The hazard ratio is a relative effect measure for time-to-event data in clinical trials. It compares the instantaneous event rates, also called hazards, of two groups over the observed time and is frequently used for endpoints such as overall survival or time to progression. Its value and its confidence interval describe the direction and precision of an estimated treatment effect, but not directly an absolute probability of an event.<\/p>\n
A hazard is the conditional rate at which an event occurs at a specific point in time in individuals who have remained event-free up until immediately prior to that. The hazard ratio puts this hazard in the investigational arm in relation to the hazard in the comparator arm. With the usual coding, a value below 1 indicates a lower instantaneous event hazard in the investigational arm, a value above 1 a higher one. Which direction is favorable always depends on the definition of the event.<\/p>\n
A hazard ratio of 0.70 does not mean that exactly 30 percent fewer participants suffer an event at any time point or that the individual risk is reduced by 30 percent. It summarizes a model for the ratio of event rates over the follow-up. For the interpretation, the definition of the endpoint, the censoring rules, the observation period and the underlying assumptions are therefore indispensable.<\/p>\n
In many trials, the hazard ratio is estimated using a Cox model. For simple interpretation, this model assumes that the hazards of the groups behave proportionally. If treatment effects change over time or survival curves cross, a single hazard ratio can only incompletely summarize the clinical development. The statistical analysis plan should therefore determine the model, covariates, handling of intercurrent events and sensitivity analyses in advance.<\/p>\n
The confidence interval shows the statistical uncertainty of the estimation. For a ratio whose reference value is 1, an interval that does not include 1 speaks against the corresponding null hypothesis under the predefined test procedure. A statistical conclusion, however, does not replace the evaluation of the absolute frequencies, the follow-up duration and the clinical relevance of the effect.<\/p>\n
According to ICH E9(R1), planning begins with a precise clinical question and an estimand, i.e., a description of the treatment effect to be estimated. The hazard ratio is not an estimand in itself, but can be a parameter used to estimate an effect related to time-to-event data. Discontinuation of the assigned treatment, subsequent therapies or death can influence the definition and interpretation of the effect and must be taken into account in advance.<\/p>\n
Kaplan-Meier curves supplement the parameter because they show the temporal course of the estimated event-free probability. They neither prove the proportionality assumption nor replace a complete analysis. The clinical study report must comprehensibly present the pre-planned analysis, all relevant deviations and the supporting evaluations.<\/p>\n
The hazard ratio is not to be confused with the risk ratio. The risk ratio compares the cumulative probability of an event in two groups at a specified time point or within a specified time period. It therefore requires a clear definition of the time point considered and does not treat the time points of individual events themselves as information.<\/p>\n
The hazard ratio, in contrast, uses the temporal location of the events and can take censored observations into account under its assumptions. Two trials can yield different hazard ratios with the same proportion of events at the end of the follow-up if events are temporally distributed differently. Conversely, neither a risk ratio nor an absolute risk difference can be derived from a hazard ratio alone. With competing risks, it must additionally be examined whether the chosen analysis adequately reflects the clinical question.<\/p>\n
For time-dependent endpoints, the definition of event and censoring, the quality of follow-up and the choice of analysis time point essentially influence the validity. Incomplete or differently documented follow-up can distort group comparisons. Therefore, the protocol and statistical analysis plan must consistently determine the analysis population, the evaluation of missing data and the handling of intercurrent events.<\/p>\n
Full-service CROs such as Mediconomics support the translation of the study objective into an estimand, endpoint definition and statistical analysis plan, the programming of time-to-event analyses as well as tables, figures and listings for the clinical study report. This also includes data checks on event data, the documentation of censoring rules and the planning of justified sensitivity analyses.<\/p>\n
Is a hazard ratio below 1 always an advantage?<\/strong><\/p>\n Only if the event is defined unfavorably and the quotient is formed of investigational arm to comparator arm. Endpoint definition and calculation direction must always be read along.<\/p>\n Can an absolute risk reduction be calculated from the hazard ratio?<\/strong><\/p>\n No. For this, absolute event probabilities at a defined time point and further assumptions are additionally required.<\/p>\n Is the hazard ratio also usable with non-proportional hazards?<\/strong><\/p>\n It can be estimated, but its summarizing interpretation is then limited. Predefined supplementary analyses can be more appropriate.<\/p>\n The hazard ratio is a relative effect measure for time-to-event data in clinical trials. It compares the instantaneous event rates, also called hazards, of two groups over the observed time and is frequently used for endpoints such as overall survival or time to progression. Its value and its confidence interval describe the direction and precision […]<\/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":"set","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-6909","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\/6909","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":2,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary\/6909\/revisions"}],"predecessor-version":[{"id":7462,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary\/6909\/revisions\/7462"}],"wp:attachment":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/media?parent=6909"}],"wp:term":[{"taxonomy":"glossary-cat","embeddable":true,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary-cat?post=6909"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Regulatory references<\/h2>\n
\n