Writing a Protocol

The background and rationale should be no more than 3 - 5 pages. It is appropriate to refer back to the Investigator's Brochure, the full grant, or 1 - 3 attached key references from the literature for more detailed information. Information about the study intervention, summary of the non-clinical data and clinical data, and a review of the literature are all part of this section. If the study involves the administration of a drug or biologic, a justification of the dosage and dosing interval should be provided.

A compliance statement is included in the template. This statement is not a review of the ethical issues raised by the study design. It simply is a statement that the research will be conducted in accordance with regulations and hospital and IRB policies.

The study objectives (or aims) should be broken down into primary and secondary objectives. The objectives need to be specific to the intervention - for a drug, they should specify the dosage level, route, frequency and duration of administration and in a defined population - "does the study drug reduce blood pressure when administered twice a day as an oral liquid, at a dose of 50 mg, in children age 6 - 12 years who have moderate hypertension". The objectives are usually written as bullet points.

• The primary objective is to determine (or evaluate) the... (efficacy, pharmacokinetics, safety, etc.)... of the study intervention...

The primary objective serves as the basis of the sample size. Secondary objectives may be exploratory or hypothesis generating and the study may not be powered to achieve these objectives.

Each objective has a corresponding endpoint and a corresponding analysis plan. The endpoints in the IRB's protocol templates are located in the analysis section. Some protocols include a table that maps each objective to its endpoint and corresponding analysis plan.

Objectives for a PK Study:

Objectives for a Clinical Trial:

Objectives and Endpoints in a Table:

General Schema of Study Design:

This section provides a brief overview of the entire study design including various phases of the study (if more than one). The details are provided later in this section.

Randomization and Blinding:

Should be discussed as an overview including how randomization schedules are generated, who executes the schedule, how the schedule is concealed and the study blinding procedures. The details of the execution of randomization and blinding procedures should be included in the Statistical Analysis Section (Section 9)

Study Duration, Enrollment, Number of Sites:

Describes the burden on subjects (duration of their participation) and should also address study feasibility. Research that is not feasible is not approvable.

Study Population:

Defines who will take part in the research. This section is mandatory and key for every study, regardless of design. The inclusion and exclusion (enrollment) criteria define the population that will be participating. The risks and benefits of the study depend on who the participants are and the enrollment criteria also must be constructed to ensure an equitable selection of subjects.

If there are any situations where there might be flexibility in the enrollment criteria, this should be stated explicitly in this section. Enrolled of subjects that do not otherwise meet one or more of the enrollment criteria constitutes a Major Protocol deviation that will need to be reported promptly to the IRB.

This section is a Visit Schedule which is visit-by-visit listing of all the procedures that will take place at each visit. If the study will have multiple procedures in one day (e.g. blood draws for a PK study) then the timing of each of these should be included.

The section on unscheduled visits may be applicable when patients are anticipated to require rescue therapy or might need interim care due to chronic illness.

The examples for the Study Visits to the right and below come from different clinical trials.

Concomitant Medications:
Should list all medications that are permitted or explicitly forbidden during study participation.

Withdrawal from Study:
Since subjects are permitted to change their minds withdraw from participation from a research study at any time, the procedures that will be followed to provide for an orderly transition from the research to routine care should be outlined. For studies without an intervention this section may not be applicable but subjects may not be able to abruptly discontinue some medications and a plan needs to be in place to manage this eventuality.

Screening Visit:

Day 0 Study Visit:

Day 3 and Subsequent Visits:

The Study Measurements should provide the detailed descriptions for how each measurement will be made including. For example, if blood pressure will be measured, it is important to state the subject"s condition (resting for at least 5 minutes), position (seated), arm (left) and technique (mercury sphygmomanometer)..

The section on unscheduled visits may be applicable when patients are anticipated to require rescue therapy or might need interim care due to chronic illness.

The examples for the Study Measures to the right and below come from different clinical trials.

Psychological or other measurement scales that will be used should be described. Those that have been validated and that are on the IRB's listing of Validated Instruments may simply be referenced. Those that are not on the list should be included in the appendix or uploaded into eIRB.

It is not necessary to include a copy of the case report form if all of the study measurements are listed in this section of the protocol. The protocol template also includes an alternative table format that may be easier for certain measurements such as laboratory tests.

Physical Exam:

Psychological Instruments:

Lab Tests:

Study Endpoints or Outcomes translate the study objectives into explicit statements that describe the comparisons to be made.

For each endpoint or outcome, the trial protocol should define four components:

• the specific measurement variable, which corresponds to the data collected directly from trial participants (eg, Beck Depression Inventory score, all cause mortality);
• the participant-level analysis metric, which corresponds to the format of the outcome data that will be used from each trial participant for analysis (eg, change from baseline, final value, time to event);
• the method of aggregation, which refers to the summary measure format for each study group (eg, mean, proportion with score > 2); and
• the specific measurement time point of interest for analysis.

If the objective is to determine the efficacy of Drug A compared to placebo for the treatment of hypertension, the study endpoint might be the change in the mean (method of aggregation) systolic blood pressure (measurement variable) for Drug A compared to placebo between Visit 1 and Visit 4 (participant analysis metric and time point of interest).

Sample Size and Power: All studies require a justification for the chosen sample size. Often the sample size is based on a formal power calculation. When this is the case, the planned sample size should be large enough to have a high probability (power) of detecting a true effect of a given magnitude, should it exist. When a power calculation is performed the protocol should include the following:

• the primary endpoint (outcome);
• the values assumed for the outcome in each study group (eg, proportion with event, or mean and standard deviation);
• the planned statistical test;
• alpha (type 1 error) level;
• power (usually at least 80%); and
• the calculated sample size per group - both assuming no loss of data and, if relevant, after any inflation for anticipated missing data modified from SPIRIT 2012 explanation and elaboration: guidance for protocols of clinical trials

Some studies, such as exploratory studies, studies of rare diseases or pharmacokinetic trials may have a sample size that is based on more pragmatic criteria rather than a power analysis.

Regardless of whether or not there is a formal power calculation, the Sample Size section should provide sufficient information to explain the why the study is proposing to enroll the specified number of subjects and not more or fewer. The number of subjects needs to be sufficient to enable the study to achieve its objectives without being any larger than necessary. The IRB is required to ensure that the proposed study is feasible and minimizes risk. A study that is not large enough to achieve the stated objectives is not considered scientifically valid. A study that is larger than necessary exposes more subjects to risk and inconvenience than required to achieve the scientific aims.

Example Endpoints:

Example Sample Size Section:

Analysis Plan provides the specific plan for how each component of the analysis will be performed. It is insufficient to merely list the tests to be used without including a discussion of exactly what comparisons will be performed. The Analysis Plan provides a justification for the data that will be collected. The IRB will not approve collection of extraneous data without a plan for how it will be used.

Inadequate analytic plans are the most frequent shortcoming of investigator-initiated research. Statistical consultation should be sought prior to submission of the protocol not after completion of the study. Descriptive studies are limited to summaries of the data (means, medians, standard deviations, etc.) and do not require a very detailed plan. The analysis of observational research study designs is usually more complex than for clinical trials. A plan is need to deal with all of the sources of bias and for confounding variables.

If the study will have an interim analysis, then the stopping rules for safety and benefit should be included.

How much detail should there be in the statistical analysis plan?

❡ The SAP should include enough detail that another statistician familiar with the data set (or their own independent data) could replicate the analysis. This implies that the SAP should delineate populations (exclusion criteria); end points; descriptive objectives; testable hypotheses; modifications or derivations of standard variables; statistical methods, including handling of missing data, correlated data, bias, and confounding; subgroups; interactions; and sensitivity analysis. While the SAP should be finalized prior to data analysis, authors may make changes to the analytic plan in response to subsequent findings. These changes to methods, hypotheses, or both should be noted in the SAP to capture when and why components were modified during analysis. Once created, the SAP should be cited in the methods, submitted along with the manuscript for review, and potentially made available as an online appendix to a published article.

From Thomas and Peterson's paper in JAMA 2012 The Value of Statistical Analysis Plans in Observational Research Defining High-Quality Research From the Start.

Example of Analysis Plan for Primary Endpoint:

Example Stopping Rule:

This section should include the details regarding the packaging, dosing, storage and accountability of study medications or other interventions. For studies that will test a device or diagnostic test, the information regarding the device or the test might be best placed at this point in the protocol, particularly if the device or diagnostic is being studied under an IDE, HDE or IND.

Description of Non-Pharmacologic Interventions:
Protocols (and clinical trial reports) frequently omit key descriptions of the study intervention. Enough detail should be present in the protocol to allow another investigator to replicate the study without consulting the investigator.

Table from Hoffman et al. Poor description of non-pharmacologic interventions. analysis of consecutive sample of randomised trials. BMJ 2013

This section is not in the Observational Template since there aren't any interventions.

Greater than Minimal Risk Safety Management:
A generic safety plan for greater then minimal risk research is included in the protocol templates. The plan includes the definitions of an adverse event, serious adverse event and summarizes the reporting requirements and timelines. A thorough discussion of reporting requirements for unanticipated problems involving risks to subjects and others, including Serious Adverse Events is available on the IRB webpage on Reportable Events.

Minimal Risk Safety Management:
When a study is limited to procedures that are not greater than minimal risk, the Safety Management section of the protocol may be simplified.

This section enumerates the procedures that will be used to run the research study. Sufficient detail is required to ensure that the research plan will produce valid results.

Treatment Assignment:
The CONSORT reporting standards require that the details of randomization and blinding be reported. These elements must therefore be in the protocol in sufficient detail to assure that the trial results will be unbiased. If the study is not a randomized trial, it may still be appropriate to include the details for how measurements, tests or radiological images will be blinded prior to interpretation.

Randomization Sequence Generation and Concealment Procedures:

Data Collection and Management:
Care in data management is a requirement to assure valid study results. Sloppy data handling will add variability to the study assessments and will affect the outcome of the study.

Regulatory and Ethical Considerations:
This section should identify the areas of risk to human subjects and address how the study will minimize those risks and maximize the potential benefits to subjects.

Informed Consent and Assent:
The protocol should provide sufficient information regarding will obtain consent, what information will be transferred (consent form, verbal script, video presentation, booklet, etc.), when the consent process will occur, where it will occur and how (in person, by telephone, etc.). The consent form by itself does not address the circumstances under which consent will be obtained. The steps that will be taken to assure comprehension (informed consent and not just consent) and lack of coercion should also be addressed.

Waivers:
The Informed consent section should also outline and provide a justification for all of the waivers (if any) that apply to the research. These include waivers of consent and assent, waivers of documentation of consent and waivers of parental permission. If HIPAA applies to the research then there should be an explanation of each requested waiver, alteration or partial waiver with a justification for each. It is insufficient to simply reiterate the regulations when justifying a waiver request; the justification needs to explain why the research satisfies the conditions for each waiver.

Safety Monitoring Plan:
Should address how those in charge of the management for the study will assess the evolving study progress including adverse events and data quality. Those in charge should assess the impact of unanticipated problems on current and future subjects. Provided that the study intervention(s) and monitoring procedures are of reasonably low risk, early Phase trials might have a plan as simple as having the investigator review all of the adverse event reports as they occur. Studies that involve a life-threatening condition or late Phase 2, 3 or 4 studies may require an independent Data Monitoring Committee (DMC), which are also referred to as Data Safety Monitoring Boards (DSMBs).

Payments to Subjects:
If subjects or parents/guardians are to be paid for the inconvenience of participating in the study, the amount of payment(s) must be stated in the protocol. The amount paid to parent/guardians should be separated from the amount paid to subjects. The IRB must review both the amount and method of payment to subjects to insure that neither presents an undue influence on the trial subjects. Subjects not completing the study, for matters of choice (subject or investigator), may be paid on a pro rata basis. Subjects that are forced to discontinue due to an adverse experience, should be compensated the entire amount.

Research that will not be published does not contribute to generalizable knowledge. This violates the first of Emanuel's requirements for what makes clinical research ethical - namely that the research have social (and scientific value). The plans for publication should be provided as a statement of the investigators assurance that the results will be made public and shared.

1. Mann H. Research ethics committees and public dissemination of clinical trial results
2. Antes and Chalmers. Under-reporting of clinical trials is unethical
3. Chalmers I, et al. All trials must be registered and the results published
4. Decullier E et al. Inadequate dissemination of phase I trials: a retrospective cohort study
5. Drazen JM. Transparency for clinical trials - the TEST Act
6. Outterson K. Clinical trial transparency - antidote to weaker off-label-promotion rules?

List key references. Generally this should include no more than 10 - 20 key references that demonstrate a thorough review of the literature and provide support for the methodology, dosage choice, measurement techniques, etc.

The Investigator's Brochure or the Grant application generally provide extensive reference material that need not be repeated in the protocol.