College Answer

Standardised Mortality Ratio is defined as the observed mortality rate/expected mortality rate. Need to estimate expected mortality rate using a scoring system (eg. APACHE II or III, SAPS II or MPM). Better than comparison of non-adjusted mortality data.

The potential limitations of the system are multiple including: inconsistencies and inaccuracies associated with collection of data and scoring (eg. GCS, recording of parameters); problems of missing data limiting inclusion of all patients; problems of patient mix not adequately accounted for by the original population used for calculation of formulae (eg. transferred patients or delays before admission); small numbers of patients (increasing the error of the SMR estimate); accuracy of the prediction model; relying on mortality as a surrogate marker for quality of care; cost of use of proprietary system; etc.

Discussion

This question closely resembles Question 30 from the second paper of 2006.

Definition of the SMR

• This is the ratio of the observed hospital mortality vs. predicted hospital mortality for a specified time period.
• One can use this to compare hospitals and ICUs
• One needs to first calculate the predicted hospital mortality using an illness severity scoring system.
• An SMR of 1 means the mortality is as expected.
• An SMR of < 1 is better than expected, and >1 is worse than expected.

Limitations of the SMR

• Acceptable deviations from the SMR are not defined
• Suffers from inaccuracies associated with data collection
• SMR may be influenced by ICU admission and discharge practices (eg. discharging patients who are palliated, or admitting patients who are inevitably going to die).
• Accuracy of the SMR as a quality assessment tool may be influenced by patients who have been predominantly cared for at another ICU, and who have been received as a transfer.
• Mortality is not a surrogate for quality of care
• The populations used to calculate the predicted hospital mortality are potentially non-representative (i.e. the population may also contains a number of dying critically ill patients, or it may contain an unusually large proportion of people in robust health).

Limitations of comparing ICUs with the SMR:

• The SMR assumes all pre-ICU care is identical
• Ignores differences in case mix
• Sample sizes need to be large enough to obey the laws of logistic regression
• Data is assumed to be flawless and complete