient from 20 cohorts with OA patients data (n = 1783 TKA and n = 2400 THA) in the Netherlands. We examined the influence of age, gender, BMI and preoperative values of HRQoL, functioning and pain on postoperative status and total improvement. Linear mixed models were used to estimate the effect of each preoperative variable on a particular outcome for each cohort separately. These effects were pooled across cohorts using a random effects model. RESULTS For each increase in preoperative point in HRQoL, the postoperative HRQoL increased by 0.51 points in TKA and 0.37 points in THA (SF-36 scale). Similarly, each point increase in preoperative functioning, resulted in a higher postoperative functioning of 0.31 (TKA) and 0.21 (THA) points (KOOS/HOOS-ADL scale). For pain this was 0.18 (TKA) and 0.15 (THA) points higher (KOOS/HOOS-pain scale) (higher means less pain). Even though patients with better preoperative values achieved better postoperative outcomes, their improvement was smaller. Women and patients with a higher BMI had more pain after a TKA and THA. Higher age and higher BMI was associated with lower postoperative HRQoL and functioning and more pain after a THA. CONCLUSIONS Patients with a better preoperative health status have better outcomes, but less improvement. Even though the independent effects may seem small, combined results of preoperative variables may result in larger effects on postoperative outcomes.