Date Published: April 11, 2019
Publisher: Public Library of Science
Author(s): Sarah K. Orr, Karen Tu, Sarah Carsley, Hannah Chung, Laura Holder, Shirin Jabbari, Daniel W. Harrington, Heather Manson, José Guilherme Behrensdorf Derraik.
The Healthy Kids Community Challenge is a large-scale, centrally-coordinated, community-based intervention in Ontario, Canada that promotes healthy behaviours towards improving healthy weights among children. With the goal of exploring tools available to evaluators, we leveraged electronic medical records from primary care physicians to assess child weights prior to launch of the Healthy Kids Community Challenge. This study compares the baseline (i.e. pre-intervention) prevalence of overweight and obesity in children 1–12 years of age living within and outside Healthy Kids Community Challenge communities.
Cross-sectional analysis of a primary care patient cohort.
Electronic Medical Record Administrative data Linked Database (EMRALD) in Ontario, Canada.
A cohort of 19 920 Ontario children who are rostered to an EMRALD physician. Children were 1–12 years of age at a primary care visit with recorded measured height and weight, between January 1, 2014 and December 31, 2015.
Overweight and obesity as determined by age- and sex-standardized body mass index using World Health Organization’s Growth Standards.
In Healthy Kids Community Challenge communities, 25.6% (95% CI 24.6–26.6%) of children had zBMI above normal (i.e. >1) compared to 26.7% (95% CI 25.9–27.5%) for children living outside of Healthy Kids Community Challenge communities.
Despite some differences in sociodemographic characteristics, zBMI of children aged 1–12 years were similar inside and outside of Healthy Kids Community Challenge community boundaries prior to program launch.
Public health policy-makers and practitioners are increasingly expected to demonstrate the impacts of their programs on population health outcomes. However, assessing the impacts of large-scale programs intended to improve the health of the population can be challenging.[2,3] The publically-funded nature of such programs means that they are often working within constrained resources, with tension between funding the program and its evaluation. Evaluators must remain nimble and seek timely and feasible solutions to maintain rigour in order to meet the public health policy-makers’ and practitioners’ needs. With the goal of exploring tools available to evaluators, we used electronic medical records from primary care physicians to assess child weights prior to the start of the Healthy Kids Community Challenge (HKCC).
Table 1 shows the baseline characteristics of children in the cohort by community type, specifically HKCC (n = 7 382) and non-HKCC (n = 12 538). Overall, baseline characteristics show that these are two relatively comparable populations. However, there were a few notable differences between children living in HKCC communities versus outside of HKCC communities. First, there was a higher proportion of children in the youngest age group living in HKCC communities (36.0% 1–3 y vs. 32.9% in non-HKCC communities); in contrast to a higher proportion of children in the oldest age groups in non-HKCC communities (22.7% 9–12 y vs. 18.5% in HKCC communities). Second, HKCC communities in the cohort had a higher proportion of children living in rural communities (19.3% vs 15.8% in non-HKCC communities). Third, there was a higher proportion of children in the lowest neighbourhood income quintile living in HKCC communities (18.5% vs. 10.1% in non-HKCC communities). Fourth, HKCC communities in the cohort had a higher proportion of children living in immigrant households (14.8% vs. 11.1% in non-HKCC communities). Finally, there was a lower proportion of children in HKCC communities with asthma (8.8% vs. 10.1% in non-HKCC communities) or a mental health condition (17.9% vs. 20.6% in non-HKCC communities), however, there was a higher proportion of children with a complex chronic disease or congenital disorder (7.2% vs. 6.4%).
This study demonstrates the utility of electronic medical records for the baseline assessment of a large-scale provincial intervention when it is not feasible to collect primary data, and existing surveys do not adequately cover the outcome (directly measured BMI), population (children aged 1–12 years), or level of exposure (community-based) over the time period of the program. Our study shows that baseline BMI z-scores were similar between children living within and outside of HKCC intervention communities, despite some differences in sociodemographic characteristics. Further, the prevalence of overweight and obesity in the cohort overall was similar to national estimates from surveys using objective measures, which is 27.0% for 3–19 year olds. Together, these findings suggest that prior to the program launch, the weight status of children living in HKCC intervention communities was comparable to Ontario children living outside of HKCC communities.