Can Payers Trust Our Data?
Analyses from RWD are playing an increasingly important role, but will payers accept pharma’s figures?
The potential for real-world data to be used on a mass scale to demonstrate the effectiveness and value of medicines is great, promising benefits for all parties including better health outcomes and lower premiums for patients and higher prices for those selling the most effective treatments.
Yet, while the use of real-world evidence is growing, a major impediment to its broader and faster adoption is the potential trust gap. After all, how can payers be sure that pharma’s data tell a true and accurate story about patient outcomes?
Payers are increasingly questioning the data they are provided with, says Michelle Hoiseth, Corporate VP at PAREXEL International, asking providers on what basis it is trustworthy and suitable.
In order to meet the test of trustworthiness, we need to evaluate contradictions or conflicts in the data, whether the data are collected with integrity, the timing and the context of the original data collection, whether it is possible to change the data between its origin and the source being used for the analysis, whether data are coming from clinician sources or other sources, if data used are in unstructured fields and, if so, how they were extracted.
“All those kinds of variables come into play,” says Hoiseth.
There are several reasons for their lack of faith in the analysis from pharma companies not least among them is the payers’ perception that there is a gap between what’s being reported and what’s being withheld when reporting a result, says Aaron Mitchell, Managing Principal and Global Lead for R&D Excellence Practice at ZS.
“Some outcomes can be selectively revealed or withheld. There have been numerous studies demonstrating a reporting bias within healthcare, where there is a bias toward publishing statistically positive results favoring an experimental treatment.”
As a result, payers, who may lack the capabilities to assess the evidence, fear data may have been mined selectively in ways that support a particular claim, says Mitchell. “We don’t know the prevalence of this but it is at the back of payers’ minds. My view is it that is happens infrequently but payers don’t know when it’s happening and when it’s not.”
Part of the reason is that real-world evidence is still an emerging field and payers are still developing their own understanding of it and so may not fully value it. In one 2016 study of a group of US payers, it emerged that only 5% of P&T monographs referenced already published real-world evidence studies, says Mitchell. “They are not used to using real-world evidence yet and even then they have to be convinced the evidence is accurate.”
The lack of a commonly accepted framework for assessing when this type of evidence is sufficiently reliable is another significant stumbling block, says Mitchell.
Whether or not a payer trusts the data being presented to them depends in part on how it is laid out, says Joseph Dye, head of HEOR, Neurology US at UCB.
There is an expectation among payers that providers of real-world evidence will have done appropriate due diligence on the data but they will not trust a report if the provider can’t answer the right questions.
“The ones that build the story have much greater impact than those just laying out the results. As a researcher, if I see a slide that makes an HEOR claim and does not tell me the data source, or if it doesn’t tell me what kind of research design or analysis was used, then I am automatically suspicious. On the other hand if I see someone putting together evidence that uses a fair and reliable method, then I feel like I can trust those results. If you lay out a compelling story for why you are presenting this information to them they become more receptive," says Dye.
Another issue effecting their trust in real-world evidence is their capacity to do their own analysis at scale. While the larger payers are starting to build their own ability to do due diligence on real-world evidence, as a group they generally still lack capacity, says Mitchell.
“The challenge is that you have so many therapies in so many areas that you can’t go through every area to evaluate every study or run enough of your own analysis. In this respect it is less of a capability issue than a scale issue. So it happens when it is important. If it is a therapeutic area that makes up a significant portion of the budget they will do the analysis.”
This lack of analytical bandwidth can end with payers making partly or largely subjective decisions rather than basing them on their own data, says Mitchell. “If I don’t have the capacity or clear standards for how to evaluate a study, I have to make decisions based on my own perceptions."
This will become a growing problem as healthcare data continues to grow exponentially and the requirement to find, manage and analyze information grows with it. Things are changing, however, and the bigger payers are now building their capabilities in assessing data for themselves and are assembling teams of highly skilled researchers, says Dye.
“They are gathering teams of people at masters and PhD level in pharmacoeconomics, HEOR, epidemiology as well as biostatisticians. They are starting to pull those kinds of people in to improve their decision making.”
Payers are also increasingly forming partnerships to provide analysis. Medical and healthcare system partnerships are developing along these lines and the large integrated delivery networks (IDNs), such as Boston’s Partners HealthCare, have their own academic research teams to call on, says Dye.
“I’m hearing of it more often now. Where some organizations don’t have their own internal capacity, if they affiliate with an academic center there is more opportunity to collaborate on research projects or access data to answer specific questions for the payer.
“I think that it is going to become more advantageous for those large IDNs to affiliate with schools and centers of public health. There is a big opportunity there to leverage data to better answer the important questions.”
The path towards a greater degree of trust in the data lies in trying to move away from the adversarial nature of many of these interactions, says Sean McElligott, Director, Global Market Access Lead Dermatology, at Janssen.
Rather than each party gathering and conducting their own analysis, it could be done in a much more constructive way by pooling their capabilities in these areas. “In terms of capabilities at the micro level-every time I have worked with an insurance company they have strong capabilities and it would be mutually beneficial for a more explicit partnership.”
He continues: “The way data is being used reflects that adversarial relationship when we should all be rowing in the same direction. We have to align our incentives so that we move towards the end goal of us trying to help patients live better lives. Data is the boat that will get us to that shore.
“Our shared objective is to make sure people suffering with disease get the drugs they need and do that in cost-efficient manner by using the data and our ability to analyze it. If we could bring our analytical capabilities together it would be jointly beneficial,” says McElligott.
Despite the doubts that may exist, closer collaboration between payers and pharma can only benefit both parties when it comes to RWE, says Mitchell. “There is way more payers and pharma could gain by collaborating using real world evidence than would be lost. I don’t think mistrust is warranted. There is a huge opportunity for both parties.
“If we build that trust it could be huge opportunity for us to improve the efficiency of care, to align on appropriate uses of therapies in specific patient populations, to align economic incentives to outcomes, to expand the utilization of theories based on real-world evidence, to expand the use of therapies in new patient populations in a way that will be cost effective for payers and also meet patient needs and outcomes. If we can overcome the trust gap and improve health there is a lot to be gained.”
Another important but seldom-included source of information is from the patients themselves. Including the patient perspective would add valuable insights into the real-world effectiveness of particular drugs, and so would build trust in the claimed outcomes, says Dye.
“The other thing missing from claims data and medical records is the patient’s voice. We often have no idea from these data sources if the patient is satisfied with the care they are receiving and if they are able to live out the best possible quality of life that is possible for them. If you don’t measure those end points then you are going to have a hard time backing up the claim that you are a patient-centered company.
“My aspiration is that we can include the patient perspective in the design of the study and in the measured outcomes using meaningful patient-reported outcomes so that it can impact decision-making. I don’t think we’re there yet.”