Acase series (also known as aclinical series) is atype ofmedical research study that tracks subjects with a known exposure, such as patients who have received a similar treatment,[1] or examines their medical records for exposure and outcome. Case series may beconsecutive[2] ornon-consecutive,[3] depending on whether all cases presenting to the reporting authors over a period were included, or only a selection. When information onmore than three patients is included, the case series is considered to be a systematic investigation designed to contribute to generalizable knowledge (i.e.,research), and therefore submission is required to aninstitutional review board (IRB).[4] Case series usually contain demographic information about the patient(s), for example, age, gender, ethnic origin. etc.
Case series have a descriptivestudy design; unlike studies that employ an analytic design (e.g.cohort studies,case-control studies orrandomized controlled trials), case series do not, in themselves, involvehypothesis testing to look for evidence ofcause and effect (thoughcase-only analyses are sometimes performed ingenetic epidemiology to investigate the association between an exposure and agenotype[5]). Case series are especially vulnerable toselection bias; for example, studies that report on a series of patients with a certain illness and/or a suspected linked exposure draw their patients from a particularpopulation (such as a hospital or clinic) which may not appropriately represent the wider population.
Internal validity of case series studies is usually very low, due to the lack of a comparator group exposed to the same array of intervening variables.
For example, the effects seen may be wholly or partly due to intervening effects such as theplacebo effect,Hawthorne effect,Rosenthal effect, time effects, practice effects or the natural history effect. Calculating the difference in effects between two treatment groups assumed to be exposed to a very similar array of such intervening effects allows the effects of these intervening variables to cancel out. Hence only the presence of a comparator group, which is not a feature of case-series studies, will allow a valid estimate of the true treatment effect.[6]