IHI.org - A resource from the Institute for Healthcare Improvement
Header Image






Focus on Variation:
Stop Tampering

Tampering is defined as "interfering so as to weaken or change for the worse." In many situations, changes are made based on the last result observed or measured. Often these changes actually increase the variation in a process or product. The methods of statistical process control can be used to decide when it is appropriate to make changes based on recent results. Adjustments to a stable process based on the previous result will usually make performance worse. However, when special causes are present, adjustments can be useful.




Examples of Tests of this Change

A clinic team that is trying to match its capacity (availability of providers) with demand (number of patients requesting appointments on any given day, week, etc.) measures its daily (or weekly) demand and makes predictions about staff levels based on those predictions. Once the staff levels are set, adjustments sometimes need to be made if there are unexpected surges in demand. For example, staff may need to work through lunch or stay later to accommodate an unusual and temporary increase in demand. However, there is no need to change the regular staffing assignments unless there is a clear trend over many days and/or weeks that would indicate an adjustment is needed to meet a demand that has reached a new level (outside of the upper or lower control limits on a control chart).


What others are saying
Post your comments about this item.
View All Comments
Post Your Comments