Social media: The evolving impact on clinical trials
January 19, 2016
This post was written by Lindsay McNair, MD, MPH, MSB, Chief Medical Officer, WIRB-Copernicus Group
Social media represents both opportunities and threats for clinical trials. On the positive side, social media interactions can build engagement and satisfaction among trial participants, allowing them to share experiences and be part of a community where their contributions are recognized. Before a study, social media can boost study recruitment, for example, Novartis used a Twitter feed to boost awareness of a phase 2 trial involving stomatitis and breast cancer, and others have used text messaging.
For individuals interested in participating in phase 1 trials, JALR.org provides links to commercial sites, including details about factors such the quality of the Wi-Fi and food, and the speed of payment, enabling participants to be ‘informed shoppers.’ RateClinicalTrials.co.uk and Yelp also post reviews of facilities that conduct clinical trials. Many sponsors are comfortable with and benefit from these types of pre-trial social media interactions, which have been taking place for several years.
Online, real-time sharing of study experiences
A less positive trend is for clinical trial participants to talk about their experience during a study through Twitter feeds or other social channels. Patients have posted structured data on efficacy and safety on an investigational drug to public and social websites. MedHelp.org has communities for people with various diseases and conditions. In one example, people with hepatitis C self-identified as being involved in a particular trial, and then posted discussions aimed at identifying who was receiving active drug, and who was receiving placebo – including postings of photos of the medication involved. Sample threads involving communication between subjects included:
Discussions may also center on side effects, with community members – who use pseudonyms or avatars, making follow-up by the trial sponsor impossible – advising others to stop taking the study medication. Threads have also included discussion about how and when to get certain lab tests outside of the study to determine whether subjects were receiving placebo, so that they could drop out of study if not on the investigational agent.
Discussions also appear in forums such as Yahoo! Chat, reflecting the fact that study participants may purchase stock in the company making promising new products, and financial analysts may be interested in predicting likely outcomes from trials to tailor their buy/hold/sell recommendations for investors, based on information that is not yet public.
PatientsLikeMe.com, a for-profit company, also has disease-related communities. Trial participants are sometimes asked to rate their experiences with various treatment regimens that are in clinical trials. There is also the website, ‘waiting for p<0.05,’ which has done analyses of data reported by self-identified trial participants, also in an attempt to predict trial outcomes.
Concerns about patient safety and bias
This trend is causing concern among sponsors, investigators and CROs, primarily due to the risks to patient safety and safety reporting, the possibility that patients may discuss adverse events online but not inform Principal Investigators, and also because of potential for bias. Data may be aggregated in attempt to predict outcome of trials still underway, with non-subjects publishing conclusions about results while the study was still ongoing (e.g., “10/20 subjects report online that they have responded, response rate is 50%...”)
This issue was highlighted in a Wall Street Journal article, ‘Researchers Fret as Social Media Lift Veil on Drug Trials: Online Chatter Could Unravel Carefully Built Construct of 'Blind' Clinical Trials.’ “On Facebook groups, online forums and blogs, some patients are effectively jeopardizing the blind. In trials for hepatitis C, multiple sclerosis and ALS…patients have been sharing details of their reactions and trying to figure out whether they are getting the active drug. Patients also swap tips on how to get accepted into trials, even if they don't meet all the requirements. Some who are in trials collect and analyze their medical data and others' to get an early indication of whether a drug will be a success.” The newspaper quotes Craig H. Lipset, head of clinical innovation at Pfizer as saying patients who share too much "could effectively chill a new drug before it ever gets to patients by misinterpreting early signals." In an opinion piece in Nature Medicine, Lipset asserted that “clinical trial sponsors must work with regulators to define pathways to monitor social media use by trial participants to understand if conversations on the internet will affect their interpretation of study results.”
The future: A case-by-case ‘risk calculator’
Looking ahead, decisions on how to proof trials as far as possible to these potential negative impacts will have to be made on a case-by-case basis. Sponsors might develop an informal ‘risk calculator’ to assess the chance of negative impacts from social media in a particular trial. This would reflect the fact that that not all indications, patient populations and protocols bear the same risk. Factors to consider might be whether the disease involved has an active advocacy group, or is a rare disease, both of which may predispose participants to online data sharing. In these cases, sponsors might consider creating specific chat-rooms that are moderated, allowing erroneous or inappropriate content to be removed. A request for patients only to disclose their trial experiences to immediate friends and family could also be included in the informed consent form. In contrast, diseases such as osteoporosis – where the patient population is typically elderly, with a limited online presence and no strong advocacy organization – might pose a lower threat of online data sharing and associated risks involving patient safety and bias.
Similarly, trials with entirely objective study endpoints will likely not be affected by online discussions, while those with subjective outcomes (such as pain in migraine) could potentially see a significant impact on outcomes. The size of the study is also a key factor: one with several thousand participants may be diffuse and diverse enough for communication to have little impact. In a 12-person study, in contrast, one participant who posts online about a negative experience may have a significant impact on both the study outcomes and the ability of the study to recruit additional patients.
Educating participants about the potential impacts of their online communications will be important. A website, Smart Talking About Clinical Studies, developed by the Center for Information & Study on Clinical Research Participation (CISCRP) with support from Shire, highlights the risks of sharing clinical trial experiences online. The website notes “the surprising effects of words” on the outcome of clinical studies, illustrating the potential impact of discussing eligibility criteria, sharing side-effect information, and the placebo effect. Messaging includes the fact that clinical studies “help us develop new medicines. And if you join one, you might want to tell friends, family and people online about it. That’s great. Why? Because clinical studies need volunteers and telling them will help get people more involved. But words are powerful – so powerful they can change:
In future, social media use among clinical trial participants is likely to increase. Sponsors and researchers need to be prepared to address participants’ questions, encourage clear ongoing communication with clinical sites, and be aware of any potential impact of social media participation on study conduct and outcomes.