New ICU Mortality Prediction Model for Low-Income Countries

Rwanda landscape, credit Pixabay
share Share

A model to predict ICU mortality has been designed for use in low-income countries. The development team say that its use of a few easily collected variables  make it suitable for use in sites that care for critically ill patients in low-income countries. This is the first time since 1989 that a model has been developed or validated in a low-income country.

Elisabeth D. Riviello, Department of Medicine, University of Rwanda, College of Medicine and Health Sciences, Kigali, Rwanda and Department of Medicine, Division of Pulmonary, Critical Care and Sleep Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, USA, with colleagues from Canada and Rwanda report on validation of the Mortality Probability Admission Model, version III (MPM0-III) in two public ICUs in Rwanda and subsewuent development of the Rwanda Mortality Probability Model (R-MPM) for use in low-income countries.

The MPM0-III model is the only ICU risk prediction that has been validated in a large cohort and is not dependent on laboratory values. In this study it was validated using data on all adult patients admitted to the two public ICUs in Rwanda between 19 August 2013 and 6 October 2014. In=hospital mortality in the 425 adults admitted consecutively was 48.7%.

See Also: Quality Care for Critically Ill Should be Goal for All

Patients were most commonly admitted for respiratory failure or for endotracheal intubation (72.8%).  Sepsis diagnosis within 24 hours included 42.2% with  sepsis, 33.0% severe sepsis, and 20.8% septic shock. ARDS criteria according to the Berlin definition were met in 12.9% of all patients.

The MPM0-III predicted mortality with area under the receiver operating characteristic curve of 0.72 and Hosmer-Lemeshow chi-square statistic p = 0.024.

The Rwanda Mortality Probability Model (R-MPM) was developed using five variables: age, suspected or confirmed infection within 24 hours of ICU admission, hypotension or shock as a reason for ICU admission, Glasgow Coma Scale score at ICU admission, and heart rate at ICU admission. The R-MPM predicted outcomes with area under the ROC curve of 0.81 with 95%confidence interval of (0.77, 0.86), and Hosmer-Lemeshow chi-square statistic p = 0.154. The article outlines incremental change in mortality expected for a change in each independent variable, based on the odds ratio.

Riviello and co-authors suggest that the R-MPM, if validated in other critically ill patients in a broad range of settings, could potentially improve the reliability of comparisons used for critical care research and quality improvement initiatives in low-income countries.

Image credit: Pixabay

«« Open Clinical Trials: Are We Nearly There Yet?

End-of-Life Care in the ICU: WFSICCM Task Force Report »»


Riviello ED, Kiviri W, Fowler RA, Mueller A, Novack V, Banner-Goodspeed VM, Weinkauf JL,  Talmor DS, Twagirumugabe T (2016) Predicting mortality in low-income country ICUs: the Rwanda mortality probability model (R-MPM). PLoS One. 2016 May 19;11(5): e0155858. doi: 10.1371/journal.pone.0155858. 

Published on : Wed, 22 Jun 2016

Related Articles
New Way Old Way written on blackboard credit Pixabay

Relocating an ICU can involve not only a physical move but changes to work practices. Relocation should be planned with a systematic... Read more

Drop of blood

Guidance on how to implement patient blood management (PBM) in a hospital is provided in an article just published in Transfusion... Read more

two hands, credit pixabay

While the evidence for the benefits of early mobilisation in the intensive care unit (ICU) is clear, practical guidance on tools... Read more

Mortality, ICU, prediction, models, Rwanda A model to predict ICU mortality has been designed for use in low-income countries. The development team say that its use of just a few variables that are easily collected make it suitable for use in sites that care for critically ill patients in low-inco

No comment

Please login to leave a comment...

Highlighted Products