A controlled before-after study to evaluate the effect of a clinician led policy to reduce knee arthroscopy in NSW

BACKGROUND: Clinical evidence shows knee arthroscopy has little benefit for degenerative conditions and considerable variation in the incidence of knee arthroscopy in Australia has been identified. This study aimed to evaluate a clinician-led evidence-based policy which was implemented in one local health district in New South Wales (NSW) in 2012 to reduce the use of knee arthroscopy for patients aged 50 years or over. METHODS: Trends in rates and volume of knee arthroscopy for patients 50 years or over in NSW between 2004 and 2015 by district were examined. Changes at four hospitals that adopted the policy were assessed by a quasi-experimental before and after study design with control groups, using the generalised estimating equations (GEE) Poisson model. Each case hospital was matched with four control hospitals in terms of the volume of knee arthroscopy surgeries performed in the five years prior to the intervention. RESULTS: Between 2004 and 2015, the number of knee arthroscopies in NSW initially increased and then decreased after 2011, with considerable variation across districts. While an overall reducing trend in NSW was observed between 2011 and 2015 (39%), a 58% reduction (95% CI: 55–62%) was found in the intervention district, including the private sector, being the greatest reduction found in all districts. The GEE Poisson results show that, compared with control hospitals, the number of knee arthroscopy was significantly reduced by 56% (95% CI: 11%–79%) at four hospitals that adopted the policy during the follow-up period (p = 0.02). CONCLUSIONS: Clinicians in one local health district initiated a policy to restrict knee arthroscopy for patients aged 50 years or over, which may explain the greater reduction seen in that district compared to all others, despite an overall decrease noted in the state. A significant reduction found at intervened hospitals proved the effect of the policy, suggesting that the implementation of a simple clinical governance process may help reduce inappropriate surgery. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12891-018-2043-5) contains supplementary material, which is available to authorized users.