Clinical Trial Data Analysis – Does A Globally Distributed Model Through Outsourcing and Offshoring Help Biopharmaceutical Companies and How?

Dr. Chitra Lele Ph.D.

Over the past decade BPOs, KPOs, SPOs and CROs have been instrumental in addressing industry challenges by providing clinical development services such as clinical operations, clinical data management, pharmacovigilance, regulatory operations, statistical programming and analysis and medical writing1. Amongst these, the domain of statistical programming and analysis, or biometrics, requires a niche skill for service delivery. But a global delivery model for this activity also has distinct advantages if offshore delivery is effectively managed.

Statistical activities such as study design, preparing the analysis plan etc. are still largely done onsite or near-shore, with internal resources or consultants (unless the entire clinical trial or the entire clinical development program is outsourced to a CRO, end-to-end). This is largely due to the preference for having the statisticians co-located with the clinical development teams for close and frequent interactions, and also due to the paucity of resources with the required level of skill and experience available in emerging markets. However, some amount of offshoring of statistical activities does happen. It is mostly through captive units run by pharmaceutical companies in locations such as China and India, and majority of it is of low complexity. There are just a few service providers with the required expertise and experience to deliver complex statistical services from remote locations and they are able to add value.

On the other hand, considerations for clinical programming are different. Due to the maturity of outsourcing that the IT industry had attained by the start of the century in countries like India, pharmaceutical companies were keen to use the available infrastructure and resource pool to get their clinical trial programming work done in a cost-effective manner through a globally distributed model. Despite the emergence of data standards and advances in software, clinical trial programming is still a resource-intensive activity, especially with the focus on validation and quality control checks at every step of the process. When critical decisions need to be made for the clinical development program, time is of the essence in analysis and reporting of clinical trial data and so is the accuracy of the results. Off-shore delivery teams help meet these demands through access to a large talent pool and follow-the-sun model.

Whilst offering increasingly good infrastructure in emerging markets, service providers can offer niche skill sets (competent SAS programmers) and computational capabilities together with better cost structures. Within the scope of programming, high complexity work such as statistical programming (for efficacy analysis) is more challenging to deliver from an emerging geography than low complexity work related to database programming, safety reporting or data mapping. There are examples in the industry where such a global approach hasn’t worked out as planned, from a combined cost-and-quality perspective, despite all the maturity and experience of global execution. Or, the sponsor organizations have had to invest heavily, and a lot more than planned, to achieve a minimum acceptable level of success. These examples are not only limited to situations where remote teams are provided by vendors, but also where remote teams are the biopharmaceutical company’s captive units, and these experiences haven’t been limited to one particular country. This underscores the need to apply criteria during selection of vendors or planning offshore captive units, which are beyond the standard ones that were used in case of more process-driven activities like Clinical Data Management (CDM) or safety programming/reporting. An important learning from these experiences, whether it’s the captive unit or a vendor unit delivering programming services from cost-effective locations, is that placement of work should be largely guided by the complexity of the tasks, a realistic evaluation of the skill-set available at the remote locations and the time it will take to build the required capacity of trained and competent resources. Also, regardless of the type of programming work that is done offshore, it is important to understand how it should be sustained and optimized in order to successfully achieve the desired results. This deserves a separate discussion.


  1. Khan, I. (2014). ‘Competent Statistical Programmer: Need of Business Process Outsourcing Industry’. Perspectives in Clinical Research, 5(3), pp. 95-100