It seems that every time I read a research paper, it is described in an entirely different way, with each kind of research having value in its own right. Some are useful for some things, others are useful for other things!
So, to make my life a little simpler, and hopefully to help you make sense of the terminology, I’ve summarised a few of the main approaches to research that I keep coming across. Understanding the approach taken during research saves time and effort when reading, as one can quickly assess the likely strengths and weaknesses of a particular paper before one reads it.
So, let’s start at the top of the research chain:
A Systematic Review
A systematic review looks at all the research available to answer one question, focusing on one treatment modality or technique to treat one condition in one species. A systemic review gives an overview of the strengths and weaknesses of each research paper available, and comes up with a conclusion; does this treatment modality work for this specific condition, or not? It will also point out if there is not enough evidence to draw a solid conclusion, and will highlight the weaknesses in the overall trend in the research. Perhaps there are methodological weaknesses in the design of some papers, or the research is not quite independent, or perhaps there is a lack of sufficient randomized controlled clinical trials.
In developing a research question for a systematic review, there are four key elements that need to be considered: the population to be studied, interventions being evaluated, comparison or control groups, and outcomes measured.
For a systematic review to be performed, there needs to be a sufficient body of evidence that has been evaluated and pertains to the research question asked. Available literature also needs to be of a high enough standard to be included in the review, meeting specific requirements that will allow the research to be comparable to others.
A Scoping Review
A scoping review identifies an area where a sufficient body of evidence exists to support a systematic review within a larger category of literature. For example, a scoping review will look at all the literature pertaining to one treatment modality, possibly in its various forms, and categorize the literature according to criteria such as population, outcome measures, conditions treated, or any other criteria that may be identified. By reading a scoping review one can identify the categories in which there is a sufficient body of evidence to evaluate the literature further. One could also look at such a review from a different perspective and identify where research is still lacking, and then move to fill those gaps.
A scoping review does not evaluate the literature to extract results and synthesize the evidence, or do a risk of bias assessment. It simply categorizes the available research on a specific subject matter or modality.
Randomized Controlled Trials
Randomized controlled trials, or RCTs, are studies that allow a treatment intervention to be compared to a control group. Very importantly, the population studied is randomly assigned to either the control or treatment group (there may be more than one treatment group) which reduces the risk of selection bias from the researchers. This allows the comparison of results from one or more treatment interventions with results from no treatment intervention or from a placebo.
Information bias can be controlled in these research papers by implementing a blinding or double blinding, where neither researchers nor practitioners know who forms part of the treatment group and who, the placebo group. In addition, the subjects themselves may be “blinded” with regard to the group they are placed in.
In a non-randomized trial the participants are not assigned to a treatment or non-treatment group by chance. Either the patient chooses their intervention, or the researcher assigns them to a group. This can increase the chance of bias and can affect the outcome of the research. The reason for the non-randomization needs to be evaluated and considered when assessing the results of this sort of study.
An experimental study will evaluate two groups of interventions: a control and an intervention group. Participants may be allocated to their groups in one of three ways:
The first is known as independent measures, and will assign different participants to each group, most likely by random allocation. In this way there is a clear test group and control group.
The second is known as repeated measures, where both groups participate in each section of the test. In this way the participants act as their own control, and fewer participants are needed altogether. There is a risk that the test will be performed better in the second trial as the participant will have had an opportunity to practice; alternatively the participant may perform worse, due to fatigue. Counterbalancing may be used to reduce this risk, where the participants are divided into two groups, with one performing the experiment first and then the control, and the other the control first, then the experiment.
The third design is by matched pair. This allows similar but different participants to be included in the study. Participants will be paired in terms of any factors that could affect the outcome measured, such as breed, age, size or activity level. Each participant in the pair is then randomly assigned to a treatment or control group.
Case Reports or Case Series
Case reports or series are highly valuable in describing novel, unusual or atypical conditions or patients. They describe in depth the findings and interventions in a patient or a group of patients in a real-world clinical setting, and can be instrumental in generating research questions.
Sharing cases is one of the best ways of sharing knowledge and experience, enabling us to come up with treatment interventions for rare conditions.
A pilot study is often a first-stage study, asking the question of whether a specific research project can be performed, and if so, how. It is performed on a small scale, with a smaller population than the anticipated main study, and evaluates the safety and quality of the study, revealing areas of weakness in the study design. These areas are then tweaked, thus ensuring the best possible design for the main study.
Quantitative vs Qualitative
Qualitative research provides data that is not numerical. It allows a more versatile data collection process and can provide depth to understanding the outcomes achieved. Quantitative research provides numerical data and statistical values, measuring specific outcomes in numerical terms.
In Vitro vs In Vivo
I don’t know about you, but I always have to go back and check which one is which! In vivo refers to anything performed inside a living organism, and in vitro is performed outside a living organism.
An easy way to remember is that ViTro has a T for Test tube. ?
Prospective vs Retrospective
In a prospective study, the population is evaluated going forward. They are followed from the time of a treatment or intervention, and outcomes are measured over time.
In a retrospective study, information from the past is evaluated. This may be performed by looking at patient records for a specific condition or intervention performed in the past.
One thing is for certain – we need to be reading the research. Ours is an ever-evolving field, and if we’re not keeping abreast, we’re slipping behind!