Data: The Difference Between Life, Death And Profit?

On February 4^th, 2020, Gilead’s Remdesivir entered clinical trials to treat the coronavirus. The drug was originally developed to combat the Ebola outbreak but is now being deployed for stage 3 trials in cooperation with Chinese authorities. The good news is that the mortality rate of Covid-19 (official name of this disease) appears to be much lower than the SARS and MERS epidemic. In general, infectious disease clinical trial success rates are quite high at about one-in-three. However, time is of the essence, especially as treatment development takes longer than most people think. Drug development requires a great deal of trials and collaboration, in this case even across borders and drastically different political and healthcare systems.

This also hits home for me personally as I recently was made aware by old and new friends about their health and financial struggle to fight with various forms of disease including cancer. Their participation in early-stage trials for rare forms using orphaned drugs, made me look harder at this topic.

Ultimately, time is the precious commodity in these scenarios and any tool we can use to shave off even a day to release a treatment option, not only saves or improves lives but it can even cut cost for the drug company and society as a whole if passed along.

Falling back on my initial research from a few years back, I wanted to understand how the pharmaceutical industry supports clinical research today around trial sites and subjects. Back then, the data quality and misalignment were seriously undercutting processes affecting the time-to-market, success rate and financial outcome of drug trials. I would also posit that other R&D-centric industries suffer these challenges but they can certainly utilize a similar approach in their engineering organizations.

There are many factors in play when a life sciences company decides to start the research and development for a new treatment. To give you an idea about the size and complexity of the challenge, here are some basic facts:

·     As of February 2020, over 330,000 studies were registered including 52,000 just for patient recruitment. Global patient recruitment numbers also fluctuate significantly ever year.

·     A decade ago, roughly 11,000 trials were trying to recruit 2.8 million subjects. However, with trial length being 1.5 to 13 years and the average FDA approval process taking over a decade –  some even longer – pharmaceutical firms need to be selective with their investment.

·     The probability-of-success (POS) is about 12% on average for drugs. As mentioned earlier, it can be as high as 33.4% for infectious disease vaccines and as low as 3.4% for oncology. With FDA approval changes these rates seem to be improving, especially for rare diseases. However, making POS-based investment decisions on top of any acceleration of the process means big money for a pharma company as well as lives improved, sometimes saved, for us all.

·     Consider also that 85% of clinical trials are delayed, 95% of them over a month.

·     The underperformance of trial sites created excess cost for the industry of $2 billion between 2006 and 2012. This problem has not changed substantially as recent data shows a 100-fold cost difference between trials. The median trial cost per trial subject is $41,117 and it takes between ten to sometimes thousands of patients for a single trial depending on stage.

·     Up to 10% of drug development cost are related to patient recruitment. The absolute numbers generated from delays in identification, acquisition and management of clinical trial sites comprised of healthcare providers (physicians) and subjects (patients) are staggering. They more than doubled over five years in the past. Clinical Research Organizations (CROs) and Networks (CRNs) are then used to optimize this administrative burden.

·     Research indicates that a delay in market introduction of a drug with just modest market appeal can cost $500,000 in lost revenue per day. For a block buster it is upwards of $8 million per day. It may sound callous but it factors into a firm’s decision if and when to invest into R&D.

This brings me back to my original point about how pharmaceutical firms are starting to tackle costly and delayed patient and site-related processes.

Finding enough and the right volunteers by just connecting the data crumbs

Clinical trial cost is very much driven by problems with patient recruiting as 70% of patients live more than two hours away from their trial site where screenings and follow-up visits take place. This means that many just don’t show up. Nearly half of all studies suffer from recruitment challenges delaying efforts by six months or more creating an improvement opportunity to save $25 million a month. “Siteless” or “Patient-centric” trials using telemedicine are trying to combat these shortcomings by addressing a typical 30% patient drop-out rate and reduce the initial time to recruit patients for a stage II and III trial from seven to four weeks.

What is surprising to me is that less than 5% of patients with cancer enroll in trials, which certainly contributes to 80% of trials failing recruitment targets? How can this be if 72% of eligible subjects are effectively known as they are within the site’s (clinic, doctor’s office) existing patient population?

There are many important improvement vectors available: regulatory reform, patient compensation improvements, patent protection changes, etc. I argue that an untapped value lever is the way data supporting a clinical trial is created, managed and shared using automated and smart data management techniques integrating supply, operations, patient, consent and protocol databases.

CROs and pharmaceutical firms are already pulling together and harmonizing so called HCO/HCP master repositories (Healthcare Organization – aka physician group or alike, Healthcare Provider – aka physician). This data pot is often initially created to align the sales and marketing effort to drive account-based marketing synergies. However, it can also enable the organization’s R&D division to understand licensing and specializations as well as historical performance of investigators (Clinical Research Associate or CRA) and doctors as it relates to specific trial protocols, diseases and patient populations. If this repository extends to validated subject (aka patients) information; including biomarkers, disease encoding as well as compounds, a more complete picture emerges. It shows potential conflicts-of-interest, a similar study’s close proximity, the risk for adverse reactions, side effects, screening errors and even patient consent omissions.

Serving this single, extended view of a clinical trial via different facets to researchers, compliance staff, financial analysts, 3^rd party investigators and providers is not only more efficient but also effective. As site investigators and their training changes frequently over the course of a trial, experts tell me that the odds of perfect data capture through legacy technology approaches is near zero.

Given data’s impact on key components of a typical cost breakdown of an average $19 million clinical trial (low of $2 million, high of $347 million), utilizing this strategy can create savings of around 12% for a stage III trial. If we use an example, such as the cholesterol-drug Zetia by Merck, which sold for $10 per day and was ranked number 148 in terms of sales early in its lifecycle, we are looking at $171,000 in additional top-line revenue per day gained. Given the average company running at least 50 trials at any point and savings of $1.6 million per trial, the scale becomes clear on the bottom line as well. Not only does it escalate quickly for a single company but across the industry, directly affecting the cost of prescription medication. Which brings me full circle to my original point; as it turns out, data provides financial leverage on top of saving lives.