Site icon

Measuring Subjective Outcomes Objectively – How Technology Can Boost Data Quality and Endpoint Reliability

Measuring Subjective Outcomes Objectively - Anthony Everhart

Dr Todd Everhart is Signant Health’s Clinical Vice President, newly appointed to accelerate intelligent eCOA and Predictive Data Analytics in Medical Clinical Research.

He speaks to the Journal of mHealth about the need to improve endpoint quality, and how technology holds the key to collecting meaningful, quality data.

What is endpoint quality and why does it matter?

For a trial to yield meaningful results, the most relevant outcomes need to be measured in the most accurate and consistent manner. Endpoint quality is about ensuring that happens, and it matters because a lot of time and money is being spent on studies that fail.

In psychiatry, for example, clinical trials have a very modest success rate. Just 24% of phase II and 56% of phase III psychiatric studies meet their primary endpoint1, in large part, because it is so difficult to show a drug/placebo separation.

The question we need to ask ourselves, as an industry, is do these medications, which have moved through so much of the drug development pathway, really not work?

Analyses recently carried out by Signant Health’s clinical data experts2,3 have determined that when researchers can identify anomalous data, the root causes are often one of two things – the collection of poor quality data, or poor data collection practices. In fact, the work showed a number of examples where bad data or bad data collection, rather than lack of efficacy, had been the reason for study failure.

Luckily, the intelligent use of today’s array of technological solutions, from electronic clinical outcome assessment (eCOA) to remote monitoring and data analytics, can eradicate both of these issues.

Eliminating poor data collection practices is becoming easier as technology evolves to replace paper scale administration with intelligent electronic systems that prevent missing data and cue investigators to correct data inconsistencies in real time while the subject is still available.  Further elimination of poor quality data is achievable utilising clinically-informed, blinded, evidence-based data analytics that detect patterns of measurement associated with poor trial outcomes and enable rapid remediation of aberrant sites and raters.

Achieving both of these aims improves data quality and endpoint reliability, saving sponosors precious time and money, and getting new drugs to the people who need them as quickly as possible.

What are the components of an endpoint quality solution/program?

Bringing clinical trials into the 21st century by utilising digital health solutions is the key to improving endpoint quality and reaping the rewards of more sophisticated, patient-centric trials.

There’s no silver bullet – rather sponsors need to embrace an integrated approach that combines the advantages of efficient data collection and analysis tools across the development pathway.

eCOA:

There is a growing body of evidence4 to show the systematic monitoring of symptoms as part of routine clinical care can result in significantly better outcomes – it can improve communication, drive satisfaction and ease symptom management5,6. This, combined with regulators and payers shifting towards a value-based approach to licencing and reimbursement, means patient-reported outcome (PRO) measures have never been more important.

Electronic clinical outcome assessment (eCOA) platforms and wearable trackers capture treatment-related improvements that matter to patients, in real time. Not only does this boost engagement but it provides sponsors with the evidence of patient-centred development that regulators are increasingly asking for.

Rather than asking clinicians to perform paper-based assessments and calculations, quality eCOA systems are intelligent data capture devices to identify data integrity issues. It can be set up with mandatory fields and programmed to provide memory-nudging hints and tips on performing the scale at hand, for example.

Rater training and certification:

Sponsors and CROs need to know that every rater understands the endpoints in the same way, and administer rating instruments the same way every time. This ensures like-for-like data is collected across multiple sites, in multiple languages and cultures, no matter who carries out assessments.

Best practice, systematic rater training and certification means sponsors can be sure all investigators understand the scoring principles and are collecting accurate, consistent, reliable data.

Data quality monitoring:

Integrated clinical trial software combined with clinical expertise and sophisticated data analytics can improve endpoint reliability.

Inconsistencies suggesting  data quality issues can be flagged and dealt with straightaway with raters being retrained or, if necessary, removed, before data quality can become a study-breaking problem.

Using blinded data analytics, experienced clinicians can asses the entirety of a site’s data compared to that of the study overall and look for key indicators that may point to data quality concerns.

Where has endpoint quality been successful?

Improving endpoint quality has been shown to be successful in several areas of research, including central nervous system (CNS) and psychiatry.2

These are areas where “soft” endpoints, such as fatigue or cognition, are more commonly measured, meaning that success is reliant on the rater’s ability to record subjective endpoints objectively.

Rather than measure tumour response, as is traditional in oncology, for example, these studies ask investigators to assess the severity of a patient’s schizophrenia, Alzheimer’s or Parkinson’s disease, and that relies on their ability to assess the disease and score it accurately.

Are there other therapeutic areas which would lend themselves to similar endpoint quality solutions?

Any field that has difficulty showing a drug/placebo separation and relies on rater-driven endpoints and patient reported outcomes stands to benefit from utilising endpoint quality solutions. Immune mediated diseases, such as rheumatoid arthritis and inflammatory bowel disease fit into this category.  Dermatology, which can be very reliant on rater-driven data, is another interesting area that could benefit from this approach.

Oncology developers are also prime candidates for endpoint quality improvements. Huge leaps forward in medical understanding mean more people are living longer after a cancer diagnosis than ever before, and today’s drugs need to offer tangible quality of life benefits that can only be measured through “soft” endpoints.

References

  1. Clinical Development Success Rates 2006-2015 – BIO, Biomedtracker, Amplion 2016.pdfhttps://www.bio.org/sites/default/files/Clinical%20Development%20Success%20Rates202006-2015%20-%20BIO,%20Biomedtracker,%20Amplion%202016.pdf Accessed April 9, 2019.
  2. Kott, Alan; Lee, Jeniffer; Forbes, Andy; Pfister, Stephanie; Ouyang, John; Wang, Xingmei, Daniel, David. (2017): Effect of High Within Subject Variance at Research Sites on Placebo Response and Drug Placebo Separation in Two Acute Schizophrenia Trials – A Post Hoc Analysis. Poster presentation. International Society for CNS Clinical Trials and Methodology (ISCTM). Paris, France, 9/1/2017.
  3. Daniel, David; Wang, Xingmei; Sachs, Gary; Kott, Alan (2017): Association of Within Person Variance with Data Quality Issues in Schizophrenia Clinical Trials. Poster Presentation. ACNP 56th Annual Meeting. Palm Springs, California, USA, 12/3/2017.
  4. Stepanski, Edward. (2016): Incorporating Routine Patient-Reported Outcomes Assessment Into Cancer Care: Building Momentum. https://www.journalofclinicalpathways.com/article/incorporating-routine-patient-reported-outcomes-assessment-cancer-care-building-momentum
  5. Feinberg, Bruce; Braverman, Julia; Fillman, Jennifer (2016) The Pivot Toward Patient-Centeredness in Medicine and Oncology. https://www.journalofclinicalpathways.com/article/pivot-toward-patient-centeredness-medicine-and-oncology
  6. Bacsh E, et al. (2014) Patient-reported outcome performance measures in oncology. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5527827/

About Dr Todd Everhart

Dr Todd Everhart, MD, FACP, is the internal medicine leader at Signant Health. Dr. Everhart is board-certified in internal medicine and a fellow of the American College of Physicians with over 23 years of experience in the practice of medicine and over 12 years of experience in clinical development. Prior to joining Signant, Dr. Everhart held positions of Vice President, Medical Affairs and Vice President, Medical Informatics at Chiltern and Covance, and consulted independently in the areas of medical monitoring, medical data review, data analytics, and physician adoption of technology. He has worked in all phases of clinical development in numerous therapeutic areas and is one of Signant Health’s endpoint quality experts specializing in rater training, placebo response mitigation, and blinded data analytics.

Exit mobile version