The persistent lack of diversity among clinical trial participants, and the resulting lack of data on multi-ethnic patient bases, requires us to consider the implications of this situation before we can address possible solutions.
It may seem obvious, but it is worth stating: when segments of the population are not included in clinical trials at levels that do not meet statistical significance, it is impossible to determine if the therapy in question will work equally, better, or worse as for the population in which it is studied.
We simply don’t know these answers until the therapy has been on the market. Side effects, efficacy and safety become better appreciated and understood only after the therapy has been on the market for some time, given that FDA approvals allow for wider prescribing that is no longer limited to the population represented in the data. Even then, adverse event reporting is unwieldy and most likely does not fully capture how medications affect larger communities not represented in the data.
The article, Racial and Ethnic Differences in Response to Medicines: Towards Individualized Pharmaceutical Treatment, states, “Pharmacogenetic research in the past few decades has uncovered significant differences among racial and ethnic groups in the metabolism, clinical effectiveness, and side-effect profiles of many clinically important drugs.” While it is important to acknowledge these differences, they don’t become apparent often until these negative effects may have already caused damage.
The consequences could lead to poorer outcomes across several aspects, including poorly controlled symptoms, further disease progression, unwanted and dangerous side effects, more doctor visits, more time away from work, increased out of pocket costs, and additional hospital stays. Those are the calculable effects and don’t include the mental and emotional toll on patients and their families.
An example of the consequences of this persistent lack of data is the anti-HIV drug, efavirenz. It is metabolized more slowly among people of African descent. Since the numbers of African Americans in the study were too few to be able to discover this, the metabolism difference led to resistance issues for many people taking the drug. Had more data been collected at the outset, it is possible that this particular complication could have been avoided for many people.
So, let’s take a hypothetical situation. Let’s suppose that U.S. Hispanics account for 25 percent of all diabetes patients in the country. That amounts to 7,500,000 people. Now, let’s say that there is a new drug, DiabetesControl (a fictional name), and roughly 1 percent – 2 percent of the U.S. clinical trial participants were Hispanics. If the trial included 2,000 people, at 2 percent (which would be consistent with current trends) only 40 Hispanics would have been included. This number does not meet statistical significance to tease out any pharmacokinetic or pharmacogenetic differences.
Now let’s look a little more closely at the U.S. Hispanic diabetes population. There is quite a bit of evidence showing that Hispanics with diabetes tend to be diagnosed at younger ages, yet with more advanced disease progression. A1C levels also tend to be higher at diagnosis than for their non-Hispanic counterparts.
Diabetes is a challenging condition to manage, with many variables that must be considered, but let’s suppose that DiabetesControl simply does not perform very well in Hispanics due to a metabolic issue that did not present itself for the vast majority of clinical trial participants, and therefore did not raise any red flags requiring special prescribing information in the label, nor did it prevent the drug’s approval.
Now, let’s suppose that of the 7.5 million Hispanics with diabetes, 10 percent (750,000) have been prescribed and are taking this new drug. Let’s say that 10 percent of this group (75,000) have all been hospitalized at least once for diabetes-related issues and let’s say 10 percent of this group (7,500) have been re-hospitalized several times within a year. Yet, since no one really knows that there is an issue with the drug, no one is looking for this to be a confounding factor in the health of these individuals.
Which bring us to the public health issue. Since I hope we can agree that the goal of treatment is to keep people as healthy as possible and out of the hospital, and that each hospitalization costs the whole health care system a great deal of money – we now have a growing problem with escalating costs, especially in light of the growing diabetes epidemic. In fact, according a recently published study, the American Diabetes Association reports that in 2017, diabetes care cost the U.S. health care system $327 billion, a 26 percent increase from 2012. This figure includes $90 billion in lost productivity.
These costs also included hospital care (30 percent), prescription medications to treat complications (30 percent), prescription medications to treat diabetes (15 percent), and physician office visits (13 percent).
The Widespread Paucity of Data
So now let’s look at some more numbers. Let’s say that of $327 billion annual health care spend, $87.75 billion is spent among Hispanic diabetes patients (which represents the hypothetical 25 percent Hispanic group). If 30 percent is for hospital care alone, then $24.525 billion is spent annually among this group. Using the above scenario, then $245.250 million might have been saved had the metabolic issue with DiabetesControl had been known. That might not seem like a lot of money, but remember, that is simply in hospital care and does not account for lost productivity, or the medications to manage the complications of the condition.
Moreover, what if there were data that supported a more effective medication that could have been used with this group and might have avoided the need for additional medications, and doctor visits? I wonder how many complications could have been avoided, to say nothing of the out-of-pocket costs and the health and well-being of these hypothetical patients had there been more information available.
We are using diabetes only as an example, but let’s remember that the clinical trial participation rate among all minorities in the U.S. ranges from approximately 1 percent – 5 percent and is consistent across almost all disease states. This level of participation, or more accurately, lack of participation, has remained constant over the decades since ethnicity data has been captured in clinical trial data.
Curiously, it is only until recently that the FDA has begun publicly posting the percentages of ethnicities represented in clinical trials. This helps in understanding the widespread paucity of data collected but does not address why this situation persists. While there are some historical reasons for the lack of participation among non-white communities, we must move beyond them and start doing things differently if we are going to make an impact in who is included in the science that should benefit everyone.
Taking the Trial to the Participants
Clinical trials are complicated. They are costly. They involve many stakeholders, including consumers, sponsors, health care systems, physicians, nurses, labs – the list is endless. Yet, there are three sectors we can discuss to change this persistent dynamic:
1) Clinical Trial Design
2) Investigator Training
3) Consumer Education
Let’s start with clinical trial design and the logistics of site location. I recognize that designing clinical trials is no easy task, however, implementing simple logistics that would make it easier for participants to join and remain in studies would not change the scientific parameters, but might very well increase participation rates for those who could benefit. Let’s face it, all stakeholders benefit when a trial is fully enrolled, when patients remain in the study and when data is representative of all groups who suffer from the condition.
What can we do? Well, imagine using existing physical infrastructure to conduct patient visits, such as drug stores where they have express clinics, or even the YMCA, which is in many communities across the country with easy neighborhood access. Since many clinical trials tend to be in cities where there are research universities, there is nothing to prevent those drugstore chains and universities from coming to an agreement to use those physical spaces for patient visits.
There are very good reasons why many clinical trials are conducted at large research institutions, including the rigorous and excellent work they do. However, most people don’t live near research institutions. Even if they do, they often can’t take time off from work to go to clinical trial visits. So why not make clinic hours more convenient? This is where using sites like drug stores, which are open later and on weekends, is a good option.
Perhaps sponsors and investigators could build protocols that would facilitate clinical trial staff scheduling study visits two or three days a week during hours that accommodate participants’ work schedules at these convenient locations where people are already comfortable going and that are closer to their homes and work places. Granted, depending on the trial, there may be certain visits that require a participant to go to where the investigators are, but I imagine that much could be done by taking the trial to the participants instead of making the participants come the trial.
Understanding the Cultural Nuances
Before we look at some persistent myths about the perception of minority patients, let’s address an often overlooked factor. One of the reasons minorities do not take part in clinical trials is simply because they are not asked. Part of the reason they are not asked is because assumptions may be made about their willingness to participate based on their ethnicity, language spoken, and/or cultural background.
Yet, when patients trust their doctor, understand what clinical trials are and how they or their families might benefit, most are open to the discussion, and their ethnicity/language/culture is not a barrier. Physicians and nurses may need more training to understand different cultural nuances that affect patients’ approach to health care, yet these relatively minor issues should not be a barrier to greater inclusion of trial participants. We need to and can encourage these health care professionals to have those conversations with their patients so that they appropriately ask patients to join or refer them to trials.
Finally, let’s look at how we can bring more understanding to consumers about clinical trials and the important role they play in science and drug development. This is where public health campaigns can make a big difference.
I remember as a young child the anti-smoking campaign in California. They were very effective and combined with legislation and continued initiatives to keep people smoke-free, they helped make California one of the states with a smoking rate 11.7 percent for adults, compared to the ~17 percent national average.
There have been public health campaigns for TB, spitting, hand-washing, tooth brushing, getting rid of flies, vaccines, and forest fire prevention (I still remember Smokey the Bear) – the list is long. We need a public health campaign for clinical trial participation. We cannot expect the sponsors of clinical trials to do this alone since the problem is bigger than any one pharmaceutical company or research university can address.
Given how much money is spent annually on health care, our nation’s health is at stake. Knowing that the population has changed from mostly Anglo to multi-cultural, we can no longer pretend that we can gather information from the same group and expect these new and often life-saving therapies to perform the same among all people.
If we continue doing nothing to change who is included in clinical trials, the disparities of health that already plague this nation will continue, will grow, and will increase costs beyond capacity. We have made great strides in health care and with concerted efforts across public and private institutions, we can create a more inclusive ecosystem where the art and science of clinical trials benefit the lives of all people.