Cochran, J. & Baus, A.(2015). Developing Interventions for Overweight and Obese Children using Electronic Health Records Data. Online Journal of Nursing Informatics (OJNI),19 (1).
Childhood obesity and the resulting co-morbid conditions have become a massive burden for primary caregivers in West Virginia and across the United States. Building culturally sensitive clinical interventions to meet population health needs presents a multi-factorial challenge. Historically it has been difficult to accurately assess the demographic characteristics of clinic populations. Fortunately, the use of electronic health records (EHR) has created a major shift in clinical documentation and the ability to routinely collect essential demographic and clinical data. While data for Meaningful Use under the Centers for Medicare and Medicaid Services is increasingly used to evaluate clinical care and outcomes, use of EHR data outside of the Meaningful Use umbrella has not received sufficient attention.
This study was designed to explore the use of EHR data beyond Meaningful Use to obtain demographic characteristics of an obese and overweight pediatric population in a rural primary care center for the purpose of informing appropriate, locally relevant intervention strategies. Findings show that the breadth and depth of information recorded on each patient can collectively provide valuable information to describe and evaluate the clinic population, identify priority areas to address, and measure change over time. Application of EHR data to understand the demographic characteristics of this particular patient population highlights the ability to identify target patient populations, uncover critical patient-level and population-level outcomes, inform intervention development and implementation, and add value to efforts in quality improvement systems transformation. Use of EHR data outside of the Meaningful Use umbrella needs increased attention in primary care.
Childhood obesity continues to be a health concern for children and adolescents. Historically, the rate of obesity has accelerated in children and adolescents from 13.9% in 1999 (de Ferranti et al., 2004) to 17.1% during 2003-2004 (Ogden, Carroll, & Flegal, 2008). The most recent reporting period of 2011-2012 shows no significant change in the prevalence of obesity among youth (Centers for Disease Control and Prevention, 2014; Ogden, Carroll, Kit, & Flegal, 2014).
Childhood obesity is especially problematic in rural areas. The southern states of West Virginia, Kentucky, Tennessee, North Carolina, Texas, South Carolina, Mississippi, and Louisiana have the highest rates of obesity (Liu et al., 2007). Respectively, more rural than urban children are overweight/obese (35.4% v. 29.3%) and obese (18.6% v. 15.1%) (Liu et al., 2012). The majority of these states are in the Appalachian region (Appalachian Regional Commission, 2007).
Previous studies demonstrate that parents often inaccurately perceive their child's weight status(Cochran, Neal, Cottrell, & Ice, 2012). The discrepancy in agreement of parental perception of their child's weight status could be related to the influence of Appalachian culture and beliefs. In Appalachia, children that are heavier may be perceived to be stronger and healthier. As the child ages, parents may perceive that the size of the childdemonstrates sustenance and hardiness (Cochran, 2006). Thin or even normal weight children are often seen as disadvantaged with less ability to withstand illness and other adverse situations. Some parents believe that having a heavier child demonstrates adequate nurturing skills (Cochran, 2006; Cochran, 2008; Trombini et al., 2003).
Medicare and Medicaid EHR Incentive programs provide financial incentives for meaningful reporting of various outcome measures, such as childhood obesity (Centers for Medicare & Medicaid Services, 2014). Reporting requires documentation of the child's weight status in the child's medical record as part of obesity screening guidelines (US Preventive Services Task Force, 2010). The weight status is then documented on the visit summary and given to the parent. Documentation and discussion of their child's weight status in the primary care setting could help motivate parents to make healthy changes in their child's diet, secondary to the medical diagnosis of overweight/obesity and suggested interventions. In the past, multiple barriers prevented community interventions for families of obese children from being successful (Sonneville, La Pelle, Taveras, Gillman, & Prosser, 2009; Vannucci & Wilfley, 2012). Interventions that relate to demographic characteristics of the group may be the problem. Previously, demographic analysis of disease specific groups has been by the perception of the provider and not supported by population analysis within the clinical practice. The purpose of this project is to assess the usefulness in EHR data to adequately describe the overweight/obese population of a rural West Virginia primary care center to promote better clinical and community interventions based on characteristics of the patients.
Data from the pediatric population ages 1-17 of a rural West Virginia primary care center were identified and extracted from the clinic's EHR based on the documentation of an overweight or obesity diagnosis as per International Classification of Diseases version 9 (ICD-9) codes using Practice Analytics software. This extraction provided 1,122 overweight or obese pediatric patients. Among this subset of patients, insurance information, gender, age, ethnicity and the first three digits of the ZIP codes were identified. ZIP code data were limited to the first three digits to preserve patient confidentiality. ICD-9 codes for co-morbid conditions hypertension, hyperlipidemia, and dysmetabolic syndrome were also obtained. Frequencies were generated for specific demographic characteristics. ZIP codes were entered into a geographical information system program for mapping from the National Center for the Analysis of Healthcare Data.
Data were analyzed using the Chronic Disease Electronic Management System (CDEMS). CDEMS is Microsoft Access-based public-domain registry software modified by the West Virginia University Office of Health Services Research for use in chronic disease identification, tracking, and identification of patients at-risk for chronic health conditions. This registry has been used successfully in identifying patients undiagnosed with hypertension (Baus, Pollard, & Hendryx, 2012) and in identifying patients at-risk for diabetes (Baus, Wood, Pollard, Summerfield, & White, 2013). Moving the EHR data to an external system allows for data transparency in that key data within the EHR, such as patient demographics, diagnoses, and results, can be queried for coding consistency and completeness.
This study included no identifying patient information, and was deemed non-human subjects research by the West Virginia University Office of Research Integrity and Compliance (protocol number 1406335154).
Gender was distributed fairly equal among the overweight and obese pediatric patients, with 568 males (50.6%) and 554 females (49.4%). Among all patients, 955 (85.1%) reported ethnicity. Of those reporting, only 3 (0.3%) described themselves as Latino/Hispanic. Reporting of racial background was found to be more complete, with only 37 patients (3.3%) without a documented racial category. Among those with race documented, the majority of pediatric patients (91.0%) are Caucasian.
Less than half (42.3%) of the overweight/obese pediatric patients were 12 years of age or older. Among those, 350 patients (31.2%) were between the ages of 12-15, and 125 (11.1%) were between the ages of 16 and 17. Approximately 41% of the pediatric patients with obesity were between ages 6 and 11, while 16.8% of the population was less than 5 years of age. Pediatric patients in the 6 to 11 years of age category (40.9%) comprise the greatest proportion of patients deemed overweight or obese (Table 1).
Only 80 (7.1%) of the overweight and obese pediatric patients had an additional ICD-9 code representing a co-morbid condition (i.e., essential hypertension, dysmetabolic syndrome x, or hyperlipidemia). Of these co-morbid conditions, essential hypertension (4.5%) was the most frequently occurring co-morbid condition among this patient population (Table 2).
Geographical mapping for the incidence of overweight and obese pediatric patients was performed using the first three digits of the ZIP code. The majority of pediatric patients (80.8%) identified for inclusion in this study reside in the eastern portion of the county in which the primary care center is located (Appendix).
Insurance coverage for patient care was obtained and categorized into private insurance, Children's Health Insurance Program, Medicaid, and other. Medicaid (46.1%) and private insurance (43.9%) were the most frequent occurring insurance carriers among these patients. The Children's Health Insurance Program was 8.6%. The remaining types were self-pay, sliding fee, or other (1.4%).
This application of EHR data highlights two broad, practical areas of consideration for primary care and public health. First, while gender and age were well represented, the 40.9% of overweight/obese children ages 6 to 11 may be reflective of pubertal changes and/or the beginning of metabolic syndrome. This is a critical public health finding, and supports the need to intervene early for prevention of obesity and its complications. Further, examining clinic-wide EHR data from this rural, pediatric patient population demonstrates the ability to make fuller use of routinely collected data for targeted patient care. This underscores the need for primary care to look beyond the expectations of Meaningful Use and to the necessity of an active application of clinical data to chronic disease prevention. Secondly, a potential area for improvement in office systems involves poor reporting of ethnicity which could be related to inaccurate intake of information or the assumption on the part of the patient that it was not necessary information. A review of information intake on ethnicity should be revisited, with subsequent analyses measuring any change in improving the collection of this important data.
A high percentage (80.8%) of this pediatric overweight and obese population resides in the same geographic area, in close proximity to the clinic. While using ZIP codes as a proxy for location may be less accurate in rural areas due to consolidation of postal areas, this method does help outline the geographic areas in highest need of targeted intervention. In this case, EHR data allows the clinic team to transform raw data into information, and that information transforms into strategic actions to help curb the increasing obesity prevalence among pediatric patients.
The diagnosis of co-morbid conditions in the overweight/obese populations may not be reflective of the actual conditions. Laboratory testing in certain age groups with obesity requires a diagnosis. Screening for lipids and metabolic syndrome begins at age ten. Most insurance companies will not reimburse for a visit or charge coded under obesity without another diagnosis code. Further studies on patients with diagnoses related to lipids and dysmetabolic syndrome should be performed to help assess confirmation of the diagnosis. Reviewing the data for children who meet criteria for laboratory screening should be performed to determine compliance with guidelines and effectiveness in identifying co-morbid conditions (Shah, Kublaoui, Oden, & White, 2009).
Healthcare coverage among these children was surprising. Medicaid and private insurance were nearly the same. Previous studies show a greater prevalence of obesity among lower socioeconomic populations (Ogden, Carroll, Curtin, Lamb, & Flegal, 2010), which would predict a higher number of overweight and obese children covered by Medicaid. The number of children covered by Medicaid may be reduced due to the implementation of the Children's Health Insurance Program which insures children who may have previously been covered by Medicaid or not covered by any plan.
Collecting secondary data from EHRs presents multiple issues. Obtaining accurate patient information and proper entry into the EHR has been a major undertaking in health care settings. Extracting the data for analysis requires an intricate knowledge of the electronic system as well as clinical expertise and research knowledge. De-identifying the data for analysis requires a secure network and clinical information systems which may not be easily obtained or used by beginning researchers. Very few rural health clinics have the financial means or available personnel to extract, analyze, and interpret the data. Extracting data from the EHR outside of the Meaningful Use program therefore presents challenges. Further study of the intersection of health informatics with rural primary care is needed to more completely characterize patient outcomes in the midst of primary care transformation.
Demographic analysis of the overweight/obese patients in a specific clinic validates the provider's concept of the population and adds to the specificity of location, medical coverage, and age. Program planning can be more specific based on the age and gender of the population. Patients in certain geographic regions of the catchment area can be networked with existing, local programs. Patients with insurance covering weight management programs can be identified and referred to existing programs close to their homes. Most importantly, early identification of co-morbid conditions in this pediatric population may prevent long term complications. Successfully designing and implementing public health programs to address the specific needs of the patient population is supported through thoughtful application of clinical data. Health analytics stands to help improve disease outcomes on population and patient level metrics.
The use of EHR data outside of the Meaningful Use umbrella needs increased attention in primary care. Multiple opportunities exist for primary care centers to utilize this process for population identification and program development. The breadth and depth of information recorded on each patient provides critical insight into intervention development, and affords opportunities to measure changes over time. The application of EHR data to understand the demographic characteristics of this particular patient population is one example indicating the strength of rural primary care to leverage contextual factors such as increased adoption of EHRs and attention to Meaningful Use for the direct benefit of prevention and amelioration of priority health disparities such as pediatric obesity.
Appalachian Regional Commission. (2007). Maps. Retrieved from http://www.arc.gov/maps
Baus, A., Hendryx, M., & Pollard, C. (2012). Identifying patients with hypertension: a case for auditing electronic health record data. Perspectives in Health Information Management, Spring, 1-15.
Baus, A., Wood, G., Pollard, C., Summerfield, B., & White, E. (2013). Registry-based diabetes risk detection schema for the systematic identification of patients at risk for diabetes in West Virginia primary care centers. Perspectives in Health Information Management, Fall, 1-10.
Centers for Disease Control and Prevention. (2014). Obesity and overweight for professionals: childhood obesity facts. Retrieved from http://www.cdc.gov/obesity/data/childhood.html
Centers for Medicare & Medicaid Services. (2014). 2014 definition stage 1 of meaningful use. Retrieved from http://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Meaningful_Use.html
Cochran, J. (2006). Bringin' it home: Appalachian culture. Paper presented at In Rural Culture: West Virginia's Legacy, Morgantown, WV.
Cochran, J. (2008). Empowerment in adolescent obesity: state of the science. Online Journal of Rural Nursing & Health Care, 8(1), 63-73.
Cochran, J., Neal, W., Cottrell, L., & Ice, C. (2012). Parental perception of their child's weight status and associated demographic factors. Online Journal of Rural Nursing and Health Care, 12(2), 11-29.
de Ferranti, S. D., Gauvreau, K., Ludwig, D. S., Neufeld, E. J., Newburger, J. W., & Rifai, N. (2004). Prevalence of the metabolic syndrome in American adolescents: findings from the third national health and nutrition examination survey. Circulation, 110(16), 2494-2497.
Liu, J., Bennett, K., Harun, N., Zheng, X., Probst, J., & Pate, R. (2007). Overweight and physical inactivity among rural children aged 10-17: a national and state portrait. Retrieved from South Carolina Rural Health Research Center website: http://rhr.sph.sc.edu/report/%287-1%29Obesity%20ChartbookUpdated10.15.07-secured.pdf
Liu, J. H., Jones, S. J., Sun, H., Probst, J. C., Merchant, A. T., & Cavicchia, P. (2012). Diet, physical activity, and sedentary behaviors as risk factors for childhood obesity: an urban and rural comparison Childhood Obesity, 8(5), 440-448. doi:10.1089/chi.2012.0090
Ogden, C., Carroll, M., & Flegal, K. (2008). High body mass index for age among US children and adolescents, 2003-2006. Journal of the American Medical Association, 299(20), 2401-2405. doi:10.1001/jama.299.20.2401
Ogden, C., Lamb, M., Carroll, M., & Flegal, K. (2010). Obesity and socioeconomic status in children and adolescents: United States, 2005-2008 (Data Brief No. 51). Retrieved from National Center for Health Statistics website: http://www.cdc.gov/nchs/data/databriefs/db51.pdf
Ogden, C., Carroll, M., Kit, B., & Flegal, K. (2014). Prevalence of childhood and adult obesity in the United States, 2011-2012. Journal of the American Medical Association, 311(8), 806-814. doi:10.1001/jama.2014.732.
Shah, S., Kublaoui, B., Oden, J., & White, P. (2009). Screening for type 2 diabetes in obese youth. Pediatrics, 124(2), 573-579. doi:10.1542/peds.2008-2949
Sonneville, K., La Pelle, N., Taveras, E., Gillman, M., & Prosser, L. (2009). Economic and other barriers to adopting recommendations to prevent childhood obesity: results of a focus group study with parents. BioMed Central Pediatrics, 9(1), 81. doi:10.1186/1471-2431-9-81
Trombini, E., Baldaro, B., Bertaccini, R., Mattei, C., Montebarocci, O., & Rossi, N. (2003). Maternal attitudes and attachment styles in mothers of obese children. Perceptual & Motor Skills, 97(2), 613-620.
US Preventive Service Task Force. (2010). Screening for obesity in children and adolescents: clinical summary of US Preventive Service Task Force recommendation. Retrieved from http://www.uspreventiveservicestaskforce.org/uspstf10/childobes/chobessum.htm
Vannucci, A., & Wilfley, D. E. (2012). Behavioral interventions and cardiovascular risk in obese youth: current findings and future directions. Current Cardiovascular Risk Reports, 6(6), 567-578. doi:10.1007/s12170-012-0272-y
Jill Cochran has been involved in rural health for 35 years. After obtaining her Masters Degree in Nursing, she practiced in small rural clinics in West Virginia. She is a visiting lecturer for West Virginia School of Dentistry and has been involved in educating health science students in Appalachian values and health care since 1992. She completed her doctoral studies in 2011 at West Virginia University and conducted research in children's obesity. She completed a fellowship with the National Rural Health Association in 2009 and has continued as an advocate for rural health. She is presently assistant professor at West Virginia School of Osteopathic Medicine and has been a nurse practitioner in pediatrics at Robert C. Byrd Clinic for 11 years.
Adam Baus, MA, MPH
Adam Baus is the Assistant Director of the West Virginia University Office of Health Services Research within the School of Public Health. He is also the Network Coordinator for the West Virginia Practice Based Research Network within the West Virginia Clinical and Translational Science Institute. He has over 11 years of experience in working directly with primary care centers on quality of care improvement and research efforts. He has a Masters degree in applied social research, a Masters degree in public health, and is a Doctoral candidate in public health sciences.