Using Canadian administrative health data to examine the health of caregivers of children with and without health problems: A demonstration of feasibility.

INTRODUCTION: Caregivers of children with health problems experience poorer health than the caregivers of healthy children. To date, population-based studies on this issue have primarily used survey data. OBJECTIVES: We demonstrate that administrative health data may be used to study these issues, and explore how non-categorical indicators of child health in administrative data can enable population-level study of caregiver health. METHODS: Dyads from Population Data British Columbia (BC) databases, encompassing nearly all mothers in BC with children aged 6-10 years in 2006, were grouped using a non-categorical definition based on diagnoses and service use. Regression models examined whether four maternal health outcomes varied according to indicators of child health. RESULTS: 162,847 mother-child dyads were grouped according to the following indicators: Child High Service Use (18%) vs. Not (82%), Diagnosis of Major and/or Chronic Condition (12%) vs. Not (88%), and Both High Service Use and Diagnosis (5%) vs. Neither (75%). For all maternal health and service use outcomes (number of physician visits, chronic condition, mood or anxiety disorder, hospitalization), differences were demonstrated by child health indicators. CONCLUSIONS: Mothers of children with health problems had poorer health themselves, as indicated by administrative data groupings. This work not only demonstrates the research potential of using routinely collected health administrative data to study caregiver and child health, but also the importance of addressing maternal health when treating children with health problems.