The pharmaceutical industry presents unique challenges for business research. Like most enterprises today, quality control and zero defects is essential. But drug manufactures are dealing with life-and-death matters on a daily basis regarding product consistency: an individual who takes too much or too little of his or her needed dose because of a deteriorated drug could have his or her health severely compromised, and legal repercussions are the likely result for the company. Additionally, products even within batches can change over time, so even an initially 'perfect' product may show signs of degradation. In the interests of consumer safety, the effects of such degradation must be anticipated through statistical means, but it remains controversial as to what is the best way to accomplish this.
The article uses hypothetical examples, rather than a real world case study to illustrate how different statistical...
Ultimately, the authors conclude that more work needs to be done. "Identifying OOT stability results is a growing concern for FDA and the pharmaceutical industry. Ideally, the method to determine an OOT alarm should not be too complex. However, something too simplistic," such as a certain percentage change for all properties of a drug, "may not be sensitive enough to identify a true OOT or may give a high rate of false signals" (Alasandro et al., 2003, p.48). Current methods, of testing for quality and depreciation of quality over time tend to err on one side or the other. Stability of the drug is usually the most reliable component to test, however other aspects, particularly within-batch results, are troublingly more difficult to implement than the between-batch methods "because of the sparse data within a batch, especially at early time points" (Alasandro et al., 2003, p.52).
