Observational cohort clinical studies, or patient registries, are common studies that generate data required not only by regulators, but, also payers and prescribers. While patient registries seem simple, a poorly designed one can lead to the collection of poor-quality data that ultimately results in poor quality evidence. In this interview, Ravi Jandhyala, Consultant Pharmaceutical Physician at Medialis and an expert in rare disease patient registries, who is also responsible for clarifying the formal definition of a patient registry¹, will discuss how to overcome clinical trial registry challenges.

Moe Alsumidaie: What is the most difficult challenge of setting up a registry?

Ravi Jandhyala

Ravi Jandhyala: To my mind, there are a number of them, however, the challenge that stands out most relates to data ownership, so who has the ultimate control of the data that goes into the study. We know from our own research and experience that this is a major concern from the investigators’ perspective in the observational clinical study or patient registry setting. In an interventional clinical study, it’s rather cut and dry, in that the sponsor is the data owner-or data controller. However, investigators are clear that they do not want their data ‘stolen’ and withheld from them for their own research in the observational setting.

The second biggest challenge is the data that is to be collected. Essentially, an observational clinical study-of which a patient registry is a type-is limited by the fact that no data can be mandated by the protocol. Therefore, all data must be collected as part of standard diagnostic and monitoring practice for the center. As a result, the data collected versus that which is requested is smaller in volume, and therefore reduces the quality of the overall data set through the missing data. Missing data is a problem across the board; however, it is of greater impact in rare diseases where patient numbers are smaller. The core data set is something I like to call the ‘Holy Grail’ of observational clinical studies. I’m pleased to say that we have addressed this and have developed a reliable approach to develop such a data set; a particularly high-profile example will be published soon.

The final challenge I’d like to convey here is data entry. Again, in contrast to a clinical study, there is not as much of a draw in terms of return on investment for investigators, and therefore the sites don’t spend much time with a patient registry, so there is a need to rely very much on their goodwill for their buy in. This can be supported with appropriate reimbursement for data entry for either an external function or a pre-existing internal researcher at the site.

MA: How did you overcome the challenges with so much hesitation from study sites and investigators?

RJ: First, we really needed to listen to the concerns and understand the background to them. This involved many face-to-face meetings and a fair degree of research into ‘failed’ examples of patient registries. The solution in this case was actually quite straightforward, which was to define an access agreement that accommodated their concerns and assign the data controller, or to use the old-fashioned term ‘data owner’, as the patient. This meant that ownership of the aggregated data set remained with the individual patients and consent was provided for its use under certain predefined circumstances. All these materials required both national regulatory and local ethics approval, and importantly, the consent to use an individual’s data could be withdrawn at any stage. 

The control and the consent allowed each patient's data to be used by the investigator, and then the sponsor, in this aggregated dataset. 

This dataset could then have research conducted on it and, of course, the investigators or steering committee-drawn from all of the participating European countries-would sit on the adjudication committee that would review applications for research to be conducted on the larger aggregated European data set.

They would feel that they could contribute to, or at least have some understanding and control over, what research is then done on the aggravated data set; the process of reviewing these research studies would be transparent, alongside the results, or outcomes, of these applications. 

Once research questions have been submitted and anonymized, aggregated tables provided, any evidence that is generated must be published. It can't be left on a hard drive somewhere. It must be made available as part of the transparency associated with a pharmaceutical company primarily sponsoring this type of research. It must also be made available, i.e. published within a peer review journal, or made available as a report on a company website, within a certain time period.

This meant that the scenario of a sponsor preventing researchers conduction data on the aggregated data set would not arise. Simple.

MA: When you go into designing these studies, how do you decide what data to collect? How do you set up the study designed to create this data?

RJ: Well, we need to look at this from the perspective of the stakeholder whose requirements the evidence must meet. We promote a multi-stakeholder approach to medical planning and Real-World Evidence generation, and have identified the patient, prescriber, payer, regulator and sponsor as relevant stakeholders, each with discrete needs.

Examples on how these needs may differ are as follows:

• Payers are traditionally not so concerned about changes in blood markers but are more focused on improvements in quality of life and burden of disease. 

• Regulators may need reassurance on certain safety aspects, such as cardiovascular risk, and require these to be assessed as post approval safety commitments.

• Prescribers are driven by evidence of effectiveness and safety outside the clinical trial population and the knowledge that their peers are using the medicine and generating Real-World Evidence. 

We spend a significant amount of our time and resource in generating this evidence from expert opinion; we’ll be publishing this later this year.

You mentioned the ‘design’ in your question. Ideally, this should begin with the design of the Real-World Evidence generating medical plan for a medicine. This is a document identifying strategic objectives, critical success factors and aspirational statements that the study is designed to generate the evidence to support. 

The real skill here lies in understanding how to build the market for the medicine and differentiate it either within its class or against current standard of care.

So, going back to your question, the evidence needs must be defined first from a stakeholders’ perspective and then the parameters are selected through desk research and direct stakeholder engagement.

MA: How do you select those stakeholders?

RJ: The stakeholders select themselves. For example, if we are preparing a Real-World Evidence generating medical plan for a rare disease, we will work through the list.

Patient stakeholder engagement will involve identifying any relevant patient organizations, and key personnel within them, who would have the experience to be able to help. They may, on occasion have developed a list of research priorities for their disease with their members and this type of document is invaluable for us in ensuring we generate data that is aligned to it. If this list doesn’t currently exist, we may develop one with the relevant patient organization which will feed into this Real-World Evidence medical plan; indeed, we are currently in the process of developing one of these documents.

Prescribers tend to be identified from the clinical trialists network or their recommendations. With any new launch, centers that treat a certain disease of interest can be identified through desk research, including literature reviews, but also through national and international organizations and their congresses. In terms of engagement of this particular stakeholder group, we do need to work with the other demands on their time, something that does add significantly to delivery timelines.

Payers are even easier to identify, for example NICE in the UK. There will be a clear understanding by the Market Access function within the sponsor as to the routes of approaching each of these in their territory. It is important to remember that reimbursement varies between North America and the constituent countries of the EU.

Regulatory engagement, so the FDA in the US and the EMA in the EU, is part of the initial approval process for any medicine; with recent changes to the EU member states, the MHRA will be exclusively responsible for the UK. Real-World Evidence is playing an increasingly important role at this stage. 

On reflection, the real challenge, is that the approach-this multi-stakeholder approach-is something which is not done well at all across the industry, and I can give you a little background on why. 

Within a pharma company, there are aligned internal functions who engage with these external stakeholders. The regulator is handled by regulatory affairs or pharmacovigilance, the commercial team engage with the prescribers and the market access or health outcomes research teams within the pharma companies tend to engage with the payer stakeholder. The Medical Affairs function-of which I’m a member-are unique, in that our background in treating patients and training in conducting clinical research on patients, including understanding how to write a protocol, a patient information leaflet, informed consent documents and a statistical analysis plan, is central to evidence generation and this multi-stakeholder process. We are able to engage with all stakeholders and are qualified to design and run these observational studies.

Therefore, Medical Affairs is the best placed to sit at the hub of the spoked wheel, supporting the design of these evidence generating strategies. Yet, in reality, these departments-regulatory, market access and commercial-tend to work in silos, and there is not this joined up thinking. 

If that mindset could be changed to a more coordinated and unified planning process, with Medical Affairs in the center, then I think we stand the best chance of providing each one of these stakeholders with the right evidence at the right time. This would avoid wasting time and resource in generating lots of individual items of evidence by the separate functions, that are limited in scope and don’t meet the full range of stakeholder needs efficiently. This sounds very simple, it sounds easy to get people to sit round a table and talk to each other, but it simply doesn’t happen. Changing that mindset is really the biggest hurdle; the take home message is: stop working in silos.

The medical plan, having it centralized, having it conducted through medical affairs, where we have this engaging multi-stakeholder needs assessment, that really is where I'd begin.

Moe Alsumidaie, MBA, MSF, is a thought leader and expert in the application of business analytics toward clinical trials, and Editorial Advisory Board member for and regular contributor to Applied Clinical Trials.

References

1. Jandhyala R, Drug Saf. 2019 Dec;42(12):1515-1516. doi: 10.1007/s40264-019-00862-x