A measure to assess trust in public health in a disaster

Background: Trust in public health has rarely been defined and measured empirically in population surveys. We developed and tested a scale to measure public health disaster-related trust.

Methods: We postulated that public-health disaster-related trust has 4 major domains: competency, honesty, fiduciary responsibility, and confidentiality. We developed 4 items scored on a 4-point Likert scale that inquired about each of these domains. Items were translated into 5 languages and fielded as part the 2004 Public Health Response to Emergent Threats Survey (PHRETS), a random-digit-dialed telephone survey of the Los Angeles County population.

Results: Scale scores ranged from 4 to 16 out of a possible 4 to 16. Scores were normally distributed and the mean was 8.5 (SD 2.7). Cronbach's alpha = 0.79. Principal components analysis disclosed one component with an Eigenvalue of 2.45 explaining 61% of total variance. Construct validity was assessed testing the hypotheses that lower trust is associated with lower likelihood of following public health recommendations in an event and that lower trust is associated with lower likelihood of household preparedness. As hypothesized, trust scores were directly related to willingness to follow public health recommendations in an event and to household preparedness (all p<0.001). These associations remained in language and ethnic sub-group analyses.

Conclusions: The Public Health Disaster Trust scale will facilitate studies of the prevalence, causes, and effects of distrust in public health disaster policies and recommendations.