{"id":6746,"date":"2026-04-30T17:50:28","date_gmt":"2026-04-30T15:50:28","guid":{"rendered":"https:\/\/mediconomics.com\/glossar\/relative-risk\/"},"modified":"2026-04-30T17:50:28","modified_gmt":"2026-04-30T15:50:28","slug":"relative-risk","status":"publish","type":"glossary","link":"https:\/\/mediconomics.com\/en\/glossar\/relative-risk\/","title":{"rendered":"Relative Risk"},"content":{"rendered":"

Relative Risk<\/strong> (RR, relative risk) describes, in clinical research, the ratio of the probability of an event in an exposure or intervention group to the probability in a comparator group. It is one of the key effect measures in randomized clinical trials, cohort studies, and benefit\u2013risk assessments. An RR of 1 indicates no difference; values >1 indicate a higher risk under the exposure, and values <1 a lower risk. <\/p>\n

For stakeholders outside biostatistics, RR is particularly helpful because it can be communicated intuitively as \u201cx times higher\u201d or \u201cx times lower.\u201d At the same time, this apparent simplicity is a common source of error: without context on baseline risk, population definition, and observation period, RR can be misinterpreted and over- or underweighted in decision-making discussions. <\/p>\n

Definition and calculation<\/h2>\n

Formally: \\(RR = rac{p_1}{p_0}\\), where \\(p_1\\) is the probability of the event in the intervention or exposure group and \\(p_0\\) is the probability of the event in the control or non-exposure group. The event may be defined, for example, as the occurrence of an adverse event, disease progression, or a primary endpoint. For practical interpretation, it is important that RR is a relative<\/em> measure: it scales with the baseline risk of the comparator group. <\/p>\n

In study reports, RR is often presented together with confidence intervals and p-values. The confidence interval is typically calculated on the log scale because the distribution of RR estimators is asymmetric. For rare events, RR can become unstable; in such cases, exact methods or appropriate models (e.g., Poisson or log-binomial regression) are commonly used. <\/p>\n

It is also important to consider which<\/em> population forms the basis of the calculation. An RR in the intention-to-treat population answers a different question than an RR in a per-protocol population or a safety population. In audits and inspections, it is often checked whether population definitions have been applied consistently across the protocol, SAP, tables\/listings, and the Clinical Study Report. <\/p>\n

Interpretation: clinical relevance vs. statistical significance<\/h2>\n

RR only provides a robust clinical interpretation when considered together with absolute risk. Example: if a rare event increases from 0.1% to 0.2%, this corresponds to an RR of 2, but in absolute terms it is often less serious than an increase from 10% to 20% (also RR=2). Therefore, decision-making processes often use additional metrics such as absolute risk difference and number needed to treat or harm. <\/p>\n

For regulatory decisions and HTA assessments, not only statistical significance matters, but also clinical relevance: effect size, precision, consistency across subgroups, and plausibility in the context of mechanism of action and safety profile.<\/p>\n

A practical checkpoint is robustness to sensitivity analyses: if RR changes materially under alternative counting rules (e.g., confirmed endpoints vs. investigator-reported), different approaches to missing data, or adjustment for relevant covariates, the interpretation should be correspondingly cautious.<\/p>\n

Distinction from odds ratio and hazard ratio<\/h2>\n

RR is directly interpretable as a ratio of probabilities and is preferred when the event probability is observable (e.g., in RCTs or cohorts). The odds ratio<\/strong> compares odds rather than probabilities and can differ substantially from RR when events are common. In case-control studies, the odds ratio is often the only effect measure that can be estimated directly, but it is sometimes misunderstood when interpreted as RR. <\/p>\n

The hazard ratio<\/strong> comes from time-to-event analyses and compares instantaneous event rates over time. It is not identical to RR, but under proportional hazards and certain assumptions it may indicate a similar direction. In clinical trials with censored data (e.g., progression-free survival), the hazard ratio is often the standard. <\/p>\n

For communication in the Clinical Study Report, it is advisable to explicitly describe in the Methods section which effect measure is primary and why. This reduces the risk that readers overlook metrics or make invalid comparisons, such as comparing RR from binary endpoints with hazard ratios from time-to-event analyses. <\/p>\n

Typical use cases in CRO and sponsor settings<\/h2>\n

In clinical project management and statistical planning, RR plays a role in several phases: operationalizing endpoints, sample size planning (when binary endpoints are used), and interim and final analyses. In safety reviews, RR is used to compare adverse events between treatment arms, particularly when events are reported as incidences per patient group. <\/p>\n

In non-interventional studies and real-world evidence analyses, RR also serves as an easily communicable metric, although confounding and bias require greater consideration. Adjusted models are often used here, such as propensity score methods or multivariable regressions, to estimate an approximately causal RR. <\/p>\n

A common misunderstanding in project teams is that a \u201cgood\u201d RR automatically constitutes regulatorily convincing evidence. Regulators, however, also assess data quality, handling of protocol deviations, missing-data strategy, and consistency with secondary endpoints and safety signals. Therefore, RR should always be interpreted as part of the overall evidence picture. <\/p>\n

Regulatory and methodological context<\/h2>\n

From a regulatory perspective, no specific effect measure is mandated; what matters is a coherent, pre-specified statistical methodology in the protocol and transparent reporting in the study report. Relevant frameworks include ICH E9 (Statistical Principles for Clinical Trials) and ICH E3 (Clinical Study Report). For the EU, documentation and transparency requirements under the Clinical Trials Regulation (EU) No 536\/2014 are also relevant; for medical devices, methodological aspects of the MDR (EU) 2017\/745 may play a role when clinical data are used for performance and safety evaluation. <\/p>\n

In practice, sponsors should ensure that effect measures are used consistently across the protocol, SAP, CSR, and publications to avoid interpretive conflicts. In addition, a clear definition of the event (e.g., treatment-emergent adverse event, confirmed endpoint) is necessary because RR is sensitive to different counting rules. <\/p>\n

FAQ<\/h2>\n

When is relative risk better than an odds ratio?<\/strong><\/p>\n

When the event probability is directly observable and events are not extremely rare, RR is usually easier to interpret and avoids exaggerations that can arise with odds ratios for common events.<\/p>\n

Can relative risk be calculated in case-control studies?<\/strong><\/p>\n

In classic case-control designs, RR cannot be estimated directly because the sample is drawn based on outcome. The odds ratio is typically reported; for rare events, it approximates RR. <\/p>\n

How should RR be presented in the study report?<\/strong><\/p>\n

Common practice is to report RR with a 95% confidence interval, p-value, and clear information on the population definition (e.g., ITT) as well as the event definition and analysis method.<\/p>\n

Regulatory references (selection):<\/strong> ICH E9; ICH E3; Regulation (EU) No 536\/2014; Regulation (EU) 2017\/745 (MDR).<\/p>\n","protected":false},"excerpt":{"rendered":"

Relative Risk (RR, relative risk) describes, in clinical research, the ratio of the probability of an event in an exposure or intervention group to the probability in a comparator group. It is one of the key effect measures in randomized clinical trials, cohort studies, and benefit\u2013risk assessments. An RR of 1 indicates no difference; values […]<\/p>\n","protected":false},"author":10,"featured_media":0,"parent":0,"template":"","meta":{"_acf_changed":false,"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"glossary-cat":[],"class_list":["post-6746","glossary","type-glossary","status-publish","hentry"],"acf":[],"related_terms":"","external_url":"","internal_reference_id":"","_links":{"self":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary\/6746","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary"}],"about":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/types\/glossary"}],"author":[{"embeddable":true,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/users\/10"}],"version-history":[{"count":0,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary\/6746\/revisions"}],"wp:attachment":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/media?parent=6746"}],"wp:term":[{"taxonomy":"glossary-cat","embeddable":true,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary-cat?post=6746"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}