Although there are specific principles behind every research but every project or study is unique in its situation and requires special attention for its design and execution.
Quality of any research study depends on how well both design and execution phases of the project are accomplished. There is a positive relation ship between the research protocol and overall quality and acceptance of the resulting research study .
A research protocol is roadmap that guide and investigator from conceiving an idea to publishing results. To be brief every study should be planned. An unplanned study according one of my teachers is a fishing expedition.
Planning is done before the conduct. How ever before you plan to do some thing you must know what you want to do i.e., `idea`, a thought that can be your own or generated in some one else mind.
This idea is transformed in to a research question that you want to answer.
Now before just jumping into planning look for the things other have done, i.e., literature review. It is very important and essential component that tells you what other has done and how they have done. If question conceived by you is already answered, there is no need of doing this because it will not add any thing new to the existing literature.
Now the idea you have conceived should be phrased and formulated in the answerable scientific research question. It should not be too big or ambiguous.
Methodology of any study is based on the way research question is posed and designed is also guided by this.
When you have a good protocol, operational plan and budget to conduct study you can do a good study.
It is excellent that protocol is reviewed by those who are expert in research methodologies (Epidemiologist) and some person in the relevant field.
The components of a Research Protocol
I will elaborate some of these things and for more detail can be found in Textbooks and literature in journals.
Some Components are as follows:
It should be brief and explanatory. Better to include type of study and setting.
Like `A case-control study of risk factors of acute hepatitis B in Karachi`.
Introduction & Background
Relevant details about the problem. The pathophysiology of the disease, global burden, similar settings and than local (if available)
A description about what other in this field in other countries and in your country has done. Literature review should be in depth to identify potential cofounders and effect modifiers and to determine the areas in which knowledge is needed.
It is very important. It should tell that why you want to do this study. Is there any gap in the existing knowledge that you want to fill? The study has been conducted in other setting and the finding can be different in your setting (but this is not a good justification).
Study objective and /or research questions
It should identify the immediate purpose of study and be very clear. Objective always start with `to` (word of accomplishment)
Like `To identify the risk factors of acute hepatitis B in Karachi`
It should include the detail description of methods with reasoning.
Important designs are:
Details of the designs can be found in any epidemiologic text
Just naming of the design is not sufficient. It should include the details what you want to do, how you are going to recruit subject, how going to follow etc.
Study population and its recruitment
Description of study population. Inclusion and exclusion criteria (Exclusion criteria apply after those who are included in study).
Description of the setting whether it is a community or a hospital. In case of hospital what are the important characteristic of this/these hospital which type of patient visit this what is its catchment area, how big it is and what diagnostic and treatment facilities it have. Is it a special centre for some specific disease.
If you are mentioning the field work duration specify it
Sample size determination
How and according to what assumptions sample size is calculated. It is very important to detect what you intend to do with adequate power.
How sampling will be done; whether is random (type of random sampling) or convenient sampling. How cases and controls will be identified.
What variables you want to measure.
Methods of data collection
What method you are going to apply as records, questionnaire and in person interview, mailed questionnaire etc What test you are applying and what is reason for applying this test. Specificity & sensitivity of the test, if some new test is to be applied.
Plan of Analysis
How you are going to analyze your data. For what you are looking for and which statistical method will apply.
Whether you run descriptive analysis or go for hypothesis testing and may be multivariate analysis.
It is better to get help from biostatistician for sample size and data analysis.
Ethical considerations & humane subject protection
This is very important consideration. Now every study should be subjected to ethical review by Institutional review Board.
Ethical consideration include:
Obligations to study subjects
Protection of their interest. Complete information regarding risk benefit for participating in any research study is complete right of every subject; no concealment. Maintaining confidentiality of information obtained from subjects.
Obligation to society
Avoiding conflict of interest, partiality; communicating finding; data sharing etc.
Obligation to funders or employers
Specifying obligations in contract how research be conducted and it will involve ethical, technical, administrative or legal responsibilities.
Obligations to colleagues
Promoting and preserving public confidence in epidemiologic research while not over and under estimating the epidemiologic methods. Disseminating the study findings.
How quality control checks will be introduced and maintained.
Some of the key points are.
Expected benefits from the study
Who will and how benefit from the study
Operational plan & budgeting
Detail time line and budget for the study
Plan of dissemination of results
How the findings will be shared with others.
By some presentations, publishing results in journal etc.
References should be quoted in Vancouver style.