Reference pricing and health outcomes: Evaluating the impact of generic substitution in New Zealand

Abstract

Background A fundamental predicament faced by healthcare policy makers is balancing access to the most effective and oftentimes expensive medicines whilst containing total pharmaceutical expenditure; meeting the health needs of the individual, yet not at the expense of society. New Zealand’s Pharmaceutical Management Agency (PHARMAC) has consistently kept expenditure on community pharmaceuticals within budget, employing pharmacoeconomic technology assessments and a mix of pricing strategies to achieve this. However, the advent of complex, biologically‐derived, expensive pharmaceuticals has commenced and the population is increasingly weighted with an elderly, medicine‐consuming sector. These factors will challenge PHARMAC’s ability to contain pharmaceutical expenditure. As patented medicines come to the end of their exclusivity period and generic equivalents become available, opportunity exists for considerable savings to be made by purchasing generics or using the price of the generics as leverage. This ‘generic reference pricing’ is the simplest form of benchmarking strategies used by funders such as PHARMAC; and will be a valuable tool in providing access to expensive biopharmaceuticals in the future. However, it necessitates switching between different brands of medicines, and as evidence of the safety of this practice is limited, resistance to it exists amongst health professionals and patients alike. Aims This thesis tested the hypothesis that patients who switch from an originator brand to a generic equivalent medicine fare no worse than those who do not make a switch. The objectives were: firstly, to measure patient health outcomes consequent to the implementation of generic reference pricing and any associated change in costs; and secondly, to explore the perspective of patients affected by brand switches, particularly their involvement in decision‐making and willingness to pay for choice of medicine brand. Methods A series of natural experiments was conducted to meet the first objective (Part 1), using centralised government health datasets, and selecting patients using lamotrigine, olanzapine, risperidone or venlafaxine as the study medicines. Outcomes measures included use of emergency services, hospitalisations and referrals to specialist services, as well as rates of deaths and reports to the national adverse reactions centre. Data were retrieved from the national datasets for periods before and after each intervention, for both a study group and a comparator group, and difference‐in‐differences comparisons were made. To determine the patients’ perspective (Part 2), a survey was conducted with the assistance of participating pharmacies throughout New Zealand. Findings No change in health services use was found following generic reference pricing, and hence no additional costs for heath service use could be attributed to the policy for any of the four medicines. However, low rates of substitution to the less expensive generic version were observed, and barriers to greater generic uptake were revealed. The role of patients in deciding to switch brands appears to be minimal and their knowledge regarding the objectives of PHARMAC purchasing strategies appears weak. Conclusions Concerns regarding negative health impacts due to brand switching are not supported by the evidence presented in this thesis. Although generic reference pricing as applied in NZ is effective in containing costs, there is more that can be done to increase generic uptake and capitalise on financial gains, such as incentivising pharmacists and educating the public. Additionally, in the evaluation of purchasing policies, this thesis finds that using an epidemiological approach to mine national health datasets is feasible, although there are deficiencies in the data with respect to community‐based services. The findings presented in this thesis augment the international literature on generic medicines and reference pricing policies as well as the use of natural experiments in examining effects of health policies.