Dear Waleed,

you might want to start with this paper from 2004:

Wunsch, Hannah et al. Impact of exclusion criteria on case mix, outcome, and length of stay for the severity of disease scoring methods in common use in critical care. Journal of Critical Care , Volume 19 , Issue 2 , 67 - 74

http://www.ncbi.nlm.nih.gov/pubmed/15236138

It looks at the original exclusion criteria from the original valididation papers for APACHE II, APACHE III, SAPS II and MPM II and their impact on disease severity, mortality, case mix of individual ICUs. You will find more relevant links in this article.

Best wishes,

Immo

Thank you very much, but maybe I did not make my question clear. What I was asking about was:

When I calculate crude mortality rate in ICU the formula is:

total deaths / total discharges in the same period, but some exclude ( as far as I know ) DNAR cases from the numerator, are there any other exclusions.

Mortality rate is deaths/admissions. There are no exclusions.

Mortality rate is calculated using the formula as follows;

Deaths / Admissions.

However, in our study (vide attachment), we excluded some patients such as patients who were hospitalized in the ICU for less than 24 hours or were younger than 18 years.

Dear Waleed,

I suggest you to consider the following points.

Mortality is computed on patients, not on admissions or discharges because a patient can be admitted to or discharged from ICU more than once during the same hospital stay but he can die only once. Therefore, if you rely on discharges to compute ICU mortality, you should exclude from the discharges those of the patients ICU readmitted during the same hospital course. The easiest way to count the denominator could be to consider only the last ICU admission for the patients readmitted, because some patients may be admitted to ICU more than twice.The formula of the crude ICU mortality rate will be:

Number of patients died / Number of patients discharged = Number of deaths / Number of discharges (excluding all discharges before the last for the patients readmitted to ICU during the same hospital course)

You propose to exclude DNAR cases from the numerator, but I am not sure that they should be excluded. The decision do not attempt resuscitation (DNAR) is taken when  cardiac or respiratory arrest is an expected part of the dying process and will help to ensure that the patient dies in a dignified and peaceful manner (according to good medical practice at http://www.gmcuk.org/guidance/ethical_guidance/end_of_life_DNACPR_decision.asp). This DNAR decision is usually taken when the patient does not improve despite the intensive treatment that has been done before the decision. Therefore, I think that these patients have to be included because they received intensive treatment anyway.

Possibly, the most important point is the reason why you want to measure ICU mortality. The aim of hospital admission is to return the patients to their families and to the society. So it is relevant not to survive ICU, but to survive hospital and to go back home (possibly in good condition, but this is another story). Accordingly, severity scores have been created to predict hospital mortality. If you want to analyze the performance of your ICU, you have to rely on standardized hospital mortality rate (Rhodes A et al Prospectively defined indicators to improve the safety and quality of care for critically ill patients: a report from the Task Force on Safety and Quality of the European Society of Intensive Care Medicine attached file).

Of course, other exclusions can be based on age and specific illnesses (i.e. burns) if you use severity scores because you need to have a population similar to that where the severity scores were validated. However, exclusions have to be performed according to the aim of your study.

I hope my comments may help you.

dear Waleed, dear All, maybe it would help in answering the question, if we would know what research or administrative answer you would like to get from calculating ICU mortality ratios. And it would also help to know what kind of ICU we are talking about. Crude mortality ratio's usually do not have a lot to contribute, without severity scoring systems. 

kind regards to all

Der Waleed,

Crude mortality rate is calculated,

Deaths / Admissions for PICU

Dear Ayes, yes I'm aware of that and that is what I'm doing, but I'm questioning its reliability as an indicator of the service.

Dear Waleed,

If you are looking for a measure to evaluate the performance of your ICU crude mortality rate will not help you. Instead you should calculate standardized mortality ratio (SMR), that is observed mortality/predicted mortality. To calculate predicted mortality you have to calculate physiologic status score (eg. APACHE II) of your patients. See the link http://ccforum.com/content/9/6/R645.

Dear Yalim Dikmen, thank you very much, I agree that this is a better measure, and I suggested that to my board, but till now the official measure is crude mortality.

Thank you

Dear Waleed

Look at the following link. I found it helpful

http://healthpolicy.ucsf.edu/content/icu-outcomes#HOSPITAL%20AND%20PATIENT%20SAMPLE

Mortality should be total deaths/total admission there should be no exclusion . It will be purest icu mortality. Of course one has to decide for what purpose rhe ICU mortality is being considered.