"Sadly many good ideas fail simply due to bad study design."
By Ahmed Adlan
16 September 2016
When conducting research one must choose the most appropriate study design to adequately answer the research question.
Why is study design important?
Choosing the right study design is key to success. No amount of clever statistics can fix a poorly designed study. An inappropriate study design can significantly hamper your chances of publication, particularly in high impact journals. In fact the wrong study design may prevent your work from getting published altogether, which would be a complete waste of your time and valuable resources. Sadly many good ideas fail simply due to bad study design.
Who is responsible for ensuring correct study design?
Before research can be conducted, ethical approval must be obtained from the relevant research ethics body. One important duty of the ethical review is to ensure that the most appropriate study design has been proposed. This ensures that resources are used sensibly and minimises participant harm. It is important that the proposed researcher has a good understanding of different study designs in order to choose the appropriate one. Ultimately it is the researcher's responsibility to choose the correct study design.
What types of studies are there?
Figure 1. Types of research studies
Which is the best study design?
The best study design largely depends on the research question, the population of interest and resources available. In general meta-analysis of randomised controlled trials and randomised controlled trials are the highest level of evidence available. Non-randomised case-control trials, cohort studies and systematic literature reviews have an intermediate level of evidence, whilst case series and case reports constitute the lowest level of evidence.
Randomised controlled studies are best to assess the effects of an intervention and should be double-blinded. Placebo-controlled studies are best to confirm efficacy of a new treatment, whilst active-controlled studies are best to compare a new treatment to current best-practice.
Cohort studies are best to determine the prevalence of a condition and to characterise a specific disease group. Case-controls studies are best to compare a target group to normal individuals.
Before beginning research it is worthwhile performing a systematic literature review. There are many potential benefits including: to review and evaluate all available evidence to date; to ensure that your research question has not already been answered; to highlight gaps in current knowledge; and to guide potential future research. A meta-analysis should always be attempted if feasible.
Balancing best study design v feasibility
It is not always feasible to employ the best study design, and often there is a trade off between an ideal design and what is practical in the real world. One useful approach is to perform a thought experiment. Think of the study design that would answer your research question the best. If not feasible or ethical to perform that study then think of the next best way. Again if still not feasible then think of the next best design. Once you have decided on a design it is best to discuss this with a trusted senior researcher (or two). Often small concessions may need to be made so it is important to remain flexible and open to criticism.
In summary
It is always worth spending a bit of time thinking about the most appropriate study design to answer your research question.
Discuss your ideas with a trusted senior researcher before embarking on a new study.
To improve your chances of succeeding in research and publishing your work in high impact journals try to choose the best study design possible to answer your research question.
Begin with a literature review (ideally systematic) and if possible perform a meta-analysis.
By Ahmed Adlan
16 September 2016
When conducting research one must choose the most appropriate study design to adequately answer the research question.
Why is study design important?
Choosing the right study design is key to success. No amount of clever statistics can fix a poorly designed study. An inappropriate study design can significantly hamper your chances of publication, particularly in high impact journals. In fact the wrong study design may prevent your work from getting published altogether, which would be a complete waste of your time and valuable resources. Sadly many good ideas fail simply due to bad study design.
Who is responsible for ensuring correct study design?
Before research can be conducted, ethical approval must be obtained from the relevant research ethics body. One important duty of the ethical review is to ensure that the most appropriate study design has been proposed. This ensures that resources are used sensibly and minimises participant harm. It is important that the proposed researcher has a good understanding of different study designs in order to choose the appropriate one. Ultimately it is the researcher's responsibility to choose the correct study design.
What types of studies are there?
Figure 1. Types of research studies
Figure 2. Important terminology
Which is the best study design?
The best study design largely depends on the research question, the population of interest and resources available. In general meta-analysis of randomised controlled trials and randomised controlled trials are the highest level of evidence available. Non-randomised case-control trials, cohort studies and systematic literature reviews have an intermediate level of evidence, whilst case series and case reports constitute the lowest level of evidence.
Randomised controlled studies are best to assess the effects of an intervention and should be double-blinded. Placebo-controlled studies are best to confirm efficacy of a new treatment, whilst active-controlled studies are best to compare a new treatment to current best-practice.
Cohort studies are best to determine the prevalence of a condition and to characterise a specific disease group. Case-controls studies are best to compare a target group to normal individuals.
Before beginning research it is worthwhile performing a systematic literature review. There are many potential benefits including: to review and evaluate all available evidence to date; to ensure that your research question has not already been answered; to highlight gaps in current knowledge; and to guide potential future research. A meta-analysis should always be attempted if feasible.
Balancing best study design v feasibility
It is not always feasible to employ the best study design, and often there is a trade off between an ideal design and what is practical in the real world. One useful approach is to perform a thought experiment. Think of the study design that would answer your research question the best. If not feasible or ethical to perform that study then think of the next best way. Again if still not feasible then think of the next best design. Once you have decided on a design it is best to discuss this with a trusted senior researcher (or two). Often small concessions may need to be made so it is important to remain flexible and open to criticism.
In summary
It is always worth spending a bit of time thinking about the most appropriate study design to answer your research question.
Discuss your ideas with a trusted senior researcher before embarking on a new study.
To improve your chances of succeeding in research and publishing your work in high impact journals try to choose the best study design possible to answer your research question.
Begin with a literature review (ideally systematic) and if possible perform a meta-analysis.
No comments:
Post a Comment