Section
7
Measuring Body Composition in Population Health Research: Case Studies
Case Study 5 Assessing, Analyzing, and Presenting Health Data from Electronic Health RecordsĀ (EHRs)
Background
A pediatric primary care system (PCS) has recently decided to address childhood obesity as a population health issue among the child and adolescent patients that it serves. Health care providers are increasingly using electronic health records (EHRs) and related electronic infrastructure to prevent and treat chronic diseases. As a result, PCSs, Federally Qualified Health Centers, hospitals, and health systems are in a unique position to analyze relevant patient panel EHR data on BMI, obesity-related comorbidities, the response to lifestyle or medical interventions, and trends over time. The PCS leadership convenes to discuss this approach with stakeholder representation from the information technology office, provider representatives, and community liaisons. One of the goals is to assess changes in adiposity over time in their patient population.
Considerations
The PCS and its partners consider how they can analyze, assess, and present data, including information from the EHR, in a way that informs population health. Providers in the PCS routinely use the EHR for basic functions, such as to calculate BMI, plot the BMI on the appropriate growth charts, display BMI percentile, and use clinical decision supports (CDS) to prompt appropriate screening, prevention, and management decisions. The PCS has successfully developed standardized EHR reports on a number of parameters, including the proportion of children being screened for BMI, the prevalence of overweight and obesity in their patient population over time, the proportion of children with overweight or obesity and a co-occurring condition (e.g., asthma), and the proportion of children with overweight or obesity referred to weight management programs. As part of these efforts, the PCS was assisted by informatics and clinical experts to identify the most appropriate protocols for data cleaning and defining which data points will be used in an analysis.
PCS leadership and stakeholders decide to develop relationships with local weight management programs and other community resources to better track and understand how members of the PCS patient panel access and use community resources. This also will inform how these community resources may improve the health of patients. The PCS leadership and stakeholders identify data use agreements with these groups as a critical step. In doing so, they identify key measures of interest, as well as potential data cleaning and analysis protocols for these new data as the PCS group works with community resources to inform information technology infrastructure and final selection of measures.
Method Selection
The PCS leadership and stakeholders identify two change-sensitive weight-related measures from the EHR as candidates for outcome measures. These measures include an age- and sex-specific measure of BMI (see Section 5). Weight, height, and calculated BMI may be collected from community partners as well as PCS providers. They will highlight changes in the prevalence of overweight and obesity over the course of months to a few years as one important outcome in their patient population.
Several other outcome measures in children with overweight or obesity are selected as well, including systolic and diastolic blood pressures, hemoglobin A1C, and liver function tests. Data cleaning protocols specific to each of these measures will need to be used or developed as well. For children aged 2ā19 years, the Daymont Method (see Automated Identification of Implausible Values in Growth Data from Pediatric Electronic Health Records in the Resources) can be utilized for cleaning BMI.116 The group will analyze and describe the relationship between changes in weight-related measures and improvements in the co-occurring conditions.