Show pageBacklinksCite current pageExport to PDFBack to top This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong. ===== Threshold Value ===== The **threshold value** in CUSUM analysis defines the alert level at which the cumulative sum of deviations signals a statistically significant change in performance. It serves as a boundary for action: * When the CUSUM curve **crosses the threshold**, it may indicate deteriorating performance (e.g., rising complication rate). * Staying **below the threshold** suggests performance remains within acceptable limits. ==== How It’s Determined ==== Threshold values can be: * **Statistical**, based on type I/II error probabilities and decision intervals (commonly used in industrial applications). * **Empirical**, based on institutional experience or expert consensus. * **Graphical**, set arbitrarily for visual feedback (e.g., 2.5 in our ICP example). In our analysis of ICP monitor placement, we used a **threshold of 2.5** to flag deviations from the acceptable complication rate of 10%. This value is not fixed and can be adjusted depending on: * Procedure complexity * Acceptable risk tolerance * Training level of the surgeon ==== Clinical Application ==== Crossing the threshold may prompt: * Case review or morbidity and mortality (M&M) analysis * Temporary increase in supervision or training * Modification of technique, workflow, or equipment Using CUSUM with a clearly defined threshold value transforms raw outcome data into a **real-time decision-support tool** for improving patient safety and clinical performance. ---- Institutions may modify the threshold value based on: * [[Historical complication rate]]s * [[Supervision level]] * [[Case complexity]] * [[Acceptable risk policy]] threshold_value.txt Last modified: 2025/04/08 18:24by 127.0.0.1