Sorry, you need to enable JavaScript to visit this website.

Transforming Clinical Trials with advanced Data Sciences

Transforming Clinical Trials with advanced Data Sciences
May 25, 2017

The landscape as it stands and trends

Data Sciences is an interdisciplinary approach that helps analyze voluminous data sectors, generated by digital sources, and thereby deriving actionable insights. While Data Sciences has been successfully deployed across multiple industries, pharma is gradually picking up on its transformative potential.

Pharma is undergoing a paradigm shift in the way clinical research is conducted, studies are designed, data are collected, care is provided and outcomes are achieved. The industry is looking for ways to leverage digital transformation to conduct trials and make key decisions faster. Clinical Data Analytics would play a significant role in turning new and existing data into insights. The clinical data analytics market is growing at a CAGR of 44.15% and will be worth $12.26 billion by 2019.

Pharma industry has various data sources at its disposal to glean information from – historical trials and research data, site performance data, investigator/key opinion leader (KOL) data, lab data, biographic and biometric data, patient behavior reports, social networking data, trial operations data, and external data sources. The ability to gauge outcomes and trial experiences accurately and incorporate the same into clinical trial design is a key factor in realizing patient centricity.

Let us look at some of the major trends impacting clinical research –

  • The expenses relating to patient recruitment and retention form a major share of trial costs.
  • Despite CRO’s best efforts, patient enrollment has a delayed start of around 85% for all clinical testing on humans. This, consequently, results in loss of sponsor revenue to the tune of $40 million per month.
  • According to a survey conducted to find out obstacles to recruitment – 81% of respondents cited ineligibility, 67% attributed it to lack of time, 66% blamed it on consent forms, and 60% held responsible protocol requirements.
  • Between 2003 and 2013, there was a 56% drop in patient retention rates during clinical trials, resulting in additional cost escalation implications.
  • Patient and site monitoring costs could be optimized through data driven insights; a holistic risk based monitoring (RBM) approach could result in 40-45% cost benefits.

It is, therefore, evident that recruitment and retention is a major challenge, leading to significant delays in clinical trials. According to statistics, 11% of the sites in a multi-center study fail to recruit even a single person. A pharmaceutical company, during phase three of a trial, can register losses of over $37,000 in operational costs and $1.1 million in opportunity costs due to a single day of delay.

Sponsors and CROs are currently looking to capitalize on vast volume of patient data available (Lab, EMR/EHR, Social Media, Syndicated and Claims Data) in order to address enrollment challenges. They have adopted an innovative design-thinking approach and identified certain key areas, like integrating patient experiences in trial designs, to identify shortcomings in the enrollment process along with how to overcome them.

Data Sciences – and the immense potential in clinical trials

As an example, a strategy devised by leveraging Data Sciences can help in a rare disease trial involving a small patient base. In such a scenario, sustaining an adequate sample size throughout the duration of the tests will be a major problem. However, Data Sciences helps in identifying, retaining, tele-monitoring, remote data-gathering, designing, and maintaining an updated record of the subjects. Further, by deploying modelling and simulation tools to support scenarios, as well as crystallize advanced insights, rare disease focused programs can be accelerated.

Moreover, Data Sciences can help curate data obtained from wearables and sensors to identify patients for common conditions despite variations within a (relatively large) sample size. The information, thus, gathered will help in early detection and diagnosis.

Unstructured data proliferates as much as its structured counterpart, largely due to social media platforms. Every individual leaves a social media footprint, and these channels provide a great opportunity to assess the impact of an outreach programs in real time and rectify any inherent shortcomings. Additionally, Artificial Intelligence (AI) and natural language processing (NLP) techniques can be used to decipher the sentiments of patients and physicians present on a clinical trial site in order to derive hitherto inaccessible insights.

Adversities and Advantages – future capabilities    

In today’s dynamic environment, business processes driven by data are becoming the new norm. As the cost of generation, collection, and storage of data falls, there is greater emphasis on propelling analytics and obtaining business insights.

In fact, most pharma companies have established dedicated units across the value chain to focus on Data Sciences However, certain challenges to implementation of an effective Data Sciences program continue to impede the transition process:

  • Managing bulk volumes, data stream and associated noises
  • Approaching Data Sciences with a problem solving mindset
  • Choosing the right set of tools
  • Employing the right people
  • Ensuring data integrity
  • Enabling data sharing and collaboration

We at HCL offer ‘Data Sciences as a Service’ and are working with various clients in advisory and implementation of clinical analytics driven solutions in solving client specific needs.

You can read more about this in our whitepaper on “Data science as a service” - /white-papers/life-sciences-and-healthcare/data-science-clinical-research