Evaluating Perception of Child Health and Safety Needs in a Community Through the Perspective of Caregivers and Physicians

Abstract

Objective: To identify and compare child health and safety concerns in the community from the perspective of physicians and caregivers in Louisville. Design: We administered surveys to physicians and caregivers who were asked to list, via open-ended questions, the top three health or safety concerns they had for children in three separate age groups (≤ 3, 4-11, and ≥12 years old). Four reviewers assigned responses to a pre-specified category/subcategory through a schematic designed by the study team and a fifth reviewer addressed any discrepancies and assigned final categories. Agreement in assigning overall categories was measured using Fleiss Kappa and descriptive statistics were used to characterize the responses. Setting: Surveys were administered to caregivers who attended community events in Park Duvalle, a lower socioeconomic neighborhood in Louisville whose residents are predominantly black. Surveys were simultaneously administered to physicians throughout the University of Louisville/Norton Healthcare system. Participants: A total of 41 caregivers with at least one child and 43 physicians were surveyed in this study. Results: Overall categorical agreement between reviewers was k = 0.839, indicating good agreement on assignment. Injury/violence was identified by both caregivers and physicians as the top concern for children across all age groups. Wide variation was found when analyzing specific injury/violence subcategories, with caregivers more worried about household safety (17-40%) and gangs/unsafe neighborhoods (10-50%) while physicians were more concerned about sleep safety (36%), domestic violence (8-27%), and unintentional injury (10-20%). Conclusions: Survey administration was a feasible method of comparing health and safety concerns among caregivers and physicians. We identified injury/violence as the top health and safety concern across all age groups in our community but noted variation on the specific type of injury/violence. Replication of our study methodology in other neighborhoods may allow for improved comparison of caregiver and physician concerns.