In testing treatments and medications, randomized clinical trials are designed to verify effectiveness and side effects of a treatment. The goal is to determine whether there is a statistically relevant improvement from the treatment compared with no treatment, a placebo, or the current, standard course of treatment. But can we leverage Real World Data to improve insight gained from clinical trials? Can we actively analyze the effectiveness and safety of medical treatments prescribed outside of clinical trials?
The answer to both questions is YES! As discussed in the last article, Wearables as Clinical Devices, clinically accurate data from wearable and embedded devices is already available and may be prevalent in a few years. Now, let’s explore how garnering evidence from Real World Data can improve the practice of medicine and randomized clinical trials.
Randomized Clinical Trials
With clinical trials, the practice of medicine embraced the scientific method for validating treatments. While a tremendous step forward, randomized clinical trials are more complicated and imprecise than typical scientific tests. First, most scientific tests are done on inanimate objects and will not directly impact anyone’s health. Quite the opposite are clinical trials which are testing the impact on participants’ health.
Second, a control group is essential in proving the clinical effectiveness of a treatment. But when the participant is sick and given a placebo, you are not trying to improve their health. The control group which does not get treatment for the illness can pose a serious ethical dilemma. If there is a standard course of treatment, the clinical trial can use the standard course of treatment as a control group to address this issue.
Third, as the clinical trial is trying to assess side effects along with effectiveness, the participants should be representative of the population that will use the medication. But you do not want to endanger children or pregnant women or even women that might become pregnant during the trial. Actually, you have to limit the clinical trial whenever a class of participants might be too risky to be in the trial which reduces the breadth of the population tested. The gender bias has been significant enough that US regulations require clinical trials to address gender as part of the trial.
Fourth, the company benefiting from a positive result for the clinical trial is usually funding the clinical trial. Although there are efforts to remove this bias and ensure the accuracy of the trial, the amount of money at risk with a clinical trial is enormous. The development cost of bringing a treatment from research through clinical trials is millions of dollars. The revenue from the treatment is expected to be many times more than the development cost to provide a return on this investment. With this much money at stake, it is very difficult to avoid conflicts of interest and bias completely.
Evidence Based Medicine
Although clinical trials must overcome these issues, they are the best available evidence for treatment and are used to define clinical standards. Along with scientific research, clinical trials are the evidence that drives Evidence Based Medicine. As it combines the best evidence with clinical expertise and input from the patient, physicians, health systems, and health insurance plans have embraced Evidence Based Medicine to guide clinical decision making. But how can the evidence surrounding health outcomes be improved or expanded? How can research extend past clinical trials and incorporate ongoing diagnosis, treatment plans, and outcomes from the broader population?
Health Outcome or Failure
Currently, health insurance companies and government health plans strive to understand the effectiveness of ongoing treatments to their members. These health plans use health outcomes to assess if the treatment was successful and effective. In chronic illnesses, care management staff may engage with the patient and discuss the impact of the treatment and changes to the treatment on the patient’s health and chronic condition. Unfortunately, most treatments are not followed up by the health plan or clinical staff to determine the effectiveness of the treatment. When was the last time someone followed up with you regarding your treatment? Sometimes a follow-up visit is incorporated for serious conditions or surgeries which allows for confirmation of recovery or adjustment of the course of treatment. Unfortunately, clinicians rarely have time to follow-up on most treatments if you do not return for another visit. Unless the patient is part of a care management program, health plans do not have visibility to the success or failure of the treatment other than another visit to the clinic, pharmacy, or hospital. In reality, the healthcare industry tracks treatment and diagnosis failures not true health outcomes.
Most people are likely to classify their health outcome as successful if they a) returned to their normal activity, b) had no signs of the issue or illness, or c) the symptoms associated with the condition were alleviated due to the treatment. If the treatment was unsuccessful but the patient was able to recover on their own, should the treatment be considered successful? Currently, how does anyone know other than the patient? The clinician is unaware – as is the health plan – when there is no additional visit, expense, or follow-up.
Real World Data and the 21st Century Cures Act
From HHS, the Department of Health and Human Services, Real World Data is information from health monitoring and mobile devices, patient related activities in both outpatient and in-home settings, electronic health records (EHRs), drug and disease registries, and medical claims and billing data. This data can improve and augment typical clinical trial data and, due to the 21st Century Cures Act, it may be used to support regulatory decision making. This regulatory validation by the 21st Century Cures Act should accelerate Real World Data’s use across healthcare
Even more interesting is the use of Real World Data to monitor ongoing treatment of illness post clinical trial. Clinically accurate wearables, health tracking apps, and even the geolocation capabilities of your smartphone become valuable evidence in defining whether the patient has returned to normal health and activity. The Real World Evidence, which is evidence gained from Real World Data, will provide broader insight regarding ongoing diagnosis and treatment of illness. The potential expansion of evidence is immense. With the ability to view and adapt to responses to treatments from the general population, the reaction time of the healthcare system to epidemics and new mutations of illnesses will be much more rapid.
Impact on Care Plans
As we gather electronic health records along with health outcome information for a broader population of patients, the ability to identify impacts of physiological differences on care plans, or treatment plans, will also increase. If the information from the electronic health records includes genetic information along with the patient’s medical history, the ability to define and validate care plans becomes even more precise and personalized. This is the definition of precision medicine or personalized medicine – being able to adapt a care plan to a specific individual based on their physiology and genetic makeup.
As access to true health outcomes, individual physiology, diagnosis, and care plans increases, a virtuous cycle of using Real World Evidence to vet both diagnosis and care plans can emerge. The patient must embrace monitoring their health more closely through manual efforts and automated monitoring. Also, both the clinical team and the patient or caregiver must collaborate to ensure that the follow-up is done and documented. But with this additional insight and access to the information, the clinical team or clinical data analysts can build a feedback loop to continually validate and improve diagnosis and care plans.
The virtuous cycle of improving the accuracy of diagnosis and effectiveness of care plans will reduce mistakes and accelerate recovery. Along with the improvement in the safety and quality of life for patients, these improvements will lower the costs as they remove unnecessary, ineffective, or inappropriate treatments. There are definitely hurdles in addressing patient monitoring and ensuring timing follow-up, but technology will reduce the effort and costs of these activities. The biggest challenge will be analyzing the tremendous amount of data flowing from these patients, clinicians, and events to uncover or infer the key insights that will improve healthcare. To address this challenge, the rapid advancements in big data analysis and machine learning will be crucial.
Insight Not Data
In the next article, the use of big data, machine learning, and artificial intelligence in healthcare will be explored. Check back as “Insight, Not Data” will be published on October 24, 2017.
Join Our Newsletter
If you would like to get updates on our published articles and get our quarterly newsletter, please sign up at Join Our Newsletter.
by Matt Larsen, Principal, Healthscient
Published on Healthscient.com: October 3, 2017