From The Editor | March 1, 2021

Uncertainty In Quality: A Primer For Emerging Biotechs


By Louis Garguilo, Chief Editor, Outsourced Pharma


Outsourced Pharma has a record of presenting readers with aspects of quality and risk management in drug development and manufacturing outsourcing.*

Let’s add a basic introduction of a concept within quality risk management – Uncertainty – specifically for emerging biopharma organizations less attuned to the intricacies of quality-related deliberations.

I asked Yash Sabharwal, CEO of CherryCircle Software, Inc., to help us out after reading his organization’s QbDVision whitepaper titled: “Quality Risk Management: Understanding Uncertainty.”

4 Steps To Uncertainty

1.  ISO 14971 Run-Up

The oft-mentioned International Organization for Standardization (ISO), is an independent, non-governmental, international organization based in Geneva, Switzerland, boasting a membership of 165 national standards bodies.

Sabharwal starts us off with a standard on risk assessment called ISO 14971 – actually a document developed for manufacturers of medical devices, “based on established principles of risk management as evolved over many years.”

Since those standards have been adopted by the FDA – in late December 2020 the FDA recognized ISO’s revised standard (14971:2019) soon after it was released – it’s important to look at some of the same concepts referenced in ICH Q9 (explained next) for risk assessment of pharmaceutical products.

Given how the med-device and biopharma industries continue to nexus, this is certainly a good place from which to proceed with our introduction of Uncertainty.

2.  ICH Q9 Departure

Getting to the heart of the QbDVision whitepaper, Sabharwal detects a potential incongruence between ISO 14971, and the quality risk management standards much earlier recognized and adopted by the FDA and our biopharma industry –  ICH Q9.

The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) brings together regulatory authorities and the pharmaceutical industry to discuss scientific and technical aspects of drug registration.

It has evolved “to respond to the increasingly global face of drug development.” 

According to Sabharwal, quality risk management standards in ICH Q9 depart from ISO 14971 in at least one significant – but as we’ll see reconcilable – way.

3. Vive La Différence

ICH Q9, explains Sabharwal, “utilizes Impact and Uncertainty to assess the Criticality of a product’s quality attribute being out-of-specification.”  

(Editor’s sheet cheat: Impact x Uncertainty = Criticality)

Yash Sabharwal

“However,” he continues, “the official standard on risk assessment in ISO14971 refers to Severity Of Harm and Likelihood Of Severity of Harm, together with the Probability of Hazardous Occurrence to define overall Risk.”

(Severity of Harm x Likelihood of Harm x Probability Hazardous Situation = Risk)

“While the ICH Q9 Impact is analogous to the ISO 14971 Severity of Harm, unfortunately the ICH Q9 Uncertainty departs from the ISO 14971 Likelihood of Harm definition.”

According to Sabharwal, “in the pharma world [as opposed to med device], Uncertainty is dependent on the underlying data supporting the Impact assessment.”

“If the assessed Impact is verified in clinical trials, the Uncertainty will be low; if the assessed Impact is based only on literature or there is little basis for the assessment, then the Uncertainty will be high.

“So a high Uncertainty helps drive a high overall Criticality of product quality attributes.”

In the QbDVision report, instead of a major point of departure, this Uncertainty layer of risk is seen as helping to serve a dual purpose:

Certainty of Knowledge - It may describe the gaps in knowledge about how the

variability of an attribute or parameter impacts patient safety/efficacy or the downstream process.


Likelihood of Severity - It may describe the likelihood of impact/harm in the event of the hazardous situation.

Explains Sabharwal:

“The Uncertainty layer is more prevalent in in the risk assessment of drug products, because it can be difficult to estimate the likelihood of the severity of harm across a wide range of attributes and parameters.

“However, ICH Q9 suggests various approaches to the assessment and risk without specifically prescribing a specific one. 

“Therefore, even though ICH Q9 and public case studies depart from the ISO 14971 standard for drug products, the philosophical approach remains consistent. Namely, that any assessment of Impact or Harm should be weighted, whether it is by the degree of understanding of the Impact, or the Likelihood of Harm

“More important,” thinks Sabharwal, “is the idea that whichever framework is selected, it should be applied consistently.” 

He adds that because of an apparent lack of clarity in existing risk assessment frameworks like the two above, some quality experts in the industry suggest yet-again new frameworks.

But as he demonstrates above: “The existing frameworks can be just as effective as long as the definitions are clear and the application consistent.”

“There is little value,” he says, “in creating new frameworks with new definitions if the central problems are actually poor understanding, and inconsistent application of existing frameworks.”

No need for more frameworks probably sounds good to most of us.  

4. A Larger Role For Uncertainty

According to Sabharwal, Uncertainty assessment is required for product quality risk management, “but it is generally not used in the assessment of process risk.”

“Nonetheless,” he says, “it can be useful to consider Uncertainty when assessing the Criticality of process parameters, and material attributes (a point made in the whitepaper).

This is because the use of Uncertainty can:

  • help identify gaps in underlying knowledge of process variables that should be assessed during process characterization
  • also lead to better segmentation of process variables that are truly critical, focusing process development on those activities that yield the greatest risk mitigation

“In complex manufacturing processes,” says Sabharwal, the ‘unknown-unknowns problem’ represents one of the greatest risks to realizing deep process understanding and intelligence.

“The use of the Uncertainty risk layer can be effective in recognizing the areas of the process where the existence of ‘unknown-unknowns’ may be more prevalent.”

A Quality Chat

I selected this sub-topic to illustrate to those of us not overly versed in Quality that:

  • risk assessment is not a checkbox exercise; it provides significant value when applied consistently
  • the type and depth of thought, consideration, and discussions that transpire
  • apparent discrepancies can be resolved; and mostly
  • as fixed and regulated as the Quality function and parameters may appear, the human element remains crucial in your drug development processes and manufacturing operations, and that more than ever includes at your outsourcing partners. 


*Try a search on our homepage for “quality risk management,” or check out articles from Mark Durivage, Barbara Unger, or Bikash Chatterjee. (For pure provocateur: see contributor Mark F. Witcher.)