Predicting poor outcome following total hip and knee arthroplasty
Research type
Research Study
Full title
A prospective observational (C) study in adult patients with primary osteoarthritis (P), to identify the physical and psychological risk factors predicting poor patient satisfaction and function (O) following primary total hip and knee arthroplasty (I) I - Intervention P - Population O - Outcome C - Comparator
IRAS ID
64866
Contact name
Joseph Dias
Contact email
Sponsor organisation
University Hospitals Leicester NHS Trust
Research summary
Summary of Results
The objectives of POSt-Plus including the pilot study were to incorporate patient public involvement (PPI) in order to help with the design and conduct of the main study (POSt-Plus), determine the feasibility of the study, explore ways of managing and mitigating against missing and incorrect data and explore the use of accelerometers. The POSt Plus study also analysed data looking into the factors that might help predict a poorer than expected outcome after hip or knee replacement surgery. A PhD thesis has been successfully completed and a PhD awarded based on some of the pilot work that led to the main POSt Plus study.
In summary, 532 patients were recruited to the study and 469 completed the study.
One of the issues raised from the pilot study involved the use of an incorrect questionnaire in the pilot study, in this case, the TSK-11 questionnaire. We analysed a method for managing this issue using a statistical technique to enable the data to still be used. We also provided a detailed plan for avoiding such an error in the main study and also explored options for mitigating against such an error if this were to occur again.
As with any clinical study, missing data can be a significant limitation in determining the results of a study. We assessed the levels of missing data in the variables Body mass index (BMI) and patient reported outcome measures such as the Oxford hip score (OHS) and Oxford knee score (OKS) and provided a detailed plan on how to minimize missing data in terms of how POSt was designed and conducted. We also used a specific statistical method for dealing with such missing data called multiple imputation. The amount of missing data at different time points during the study ranged from 0% to 12%.
Risk factors for predicting poorer outcome From the POSt-Plus data, several factors have been identified as being a significant influence on the initial outcome of patients having a hip or knee replacement. These include some psychological aspects of managing pain such as kinesiophobia and self-efficacy as well as the issue of sleep disturbance after joint replacements. Sedentary behaviour and low levels of activity prior to surgery was also found to influence outcome and interestingly, although hip and knee replacements are performed to help patients improve their function and mobility, this did not appear to be the case. The results from the accelerometers (activity monitors) showed that activity levels remained the same or indeed reduced after surgery.
Patients spent almost all of their time (97.5%) in either sedentary or light activity but only spent 2.1% of their time in moderate to vigorous physical activity (MVPA). After their knee replacement, at three months, there was a significant decrease in physical activity. The mean number of daily steps taken decreased significantly from 3191 to 2133 and percentage time spent in sedentary activity increased significantly from 79.4% to 85.1%.
For hip replacement patients, the mean number of daily steps taken increased from 3460 to 3599 and percentage time spent in sedentary activity remained unchanged 78%. Hip replacement patients also spent 98.1% of their time in either sedentary or light activity. Only 2% of their time was spent in MVPA.
Mean Oxford knee (OKS) and hip scores (OHS) did improve significantly however from 16 to 35 and from 16 to 37 respectively. This provides good evidence that from a patient’s perspective, hip and knee replacements did help relieve the pain associated with arthritis and helped with activities of daily living.
Work is currently ongoing in trying to develop a screening tool to help predict poor outcome and will need further validation studies to test the tool’s efficacy in a clinical setting.
REC name
East Midlands - Leicester Central Research Ethics Committee
REC reference
13/EM/0061
Date of REC Opinion
13 Feb 2013
REC opinion
Favourable Opinion