HA425 Unit 9 Control Chart Seminar Reflection Example

One Point Above the Line and a Holiday Nobody Mentioned: A Seminar Reflection on Reading a Control Chart for a Composite Hospital's Discharge-Before-Noon Push, From Declaring Success to Seeing a Special Cause With a Different Name

[Student Name]

Purdue University Global

HA425: Operational Analysis and Quality Improvement

Unit 9 Seminar Reflection

[Instructor Name]

[Date]

The hospital, the chart and the weekly figures are composite and bracketed; the reflection shows the form of a seminar response and does not describe a real initiative.

The Chart

The instructor posted a weekly control chart for medical-surgical units at a composite 180-bed hospital. Each point was the percentage of discharges completed before noon. For [20] weeks before the initiative, the points centered near [14] percent, with control limits at about [7] and [21]. The initiative launched with a kickoff meeting and a daily huddle. The next week's point was [31] percent, well above the upper limit. The four weeks after that were [15], [17], [14] and [16] percent.

My First Reading

Before the session I wrote: "The initiative worked. A point above the upper control limit is a signal, and this one is far above it." I reasoned that a jump that large could not be chance, and the chart's own rules said so. I was right that the point was a signal and wrong about what it was a signal of.

What the Session Applied

The instructor asked us to apply the chart's rules fully, not just the first one. A single point outside the limits signals a special cause, something outside the usual system. But a sustained improvement from a change in the system would show as a shift: a run of points on one side of the center line. Standard guidance reads a long unbroken run on one side of the center, often set at eight weeks in a row, as evidence that the underlying process has moved (Provost & Murray, 2011). After the spike, the next four points sat right back at the old center. Then a classmate asked what else happened that week. The launch week was the week of a national holiday. Elective surgeries were canceled, census was low, and families were available to pick patients up in the morning. That is a special cause, but not the initiative.

Why My First Reading Was Tempting

I wanted the initiative to work, and the chart gave me a dramatic point. Statistical process control is meant to separate signals from noise, but it cannot name the cause of a signal; people do that, and people tend to name the cause they hoped for (Benneyan et al., 2003). The leaders who declared success were doing what I did.

What a Better Analysis Would Do

1. Annotate the chart with events, such as holidays and census changes, before interpreting it.

2. Wait for enough points to see whether a shift has occurred.

3. Look at a run chart of the same data for runs and trends, which can show a change sooner than waiting for points outside the limits (Perla et al., 2011).

4. Check a balancing measure: if discharges before noon rose, did readmissions or patient complaints rise too?

What the Initiative Might Still Achieve

The chart does not show the initiative failed. Four weeks is too few to judge a shift. If the daily huddle is working, the next several points might sit above [14] percent. But the evidence used to declare success says nothing about the initiative.

Connecting It to My Own Project

Our emergency department project will face the same temptation. The first week after the evening shift change might show a dramatic drop in waits, or a dramatic rise if a busy flu week arrives. I now plan to mark the start of each test cycle and any unusual events, such as a nearby hospital closing its emergency department for a night or a regional storm, directly on the run chart, and to wait for a run of points before telling anyone the change worked. I will also report the morning balancing measure on the same page, so a gain in the evening is never shown without the cost.

The Group's Revised Reading

By the end of the session the group agreed on a short statement for the hospital's leaders: the launch week was a holiday special cause, the following four weeks show no change yet, and the initiative should be judged after at least eight more weeks of data, with holidays marked.

What I Will Do Differently

In my own department I will annotate every run chart with events, use the shift and run rules before declaring a change and ask "what else happened that week?" before naming a cause.

What Still Bothers Me

Leaders often need to show success quickly to keep support for a change. Waiting eight weeks may feel too slow. I am not sure how to balance honest reading of data with the need to keep momentum, except perhaps by reporting progress on process measures, like huddle attendance, while the outcome data accumulate.

References

Benneyan, J. C., Lloyd, R. C., & Plsek, P. E. (2003). Statistical process control as a tool for research and healthcare improvement. Quality and Safety in Health Care, 12(6), 458-464. https://doi.org/10.1136/qhc.12.6.458

Perla, R. J., Provost, L. P., & Murray, S. K. (2011). The run chart: A simple analytical tool for learning from variation in healthcare processes. BMJ Quality & Safety, 20(1), 46-51. https://doi.org/10.1136/bmjqs.2009.037895

Provost, L. P., & Murray, S. K. (2011). The health care data guide: Learning from data for improvement. Jossey-Bass.