This month's featured paper is from the lab of Jennifer DeVoe, M.D., DPhil, MPhil, MCR, associate professor of family medicine and medical informatics and clinical epidemiology, and is titled, "Linkage methods for connecting children with parents in electronic health record and state public health insurance data." It was published in the Maternal and Child Health Journal.

March 27, 2014

Who are your parents? Most of us can answer this question quickly. But asking the same question of an electronic health record (EHR) or health insurance data is a more complex issue. Why? Health insurance is increasingly fragmented within a family, with parents and children often carrying different insurance coverage through different programs. Linking data from parent to child, therefore, becomes a much more difficult task. This month's featured paper, "Linkage methods for connecting children with parents in electronic health record and state insurance data," published in the Maternal and Child Health Journal, highlights the research methodology crafted by a multi-institution research team led by Jennifer DeVoe, M.D., DPhil,  MPhil, MCR, associate professor of family medicine, to accomplish this undertaking.  

The ability to link children to their parents through electronic health records and health insurance data has significant ramifications for the efficacy of health care delivery.  "When parents have health insurance, it is more likely that their children are insured and receive guideline-appropriate healthcare," said Heather Angier, M.P.H., lead author, research associate in the department of family medicine. "Because of this strong association between coverage for parents and optimal health insurance and care for children, efforts to optimize children's health must be informed by coverage and health care services use data from both children and their parents." 

Obtaining such family-level information, therefore, requires the ability to link children with their parents in large, health care-related databases. "While data on individual patients has been linked across multiple datasets for various purposes, we knew of no published methods for achieving linkages between children and their parents in EHR or public health insurance administrative data," said Ms. Angier. 

A collaborative team of researchers from OCHIN, OHSU and Kaiser Permanente developed a methodology for successfully linking children with their parents in two large datasets: 1) EHR data from Oregon OCHIN clinics and 2) administrative data from the Oregon Health Plan (OHP) for individuals enrolled in Oregon's Medicaid and Children's Health Insurance programs. OCHIN is a technology innovation and dissemination organization that centrally hosts and maintains an EpicCare© EHR system, shared among many community health centers across the country. OCHIN uses EHR systems and EHR data to facilitate care innovation and improvement in clinics that serve vulnerable populations. This study utilized EHR data from 141 clinics in the OCHIN network serving both children and adults in Oregon. With the OCHIN dataset, this research team used information regarding either the child's emergency contact or their 'guarantor' (financially liable adult) for children's visits between 2002 and 2010. Of the nearly 250,000 children in the dataset, 62,967 had a 'linkable' parent in the EHR. In the OHP data, client and household identification numbers were used to identify links; from 889,452 household ID numbers assigned to at least one child, 525,578 had a 'linkable' parent (272,578 households). 

"We then used OHP individual client ID numbers to assess the quality of our links through cross-validation," said Dr. DeVoe. "Thus, 99.8% of EHR links were validated in OHP, and 97.7% of OHP links were validated in OCHIN's EHR." The process of identification and verification of the parent-child links required both a large-scale computational approach and rigorous attention to detail in cleaning and standardizing the datasets.

"Dr. DeVoe and Ms. Angier's research has set the stage for what will be an increasingly important research area – understanding how well health care is provided, in this case to children and families," said Dr. Eric Orwoll, associate dean for clinical science and director of the Oregon Clinical Translational Science Institute.  

The ability to link children to their parents in health care-related datasets will inform research, policy and practice in many different ways. "For example," said Ms. Angier, "it could allow for treatment that targets the whole family. Obesity is often a problem for children and their parent(s), so treatment options could be directed at the whole family if clinicians are aware of both the child and parent(s) weight. It could also help health care recommendations for disease. For example, if the EHR indicates the child's parent is a smoker, this information could prompt a provider to offer additional education on how this could impact the child's asthma and offer resources to help a parent quit." 

Dr. DeVoe points out that, "Having known linkages between parents and children in their medical records might facilitate better coordination of care for families and health reminders. A provider could be electronically prompted to remind a mother at her visit about her child's immunizations that are overdue and help get these scheduled. Importantly, these linkages could be used to better understand and diagnose children by being able to see the medical issues suffered by their parents." 

Dr. Orwoll says he chose this paper, in part, because this team is pioneering new approaches for combining information from several sources of big data. "Their methods will enable studies with important practice and policy implications," he said. "Moreover, their collaborative team drew from OCHIN, OHSU and Kaiser Permanente NW Center for Health Research, and is a model for trans-system health services research."  

Congratulations to Ms. Angier, Dr. DeVoe, and the collaborative research group on a groundbreaking paper that will help shape our future understanding of EHR and health insurance data.

Particulars:

Angier H1, Gold R, Crawford C, O'Malley JP, Tillotson CJ, Marino M, DeVoe JE
Matern Child Health J. 2014 Feb 23. [Epub ahead of print]

Authors:

Heather Angier, MPH, research associate in the Department of Family Medicine, OHSU
Rachel Gold, Ph.D., MPH, assistant investigator, Kaiser Permanente Center for Health Research
Courtney Crawford, senior research assistant in the Department of Family Medicine, OHSU
Jean O'Malley, MPH, research associate in the Department of Public Health & Preventive Medicine, OHSU
Carrie Tillotson, research associate in the Department of Public Health & Preventive Medicine, OHSU
Miguel Marino, Ph.D., assistant professor in the Departments of Family Medicine and Public Health & Preventive Medicine, OHSU
Jennifer DeVoe, M.D., DPhil, MPhil, MCR, associate professor in the Department of Family Medicine, OHSU

• Read the paper
• Read more about Dr. DeVoe

More Published Papers

• Previous School of Medicine Papers of the Month
  
• Recent OHSU published papers

Pictured

Top: Jennifer DeVoe, M.D., DPhil, MPhil, MCR

Bottom: Carrie Tillotson, Rachel Gold, Ph.D., MPH, Miguel Marino, Ph.D., Jean O'Malley, MPH