IHP 435 Module 2 Measurement Short Paper Example

Measuring Emergency Department Waits: Building a Complete Family of Measures

[Student Name]

Southern New Hampshire University

IHP 435: Performance Improvement Measurement and Methodologies

Module Two Short Paper

[Instructor Name]

[Date]

Measuring Emergency Department Waits: Building a Complete Family of Measures

Cedar Point Medical Center's emergency department sees about 48,000 visits a year, and leaders have set a goal of shorter waits. The only measure on the monthly dashboard, however, is average length of stay, which blends patients discharged after a quick visit with those waiting hours for an inpatient bed. When the number goes up or down, no one knows why. This paper builds a family of measures for the waiting problem, explains how each is classified and defined and shows how the measures should be displayed so leaders can tell real improvement from normal fluctuation.

Why a Family of Measures

No single measure can tell a team whether an improvement worked. An outcome measure shows whether patients are better off but changes slowly and is influenced by many factors. A process measure shows whether the team is doing what it planned, which helps explain outcomes. A structure measure shows whether the resources needed are in place. And a balancing measure warns if improving one part of the system has made another part worse. Together, they tell a coherent story; alone, each can mislead.

Donabedian's Classification

Donabedian (1988) offered a three-part lens that is still the backbone of health care measurement. The first part looks at the setting: the people on duty, the rooms and equipment, the way the department is organized. The second looks at the work itself, the steps clinicians and staff take with each patient. The third looks at what happens to patients as a result, including their health and how they experience care. In his view the three parts are connected by likelihood rather than certainty: a better-equipped setting raises the odds that the work is done well, and better work raises the odds of better results. For the emergency department, adding a triage nurse changes the setting, but it will shorten waits only if it changes how patients actually move through the first hour.

What Makes a Good Indicator

Mainz (2003) described how clinical indicators should be defined and classified. Good indicators have a precise definition, including what is counted in the numerator and denominator, clear inclusion and exclusion rules and a specified data source and time period. They should be based on evidence or professional consensus that they reflect quality, be valid, meaning they measure what they claim, and be reliable, meaning they give consistent results when measured again. He also distinguished rate-based indicators, which track how often something happens, from sentinel indicators, which flag individual serious events for review.

Applying these criteria forces the team to decide exactly when the wait starts and ends, which patients count and where the data come from, decisions the current dashboard never made.

Outcome Measures

Two outcome measures matter most to patients. The first is the percentage of patients who leave without being seen by a provider, defined as patients who registered and left before a physician, physician assistant or nurse practitioner documented an evaluation, divided by all registered patients each week. Patients who leave may return sicker or not return at all. The second is patient-reported wait experience, measured by the survey item asking how long patients waited before seeing a provider, reported monthly as the percentage answering in the top category.

Process Measures

Two process measures explain the outcomes. Door-to-provider time is the median minutes from arrival time recorded at the kiosk to the first provider evaluation documented in the record, reported weekly for all patients. Arrival-to-triage time is the median minutes from arrival to completion of the triage assessment. These are the steps the improvement will change, so they should move first if the changes work.

Structure Measures

Structure measures confirm that the resources the improvement depends on exist. The first is the number of provider hours scheduled between 11 a.m. and 11 p.m., the busiest period, per week. The second is the percentage of hours a dedicated triage nurse is staffed at the front door. If process measures do not improve, structure measures help show whether the plan was carried out.

Both come from records the department already keeps, so collecting them adds almost no work.

Balancing Measures

Balancing measures ask what might get worse. Moving patients faster to a provider could lead to hurried evaluations, so the 72-hour return rate, patients who come back within three days, is tracked as a check on quality. Faster front-end processes may shift pressure elsewhere, so the median boarding time for admitted patients and staff overtime hours are tracked as checks on workload and flow. These measures protect patients and staff from well-intended changes with hidden costs.

The Measure Family at a Glance

Table 1 brings the measures together. Each has a type, a definition, a source and a reporting frequency, so anyone reading the dashboard knows what the number means and where it comes from.

Table 1. Family of Measures for Emergency Department Waits

Note. Definitions follow the criteria for clear indicators.

Displaying Measures Over Time

Perla et al. (2011) explain that a run chart, a time-ordered line of values drawn against their middle value, lets teams see whether changes coincide with real shifts in performance, using simple probability rules to separate signals from random variation. Comparing one month before a change with one month after cannot do this, because a single good or bad month may be chance. Each measure in the family should therefore be plotted weekly or monthly with changes marked on the chart, starting with at least twelve baseline points where data allow.

Conclusion

The current dashboard's single average hides more than it shows. A family of measures classified by Donabedian's model, defined to Mainz's standards and displayed on run charts gives Cedar Point a way to know whether changes to its emergency department actually shorten waits without causing new problems.

References

Donabedian, A. (1988). The quality of care: How can it be assessed? JAMA, 260(12), 1743-1748. https://doi.org/10.1001/jama.1988.03410120089033

Mainz, J. (2003). Defining and classifying clinical indicators for quality improvement. International Journal for Quality in Health Care, 15(6), 523-530. https://doi.org/10.1093/intqhc/mzg081

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