Does the removal of financial incentives from Australian general practices affect immunisation rates?

Abstract

Background Financial incentives are commonly used to drive improvements in health care delivery. There is a large body of evidence around the impact of the delivery of financial incentives on health outcomes, quality and prevention. Results from a systematic review of published reviews found that randomised controlled trials and time series studies undertaken between 1980 and 2000 have contributed most of the evidence. Most of the studies looked at associations between the introduction of incentives and improvements in quality. However, poor study design and unaddressed methodological errors have significantly limited the drawing of sound conclusions from the evidence to date. There is a paucity of published research on the removal of financial payments to general practitioners (GPs) and its effect on preventive care.

Aim This thesis aims to: (i) determine if the removal of general practitioner financial incentive payments for immunisation are associated with a reduction in paediatric immunisation coverage; (ii) determine if there is evidence of associations between levels of socio-economic disadvantage and paediatric immunisation coverage after removal of financial incentive payments to general practitioners, and; (iii) determine if there is evidence of an association between levels of relative remoteness and paediatric immunisation coverage after removal of financial incentive payments to general practitioners.

Method Childhood immunisation has been linked to the delivery of two financial incentives provided to Australian general practitioners for the reporting of paediatric immunisation coverage to the Australian Childhood Immunisation Register (ACIR). The General Practice Immunisation Service Incentive Payments (SIP) and Outcomes Bonus Payments (OBP) were removed incrementally in 2008 and 2013. A systematic review of reviews was undertaken on current evidence of the effect of incentives on immunisation. Paediatric immunisation coverage data for all Australian children under the age of seven was used to examine the impact of the removal of these incentives on immunisation rates. Interrupted segmented regression times-series analysis (ITSA) was performed on postcode-level immunisation coverage data across 64 quarters between 2000 and 2016. Sub-group analysis was performed on the data stratified for socio-economic disadvantage and rurality. The data were transformed using an autoregressive integrated moving average (ARIMA) model to address autocorrelation.

Results The results revealed an immediate and statistically significant drop in immunisation coverage following removal of the second of two incentive payments to GPs, the OBP. This drop persisted even after the data was modelled to control for heteroscedasticity and the inflationary effects of the ‘No Jab No Pay’ policy. The effect was still evident after the series had been transformed with weighting coefficients to address the issue of small cells. The analysis also showed an additional 19,441 children would have been reported as being fully immunised between January 2000 and 1 January 2016 had the SIP and OBP incentives not been removed. A reduction in immunisation coverage was evident across all socio-economic quintiles after the national series had been stratified across areas of disadvantage. This effect was found to be larger in areas of greater disadvantage. The same outcome was observed across all geographic regions of Australia, where major city, regional and remote areas showed a statistically significant reduction in immunisation following removal of the OBP. The postcode areas which showed the greatest decrease in immunisation were those outer regional to very remote regions of Australia. There was no statistically significant change in coverage after removal of the SIP.

Conclusion This thesis establishes evidence of a strong relationship between the removal of incentives to GPs and the provision of preventive services. Not only does this analysis provide estimates which are statistically significant, it also provides multiple analyses which test the validity of the outcomes. To the author’s knowledge, the innovative methods which have been used in this thesis to measure the effects of the removal of financial incentives to GPs, has not been previously applied to the research around incentive effectiveness in either Australia or overseas. This places this research in a unique position to both contribute evidence and influence future policy.