Running protocol development
A clinical study protocol is produced by a team whose experts each hold part of what it must say, with a medical writer turning their input into one document. Organising that work covers whether the team settles a short study concept before the full protocol is drafted, how far the writer's part reaches into the design, how writing teams fit an institution's approval process, and the timelines, tools, templates and checklists that govern drafting and review.
Why the protocol needs a writer
The protocol is the recipe for performing a study, an instruction book that tells the doctors and nurses responsible for a clinical trial how to perform it. A trial may have 50 investigator centres around the world, and all of them have to perform the study in exactly the same way, because if each does it a little differently the data will not mean anything at the end. The way to ensure that every centre performs the trial in the same way is a very well-written protocol that is extremely clear and contains no errors or sources of confusion.
The knowledge it sets down is spread across the team. The clinical project manager knows how the study should be performed, the medical director knows which patients should take part, what dose of the drug they should receive, how often and which assessments should be performed, and the statistician knows how the data should be analysed. These people may not be good at writing things clearly enough for the doctors to understand how to perform the trial, and the medical writer obtains the information from each of them and creates one clear, well-written document. The patient informed consent form is written in parallel with the protocol.
A protocol team normally includes a medical and clinical pharmacology expert, a statistician, a pharmacokinetics expert, regulatory consultants, a study manager, a project manager and a medical writer, whose role is to coordinate the input of all the experts to compile the protocol. Planning can avoid many problems and delays, and the efficient production of protocols requires good teamwork, local or international, that makes optimal use of each member's expertise under experienced leadership.
The clinical study report has been part of the classical medical-writing repertoire since medical writing became a professional discipline, while the protocol reached medical writing's radar considerably later and in many organisations still has not. The protocol is a document type that can greatly benefit from medical-writing expertise, both in the organisation of the complete document and in the streamlining of the process that develops it.
Maturing the study concept before the full protocol
Medical-writing participation in protocol development is not common practice, and writers who take part have mixed experiences. For too many writers, membership of a protocol team can amount to weeks or even months in a torture chamber. Many find themselves producing an endless series of draft versions of the full protocol to accommodate continual input from the team, only to be surprised by a decision after the tenth draft to add a further treatment arm, a preceding wash-out period and three more visits during the treatment period, and to change the statistical approach from superiority to non-inferiority on the basis of a redefined primary variable. Given the complexities and redundancies of a full protocol, implementing such modifications is a high burden.
Almost every trial matters greatly to its sponsor. A clinical study is typically a substantial investment of resources, sometimes to the limit of the sponsor's capabilities or beyond, and occasionally the sponsor's economic survival may depend on the positive outcome of a single study, so the burden on the people who make the design decisions may be enormous. Multiple stakeholders within the sponsor, potentially representing conflicting positions, may want a say in the objective, design and setup of the study, and getting everyone to agree on and commit to one final protocol can be a challenge. The sponsor's internal thought maturation is therefore rarely a straightforward path, and the team may frequently change its mind and even turn repeatedly in circles. There is nothing wrong with this in principle, and it should not be fought, but it should take place on the basis of the right document type.
That document is a study concept, also called a protocol outline, and whatever it is called, it has two features. It is restricted to the main medical-scientific content items, in particular the objectives, the key inclusion and exclusion criteria, the treatment arms, the complete schedule of activities, and the definition of the main variables and their statistical analysis. More importantly, it is almost completely devoid of the built-in redundancies of a full protocol, so capturing multiple substantial content modifications in it is easy, while incorporating them into a complete protocol can be a real nightmare. All maturation steps of the main study features should take place with the study concept as the basis for discussion. Maintaining a brief, concise study concept that reflects all the team's decisions, is a valuable service medical writing can offer, and collecting and consolidating sometimes conflicting contributions efficiently is an expertise it routinely provides.
Work on the complete protocol should start only once the study concept is finally approved, and that agreement should include input from external sources such as key opinion leaders as well as approval from the sponsor's internal governance bodies. Ideally, no comment raised at the final review of the full protocol could already have been raised on review of the outline. The study-concept approach is strongly recommended even when a new study will be very similar to a completed one, rather than starting on the full protocol by editing the earlier study's approved protocol.
Many protocol teams want to see a complete draft protocol sooner rather than later, so an education process is needed to raise their comfort with the study-concept approach. The main arguments for it are that the information in a study concept suffices to support practically all discussions and decisions on the main study features, and that expanding an approved study concept into a complete protocol is very straightforward and can be done by a medical writer very quickly, so the approach does not extend the overall development timeline.
Where the writer's part stops
Writers are sometimes asked to help develop the study design and the statistical section. One medical-writing company usually declines both: the statistical section because none of its writers are statisticians, and the study design because they are not doctors. One experienced writer there holds that key opinion leaders and investigators, not medical writers, should be involved in the study design. What the writers can do is review the literature, look at how studies in the field have been designed and make recommendations to the client, such as reporting that all the published studies use a particular primary efficacy criterion and particular secondary ones and recommending the same.
The same company expects the client to provide a final version of the protocol synopsis before it starts writing the full protocol. In reality clients often provide a synopsis that is not final, and the writers find themselves involved in the study design during the writing.
Protocol writing teams in research institutions
The National Cancer Institute's Division of Cancer Treatment and Diagnosis established a Project Management Office in 2006 to integrate formal, business-focused project management practices into early-phase drug development. A Writing and Editing Team within the office was responsible for clinical protocol writing and support, manuscript writing and editing, and project management document support. To let clinical investigators devote more time to research and patient care, the team collaborated with the principal investigators of the Developmental Therapeutics Section of the Medical Oncology Branch to draft clinical protocols and protocol amendments, prepare submissions and continuing reviews, and shepherd protocols through the institute's review committees.
The team developed a streamlined quality assurance and quality control process that documented every protocol preparation and submission step required for approval. Heavy emphasis on good writing and editing ensured that the protocols and consent forms sent to review committees contained all the required regulatory information, so reviewers could focus on the science and ethics rather than the format, which minimised the number of stipulations and re-review cycles. Through careful planning the team anticipated submission deadlines and responded quickly to review committees, which led to faster approvals. These strategies were essential to its management of more than 15 clinical protocols active in the clinic and about 5 in development, and a web-based tool, Project Web Access, facilitated its communication with principal investigators, project managers and senior management.
The National Institute of Allergy and Infectious Diseases founded its Protocol Development Program in 2009 to deliver early interventions in protocol development by expediting processes, managing administrative and regulatory requirements, and providing writing support within the institute. Its team consists of protocol navigators, who assist with the logistical aspects, and medical writers, who write and edit clinical protocols from study concepts (hypotheses, study objectives and design), informed consent documents, standard operating procedures and publications.
The steps and milestones of the institute's protocol development process were mapped using the Six Sigma approach, into a functional flowchart of 48 steps and 8 milestones involving 12 departments or functions, and the medical writer plays an integral role in 16 of the 48 steps. The process begins once a concept of the planned protocol is approved and timelines are set to accommodate the needs of the study. The medical writer collaborates with the principal investigator to draft the protocol, which is then submitted for an intramural scientific review. Following approval, the writer drafts the informed consent, and all the documents are submitted for a pre-institutional review board review and an institutional review board review, and also for a simultaneous review by other committees if the study design warrants it. Throughout the process, the writer works closely with the principal investigator to help address the comments and stipulations from the review committees, while the protocol navigator ensures that all the logistical and regulatory requirements have been met. The programme successfully implemented a new approach to improving processes for the institute's intramural research, and the medical writer plays a pivotal role in it.
Timelines
Medical writers routinely face accelerated timelines to complete drafts, and realistic schedules are often a luxury. Lack of coordination among team members or any lag during development could lead to a cascade of events that may result in study delay, so sufficient time needs to be invested in the protocol to account for multiple reviews, edits and adjustments after reviews. When sufficient time is not allotted, protocol development is likely to become excessively long, because truncated initial timelines require more extensive editing or review, and the content could be repetitive, verbose, ambiguous, inconsistent and riddled with errors. A lean writing process that emphasises timeliness and quality is beneficial for producing a clear and concise document with a well-organised message.
Clear descriptions of the roles and functions of each team member are vital, so that each member understands the contribution required of them and so that the time needed for that contribution can be determined. Alongside the multidisciplinary team that contributes the content, the editorial and regulatory publishing teams manage the redactional quality of the final document. A standard, full-process timeline streamlines every part of the process, from synopsis preparation to the published document, and guarantees fewer process conflicts. Standard timelines have evolved over time, each step refined by the collective experience of colleagues across departments and functions, and they have built-in flexibility for variable workloads, since each team member may work on several projects at once. Each step is divided into rounds, and development generally proceeds from drafting to review or quality control and then to drafting again, where review comments can be addressed. Standard timelines may need additional flexibility, such as an ad hoc meeting to address unresolved issues. The drafting step ends with the approval of the protocol, which is followed by the publishing of the common technical document in a format consistent with regulatory submissions, such as the ICH CTD format.
ICON's standard protocol development timeline includes 10 to 15 days to prepare the synopsis, 47 days for the body of the protocol and 3 days to publish the CTD-ready document. The synopsis is not always needed, because the client may have prepared it during the conceptual development of the study, and the client may prefer to publish the protocol through another publishing service; the standard timeline then includes only the 47-day protocol preparation. Adding a synopsis and publishing changes the timeline but not the protocol development process.
A 10-question online survey asked 75 medical writers, 54 (72%) at ICON and 21 outside it, which timeline-saving tools and approaches they used most; 41% of them were early-phase writers, 37% late-phase writers and 21% specialised in other areas such as manuscript writing. Most said they needed 5 to 7 weeks (49%) or more than 7 weeks (35%) to develop a protocol, including draft preparation, review cycles, administrative tasks and finalisation. For each review cycle, 51% allocated an average of 3 to 4 days, 46% needed 5 days or more and 3% needed 1 to 2 days.
To meet aggressive timelines, 50% of the writers had worked collaboratively across time zones to develop a protocol, and they believed the approach helped reduce development timelines. The key factors for a successful collaborative approach that they selected were continuous work around the clock, parallel and continuous quality control of the work, and a good relationship among the authors. The primary factors they named as able to harm collaborative writing were lack of communication, conflicting opinions and lack of flexibility. Efficient collaborative work across time zones had a marked impact on shortening timelines while maintaining high-quality documents.
Software tools
As the number of content contributors grows, so does the potential for inconsistencies and errors, which affects the efficiency of document review and the overall schedule. Document review and quality control checks can take hours, days or even weeks depending on the complexity and quality of the document, and they are typically the most expensive and time-consuming part of protocol development. Lean writing workflows and software applications can reduce the time lost in review without sacrificing document quality.
A proofreading add-in for Microsoft Word, such as PerfectIt, takes over some of the most tedious and time-consuming editing. It catches inconsistencies in hyphenation and numbers, checks punctuation and capitalisation, corrects double spaces, removes spaces after paragraphs and before punctuation, and ensures that every abbreviation is defined only once, at its first presentation. It saved time in hunting for typographical and grammatical errors and style inconsistencies, giving a higher-quality draft and less overall quality control time.
In large documents with many references, references can be lost, moved or unlinked from the body, and style inconsistencies can appear as contributors add sections. A citation manager such as EndNote lets writers insert in-text citations while it organises the bibliography in the preferred style and updates and cross-links citations with references throughout drafting. It keeps bibliographies and literature in searchable libraries, saves references from databases such as PubMed or Google Scholar, and shares citation libraries across a team. It saved time by reducing the writers' referencing workload, including tracking and formatting citations throughout large documents.
With many authors and reviewers, reviewers may give conflicting comments, and consolidating comments across documents and controlling which edits are made is time-consuming. An online platform for co-authoring, reviewing and consolidating comments, such as PleaseReview, lets reviewers review simultaneously in a single interface, see each other's comments and discuss them. It saves the entire review process in an audit-ready log, and the comments and proposed changes can be found in the Word file as tracked changes. The author has full control over the proposed edits, and no changes or comments can be lost or overwritten by other reviewers. It also gives an overview of the process, including each reviewer's review status, reconciliation reports and summaries of all comments and actions.
Inside ICON, most surveyed writers used PerfectIt (91%), PleaseReview (53%) and EndNote or another reference manager (49%) during protocol development, while outside ICON EndNote was preferred (63%), followed by other software and PleaseReview (37% each) and PerfectIt (16%); these differences likely reflect company preferences. Of the writers who used one or more of these tools, 45% considered an application such as PleaseReview more helpful than the others for protocol development and reducing timelines (52% at ICON and 29% outside it), likely because it directly addresses two of the negative factors the writers identified, conflicting opinions and communication. The surveyed writers estimated that the tools together can help save up to 35% of the time needed to develop a protocol. Software tools are widely accepted by medical writers, and 92% of the surveyed writers already used the tools.
Templates and checklists
At ICON, the writing team is responsible for finding an adequate protocol template as the basis for development. Options such as the TransCelerate common protocol template and the National Institutes of Health template incorporate feedback from multiple industry stakeholders. Templates may vary with the foreseen study design, study population or therapeutic area. A template usually contains already-approved regulatory language, standard laboratory assessments, inclusion and exclusion criteria, a study diagram and one or more schedules of assessments, with sections prefilled with the most common and routine assessments and language. Prototyping and project-specific protocol templates, such as a synopsis structure and consistent table and figure formats for both clinical pharmacology and therapeutic studies, can greatly facilitate the rapid production of good-quality protocols. Standard text components are also valuable, such as background information on the clinical indication, the nature of the compound, the drug development plan and the main nonclinical information. To save time, standard templates that follow the regulatory guidelines should be developed.
A protocol checklist can reduce content errors across sections. It serves as a tip sheet for writers and peer reviewers working on early clinical protocols, and it can help ensure consistency of content across sections, identify sections prone to repetition and reduce the likelihood of amendments. Writers of early clinical protocols in one group use it to check their own work before delivering drafts for team review and as a last self-check before finalisation, and its use has proven to reduce time during document development and quality checks. Many of its tips come from inconsistencies uncovered in past protocols after finalisation or raised during the development of the informed consent form.
Compressed timelines in practice
ICON's medical writing team regularly responds to sponsors' requests to compress standard timelines and deliver high-quality protocols sooner, and it uses software tools, templates and checklists and works collaboratively across geographically dispersed teams to do so. In five real phase 1 protocols of moderate complexity, two drug–drug interaction studies, two single- and multiple-ascending-dose studies and a food-effect study, a collaborative writing approach was used throughout, and the writers had 5 or more years of experience and used one or more of the tools. In these cases, protocol development time under reduced timelines fell by about 26% compared with standard timelines, without compromising the quality of the document. The five reduced timelines averaged 26.5 days, with a range of 18 to 33 days. Most of the surveyed writers (62%) had expected a reduction of 25% or less, so software tools, templates, checklists and collaborative writing had more impact than they expected. With lean workflow processes and useful tools, medical writers can reduce the time and personnel hours each step takes, and with clear processes, roles and responsibilities backed by such tools, they can overcome common development challenges.