Shahid Manzur is principal, Consulting, EY, shahid.manzur@ey.com

Shahid Manzur

On April 20, 2021, FDA released redacted details from an inspection at a manufacturing site producing the COVID-19 vaccine (1). The inspection report indicated a failure to investigate a deviation that was already recorded in the system, but not thoroughly investigated.

Inadequate investigation and lack of root-cause analysis tends to be a common observation cited by regulators, and this report underscores the crucial need to perform a critical function where traditional methods fall short. These are complex problems that require enablement beyond the traditional training, standard operating procedures (SOPs), and quality management systems. Although enterprises are actively exploring technologies and investing to improve results, increasingly there are analytics without insight, process management without process improvement, and investments that will deliver value in the future, but not today. This article discusses three ways quality organizations can move the needle on these challenging problems.

As organizations continue to promote data-driven decision-making, it is still hard to come by important data-driven insights. This challenge can be due to a variety of reasons, but three issues often rise to the top.

First, the most basic prerequisite is the availability of reliable data. Quality professionals deal with a significant amount of data from an increasingly large number of sources. Yet, quality data platforms in most organizations can be limited when compared to other areas, such as commercial or finance. Key to powerful insights is the ability to integrate multiple datasets. To effectively investigate a product complaint, you will need to review where the product was manufactured, packaged, and distributed, as well as current inventory levels and any quality issues associated with the batch. This information will come from a variety of systems, including quality management systems, enterprise resource planning, manufacturing execution systems, and laboratory information management systems.

Additionally, there must be the ability to drive advanced analysis beyond traditional reporting and analytics. In addition to technical capabilities, the skillsets necessary can be rather unique. Data scientists with quality backgrounds can be difficult to come by, and there is almost always a steep learning curve for both analytics and the business teams.

Third, and in the author’s opinion, one of the most important is the operating model. How do you work on your data? How do your analytics and business teams collaborate and test various hypothesis quickly? Increasingly, there is a trend toward data and insights as a service. While data are assets, they only start to generate value when you turn them into critical insights in a timely fashion. Working with a partner can help accelerate that process.

Another common issue that impedes a quality team’s performance is the nature of certain tasks: collecting and analyzing hundreds of data points across systems, documenting results, and preparing reports. Manual handling not only introduces room for error, but it is time-consuming—consider the manpower spent on annual product quality review (APR/PQR).

Another classic example is monitoring of chemical, manufacturing, and control (CMC) compliance. Certain types of change controls have regulatory implications and can impact product release in a market. Compliance creates a significant overhead for batch disposition teams, because these regulatory assessments are often performed manually, and the market implementation constraints must then be calculated manually as well. This process is another example where assisted automation and cross-system process orchestration can be useful. Automation of APR, field alert reports, and lot acceptance protocols, as well as various steps in the batch disposition process are examples of processes that can benefit from automation.

Enable collaboration beyond quality agreements

There is a rising trend in leveraging contract service providers to support the diverse needs in the industry. Thus, the capability to effectively enforce quality agreements will also need to evolve to monitor compliance. Often the data are exchanged manually with vendor quality managers, and latency tends to be an issue. Deeper collaboration will require the ability to exchange data and monitor protocols per the agreements, including direct system integration, collaboration portals, and tools that enable batch traceability throughout the process.

Quality lapses are not uncommon, despite rigorous training and SOPs. In the author’s experience, a digitized, quality-focused platform to help investigate and resolve quality issues will be a competitive advantage for pharma manufacturers.

Articles

Automated systems help detect and resolve quality issues.