Randomised controlled trial of the cost-effectiveness of water-based therapy for lower limb osteoarthritis

To determine the efficacy of community water-based therapy for the management of lower limb osteoarthritis (OA) in older patients.

A pre-experimental matched-control study was used to estimate efficacy of water-based exercise treatment, to check design assumptions and delivery processes. The main study was a randomised controlled trial of the effectiveness of water-based exercise (treatment) compared with usual care (control) in older patients with hip and/or knee OA. The latter was accompanied by an economic evaluation comparing societal costs and consequences of the two treatments.

Water exercise was delivered in public swimming pools in the UK. Physical function assessments were carried out in established laboratory settings.

106 patients (93 women, 13 men) over the age of 60 years with confirmed hip and/or knee OA took part in the preliminary study. A similar, but larger, group of 312 patients (196 women, 116 men) took part in the main study, randomised into control (159) and water exercise (153) groups.

Control group patients received usual care with quarterly semi-structured telephone interview follow-up only. The intervention in the main study lasted for 1 year, with a further follow-up period of 6 months.

Pain score on the Western Ontario and McMaster Universities OA index (WOMAC). Additional outcome measures were included to evaluate effects on quality of life, cost-effectiveness and physical function measurements.

Short-term efficacy of water exercise in the management of lower limb OA was confirmed, with effect sizes ranging from 0.44 [95% confidence interval (CI) 0.03 to 0.85] on WOMAC pain to 0.76 (95% CI 0.33 to 1.17) on WOMAC physical function. Of 153 patients randomised to treatment, 82 (53.5%) were estimated to have complied satisfactorily with their treatment at the 1-year point. This had declined to 28 (18%) by the end of the 6-month follow-up period, during which support for the intervention had been removed and those wishing to continue exercise had to pay their own costs for maintaining their exercise treatment. High levels of co-morbidity were recorded in both groups. Nearly two thirds of all patients had a significant other illness in addition to their OA. Fifty-four control and 53 exercise patients had hospital inpatient episodes during the study period. Water exercise remained effective in the main study but overall effect size was small, on WOMAC pain at 1 year, a reduction of about 10% in group mean pain score. This had declined, and was non-significant, at 18 months. Mean cost difference estimates showed a saving in the water exercise group of pound123--175 per patient per annum and incremental cost-effectiveness ratios ranged from pound3838 to pound5951 per quality-adjusted life-year (QALY). Net reduction in pain was achieved at a net saving of pound135--175 per patient per annum and the ceiling valuation of pound580--740 per unit of WOMAC pain reduction was favourably low.

Group-based exercise in water over 1 year can produce significant reduction in pain and improvement in physical function in older adults with lower limb OA, and may be a useful adjunct in the management of hip and/or knee OA. The water-exercise programme produced a favourable cost--benefit outcome, using reduction in WOMAC pain as the measure of benefit. Further research is suggested into other similar public health interventions. Investigation is also needed into how general practice can best be supported to facilitate access to participants for research trials in healthcare, as well as an examination of the infrastructure and workforce capacities for physical activity delivery and the potential extent to which healthcare may be supported in this way. More detailed research is required to develop a better understanding of the types of exercise that will work for the different biomechanical subtypes of knee and hip OA and investigation is needed on access and environmental issues for physical activity programmes for older people, from both a provider and a participant perspective, the societal costs of the different approaches to the management of OA and longer term trends in outcome measures (costs and effects).