Skip to content

Using evidence from outside randomized controlled trials

May 29, 2017

Observational trials and large datasets are increasingly being used to answer research questions that randomized controlled trials cannot, according to experts at the Big Data in Biomedicine 2017 Conference in Stanford, California.

Certain issues are not appropriate for randomized controlled trials, such as the value of breast-feeding and the optimum weight for good health, said Mark Cullen, MD, director of the Center for Population Health Sciences at Stanford University.

Observational studies can be faster, less expensive, and better able than large clinical trials to deliver on the promise of precision medicine, he noted.

Each type of study has its place, said Allison Kurian, MD, associate professor of medicine and of health research and policy at Stanford University.

There is “nothing like a randomized controlled trial” for the assessment of novel therapies, “but there are important questions for which observational studies may do as well, or better,” she told Medscape Medical News.

In her work, observational data have led to an online decision tool that can help women who carry the BRCA gene mutation manage their risk for breast cancer.

Projecting Breast Cancer Survival

Dr Kurian and her colleagues used data from the Surveillance, Epidemiology, and End Results (SEER) national cancer registry to develop simulation models that project outcomes, by age, if a woman has her ovaries or one or both breasts removed or if she opts for intensive annual screening.

“There’s no way you could compare all those things in a randomized trial,” Dr Kurian pointed out.

She and her team are currently working on a project that links SEER data from California and Georgia — two states with distinctly different population and geographic mixes — with SEER registries from all states and genetic testing data from the laboratories used by those two states. The project is supported by a grant from the University of Michigan.

“All the richness of patient demographics and tumor characteristics that we can get out of SEER” for the 160,000 patients diagnosed with breast or ovarian cancer in the past 4 years is then linked with “very detailed genetic information from the laboratories themselves,” she explained.

“This will be an incredibly powerful tool for understanding what happens when a person with a genetic mutation is treated, what the impact is, and how common these are,” she said. “This was impossible to find out, really, until we began to merge these big datasets.”

In another ongoing trial, known as ADAPTABLE — Aspirin Dosing: A Patient-centric Trial Assessing Benefits and Long-term Effectiveness (Curr Cardiol Rep. 2016;18:81) — daily aspirin is being examined to see whether 325 mg or 81 mg is better for the prevention of heart attack, stroke, and bleeding in people living with heart disease.

Clinical trials typically require people to come to researchers, but in this trial, patients are followed as they come in for regular healthcare appointments, said colead investigator Adrian Hernandez, MD, director of outcomes and health services research at the Duke Clinical Research Institute in Durham, North Carolina.

Information is tracked through patient portals and healthcare providers, he explained.

Bringing the Study to the People

The 3-year trial, which started last year, is being conducted through the National Patient-Centered Clinical Research Network, or PCORnet, which is a community of patients, clinicians, and health systems.

“We’re seeing that you can actually identify and recruit participants in a more rapid way. The average cardiovascular trial recruits one participant per site, or less, per month,” Dr Hernandez reported, but “we’re seeing about 40 or 50.”

All stages — identifying patients, approaching patients, getting consent, and randomizing patients — are conducted electronically. Researchers can reach out to patients without internet access directly, by phone, but most people generate the data as soon as they walk through the doors of the health system or information is entered in their portal.

Last year, 40 million potential study participants walked through the doors of the PCORnet health system, he pointed out.

“Doing it in the real-world clinical setting is important,” Dr Hernandez said, and the system can be reused for other trials.

Dr Cullen has disclosed no relevant financial relationships. Dr Kurian reports that her institution receives funding from Myriad Genetics and participates in unfunded research collaborations with Genomic Health, Invitae, GeneDx, and Ambry Genetics. Dr Hernandez reports receiving research grants from Portola and AstraZeneca.

Big Data in Biomedicine 2017 Conference. Presented May 25, 2017.


From → Uncategorized

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: