IHP 435 Module 6 Data Display Short Paper Example

Stop Reacting to Last Month: Displaying Falls Data with Run and Control Charts

[Student Name]

Southern New Hampshire University

IHP 435: Performance Improvement Measurement and Methodologies

Module Six Short Paper

[Instructor Name]

[Date]

Stop Reacting to Last Month: Displaying Falls Data with Run and Control Charts

On Cedar Point's fourth-floor medical unit, the monthly falls report is a single slide: last month's fall rate beside this month's, with an arrow showing direction. In March the rate fell from 4.1 to 2.6, measured per thousand patient days, and the manager thanked staff for their vigilance. In April it rose to 4.4, and staff were told to recommit to hourly rounding. Nothing about the unit's fall prevention practices had changed. This paper explains why two-point comparisons mislead, how run charts and control charts display data correctly and how leaders should respond to what the charts show.

Why Two Points Mislead

Every process varies. A unit with a stable fall prevention program will still have more falls in some months than others, because patient mix, census and chance all vary. Comparing two months treats that ordinary variation as meaningful, which leads leaders to reward or punish staff for random fluctuations. Over time, staff learn that the numbers are arbitrary and stop trusting them. The solution is to display many points over time, so that ordinary variation can be seen for what it is and real change stands out.

Run Charts

Perla et al. (2011) describe the run chart as a line graph of a measure over time with the median drawn across it, simple enough for any team to build. They explain probability-based rules that help decide when a pattern is unlikely to be chance: a shift, when at least six readings in a row land above, or all below, the middle value; a trend, when five or more climb or drop without a break; too few or too many runs, where a run means an unbroken stretch of readings above or below the middle; and an astronomical point, one obviously different from the rest. Points exactly on the median are not counted toward shifts. These rules work without complicated statistics and are suitable for small numbers of data points.

Control Charts

Benneyan (2003) explains that control charts, the core tool of statistical process control, add upper and lower control limits to a time series, typically about three standard deviations from the center line and calculated from the process's own data. Points outside the limits, or certain patterns within them, indicate special-cause variation, a signal that something has changed and deserves investigation. Points within the limits without such patterns reflect common-cause variation, the normal noise of a stable process, which should not be chased point by point but can be reduced only by changing the system itself. He emphasizes that control charts are useful both for monitoring and for judging whether an improvement produced a real change.

Choosing the Right Chart

Mohammed et al. (2008) offer practical guidance on selecting and plotting basic control charts. The choice depends on the type of data. For counts of events where the opportunity varies, such as falls per patient day, a u chart is appropriate, because it adjusts limits for the changing number of patient days each month. For proportions, such as how often nurses finish the admission fall screen, a p chart fits. For continuous measurements, such as minutes, an individuals chart is often used. They also stress practical points: use enough data to set limits, usually twenty or more points, plot in time order and avoid recalculating limits every time a new point arrives.

Replotting the Falls Data

The quality analyst plotted the unit's falls for the past twenty-four months on a u chart. The center line sat at 3.6, and the monthly limits ranged roughly from 0.8 to 6.4, varying slightly with patient days. Every point, including March's 2.6 and April's 4.4, fell inside the limits, and no run chart rule was triggered. The chart showed a stable process: the unit generates roughly 3.6 falls for each thousand patient days, month after month, and the swings that prompted praise and criticism were noise.

Table 1. Matching Data to Charts

Note. Chart choice follows the type of data.

What the Chart Means for Leaders

A stable process at that rate will keep producing that rate until the system changes. Exhorting staff after a bad month will not help, and praising them after a good month rewards luck. If leaders want fewer falls, they need to change the process, for example by testing a new approach for high-risk patients, and then watch the chart for a signal such as a shift below the center line. The chart also protects staff: a month above average is not evidence of failure unless it breaks the limits or triggers a rule.

Teaching Managers to Read Charts

Charts help only if managers read them correctly. The quality department should offer a one-hour session for unit managers on reading run charts and control charts, using their own units' data rather than textbook examples. Managers would practice deciding whether a new point is a signal, explaining the chart to staff in plain words and identifying when a change to the system, not a pep talk, is needed. A one-page guide posted beside each unit's chart can remind readers of the rules. Over time, the conversation at staff meetings should shift from why this month was worse to what the unit is testing to change its baseline.

A Response Policy

The unit should adopt a simple policy. Monthly falls data are displayed on the u chart with at least twenty-four months of history. A point outside the limits or a run chart rule triggers a review within two weeks to look for a special cause. Points within limits without a signal are not discussed as successes or failures. When a deliberate change is made, it is marked on the chart, and the team watches for a shift. Limits are recalculated only after a sustained shift confirms that the process has changed.

Conclusion

Two-point comparisons turned random variation into praise and blame on the fourth floor. Run charts and control charts show that the unit's falls process is stable, which means improvement will come from changing the system, not from reacting to last month. Displaying data correctly is itself a management improvement.

References

Benneyan, J. C. (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

Mohammed, M. A., Worthington, P., & Woodall, W. H. (2008). Plotting basic control charts: Tutorial notes for healthcare practitioners. Quality and Safety in Health Care, 17(2), 137-145. https://doi.org/10.1136/qshc.2004.012047

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