DNP/730 Week 4: Improvement Science and the Model for Improvement, sample paper

Three Questions and Four Small Tests: Applying the Model for Improvement to Morning Discharges, and Avoiding the Ways PDSA Cycles Usually Go Wrong

[Student Name]

University of Phoenix

DNP/730: Organizational and Systems Leadership

Week 4 Assignment

[Instructor Name]

[Date]

The units, tests and data are composites written for a model paper.

Our two medical units discharge only 9% of patients before noon. Having assessed the system, culture and leadership approach, I turned to method. This paper applies the Model for Improvement on 5 East.

The Science of Improvement

Berwick (2008) described the science of improvement as distinct from traditional clinical research: improvement work is embedded in complex social systems, relies on learning from rapid cycles in context and requires methods suited to asking not only whether a change works but how and in what circumstances. He argued that evaluation of improvement should use a wider range of methods than randomized trials alone.

Question One: Aim

What are we trying to accomplish? Increase discharges before noon on 5 East from 9% to 30% within six months, without increasing 30-day readmissions. The aim is specific, measurable and time-bound. The target is informed by a published hospital effort in which the discharge-before-noon rate rose from 11% to 38% over 13 months (Wertheimer et al., 2014).

Question Two: Measures

For the second question, about knowing whether a change helps, the outcome measure is weekly percentage of discharges before noon. Process measures include the percentage of next-day discharges identified by 3 p.m. and the percentage of discharge prescriptions completed the evening before. Balancing measures are 30-day readmissions and patient-reported readiness for discharge.

Question Three: Changes

What changes can we make? From the system map, staff proposed: afternoon identification of likely next-day discharges, morning rounds starting with those patients, evening completion of prescriptions and discharge teaching begun the day before.

How PDSA Goes Wrong

Taylor et al. (2014) systematically reviewed 73 articles reporting PDSA cycles and found that many did not follow the method's key features. Fewer than 20% documented a sequence of iterative cycles, few started with small-scale changes and very few let data collected at least monthly steer the next cycle. Reed and Card (2016) argue that PDSA is often treated as a simple tool when it requires skill, resources and attention to prediction, learning and scaling, and that projects often jump to large changes without adequate testing.

The first test involved one hospitalist, one day and three patients; it was supposed to be small enough to fail cheaply.

Choosing the First Change

Of the four change ideas, afternoon identification of next-day discharges was tested first because every other change depends on knowing who will leave. Testing the foundation first avoids optimizing downstream steps for patients who are not actually leaving.

Cycle One

Plan: one hospitalist will identify likely next-day discharges at 3 p.m. with the charge nurse. Prediction: at least two patients will be identified and discharged before noon the next day. Do: on Tuesday, the hospitalist and charge nurse identified three patients. Study: two were discharged before noon; one waited for a physical therapy evaluation. Act: adapt by adding physical therapy to the afternoon conversation.

Cycle Two

Plan: the same hospitalist, charge nurse and a physical therapist meet for five minutes at 3 p.m. for three days. Prediction: therapy evaluations for identified patients will be completed by 10 a.m. Do: completed as planned. Study: seven of nine identified patients were discharged before noon; two waited for prescriptions. Act: adapt by adding pharmacy.

Cycle Three

Plan: pharmacy will fill prescriptions for identified patients the evening before, for one week. Do and study: all prescriptions were ready by 8 a.m.; morning discharge rate for identified patients reached 81%. Act: adopt for this hospitalist; plan to test with a second hospitalist.

Cycle Four

Plan: expand to two hospitalists for two weeks. Study: the second hospitalist rounded on new admissions first, so identified patients were seen late. Act: adapt by agreeing that identified discharges are seen first unless a patient is unstable.

Staff Feedback During Cycles

After each cycle, I asked the charge nurse, hospitalist and therapist what was easy, what was hard and what they would change. Their comments shaped the next test as much as the numbers did. One suggestion, a whiteboard column labeled tomorrow, became part of cycle three.

Data Over Time

Weekly data on the unit's discharge-before-noon percentage rose from a baseline median of 9% to 21% by week eight, shown on a run chart with six consecutive points above the baseline median. Readmissions did not change, and patient-reported readiness scores held steady, which suggests the earlier timing did not rush patients out before they were prepared.

A Failed Test Worth Keeping

In a side test, we tried having nurses complete discharge teaching at 7 a.m. before rounds. It failed: patients were groggy and families absent. We abandoned it and moved teaching to the prior evening instead. Recording failed tests prevents others from repeating them.

Avoiding the Pitfalls

Heeding the systematic review (Taylor et al., 2014), each cycle had a prediction, was small, was documented and used weekly data. Heeding the critique of PDSA practice (Reed & Card, 2016), we planned for learning, allocated time for the team and did not scale until tests worked reliably.

What We Learned Beyond the Numbers

The cycles revealed dependencies no one had mapped: physical therapy timing, pharmacy workflow and hospitalist rounding order. Each was discovered because a small test failed in a specific way. Large rollouts often fail without revealing why; small tests fail informatively.

Team and Time

The core team met for 20 minutes twice a week, and I protected the charge nurse's time for the 3 p.m. meeting by adjusting her assignment. Reed and Card (2016) note that PDSA requires resources; without protected time, the tests would have been squeezed out by daily work.

Documenting the Cycles

Each cycle was recorded on a one-page form with the plan, prediction, what happened, what was learned and the decision to adopt, adapt or abandon. The forms became the project's history and helped new team members understand why the process looks the way it does.

Spreading to 5 West

On 5 West, the tested changes will be introduced as starting points, not mandates, with the unit's own PDSA cycles to adapt them. Staff there will run their first test with one hospitalist and their own charge nurse, preserving the ownership that made the 5 East tests work.

Scaling

Scaling to all hospitalists and to 5 West requires more than repeating the test. Each group's context differs, and further cycles will adapt the changes.

Predictions Matter

Writing a prediction before each test forced the team to state what it believed would happen. When predictions failed, as when the second hospitalist rounded differently, the gap between prediction and result pointed straight to what needed to change.

Conclusion

Applying the Model for Improvement to morning discharges produced a clear aim, measures and changes tested through four small, documented PDSA cycles, raising discharges before noon from 9% to 21% so far without increasing readmissions. Evidence that PDSA is often misapplied shaped the cycles to be small, predictive, iterative and data driven.

References

Berwick, D. M. (2008). The science of improvement. JAMA, 299(10), 1182-1184. https://doi.org/10.1001/jama.299.10.1182

Reed, J. E., & Card, A. J. (2016). The problem with Plan-Do-Study-Act cycles. BMJ Quality and Safety, 25(3), 147-152. https://doi.org/10.1136/bmjqs-2015-005076

Taylor, M. J., McNicholas, C., Nicolay, C., Darzi, A., Bell, D., & Reed, J. E. (2014). Systematic review of the application of the plan-do-study-act method to improve quality in healthcare. BMJ Quality and Safety, 23(4), 290-298. https://doi.org/10.1136/bmjqs-2013-001862

Wertheimer, B., Jacobs, R. E. A., Bailey, M., Holstein, S., Chatfield, S., Ohta, B., Horrocks, A., & Hochman, K. (2014). Discharge before noon: An achievable hospital goal. Journal of Hospital Medicine, 9(4), 210-214. https://doi.org/10.1002/jhm.2154