OBJECTIVES: The recent surge in the number of targeted cancer treatments being launched has increased concern about rising costs and uncertainties in real-world drug effectiveness. To mitigate these concerns, several Asian countries have implemented patient access schemes (PAS). Few studies assess their impact on national drug policies (NDP). This study reviews PAS used in China, South Korea, Taiwan, and Thailand, to understand the impact on healthcare costs, drug utilization, and clinical outcomes.

METHODS: A targeted literature review of PAS implemented for oncology products over the past 10 years (January 2009 to June 2019) was conducted using publicly available databases, web and grey literature. To validate the findings and further explore key issues, stakeholders responsible for drug access from the four markets were interviewed. Findings regarding the impact on cost, utilization, and outcomes were extracted and synthesized.

RESULTS: Overall, 28 articles were identified, and five stakeholders were interviewed. A range of PAS were identified in the markets, with financial schemes more commonly used than outcomes based schemes. As budget impact is the primary concern of payers, impact evaluation mostly focused on cost and utilization. Across the four markets, PAS increased drug utilization by up to 52% with peaks mostly observed during the first 2 years of implementation. PAS reduced drug prices by up to 69% and generated cost savings by up to US$165 million. However, there was little evidence suggesting that PAS improved health outcomes. Interviewees agreed that measuring PAS impact against long-term health system goals is challenging.

CONCLUSIONS: PAS are an important policy option to enable patient access while mitigating uncertainties in cost-effectiveness and budget impact. While early experiences in Asia show that PAS can be effective in achieving intermediate NDP objectives on drug utilization and cost-savings, long-term impact on containing healthcare costs and improving clinical outcomes remains uncertain.