guessology

Last edited 10/2019 and last reviewed 04/2022

Guessology is a term I have used with my GP trainees when discussing the use of evidence in clinical medicine (1). The term is there to emphasise that evidence is an essential part of clinical decision making - but that evidence also has limitations that we, as primary care clinicians, have to evaluate and apply in order to use in clinical practice.

Examples of "guessology":

  • criteria that studies have for patients being included in particular clinical studies are always defined. However in clinical practice, clinicians are often applying evidence from clinical studies to patients who, given the strict inclusion and exclusion criteria for such studies, would not be included in a study. We are in effect, inducing a hypothesis of effect, given the patient in front of us would not be included in the study that we are basing a therapeutic action on.

  • similarly some actions undertaken by clinicians are not based on any randomised blinded controlled trial - but are accepted based on clinical practice and consensus - such as the use of intravenous loop diuretics in a presentation of acute left ventricular failure - there has never been a randomised blinded placebo controlled trial in this scenario but this is an intervention that is accepted based on cause and effect seen in countless emergency medicine scenarios.

Medical therapeutic decisions can then never be totally deterministic, with simple predictable cause and effect, as each patient we are dealing with is an individual and we are basing evidence on populations where an intervention has been made. Also we cannot increase the accurary of applying evidence to real-life clinical situations by including multiple variables because of the constraints of the "combinatorial explosion" (2).

Therefore "guessology" is really a term reflecting the art of clinical medicine - and that the art of clinical medicine works cooperatively with the science in order to determine decisions in the best interests of individual patients.

Reference: