PUBLIC TRADE-OFFS IN RESPIRATORY PANDEMIC RESPONSE POLICIES IN NEW ZEALAND: FINDINGS FROM A NATIONAL DISCRETE CHOICE EXPERIMENT
Author(s)
Hui Yee Yeo, PhD, MSc (Health Economics), MPharm (Hons), RPh1, Nikki Turner, MBChB, MD, FRNZCGP2, Tony Dowell, MBChB3, Samantha Marsh, PhD, MPH, BCom/BSc2, Isabella Hills, BHSc2.
1Immunisation Advisory Centre, Gen. Practices & Primary Hlthcare, and Liggins Institute, University of Auckland, Auckland, New Zealand, 2Immunisation Advisory Centre and Gen. Practices & Primary Hlthcare, University of Auckland, Auckland, New Zealand, 3Primary Health Care, University of Otago, Wellington, New Zealand.
1Immunisation Advisory Centre, Gen. Practices & Primary Hlthcare, and Liggins Institute, University of Auckland, Auckland, New Zealand, 2Immunisation Advisory Centre and Gen. Practices & Primary Hlthcare, University of Auckland, Auckland, New Zealand, 3Primary Health Care, University of Otago, Wellington, New Zealand.
OBJECTIVES: Future respiratory pandemics will require governments to balance infection control measures against their broader economic and social consequences. This study investigates public preferences for pandemic response strategies in New Zealand and explores how people trade off health, social, and economic outcomes.
METHODS: A Discrete Choice Experiment (DCE) was developed using a D-efficient experimental design (D-error=0.016). The survey includes 48 choice tasks divided into eight blocks and will be administered in two nationally representative cross-sectional waves: winter 2026 (N=500) and summer 2027 (N=500). Attribute development was informed by evidence synthesis, community focus groups, policymaker consultation, and expert consensus using the nominal group technique. The final design includes mitigation measures (mask requirements, school closures, business and essential service restrictions, and gathering limits), policy duration, and projected outcomes including mortality, long-term health effects, and household income reduction. Implausible combinations were excluded using constrained modified Fedorov algorithms. A pilot survey (N=50) conducted in May 2026 generated priors for optimisation of the final design.
RESULTS: Pilot conditional logit analyses showed the model was statistically significant (LR χ²=56.79, p<0.001) with acceptable explanatory power (pseudo R²=0.137), supporting the face validity of the pilot DCE. Most coefficients were in the expected negative direction, indicating respondents generally preferred fewer restrictions, fewer deaths, fewer long-term impacts, and lower household income loss. Stronger business closures (β=-0.60, p=0.006), longer policy duration (β=-0.47, p=0.044), higher mortality (β=-0.083 per 1,000 deaths, p=0.001), greater long-term impacts (β=-0.039 per 10,000 affected, p=0.005), and larger income reductions (β=-0.029 per 1% decrease, p<0.001) were associated with lower preference. Mask requirements, gathering limits, and school closures showed weaker or non-significant effects, suggesting differing perceived importance across measures.
CONCLUSIONS: Results from the winter wave are anticipated before conference presentation. This study will provide evidence on socially acceptable pandemic response strategies to support future preparedness planning and risk communication in New Zealand.
METHODS: A Discrete Choice Experiment (DCE) was developed using a D-efficient experimental design (D-error=0.016). The survey includes 48 choice tasks divided into eight blocks and will be administered in two nationally representative cross-sectional waves: winter 2026 (N=500) and summer 2027 (N=500). Attribute development was informed by evidence synthesis, community focus groups, policymaker consultation, and expert consensus using the nominal group technique. The final design includes mitigation measures (mask requirements, school closures, business and essential service restrictions, and gathering limits), policy duration, and projected outcomes including mortality, long-term health effects, and household income reduction. Implausible combinations were excluded using constrained modified Fedorov algorithms. A pilot survey (N=50) conducted in May 2026 generated priors for optimisation of the final design.
RESULTS: Pilot conditional logit analyses showed the model was statistically significant (LR χ²=56.79, p<0.001) with acceptable explanatory power (pseudo R²=0.137), supporting the face validity of the pilot DCE. Most coefficients were in the expected negative direction, indicating respondents generally preferred fewer restrictions, fewer deaths, fewer long-term impacts, and lower household income loss. Stronger business closures (β=-0.60, p=0.006), longer policy duration (β=-0.47, p=0.044), higher mortality (β=-0.083 per 1,000 deaths, p=0.001), greater long-term impacts (β=-0.039 per 10,000 affected, p=0.005), and larger income reductions (β=-0.029 per 1% decrease, p<0.001) were associated with lower preference. Mask requirements, gathering limits, and school closures showed weaker or non-significant effects, suggesting differing perceived importance across measures.
CONCLUSIONS: Results from the winter wave are anticipated before conference presentation. This study will provide evidence on socially acceptable pandemic response strategies to support future preparedness planning and risk communication in New Zealand.
Conference/Value in Health Info
2026-09, ISPOR Asia Pacific 2026, Bangkok, Thailand
Value in Health, Volume 55, Issue S1
Code
PCR28
Topic
Patient-Centered Research
Disease
SDC: Infectious Disease (non-vaccine)