This article is contributed wholly or in part by OrthoEvidence (OE). Section editors and authors for OrthopaedicsOne may have edited its content or added new information. The use of information from OrthoEvidence should not be construed as support for or endorsement by that organization for any new information added by OrthopaedicsOne members, or for any editing of the original content. Learn more about OrthoEvidence

Absolute risk increase

Difference in the absolute risk (percentage or proportion of patients with an outcome) in the exposed versus the unexposed intervention. Typically used with a harmful exposure.

Absolute risk reduction

Difference in the absolute risk (percentage or proportion of patients with an outcome) in the exposed (in intervention) (experimental event rate) versus the unexposed (control event rate). Use restricted to a beneficial exposure or intervention.

Adjusted analysis

An adjusted analysis takes into account differences in prognostic factors between groups that may influence the outcome. For instance, in comparison between experimental treatment and control groups, if the experimental group is on average older, and therefore at higher risk of an adverse outcome than the control group, the adjusted analysis will show a larger treatment effect than the unadjusted analysis.

Alpha error

The probability of erroneously concluding there is a difference between two treatments when there is no difference. Typically, investigators decide on the chance of a false-positive result they are willing to accept when they plan the sample size for a study.

Baseline risk

The risk of an adverse outcome in the control group of an experiment. Synonymous with control event rate.

Bayesian analysis

An analysis that starts with a particular probability of an event (the prior probability) and incorporates new information to generate a revised probability (a posterior probability).

Before-after trial

Investigation of an intervention in which the investigators compare the status of patients before and after the intervention.


A systematic tendency to produce an outcome that differs from the underlying truth.

(1) Channeling effect or channeling bias

The tendency of clinicians to prescribe treatment based on a patient’s prognosis. As a result of the behavior, comparisons between treated and untreated patients will yield a biased estimate of treatment effect.

(2) Data completeness bias

Using the information system to log episodes in the treatment group and using a manual system in the noncomputer decision support system group can create a data completeness bias.

(3) Detection bias

The tendency to look more carefully for an outcome in one of two groups being compared.

(4) Incorporation bias

When investigators study a diagnostic test that incorporates features of the target outcome.

(5) Interviewer bias

Greater probing by an interviewer in one of two groups being compared.

(6) Publication bias

Publication bias occurs when the publication of research depends on the direction of the study results and whether they are statistically significant.

(7) Recall bias

Recall bias occurs when patients who experience an adverse outcome have a different likelihood of recalling an exposure than the patients who do not have an adverse outcome, independent of the true extent of exposure.

(8) Surveillance bias

Synonymous with detection bias; the tendency to look more carefully for an outcome in one of two groups being compared.

(9) Verification bias

Results of a diagnostic test influence whether patients are assigned to a treatment group.

Blind (or blinded or masked)

The participant of interest is unaware of whether patients have been assigned to the experimental or control group. Patients, clinicians, those monitoring outcomes, judicial assessors of outcomes, data analysts, and those writing the paper all can be blinded or masked. To avoid confusion, the term masked is preferred in studies in which vision loss of patients is an outcome of interest.

Bootstrap technique

A statistical technique for estimating parameters such as standard errors and confidence intervals based on resampling from an observed data set with replacement.

Case reports and case series

Descriptions of individual patients. A study reporting on a consecutive collection of patients treated in a similar manner, without a control group. For example, a surgeon might describe the characteristics of an outcome for 100 consecutive patients with cerebral ischemia who received a revascularization procedure.

Case-control study

A study designed to determine the association between an exposure and outcome in which patients are sampled by outcome (some patients with the outcome of interest are selected and compared with a group of patients who have not had the outcome), and the investigator examines the proportion of patients with the exposure in the two groups.

Chi square test

A statistical test that examines the distribution of categorical outcomes in two groups, the null hypothesis of which is that the underlying distributions are identical.

Clinical prediction rules (or clinical decision rules)

A clinical prediction rule is generated by initially examining, and ultimately combining, numerous variables to predict the likelihood of a current diagnosis or a future event. Sometimes, if the likelihood is sufficiently high or low, the rule generates a suggested course of action.


Interventions other than treatment under study that may be differentially applied to experimental and control groups and, therefore potentially bias the results of a study.


A group of persons with a common characteristic or set of characteristics. Typically, the group is followed up for a specified period to determine the incidence of a disorder or complications of an established disorder (prognosis).

Cohort study (or cohort analytic study)

Prospective investigation of the factors that might cause a disorder in which a cohort of individuals who do not have evidence of an outcome of interest but who are exposed to the putative cause are compared with a concurrent cohort who also are free of the outcome but not exposed to the putative cause. Both cohorts then are followed up to compare the incidence of the outcome of interest.

Conditional probabilities

The probability of a particular state, given another state. That is, the probability of A, given B—Probability (A/B).

Confidence interval

Range of two values within which it is probable that the true value lies for the entire population of patients from whom the study patients were selected.


A factor that distorts the true relationship of the study variable of interest by virtue of also being related to the outcome of interest. Confounders often are compared. Randomized studies are distributed unequally among the groups being less likely to have their results distorted by confounders than are observational studies.


Contamination occurs when participants in either the experimental or control group receive the intervention intended for the other arm of the study.

Continuous variables

A variable that theoretically can take any value and in practice can take a large number of values with small differences between them.


The magnitude of the relationship between different variables or phenomena.

Correlation coefficient

A numerical expression of the strength of the relationship between two variables, which can take values from -1.0 to 1.0

Cost analysis

If two strategies are analyzed but only costs are compared, this comparison would inform only the resource-use half of the decision (the other half being the expected outcomes) and is termed a cost analysis.

Cost-benefit analysis

A form of economic analysis in which the costs and the consequences (including increases in the length and quality of life) are expressed in monetary terms.

Cost-effectiveness analysis

An economic analysis in which the consequences are expressed in natural units. Some examples would include cost per life saved or cost per unit of blood pressure lowered.

Cost-minimization analysis

An economic analysis conducted in situations where the consequences of the alternatives are identical, and so the only issue is their relative costs.

Cost-utility analysis

A type of cost effectiveness analysis in which the consequences are expressed in terms of life-years adjusted by peoples’ preferences. Typically, one considers the incremental cost per incremental gain in quality adjusted life-years.

Cox regression model

A regression technique that allows adjustment for known differences in baseline characteristics between experimental and control groups applied to survival data.


Searching a data set for differences between groups on particular outcomes, or in subgroups of patients, without explicit a priori hypotheses.

Decision analysis

A systematic approach to decision-making under conditions of uncertainty. It involves identifying all available alternatives and estimating the probabilities of potential outcomes associated with each alternative, valuing each outcome, and, on the basis of the probabilities and values, arriving at a quantitative estimate of the relative merit of the alternatives.

Decision under risk

A decision against nature in which a probability distribution on the states of nature is known.

Decision under uncertainty

A decision against nature with no knowledge about the likelihood of the various states of nature.

Decision tree

Most clinical decision analyses are built as decision trees. Articles about clinical decision analyses usually will include one or more diagrams showing the structure of the decision tree used for the analysis.

Dichotomous outcomes

Yes or no outcomes that either happen or do not happen, such as reoperation, infection, and death.

Dichotomous variable

A variable that can take one of two values, such as male or female, dead or alive, having an infection or not having an infection.

Direct costs

The costs of all resources that can be traced to a particular intervention.

Economic analysis

A set of formal, quantitative methods used to compare two or more treatments, programs, or strategies with respect to their resource use and their expected outcomes.

Economic evaluation

Comparative analysis of alternative courses of action in terms of their costs and consequences. The effect size is the difference in outcomes between the intervention and control groups divided by some measure of variability, typically the standard deviation.

Effect size

The difference in the outcomes between the intervention and control groups divided by some measure of the variability, typically the standard deviation


Achieving the maximal increment in health benefit for a given quantity of resources.

Evidence-based medicine

The conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients. The practice of evidence-based medicine requires integration of individual clinical expertise and patient preferences with the best available external clinical evidence from systematic research.

Fold back analysis

The process of solving a decision tree by working backward.


The ability to generalize the findings of a study to a larger group of similar people.

Hawthorne effect

Human performance that is improved when participants are aware that their behavior is being observed. In a treatment study, the treatment is deemed effective when it actually is ineffective. In a diagnosis study, the patient does not suffer from the target condition, but the test suggests the patient does.

Hazard ratio

Investigators may compute the relative risk with time, as in a survival analysis, and call it a hazard ratio, the weighted relative risk over the entire study.

Health-related quality of life

Measurements of how people are feeling or the value they place on their health state.


Number of new cases of disease occurring during a specified period; expressed as a percentage of the number of people at risk.


Differences between patients or differences in the results of different studies.

Intention-to-treat principle or intention-totreat analysis

Analyzing patient outcomes based on which group into which they were randomized regardless of whether they actually received the planned intervention. This analysis preserves the power of randomization, thereby maintaining that important unknown factors that influence outcome are likely equally distributed in each comparison group.

Linear regression

The term used for a regression analysis when the dependent or target variable is a continuous variable and the relationship between the dependent and independent variables is thought to be linear.

Logistic regression

A term used for a regression analysis in which the dependent or target variable is dichotomous and which uses a model that relies on logarithms.


An overview that incorporates a quantitative strategy for combining the results of several studies into one pooled or summary estimate.

Multivariable regression equation

A type of regression that provides a mathematical model that explains or predicts the dependent or target variable by simultaneously considering all of the independent or predictor variables.

Multivariate analysis

An analysis that simultaneously considers a number of predictor variables.

Null hypothesis

In the hypothesis-testing framework, the starting hypothesis the statistical test is designed to consider and, possibly, reject.

Number needed to harm

The number of patients who would need to be treated during a specific period before one adverse side effect of the treatment will occur. It is the inverse of the absolute risk increase.

Number needed to treat

The number of patients who need to be treated during a specific period to prevent one bad outcome. When discussing number needed to treat it is important to specify the treatment, its duration, and the bad outcome being prevented. It is the inverse of the absolute risk reduction.

Observational studies (or observational study design)

Studies in which patient or physician preference determines whether a patient receives treatment or control.


A ratio of probability of occurrence to nonoccurrence of an event.

Odds ratio

A ratio of the odds of an event in an exposed group to the odds of the same event in a group that is not exposed.

Opportunity cost

The potential benefit given up when the choice of one alternative precludes the selection of a different alternative.


In a comparison of two interventions, the ability to detect a difference between the two experimental conditions if one in fact exists.

Prognostic factors

Patient or study participant characteristics that confer increased or decreased risk of a positive or adverse outcome.

Prognostic study

A study that enrolls patients at a point in time and follows him or her forward to determine the frequency and timing of subsequent events.

P value

The probability that results as or more extreme than those observed would occur if the null hypothesis were true and the experiment were repeated over and over.

Quality-adjusted life-year

A unit of measure for survival that accounts for the effects of suboptimal health status and the resulting limitations in quality of life. For example, if a patient lives for 10 years and his or her quality of life is decreased by 50% because of chronic lung disease, his or her survival would be equivalent to 5 quality-adjusted life years.

Random allocation

A sample derived by selecting sampling units (individual patients) such that each unit has an independent and fixed (generally equal) chance of selection. Whether a given unit is selected is determined by chance, for example, by a table of randomly ordered numbers. Allocation of individuals to groups by chance, usually done with the aid of a table of random numbers. Not to be confused with systematic allocation (on even and odd days of the month) or allocation at the convenience or discretion of the investigator.

Randomized trial

Experiment in which individuals are allocated randomly to receive or not receive an experimental preventative, therapeutic, or diagnostic procedure and then are followed up to determine the effect of the intervention.

Relative risk

Ratio of the risk of an event among an exposed population to the risk among the unexposed.

Relative risk reduction

An estimate of the proportion of baseline risk that is removed by the therapy, it is calculated by dividing the absolute risk reduction by the absolute risk in the control group.


Refers to consistency or reproducibility of data.


The proportion of people who truly have a designated disorder who are so identified by the test. The test may consist of, or include, clinical observations.

Sensitivity analysis

Any test of the stability of the conclusions of a healthcare evaluation over a range of probability estimates, value judgments, and assumptions about the structure of the decisions to be made. This may involve the repeated evaluation of a decision model in which one or more of the parameters of interest are varied.


The proportion of people who truly are free of a designated disorder who are so identified by the test. The test may consist of, or include, clinical observations.

Studies or study design

The way a drug study is organized or constructed.

(1) Phase I studies

Studies that investigate a drug’s physiologic effect or ensure that it does not manifest unacceptable early toxicity, often done with healthy volunteers.

(2) Phase II studies

Initial studies on patients, which provide preliminary evidence of possible drug effectiveness.

(3) Phase III studies

Randomized control trials designed to definitively establish the magnitude of drug benefit.

(4) Phase IV studies or postmarketing surveillance studies

Studies done after the effectiveness of a drug has been established and the drug marketed, typically to establish the frequency of unusual toxic effects.

Treatment effect

The results of comparative clinical studies can be expressed using various treatment effect measures. Examples are absolute risk reduction, relative risk reduction, odds ratio, number needed to treat, and effect size. The appropriateness of using these to express a treatment effect and whether probabilities, means, or medians are used to calculate them depends on the type of outcome variable used to measure health outcomes. For example, relative risk reduction is used for dichotomous variables, and effect sizes are normally used for continuous variables.

Utility measures

Measures that provide one number that summarizes all of health related quality of life are preference- or value weighted; these have the preferences or values anchored to death and full health and are called utility measures.