It is often said that perception is reality. However, perception may not always be the truth. Making decisions about how to spend time, money, and resources based on perception, rather than truth, can have a negative impact on your organization.
Over the past year, staff at the Oregon State Hospital (OSH) have been doing their best to respond to the COVID-19 crisis. This pandemic has forced us to quickly adapt to the incredible challenges of keeping our patients and staff safe while continuing to provide excellent care for our patients. Some of the safety measures have been difficult for our patients. In-person visitations from loved ones, and activities outside the hospital have been canceled. Treatment Mall offerings for our patients have been reduced to accommodate social distancing. Patients have been transferred to other units to allow space for OSH to establish admission monitoring units and quarantine/isolation units. During all this uncertainty, change, and fear, OSH’s Emergency Operation Center (EOC) perceived that our patient aggression levels were increasing. The EOC commissioned a task force to reduce the increased aggression caused by COVID-19.
Our team, which included two Lean Leaders, a Data Analyst, and a Program Director, began its work by asking the question “Has patient aggression increased as the result of the COVID-19 crisis?” To answer this question, we turned to data. The brilliant thing about data is that it doesn’t lie, have bias, or an agenda. It simply tells the truth. Collecting baseline data in the discovery phase of a project is an essential element in the success of any improvement project. Proving there is a problem by using data allows you to understand the severity of the problem, helps you determine the root cause of the problem, and if the improvement implemented fixed the problem.
The task force analyzed two variables, patient to patient aggression, and patient to staff aggression, before and after COVID-19, to determine if patient aggression had increased. We utilized two tools to analyze the data.
(1) Control charts, which are a statistical tool used in quality control to analyze and understand process variables. In our case, we looked to see if control limits during COVID-19 months increased by three standard deviations.
(2) A two-sample t-test, which is often used for evaluating the means of two variables or distinct groups. The null hypothesis for the t-test is there is no difference between the two samples. The alternative hypothesis is there is a difference. You run the two-sample t-test and look at the p value. If the p value is less than 0.05, you reject the null hypothesis and accept the alternative.
In the example below, the p value is higher than 0.05, so you cannot reject the null hypothesis; there is not enough evidence in this sample to suggest that the aggression is higher in the after-data relative to the before-data.
The two-sample t-test is a simple tool you can run in Excel 2016. Click on Data Analysis in the Analysis group on the Data tab. If the Data Analysis command is not available, you need to load the free Analysis Toolpak add-in program.
After comparing the data on 30 different units, the team discovered that there was no statistical increase in aggression on any of the units as the result of the COVID-19 crisis. Some units experienced some increase, however they were within the control limits of the Control Charts and above .05 on the p value of the Two Sample t-test. The team used data, rather than perception, to understand the truth. The leadership at OSH was then able to objectively look at the data and decide not to spend time, money, and resources on a false perception. Perception may be reality, but data tells the truth.
Steve Unwin, Lean Six Sigma Black Belt
Office of Performance Improvement
Oregon State Hospital