Funnel Plot Control Limits
Wider control limits reduce false-positive outlier labelling but also reduce sensitivity to persistent differences. Failure to reject a true SMR of 1.00 does not prove that an ICU performs equally to other ICUs. Annual funnel plots classif…
1 sources - 4 claims
Wider control limits reduce false-positive outlier labelling but also reduce sensitivity to persistent differences. Failure to reject a true SMR of 1.00 does not prove that an ICU performs equally to other ICUs. Annual funnel plots classify ICUs inside control limits as in control and those outside as potential high or low outliers. Conservative 99.8% limits with overdispersion adjustment may leave ICUs with repeated moderate excess mortality unflagged.