The ABCs (and D) for Avoiding Out of Specification Results


The Basics… and the Consequences

Out of specification results (OOS) include all test results that fall outside the specifications or acceptance criteria established in drug applications, drug master files (DMFs), official compendia, or by the manufacturer. Even in the best-case scenario, such events can be detrimental to a firm’s success, as the discovery of an OOS result prompts an investigation, which, until completed, halts all manufacturing processes related to the violation. At worst, an OOS result can delay a batch release.

It is needless to say that such bottlenecks should be avoided at all costs. Follow these four tips to curtail OOS results.

A.) Thoroughly investigate any instances of OOS results

The first, and perhaps most obvious step of avoiding future OOS results is to thoroughly investigate any such instance which occurs. More importantly, these internal inquiries in the event of OOS results are not optional, but rather legally mandated. As a rule of thumb, in order to spare further regulatory scrutiny and uncover the root of the issue, investigations must adhere to the industry’s 5 Golden Rules of OOS investigations:

“Do it because you want to, not because you have to”

As previously implied, laboratories don’t face a great deal of choice when forced to inspect any potential cause of an OOS result. Although the mandated investigation doesn’t require enthusiasm on the part of the scientists running it, there are several reasons why it is still crucial to the endeavor’s success. For one, even a shade of reluctance can impede auditors from investigating thoroughly enough to uncover every error, which will only lead to another OOS result down the line.

Furthermore, pharmaceutical firms should recognize the reality that they are, in fact, most benefited by such exhaustive inspections, as they ensure a product’s continued success.

“Always follow the evidence”

This is a rather self-explanatory mantra as, ultimately, the conclusions derived from OOS investigations must be grounded in sufficient evidence to be warranted.

Keeping in mind the first “Golden Rule”, a pharma firm faces no benefit from coming to false conclusions throughout a lengthy audit. For one, if the culprit, so to speak, was incorrectly identified, then a great deal of money would be spent both on implementing a “fix” and inevitably repeating the investigation once the OOS results continue.

Moreover, the regulatory bodies which mandate such inquiries may take action against what they consider to be a faulty investigation. It is also important to keep in mind that such investigations aren’t cheap, and pursuing a red herring costs the company time and money.

“Say what you mean, precisely”

The language used throughout the investigation must reflect the strength of available evidence. After all, there is a big difference between a proven lab error and one that is simply suspected. Impactful words, such as “proven” and “demonstrated” should be reserved for concrete, definitive evidence.

If evidence exists, but is not strong enough to support their usage, then terms such as “suggests” or “likely” are more appropriate. This is essential in establishing a long lasting scientific record of the investigation that is clear and understandable.

“It’s not over till it’s over”

Once a reason for the OOS result is discovered, labs must resist the temptation to end investigations as quickly as possible. The discovered reason for the OOS result may not be the same as the underlying root cause, as often there is more than one that needs to be fully investigated to prevent reoccurrence.

“Don’t let it happen again”

One surefire way to let an intensive and expensive investigation go to waste is by ultimately repeating the same OOS result and thus necessitating another one. Following SOPs and properly training auditors is an essential aspect of smooth investigations down the line.

Repeat instances of the same problem, however, is a clear indication that labs have not embraced opportunities for improvement.

B.) Provide operators with thorough training and properly qualified instrumentation/compliances (21 CFR software)

As previously mentioned, OOS results are practically unavoidable if labs refrain from implementing extensive training of their operators. Furthermore, regardless of how well trained they happen to be, researchers require properly qualified instrumentation that is regularly maintained and calibrated. Consequently, the instruments supplied to laboratories, and the SOPs behind their operation, are essential.

For instance, since key regulations, such as FDA Title 21 CFR Part 11, perpetually seek to prevent falsification, operators must have access to 21 CFR software while also adhering to company standards which discourage the falsification of evidence.

C.) Use controls to essentially prevent the likelihood of OOS results

As discouraging as labor-intensive investigations are, it is certainly encouraging that OOS results can be essentially prevented altogether by simply implementing a tight set of controls. While somewhat restrictive, controls minimize the likelihood of operator error. Also, another method of limiting OOS results is to use control charts, which are used in separating random from non-random variation.

And D.) Manually analyze lab data and other relevant materials

Continuing with the established theme of covering bases, laboratories should not exclusively rely on SOPs, instrumentation, and 21 CFR software as the ultimate safeguard against OOS results. While each of those go far in protecting firms against the strain of eventual OOS investigations, it is still crucial to double check results and data.

After all, while 21 CFR software is certainly effective, it must still be double-checked for optimal assurance.


  1. Center for Drug Evaluation and Research. (2006, October). Investigating Out-of-Specification Test Results for Pharmaceutical Pro. Retrieved from
  2. McPolin, Oona, and Mourne Training Services Ltd. “5 Golden Rules For Effective (And Inspection Ready) OOS Investigations.”, 4 July 2018,
  3. Nunnally, B. K., McConnell, J. S., & Nunnally, D. F. (2008, June 1). Avoiding the Pain of Out-of-Specification Results. Retrieved from