PDA Letter Article

Investigating Out-of-Specification Results for BET

by Hayden Skalski, Sievers Instruments

How important are out-of-specification (OOS) investigations in a quality control lab? The short answer is: Extremely important. For endotoxin testing, an OOS is a result that exceeds an endotoxin limit for a product and can potentially be very costly in two ways:

  1. Time and resources spent performing an investigation
  2. Product on hold until the root cause is determined and confirmed in the OOS investigation

Getting down to the root cause of an OOS result often requires a laboratory to utilize in-depth investigative tools to rule out other potential causes and make a scientific decision about the path forward. One example of an OOS result would be having an LAL bacterial endotoxins test limit of <1 EU/mL for a product but recovering 2 EU/mL. All OOS results necessitate investigation with the primary objective being to determine whether an assignable cause exists or an assignable cause can be identified. OOS endotoxin results can be particularly concerning in terms of patient safety, as high levels of endotoxin in a drug product can cause adverse reactions if administered. Therefore, it is extremely important to investigate these results and find a root cause before releasing or discarding a batch. The U.S. FDA Guidance for Industry: Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production is referenced throughout this document, and this article is intended to summarize and recap that guidance.

An OOS investigation can be split into two parts (1):

  • Phase 1: Laboratory Investigation
  • Phase 2: Full-Scale OOS Investigation

The investigation should start in the laboratory, then move to include all other relevant departments if no clear and obvious lab error is identified. Phase 1 (laboratory investigation) of an OOS should include an initial assessment of the accuracy of the laboratory’s data, including a review of the analysts’ training records, test forms, expiry of reagents, equipment used and test controls. Whenever possible, this should be done before test preparations, such as sample dilutions, are discarded (1). This way, hypotheses regarding laboratory error or instrument malfunctions can be tested using the same test preparations.

If this initial assessment indicates that no causative errors were made in the analytical method used to arrive at the data, a full-scale OOS investigation (Phase 2) should be conducted (1). Upon identification of the OOS result by the analyst, it should be immediately documented, and any potential analyst errors ought to be identified. The testing analyst’s responsibilities begin with achieving accurate endotoxin results. Trained analysts should be aware of potential problems that could occur during the testing process. The analyst should also ensure data is complete and accurate prior to discarding test preparations and upon informing the laboratory supervisor to assess the OOS results. Once identified, the OOS assessment by the supervisor must be objective and timely. An OOS assessment should include:

  1. Confirmation of the analyst’s knowledge
  2. Review of raw data
  3. Verification of calculations
  4. Confirmation of the appropriate use of reagents, standards and instruments
  5. Formal documentation for all of the above

When clear evidence of laboratory error exists, laboratory testing results should be invalidated (1). When no obvious lab error exists and an OOS result is confirmed, it must be reported to Quality Assurance within the appropriate or specified time frame in the firm’s procedures. Once notified, there should be a cross-functional notification with different departments to determine severity and a full-scale investigation. The description of the investigation could include the hypothesis, what samples to be tested, how many replicates, and how it will be documented. It is vital to distinguish between an OOS result and an invalid result. Invalid tests are those where system suitability parameters do not function as expected and therefore may affect the accuracy of the test results. Such system suitability parameters could be:

  • Negative controls (positive gel or reaction time within the range of the standard curve)
  • PPCs (outside 50%–200%)
  • Generation of a linear standard curve
  • %CV (outside set limits if applicable)

An invalid assay, which fails the validity criteria above, would result in a retest from the original sample. Confirmed invalids should still be investigated, but as a lab investigation only, not an OOS. Confirmation based on a retest that proved the same result as the first test, such as spike recovery lying outside the 50%–200% range, would require further investigation.

The goal of Phase 2 is to determine why the product failed if the initial assessment in Phase 1 did not determine that laboratory error caused the OOS. This phase may consist of a production process review and/or additional laboratory work. The objective of such an investigation should be to identify the root cause of the OOS result and take appropriate corrective and preventative action (CAPA) (1). Retesting the sample will be important in this phase as this can determine if there is truly an OOS result or if some type of compromise has happened in the process. Retesting should be done on the same sample aliquot by an analyst who did not perform the initial testing. Retests should be specified in the firm’s standard operating procedures (SOP) on the number of retests before an investigation is launched and concluded. This will help to avoid testing into compliance, which is testing a sample a certain number of times until a passing result is received and the OOS is disregarded. If a retest confirms an OOS result, a further deep dive into the manufacturing process and batch review will be necessary to get to root cause. This may include, for example, examining raw materials or noting any changes in components for that specific batch. Use one of the root cause analysis tools, such as the fishbone diagram (Ishikawa) or Five Whys, is recommended to determine the root cause of an OOS result.

It is important to have a state of control in quality control labs. Processes can be controlled using established quality management systems such as training, SOPs and documentation, investigations, CAPAs, and process monitoring. All labs generate and collect data, endotoxin results for instance, and this data can be used to reduce the variability and increase the effectiveness of endotoxin processes such as the successful removal through downstream and purification activities in manufacturing. Investigations are important and vital to analyzing and improving processes. Following a root cause analysis tool, such as a fishbone diagram or Five Whys, can help a lab nail down root cause and determine the severity of the OOS result. Once a true OOS is thoroughly investigated and confirmed or denied, the Quality Control and Quality Assurance teams can decide on the path forward—whether that is to reject or accept and release.

In conclusion, it is critical to remember that the main goal of quality control testing is the safety of the patient. Endotoxin is a critical release test, so all final-release drug products must pass a lab’s specified endotoxin limit. If a product fails an endotoxin limit test, the result must be thoroughly investigated prior to product release to achieve that goal of patient safety.


  1. U.S. Food and Drug Administration. May 2022. Guidance for Industry: Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production. Silver Spring, MD: U.S. Department of Health and Human Services.

About the Author

Hayden SkalskiHayden Skalski is the Life Sciences Product Application Specialist for the Sievers Instruments product line, specializing in bacterial endotoxins testing. Hayden has over eight years of experience in the pharmaceutical industry and quality control microbiology and has presented on numerous topics surrounding endotoxin testing. Previously, Hayden held roles at Charles River Laboratories, Regeneron and Novartis, validating and executing method development protocols for endotoxin testing, providing customer support, troubleshooting and supporting high-volume product testing. Hayden has a B.S. from the State University of New York at Albany (SUNY) in Biology.