Why is Study Design Important in Clinical Trials Research?
The importance of a well-designed study including a statistically valid analysis plan cannot be understated in clinical trials. A good study design, appropriate to the population, therapy and question will:
- allow the value of new therapies to be isolated from random variations or fluctuations
- identify which participant attributes or characteristics (e.g. Age/Gender/Body Mass Index) will benefit from a new therapy
- provide valuable data on the mechanism of action of a new therapies
- allow the study to identify the magnitude and appropriateness of treatment benefits for human participants
The two principal types of medical investigations that are undertaken in health research are experimental and observational studies.
Observational studies involve various designs where researchers may observe risk factors (and often exposures) and outcomes without any intervention.
Conversely, in experimental studies, researchers have a direct and active role in allocation of therapies, treatment or interventions under investigation to the participants and cohorts involved.
In a case-control study, a group of participants with a specific disease or outcome, ‘cases’; and a group of participants without the disease or outcome, ‘controls’ are selected. Comparisons are then made between the two groups with regard to a risk factor or exposure. Retrospective data collection is obtained from participants regarding the risk factor or exposure status of interest in the investigation. The hypothesis of interest in case-control studies is associations between exposure and being a case.
Cohort Studies involve following and assessing, over a predetermined period of time, a sample population or cohort with or without a risk factor of interest. The basis of this design is then to assess outcomes in both cohorts, thereby providing a measure of outcome incidence in different groups (e.g. exposed, non-exposed). Cohort studies are generally prospective, whereby individual characteristics of interest to the study are measured progressively at baseline and at periodic follow up visits.
The aim of cross sectional studies is to determine associations between outcomes and risk factors. This type of study provides a measurement at one point in time and is valuable in estimating prevalence of diseases. As there is no time factor, cause and effect may not be established but similarly to case reports/series, cross sectional studies can be useful for hypothesis generation.
Case Series/Case Report
Often an interesting, rare or unusual case, reported as a detailed patient history or chronicle of pertinent events in a single or series of patients. A case report or series is often worthwhile for hypothesis generation but do not allow for statistical analysis.
Randomised Controlled Trials
Randomised controlled trials (RCTs) involve participants with distinct conditions, diseases or attributes of interest being randomly assigned to different groups – such as two different types of therapies or medications – and then followed over time to assess certain outcomes. Comparisons are made, using statistical methods, between these outcomes in each group. RCTs allow researchers to make inferences about the efficacy of the treatments under investigation in the populations from which the participants were drawn. RCTs provide the highest level of evidence of any single study.
- Greenland KJ & Rothman S: Modern Epidemiology, Lippincott-Raven 1998, Chap 6 and 7 p96-105.
- Woodward M: Epidemiology Study Design and Data Analysis 2nd Ed, Chapman & Hall/CRC 2005; Chap 5 p215-227; Chap 6 p273-277, p285-296, p298-301 and Chap 7 p353-355.
- Gibbs E & Kirby A: Study types, levels of evidence. NHMRC CTC, 2015 University of Sydney.