PCORnet Blog

Leveraging PCORnet and Health Data to Make Gains against Obesity

Date: August 18, 2015
Authors: Rachael Fleurence, PhD, Program Director, CER Methods and Infrastructure, PCORI and Joe Selby, MD, MPH, Executive Director, PCORI

Obesity is at epidemic levels in the United States. About 30 percent of American adults are considered obese, with 5 percent severely obese, and among adolescents (ages 12-19), 20 percent are obese. This makes obesity second only to tobacco use as the top cause of preventable death. Obesity is associated with a long list of diseases and conditions that impose an enormous burden on patients, their families, and the healthcare system as a whole—including diabetes, heart disease, certain cancers, musculoskeletal conditions, and sleep disorders.

As part of our mission to help people make better-informed healthcare decisions, PCORI has funded a range of research projects on obesity. Our latest effort will apply the power of PCORnet, our developing national patient-centered clinical research network, to conduct two compelling comparative effectiveness research (CER) studies in this area—one comparing the health benefits and safety of three available surgical approaches to weight loss and the other looking at whether antibiotic treatments given early in life differ in their tendency to cause excess weight gain in children. Our Board of Governors approved awards for both studies at its August 18 . These projects are the latest designed to demonstrate PCORnet’s capacity to answer important health questions. They are observational studies, meaning they will rely on analyzing data collected in electronic health records (EHRs) as a part of everyday care. Researchers will review this information representing patients across the country to assess the outcomes that patients experienced from the care they received. To protect patient privacy and ensure data security, the EHR information is securely maintained behind the firewalls of the home institutions, and the researchers work only with general information combined from many records.

PCORnet is well positioned for these new studies. Because obesity is such a major health problem, we required all of the 11 CDRNs in Phase I and the 13 CDRNs in Phase II (launching in October 2015) to include obesity as one topic of focus. Since the networks are so large and represent a wide variety of patient groups, researchers can assess whether treatments affect groups differently.


Which Surgical Approach Works Best?

Diet and exercise are the most common prescriptions for losing weight, but that approach doesn’t work for many persons with severe obesity. So some very obese people turn to bariatric surgery, procedures that restrict food intake and, in some cases, interrupt digestion. Studies suggest bariatric surgery, followed by healthy eating and regular exercise, not only reduces weight but lowers bariatric surgeryrates of diabetes and even death. The PCORI-funded study will expand current evidence to include adolescents and older adults and follow patients for a longer period.

Led by Group Health Cooperative in Seattle, the new study will compare outcomes of patients who have received one of three surgical treatments: Roux-en-Y gastric bypass, adjustable gastric banding, or sleeve gastrectomy. The study will document changes in body mass index (BMI, a number that describes weight related to height) and rates of diabetes remission and relapse, as well as such major adverse events as additional hospitalizations and surgeries, hernias, and death. In addition to providing an overall comparison of the effectiveness of the different bariatric surgery approaches, the study will break out results for subgroups, including racial and ethnic minorities, older adults, and adolescents.

Through a series of focus groups made up of adolescents and adults with severe obesity, the study will also develop a database of patient preferences and opinions on such topics as whether to undergo weight-loss surgery, which type of surgery to receive, and how follow-up care is provided.

The project will use information from medical records from 10 PCORnet Clinical Data Research Networks (CDRNs), involving 53 healthcare organizations and more than 60,000 patients who have undergone bariatric surgery in the past years; half of these patients underwent gastric bypass, 10 percent gastric banding, and 40 percent sleeve gastrectomy procedures. The study includes the largest group of adolescent bariatric patients (900) ever studied, as well as more than 17,000 patients with diabetes. It’s also the first large study with substantial numbers of patients who received sleeve gastrectomy, a newer procedure.

To protect privacy and ensure data security, each CDRN has its own governance system and maintains its own data, securely protected by an institutional firewall. Each partner network has developed security measures, policies, and procedures that it deems right for its members, with patients involved in an ongoing basis.

Do Antibiotics Play a Role?
Because obesity is so difficult to treat, it’s also important to look at how to prevent it. Some research has suggested that treatment with broad-spectrum antibiotics is associated with obesity in children. Antibiotics are among the most valuable treatments available, but recent research has raised concerns about the possible long-term impacts of certain antibiotics widely prescribed in early infancy for routine conditions such as ear infections. So the second study approved by our Board will look at whether use of these antibiotics causes excess weight gain in early childhood, compared with more specific, or narrow-spectrum agents, and, if so, at which ages and among which patient subgroups.

This project, being conducted by Harvard Pilgrim antibioticsHealth Care in Boston, will look for a connection between antibiotic use and patterns of childhood growth. The study will assess various types of antibiotics, frequency of use, and the timing of treatments in the first two years of life. It will follow children’s growth trajectories from infancy to age five and measure obesity at ages 5 and 10. It will also examine the correlation between antibiotic use and obesity among subgroups, including racial and ethnic minorities, and whether the child’s mother received antibiotics during pregnancy.

The study uses information from medical records of 600,000 children across 42 healthcare systems participating in nine PCORnet CDRNs. This information remains behind the firewalls of the home institutions.

The study also will employ focus groups of parents and in-depth interviews of clinicians to explore how best to put the findings into everyday practice.

PCORnet as a Platform for Obesity Research

We’re excited by the role these studies can play in helping patients and those who care for them make better-informed decisions about obesity prevention and treatment. But we’re also pleased by the chance to further demonstrate PCORnet’s capacity to harness the power of partnerships and health data to conduct critical comparative clinical outcomes research faster, more efficiently, and less expensively than is possible now. These new trials follow our first PCORnet demonstration project, which seeks to determine the optimal dosage of aspirin in preventing heart attack and stroke in adults with heart disease.
We’re confident that the results of these and other demonstration studies that we expect to roll out in coming months will show that PCORnet can be an important new tool in the effort to provide evidence-based answers to healthcare questions that patients, their families, and clinicians face daily. We hope you’ll keep track of our progress.