An Ideal SOP System


Applied Clinical Trials

A "can’t live with them, can’t live without them" attitude is revealed in survey of standard operating procedures in the management and conduct of clinical trials.

This article aims to define guiding principles and to establish a framework to set up an ideal standard operating procedure (SOP) system. Only a few articles dealing with what constitutes ideal SOPs have been published.1 Rather than examining a few SOPs in a certain area or just the documents that carry such titles, it may be advantageous to take the entirety of all documents that describe relevant processes and call this an “SOP system.” SOPs are not usually standalone documents, but they are linked to each other and to other systems, such as policies, handbooks, guidance papers, or instruction documents. The definition of those items unfortunately varies and are used by different companies or institutions in different ways. In some cases, policies are above SOPs in hierarchy, which, in turn, are above guidance documents and, lastly, instruction papers. Common to all is a hierarchical structure of governing and dependent documents. An SOP system is described as the entirety of all above-mentioned documents that include the processes to generate and alter, distribute, train, implement, and maintain them. One publicly available example is The European Medicines Agency’s Integrated Quality Management System, which consists of six hierarchical levels, starting with the quality policy and the mission statement on level 1 down to “documents resulting from the process and needing to be controlled and stored in accordance with relevant agency policies and SOPs.”2

A survey3 conducted by a subgroup of the German Association of Research–Based Pharmaceutical Manufacturers (Verband Forschender Arzneimittelhersteller, vfa) has shown that in the pharmaceutical environment, the number of SOPs and related documents is increasing. There is no indication that the objectives, such as quality or efficiency of underlying processes, are achieved or improved this way. Also, there is a large variety both in size and structure of the SOP system across companies. The survey revealed that there are major deviations between what the SOPs describe and how processes are implemented in real life. Several hypotheses/questions should therefore be explored:

  • Is there a lack of knowledge due to the large amount of SOPs that are produced and changed frequently?

  • Is there a lack of discipline to adhere to SOPs?

  • Are the SOPs simply inadequate, leading to the need for deviations?

  • Are deviations tolerated actions to avoid major problems in these situations?

  • Are SOPs too complicated to allow timely reactions to problem, thus procedural shortcuts are implemented?

  • Are decisions not taken, although the process to arrive at them has been described in the SOP?

  • Does management not care about supervising the execution and adherence of SOPs?

Within this survey, we learned that an ideal SOP system does not exist. Ways of thinking, expectations, and need for guidance differ considerably between functions.

Experienced people require broader margins for individual decisions while those with limited experience ask for more and stricter guidance.

The survey consisted of two sets of questions, one of which was directed to the management of clinical research, the other to the operations staff conducting the clinical trials (see charts at left and below; click on each to enlarge). In summary, the answers showed that this survey is a good starting point to generate ideas on how an ideal SOP system could look.

While we shall briefly discuss the characteristics of

ideal processes, the focus will be on the ideal SOP system-which requires that the individual processes are efficient by themselves. Regarding optimality, the requirements were, in summary, as follows ahead. It should be considered that there was a wide range of opinions, and this summary simply constitutes their majority. Our ideas on how to create an “ideal” SOP system are also vindicated by the results of the survey. The SOP system should cover all regulated processes, but only in a degree that allows optimization of processes easily.


‘Ideal processes’

The definition of ideal processes is taken from business process excellence (e.g. ASQ4, EFQM5). There, an ideal business process is seen both from the outcome and the necessary inputs.

Efficiency is defined as the optimal ratio of input vs. output. SOPs should reach their aim without the need to consume many resources. In fact, they should, through standardization, lead to fewer resources being consumed when achieving the same level of quality. Such questions could be answered by benchmarking with other organizations, or by consensus processes. It will be difficult in practice to find the minimum level of effort needed to achieve the necessary quality. In site monitoring, thresholds can be given; however, to transform this into an SOP is difficult, as there may be several different routes to arrive at this goal, and not all of them can be described in detail. Effort must be taken to measure efficiency in a reliable way in order to follow up on progress.

However, other criteria such as simplicity and adaptability have also been quoted. The following paragraph attempts, on the basis of experience, to develop a set of guiding principles. These will not be hierarchically ordered, however, some may carry more weight than others depending on the circumstances.



Guiding principles for SOP systems

1. Necessity/need

SOPs should only be defined where necessary. If they focus on rare situations, or situations which are not important for the quality of the study, they remain useless. Such situations should be managed depending on the circumstances of the individual case.

  • Is the regulated process important?

  • Is it frequently occurring?

  • How often is it executed? Does it bear a significant risk?

  • Is this quality level really needed?

  • What would happen without regulated process?

An analysis should be performed as to which processes need detailed description and standardization for legal or regulatory reasons, which guides risks that need to be managed and to what degree, and whether SOPs are an adequate tool to do so. In some instances, the risks are minimal in comparison to the effort needed to avoid them.

Such a risk analysis describes the magnitude of risk, its probability of occurring, and the risk reduction through SOPs (and other measures) in two key dimensions: patient safety and data integrity.

In practice, the identification of processes without a need for an SOP is challenging. This is also reflected in the results of the survey that showed that “the vast majority was unable to identify SOPs that could be done without.” In consequence, the focus should be on how detailed a process must be covered by an SOP and/or which process steps can be eliminated in order to achieve the needs of the different users/user groups.

2. Expediency/suitableness/effectiveness

Does this regulated process do what it is intended to do? This may also be called functional reliability or effectiveness. SOPs should be written in such a way that the objective of the underlying process is supported. One needs to avoid formal hurdles that hinder the objectives of the process being reached. One example is the reporting of serious adverse events (SAEs), where the need to report is more than obvious and the SOP, including any templates/forms, needs to provide an easy way as to how to best support the reporting. Every element, the SOP itself, the forms and the templates, needs to support a smooth reporting process and should not add problems to it.

Effective SOPs describe the processes in a way that their efficiency is not impaired by lack of clarity or understanding, and that they achieve the deliverables with a minimum of effort. Implementation of a new SOP, or SOP revisions should follow a pragmatic approach. If the need is there, implementation is more accepted by the users.

3. Simplification, transparency, and ease of acquiring

KISS-“keep it short and simple.” An SOP should describe the standard situation of a process. Therefore, special rules and exceptions should be avoided. SOPs need not consider all imaginable situations. SOPs and associated workflows should be kept as simple as possible. As a rule, an employee should be able to easily memorize the relevant processes. Critical processes such as safety reporting should be easily available for everybody and everywhere. For less important processes or areas, an overview should be easily accessible for staff on where to find the respective SOPs. To support easy access a logical and easy to remember numbering system can be of great help.

As an example, the EMA groups together documents according to their scope,e.g.: 

  • 0001-0999 EMA (cross-agency)

  • 1000-1999 PDM (Product data management)

  • 2000-2999 INSP (Inspections)

  • 3000-3999 H (Human)

4. Stability and predictability

“Less is more.” Changes to SOPs should only be made for two reasons: if legal or regulatory changes mandate it or if major process improvements can be gained. Therefore, it may help to avoid detailed information, which are not process relevant but subject to frequent changes, for example, version numbers of supporting documents. In addition, highly variable elements, such as templates, may be attached to an SOP to be adapted without changing the SOP itself.

A regular review cycle of the SOPs (we suggest two to three years) should be established in order to evaluate whether modifications of processes or changes in the organization have taken place and are not covered by the SOPs.

5. Global reach

As almost all registration studies are multinational, SOPs need to be globally usable. Therefore, global SOPs should describe all globally defined processes to ensure harmonization and efficiency across the whole organization. However, regional or local amendments to global SOPs need to be possible but should only be introduced if required by regional/local law/regulations or organizational structures of affiliates.


Pitfalls of SOP improvements

1. Size of SOPs

There is an ongoing misconception that the size of SOPs is critical for the acceptance of the user. As we learned from the survey, more important than the size of an SOP or SOP system is how to navigate within and how the connections between them can be highlighted. This may explain why the satisfaction with the SOP system is not associated to its size.

2. Number of SOPs

Simplification of SOP systems often aims to reduce the number of SOPs based on the misbelief that the number is an important indicator of the quality and usability of the system. Respective activities are merging several SOPs into one document, renaming of SOPs to, e.g., working instructions, best practice documents, functional guidelines, and monitoring manuals. This is not the way to handle the problem, as the results of the survey pointed out that the number of SOPs does not matter at all.



3. 'Parallel SOP world'

There is a strong tendency to have, e.g., best practice documents outside of the official SOP system. This is born out of the misbelief that these documents are not in the scope of audits and/or inspections. According to ICH GCP, chapter 1.29,7 the inspector can review every document deemed relevant.

4. Creation of a ‘perfect SOP system’

SOP improvement projects almost always build on an existing system. They are adding further requirements and restrictions to already overregulated processes, with the aim of coming as close as possible to a perfect solution. Very often the current system is not questioned and, therefore, complexity is increased, slowing down changes, e.g., prolonged update cycles. PARETO rule (80:20)8 should guide us also in this area.

5. Extent of SOP distribution

It is a misconception that distributing SOPs to a broader audience allows them to understand the big picture. This leads to an information overload, which can lead to frustration and resistance. The risk may be that relevant items are diluted by the vast amount of information.

6. SOPs do not 'stand alone'

The assumption that following the SOP ensures quality in itself is dangerous. The processes described in the SOPs need to be transferred into the individual daily business activities. It still requires lateral thinking and does not replace interactions and communications in specific situations.

7. SOPs do not claim the 'perfect solution'

The assumption that the SOP describes the perfect solution is misleading. The aim of an SOP is to provide a reliable, robust, and standardized way of reaching a defined outcome. “All roads lead to Rome.”9 There are always several potential routes to getting a result. The SOP defines one route for which we know how to manage hurdles and challenges.

8. Number of signatories

To ensure commitment and acceptance of SOPs, often all concerned line managers are included in the sign-off of the SOP during the release cycle. This is based on the assumption that signing an SOP increases acceptance and commitment of the user. Finally, it prolongs the release cycle and slows down the update of SOPs. Thus, the number of signatures should be minimized. Nevertheless, process modifications always have to be aligned with the relevant functions.

9. Can everyone write SOPs?

Writing an SOP is challenging and requires deep insight into how to analyze processes and how to describe and visualize them. The SOP system in its entirety should appear in a harmonized and standardized format. Preferably an SOP writer is recommended, who acts as moderator and even as change manager during the SOP generation process.



1. Format of an ideal SOP

An SOP should follow the logical flow of a process and describe the associated roles and responsibilities. Flow chart(s) could be of added value to provide an easy and quick overview. This can be implemented using business process modeling software tools, which can help identify shortcomings in the underlying process.

Any process step should have one responsible role that performs the activity. Further contributors can be visualized using a RACI matrix (R=responsible, this function does the actual work, A=accountable, this function supervises and decides; C=to be consulted; I=to be informed).10,11

If further governing or dependent documents are associated with SOPs, e.g., forms, templates, standard letters, related SOPs, checklists, or literature references, they need to be clearly referenced. Interfaces to other SOPs should also be mentioned. The interfaces may transgress the boundaries of a department, and may, therefore, have to be aligned with processes in other departments. Preferably a process should be described in one SOP only (end to end). If hyperlinks are used, the selection of the current version should be ensured to avoid outdated links. In addition, the SOP must contain transition rules, and version change details.



2. Who should write SOPs?

SOPs should be written by specialized/professional SOP writers who interview the process owners to translate the steps into a well-arranged process flow. Clarity, simplicity, and intuitive logic should guide them.

As alternative, specialized SOP facilitators (i.e., experts who support the subject matter experts) can be used when the process owners try to write the SOPs as they do it. In any case, the overall responsibility for the appropriateness of the business processes resides with the operating functions.

3. Implementation and training

Who should receive SOPs? Only persons that appear in the RACI chart as “R” (responsible: the one who does the work) or “A” (accountable: the one that supervises and makes decisions) should be recipients of respective SOPs.

Implementation of SOPs/updated SOPs always includes training and should therefore be reflected in the individual function training plans. The type of training is determined by complexity of changes and criticality of underlying processes. This can be self-reading (read and acknowledge), e-learning, web seminar, or face-to-face training. Training should be completed ideally prior the SOP coming into effect. However, the very latest point of training completion should be before the activity described in the SOP is performed by the individual. This needs to be documented and may be, for example, covered by a learning management system.

Independent from type of training, a plausibility check should be performed at least as to whether the training has been effective:

  • In a web session or face-to-face meeting, the use of the TED system during training maintains anonymity but shows how much has been learned and what may be sources of misunderstanding.

  • For online training courses, a quiz requiring a predefined success rate can be useful.

  • For self-reading, the confirmation of ‘read and understood’ might be sufficient.

The overall training effect has to be supervised by the line manager to verify that the training positively influenced the daily business. Another option could be assessments some months after the SOP becoming effective.

4. Control of SOP adherence and continuous improvement

There are different approaches to validating and developing a system. From within the own organization, feedback and proposals of users will help as much as the thorough analysis of results from audits and self-assessments and quality checks, especially with regard to SOP violations. From external sources, the result from regulatory inspections as well as SOP violations by external partner (e.g., investigators) are noteworthy. In larger organizations, an electronic tracking system will become necessary.

In each instance, it is necessary to perform detailed root-cause analysis to differentiate, for example, between an inadequate process/SOP and an appropriate process/SOP but with major problems in training and discipline in adhering to the SOP. Use of the “five whys”12 is often helpful in order to identify the underlying problems.

This information needs to be fed into a continuous improvement cycle (PDCA)13 in order to optimize the SOP system.



Discussion and conclusion

Obviously, our project to achieve an “ideal” SOP system is very ambitious. We’ve learned from the survey that the majority of users feel SOPs are too extensive, too restrictive, and too complex. Nevertheless, at the same time, they consider SOPs useful, necessary, and satisfactory. Consequently, there are competitive endeavors and requirements in both directions of enlarging and reducing levels of detail of an SOP system (and a lot more actions). At the same time, nobody is willing or ready to waive a process or standard procedure. Thus, it seems that there is no easy fix.

Knowing that, we chose the basic approach to make efforts toward improving SOP systems more rational. It is the nature of continuous improvement that volume and complexity of SOP systems increases, because details are supplemented in order to comply, as a rule, with audit or inspection findings or to implement more complex requirements. Most often we lack the courage and not only time and effort to redesign an operating process or the SOP system from scratch.

We are convinced that challenging our “improvements” with respect to necessity, effectiveness, clarity, simplicity, and reliability can help us to optimize our SOP systems pragmatically.

Sometimes we will reach the outer limits of the doable and sometimes only the courage to challenge the fundamentals of our systems will open new options.

Many of our ideas are characterized from the perspective of large sponsor organizations. But they may be transitioned also to smaller entities. A special case, and thus not reflected in our recommendations, may be so-called globally outsourced trials, where the definition of deliverables is advantageous when working with CROs that may have different processes. Time, cost, and quality should define the desired end result and ensure that legal and regulatory principles are not violated.

In addition, it may be considered even within an “own” SOP system whether a description of the deliverables (final state of outcome or quality level) should complement a precise description of processes in a SOP. In reality, a combination of both approaches may be needed. For instance, to make the degree of control of protocol deviations meaningful, the allowable percentage of major and minor deviations could be described.

The SOP system should always be part of a quality management system. Therefore, also structurally, other means of assuring and promoting quality should always be considered. The development and use of SOPs should be aligned with all those other processes, such as audits and inspections, quality management reviews, and self-assessments. Ideal SOP systems require that their use is not associated with information overload and complexity, as this will cause an increase in errors. Therefore, some principles have been described, which can be checked against when changes in the “own” SOP system have been contemplated. Finally, an ideal SOP system is ideal at a certain point in time, and needs constant renewal to stay like that.


Gabriele B. Schmidt, PhD, Country Clinical Quality Manager, MSD Sharp & Dohme GmbH; Dieter Baier, PhD, General Secretary, Medical and Scientific Excellence, Roche Pharma AG; Arthur Hecht, Quality Medicine, Boehringer Ingelheim Pharma GmbH & Co. KG

* In memoriam: This article is dedicated to the memory of Dr. Michael Herschel (1953-2014), who initially chaired the team of authors.



1. The Ideal SOP, Posted on: 09 Feb 09, Jo Hills - Field Reporter, retrieved July 15, 2015; [cited 2015-Dec-03] available from

2. The European Medicines Agency’s Integrated Quality Management System POLICY/0001 EMEA/MB/355781/2007 Rev.1; 10-MAR-2009

3. Gabriele Schmidt, D. Baier, Arthur Hecht, and M. Herschel, SOPs in Clinical Research, Applied Clinical Trials July/August 2013.

4. ASQ - American Society of Quality [cited 2015-Dec-03]. Available from

5. EFQM - European Foundation for Quality Management [cited 2015-Dec-03], available from

6. Creating and maintaining SOP/WIN folders in EDMS WIN/EMA/0035, 09-Jul-2010

7. ICH Harmonized Tripartite Guideline, Guideline for Good Clinical Practice E6 (R1), current step 4 version dated 10-Jun-1996 [cited 2015 Dec 03]. Available from

8. According to Vilfredo Pareto (1848–1923) [cited from Wikipedia 2015-Dec-03] ; Pareto, Vilfredo; Page, Alfred N. (1971), Translation of Manuale di economia politica ("Manual of political economy")

9. McGraw-Hill Dictionary of American Idioms and Phrasal Verbs, © 2002 by The McGraw-Hill Companies, Inc.

10. Smith, Michael L., James Erwin, and Sandra Diaferio. "Role & responsibility charting (RACI)", abgerufen am 17 (2007): 2007, [cited 2015-Dec-03] available from http://www. 

11. Morgan R. How to do RACI charting and analysis: a practical guide, 22-Sep-2008 [cited 2015-Dec-03] available from

12. Duke Okes: Root Cause Analysis: The Core of Problem Solving and Corrective Action, 2010

13. Deming, W.E.: Out of the crisis. Massachusetts Institute of Technology, Massachusetts, USA, 1986

Related Content
© 2024 MJH Life Sciences

All rights reserved.