association – “A relationship between an exposure (or a characteristic) and a disease that is statistically dependent; that is, the presence of one alters the probability of observing the presence of the other. Association is a necessary condition of a causal relationship, but many exposure-disease associations are not causal. If there is no association, the variables are said to be independent.”1
bias – “A systematic (as opposed to random) error in a research study leading to the deviation of results from the truth in one direction. Bias can result from many sources including sample selection, measurement methods, and interpretation.”1
case control studies – Observational studies that are retrospective in nature where one group of individuals has already contracted a particular disease or outcome of interest and is compared to a group that has not contracted the outcome in an attempt to determine what the exposure was that caused the disease in the one group and not the other. Provides Level 3 Evidence.
causality – The determination of “cause and effect”. The ability to be able to state with certainty that “A” causes “B” (a specific exposure has been shown to cause a specific outcome).
causal inference – The actual process of identifying causal relationships.
cohort studies – Prospective observational studies where one group of individuals has been exposed to a putative causal agent or risk factor (e.g., tobacco), while the other group has not (no exposure to tobacco). Both groups are then followed by the researcher to measure the “incidence” or development of the disease/outcome of interest, e.g., lung cancer, and to determine if there is a significant difference between the exposed and unexposed groups. Provides Level 2 Evidence.
correlation – “Linear association between two continuous (numerically-based) or ordinal (rank based) variables. The measure of the correlation is the correlation coefficient, which ranges from 1 (perfect positive/direct association) through 0 (no association) to a -1 (perfect negative/indirect association).”1 The closer the value is to a 1, (eg. r = 0.8) the stronger the association, the closer the value is to a 0, the weaker the association (eg. r = 0.3).
dose-response relationship – The higher the exposure to a suspected causal agent, the odds of getting the disease increase, conversely if the exposure to the agent is lowered, the odds of getting the disease (outcome) are reduced. (ie. number of cigarettes smoked).
literature (narrative) reviews – Publications where the author or several authors have conducted a search of the literature on a particular topic of interest, which is often broad in nature. It entails a discussion of their findings from a theoretical or contextual view but is primarily based on the opinions of the author(s). These reviews are difficult to reproduce since pre-established criteria for inclusion of articles often are not identified. Literature reviews are not considered original research and thus fall into the lowest level of evidence along with case studies, expert opinions and editorials.
meta-analysis – A statistical method used in systematic reviews to combine the data from a number of homogenous studies to quantitatively analyse their combined outcome. (Considered a Secondary Study and provides Level 1 Evidence).
odds ratio (OR) – “Measure of association obtained from a case-control study.”1 The OR is the proportion of patients with the target event divided by the proportion without the event.
randomized controlled trial (RCT) – An experimental prospective study where individual study participants are randomly assigned to either an intervention group(s) or control group(s) for a specified period of time after which a particular outcome(s) is/are measured. Considered as Level 1 Evidence.
relationship – Analogous to association.
relative risk (RR) – “Used in the reporting of cohort studies and RCTs. The RR is the ratio of the incidence of the disease or death among the exposed to the incidence of the unexposed.”1
risk factor – A specific exposure that has been determined by application of specific criteria such as the Bradford Hill criteria, or other accepted criteria, to be causal of a particular outcome or disease (causality has been determined).
statistical significance – “A measure of whether a particular result is unlikely to be due to chance. Statistical significance is generally specified by a value of P < 0.05, that is, less than 1 chance in 20.”1
systematic review – An analysis of the quality of the highest levels of evidence available in order to answer a very specific predetermined question that is developed using a PICO process where P = Population/Problem, I = Intervention, C = Comparison group, and O = Outcome of the study. Additionally, very strict inclusion/exclusion criteria are applied for the execution of the literature search (Considered Secondary Studies and provides Level 1 Evidence).