For Professionals (Shop Floor) > Research >

How To Run



How to Undertake a Research Project

A successful research project requires a structured and disciplined approach. There are a series of stages to go through. These have to be done in the right order - it may not be immediately obvious to you why this has to be done, and it may seem like a waste of time to give so much attention to the details of working out the question, the exact methods and the way that the data will be analysed before actually collecting data. However, if you follow this advice you are much more likely to actually complete a successful project. This advice comes from people who have made the mistakes so it is well worth following.

You should have planned all stages of your project, including the analysis and presentation of results before starting data collection (as data collection comes fairly late on in the structured approach to research). Planning may include the collection of some pilot data.
It is often a good idea to write up the project before you start data collection! This may sound odd – but it is the best way of making sure that you have undertaken a structured approach (writing a research grant application has the same effect). First write the Introduction (for which you will need to do a literature search and formulate the research question), then write the Methods (being specific about all aspects of the research), then write the Results, leaving gaps to be filled in when you have collected the data (this will make you think about data analysis and presentation). All of this needs to be done before any data is collected!

Planning your Project


When starting to think about your project there are 10 things that you should consider from the beginning:

1 Define your research question

Research should be driven by the need to answer a question. The question should be well-defined, important and relevant. Having a well-defined question will help guide you through the many methodological problems that lie ahead. Issues such as who to study, what intervention to study, and what outcomes to measure should be determined by the research question. See our document - Defining and refining your research question.

2 Make sure your idea is original

If your idea is a good one someone else may already have thought of it. Do a reasonably thorough literature review using Medline and the Cochrane database. Also, check the National Research Register for work in progress. If someone has already investigated your idea, don’t necessarily give up. A lot of research is flawed or irrelevant to local or emergency practice. Maybe you can do it better or in a more relevant setting?

3 Draw up a research proposal

Use your well-defined research question to guide you and our document Writing a research proposal to help you. Writing a proposal is a good way of working out what difficulties you will face before they arrive.

4 Decide what methodology you should use

Your choice of method should be determined by your research question, not vice-versa. A basic introduction to the various types of research methods is available on this website. Work out what type of question you are asking and then look at the relevant document.

5 Find out what skills you will need

Research is a professional business, you need to know what you are doing if you hope to be successful. Skills can be acquired from many sources, but there is no substitute for talking to someone who has been there before. This website has some basic resources aimed at helping you to communicate effectively with experienced researchers. Use them to develop your idea and write your proposal.

6 Work out what resources you will need

All research consumes resources. Mostly it consumes your time and energy. Do not assume that either are boundless and use them wisely. Plan your project carefully then double all your estimates of the time each stage will take. Estimate your sample size and plan recruitment. Acquiring major research funding is a difficult business, but many smaller grants go unused. Ask your research office or look at our sources for funding.

7 Work out what help you will need

Doing research on your own is hard work. You can easily make mistakes without realising. Formal research supervision is the best option but very difficult to come by. Hospital and health authority research offices will be able to point you towards taught courses and the resources to fund them. University departments of epidemiology, statistics and health economics can often give up to an hour of advice without charge. Contact experienced researchers in emergency medicine or related specialities. Whatever you do, make sure you have a well-defined idea, have searched the literature and have at least attempted to write a protocol before you seek help.

8 Get ethics committee approval

Always consider the need for ethics committee approval. Even retrospective case note review carries issues of consent and confidentiality. Projects that do not directly involve patient contact (e.g. case note review) will usually be approved by Chairman’s action, without full committee review. Any project involving patient contact, even to simply complete a questionnaire, will require ethics committee approval. See our document on How to get ethics committee approval.

9 Establish proper research governance

Things can go wrong in research- adverse events can occur, data may be lost, patients or staff may complain or results may be flawed. Just as with clinical practice, make sure you have the skills to manage your project, the supervision you require and everyone understands who will carry the can if it all goes pear-shaped. Ensure that you know what you are responsible for and who you are accountable to. See the Department of Health guidance on research governance.

10 Pilot or die

No matter how carefully you have planned your research you will almost certainly have overlooked something. The only way to find out is to get out there and start doing your research – and then stop before you waste too much time. Analyse the data you have collected so far. You will almost certainly find out something very important that you wish to change in your protocol. Nobody ever regrets pilot work.