Please visit our website www.ccmacademics.com for more detailed analysis:
Hello and welcome everyone for series on statistics today’s presentation is on observational and interventional studies what are the types of studies so in this presentation we will define the types of the studies and we will tell you the difference between association or correlation and causation so studies are basically divided into two they can be observational or
Interventional in observational we collect data in a way it doesn’t interfere how the data arises which means that the researcher doesn’t interfere with this study subjects the results that we get from an observational study helps us in establishing correlation there can be two types of observational studies the first is retrospective it means that we retrospectively
Collect the data from what has already happened or prospective which means as this study goes along we keep on collecting the data this second type of study is interventional which means that the researcher directly interferes with the study subjects and randomly assigns them to each groups this type of study that is an randomized control trial helps in establishing
Causation let’s take an example of an observational study here are six patient admitted to the icu we decide to give them polymyxin and look for acute kidney injury so three patients receive polymyxin and out of these three patients two patients develop acute kidney injury now what did we get in this particular study what was our conclusion does it mean that polymyxin
Is associated with a ki or we can say polymyxin causes a ki i don’t think we can say here that polymyxin causes a ki as we are not sure whether all the patients are which means it was the clinician who decided to start the polymyxin based on his clinical experience it means that most probably polymyxin was given to patients who were already very sick and it means
They already had a higher risk of developing acute kidney injury so in this particular study what we can say is polymyxin is associated with acute kidney injury which means that most of the patients who received polymyxin had a correlation with the akh we cannot causally associate polymyxin with the development of a ki now let’s take an example of an interventional
Study against six patients admitted to the icu we give them polymyxin and look for a ki but here we randomly assign the patients and with each patient having equal chance of getting selected to the interventional group so what was the conclusion in this study in this type of randomized trial where we select patients and randomly divided into two groups we can
Establish causation as other factors which can affect the study results i thought to be taken care of which means that the risks of bias have been almost eliminated only then we can say that polymyxin is a causative agent for developing acute contingent so previously we thought that correlation is implied by causation actually that’s not the case the correlation
Is a separate thing and to establish causation we need to do randomized controlled trials because what we forget is there are confounding factors these confounding factors are mostly present in the observational studies which make it difficult for us to establish causation thank you for your patience and subscribe to our channel for further information and more such videos on statistics
Transcribed from video
Types of Studies: Basic Biostatistics By CCM Academics