How to Write a Research Protocol1

On which areas should one do research? This depends on multiple factors. The reasons for doing research may vary from educating parts of the brain seldom exercised (probability of success = 0.9) to winning the Nobel Prize (probability of success ≤ 0.000001). It depends upon the expertise and interests of the researcher. Clinical research is usually driven by the knowledge gaps or challenges faced by a clinician in day to day clinical practice. The range is vast, varying from help seeking behavior, diagnostic evaluation and efficacy of different interventions, to relapse prevention. When there is a barrier in accessing the health care facilities in diseases like mental illness, substance use disorders or other stigmatizing illnesses like leprosy, most of the patients can only be identified in epidemiological studies. Sometimes, the research 2is driven by the National Programs and policies of the government. As now-a-days research is gradually becoming an expensive affair, the availability of funds and the priority areas of the funding agencies are likely to influence the area of research. Recent trends in the research, supported by literature review, can also generate some ideas for a research study. The investigators should also decide whether to do multidisciplinary research or in one particular specialty, basic or applied research, quantitative or qualitative research, or research in health economics.

A good research topic should be Feasible, Interesting, Novel, Ethical and Relevant. These criteria have been called the FINER formula. After identifying an area for research, a detailed protocol is prepared. The protocol should explain what the research plans to achieve, how the topic of the study is important, and how the knowledge gained from this research can address the current knowledge gap in national and international scenarios. The research design should be formulated by taking into account the feasibility and practical aspects versus robustness or soundness of the methodology of the study which can answer the research question.

The protocol is a comprehensive action plan of the proposed research study, and needs to be prepared carefully and not in haste. The protocol is an essential part of a good quality research. It is required not only to get approval from ethics committee but is also a guiding tool for the whole research study. Writing a research protocol of high quality and relevance is also important to apply for funding a research. The research proposal is supported by findings from review of literature or from a pilot study. The protocol explains in detail the procedures to be carried out in the study, including the problems likely to occur during the conduct of the study and the methods to overcome them. The investigators need to discuss among themselves or with other colleagues in the respective field to improve or strengthen the proposal.

In studies where multiple investigators are involved, the role of individual investigators in the research study should be clearly delineated and all investigators should follow a uniform approach to maintain the fidelity. In a multicenter study 3or in a study which multiple investigators are involved or the funded projects, the publication protocol, order of authorship, dissemination of results, data sharing and data ownership should be clarified in the protocol. A good research study should strictly adhere to the protocol. Occasionally, the protocol may need modifications after the study has been started, due to change in circumstances or any other reason. If such a situation happens, approval would need to be taken from the ethics committee regarding the changes made, and details about the change and reason for making a change in the protocol would also need to be mentioned in the methodology as well as the results and the final report.

The different components are discussed in detail in this section.

After the title, names of the investigators and personnel involved in the study should be mentioned. This includes who will be the principal investigator, coprincipal investigator and coinvestigators, their affiliations and the institution (s)/sites where the study will be carried out.

For the purpose of this chapter, we will be using the example of a study that measures the plasma levels of omega-3 fatty acids in patients with depressive disorders and healthy controls. In this study, the title may be written as ‘A controlled study of plasma omega-3 fatty acids in patients with depressive disorders’.

Some protocols, particularly the larger ones, include a synopsis or summary which summarizes all the elements of the protocol.

While writing the review, attention should be drawn to the good points and deficiencies of the available studies. In terms of writing style, it is a good practice to make the writing flow rather than simply going through a list of papers. The current knowledge gap in the respective field should be highlighted here and how this piece of research can fulfill this knowledge gap needs to be mentioned. One may structure the review of literature under subheadings.

As per the example title given earlier, details about omega- 3 fatty acids, and their relevance to psychiatric illnesses, particularly depression, needs to be reviewed.

Based on the literature review and current understanding, research hypotheses or rationale for the study should be formulated. Example of rationale (as per the above study title): ‘There is no Indian data on relationship between omega-3 fatty acids and depression, or correlates of levels of omega-3 fatty acid level with severity of depression’.

The hypothesis is a tentative prediction of presumed relationship between two or more variables. It reflects what result is expected from the research by the investigators. It guides the scientific inquiries and often follows directly with the question being asked or the theory being tested. It provides direction to the research design and assists in the collection, analysis and interpretation of data. The hypothesis should be stated in simple and understandable words. In general, hypotheses are stated in the null form, as a statement which states that no relationships or differences exist. Then the hypothesis of no significant difference is challenged. The result of statistical testing gives the probability that the hypothesis of no difference is true.6

An example of null hypothesis is, there is no difference in levels of plasma omega-3 fatty acids in patients with depressive disorders and healthy controls.

An alternate example of hypothesis could be, there is a difference in levels of plasma omega-3 fatty acids in patients with depressive disorders and healthy controls.

Or

While writing the methods, fluency must be ensured. A common mistake that makes the write-up appear disjointed is use of both passive as well as active voice within the same section. One must stick to either the point of view of the experiment (passive voice) or the point of view of the experimenter (active voice) throughout a paragraph to ensure coherence. This section should have all the information, so that the reader who needs to replicate the study/experiment can do it without access to the detailed study protocol. The methodology can be presented in structured or unstructured format. Structured representation makes it clearer and objective for easier comparability between different studies. Though unstructured format lacks objectivity, it may be a better option for descriptive and qualitative studies.

In methodology, one needs to state the study design, study population, setting of the study, selection criteria, study procedure, data collection, assessments, and interventions, if any.

It is a general practice to refer the participants as patients. But, when we include a control group, they may not be patients. So, preferably, the term subjects can be used. The specific selection criteria of the participants should be defined here. The participants can be patients, healthy controls or general population (in case of epidemiological study), caregivers, or health care providers. The study settings/universe of the study too should be clearly mentioned. They can be subjects with a particular diagnosis, all the subjects attending the outpatient services, or inpatients. If the subjects belong to a particular diagnosis, it should be clearly operationalized as per standard practice guidelines or any standard reference or sometimes by proxy measures. For example, a stable patient may be defined as one who is on the same dose of medication for a particular period of time. The inclusion and exclusion criteria should be clearly mentioned, taking the possible biases into the account.9

In longitudinal studies, the strategies to address the confounding factors in selecting or allocating the participants and the attrition of subjects should also be described.

Instruments or assessment methods used to study the objectives should be described. The cross-cultural variation, and validity and reliability of an instrument in a different language should be addressed. Details of the measurement procedures or instruments need to be given in detail.

In general, the variables in a study could be classified into four:

The main predictors and all outcome variables should be described with sufficient detail, so as to allow replication and assessment of the quality of the measurement or classification.

In some of the research projects, particularly the funded projects, certain specific issues also need to be specified. The project milestone can act as a guide or reminder to make the task time-bound. The methods of dissemination of research findings allow the readers to know how the results would be accessible to the research community. It can be through presentation in conferences, academic forums or through the publication in academic journals. In funded projects, the funding agency will release the funds for the project as per the milestone. Funds can also be allocated for dissemination of the research findings,. A list of all the resources likely to be required for successful completion of the project needs to be included in the protocol. The resources required can be personnel, instruments, stationary and contingencies. As the resources have cost implications, adequate justification must be given to the funding agency for the same.

Exact methods planned and the tests to be used to analyze the data must be mentioned beforehand. It should not be the case that data analysis would be planned after collecting the data, as it can introduce bias. Details on how the primary and secondary outcomes will be analyzed should be included. The computer software that will be used for data analysis should also be mentioned.

The choice of statistical analysis is dependent on the purpose of investigation, mathematical characteristics of the variables involved, the method of data collection and the statistical assumptions made about these variables. Descriptive statistics can be used to summarize the data to describe the sample. Basic examples of numerical descriptors include the mean and standard deviation. Inferential statistics is used to model patterns in the data, accounting for randomness and drawing inferences about the larger population. These inferences may take the form of answers to yes/no questions (hypothesis testing), estimates of numerical characteristics (estimation), descriptions of association (correlation), or modeling of relationships (regression).

Internationally recognized ethical guidelines a include Declaration of Helsinki—a set of ethical principles regarding experimentation involving human subjects developed for the medical community by the World Medical Association; and the Indian Council of Medical research (ICMR) guidelines on research on human subjects (available from http://icmr.nic.in/ethical_guidelines.pdf).