Table 6

Key challenges to address in future analyses of the LongSHOT cohort exploring the relationship between handgun ownership and mortality

Mismeasurement of exposureNon-handgun owners may in fact be owners due to, for example, failure to match their purchases in probabilistic linkage, unlawful handgun acquisition and purchases made prior to 1985. Non-handgun owners may own unobserved long guns.
Unobserved confoundingRelevant differences may exist between handgun owners and non-owners that are not measured in the linked data (eg, incidence of mental illness, risk-taking propensity).
Restriction of cohort to voter file registrants in CaliforniaGeneralisations to non-registrants in California and to residents of other states may be impaired by relevant unobserved heterogeneity.
  • LongSHOT, Longitudinal Study of Handgun Ownership and Transfer.