JAMAevidence Glossary

Terms are derived from Users' Guides to the Medical Literature: A Manual for Evidence-Based Practice, 2nd Edition and The Rational Clinical Examination: Evidence-Based Clinical Diagnosis. Updated December 2009.
Download a PDF of the glossary (179 KB). (Adobe Reader is required to open or print PDF files.)





T-test
A parametric statistical test that examines the difference between the means of 2 groups of values.

Target condition
In diagnostic test studies, the condition the investigators or clinicians are particularly interested in identifying (eg, tuberculosis, lung cancer, or iron-deficiency anemia).

Target endpoints
In intervention studies, the condition the investigators or clinicians are particularly interested in identifying and in which it is anticipated the intervention will decrease (eg, myocardial infarction, stroke, or death) or increase (eg, ulcer healing). See also Cohort study.

Target-negative
In diagnostic test studies, patients who do not have the target condition.

Target-positive
In diagnostic test studies, patients who do have the target condition.

Tertiary care
Medical care provided to a patient when referred by one health professional to another with more specialized qualifications or interests. There are two levels of referred care: secondary and tertiary. Secondary care is usually provided by a broadly skilled specialist such as a general surgeon, general internist, or obstetrician. Synonymous with Referred care.

Tertiary care center
A medical facility that receives referrals from both primary and secondary care levels and usually offers tests, treatments, and procedures that are not available elsewhere. Most tertiary care centers offer a mixture of primary, secondary, and tertiary care services so that it is the specific level of service rendered rather than the facility that determines the designation of care in a given study. See also Referred care; Primary care.

Test threshold
The probability below which the clinician decides a diagnosis warrants no further consideration. See also Treatment threshold.

Theoretical saturation
The point at which iterations among data collection, analysis, and theory development yield a well-developed concept, and further observations yield minimal or no new information to further challenge or elaborate the concept. See also Informational redundancy.

Theory
Theory consists of concepts and their relationships.

Theory triangulation
Theory triangulation is a process whereby emergent findings are corroborated with existing social science theories. See also Triangulation.

Threshold NNT
Maximum number needed to treat (NNT) or number needed to harm (NNH) accepted as justifying the benefits and harms of therapy. See also Test threshold.

Time-series design
In this study design, data are collected at several times both before and after the intervention; data collected before the intervention allow the underlying trend and cyclical (seasonal) effects to be estimated. Data collected after the intervention allow the intervention effect to be estimated while accounting for underlying secular trends. The time-series design monitors the occurrence of outcomes or endpoints over a number of cycles and determines whether the pattern changes coincident with the intervention.

Tinel sign
Paresthesias in the distribution of the median nerve when the clinician taps on the distal wrist crease over the median nerve.

Treatment effect
The results of comparative clinical studies can be expressed using various intervention effect measures. Examples are absolute risk reduction (ARR), relative risk reduction (RRR), odds ratio (OR), number needed to treat (NNT), and effect size. The appropriateness of using these to express an intervention effect and whether probabilities, means, or medians are used to calculate them depend on the type of outcome variable used to measure health outcomes. For example, ARR, RRR, and NNT are used for dichotomous variables, and effect sizes are normally used for continuous variables. See also Absolute risk reduction; Relative risk reduction; Odds ratio; Number needed to treat.

Treatment target
The manifestation of illness (a symptom, sign, or physiological abnormality) toward which a treatment is directed. See also Endpoint.

Treatment threshold
Probability above which a clinician would consider a diagnosis confirmed and would stop testing and initiate treatment. See also Test threshold.

Trial of therapy
In a trial of therapy, the physician offers the patient an intervention, reviews the impact of the intervention on that patient at some subsequent time, and, depending on the impact, recommends either continuation or discontinuation of the intervention.

Triangulation
1. More than one investigator collects and analyzes the raw data, such that the findings emerge through consensus among investigators. 2. In qualitative research, an analytic approach in which key findings are corroborated using multiple sources of information.

Trigger orders
Orders in response to which the computer decision support system (CDSS) would initiate action.

Trim-and-fill method
When publication bias is suspected in a systematic review, investigators may attempt to estimate the true intervention effect by removing, or trimming, small positive-result studies that do not have a negative-result study counterpart and then calculating a supposed true effect from the resulting symmetric funnel plot. The investigators then replace the positive-result studies they have removed and add hypothetical studies that mirror these positive-result studies to create a symmetric funnel plot that retains the new pooled effect estimate. This method allows the calculation of an adjusted confidence interval and an estimate of the number of missing trials.

True negative
Those whom the test correctly identifies as not having the target disorder.

True positive
Those whom the test correctly identifies as having the target disorder.

Type I error
An error created by rejecting the null hypothesis when it is true (ie, investigators conclude that a relationship exists between variables when it does not).

Type II error
An error created by accepting the null hypothesis when it is false (ie, investigators conclude that no relationship exists between variables when, in fact, a relationship does exist).