PW 1811 Evaluating the period of purple crying program: problems of no statistical power

Background Shaken baby syndrome (SBS) is a devastating form of infant abuse, with significant impact on families and society. The Period of PURPLE Crying program in British Columbia (BC), a program to prevent infant abuse, offers crying education and materials to all parents of newborns through maternity and public health services and through an annual public education campaign. PURPLE was evaluated for its effectiveness to change knowledge and attitudes in a randomized control trial. Objectives The PURPLE program monitors a number of metrics, however, since the incidence of SBS is very low, achieving statistical power for the purposes of evaluating the effect of PURPLE is not attainable in a reasonable time frame. The presentation will present methods used to evaluate the PURPLE program in BC; discuss issues concerning program delivery fidelity; and provide practical solutions to overcome challenges. Methods The incidence of SBS among 0 to 24 month olds was identified through retrospective and prospective surveillance. Program delivery fidelity was monitored through regularly scheduled surveys of maternity and public health nurses, and parents. Changes in SBS hospitalization rates for 0–24 month olds pre- to post-implementation were measured. Findings A 35% reduction in SBS hospitalization rates was found following the implementation of the program. Survey results showed varying program fidelity by region which resulted in the development and implementation of action plans with key delivery stakeholders. Fidelity improved significantly as a result of this action plan. Conclusions Despite a low incidence of SBS, requiring decades of data collection to power an evaluation of the PURPLE program, there are metrics that can be applied to measure program process, impacts and fidelity.