Willingness to pay for emergency referral transport in a developing setting: a geographically randomized study.

OBJECTIVES: The objective was to identify the correlates of willingness to pay for ambulance transports from a rural city to a regional hospital in Guatemala.
METHODS: An innovative methodology that utilizes a novel randomization technique and satellite imagery was used to select a sample of homes in Santiago Atitlán, Guatemala. The respondents were surveyed at these homes about their willingness to pay for ambulance transport to a regional hospital. A price ladder was used to elicit respondents' willingness to pay for ambulance transport, depending on the level of severity of three types of emergencies: life-threatening emergencies, disability-causing emergencies, and simple emergencies. Simple and multiple linear regression modeling was used to identify the social and economic correlates of respondents' willingness to pay for ambulance transport and to predict demand for ambulance transport at a variety of price levels. Beta coefficients (β) expressed as percentages with 95% confidence intervals (CIs) were estimated.
RESULTS: The authors surveyed 134 respondents (response rate=3.3%). In the multivariable regression models, three variables correlated with willingness to pay: household income, location of residence (rural district vs. urban district), and respondents' education levels. Correlates for ambulance transport in life-threatening emergencies included greater household daily income (β=1.32%, 95% CI=0.63% to 2.56%), rural location of residence (β=-37.3%, 95% CI=-51.1% to -137.5%), and higher educational levels (β=4.41%, 95% CI=1.00% to 6.36%). Correlates of willingness to pay in disability-causing emergencies included greater household daily income (β=1.59%, 95% CI=0.81% to 3.19%) and rural location of residence (β=-19.4%, 95% CI=-35.7% to -89.4%). Correlates of willingness to pay in simple emergencies included rural location of residence (β=59.4%, 95% CI=37.9% to 133.7%) and higher educational levels (β=7.96%, 95% CI=1.96% to 11.8%). At all price levels, more individuals were willing to pay for transport for a life-threatening emergency than a disability-causing emergency. Respondents' willingness to pay was more responsive to price changes for transport during disability-causing emergencies than for transport during life-threatening emergencies.
CONCLUSIONS: The primary correlates of willingness to pay for ambulance transport in Santiago Atitlán, Guatemala, are household income, location of residence (rural district vs. urban district), and respondents' education levels. Furthermore, severity of emergency significantly appears to influence how much individuals are willing to pay for ambulance transport. Willingness-to-pay information may help public health planners in resource-poor settings develop price scales for health services and achieve economically efficient allocations of subsidies for referral ambulance transport.