Modelling survival data from hepatorenal syndrome acute kidney injury

Hepatorenal syndrome (HRS) is a common complication of advanced cirrhosis. It is characterised by renal failure and major disturbances in circulatory function and it is caused by intense vasoconstriction of the renal circulation. Liver transplantation is the treatment of choice for HRS. However, it...

Descripción completa

Detalles Bibliográficos
Autor: Morató Catafal, Alba
Tipo de recurso: tesis de maestría
Fecha de publicación:2022
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/364893
Acceso en línea:https://hdl.handle.net/2117/364893
Access Level:acceso abierto
Palabra clave:Survival analysis (Biometry)
Medical statistics
Survival
Hazard
Competing
Landmark
Multi-state
Cirrhosis
Kidney
Hepatorenal
Anàlisi de supervivència (Biometria)
Estadística mèdica
Classificació AMS::62 Statistics::62N Survival analysis and censored data
Àrees temàtiques de la UPC::Matemàtiques i estadística::Estadística matemàtica
Descripción
Sumario:Hepatorenal syndrome (HRS) is a common complication of advanced cirrhosis. It is characterised by renal failure and major disturbances in circulatory function and it is caused by intense vasoconstriction of the renal circulation. Liver transplantation is the treatment of choice for HRS. However, it is not always possible owing to the short survival expectancy. Also, in the last years it has also been proved that the syndrome can be reversed by the simultaneous administration of albumin and arterial vasoconstrictors such as terlipressin. The main aim of this study is to assess the survival of HRS patients that are being treated with albumin, vasopressors or both of them and check whether their response to those treatments is determinant for their survival. To do so, classical competing risks analysis will be performed but also, some other techniques, such as the inclusion of time-varying covariates and the use of landmarking and multi-state models will be introduced. Finally, the results obtained with all the different methods will be compared and discussed.