Paul W. Barone

The spread of COVID-19, which is caused by SARS-CoV-2, an enveloped, single-stranded RNA virus, was declared a global pandemic on March 11, 2020, and the virus has infected millions of individuals and led to hundreds of thousands of deaths. One of the main approaches used to mitigate the spread of the virus has been to institute stay-at-home orders and close all non-essential businesses. These closures do not include pharmaceutical and biopharmaceutical companies, which are conducting essential work to rapidly develop and manufacture new treatments and vaccines to fight the pandemic as well as continuing to provide a steady supply of life-saving medicines to patients.

While the biopharmaceutical industry undoubtedly faces pressure to rapidly develop new treatments and vaccines to fight COVID-19, it also must ensure the safety and supply of all current commercial and clinical biotherapeutics and vaccines. Strategies to prevent process and product contamination by SARS-CoV-2 and to ensure continuity of manufacturing during the pandemic are, therefore, important. The Massachusetts Institute of Technology (MIT) Center for Biomedical Innovation (CBI), and its two biopharmaceutical industry consortia, the Consortium on Adventitious Agent Contamination in Biomanufacturing (CAACB) and the Biomanufacturing Consortium (BioMAN), hosted a webinar on The Impact of SARS-CoV-2 on Biomanufacturing Operations on April 22, 2020. The webinar featured a series of expert perspectives, both from invited speakers with regulatory expertise and individuals from BioMAN and CAACB member companies with viral safety and manufacturing expertise, to provide company-specific views of how the COVID-19 pandemic was impacting biopharmaceutical manufacturing.

Prior to the webinar, a 23-question survey was sent to member companies to gain an industry-wide snapshot of the impact of the COVID-19 pandemic on biomanufacturing operations. The results of this pre-webinar survey, which was completed by 15 member companies, were also presented and provided a high-level view of how biomanufacturers were approaching the global pandemic. This article presents the key points discussed at the webinar, as well as any key conclusions based on the limited information available at the time, to provide the recombinant protein, vaccine, and cell and gene therapy manufacturing community with these insights.

Prevention of process and product contamination by SARS-CoV-2

Preventing biopharmaceutical process or product contamination with an emerging virus, in this case SARS-CoV-2, hinges on first assessing the risk that the virus poses to a manufacturing process. All 15 companies who responded to the pre-webinar survey indicated that they had done this risk assessment. The key questions that must be addressed in a risk assessment are as follows:

What are the potential sources of SARS-CoV-2, and are those sources a potential route of entry into the manufacturing process?

Can SARS-CoV-2 replicate in the manufacturing production cell line?

Is it possible to detect the presence of SARS-CoV-2 with existing virus detection assays?

Can the purification process remove or inactivate SARS-CoV-2?

Potential direct sources of SARS-CoV-2 were identified as operators, cell banks, and raw materials. As a rough measure of the perceived importance of these three potential sources for SARS-CoV-2, 100% of companies who responded to the pre-webinar survey had assessed the risk of infected employees introducing the virus into their manufacturing processes and 60% had assessed the risk of contaminated raw materials as the source. Personnel are a potential source of the virus as viral shedding from asymptomatic individuals has been documented in the literature (1). Existing manufacturing controls and, in the case of recombinant protein production, the minimal number of open manipulations were thought to make the likelihood of direct transmission from personnel unlikely. Similarly, raw materials used for recombinant protein production were thought to be an unlikely source of potential SARS-CoV-2 contamination because they are not of human origin, they may be produced in a manufacturing process that is harsh enough to inactivate viruses, may inactivate virus themselves, or may be treated at the production site to reduce the risk of virus contamination.

This article was revised by the authors on Oct. 20, 2020 to reflect updated research and findings in cited references.

In comparison, as reported in the pre-webinar survey, no company had tested cell banks for the presence of SARS-CoV-2 specifically. The introduction of SARS-CoV-2 from cell banks is highly unlikely because intense screening procedures can determine general contaminants to a high degree of sensitivity as well as the fact that SARS-CoV-2 is a recently emerging virus, and most cell banks currently in use had been established prior to its first appearance. However, cell banks made during the current pandemic may be of increased risk.

A potential indirect route of entry is from the manufacturing environment via personnel, and 60% of survey respondents stated they had evaluated the environment as a potential route of viral entry. To date, two publications have investigated the persistence of the virus on surfaces. Both studies demonstrated the virus could persist in the environment for extended periods of time, with estimates of SARS-CoV-2 survival ranging between three and seven days on stainless steel and plastic (2,3). It is important to note that one report evaluated the surface stability of both SARS-CoV-1 and SARS-CoV-2. Despite their presumed similarity, both viruses exhibited different stabilities on some of the surfaces tested (3). Given the potential for the virus to survive in the environment, one publication reported the capability of a variety of common disinfectants to inactivate SARS-CoV-2. The report found that, after a five-minute incubation with all disinfectants tested in solution, except for hand soap, no virus was detectable (2).

The second question is whether SARS-CoV-2 can propagate in production cell lines. Forty percent of the companies who responded to the pre-webinar survey had evaluated, largely based on literature, the ability of their production cell lines to support the replication of SARS-CoV-2. Evidence presented during the webinar and reported in the literature indicate that Vero cells can support the replication of SARS-CoV-2, with observable cytopathic effect in a TCID50 assay (3). Conversely, evidence presented during the meeting and reported in the literature indicate that Chinese hamster ovary (CHO) cells cannot support the replication of coronaviruses, except for bovine coronavirus cultured in the presence of serum (4). Additionally, a recent publication pre-print has presented data demonstrating that CHO, HEK293, and HT1080 cell lines do not support SARS-CoV-2 replication, whereas, Vero cells do (5).

Next, is it possible to detect the presence of SARS-CoV-2 should it enter the biomanufacturing process? The 

virus assay is considered the gold standard broad spectrum detection assay, and Vero cells are one of the indicator cell lines commonly used in this assay. As noted earlier, Vero cells have been reported in the literature to measure the TCID50 of SARS-CoV-2 (3). A few companies had directly evaluated the ability of the 

 virus assay to detect SARS-CoV-2. One company noted in the pre-webinar survey that they had confirmed Vero’s ability to detect SARS-CoV-2 at a contract testing laboratory, and, during the webinar, one company noted that the standard panel of cell lines was capable of detecting SARS-CoV-2. Since the webinar, at least one publication pre-print has reported the ability of 

 virus assays with Vero cells as one of the indicator cell lines to detect SARS-CoV-2 (5). A few companies were also developing molecular assays to detect SARS-CoV-2, but as of April 2020, no company was performing routine lot testing specifically for this virus.

Finally, should the virus enter the manufacturing process undetected, can the purification process remove or inactivate SARS-CoV-2? Of the 15 companies who responded to the pre-webinar survey, 60% indicated they had performed a risk assessment of their downstream purification unit operations to inactivate or remove coronaviruses. This evaluation of downstream clearance to remove or inactivate SARS-CoV-2 was based on data from other model enveloped viruses, with similar biophysical characteristics as SARS-CoV-2, used for virus clearance studies as per International Council for Harmonization (ICH) Q5A (not a coronavirus model) (6). While no data were presented for SARS-CoV-2, validated and orthogonal viral clearance unit operations for other enveloped viruses are also expected to be effective for SARS-CoV-2, as seen in studies with previous strains of coronavirus. It was also pointed out that viral filtration is validated for its ability to clear minute virus of mice (MVM), a 20-nm parvovirus, and would therefore be expected to provide robust clearance of larger enveloped viruses such SARS-CoV-2, which has been reported to be 60 to 140 nm in size (7).

An important key question, however, is how broadly data from model viruses can be applied to other viruses, such as SARS-CoV-2. One assumption is that SARS-CoV-2 will behave similarly to SARS-CoV-1. However, as noted earlier, there are reported differences between SARS-CoV-2 and other enveloped viruses, including SARS-CoV-1. Of particular relevance is a published report that indicates SARS-CoV-2 is stable for 60 minutes at pH between 3 and 10 (2). These data highlight the potential risk in extrapolating data from one virus to another. Additionally, data from the non-coronaviruses listed in ICH Q5A that are typically used to validate downstream processes are also probably less relevant for comparison.

SARS-CoV-2 impact on advanced therapy medicinal products

Based on the discussion above, a case can be made that the risk of contamination with SARS-CoV-2 to traditional biotech products, especially those produced in CHO cell culture, is likely to be low. However, for many advanced therapy medicinal products (ATMPs), the risk is likely to be higher. First, cells may come directly from patients or donors, who may be infected, and some manufacturing processes use raw materials that are of biological, and sometimes human, origin (8). Second, the increased number of open manipulations in the manufacture of ATMPs increases the opportunity for an operator to introduce virus into the process. Third, for processes that use human cells, the potential risk of replication from human viruses is higher (9). Virus permissiveness, however, is cell-type dependent. 

Replication of SARS-CoV-2 in HEK293T cells was not observed without first transfecting the cells to express the ACE2 receptor (10), while multiple other human cell lines were found to replicate SARS-CoV-2 (11). 

This lack of replication in the human cell lines HEK293 and HT1080 was also confirmed in the literature (5). Fourth, as some ATMP products have a short shelf life, the speed of testing may be a challenge, in addition to the considerations above. Finally, some ATMPs, especially cell-based therapies, have little to no downstream clearance capable of removing an undetected and unwanted contaminant (9). One presenter assigned the highest risk to allogeneic cell therapies, a lower risk to autologous cell therapies, and the lowest risk to acellular and gene therapy products. Because of the challenges for ATMPs, it was recommended by one of the speakers with regulatory expertise that additional testing (e.g., of operators, drug product prior to release, or high-risk materials) and rigorous application of current good manufacturing practices (CGMPs) be applied to ensure product safety.

Continuity of manufacturing during the COVID-19 pandemic

Continuity of biopharmaceutical manufacturing is essential to providing life-saving medications to patients who need them, even in the face of a global pandemic. As confirmed by all the presenters during the webinar and from all but one respondent to the pre-webinar survey, biomanufacturing operations have not been stopped or paused during the COVID-19 pandemic. However, ensuring continued operation will rely on raw material availability, maintaining the health and availability of personnel, and ensuring no disruptions in product distribution or services provided by outside companies.

The lack of even one raw material can impact biopharmaceutical manufacturing as affirmed by one company’s report of pausing manufacturing due to the fact that not all raw materials were available. Broad impacts from the global pandemic have led to diversions in the upstream supply chain. Some suppliers are being asked to manufacture disposable parts for ventilators, instead of disposable bags used in biomanufacturing. Commodity precursor materials, such as folic acid from Asia, carbon dioxide in the state of Washington, and glass for vials and syringes, may be diverted for use in other industries or no longer manufactured at the same scale as broad swaths of the economy are shut down. Some critical questions that companies need to think about are whether their medium/long-term supply of critical raw materials can cover the duration of the pandemic and what is the market duration of their current supply.

It was repeated throughout the webinar that the health of personnel is crucial to the continuity of manufacturing. All companies reported utilizing social distancing and efforts to reduce the number of employees, such as only allowing “essential personnel” on-site. For employees who are required to come on site, companies are mitigating risks to employee health through contact tracing and through enhanced use of personal protective equipment (PPE). One company has given each employee three reusable face masks to be used in office areas. Because there is a shortage of PPE worldwide, this company does not allow surgical masks to be used in non-manufacturing operations. Another company stated their policy that personnel are not allowed to enter a manufacturing facility if they are ill, while another company supports voluntary temperature checks and splitting operational teams. Despite these measures, it is also imperative that companies have a plan for their manufacturing operations in a scenario where a significant fraction of personnel become sick or are quarantined (12).

Although it was not specifically addressed by the presenters during the webinar, it is important to think of the pandemic’s impact on clinical trials as well as the development and manufacturing of new clinical products. One respondent to the pre-webinar survey indicated that the manufacture of products had been prioritized over other activities, potentially impacting other non-product-related activities. In addition to this, the COVID-19 pandemic can lead to other unanticipated delays, such as difficulties recruiting clinical trial participants, especially as hospital visits for non-COVID-related maladies have dropped precipitously during the pandemic.

There has been a shift to address the needs of COVID-19, with several companies now prioritizing the development and manufacture of antiviral drugs and vaccines. It is highly likely that this shift to meet the immediate needs of COVID-19 will also lead to a number of follow-on impacts, such as delays and priorities of staff, for existing product development programs within companies.

The key insight that was repeated during the webinar is that SARS-CoV-2 and the COVID-19 pandemic pose a number of risks to the biopharmaceutical manufacturing industry. The virus itself is a potential product and process contaminant; however, the contamination risk is process and product specific. Current evidence indicates that the risk is likely to be low in the case of recombinant proteins manufactured in processes that meet the following requirements:

Proteins are produced in cell lines that do not replicate SARS-CoV-2, such as CHO cells.

The processes use validated robust downstream viral clearance.

Viral safety testing that utilizes Vero cells as indicator cell lines is routinely conducted.

However, the risk is significantly higher for products that do not meet these requirements. 

For example, autologous cell therapies have a large number of open operations that increase the possibility of introducing the virus from operators, utilize cells that are potentially at higher risk of replicating SARS-CoV-2, and lack downstream viral clearance.

In addition to the risk the virus poses to process and products, consequences from the COVID-19 pandemic itself may have impacts on biopharmaceutical manufacturing operations as a whole including impact to personnel, such as modifications in staffing levels and implementation of procedures to ensure manufacturing personnel health and safety. Disruptions to the raw material supply chain have also occurred. It was reported that upstream raw materials may no longer be available or are being diverted for the manufacture of other important supplies to combat the pandemic. Finally, there have been disruptions in clinical trials and new drug research and development.

It has only been a few months since COVID-19 was declared a pandemic by the World Health Organization (WHO), and it is feasible that some level of restrictions will continue for the foreseeable future. What, then, are companies doing to prepare for the next pandemic? Just over half of those who answered the pre-webinar survey have surveillance programs for emerging viruses. The first signals are subtle; one company’s pathogen safety group, which meets weekly, noticed on January 5, 2020 that Chinese regulatory authorities had notified WHO of a strange disease. However, a key challenge for all biopharmaceutical companies during a pandemic is rapidly finding relevant information to help assess the risk of an emerging virus to their processes and products. Experience from the COVID-19 pandemic has demonstrated that much of this work is often repeated across organizations. There is an opportunity for industry collaboration, through organizations such as CBI and the CAACB, to enable faster access to information, including potentially collecting and aggregating relevant data to help manufacturers assess the risk to their processes.

 online, DOI:10.3201/eid2607.201595 (2020).

 Microbe online, DOI:10.1016/S2666-5247(20)30003-3 (2020).

M. Francis, “Propogation of bovine coronavirus in Chinese hamseter ovary cells,” US Patent WO 00/44883, 2003.

, DOI:10.1016/j.biologicals.2020.08.010 (2020).

Q5A Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin, 

Guidance for Industry, Good Manufacturing Practice Considerations for Responding to COVID-19 Infection in Employees in Drug and Biological Products Manufacturing

Paul W. Barone, pbarone@mit.edu, is co-director, Biomanufacturing Program, and director, CAACB; Flora J. Keumurian, florak@mit.edu, is program and operations manager; Michael E. Wiebe, mewiebe@mit.edu, is lead investigator, CAACB, and principal at Quantum Consulting; Jacqueline Wolfrum, jwolfrum@mit.edu, is co-director, Biomanufacturing Program, and director, BioMAN; James C. Leung, leungjc@mit.edu, is lead investigator, CAACB, and senior research fellow; Anthony J. Sinskey, is faculty director, MIT Center for Biomedical Innovation; Professor of Biology, MIT, asinskey@mit.edu; and Stacy L. Springs*, ssprings@mit.edu, is senior director of programs and executive director of Biomanufacturing Initiatives; all are at MIT Center for Biomedical Innovation.

*To whom all correspondence should be addressed

Articles

This article presents the results of a survey conducted to gain insight on the impact of the COVID-19 pandemic on biomanufacturing operations. 

The Impact of SARS-CoV-2 on Biomanufacturing Operations

Although there is awareness among the biopharmceutical industry of the challenges and trade-offs associated with monitoring the safety of drug products, there does not seem to be a clear and consistent understanding of characteristics correlated with the presence of critical-to-quality deviations of a product at the commercial scale (1). The intent of this paper is to initiate an understanding of product and site characteristics that statistically correlate with quality performance of biopharmaceutical commercial products. Towards this objective, a comprehensive survey on related topics is underway at the Massachusetts Institute of Technology (MIT) Center for Biomedical Innovation (CBI). The interim results described in this research represent survey results from 34 commercial-scale biopharmaceutical products at 11 international sites. This research builds on analysis done by the MIT CBI team earlier this year (2). 

Some regulatory agencies, including FDA, have actively pursued a risk-based approach to manufacturing site inspection because their inspections are site-specific and not restricted to an individual product (1). Explicit inputs that FDA uses for its risk-based selection of sites to inspect include the types of products manufactured (e.g., prescription versus over-the-counter, approved versus unapproved, therapeutic classes); the control and/or contamination potential of the manufacturing environment (e.g., facility size, facility type, number of drug products, sterility requirements); and cost of inspection (e.g., domestic versus international locations). These inputs are logical for consideration of total risk to quality at the level of a site, but, in this research, the focus regarding quality is on the product. 

A comprehensive survey on topics related to site characteristics, quality approaches, quality activities, perception of relative consistency of regulators, product-level process details, and product-level compliance performance has been developed by MIT's CBI. The survey includes questions that result in a total of 57 variables across these topic areas. The developed survey was tested with industry representatives and then deployed over the past 18 months as a secure web-based questionnaire directly to biopharmaceutical manufacturing sites through several different channels. The individuals targeted for inclusion as survey respondents have been manufacturing plant managers or their representatives.  

In total, 20 sites representing 52 manufactured products have completed the survey. These responses have been filtered to focus on commercial-scale manufacturing and products for which all relevant information was provided. This reduced the number of products included thus far to 34 across 11 sites.  

The research is supported by the Alfred P. Sloan Foundation of New York City. 

Characteristics of the products and sites. 

The demographics of the 34 products included in the analysis are organized into the following groups: 

29 of the products surveyed are marketed in the US; 24 in Canada; 33 in Europe; 26 in Japan; 17 in South America; and 20 in Asia (outside of Japan).

Product type: 27 of the products surveyed are therapeutic proteins; 12 are antibodies

Processing: 25 of the products surveyed are produced by mammalian cell culture; 9 by microbial fermentation

Complexity: 15 of the products are perceived as above average in process complexity

Role of quality by design (QbD): 10 of the products are significantly impacted by use of a QbD approach.

Characteristics of the manufacturing site:

Geographic location of commercial manufacture: 6 are based in in Asia; 12 in Europe; and 16 in the US

Contract manufacturing: 14 of the products are manufactured in a site that does greater than 40% of their work as contract work 

Other biologic products at the site: 15 of the products are manufactured in a site that also manufactures more than five biologic products

Use of third-party quality inspections: 24 products surveyed are manufactured in a site that has had multiple third-party quality inspections (for practice) in the past five years.

Based on these overal demographics, the products in the MIT CBI survey represent a reasonably diverse set of biopharmaceutical products, processes, and manufacturing sites. 

Distribution of technical personnel within the manufacturing sites. 

The distribution of technical personnel within a manufacturing site provides some representation of the priorities and/or focus of the site. The sites reported the number of personnel in each of the following technical areas: quality control (QC)—average 12.6%; quality assurance (QA)—average 17.4%; engineering support and services average 8.8%; technical services—average 16.3%; and manufacturing—average 44.9%.  

Figure 1: Fraction of personnel associated with each of the technical areas at different sites. QA is quality assurance; QC is quality control; ESS is engineering support and services; and TS is technical services. (ALL FIGURES COURTESY OF THE AUTHORS) 

A histogram with the fraction of personnel in each of the first four areas is presented in 

. The histogram of fraction of technical personnel in manufacturing is shown in 

. There is a wide distribution of where technical personnel are assigned within the sites (particularly visible from the distribution of technical personnel in manufacturing) suggesting one or more of the following scenarios across the sites that completed the survey:  

Different role definitions for personnel within the technical areas

Structural differences between sites necessitating different distributions of personnel

Different priorities and/or focus of the site management

Lack of understanding regarding optimal distribution of technical personnel. 

Figure 2: Fraction of technical personnel in a manufacturing position across sites surveyed.  

 Each site was asked to identify significant sources, or drivers, of changes in quality efforts and activities at their site. The respondents indicated whether specific drivers had been experienced by the site in the past and/or are anticipated drivers for the site in the future. The results from the survey of the sites are presented in Figure 3, ordered from left to right in descending order of fraction of sites identifying the type of driver as one that has been experienced in the past. Of particular note are the following findings:  

The two types of drivers most frequently experienced were technical challenges (i.e., process, material, equipment) and regulatory challenges (i.e., change from the regulator, noncompliance).

Drivers potentially directly related to profitability of the product (e.g., cost reduction and pursuit of new markets) have been infrequently experienced.

Drivers expected to be experienced significantly more in the future than the past are those directly related to profitability including pursuit of new technology, cost reduction, and pursuit of new markets.

The number of types of drivers expected to be experienced per site is significantly higher than the number of types of quality drivers historically experienced per site, increasing from an average of 40% to 60%. For example, sites expect to have more drivers of change in quality approach and activities and/or have different outcome priorities (e.g., cost reduction and anticipated regulatory change) in the future compared with the past.

Figure 3: Experienced and expected drivers of quality activities and resources.    

Correlations with critical-to-quality product deviations. 

Using critical-to-quality defects at the commercial scale as the dependent variable, 56 other variables were compared based on likelihood ratios for each of the 34 commercial products. Of these 56 variables, 11 were found to have statistically significant different likelihoods depending upon whether a critical-to-quality defect was experienced for that product. These 11 variables of significance are shown in Figure 4 and grouped based on confidence level of statistical significance (the top row contains variables for which the difference in likelihood ratio demonstrated greatest confidence) and grouped based on whether the correlation was negatively or positively correlated with a critical-to-quality defect at the commercial scale. If the variable is in the negatively correlated section of 

, then presence of that variable is more likely to occur for a product without a reported critical defect.  

Table I: Variables that were found to be statistically different in likelihood of correlating with a critical-to-quality defect. 

For example, a product manufactured at commercial scale in a site having an above-median fraction of technical personnel working in QA was statistically less likely to have experienced a critical-to-quality defect at the commercial scale. On the other hand, a positively correlated variable is more likely to be present for a product that has experienced a critical defect. For example, a product manufactured in Asia is more likely to have experienced such a defect. 

Consistent with prior preliminary discussion regarding variables correlated with product defects, many of the statistically significant variables presented in 

 fall into the categories of history of quality issues in the product lifecycle, geographic region of manufacture, and manufacture in a contract manufacturing site (2). Of note is the dependence upon fraction of technical personnel employed in QA and employed in manufacturing. More generally, this dependence is related to the discussion of the distribution of technical personnel across a site as described above, where the lack of consistency and, perhaps, understanding of an optimal distribution was highlighted. The data presented in 

 demonstrates that above median level of the fraction of technical personnel in QA and below median level of the fraction of technical personnel in manufacturing are both correlated with lower likelihood of a critical-to-quality defect at the commercial scale.  

It is possible that greater emphasis on quality is demonstrated by fraction of personnel in QA compared with an emphasis on execution demonstrated by fraction of personnel in manufacturing. Further investigation in this area and clarity regarding the expectations of technical personnel roles could provide greater insight into this observation. 

Although there were 11 variables significantly correlated with likelihood of a critical-to-quality defect at the commercial scale, there were also 45 variables found not to be so. Several of these variables are notable as not having been found to correlate with critical-to-quality defects at the commercial scale, including:  

Type of product (e.g., protein, antibody).

Mode of production (e.g., mammalian, microbial).

Level of production complexity perceived by the site: The first three variables are related to perceived manufacturing complexity and logic could suggest that a dependence is expected, but has not been found in the current data.

Use of QbD: QbD can take on many forms and has a complex impact on measured quality performance metrics. Generally, this lack of correlation is notable as quality by design is expected to increase the likelihood of product with consistent quality over time (3).

Number of product batches manufactured in the past five years: More batches would be expected to increase probability of a single defect occurring but is logically counteracted by familiarity with the product and the possibility of increased learning from completed batches (4).

With the intent of increasing understanding of variables associated with commercial quality performance of individual biopharmaceutical products, a survey was conducted with questions related to the product lifecycle and site characteristics. The interim analysis presented here was based on 34 products across 11 sites with diverse characteristics.  

The analysis focused on determining which product or site variables were statistically more likely to be present for products that have experienced a critical-to-quality defect at the commercial scale. The themes of variables found to be statistically significant based on the likelihood analysis were:  

A history of quality issues in the product lifecycle 

The manufacture in a contract manufacturing site 

Distribution of technical personnel at the site (particularly employed in QA and employed in manufacturing). 

A number of themes not found to be statistically significant for association with commercial critical-to-quality defects, based on the data available, include:  

MIT survey results address product and site characteristics that statistically correlate with quality performance.