Writing a study protocol

A project involves an extra effort during a limited time. After the project is over we want to know if what you did was so good that it should be incorporated in the routine of our health care, school management, business (or whatever). If the answer is yes it might result in funding taken away from another part of the business to give priority to your suggested new routines. Different ideas for improvement of the business will be evaluated against each other. In order to make these decisions each suggestion for improvement need to be properly evaluated.

The most important part of planning for an evaluation of a project takes place way before data collection. This webpage describes what needs to be in the written presentation (study protocol) of a project before data is collected. (Writing a manuscript after data is collected and analyzed is described in another webpage.)

Why a written study protocol / project plan?

There are two main reasons to write a carefully planned study protocol:

You need a construction plan!

You need a construction plan!

  • For you as a project manager.
    A project can be viewed upon as the construction plan when you are building a house. There are those who manage to build a complete house from scratch without any construction plan. However, it is not recommended. In almost every project new interesting questions appear along the way. When this happens it is useful to have a look at the construction plan and read what you have written there. It must be very strong reasons to change the construction plan after the building process has commenced. New interesting ideas appearing along the way is not uninteresting. On the contrary! They can often be good ideas for the next project. The most important tasks for the construction plan is to set boundaries for the project, refine the research questions and to ensure that projects finish in a reasonable time frame.
  • For those who will assess your project.
    Those who must assess your project needs to be able to familiarize themselves with it. It could be a funding body, your employer, an ethics board, etc. Many of these have some questions they will ask. Your project plan should be designed to answer the most common questions expected to be raised. You will risk being knocked down or have an lengthy approval process if you don’t address these issues from the start.

It is advisable to finalize the study protocol before starting with the ethics application, grant applications and other approvals that may be needed. Working on the study protocol and the ethics application in parallel often ends up with some mismatch between the study protocol and  the ethics application, and the ethics committee will likely pick up that causing a delay.

Important sections in your study protocol

Title page

Write a catchy title, names of involved persons & institutions and finally contact details to one of the principal investigators.

Introduction / Background

  1. Begin with describing the topic and why it is an important topic. You may want to mention prevalences and / or costs associated with this topic.
  2. The rationale for a project is always to solve a problem. In this paragraph describe the unresolved problem associated with the topic. The problem may be (a few examples):
    -We do not have a good treatment option for Dengue fever
    -We have a high incidence of postoperative infections
    -We have insufficient information to identify high risk patients
  3. Once you have defined the topic and the associated problem then you need to state what others have done to solve the problem. This paragraph is a very short summary of publications describing studies focusing on the problem you just described. In rare situations you (and your advisors) may find that to your knowledge none has tried to solve the problem. If that is the case state that this rather than just not addressing previous publications in the topic. This section should include citations that refer to a reference list at the end of the study protocol.
  4. In the previous paragraphs you have described the topic, the remaining problem and what others have made to try and solve the problem. If others did solve the problem completely then there would not be a need for a new project. Thus, you need to describe, in light of what others already have made, what remains to be solved. This is the rationale for your project.

Aim

State in a general form what you want to achieve in your project. What do you aim for? What is the objective? This aim must link up closely to the rationale you concluded in the previous paragraph.

Research questions and / or hypothesis

If your project is evaluated using quantitative methods (statistics) then you must clarify your aim in answerable research questions or hypotheses that you can confirm or reject. Primary research questions are those questions you perceive as most important and where you have made an appropriate sample size calculation. Other research questions are secondary research questions. If you plan a purely qualitative project where no part involves statistics then it is common to refrain from describing specific research questions or hypothesis.

A common mistake when writing research questions is to include an appraisal in the research questions:
“Is it better to use treatment A compared to treatment B?” or “Are we doing the right thing?”. These aims include an appraisal that is more or less subjective. A scientific questions and its corresponding scientific answer is without any subjective appraisal. The appraisal may be added as part of the discussion around your results but should never be a part of your aims, research questions or results. The proper research questions should look like “Does treatment A reduce blood pressure more than treatment B?” or “Does the current diagnostic procedures add prognostic information regarding the outcome X?”. Treatment A might be more efficient in lowering blood pressure but it might also cost ten times more than treatment B. Hence, when answering the question if A is better than B you may use the answer to the research question “Does treatment A reduce blood pressure more than treatment B?” …but you would also take into consideration other factors such as costs, side effects, ability to deliver A compared to B, etc. Hence, A might lower the blood pressure more than B but your final appraisal may be that B is the preferred option.

The message here is avoid appraisals completely in your aim and research questions. However, this does not mean that you can’t address the question of appraisal. You may well discuss this in your discussion section when you write up the final manuscript for publication.

Methods

  1. Sampling of participants
    If your project involves animals or humans describe in detail how participants are going to be selected. From where are they going to be selected. This is where you describe detailed inclusion criteria.
  2. Sampling frame and population
    If your project involves humans then the participants in your population of interest can be seen as a sample taken from an underlying population. The sampling frame are those from your population of interest that for practical reasons are eligible for being included. The results from your project can hopefully be generalised to this underlying population. Describe to what population you expect your results to be valid. (Ignore this section if you do not involve humans or if you use qualitative approach)
  3. Sample size calculation
    In quantitative projects (analysing data with statistics) you want to extrapolate your results to the underlying population. To have a fair chance of doing that your sample must not be too small. On the other hand it should not be unnecessary large bothering more individuals than necessary. This can usually be estimated prior to the study by making some relevant assumptions and using information from previous studies. Describe in detail what assumptions you made and the exact sample size you calculated for all your primary research questions. (Ignore sample size calculation if you use qualitative approach).
  4. Exclusion criteria
    There might be criteria for later exclusion of participants in longitudinal studies where patients are followed for a short or long time. Thus patients that are included can later be excluded. Exclusion criteria are not just the mirror image of inclusion criteria. They are something else applied after patients are included. If an inclusion criteria is being male then being female is not an exclusion criteria because females were never included.  Exclusion criteria are usually not needed in pure cross-sectional studies where patients are not followed over time. Thus, this is the section where you specify specific criteria for later exclusion.
  5. Group allocation
    In randomized controlled trials describe in detail how participants were allocated to different groups. How was randomization done. Ignore this if participants are not allocated to separate groups.
  6. Interventions
    In studies comparing treatments / interventions describe the different treatments given in enough detail so a reader can replicate this. Ignore this if the study does not compare interventions.
  7. Data collection
    Describe what kind of data will be collected for analysis. This section should be detailed. Established well known questionnaires do not need to be described in detail. A reference is enough.
  8. Data analysis
    Describe in detail how analysis of collected data will be done. In quantitative approach label this section “Statistical analysis” and give details on statistical methods and variables. It is not enough to simply state that you will use Excel or SPSS. You must also give names on the statistical techniques you have chosen. If you use a qualitative approach describe the steps in your analysis and if possible label your approach (such as “Grounded theory” or Phenomenology according to Giorgi”).
Where do you get your funding?

Where do you get your funding?

Budget

Describe the anticipated costs and who pays for it. If no money is needed state this.

Time line

Describe when the different steps in the project will start and finish. It is quite common (and practical) to use a Gantt chart.

Contingency plan

There are always a chance that something goes wrong. One of the most common problems encountered is that recruitment of participants / observations is slower than anticipated. What potential problems have you identified and what contingency plans do you have in place?

Reference list

Parts of the introduction / background (and sometimes also methods) should cite previous publications. They should be listed at the end as a reference list.

Date – Version – Page number and logotypes

Sometimes you write the study protocol directly in a website where you apply for a grant or an ethics approval. Sometimes you write your study protocol as a Word file that is to be attached to an application. In the latter situation it is good practice to let each new version get its own version number that should also be dated. Version number, date and page number should be placed in the footer, visible on all pages. Logotypes should be placed in the header.

Common pitfalls when writing a study protocol

Confused?

Confused?

  • Missing to describe the topic and associated problem. It is common that the researcher are so immersed in it that this seems so obvious that there is no need to write about that. However, members of the Human Research Ethics Committees (HREC) may not be equally immersed in the topic.
  • Forget to mention anything about previous publications in the topic.
  • Forget to clearly explain why a new study is needed given all previous publications.
  • Avoid too much emphasis on the local situation. Even local findings may have a learning lessons also for people outside the local context. Try and write your proposal so the possible learning experience also for “outsiders” can be understood.
  • Writing the aim as a possible consequence of the study is not a good idea. It is much better to state the aim as “Is there a difference in symptoms compatible with nicotine abstinence between men and women trying to quit smoking” compared to “I want to improve the chances of women to be successful when they stop smoking”. The latter may be a consequence of the study once you have clarified if women have more symptoms indicative of nicotine abstinence.
  • Not describing details of how participants are selected is likely to result in questions or rejections.
  • Failing to describe details of data analysis such as names of statistical or qualitative procedures gives the impression that your study is not well planned.
  • Failing to describe a sample size calculation (if you use a quantitative approach) also gives the impression that your study is not well planned.
  • Most Human Research Ethics Committees (HREC) require you to have appropriate logotypes on top of your study protocol and a version number, date and page numbering at the bottom of each page. This is easily forgotten.

Examples of study protocols

(Under construction)
You should cite this article if you use its information in other circumstances. An example of citing this article is:
Ronny Gunnarsson. Writing a study protocol [in Science Network TV]. Available at: http://science-network.tv/writing-a-project-plan/. Accessed May 27, 2017.

Comments are closed.