May 15, 2006 (Vol. 26, No. 10)

Appropriately Designed Quality Systems Can Improve and Streamline the Research Process

The title of this article may surprise some who wonder how deviations can occur in an R&D environment and why there should be controls to prevent them. Over the years, various scientists have commented that research is a creative process that requires imagination and flexibility, arguably the antithesis of control. My experience indicates that appropriately designed quality systems in R&D, including deviation controls, can improve and streamline the research process and result in better designed marketed products. The purpose of this article is to provide some practical recommendations for controls in the R&D environment.


There are numerous examples of products that took many years to research, develop, get approved, and finally, marketed, due in part to inadequate controls during the R&D phases. Products have been approved but could not be marketed because they could not be manufactured reproducibly according to the conditions of approval. Review of data from preclinical, clinical, and other product studies and investigations into discrepant results from laboratory and clinical studies have revealed wide variations in product formulation, handling, manufacturing, and testing.

In some cases, there have been so many issues to consider that causation of variability in study results could not be determined. In other cases, the lack of adequate documentation prevented any meaningful review or investigation because an accurate history of R&D could not be reconstructed.

Too often, it is virtually impossible to accurately trace the progression of a product&#8217s development from initial phases of research, through multiple development phases, and finally, to the product submitted for approval and proposed for marketing. Sadly, since the review of product development typically occurs long after the work is completed, problems are identified late in the product life cycle and can result in redoing multiple studies.

Product Life Cycles

Various authors have written about product life cycles, which include, generally, the following phases: concept, planning, research and development, scale-up, launch, and maintenance. Applying the definition presented above, research can occur during any of the phases of a product&#8217s life cycle although, admittedly, the R&D activities can vary in scope and extent depending upon purpose.

Some R&D activities may be regulated directly by FDA according to GxPs, while others may support further work that is submitted to and ultimately reviewed and evaluated by the agency. Research scientists may assist in the investigation of manufacturing or laboratory related failures as well as the design and development of corrective actions for marketed products just as they design and develop new products and associated specifications, manufacturing processes, and controls.

Various best practices apply to research, development, manufacture, and control of products regulated by FDA and include GLPs, GCPs, GTPs, GMPs, and Part 11 (electronic records). FDA regulations and guidelines related to GxPs include, among other things, requirements for standard operating procedures. Deviations from such procedures must be identified, investigated, and corrective actions implemented to prevent recurrence. FDA-regulated companies generally understand that these requirements exist even if they do not always meet current standards in their execution, as evidenced by the content of numerous Warning Letters issued during the past several years.

Simply, a deviation is a difference from what is expected or planned. In the GMP world, as the manufacturing environment is sometimes called, the concept of a deviation is pretty straightforward. Since the GMP world is highly regimented and monitored, expectations are (or should be) clearly stated in procedures and specifications. (For example: Mix for 10 minutes.) Deviations from these expectations are, therefore, relatively easy to identify.

R&D Environment

But what about the R&D environment where experiments and investigations are conducted to obtain additional knowledge or confirm previous results? Clearly, and appropriately in these circumstances, the results cannot or should not be completely predictable. Nonetheless, such work does contain some expectations. For example, if an unknown sample is tested against a blank, a standard, or a control, there are expectations about the characteristics of the blank, standard, or control that will be used to evaluate the results from the unknown sample. As research progresses toward product development, and the body of information about a product builds, the list of expectations grows.

Deviation control in such a setting involves controlling the variables that can be controlled. When sources of variability cannot be identified or controlled, the work history (who, what, when, where, how) should be documented sufficiently to stand as institutional memory and a source of intelligence for future planning and study.

In early conceptual work, it may be appropriate to apply basic laboratory controls, as described below, and fully document the conditions under which work was performed. This documentation is invaluable to a company and other scientists as a product moves from the laboratory into toxicology studies and clinical trials, and provides a sound base from which to evaluate the impact of both subtle and distinct differences in a product through the development phases.

The validity of conclusions made from the interpretation of experimental data and other test results is built upon the level of control of known sources of variability: people, equipment, methodology, and samples. The more these elements are managed to prevent or reduce variability, the more assurance there is that the data accurately reflects the true characteristics of the product.


Basic controls for R&D

Establish a risk-based quality system for R&D work that is designed to control what needs to be controlled, rather than to set controls just for the sake of setting controls.

Establish a quality system in which the levels of control increase as a product progresses through the development phases.

Institute appropriate documentation practices that provide flexibility for scientists but result in a complete and accurate history of purpose and scope of work; how the work was performed; where the work was performed and by whom; and observations, results, or conclusions.

Establish mechanisms whereby related pieces of work can be linked or cross-referenced so that the story of product development can be assembled efficiently.

Basic controls for projects.

Develop and follow an experimental plan (part of a project plan) that sets forth the specific types of controls deemed appropriate according to the purpose and scope of the work that is planned.

Establish appropriate controls over known sources of variability including:

PeopleEstablish procedures describing how work should be performed. These procedures may be presented in memoranda, protocols, SOPs, or other forms of documentation. Provide training, instructions, or demonstrations for employees to ensure that procedures are interpreted and will be followed correctly and consistently. It is particularly important to ensure that technique-based skills (e.g., pipetting) are consistently applied. Do not assume that individuals with equivalent years of education and experience will perform critical functions or interpret instructions in the same manner.

EquipmentUse clean or new equipment, accessories, and other labware; use instruments that are in good working order and calibrated; specify as many equipment operating conditions as possible and document the actual conditions used or observed.

MethodologyDescribe analytical methods and manufacturing instructions as thoroughly and specifically as possible.

Samples and other materialsDescribe and characterize samples, controls, and ingredients in manufactured products according to the scope and purpose of the work. At a minimum there should be a record describing: source, appearance, storage, and handling conditions. As development progresses, and specifications are developed, further testing and characterization is expected.

Applying these simple principles and recommendations will improve the quality of R&D work, add credibility and validity to study results, and help to streamline the product development process.

Previous articleAnthony S. Fauci, M.D., director of the National Institute of Allergy and Infectious Diseases
Next articleDr. Lee Lynd Associate Professor of Engineering at the Thayer School of Engineering at Dartmouth College