Voice of patients missing in adverse event reporting
Patient reports could unearth problems sooner
Rather than blind, call it a "deaf spot." The adverse event reporting from clinical trials that helps inform conclusions about a drug's safety has a crucial hole in it the voices of actual patients taking the drug.
That's the opinion of oncologist Ethan Basch, who says reports from patients tend to be more sensitive to adverse events than those filtered through clinicians. Indeed, such reports could bring serious problems with a drug to light earlier.
"It's important that clinical investigators and IRBs or [data safety monitoring boards] have a real understanding of the toxicities of agents that we're evaluating," says Basch, MD, a researcher at Memorial Sloan-Kettering Cancer Center in New York City. "As such, anything we can do to improve our understanding of the toxicity profiles of drugs or devices is beneficial. Because after all, what we're interested in is efficacy and safety."
Basch published his views in a recent commentary in The New England Journal of Medicine.1 He says studies have demonstrated that compared to patients, clinicians tend to report lower severity of adverse symptoms from the same events.
Adding patient reports to the adverse events collected and studied in a trial could increase researchers' ability to pinpoint serious adverse events sooner, Basch says.
"As far as detecting symptoms, patient self-reporting is emerging as a gold standard," he says. "It's already the gold standard for efficacy evaluation and it should be the gold standard with risk evaluation as well."
Why do patient and clinician reports differ? Basch thinks it could be a combination of factors. In order for an adverse event to be reported, the patient must volunteer the information or be asked about it. Then, the clinician has to document it and a data manager has to classify the information using a system such as the Common Terminology Criteria for Adverse Events (CTCAE), and put it in a database.
Along the way, information can be lost or altered, affecting the accuracy of the data, Basch says "It's like the children's game of 'telephone' you start with a word and by the time it gets to the end of the chain, it's something else related, but different," he says. "That's what happens with the reporting of adverse events. Things get dropped or they get changed. Patients may downplay or up-play their symptoms to clinicians. The clinician may think the patient is exaggerating or downplaying and may alter what the patient is saying. And if we look from clinician to clinician, we're very unreliable, compared to each other."
Surveys by paper, phone
Basch says patient reporting could be accomplished in a number of ways through human interviews, paper surveys or electronic means such as automated phone surveys.
"Automated telephone systems are the way that we book our airline ticket or our movie ticket," Basch says. "These kinds of systems are increasingly used for administering different kinds of efficacy measures. And these could also be used for adverse event reporting."
In Basch's view, those reports wouldn't go straight to an IRB or data safety monitoring board. Instead a report would go to the investigator, with severe symptoms triggering an alert to the research team, which could evaluate them to see if they require expedited reporting. Patient reports would ultimately be submitted with all other adverse events data at the conclusion of a trial.
The research community already is coping with a large body of adverse event reports, which some say create so much "noise" that it's hard to make out the "signal" of a real problem with a drug. While adding the patient reports could pose some administrative challenges, Basch doesn't think it would increase that noise-to-signal difficulty. For one thing, he says, we'd get a better handle on the baseline symptoms that patients experience, many of which may not have anything to do with the study intervention.
"It's possible that clinician reporting of adverse events is more specific but less sensitive and patient reporting is more sensitive but less specific," Basch says. "However, when it comes to detecting adverse events, we would rather be sensitive. We would rather not miss important things. What happens with clinician reporting is that we miss important things."
No IRB requirements
Still, Basch doesn't suggest that IRBs start requiring patient reporting in protocols they review. He says it's likely that investigators and sponsors will undertake it on their own, recognizing that it improves the quality of data being collected. Eventually, he says, he would like to see regulatory action, for example in the form of an FDA guidance.
Basch and his colleagues already have developed a Patient-Reported Outcomes version of the CTCAE (called the PRO-CTCAE) for the National Cancer Institute, as part of an initiative to introduce patient-reported outcomes in NCI-sponsored trials.
"I think IRBs do have a role, but first, I think this is going to come from investigators and regulators and then IRBs will follow suit," he says."I don't know that it's for IRBs to go back to investigators and say `You have to include a screen of patients for adverse events.' That said, when there are adverse symptoms of particular concern, it is warranted for investigators or sponsors to include patient-reported measures for those particular symptoms."
If an investigator proposes using a patient-reported tool in a trial, the IRB should ensure that it is valid and reliable, and that the way it's being administered is reasonable, Basch says. There should be a mechanism for real-time review, with a clinician evaluating reports to see if something turns out to be a serious adverse event that requires expedited reporting, he says.
Reference
- Basch, E. The Missing Voice of Patients in Drug-Safety Reporting. N Engl J Med 2010 Mar 11;362(10):865-9.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content