Our study demonstrated significant changes in the rates of PPI dispensing, as a surrogate measure of GP prescribing, following two targeted interventions, using national, whole-of-population dispensing data. We found that rates of standard strength PPI dispensing declined following the interventions, while dispensing rates for low strength PPIs increased after the 2009 intervention but not the 2015 intervention. These findings suggest that the NPS MedicineWise programs were effective in improving quality PPI prescribing by GPs in Australia.
Our results are consistent with recent studies of Australian veterans, which found that interventions conducted in 2004, 2006 and 2009 by NPS MedicineWise and by the Veterans’ MATES program in 2006 and 2012 resulted in a 20.9% relative decrease in overall PPI dispensing and a 42.2% relative increase in low strength PPI dispensing 12 months after the final intervention in the veteran population [29]. Medicines dispensed to Australian veterans are subsidised through the Repatriation Pharmaceutical Benefits Scheme (RPBS), a funding body distinct from the PBS. Our data did not include RPBS dispensing records and our findings suggest that GPs may have applied new knowledge resulting from these programs in treating both their veteran and other community patients.
We estimated that the 2015 NPS MedicineWise program was associated with a 3.0% decrease in the standard strength PPI dispensing rate in the first 15 months following the intervention. This was greater than the 1.6% decrease during the 15 months after the 2009 intervention and the larger drop may be related to the Choosing Wisely Australia PPI campaign launched during the same month as the NPS MedicineWise 2015 intervention. The Choosing Wisely campaign was aimed at beginning a conversation between GPs and patients around the long-term use of PPIs; with Choosing Wisely members recommending regular attempts at lower strength prescribing or cessation of PPI therapy in patients with uncomplicated disease [17]. Although the impacts of the Choosing Wisely campaign and the 2015 NPS MedicineWise program could not be separated, it is possible that each reinforced the messages of the other, resulting in a further reduction in the observed use of standard strength PPIs.
Other studies have shown the impact of educational programs on GP prescribing, including a study conducted by May et al. Doctors participating in an educational visiting program in Adelaide that focused on better use of prescribed non-steroidal anti-inflammatory drugs (NSAID) reduced their use of NSAIDs by 9% and 28% for two different measures compared to a comparison group [30]. Other studies have found positive impacts of educational programs on GP behavior, including programs on the treatment and management of incontinence, health behaviours of elderly people and adolescent health care [31-33].
Our study highlights the benefits of engaging with practitioners to improve the quality use of medicines. The goal of Choosing Wisely is to start discussions between physicians and patients around specific therapeutic practices. NPS MedicineWise actively engages general practitioners and consumers in educational activities aimed to improve the quality use of medicines through behaviour change [34]. There are potential opportunities for additional quality use of medicines interventions, as well as on-going PPI education.
Strengths and limitations
Our study evaluated changes to dispensing of PPIs prescribed by GPs to concessional patients over time. The cost of all PPI medicines subsidised by the PBS is above the concessional co-payment threshold and, therefore, we have a complete ascertainment of PPI dispensing for these concessional patients. The DHS began collecting dispensing data for medicines where cost was below the co-payment threshold for general non-concessional patients from April 2012 and, based on dispensing data from that time, it has been estimated that approximately 70% of PPIs prescribed by GPs were dispensed to concessional patients. Nearly 88% of prescription medicines dispensed under the PBS are for concessional patients. Concessional patients include low-income earners, welfare recipients, and Health Care Card holders and are higher users of health services due to their generally poorer health status. However, we expect the intervention impacts we observed in concessional beneficiaries would be similar among general beneficiaries as we do not expect the treatment of GORD with PPIs to differ between the two subsidy groups.
Most PPIs are available over-the-counter (OTC) in Australia and PPIs that are available from pharmacies without a prescription include rabeprazole, pantoprazole, esomeprazole and omeprazole. Given the limitation of our data we are not able to determine if patients switched from prescription to OTC PPIs in the study time period. However, during the time period of the study, OTC formulations contained just a 7 day supply and cost was above that of the concessional beneficiary co-payment. We believe it is unlikely that many concessional beneficiaries would have switched to an OTC formulation although we do not know if this is the case. The PBS has a safety net which resets each year on 1 January. In 2019, the PBS Safety Net threshold was $390 for concessional card holders. Before meeting the threshold each medicine costs concessional patients $6.50 and once they reach the threshold all PBS medicines are free of charge. This is an incentive for patients to pay the concessional rates for PBS medicines rather than buying those available OTC from a pharmacist as not only are the medicines cheaper but the cost contributes to the patient reaching the Safety Net threshold. Non-concessional patients also have a safety net but due to the low cost of PPIs they do not have the cost covered under the PBS and are more likely to purchase PPIs over the counter than concessional patients. If a doctor will no longer prescribe a PPI to a patient who would still like to take them, they can purchase them OTC. However the data provided for this study does not include any information on OTC sales.
Although this study is observational and causality cannot be confirmed, given the timing of the intervention programs, and that we did not identify any other potential confounding events following the 2009 and the 2015 programs that might explain the changes we observed, we believe that the change in the rate of PPI dispensing is attributable to the NPS MedicineWise programs and the Choosing Wisely RACGP recommendation.
PBS data are maintained for the purpose of providing reimbursement to patients and pharmacies, and clinical information such as diagnoses and treatment indications are not captured. Although the programs we evaluated were aimed at improving the quality use of PPIs, we are unable to assess the appropriateness of prescribing using these data. Similarly, our prescriber-level PBS data do not allow us to evaluate the rates at which individual patients switched between PPI treatment strengths or ceased PPI therapy or to determine the duration of treatment for individuals.
Another limitation of the study is that there is no information about patient adherence to medicines dispensed.
The strengths of our study is the use of a longitudinal and complete dataset comprised of dispensing records for PPI medicines prescribed to concessional patients by every general practitioner in Australia from 2006 to June 2016. These data allowed for robust estimates of trends over time in the exact population targeted by the interventions.