Background Baby boxes provide goods to new parents and a space for infant sleeping. They were first introduced in Finland, and it has been argued that the policy helped reduce infant mortality. We evaluated the impact of the Finnish Maternity Grant (which includes the Finnish Baby Box) on infant mortality rates (IMRs) at the points of introduction (disadvantaged mothers only) in 1938 and universalisation in 1949.

Methods Maternity Grant introduction and universalisation were evaluated as distinct natural experimental events, using interrupted time series analysis. The outcome was IMR per 1000 live births. We analysed national data on all infants born in Finland between 1922 and 1975, estimating step and trend changes in the outcome following the point of intervention. Sensitivity analyses included truncating the pre-intervention period and a double break point model, incorporating terms for both introduction and universalisation.

Results Maternity grant introduction in 1938 was associated with a step-change increase (β=14.59, 95% CI 4.30 to 24.89) in Finnish IMRs. Maternity grant universalisation in 1949 was associated with a step-change decrease (β=-14.35, 95% CI -20.94 to -7.76) in Finnish IMRs. Sensitivity analyses produced corresponding associations.

Conclusions While we observed changes in IMRs associated with Maternity Grant introduction and universalisation, these changes cannot be disentangled from the impact of the Second World War or other relevant policy developments on infant mortality. Consequently, the relationship between the Finnish Baby Box or comparable contemporary interventions and infant mortality remains unclear.