Digital innovation has been revolutionizing clinical trials and has made virtual trials possible. A critical element and driver has been the ability of those in the industry to extract insights from both structured and unstructured data. In its essence, there is immense data generated throughout clinical trials to identify drug efficacy and safety.
Advancements in data management and analytics have extended the possibilities in data science by utilizing wider data sets, real world data (RWD), operational data, and technical data to power-efficient and cost-effective clinical trials.
Data Services at Navitas Life Sciences
Navitas Data Sciences provides complete support or augments existing teams. Our service structure and Functional Service Provider (FSP) model provides the required flexibility for best-fit support that aligns with specific study requirements.
Gain Key Data Insights Using Our Statistical Support
We will work with you to develop:
- Primary and Secondary Objectives
- Sample Size and Power Estimation
- Randomization and Blinding of Sponsor Team
- Global Definitions, Conventions, and Analysis Windows
- Rules for Handling Missing Data
- Definitions of Populations: Efficacy-evaluable, Intention-to-Treat (ITT), Safety and Other Subgroups
- Interim Statistical Analysis
- Demographics and Baseline Characteristics
- Subject Disposition and Compliance to Study Treatment
- Method for Analysis of Efficacy Endpoints
- Evaluation of Safety Parameter
Here is an excerpt of an interview with our Chief Operating Officer, Navitas Data Sciences, Paul Gilbert, who has shared his personal experiences and the complexity in clinical data services.
Could you give us some background about your career and how you came to be involved in data sciences?
I studied Chemistry at SUNY Albany and eventually went on to work as a bench chemist at Sterling Drug, a global pharmaceutical company in Albany, New York. There, I started doing computer programming to analyze test results. Eventually, I realized that I was more interested in programming than in chemistry. So, I accepted a Data Manager position in the data management department, where the jobs of data management and SAS-based statistical programming were combined. That is where I started my data sciences journey.
How does data services support clinical trials? What is the role of a clinical data FSP?
Our role is to assist in gathering clinical study data, organization of data in SDTM and ADaM, perform analysis, present the study results and generate the regulatory submission package. These tasks are performed by Data Managers, Statistical Programmers and Biostatisticians.
Biostatistics is more about the final analysis and presentation of the data. It is also important for designing the clinical study. This is, of course, a very simplistic view and adopting an FSP model is different from opting for full service. In utilizing the FSP model, we augment the client’s existing team, directly supporting their business.
How would you assess the importance of a ‘uniform approach’ in dealing with missing data in terms of ‘correctly’ interpreting trial data and the impact of misinterpreted data?
The approach to missing data is usually defined in the analysis for each study report, integrated efficacy, or integrated safety reports. This usually results in a uniform approach to missing data within a regulatory submission. A uniform approach can lead to a quicker delivery of the study report and a faster regulatory review.
How has data science evolved over the years?
There are two main points. The first is technology. Computer hardware continues to become much faster and storage is less expensive. New software and applications such as EDC, statistical computing environments, metadata management and define.xml generation have streamlined data management and statistical programming. Cloud computing has greatly reduced cost. The second is the implementation of standards in the industry. Standards such as ICH technical requirements, CDISC eCRF, CDASH, SDTM, and ADaM; and the acceptance of these standards by regulatory agencies.
What are some of the highlights in your career in the clinical trial industry?
In 1983, my data sciences career began at Sterling Drug in New York. The company was purchased by a drug company in Pennsylvania in 1989. The move to Pennsylvania was a big turning point because it was essentially moving to a new company and a new job. Another big turning point was fast forward a few years to 1993 when a good friend, Matthew Ferdock, and I decided to start DataCeutics, which is now a part of Navitas Life Sciences.
What is the role played by bodies like CDISC play in the industry ecosystem?
Before Clinical Data Interchange Standards Consortium (CDISC) standards were developed, every drug company submitted data in their own format. Sometimes the regulatory agency review team needed to learn each drug company’s data standards. There were also situations where the regulatory agency review team would provide a favored data format causing additional data formatting. This was very inefficient and contributed to longer data preparation and regulatory review times. CDISC CDASH, SDTM and ADaM standards provide pharmaceutical and biotechnology companies a target for collecting data, presenting source data and presenting analysis data files. They ultimately streamline the regulatory review.
Would you say companies are struggling with this or is it the norm?
Generation of SDTM and ADaM has become the norm for the big and mid-sized companies alike, across both pharmaceutical and biotechnology companies. It can be a challenge in small companies, largely due to the limited number of submissions that they carry out.
What most inspires you about working in this field?
Firstly, we are doing something good for humanity throughout the world. We are supporting the bringing of new drugs to the market. Secondly, building a company where the culture is built around a work-life balance. This goes back to when Matthew Ferdock and I started the company, we decided to build a company of experienced people who were committed, self-sufficient and did not require supervision. We began as a remote company, and probably one of the first remote companies in the pharmaceutical industry.
We started the remote model in 1993, which we continue to this day. We have grown and built our team in India around the same model, bringing in very experienced and dedicated data science experts.
How is Navitas Data Sciences unique as a clinical data FSP ?
We focus on bringing in very experienced people, so in the US, that means a minimum of 8 years of experience with an average of 19 years. In India, we are looking at a minimum of 5 years of experience, with an average of 9 years. Everyone is a remote worker, so we are able to select the best people from all over the country. We have a rigorous interviewing and testing policy to select the right person for the position. New hires are assigned a mentor for 3-6 months. We try to promote from within. We are proud to maintain a staff retention rate of 95 percent, so we know that we are doing something right.
Can you describe how Navitas Data Sciences have nurtured relationships with clients over decades?
We are always transparent with our clients and give them the facts. More so, building a long-term relationship is about doing consistently excellent work and delivering on time. Not just good work, but excellent work. Our primary focus is always presenting our clients with quality deliverables.
What are your personal values that help you be a better leader?
I always try to be fair, pleasant and look at the positive side. I believe in giving our staff the responsibility for their work and their roles in the organization.
Do you have a message for aspiring data scientists?
Enjoy your job and do your best!