Nimita Limaye, PhD

At the 2018 DIA Annual Meeting, our DIA Clinical Data Management (CDM)/Risk-Based Monitoring (RBM) Community met with many people who were implementing or preparing to implement RBM and risk-based quality management (RBQM) approaches. Our attendees were looking for support and a forum to address their concerns. They knew that there were new guidelines and that the industry was anxious to get started with RBM activities, but they also wished to address the pitfalls one might expect. To address this need for a discussion of the changes taking place in our industry and to support each other in this endeavor, our group presented a segment of our roadmap beginning with a look at pre-study risk planning, vendor selection, and change management in RBM and RBQM.

 has focused upon the implementation of RBQM activities in clinical trials. Multiple stakeholders were looking for real experiences that can be incorporated into their risk management strategies. The FDA’s comment period on the draft guidance “A Risk-Based Approach to Monitoring of Clinical Investigations”

reminded us of the need to focus on how we were implementing the RBM guidance. The comment period also offered us an opportunity to communicate our suggestions for how to interface with and incorporate the needs of our communities and stakeholders. Another starting point for our roadmap was the case study from Cyntegrity.

 To assist in this overall endeavor, we have been working on an RBM Roadmap produced by the CDM and RBM DIA Communities. We envision a set of best practices.

	ICH GCP E6 (R2) requires clinical trial sponsor companies to incorporate RBM in the risk management plan.

	Risk planning should start with the development of the target product profile (TPP)

 and then continue into the development of an individual protocol for a study. Some aspects of the TPP, such as once-daily dosing or a positive safety profile, must be taken into account when developing the risk management plan for a study.

	The most impactful operational risks to a prospective clinical trial will most often be associated with the most important processes. Yet, it is not only the critical data and processes, as highlighted by ICH E6 R2, but also the systems, (devices, sensors, or software apps) that are used in clinical trials that contribute to risk.

	Thus, to begin with, the study team identifies critical data, tools, and processes. One example of critical data is the survival data of patients that withdrew their informed consent or those that are lost-to-follow-up in an outcome trial. These inputs may come from data management (DM), biostatistics, or clinical operations (CO). Examples of critical processes include the timely reporting of serious adverse events to the sponsor and the adequate management of those by the sponsor. This process depends on multiple stakeholders, including DM, pharmacovigilance/drug safety, regulatory affairs (RA), medical affairs (MA), and CO. In addition, it may be worthwhile to check inspection findings from the FDA or European Medicines Agency (EMA) to see which have been areas of concern based on historic data. Keeping a closer eye on these findings may be very helpful in the future. As we will discuss later, each study team member plays a unique and crucial role in the risk-identification process. Thus, while RBM focuses on risk-based monitoring, the integration of a systemic risk management strategy for identifying and managing risks-from study design to the submission of the clinical study report-has resulted in a more evolved descriptor called risk-based quality management (RBQM).

	Once the critical data, processes, and systems used in a trial have been identified for the study, a core study team representing cross-functional expertise should conduct an exercise of risk identification. This includes identifying all of the potential risks that may impact patient safety, data quality, and/or the operational success of the study. Identified risks should be ranked and prioritized based on three dimensions: impact, likelihood, and detectability (ILD). The product of ILD helps derive the risk score or the risk probability number (RPN). This will help the team to focus their risk mitigation efforts on the greatest risks to the study.

	Once the operational risks have been identified and characterized, the team develops a risk mitigation action plan to address the highest-priority risks. As previously mentioned, a high rate of lost-to-follow-up or discontinued patients might be an identified risk for the study. A number of mitigation actions might be considered for this risk:

		Removing one or more ancillary procedures from the protocol requirements that are not critical to support study endpoints may reduce the relative burden on patients participating in the study.

		Incorporate additional awareness training for sites related to this risk, including the need to check proactively with patients on their burden/comfort level with continued study participation, and what can be done to improve their experience, etc.

		Implement one or more key risk indicator (KRI) metric to identify sites that may be experiencing a high rate of lost-to-follow-ups, missed or missing visits, or withdrawn consents.

	These mitigation planning items need to be documented carefully, either within the risk management plan (RMP), risk log, or integrated quality risk management plan (IQRMP), where accountabilities can be assigned. The individuals accountable for managing the various risks must be identified via the operational plans, such as the central monitoring plan, the CDM plan, the site monitoring plan, or the pharmacovigilance plan. Thus, the RMP serves as a master plan, feeding into the individual operational plans and is, hence, described as an “integrated” risk management plan.

	Once the key risks have been specified and mitigation plans established, the complete set of KRIs to be used in support of risk detection should be confirmed. The list should ideally be comprised of therapeutic area-specific and study-specific KRIs identified during mitigation planning, along with a standard set of generic KRIs that the organization has identified to be used across studies.

The study team should identify about 10 to 25 KRIs for a study. When the team finds lost-to-follow-up patients and withdrawn informed consent, it should monitor the KRI for an increasing number of missing or missed visits at sites throughout the study. Applying relative weights to each of the KRIs may be helpful in supporting decisions on which KRIs to follow up on as they emerge during the study and the urgency with which outliers need to be addressed. Not every risk in a study can or should be mitigated. This exercise, though, must be run very carefully in order to not overlook critical data or processes. Changes in local regulations is illustrative of a risk that needs to be accepted. It is also important that teams are trained not only on defining the KRIs, but on interpreting them as well.

	Vendor selection for the implementation of RBM

	Vendor selection when implementing RBM activities is not an easy task. It involves internal and external challenges.

	At the start, one needs to have clarity within the organization as to what are the driving factors for implementing RBM. Is this a strategic decision with clear objectives (i.e., decreased number of site visits and increased quality) and clear milestones (i.e., decreased number of site visits) that improve quality? Many companies have not bought into this model, but are merely testing the waters. Some are recognizing the significant weightage that the FDA, EMA, TransCelerate BioPharma, and the latest ICH E6R2 guidelines have given to RBM and are making tactical decisions to perform a pilot study. However, it is important to remember that the implementation of RBM involves a significant investment in technology-and the benefits of this approach may not be evident by performing one small RBM pilot study.

	Another internal challenge is whether to follow a one-stop-shop approach or to identify smaller, more cost-effective and amenable partners. There is often a lack of clarity as to whose SOPs should be followed and whose KRI libraries should be used-the vendors or the sponsors? Even more critical is the lack of pre-defined key performance indicators (KPIs).

 These are essential to measure the success of the RBM implementation strategy and of the partnership as well.

	A survey conducted at the 2018 DIA meeting in Boston demonstrated that around two-thirds of the participants have never been involved in RBM vendor selection, did not have internal RBM expertise, and had not defined KPIs for RBM implementation within their organizations. This shows us the challenges that the industry is facing today.

	Implementing RBM is not just about analytics, technology, clinical, or RBM domain expertise-it is really about all of them. Finding 

a vendor who really understands the science of RBM is important. It is crucial to work with a vendor who not only has a lot of experience, but is also willing to share its learnings with the sponsor company. A vendor (CRO) may have significant clinical trial experience but does not have a rich RBM implementation experience. This needs to be validated carefully. Alignment in terms of scale and geographical footprint is also important when establishing a preferred partnership. Does the vendor actually have RBM capabilities at the required locations?

	Perhaps even more important than all of the above is the cultural fit, the alignment in vision, and the flexibility and transparency offered by the vendor. This would go a long way in terms of transforming this from a sponsor-vendor relationship to a true partnership.

	Principles of change management applied to RBQM

	Why do we need to address change management as part of RBQM preparation? What do we mean by change management? John Kotter, the guru of change management, refers to the eight stages that are required for the transformation and change to take place.

		Planning for and creating short-term wins

		Consolidating improvements and producing still more changes

Change management, in the context of instituting RBM and RBQM practices, refers to more than just the relationship between the sponsors and the sites. In following the stages listed above and implementing a vision, one needs to integrate a change management strategy from the risk identification stage itself and through the entire risk management planning and vendor selection stages, as well as developing a cohesive RBQM plan, addressing the needs of each stakeholder.

	Consider changes within the organizations and how they are affected by RBQM. Successful change management for the adoption of RBQM is much bigger than our relationship with the sites. It speaks to changes in our relationship and understanding of all of the clinical trial processes and study data. It speaks to transformations in how we conduct clinical trials and how we, as a community, work together and support each other. The new RBM approach is not a competition between stakeholders, but rather a way to assist each other in minimizing risk and enhancing the quality of the output of all stakeholders.

	Bottom line-change management and the adoption of RBQM means disruption. RBQM offers a disruption that will improve the organization.

	How do we differentiate resistance versus misunderstanding? How do we address those who see the benefits but are lacking in a commitment to the rigors required of change? Do we have time to make all the changes that are necessary? How do we prepare everyone for the bumpy first pass?

	As we have seen in implementing RBQM best practices, we are involved in actualizing all eight stages at the same time. We are often not in a position to build a powerful guiding coalition before implementing the changes. Yet, the implementation of all eight stages is required for success. There is no option to not implement RBQM. How can we use Kotter’s stages to help us in our preparation and planning?

	Pharma is experiencing a sense of urgency for adopting the principles of RBQM. For DM preparing internally for the adoption of RBQM, questions about systems, case report form (CRF) designs, and the timeliness of implementation lend all data activities a sense of urgency. Clinical operations teams express a sense of urgency regarding training and education on the new terms and processes. For regulatory, quality, and program managers, the sense of urgency is two-fold: first, assessing preparedness for RBQM, and, second, demonstrating preparedness. This is equally important for quality assurance (QA), as QA needs to ensure that all SOPs are in place, teams are trained on the same, and training records are in place as well.

	It is not only the leader who is in charge of developing the KRIs, but rather a coalition or a group that can work together. Many leaders look forward to building a guiding coalition in implementing RBQM. Who has the right skills for taking the lead? Who is writing the SOP?

	Monitors often fear that they have become dispensable. They fear that communication with the site will be centralized and personal interactions will disappear. They also fear that data managers may not understand how much the sites might skew the data. For regulatory, quality, and program managers, a guiding coalition of leaders is essential. Their underlying fear is that they will not be seen as an important voice in the leadership program. The main fear is disruption and loss of quality due to the new process.

	The vendor’s experience in change management is worth assessing. Both the sponsor and the vendor need to join hands in the implementation of a strategic change management approach within the sponsor’s organization, and attempt to address the concerns of naysayers.

	Creating a vision for the implementation of RBM and RBQM depends on a succinct determination of what are considered to be best practices within the organization. With many participants offering their visions, the essential elements can be obscured or may, at times, appear too discordant to be put into practice. As discussed earlier, it is a shared vision that brings the team together.

	For the data manager, a clear vision entails updating and executing new systems for RBQM. Who is doing the testing? When does one have time for testing? How are deviations from critical data points identified, interpreted, and actioned? For monitors, there may be conflicting visions. Some individuals may not see the changes as beneficial. Changes in visit schedules and data points to be reviewed can be disruptive. RBQM may feel too open-ended-how does one check for deviations and ensure quality at the site remotely? How does one retrain the site when difficulties are noted?

	RBM results in a disruptive change, not only in monitoring methodology but also in organizational structures and processes, which is not an easy thing to accept. It is important that the sponsor recognizes the commitment that these changes will require. How does its vision for RBQM answer the requirements of the regulatory authorities as well as address the institutional procedures?

	In communicating the vision, we need to show value. What is the value that RBM and RBQM will bring to our organization? It is important to let everyone know what is being planned and to address all concerns. Concerns vary in nature for different stakeholders. Data managers may have concerns regarding who is going to define the KRIs and how they will be collected in the study. Monitors may be concerned about what they should be communicating to the site and whether the site will buy into that message. RA may have its own perspective regarding whether RBQM will really add value, or will it just pose one more regulatory hurdle? Sites are often not even informed about the change in approach, despite being the most important stakeholders after the subjects. Technology teams may have concerns around whether their platform solution will pass muster with the end users (the central monitors) and with auditors.

	An organization must have internal RBM experts (or utilize external consultants) to be able to make meaningful, well-informed decisions. A lack of internal champions may often set this up for failure.

	All stakeholders within the company and externally should be empowered to follow and contribute to the vision. As the key in the structure of implementing RBM and RBQM in the institution, central monitors are connected to all parties within and outside the company. A clear message on the implementation strategy and how structures are changing will empower everyone that is impacted by RBQM.

	The DIA survey revealed that 40% of the participants found the lack of internal champions to be the biggest challenge within their organization, followed by a lack of vision.

6-Planning for and creating short-term wins

	It is good to demonstrate success early on to show where the organization is headed and what can be accomplished. Again, showing value is essential in creating short-term wins.

	The RBQM approach is integrated right from the start of the study.

 Central monitors can show how well its worked by identifying outliers in critical data points, and alerting monitors to take corrective action. The monitors can demonstrate short-term wins by addressing these issues remotely and actually enhance the quality of the data being received from the site. Sites that are doing well, despite fewer visits, should be recognized for their achievements. For sites that need more frequent visits, the win can be to demonstrate early identification of issues and ease of follow-up.

	Once one has a short-term win, it needs to keep the momentum in place for the next set of changes. This requires preparation and planning. The first short-term wins are great, but consolidating these improvements requires vigilance. One may not have all the answers to the problems and challenges along the way. For example, a KRI that is cumulative may not be sensitive enough to show us how the site is improving or worsening, but it gives us an opportunity to consolidate our improvements and produce more changes.

	The final reminder in any change management plan is that change takes time. If one rushes through the process without validating that everything is done, it is likely that the program, in turn, will suffer. We have often seen that starting and then stopping the implementation and going back to the old ways can be more harmful than waiting until everyone is prepared to get started.

	Although companies that transition an ongoing study to the RBM approach may find it more challenging than starting with an RBM approach from the beginning, it is not impossible. But it’s crucial to get started with RBM the sooner the better.

	Institutionalizing these new approaches always takes time. The secret to success for implementing change management includes transparency and the consistent ongoing communication of a shared vision.

		ICH H. (2016). Integrated addendum to ICH E6 (R1): guideline for good clinical practice E6 (R2). Retrieved from 

https://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E6/E6_R2__Step_4_2016_1109.pdf

		FDA. (2019). A Risk-Based Approach to Monitoring of Clinical Investigations. Retrieved from 

www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM633316.pdf

		Alsumidaie M, Andrianov A. Case Study: Using RBM to Evaluate Site Engagement. 

http://www.appliedclinicaltrialsonline.com/case-study-using-rbm-evaluate-and-predict-site-engagement

		FDA. (2007). Guidance for Industry and Review Staff Target Product Profile-A Strategic Development Process Tool. Retrieved from 

www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm080593.pdf

		Telfer G. Risk-Based Monitoring Insights from the Industry: Steve Young. 2016, 

https://cluepoints.com/risk-based-monitoring-insights-from-the-industry-steve-young-2/

		Limaye N, Jaguste V. Risk-Based Monitoring (RBM) Implementation: Challenges and Potential Solutions. 

Therapeutic Innovation & Regulatory Science

https://journals.sagepub.com/doi/abs/10.1177/2168479018769284

		Limaye N. The Three Biggest Challenges in RBM Related to Key Risk Indicators. 2018, 

https://cyntegrity.com/the-three-biggest-challenges-in-rbm-related-to-key-risk-indicators/

		Limaye N. Twenty Five Key Questions associated with Vendor Selection for Risk Based Monitoring. 2017, 

http://nymro.com/twenty-five-key-questions-associated-vendor-selection-risk-based-monitoring/

		Kotter J. Leading Change: Why Transformation Efforts Fail. 

http://www.lighthouseconsultants.co.uk/wp-content/uploads/2010/08/Kotter-Leading-Change-Why-transformation-efforts-fail.pdf

		Alsumidaie M, Proupín-Pérez M, Andrianov A, Widler B, Schiemann P. Ten Burning Questions about Risk-Based Study Management. A

http://cyntegrity.com/rbm-consortium-10-burning-questions-risk-based-study-management/

, Vanderbilt University, School of Medicine, Dept. of Biostatistics; 

, Practice Lead, Life Sciences, Applied Technology Solutions, Inc.; 

, Chief Operations Officer, CluePoints; 

, Sr. Manager, Boston Scientific Corporation; 

, Consultant-Retired FDA Director, Biostatistics; 

, Chief Scientific Officer, Cyntegrity GmbH

Articles

The importance of applying change management techniques throughout the preparation, planning, and execution of RBM and RBQM approaches.

A Roadmap for Implementing RBM and Quality Management

Social media has moved beyond being a fashionable word to one that is drawing renewed attention from the pharma and healthcare industries. The power of these tools and the impact that they can have not only on brand perception and, effectively, on sales, as well as the increasing interest of the regulators in social media is resulting in this shift. In addition, tools and technology and the growth of the data sciences industry have proven to be powerful enablers. As patient centricity becomes the cynosure of attention, the need to capture their views becomes necessary. Social media brings in a pragmatic component to clinical trials, supporting the world of evidence-based medicine (EBM). Patients themselves want to draw more informed decisions and want to be participants in the decision-making process on how their health is going to be managed. 

	It has been observed in studies that of the more than 74% of Internet users that engage on social media, 80% are looking for health information,

 with 90% of the younger media-savvy 18-to-24 year-olds

 claiming that they relied upon this source, using it twice as often as the more senior population. They shared health-related conversations and patient stories on diverse topics such as how someone coped with a chronic condition, views on diet and exercise, and their choice of physician.

 The most accessed online resources for health-related information included WebMD (56%) and Wikipedia (31%).

 Patients wanted doctors to actively share updates on the disease, new pipeline compounds, their experience with different drugs, etc., on social media and 60% claimed that they trusted what the doctors were posting.

 It becomes all the more important for healthcare professionals (HCPs) to share accurate information and timely updates with the patient community. It was not only patients who were leveraging social media, 60% of HCPs were also actively watching what their counterparts are sharing on health-related issues. Doctors themselves seemed to truly believe that social media is impacting the quality of care. Forty percent of patients also relied upon social media to assess how others were dealing with chronic conditions.

	All of these situations have led to the development of numerous online patient communities such as:

 The largest online patient community, spanning 500,000 members

 and covering over 2,500 conditions. It partnered with the National Quality Foundation to leverage social media to assess the quality of life in communities with multiple sclerosis (51,000 members), chronic obstructive pulmonary disease (2,500 members), and rheumatoid arthritis (10,000 members). They also compared patient-reported outcome (PRO) data generated from specific tools provided to these communities. Analyses showed that the PROs needed to use more “patient-friendly” language to describe the symptoms and it was the first time that such a large-scale study had been performed for an assessment of this nature.

An online patient community to which the FDA granted authorization to market a direct-to-consumer genetic test, the Bloom syndrome carrier test; customers who purchase their personal genome service (PGS) would receive their ancestry information and uninterpreted raw genetic data.

A site which combines pharmacist expertise, FDA data, and real-life experience from patients.

A closed online community which provides a platform for patients and caregivers to connect amongst themselves and find disease-related guidance and answers that may be scant in other formal channels.

A British social enterprise which integrates with the National Health Service (NHS) network and allows patients control of their own medical records.

 Include those such as Doximity (has 70% of US doctors as verified members);

 Sermo (with over 800,000 members across 96 specialties);

 MomMD (women doctor’s networking site, 11,000+ active members);

	There is a whole spectrum of social media services being leveraged by pharma and healthcare, ranging across listening and analytics, marketing, and engagement. Companies are designing geography-based social media strategies based on the audience that they are targeting. Novartis, for example, thus engages with the public for its brand Gilenya

 (the once-a-day pill developed to treat multiple sclerosis) through a dedicated handle on Twitter, @GILENYAGoUSOnly. The introduction section of this Twitter handle clearly calls out that it is only for a US audience and also sets forth other guidelines of interaction, such as the response window, how to share personal details, and also the discretion Novartis would practice in responding or not responding to certain tweets.

 Internally, for a pharma company, this means the need for increased harmonization in social media efforts across countries and regions, while staying compliant with local regulations. When it comes to listening, social media tends to be more porous and it no longer matters, for instance, where an adverse event (AE) was reported, as it can find interested audiences anywhere in the world. But engagement is a different world altogether, as it has to be contextualized to the specific user, geography, local regulations, and so on.

	The likelihood of generating excess AE traffic has been one of the reasons that often dissuades the pharma industry from leveraging social media listening. The four basic elements for submission of an individual case safety report to the FDA include an identifiable patient, an identifiable reporter, a suspect drug or biological product, and an adverse experience or fatal outcome suspected to be due to the suspect drug or biological product. Pharma companies are required to publish events reported on company-sponsored websites. However, if they do become aware of an event that has been reported on another site, they should review it and determine if it requires to be reported on their’s. In recently reported industry views on the issue, Abbott felt that the entire web should be monitored for AE reports, while AstraZeneca did not align with this approach. Bayer allowed the Pharmaceutical Research and Manufacturers of America (PhRMA) to speak for the company, the group’s stand being that only company-sponsored websites should be monitored and the events meeting the reporting criteria defined by the FDA should be posted, provided the reporters were privately contactable, so as to respect patient privacy issues. Lilly also observed that national privacy laws needed to be respected and follow-up with patients on these reports should be avoided. Merck & Co. cautioned that social media reports could result in an AE being blown out of proportion, citing the example of Sanofi, which had to shut down its Facebook page when a patient who reacted to cancer drug Taxotere  posted a flood of comments about experiencing hair loss, triggering major reactions from the larger patient community. Sanofi did reopen the page later, but included terms of use. Lilly recommended that the FDA create a separate category for events reported on social media and also conducted a pilot study to demonstrate that the considerable efforts invested in monitoring social media for AEs did not yield corresponding results.

	The most significant challenge faced by pharma was identifying the reporter of AE information on social media sites. The FDA defines an “identifiable reporter” as one who is privately contactable. While Novartis believed that additional demographics were important, AstraZeneca felt that an email ID or even a Facebook contact was enough. However, both companies agreed that AEs collected from organized data collections systems should be considered as “solicited,” whereas the rest should be treated as “spontaneous” reports. Though the European Union (EU) does not mandate the monitoring of social media for AEs, it does require that events that have been observed should be reported. This may change as observations from the WEB-RADR (Recognizing Adverse Drug Reactions) project gain traction in the EU. Launched in September 2014 by the Innovative Medicines Initiative, WEB-RADR, a €2.3 million, three-year public-private project, is responsible for developing a mobile application for reporting adverse drug reactions (ADRs) to regulatory bodies in the EU region. The mobile app would help evaluate the potential of social media data in identifying safety issues.

	It has also been observed that in a study conducted on AE reported for Lipitor

 (sibutramine) in the FDA Adverse Event Reporting System (FAERS) database and on AskaPatient.com (a patient-support group website), the majority of AE reports on social media came from a younger population and focused on milder AEs on AskaPatient.com, as compared to those reported in FAERS.

	Pharma companies need to be mindful of varying regulations on the advertising of branded prescription products, while defining their product promotion strategy. It is very important, therefore, that each company develops its own social media policy guidelines. It has been observed that 23% of pharma organizations do not have policies to address data security and privacy, whereas 31% of healthcare organizations do have healthcare policies in place.

 A review of the top 100 companies listed on the London Stock Exchange demonstrated that the top three social media-savvy pharma companies are Johnson & Johnson, GlaxoSmithKline, and Pfizer, respectively This becomes especially relevant when a company is marketing its products in multiple geographies, with varying local regulations. While this is acceptable in the U.S., advertising of branded prescription products is not accepted by many other countries. 

	In addition, when tweeting, one needs to be mindful of the FDA draft guidance on 

Internet/Social Media Platforms with Character Space Limitations - Presenting Risk and Benefit Information for Prescription Drugs and Medical Devices, 

 This guidance specifically focuses on FDA-regulated medical products, the 140 character space limit associated with tweets, and the need of the pharma industry to present both benefit and risk information during product promotion while tweeting. While this does not apply to company websites, or product pages on social media websites such as Facebook or Twitter, the important take-home from this guidance is that irrespective of the character limitation, if a company is promoting the benefits of its product in a tweet, it also needs to bring to the attention of the consumer, albeit in a concise manner, the potential risks of the drugs as well, or else evaluate other platforms for product promotion. 

	Twitter, for its part, has been simplifying its rules, which gives pharma marketers some leeway. The site’s 140-character limit has revised to 280, making it easier for benefits and risks to be included together. In a survey conducted by Ogilvy Healthworld with 14 big pharma companies, a 530% increase in the number of tweets in 2014 as compared to 2013 was observed. With access significantly enhanced, the need for regulations in this regard is understandable. Interestingly, the highest increase (300%) was noted in the activity of a key German drugmaker.

	One recent example of the FDA coming down hard on pharma social media practice was with respect to the promotion of Diclegis

, a Duchesnay’s morning sickness drug, via social media posts from Kim Kardashian. The agency pointed out that while the posts highlighted the benefits of Diclegis

, the risks were not highlighted. This resulted in a warning letter being issued by the FDA’s Office of Prescription Drug Promotion (OPDP).

 It was one of 19 letters issued by the FDA, including three warning letters and 16 untitled letters.

	Not only do regulations regarding the posting of promotional information deter the pharma industry, but a large sector still perceives “listening” to be a liability. The key concern is that if companies do go down the social media path, they would need to notify regulators regarding all AEs identified by them on social media sites. The reality is that up to 90% of AEs go unreported and reporting by patients and HCPs is extremely low.

 In efforts to address these gaps, Novartis and other pharma companies have been working on Web-RADR. Over three million posts from Facebook and Twitter (55% excluded as spam) were analyzed. Two percent were termed as “proto-AEs”-information which could represent potential AEs and were coded using the Medical Dictionary for Regulatory Activities (MedDRA). Epidemico is developing algorithms to help detect potential AEs reported online and also eliminate duplicate reports. It is important to use a sophisticated, natural language processing system to eliminate the noise and draw meaningful results. About 12 drugs are being monitored by WebRADR and the data will then be analyzed by the European Medicines Agency (EMA) and the companies in the WebRADR consortium. 

	The ownership of such data-whether it should be the patient, the company, or regulator-is still being debated.

 Patient advocacy group, EURODIS, which is associated with Web-RADR and represents patients with rare diseases, recommends that all data from social media be added to EudraVigilance, the EU’s pharmacovigilance ADR database. This database is freely accessible to patients and researchers to aid signal detection. Currently, in the EU, companies are not obliged to scan social media and report AEs. However, if they do scan and find them, then they are obliged to report them to regulators. Regardless, it is mandatory for companies to monitor and report any side effects of drugs reported on their own sponsored websites.

	Drugmakers may have to reconsider their strategies for tracking AEs spontaneously, as the General Data Protection Regulation (GDPR) is set to go into effect in May. As per GDPR, the collection of any personally identifiable information (PII) by pharma would be regulated, provided the companies declare on their websites and social media properties that they are doing social listening and that the information shared by users may be used for this purpose. It is important to note that ethics-committee requirements vary from country to country and it is crucial that they also be examined so as to ensure compliance.

	A good example of an ADR being identified through social media channels was the identification of “Crix belly” syndrome, also known as lipodystrophy syndrome, which results while taking antiretrovirals to treat HIV. It was detected using social media but was not identified during clinical trials, as the study duration was 48 weeks and the side effect manifested after that.

 Patients themselves are screening multiple social media websites to find out more information about ADRs, and are sharing their experiences on sites such as PatientsLikeMe and 23andMe. An FDA-funded study, which analyzed 61,401 tweets, demonstrated that 4,041 (7.2%) could be classified as proto-AEs (posts that resembled AEs), which was three times the amount typically reported to the FDA by patients.

	While the agency has been firm in addressing critical issues such as a balance in reporting benefits and risks of drugs, it also established a research partnership with PatientsLikeMe in June 2015 to help monitor AE reports from patients.

 PatientsLikeMe has also partnered with UCB to create an online epilepsy community to track real-world experiences in dealing with the disease; the platform provides patients with an opportunity to directly report AEs to the FDA.

	Today, 11% of clinical trials are leveraging social media for patient recruitment. Biogen Idec (now Biogen), for example, had been screening an average of six patients per week in its clinical studies. That rate reportedly shot up to 800 patients within two weeks of partnering with MyHealthTeam.

 It was also interesting to note that as per a U.S. study on omni-channel recruitment outreach, apart from the benefit of significantly higher accrual speeds, it was found that the per-patient cost by direct mail was $30, versus 86 cents via the social media route.

 In addition, the opportunity cost associated with a delay per day is significant. With 37% of sites failing to meet recruitment criteria and up to 10% not recruiting a single patient during a trial, the use of social media to rapidly and cost effectively scale up patient recruitment becomes key.

	The Mayo Clinic conducted a pilot study to demonstrate that social media can be effectively used to recruit large, demographically diverse patient groups in a cost-efficient manner, which served as a key message to HCPs.

 A study of 1,516 randomly chosen tweets out of 15,346 that contained the phrase “lung cancer” demonstrated that about 18% of the tweets were about clinical trials. Interestingly, only one of the tweets provided a link to a patient recruitment website.

 With over 320 million monthly active users, Twitter could offer significant potential in driving patient recruitment.

	On the Facebook front, as the world progresses gradually toward virtual clinical trials, VERKKO, Sanofi’s successful Phase IV trial for diabetes that evaluated a wireless blood glucose meter in a remote setting, recruited 60 patients, all online through Facebook. Interestingly, the average patient age was 56, with some patients older than 70; patient satisfaction scores were 4.52 out of 5, indicating that social media is not the domain of only the young.

	Patient centricity continues to be an emerging theme in clinical research-and social media has been a key facilitator in this transformation. The dynamic between the two goes a step beyond reporting of AEs or patient recruitment, but also moves into the crowdsourcing of protocols. The first protocol to be crowdsourced with an investigational new drug approval was for Transparency Life Sciences’ antihypertensive lisinopril.

 Crowdsourcing was also used to assess the use of metformin in men with rising prostate-specific antigen after localized treatment for prostate cancer. Inputs were obtained from 43 physicians and 33 patients using Transparency’s Protocol Builder platform. As patients would provide their feedback based on their own real-life experiences, it is expected that accrual rates would be much higher, since the real needs of the patient would be addressed. Protocols for irritable bowel disease, Parkinson’s disease, and MS are also being developed using crowdsourcing.

	Where does pharma move from here? The biggest challenge the industry faces today is not that the value of social media is not well recognized or that the risk of AEs is deemed too high-it is how to reap the benefits of social media optimally. Clearly, the answer does not lie in listening alone, as most brands experience a steady state after listening for a while and no new incremental insights are observed. The industry’s plain, vanilla social media services are not the answer; instead, the evolving trend is more toward digital services such as 

mobile apps that help track drug adherence, or more complex solutions like Amazon’s Echo, a voice-enabled computer that can recite potentially lifesaving instructions for a user on say, for example, cardiopulmonary resuscitation during an emergency.

	The life sciences industry is also cognizant that today’s informed patients and caregivers are looking for more credible information and better tactics of engagement than the traditional leaflets or, more recently, the occasional tweets. Engagement through online communities, health and fitness devices, and mobile apps is what the industry is looking at today.

 Even traditional social media listening is moving into a new dimension, as pharma companies increasingly explore big data and natural language processing solutions and direct partnerships with forums such as PatientsLikeMe, with the two-pronged objective of (1) driving efficiency in the process and (2) getting real patient viewpoints.

	The power of social media to transform healthcare is substantial. While the pharma industry has long suffered from social anxiety, it is increasingly opening up to the use of social media, and both HCPs and patients are becoming more cognizant of the power the communication tool wields. Watchful steps are being taken in this direction. When GSK recently conducted a search across Facebook and Twitter, the company found 21 million mentions of its products, and this data also resulted in the recall of one of its drugs. GSK worked with Epidemico to filter out irrelevant posts and to ensure compliance with FDA-reporting requirements.

	While one may think that the cost implications are considerable and it would perhaps have been best if GSK would not have mined this data, the reality is that ADRs will be identified at some point of time, resulting potentially in more severe consequences for the manufacturer and significantly more harm for the patient population. Social media thus serves as a powerful enabler for responsible pharma companies that take ownership of patient safety, for patients who want to make informed decisions and share experiences, and for HCPs who want to engage with their patients and drive mutually agreeable decisions that are in the best interest of the patient. True patient centricity is not about ignoring commercial value, but about prioritizing the patient. Social media is the game changer in that equation.

, PhD, is CEO, Nymro Clinical Consulting Services, email: 

is SME, Social Media, Tata Consultancy Services; email: 

	1. 30 Facts and Statistics on Social Media and Healthcare. 

https://getreferralmd.com/2017/01/30-facts-statistics-on-social-media-and-healthcare/

	2. 24 Outstanding Facts and Statistics and Figures on how Social Media has impacted the Healthcare industry. 

https://getreferralmd.com/2013/09/healthcare-social-media-statistics/

	3. Stricker P. Social Media: An Emerging Communication Modality (2014) 

http://www.naylornetwork.com/cmsatoday/articles/index-v2.asp?aid=296436&issueID=30171

	4. Kotenko J. The doctor will see you now: How the Internet and social media are changing healthcare. (2013) 

https://www.digitaltrends.com/social-media/the-internet-and-healthcare/

	5. Castellucci M. Leveraging online platforms to measure how patients feel Modern Healthcare (2017) 

http://www.modernhealthcare.com/article/20170831/NEWS/170839984

	6. Raths D. NQF Turns to Online Communities on Patient-Reported Outcomes. Healthcare Informatics (2017) 

https://www.healthcare-informatics.com/news-item/nqf-turns-online-communities-patient-reported-outcomes

	7. Brunner, G. FDA authorizes 23andMe to market health-related genetic testing. (2015) 

http://www.extremetech.com/extreme/199602-fda-authorizes-23andme-to-market-health-related-genetic-testing

https://ipfs.io/ipfs/QmXoypizjW3WknFiJnKLwHCnL72vedxjQkDDP1mXWo6uco/wiki/23andMe.html

	9. Strickland M. Patients Know Best: A Changemaker Health Case Study. (2017) 

https://medium.com/change-maker/patients-know-best-a-changemaker-health-case-study-2f203b0971ae

	13. Ghinn, D. Pharma gets social: FDA guidance on twitter is irrelevant for global pharma (2013) 

http://www.pharmaphorum.com/articles/pharma-gets-social-fda-guidance-on-twitter-is-irrelevant-for-global-pharma

	14. Mack, J. Solving the Social Media Adverse Event Reporting Problem. Is it just too much ‘noise’ to sift through? (2010) Pharma Marketing News, 9.3, 1- 9. 

http://www.news.pharma-mkting.com/pmn93-article04.pdf

	15. Sukkar, E. Searching social networks to detect adverse reactions. (2015)

http://www.pharmaceutical-journal.com/news-and-analysis/features/searching-social-networks-to-detect-adverse-reactions/20067624.article

	16. Duh MS et al. Learning From Social Media For Adverse Event Reporting. (2017) 

http://www.analysisgroup.com/uploadedfiles/content/insights/publishing/learning_from_social_media.pdf

	17. Global Top 10 Pharma Companies on Social Media. (2014) 

https://www.slideshare.net/simplify360/johnson-johnson-top-pharma-company-in-social-media-gsk-and-pfizer-follow

	18. Krackov, D.A., Davar, M.V., Kailasnath, V. and Gitterman, A. FDA’s New Online Media Draft Guidance Documents. (2014) Arnolds and Porter Advisory, 1 – 15. 

http://www.arnoldporter.com/resources/documents/ADV614FdasNewOnlineMediaDraftGuidanceDocs.pdf

	19. Taylor, N.P. Bayer, Boehringer lead rapid rise in pharma’s social media output. (2015). FierceBiotechIT. 

http://www.fiercebiotechit.com/story/bayer-boehringer-lead-rapid-rise-pharmas-social-media-output/2015-04-27

	20. Perrone, M. Kim Kardashian’s social media posts prompt FDA warning. (2014) 

http://www.azcentral.com/story/money/business/consumer/2015/08/11/fda-warning-kim-kardashian-diclegis/31501143/

	21. Papple D. Up to 90 percent of adverse events go unreported, pharma companies search social media to make drugs safer. (2015)

https://www.inquisitr.com/1914900/up-to-90-percent-of-adverse-reactions-unreported-pharma-companies-search-social-media-to-make-drugs-safer/

	22. Innovative Medicines Initiative WEB-RADR Workshop Report: Mobile Technologies and Social Media as New Tools in Pharmacovigilance (2017) 

http://www.ema.europa.eu/docs/en_GB/document_library/Report/2017/02/WC500221615.pdf

	23. Points to consider about clinical trial data and transparency. (2013) 

https://globalhealthtrials.tghn.org/site_media/media/medialibrary/2014/02/points_to_consider_final-1.pdf

	24. Banks L. Social adverse event reporting – giving a voice to patients. (2015) 

https://pharmaphorum.com/views-and-analysis/monitoring-social-adverse-event-reporting/

	25. Okun, S. Sally Okun explains the new research collaboration with the FDA (2015) 

http://blog.patientslikeme.com/2015/06/16/sally-okun-explains-the-new-research-collaboration-with-the-fda/

	26. Wicks, P., Keinenger, D., Massagli, M.P., et al. Perceived benefits of sharing health data between people with epilepsy on an online platform. 

	27. Rosenberg R. Biogen Idec turns to MyHealthTeam in successful test of social networks for enrolling MS clinical trial (2014) 

https://www.myhealthteams.com/wp-content/uploads/2015/01/cww1827_myHealthTeams.pdf

	28. Patient recruitment: is social media the cure for this headache? 

https://clinicalaccelerator.com/2015/02/17/patient-recruitment-is-social-media-the-cure-for-this-headache/

	29. Donahue M. Patient Recruitment via Social Media: Lessons Learned (2012). 

http://www.pharmexec.com/patient-recruitment-social-media-lessons-learned

	30. Fassbender M Twitter for #Clinical Trials Recruitment? Researchers see the potential. (2016) 

https://www.outsourcing-pharma.com/Article/2016/03/09/Twitter-for-ClinicalTrials-recruitment-Researchers-see-the-potential

	31. Adams B. eClinicalHealth, Sanofi boast of success in their Facebook recruited study. (2016) 

http://www.fiercebiotech.com/cro/eclinicalhealth-sanofi-boast-success-their-facebook-recruited-study

	32. Mansell P. Lisinopril IND is crowdsourcing first, says Transparency Life Sciences (2012) 

http://www.pharmatimes.com/Article/12-12 20/Lisinopril_IND_is_crowdsourcing_first_says_Transparency_Life_Sciences.aspx

Outlining the growing attention and pursuits around social media in adverse event reporting and advancing patient-centric initiatives in clinical trials.   


Nimita Limaye, PhD

Articles

A Roadmap for Implementing RBM and Quality Management

How Social Media is Transforming Pharma and Healthcare