HIM 550 Module 8 Visualization Short Paper Example

Red Gauges and Honest Charts: Redesigning the Patient Flow Dashboard

[Student Name]

Southern New Hampshire University

HIM 550: Data Management and Data Quality

Visualization Short Paper

[Instructor Name]

[Date]

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

Red Gauges and Honest Charts: Redesigning the Patient Flow Dashboard

Every weekday at 8 a.m., Cypress Hollow Medical Center's leaders open a patient flow dashboard before the bed huddle. Its numbers now rest on better data, since the discharge measure has been rebuilt on nursing departure time. But accurate numbers can still mislead when they are drawn badly. This paper critiques the current dashboard, explains the principles its design violates and proposes a redesign that helps leaders see what is changing and what to do about it.

What Dashboards Can Do

A dashboard is meant to give busy people the few measures they need at a glance. Dowding et al. (2015) reviewed studies of clinical and quality dashboards and found some evidence that dashboards giving clinicians quick access to performance information improved adherence to quality standards, although the studies were limited and varied in design. The review also noted that how information is displayed affects whether users understand it. A dashboard that is misread is worse than none, because it gives leaders confidence in the wrong conclusion.

Four Problems with the Current Dashboard

The first problem is the gauge. A large speedometer shows the before-noon rate in red because it falls below the 40% target. It reports one number with no history, so leaders cannot tell whether the rate is improving, steady or slipping. The color also turns the display into a verdict rather than information.

The second is the unit bar chart. Its vertical axis begins at 25%, so a medical unit at 28% and a surgical unit at 34% appear as bars of 3 and 9 units in height, suggesting that one unit discharges three times as many patients in the morning as the other. Tufte (2001) called the ratio between the size of an effect shown in a graphic and its size in the data the lie factor, and argued that it should be close to one. Here it is far larger.

The third is the month-over-month arrow beside each measure. A green up arrow or red down arrow appears whenever the number changes at all. Most monthly changes are ordinary variation, yet leaders routinely ask units to explain a red arrow that reflects nothing but chance.

The fourth is a three-dimensional pie chart of admission sources. Pie slices are hard to compare, the tilt enlarges the front slices and missing values, now separated from Other in the warehouse, have been folded back into Other by the chart's own grouping.

Who Reads the Dashboard

Design choices depend on the audience, and this dashboard has three. Senior leaders glance at it for less than a minute before the huddle and need to know whether flow is improving and where attention is needed. Unit managers use it to compare their own unit with others and to prepare for questions. The bed placement team uses it during the day to anticipate afternoon crowding. The current design serves none of them well: leaders see a verdict, managers see exaggerated gaps and the bed team sees nothing about timing within the day. A useful redesign keeps the top of the screen for the leaders' question, places unit panels below for managers and adds an hourly departure pattern for the bed team on a second tab rather than crowding the first.

Principles for the Redesign

The redesign follows four principles drawn from the critique. Show change over time rather than single values. Start quantitative axes at zero for bars and use the same scale across comparable panels. Distinguish signals from noise with statistical rules rather than arrows. Show data quality openly, including missing values and whether last night's load completed. Tufte (2001) also urged removing decoration that carries no data, which rules out the three-dimensional effects and gauge bezels.

The Redesigned Dashboard

Table 1 compares each element of the current dashboard with its replacement.

Table 1. Current and Redesigned Dashboard Elements

Note. Prepared by the author from a review of the current patient flow dashboard.

Reading Change Honestly

At the center sits the weekly share of morning departures, plotted over 26 weeks around its median, with the 40% target drawn as its own line. Perla et al. (2011) set out a few simple rules, among them a long unbroken run of weeks on one side of the median, for telling genuine change from chance. When the discharge program begins, leaders will see whether the rate shifts rather than reacting to single weeks. Placing emergency boarding hours directly beneath it on the same time scale reflects the analysis milestone's finding that the real value of morning discharges lies in bed flow. A small note under the chart will state that the rate uses nursing departure time and will show how many departure times were missing that week, so leaders can see at once whether a dip reflects patients or documentation.

Testing the Redesign

Before release, the redesign will be tested with five regular users, including the chief nursing officer and two unit managers. Each will be asked three questions using the old and new versions: is the rate improving, which unit has the most room to improve and is today's data complete. If users answer more accurately and quickly with the new design, it will replace the old one. Their comments will also decide which measures can be removed, since a dashboard that tries to show everything shows nothing clearly.

Conclusion

The old dashboard displayed accurate data in ways that exaggerated differences, hid history and treated noise as news. The redesign shows change over time, honest proportions, visible gaps and the context leaders need, so the morning huddle can focus on what is actually changing.

References

Dowding, D., Randell, R., Gardner, P., Fitzpatrick, G., Dykes, P., Favela, J., Hamer, S., Whitewood-Moores, Z., Hardiker, N., Borycki, E., & Currie, L. (2015). Dashboards for improving patient care: Review of the literature. International Journal of Medical Informatics, 84(2), 87-100. https://doi.org/10.1016/j.ijmedinf.2014.10.001

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

Tufte, E. R. (2001). The visual display of quantitative information (2nd ed.). Graphics Press.