I’m working as a Global Clinical Development Manager at a Pharmaceutical firm based out of Mumbai. As a safety physician, I’m involved in determining safety trends (adverse events) across clinical studies conducted, providing assistance on patient recruitment strategies, analysis of huge medical data of patients with respect to safety (checking side effects between drugs) and efficacy (how well the drug works compared to others). In addition, my duties include assisting in the preparation of statistical analysis of plan (SAP) and doing reconciliation of safety events that occurred during studies, checking data listings and medical coding during trials/programs, leading data interpretation meetings with the preparation of data visualization graphs and drawing inferences and supporting budget forecasting and tracking of clinical studies. I chose to upskill in DSBA to learn how data analytics can help deliver results quickly and help ease working.
The problems faced are understanding safety and efficacy trends at the initial stage so that appropriate measures can be taken, delayed recruitment at some hospital sites, and forecasting of results (positive or negative). Data science can help to analyse trends early, and boost recruitment strategies with help of data visualization and advanced statistics.
I utilised numpy and pandas in python to analyse data quickly. Matplotlib for data visualization is another important factor which helps understand data graphically in a simple way. The key concepts taught during the lectures helped in understanding the problem and the use of appropriate tools for solving problems effectively.
We were working on adverse events of a drug used for treating dyspepsia (indigestion). It can increase levels of a hormone called prolactin which further leads to different changes in the body. Now to understand whether the drug really cause this increased trend was important by looking at the data holistically. Using data visualization strategies, the correlation between two parameters and advanced statistics – all played a role to derive useful insights from the data. I recommend the use of data science analysis software such as python in clinical trials. The concepts are still being tested and evaluated in different areas of clinical studies. But surely this is having an impact on process efficiency by approximately 30-40%, cost reduction in taking critical decisions at the right time and forecasting outcomes with respect to patient recruitment in studies.
This exercise has simplified my work simplified aided in understanding the safety and efficacy trends so that appropriate decisions at an organizational level can be taken.
