Information for Researchers

Ok, so you have made an observation, which has led you to think of some new ideas.  You now need to think of the questions that you want to answer for example:

• What are the issues?
• What is the relationship between?
• Why?
• How?

This will allow you to define the most appropriate choice of methods.

Your next step is to review the literature.  The best way to do this is to use the internet or physical journal/book entries.  Most popular sites on the internet include BIOMED and PubMED.  When doing your searches you need to select certain keywords which will enable you to focus on research that has already been carried out in your specific research area.  The easiest and less time-consuming way to find this out is by reading the abstracts.

Once you have done a thorough literature search, you need to ask if the research question/s you propose have already been answered.  If it hasn’t then you have a clear ground to go ahead with your proposed project.  However  if your question has already been answered there are other possibilities that need to be looked into before discarding your research:

• How were previous studies performed?
• Were the methods valid?
• Could the methods be improved?
• What were the results?
• Was there any bias in the results?
• Can I do a better study?

Now you can point the direction the research is going by stating the main purposes of the research. You need to know what exactly you want to achieve in the study – this is known as the AIM.  Also, how will you achieve your aims?  These will be your OBJECTIVES.  The purpose of the research should indicate the data that will be collected and the populations that will be studied.

The population that you choose to carry out your study could be very large in size, which would mean it would take a very long time for you to complete your study.  This is why we use a sample of the population.  The idea of using a specific sample size is to get the right number in the sample:

• not too many,
• and not too few so that the results can be expanded and show a true representation of the population

Your sample size is determined by using statistics, such as an alpha coefficient.

Choose the type of study.  If you have found various types of study design in your literature review, you are under no obligation to repeat a design.  It could be the case that they did not use the most appropriate method for that type of study, even if they had their results published.

There are various types of study designs in medical research.  The main types of study are Comparative studies such as:

• Controlled study (Case, placebo – to look for factors associated with being a case)
• Case-series study (to look for factors that correlate with patient outcome)
• Cohort study (systematic conducted either prospectively or retrospectively)
• Clinical trial (Randomised, to compare therapies or prevention strategies)
• Questionnaire survey
• Interview survey (semi-structured)
• Action research
• Ethnographic (cultural)
• Phenomenological (happening)

Once you know the type of study you are going to conduct you need to plan out the study.  The methods section needs to be a clear and precise description of what you intend to do so that other people should be able to read this and understand what data you will collect and how, and also be able to replicate your research without asking you.

You need to be able to define the patient population you wish to carry out your study within.  This is done by listing the inclusion and exclusion criteria.  You must avoid making the criteria too tight/narrow as it will later be very difficult to know how far to stretch the results from your sample into the wider population.  It might be worth getting an idea of how many patients you would be able to recruit into the study, taking into consideration some may refuse to take part, may be missing from data analysis or other reasons.  If you find you might be left with too small a sample number, then you may have to rethink your inclusion/ exclusion criteria.

The next question to ask yourself is what kind of data will I be retrieving?

There are two main types of data that can be collected: Quantitative and Qualitative data. 

Quantitative data is in the form of measurements and can be further sub-divided into:

• discrete data where the values go up in steps e.g. shoe size, or,
• continuous data where the values go up in intervals e.g. height, age, blood pressure.

Qualitative data is in the form of categories and can be further sub-divided into:

• Nominal data where the data is given a numerical code i.e. give a number to a category e.g. Male = 1 Female = 2, or,
• Ordinal data where the data is analysed in ordered values i.e. give increasing/ decreasing number to a category e.g. Poor = 0, Good = 1, Excellent = 2.

It is always a good idea to give supporting evidence that the key data you have chosen will be reliable and valid.  Reliability refers to consistency and repeatability, whereas validity refers to honesty and accuracy.

How will you collect the data?

Most data is collected in either of two categories:

• Questionnaires
• Data collection forms

Questionnaires are used most when collecting Qualitative, ordinal data.  Data collection forms are used most when collecting Quantitative data. 

Again, it is a good idea to give supporting evidence that the way in which you will be collecting the key data will be reliable and valid.  For example, many well-known questionnaires have been tested extensively and are accepted as a valid and reliable method for collecting data, such as the SF-36 for Quality of Life assessment.

Give thought as to how the data will end up on the computer.  You need to think about the optimal layout of record forms and how to input data into a spread sheet for analysis.

The organisational structure of the research must be defined. You will need to make sure that the right person will be at the right place at the right time to make the right measurement on the right patient and so forth.  Too many studies fall down because of bad administration and this shows up in suitable patients not being entered, low response rates to questionnaires, and missed values for tests and assessments.  A successful study requires a dedicated enthusiast to make sure everything runs smoothly.  A well-planned study does at least provide the informed researcher with the knowledge that a well-organised study will produce results of clinical value.

When obtaining your data, consideration needs to be given in regards to confidentiality and ethics (Information and Consent).  Who will collect the data?  Who will be able to see the data? Does the patient need to consent for the research to be carried out? Will the research harm or distress the patient in any way?

Once you can confirm all the above appropriately, you will then be able to put together the Protocol. 

Refer to ‘Writing a Protocol’ for more information

In order to carry out the research, you may need to use the services of another department/s (Support Service).  The departments which are likely to offer a service for research activity are (but not limited to): 

• Cardiology
• Clinical Chemistry
• Haematology
• Microbiology
• Pharmacy
• X-Ray Departments
• Physiotherapy
• Medical Physics
• Health Records
• Hotel Services
• Medical Illustration

So, once you have decided which support services you are going to need for your research, you need to contact the Heads of department and discuss your idea.  You should provide them with the following information:

• Outline of project,
• Number of patients likely to be recruited,
• Type of additional tests/samples/visits,
• Number of additional tests/samples/visits per patient/subject,
• Frequency of tests/samples/visits,
• Expected commencement and closure date.

This information will allow the Head of Department to establish whether they could accommodate the additional work.  The R&D Directorate will ensure all Heads of Department are notified and request their agreement of support as part of the R&D approval process.

Please see the below flowchart to get an understanding of the research process at a glance.