Peter Kitz is vice-president, Global Pharmaceutical Quality, Bristol-Myers Squibb.

Peter Kitz

	The European Federation of Pharmaceutical Industries and Associations (EFPIA) conducts an annual survey of inspections among its member companies, which are research-based pharmaceutical firms. The data (1) generated demonstrated the workload, the use of resources, and the outcome of inspection practices for the oversight of good manufacturing practice (GMP) and good distribution practice (GDP) by domestic and foreign regulatory authorities. Member companies are confronted with more inspections of their manufacturing sites by authorities from other countries (i.e., foreign inspections). Part II of the article describes additional challenges and opportunities to improve inspection efficiency, based on data from the survey (1).

	Industry and regulators have made the point that inspections are resource-consuming. Here are some calculations based on the data (1).

	Estimated resources involved in inspections

	Based on data from the EFPIA survey, an estimation of resources per foreign on-site inspection is given in 

. For on-site inspections, many companies were asked to submit documents prior to inspections. The data showed that the submitted documents helped to focus and reduce the inspection time on-site. The preparation effort might additionally be driven by specific requirements from individual inspectorates. Compared to the regulators, the inspected companies needed 10 times more resources for preparation, conduct, and follow-up of inspections.

	Resources calculated as the duration of inspections

	The average duration of foreign inspections by country, where more than four inspections were performed (not including paper-based inspections), is shown in 

. Based on the 2015 data of inspections at manufacturing sites, the following can be calculated:

		On average, two inspectors come for each inspection

		Foreign inspections last, on average, 4.1 days at the site, requiring 8.6 inspector days

		Domestic inspections use, on average, 3.1 days per site, requiring 6.4 inspector days

		The industry requires approximately 160 person days for preparation, conduct, and follow-up of inspections

		An average inspection fee is assumed to be €30,000.

	For inspectorates, it is assumed 20 person days per inspection (including pre- and post-inspection and travel for experts and participants from corporate departments). For the 486 inspections in 2015, the sum is estimated to be up to 9720 person days.

	The 2015 survey (1) asked seven questions on specific and general issues related to inspection practice. These questions covered, for example, noticeable change regarding how inspections have been conducted; unique question content in the scope of inspection that had not been expected; topics raised in inspections that are beyond the announced scope of the inspection; the approach taken by inspectorates to identify data-integrity topics; and if the national agency implemented a risk-based approach for scheduling/conducting inspections.

	A summary of the received responses on all the specific questions conclude the following priorities of inspection practices:

		Data integrity with focus on quality control laboratories, pharmacovigilance data, and the manufacturing environment, including the corporate environment

		Sterile manufacturing processes and operations

		Approaches for application of new and revised or not yet implemented guidelines. Some answers reflected this area to be an “unusual inspection practice” that might represent an opportunity for inspectorates for training of new staff or to challenge regulations in practice.

	Considerations to optimize inspection practice

	After 12 years of data collection, EFPIA recognizes that there are significant opportunities for better use of resources and suggests the following points for consideration to optimize inspection practice:

		Inspections should be scheduled and conducted using a risk-based methodology. The Pharmaceutical Inspection Convention and Pharmaceutical Inspection Co-operation Scheme (PIC/S) model for risk-based inspection planning in the GMP environment could be used as a basis for inspection planning (2).

		Regulatory authorities should give first priority to ensuring that manufacturers and affiliates in their own territory are compliant through adequate local oversight. To this end, information in the EudraGMDP database should be kept up-to-date.

		Foreign inspections should be conducted only if deemed necessary. First of all, such inspections can be facilitated through application of the principle of mutual reliance. Where possible, regulatory authorities should aim to rely on the partner authority’s supervisory oversight, based on risk and best use of resources.

	Currently, the exchange of inspection reports and company responses does not seem to reduce the number of inspections; instead, this information is used often as a preparation step for upcoming inspections. Resources could be better leveraged by mutual recognition or reliance on inspection outcomes from trusted inspectorates, for example, from PIC/S member inspectorates. In 2013, EFPIA proposed in a position paper that PIC/S member inspectorates should agree on equivalence of the domestic GMP requirement (3). This approach could lead to opportunities to reduce the number of redundant inspections by relying on inspections previously conducted by other members who maintain harmonized standards.

	Statutory agreements, in the form of mutual recognition agreements (MRAs) or memoranda of understanding (MoU), also offer an opportunity to reduce the number of inspections. For example, the United States is committed to bilateral and multilateral dialogue with other countries (e.g., in form of a MRA with the European Union independent of the Transatlantic Trade and Investment Partnership [TTIP]).

	In the longer term, inspection reports along with the inspected company’s responses could be shared in an international inspection database, provided confidentiality issues are addressed and respected.

	The EFPIA position paper of 2014 (4) proposes effective and balanced risk-based regulatory oversight to improve the use of resources. Inspection efficiency can be improved by harmonization of standards and regulatory processes including:

		Certification process for regulatory inspectorates

		Risk-based approach to inspection planning and scheduling

		Standard data packs for on-site and paper-based inspections for faster provision of information as well as better facilitation and use of resources

		Standard terminology for categorization and ranking of observations, and standard reporting template

		Globally accepted GMP/GDP certificate format to document compliance of an inspected siteGlobal database for sites approved by certified inspectorates. The EudraGMDP database is a source of compliance information of sites inspected globally. EFPIA encourages regulatory agencies from outside the European Economic Community (EEC) and MRA countries to use this platform. EFPIA further suggests global product numbers (International Organization for Standaradization [ISO] Identification of Medicinal Products [IDMP]) could be incorporated into this database.

	It is proposed that shared inspection outcomes may avoid duplication of effort and help facilitate early access to markets. The reasons for performing unannounced inspections are well understood; however, comparison of the outcomes from the EFPIA survey shows no difference, either for domestic or foreign inspections. The only difference is that unannounced inspections need longer to gain access to experts than announced inspections.

	EFPIA has proposed the development of a standardized preparation documentation package to enable faster provision of information and better facilitation and use of resources (4). This package could include:

		Site-related information: site master file (PIC/S SMF)

		Product related information: annual product/annual quality reviews

		Quality system-related information: quality manual (reflecting quality management system)

		Additional compliance information: GMP/GDP certificates for the site, list of inspections, list of internal audits and number of customer/contractor audits, major changes, rejected batches, out of specifications, and others.

	Interaction between regulatory authorities might facilitate waivers for inspections. Time to market could be improved if inspectorates exchanged data on inspection activities, including pre-approval inspections. The industry would benefit from a harmonized guidance for inspectorates defining documents to send for a paper-based inspection prior to an on-site inspection.

	Interaction with and within inspectorates. EFPIA encourages PIC/S member inspectorates to rely on conclusions from domestic inspections by PIC/S members. The industry understands that PIC/S does not have a legal obligation to share information, but PIC/S membership does have the opportunity to increase mutual reliance. The data show the potential to take advantage of resources that could be redirected to domestic oversight or higher risk sites. As shown in 

, 46% of foreign inspections were conducted by PIC/S member inspectorates; these inspections resources could be saved by relying on the results of inspections already conducted, such as the GMP certificate of an inspectorate that is a member of PIC/S. To create greater efficiencies, while maintaining an effective programme, the authors believe it is worth encouraging PIC/S member inspectorates to rely on conclusions from domestic inspections by PIC/S member inspectorates.

	Furthermore, the existing EU/EMA joint inspections process can be extended to include the local regulatory authority in foreign countries to provide greater efficiency with the same inspection. The continued use and success of risk-based approaches as applied by long-time PIC/S members can be a role model for new PIC/S member inspectorates to continue building confidence in other PIC/S member inspectorates.

	PIC/S member inspectorates could reference PIC/S-GMP requirements in their inspection report or in a comparison table with the local requirements. In addition, another aspect that the authors believe to be worth considering, is to extend the sharing of inspection reports among PIC/S member inspectorates at a harmonized platform (e.g., EudraGMDP). Continuous education is encouraged, and useful sources include PIC/S trainings/seminars and the PIC/S Inspectors Academy, which can help to build trust among inspectorates. Finally, PIC/S could facilitate mutual reliance on inspections by adopting local statutes.

	The pharmaceutical industry has become increasingly global, and new risks have emerged, such as falsification of medicines. Although increased cooperation between regulatory agencies has been noted, there is still significant duplication of inspection oversight at manufacturing sites. Avoidance of inspection duplication would allow resources to be directed towards higher risk areas of the supply chain.

	Although the number of foreign inspections has not increased, it has levelled at a high number (1). In the 2015 survey, 640 sites from innovative medicine manufacturers received foreign inspections, costing more than 75,000 regulator hours and €80,000,000 spent by EFPIA member companies.

	Globally accepted certification systems have proven to be efficient and effective in other industries, such as aviation, nuclear energy, and food. EFPIA advocates the use of a globally accepted GMP/GDP certificate format to demonstrate compliance of an inspected site. Such certificates should be available on a public database. EFPIA suggests that harmonization of inspections of pharmaceutical products is possible, in which the right quality in the lifecycle of inspections is balanced through making best use of resources.

	Data were evaluated by the following members of the EFPIA GMP network reporting to the Technical Development Expert Group (TDEG): Stephan Rönninger, Amgen (Moderator); Johanna Berberich, Bayer; Véronique Davoust, Pfizer; Peter Kitz, Bristol-Myers Squibb; and Andreas Pfenninger, Interpharma. Support was provided by Gerd Fischer, Boehringer Ingelheim. The authors thank Michele Hunter for technical writing.

	In addition, that authors thank the companies that provided input to the survey in 2015 and before: AbbVie, Almirall, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Biogen, Bristol-Myers Squibb, Eli Lilly and Company, Grünenthal GmbH, GlaxoSmithKline, Johnson & Johnson, Merck (MSD), Merck Serono, Novartis, Novo Nordisk, Pfizer, Roche, Sanofi, Servier, Takeda, and UCB.

	1. S. Rönninger et al., “The GMP/GDP Inspections Landscape--Part I: Data,” Pharm. Tech. Europe 29 (1) (January 2017).
	2. PIC/S, A Recommended Model for Risk-Based Inspection Planning in the GMP Environment, PI037-1 (January 2012).
	3. EFPIA, A Concept for Harmonized Reporting of Inspections, position paper (May 2015).
	4. EFPIA, 

Enhanced Good Manufacturing and Good Distribution Practices (GMP/GDP) Inspection 5

. Efficiency (May 2014), accessed July 2016.

	Pharmaceutical Technology Europe
	Vol. 29 No. 2
	Pages: 5-9

	When referring to this article, please cite it as S. Rönninger et al., "The GMP/GDP Inspections Landscape-Part II: Considerations and Opportunities," 

Articles

The authors look at challenges and considerations to continuously improve inspection efficiency.

The GMP/GDP Inspections Landscape—Part II: Considerations and Opportunities

	Good manufacturing practice (GMP) inspections are fundamental for ensuring that medicinal products are produced to the high standards required for patient safety. The European Federation of Pharmaceutical Industries and Associations (EFPIA) represents research-based pharmaceutical companies in Europe. For reasons of transparency, EFPIA conducts an annual survey of inspections among its member companies. The data generated demonstrate the workload, use of resources, and the outcome of inspection practices for oversight of GMP and good distribution practice (GDP) by domestic and foreign regulatory authorities.

	Member companies are confronted with more inspections of their manufacturing sites by authorities from other countries (i.e., foreign inspections). The purpose of this data collection is to create awareness and develop potential opportunities to redirect resources toward better site coverage and, therefore, protection of the patients. Part I of this article reports, for the first time, the results of the annual inspections survey. Part II will look at additional challenges and opportunities for improving inspection efficiency (1).

	According to the International Council for Harmonization of Technical Requirements for Human Use (ICH), inspections of a firm’s manufacturing operation are essential to evaluate commercial manufacturing capability, adequacy of production and control procedures, suitability of equipment and facilities, and effectiveness of the quality management system in assuring the overall state of control. Notably, pre-approval inspections include the added evaluation of authenticity of submitted data and link to dossier (2). Inspections of a firm’s manufacturing operations assess GMP and GDP compliance and are legally required for drug (medicinal) products in most countries worldwide. In some regions, inspections of API manufacturers are also mandatory. Furthermore, relevant regulations for quality systems (3) require companies to have an independent auditing scheme of their sites, contractors, and suppliers. Recent emerging regulations require regulatory authorities to inspect manufacturers in the countries where the medicines and APIs are produced.

	Objective, scope, and definition of the EFPIA survey

The objective of the annual survey is to provide awareness of global health authority inspection activity, focusing on duplicate inspections as a means of determining the value to the patient while controlling risk. Through the annual survey, EFPIA aims to encourage regulatory authorities to focus their resources on activities that pose higher risks to patient safety and to drive a collaborative approach among health authorities in inspection practice and sharing of inspection outcomes. EFPIA recognizes that convergence in activities between regulators is essential to effectively evaluate commercial manufacturing capabilities.

	The EFPIA member companies reported inspections at companies of research-based manufacturers. The numbers should be seen as a snapshot. Aspects of the survey are as follows:

		Regulatory GMP/GDP inspections conducted at manufacturing sites and affiliates, inside and outside the regulatory authority’s own borders, including related International Organization of Standardization (ISO) certifications (e.g., for combination products/medical devices) are included in the data.

		Only companies that own facilities are reported, as inspections of contract manufacturers might be counted twice.

		Sites that manufacture both human and veterinary products are included in the scope of this survey. Other inspection types (e.g., on pharmacovigilance aspects [GPvP], inspection on narcotics, and environmental safety) had been reported in some cases, but not assessed.

		Sites manufacturing exclusively veterinary products are not reported.

	In addition to the data collection, there are specific questions concerning inspection activities that are adapted each year to reflect areas of particular interest to EFPIA. The definitions used in the survey are available online (listed online in 

	Over the span of the survey so far (2003 to 2015), data have been collected from approximately 660 manufacturing sites annually. Conducting and receiving inspections involve considerable regulator and industry resources. There were 486 foreign inspections and 693 domestic inspections, which required 3951 foreign inspector days and 4456 domestic inspector days. It should be noted that these days were spent by regulators only; industry resources are not included in the figures. Also, these days do not include resources used for paper-based inspections or inspections where no inspector days had been reported. Less than 10% of the inspections were reported as unannounced. Whether announced or unannounced, there was a high level of compliance (more than 95% of the inspections had positive outcomes, with no regulatory action required).

	Inspectorates performing foreign inspections

	Over the past 12 years, 64 countries have performed foreign inspections. In 2015, 34 countries performed foreign inspections, compared with 40 in 2014 and 31 in 2013. The survey results showed the number of foreign inspections per country in 2015 (see 

). Group 1 is formed by countries performing more than 60 foreign inspections; Group 2 is for countries performing more than 20 and less than 60 foreign inspections. In addition to those shown in Figure 1, there were also seven countries where one foreign inspection was made.

	The data show that countries in Group 1--the United States and the European Union (where all EU countries are counted as one inspectorate)--perform approximately twice the number of foreign inspections compared with countries in Group 2. A trend analysis of survey data from 2011 to 2015 revealed:

		A downward trend in the number of foreign inspections being performed annually by Group 1 and Group 2 countries from 2011 to 2015 with some exceptions.

		An upward trend in the number of foreign inspections being performed annually by other countries (e.g., China performed five foreign inspections in 2011 compared with 16 in 2015; Belarus performed three foreign inspections in 2011 compared with 12 in 2015).

		Analysis of foreign inspection activity between the ICH founding members of the US, EU, and Japan in 2015 revealed the following:

		The US performed 62 foreign inspections in the EU, compared with 104 inspections in the previous year. It performed no inspections in Japan in 2015, compared with one inspection in 2014.

		The EU performed 35 foreign inspections in the US, compared with 36 in 2014; it conducted no inspections in Japan, compared with one inspection in 2014.

		Japan conducted 12 inspections in Europe in 2015, compared with 23 in 2014; it performed 10 inspections in the US in 2015, versus 19 inspections in 2014.

	There were significantly fewer inspections from the US in Europe (approximately 40%) and from Japan in the EU (approximately 48%). Similar reductions had been reported in the 2014 data.

	The maximum number of inspections at a single site within one year has been reported to be 15, including seven foreign inspections. The same workload for foreign inspections had been reported by five sites. Overall, 52 sites received more than four inspections in 2015, 570 sites report having between one to four inspections in 2015, and 118 did not receive any inspections in 2015. Although the number of sites with no inspection is decreasing, indicating better overall coverage, there are still sites that are not inspected. This observation might suggest that resources could be better balanced.

	The survey asked for specific data from sites receiving multiple inspections, with the aim of addressing the question: “Are multiple inspections driven by risks?” It was noted that one manufacturing site can have several manufacturing licenses, thus there can be a need for several inspections at the same geographical location. However, there was no clear indication that these multiple inspections were always driven by risk. The data also confirmed that there is always intensive domestic oversight.

	In general, the inspections cover drug products (64%, including 18% sterile operations), API (20%, including 10% biotech API sites), vaccines manufacturers (8%), others (5%), and medical devices (3%). The majority of foreign inspections target the drug product (see 

). The survey results confirm the split of product types inspected remains constant among the different inspectorates with the exception of Japan. The authors believe this could be a result of the specific mutual recognition agreement (MRA) between Japan and Europe.

	Of reported inspections, 85% were GMP related, 4% were GDP related, 3% covered ISO certification, and 8% were related to other GxP operations. The data show that:

		79% of domestic inspections concerned GMP compared with 94% of foreign inspections

		5% of domestic inspections examined GDP compared with 2% of foreign inspections

		11% of domestic inspections examined other GxP compared with 4% of foreign inspections.

	Foreign inspections focused largely on GMP processes. Inspection of distribution licenses was performed mostly by the domestic regulatory authorities.

	For the purpose of the survey, a GDP inspection is defined as an inspection that covers systems and processes within the scope of the distribution license. A significant number of new regulations and guidelines (4-7) were published in recent years aiming at better control of distribution channels to prevent entry of falsified products in the legal supply chain. Inspections of GDP compliance are, therefore, deemed to be important. They can be performed at a manufacturing site and/or at an affiliate of the company in the country of destination or at a hub. The survey shows that more of these inspections are performed each year.

	In the 2015 survey, 203 inspections were reported to include GDP in their scope:

		91% were domestic inspections--81% at affiliates and 19% at manufacturing sites; 2% of the inspections covered APIs or API/sterile and 98% covered finished products; none of these inspections reported a follow-up with interruption of supply

		9% foreign inspections were reported-58% of these inspections had been performed at manufacturing sites; 42% of these inspections at affiliates (15% of these included an ISO certification).

	Out of the 22 companies surveyed in 2015, 16 reported inspections at affiliates. The majority of inspections were conducted according to distribution licence. Other topics covered are GMPs (e.g., supply chain verification by a responsible person, repackaging, relabeling), other GxP processes (e.g., pharmacovigilance), and ISO certifications. There were 309 inspections at affiliates; 96% were domestic inspections (performed by inspectorates from 99 countries), and 4% were foreign inspections:

		83% of all inspections had no follow-up

		10% of all inspections had follow-up without interruption of supply

		6% of all inspections had follow-up with interruption of supply (e.g., distribution affected, import delay, delivery to customers affected)

		1% of all inspections had no outcome reported/outcome pending.

	The majority of inspections at affiliates named the GDP distribution license as the primary scope of the inspection and drug product as the product type inspected.

	The survey analyzed the different forms of inspections. In 2015, there were 160 inspections flagged as pre-approval inspections (PAI) of manufacturing sites (95 were foreign and 65 were domestic inspections).

	Two joint inspections were reported in 2015, which were in fact, two inspections running in parallel at the same time and were thus “concomitant” inspections rather than joint inspections: one inspection in the US from the Food and Drug Administration (FDA) and EU, and one inspection in Australia by Australia’s Therapeutic Goods Administration (TGA) and US FDA.

	In the past, the industry has described positive and negative views of joint inspections, with some companies questioning the benefits and reporting that it feels like two different inspections at the same time. From the regulators’ point of view, there is an open question if joint inspections will continue to be organized. Joint inspection pilot programmes were initiated by US FDA. Later, US FDA joined the Pharmaceutical Inspection Convention and Pharmaceutical Inspection Co-operation Scheme (PIC/S), shifting the exchange toward mutual reliance on inspection data from partner inspectorates. The PIC/S Joint Audit Programme could still be used for alignment and training purposes.

	Paper-based foreign inspections Twenty-four paper-based inspections were reported as foreign inspections in 2015. These inspections were conducted by China, Chinese Taipei, Germany, Japan, South Korea, Syria, and US FDA. Inspected companies felt a need to involve a lot of resources due to requests for specific inspection formats and questionnaires. This issue was compounded by the fact that institutionalized formats such as the PIC/S site master file were not accepted by the different regulators.

	Assessment of the outcome of inspections

	Data on compliance following inspections are shown in 

. The data indicate that domestic and foreign inspections lead to similar outcomes.

	Inspections are essential to evaluate commercial manufacturing capability, adequacy of production and control procedures, suitability of equipment and facilities, and effectiveness of the quality management system in assuring the overall state of control and can include the evaluation of authenticity of submitted data in a registration dossier. The results of the survey based on the 2015 inspection data and trends since 2003 demonstrate dublication, workload, use of resources, and the outcome of inspection practices for oversight of GMP and GDP inspections and cerifications.

	Data were evaluated by the following members of the EFPIA GMP network reporting to the Technical Development Expert Group (TDEG): Stephan Rönninger, Amgen (moderator); Johanna Berberich, Bayer; Véronique Davoust, Pfizer; Peter Kitz, Bristol-Myers Squibb; and Andreas Pfenninger, Interpharma. Support was provided by Gerd Fischer, Boehringer Ingelheim. The authors thank Michele Hunter for technical writing.

	In addition, the authors thank the companies that provided input to the survey in 2015 and before: AbbVie, Almirall, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Biogen, Bristol-Myers Squibb, Eli Lilly and Company, Grünenthal GmbH, GlaxoSmithKline, Johnson & Johnson, Merck (MSD), Merck Serono, Novartis, Novo Nordisk, Pfizer, Roche, Sanofi, Servier, Takeda, and UCB.

	1. S. Rönninger et al., The GMP/GDP Inspections Landscape--Part II: Opportunities and Considerations, 

, 2017, in press.
	2. ICH, Quality Implementation Working Group on Q8, Q9, and Q10: Training Material, www.ich.org/products/guidelines/quality/training-programme-for-q8q9q10/presentations.html, accessed July 2016.
	3. ICH, ICH Harmonized Tripartite Guideline, Pharmaceutical Quality System Q10, step 4 version (June 2008).
	4. Amending Directive 2001/83/EC on the Community Code Relating to Medicinal Products for Human Use, Falsified Medicines Directive (1 Apr. 2011).
	5. Good Distribution Practice of Medicinal Products for Human Use, Guideline 2013/C 343/01 (5 Nov. 2013).
	6. Principles of Good Distribution Practice of Active Substances for Medicinal Products for Human Use, 2015/C 95/01 (19 Mar. 2015).
	7. EC, Good Distribution Practice for Medicinal Products for Human Use, Questions and Answers, Ares(2014)968163 (28 Mar. 2014).

The European Federation of Pharmaceutical Industries and Associations (EFPIA) has conducted an annual survey of GMP/GDP inspections at sites and affiliates among its member companies from 2003 until the present time. This article describes findings of the EFPIA annual survey of site inspections among its member companies.