Comparison of Causal Relations Results from Randomized Controlled Trials and Observations

Explain why it is often easier to get knowledge of causal relations (i.e. knowledge of whether one thing causes another) from a randomized controlled trial than from an observational study (such as a cohort study or a case-control study).  Under what circumstances might it nonetheless be better to use an observational study?

Randomised Control trials (RCTs) are widely considered to be one of the most important type of experimental study used in medicine. They are considered to have higher validity as compared to most other studies (Bruno, 2016)[1], as well as this they allow a higher degree of control of the intervention which creates more certainty when establishing causality. However, they do have their limitations and there are certain scenarios in which it is better to use an observational study such as a cohort or case control study.

The design of an RCT lends itself to having high validity, by randomising which patients are given the intervention, allocation bias is reduced and so any effects observed after giving the intervention can be considered to have been caused directly by it with greater confidence. Blinding is also used (though not always) and this further eliminates bias, e.g. in a double blind RCT neither the patient’s or the doctors know whether they are receiving/giving the intervention. This eliminates both performance bias, where knowing which group they are in would affect the subject’s response and assessment bias where the effects of the intervention can be overestimated or underestimated by the doctor (Chan AW, 2013). Often, the control group is given a placebo however there are sometimes ethical issues with this in which this is not possible for example there is a clear ethical issue in inevitably raising the hopes of patient’s who are ill with no current treatment available and as such are considered terminal., giving them hope that they might receive a new novel treatment or no such treatment at all raises great ethical objections. This is alleviated in other trials where the control group can be given the treatment which is the best available at the time so that the improvement of the novel treatment over the current standard can be tested.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!
Find out more about our Essay Writing Service

Case control studies are a type of observational study that attempts to link risk factors with a disease, typically using statistics. Generally, the method involves a group of patients who have the disease (considered cases) being compared to a group of people that are healthy and do not have the disease but are otherwise identical or as close as to the patients (considered controls)., hence the name. Risk factors are then compared and this information is used to associate the chances of someone developing the disease if they have that risk factor (KA, 2006). As such, this type of study is frequently used to study diseases that are considered rare in which it would be difficult to gather a group of patients’ large enough at one time for another type of prospective study (Schulz KF, 2002). Unlike cohort studies, the disease is what is used to select the patients initially and not the risk factors, so it can be said that instead of measuring the risk of disease based on exposure the odds of exposure are measured based on disease. (Bruno, 2016)

Cohort studies are a type of observational study in which a group of individuals that represent a sample considered to accurately a population of interest is followed over a long period of time whilst data is collected on items such as risk factors (lifestyle etc.) and whether individuals develop a disease or not. Unlike case control studies, patients are disease free at the start of the study and the risk of developing the disease is measured. Unlike other types of study, cohort studies can be both retrospective and prospective, retrospective ones study events from the past up until now and look for risk factors that can be associated with the disease, this can be advantageous as you already have a large time period over which data was collected so you do not have to wait as you do in a prospective study, however there are other issues such as the reliability of the quality of data collected. (KA., 2003). Prospective cohort studies follow patients into the future and allow a more complete study to be undertaken as various factors can be incorporated into the study to create a baseline to compare to. However, the clear shortcoming is the long period of time over which the study has to be taken, often lifetimes.

Generally, RCTs are considered the ‘gold standard’ and as such observational studies are used when it is not feasible to undertake and RCT. As mentioned previously, if a disease is rare, a case control study is the most efficient study to undertake and identify potential causes. Also, sometimes it is unethical to undertake an RCT, e.g cannot deliberately expose subjects to a potentially harmful situation just to see if the effects of the intervention avoided it. (Institute for Work & Health, Toronto , 2016) If very little is known about a disease then a cohort study may be the best study to undertake so that it can be seen which risk factors were more likely to cause it do develop. However observational studies are more prone to containing confounding bias as they do not generally eliminate as many confounding factors as RCTs do. This leads to them having a lower validity generally. (Bruno, 2016). RCTs are still considered the gold standard because they leave little to chance and as such produce more reliable evidence. However unlike in the past, there is a greater awareness of the limitations of RCTs, as well as them being extremely expensive both in terms of money and time. Also, the validity of RCTs increases with the sample size and as there is a limit to the sample size that can be feasibly attained for an RCT, diseases that affect a very small population or that take a very long time to appear or be treated may not be picked up by an RCT and in these cases it is almost always better to use an observational study.

Bibliography

• Bruno, D. (2016). Study design made easy. Medical Writing.
• Chan AW, T. J.-J. (2013). SPIRIT 2013 statement: defining standard protocol items for clinical trials. Ann Intern Med.
• Institute for Work & Health, Toronto . (2016). Observational vs. experimental studies . At Work.
• KA, L. (2006). Study design V. Case control studies. Evid Based Dent.
• KA., L. (2003). Study design IV. Cohort studies. Evid Based Dent.
• Schulz KF, G. D. (2002). Case-control studies: research in reverse. Lancet.

[1] Meta-analyses and systematic reviews are still considered more valid however.