Making the Most of Drug Development Data
Improving time to market requires finding — and sharing — the right information at the right time. A benchmarking survey reveals that pharma and biopharma still have a long way to go.
By Ken Morris, Ph.D., associate head of Industrial and Physical Pharmacy, Purdue University, with Sam Venugopal, director, Life Science Business Operations, Conformia Software, Inc. and Michael Eckstut, vice president, Life Science Operations, Conformia Software, Inc.
Drug development continues to change dramatically (see Data Use and Misuse
, below). As a result, the amount of information that pharmaceutical companies generate and collect during drug development doubles every five years. Unfortunately, only 10% of this information is ever leveraged to improve overall competitiveness and compliance . We decided to find out why, in a survey (Pharmaceutical Manufacturing,
April 2005, p. 73) designed to assess the current state of information and knowledge management during pharmaceutical and biopharmaceutical process development.
The survey (see Demographics and Process Development Growth Trends
, below) benchmarked each respondent against an ideal, in which business and information processes are characterized by:
- Structured Data
This article summarizes what we found.
In brief, drug development professionals know where they need to go to optimize data management. They also have a clear idea of what they need to perform their jobs better. However, the organizations in which they operate often fail to see the big picture. They don’t integrate or align business and information processes, or create the required to transfer business, science and compliance data across development silos.Data transparency
Since data transparency is the prime prerequisite for knowledge management, the survey first assessed how visible, accessible and traceable information was in each respondent’s organization. Results indicate that most drug development organizations need to get a better handle on the data they generate. As Figure 1 (below) shows, most drug development organizations — regardless of size — are working with ad hoc
or siloed processes, in which data are not being captured consistently. This results in fire-fighting when issues come up, as well as “tribal data,” in which individuals store most of their data on individual desktops or servers.
Data visibility and traceability appear to be limited in cross-functional processes involving multiple groups (e.g., manufacturing, quality, compliance). Figures 2 and 3 (below) summarize survey results.
While a few respondents described no data visibility, most indicated that data were at least somewhat visible to various functional groups within their development organizations, indicating that their companies are at least taking the initial steps required to move beyond data silos.
However, traceability is still limited, as shown in Figure 3 (below). While half of the respondents indicated that information could be traced, this was often done by paper searches and other labor-intensive activities, throughout the organization.
Close to 29% of the respondents could trace selected data back across the development lifecycle, but 21% of the respondents indicated that they could not trace data back at all and not a single respondent indicated that full and complete traceability could be achieved in their environments today.Outside four walls
Clearly, poor data visibility, accessibility and traceability diminish efficiency and performance within any drug development organization. However, the negative impact only multiplies when this type of environment is replicated across an organization’s key alliances and partners.
As pharmaceutical partnerships become increasingly important (see Demographics and Process Development Growth Trends
, below), organizations won’t be able to sustain ad hoc
data transfer. To prevent data from getting lost through the cracks, robust, mutually-agreed-upon data transfer systems are urgently required, supported by strong and flexible processes that will allow organizations to “do more with less.”
The current FDA regulatory framework rewards companies that clearly understand their processes, whether for manufacturing or development. Clearly, if any drug development organization is to demonstrate full understanding of its development processes, it can no longer rely on disparate sets of spreadsheets, documents and siloed applications to analyze data.
Establishing context around every piece of drug development data will be essential to ensuring process understanding and regulatory compliance. Each isolated piece of information will need to be correlated and linked with other pieces of information.
For example, drug product lot data will need to be associated with production equipment and the state of that equipment (e.g., maintenance records). But the same lot data should also be associated with data on intermediates and raw materials, including vendor data and Certificate of Analysis (COA) information, as well as the process steps that were carried out, the analytical results and the operators who were on duty that day in that particular facility.