The report, published in the December edition of Current Epidemiology Reports, showed that without appropriate care, false safety signals and false safety assurances can become a real and significant risk to patients and healthcare systems.
The researchers found that many data sources do not include information on potential risk factors that affect health outcomes, such as use of illicit substances, use of over-the-counter medicines, smoking, and actual adherence to the medication. Additionally, some long-term outcomes, such as cancer, cannot be studied easily because of the relatively short-term follow-up contained in many databases.
"The monitoring of drug safety cannot yet be delegated to smart algorithms applied to healthcare databases," said Elizabeth Andrews, Ph.D., vice president of pharmacoepidemiology and risk management at RTI-HS and co-author of the paper. "With all the fervor around the potential for big data, it's critical to keep in mind the substantial differences across databases in content, coding systems and practices, duration of available medical history and follow-up time, quality of outcome information, and clinical practice patterns."
The paper reviewed the history of insurance claims, electronic health records, and disease registries used for pharmacovigilance and drug safety. The authors also explored the evolution of methods to improve the quality of this research and offer best practice recommendations for meeting the challenges of conducting the research.
"The use of healthcare databases in the evaluation of medical products has become very popular in recent years," said Suzanne L. West, Ph.D., Fellow at RTI International and senior author of the paper. "We wanted to study the evolution of database use as well as the implications of using these sources of data."
The findings showed that healthcare databases have improved in number and quality, which allows researchers to study more products in diverse settings and to better evaluate some rare outcomes.
The paper also notes that healthcare databases typically include real world evidence, or secondary sources of data, as opposed to strictly using data from clinical trials. The real world data that go into healthcare databases are often not verified against source records or controlled, which has both benefits and drawbacks.
"A drawback of clinical trials is that they are highly controlled and highly monitored to ensure strict adherence to protocol; however, that's not how people take drugs in the real world," said Andrea V. Margulis, MD, ScD, senior research epidemiologist and co-author of the paper. "Real world data have problems too because they are so uncontrolled. With this report, we wanted to explore the pros and cons of working with real world data."
RSS