The technical terms used in the reviews, explained in plain clinical language.
The measurement taken before the intervention starts, used as the reference point for calculating change.
The placebo version of a physical procedure. The device is applied and switched on, but the dose is not delivered, so participants cannot tell it from the real treatment.
Neither the participant nor the outcome assessor knows which treatment was given, preventing expectation on either side from contaminating the measurement.
A design that tests two interventions at once by combining both options of each into four groups, answering two questions with one sample.
The primary outcome is the study's main question, the one the sample size was powered for. Secondary outcomes are measured too, but without a guarantee of enough participants to detect differences.
Expresses the size of a difference in standard deviation units, allowing comparison across different scales. Conventionally 0.2 is small, 0.5 moderate and 0.8 large.
The probability of observing a difference at least as large as the one found if there were no real difference. The usual 0.05 cut-off is a convention, not a boundary between true and false.
The range of values compatible with the observed data. A 95% interval that includes zero means the study could not rule out no effect.
The tendency of extreme values on a first measurement to fall closer to the average on a second one, purely because part of the extreme value was random variation.
A way of handling missing data by estimating each missing value several times from the available information and combining the results, rather than discarding incomplete participants.
Concluding there is no difference when there actually is one, usually because the sample was too small to detect it.
A physiological or laboratory measure used in place of what actually matters to the patient. Improving the surrogate does not guarantee improving the real outcome.
How far a measurement repeats itself when nothing about the patient has changed, usually reported between different assessors and within the same assessor over time. An unreliable measure produces variation that looks like a treatment effect but is procedural inconsistency.
Rates the methodological quality of a randomised controlled trial across 11 criteria. The first criterion covers external validity and is not counted, so the score runs from 0 to 10. It measures how the trial was conducted and reported, not whether the result matters clinically. Official scores are assigned by trained PEDro raters.
A systematic review gathers, by explicit and reproducible method, all studies answering the same question; a meta-analysis combines their results into a single estimate. Comparisons made between different studies inside a meta-analysis are not equivalent to a randomised comparison within one study.
Tests whether a treatment's effect on an outcome runs through an intermediate mechanism. No mediation means the treatment may still work, but not by the assumed route.
A laboratory measure of how well the nervous system can dampen its own pain: a painful stimulus is rated alone, then rated again while a second pain is applied elsewhere (usually a hand in cold water). In a well-functioning system the first stimulus hurts less.
The counterpart to CPM: a series of identical painful stimuli is delivered and the growth in pain from first to last is measured. More growth means the system is amplifying rather than filtering.
A non-invasive technique passing a weak continuous current between two scalp electrodes. It does not fire neurons; it shifts how readily the region under the electrode fires.
The pathways running from the brain down to the spinal cord that reduce how much pain signal reaches awareness. The body's endogenous brake on pain.
A state in which the central nervous system responds in an amplified way, so pain no longer reflects the amount of tissue damage.
A disproportionate worsening of symptoms after physical activity, often delayed by hours or a day.
A fibromyalgia-specific questionnaire combining physical function, overall impact and symptoms into a single score, where higher means greater disease impact.
Separates pain intensity from how much pain interferes with activities, mood, sleep and relationships, since the two do not always improve together.
Assesses sleep quality over the past month across seven components; a higher score means worse sleep.
A self-report questionnaire for the intensity of depressive symptoms. It measures symptom severity, not diagnosis.
A questionnaire system developed by the NIH to measure patient-reported outcomes such as pain, fatigue and anxiety on standardised, comparable scales.
A continuous line on which the patient marks pain intensity between two extremes; the measure is the distance to the mark.