NU803 Unit 6 Baseline Data Summary Example

Sixteen Weeks Before Anything Changed: A Baseline Data Summary of Steroid-Treated Patients on 4 North and 4 South, Produced by Script, With Its Gaps Named

[Student Name]

Purdue University Global

NU803: Change Management and Clinical Data Analysis

Unit 6 Baseline Data Summary

[Instructor Name]

[Date]

The units and all figures are composite; numbers are bracketed and internally consistent.

How the Figures Were Produced

All figures were produced by a single script, run on the glucometer export, the pharmacy's steroid order file and the bed census for the [16] weeks before education began. The script identifies eligible steroid days from orders at [20] mg or more in prednisone terms, excludes days meeting the exclusion rules, links each day to glucose readings by patient and date, applies the confirmatory-repeat rule, and counts each measure by week. Running the same script on implementation data will guarantee that the two periods are counted by identical rules.

Volume

Across both units there were [1,440] eligible steroid days, about [90] a week, belonging to [318] patients. The median stay on steroids was [four] days. About [38] percent of patients had a recorded diagnosis of diabetes.

Monitoring

Only [612] of the [1,440] eligible days, [42.5] percent, met the definition of a monitored day, two or more readings with one taken after [1100]. A reading between [1030] and [1200] was present on [23] percent of eligible days, and a reading between [1500] and [1700] on [29] percent. Among patients without diabetes, [46] percent went without any glucose reading during their first [48] hours on the steroid.

Outcome

Of the [612] monitored days, [190] included a reading above 250 mg/dL, a rate of [31.0] per 100 monitored days. Weekly rates ranged from [27.5] to [34.2], with a median of [30.8]. The weekly series showed no shift, trend or unusual number of runs, which suggests a stable baseline suitable for comparison.

A stable baseline is the quiet finding that makes every later finding possible: if the rate had been drifting, no change after go-live could be credited to the bundle.

Response

There were [58] episodes that would have met the escalation definition, a pair of readings over 216 mg/dL inside one day. A scheduled insulin order followed within [24] hours in [20] of them, [34] percent. Nursing notes documented a call to a prescriber in [23], though paging logs could not confirm the time of the call in most cases.

Balancing

There were [16] steroid days, counting the two days that follow a final dose, with a reading below 70 mg/dL, a rate of [1.1] per 100. Weekly counts were too small to chart meaningfully, so the balancing measure will be tracked as a running total with a threshold for review, as the evaluation plan specifies.

Differences Between the Two Units

The units were combined for the main figures, but their baselines differed in ways worth recording. 4 North, which admits more patients with chronic lung disease, had [58] percent of the steroid days, and its monitored-day rate was lower, [38] percent against [48] percent on 4 South. Its hyperglycemic rate among monitored days was slightly higher, [32.6] against [29.4] per 100, though weekly counts on each unit were small enough that the difference may reflect chance. These differences matter for implementation: if the bundle works on one unit before the other, the baseline gives a reason to expect that 4 North, with more steroid days and less monitoring, will show the larger change in check completion.

Gaps in the Baseline

Three gaps affect how the baseline can be used. The largest is that nobody on either unit ever recorded when trays arrived. The bundle's instruction is to check within [30] minutes before the tray, but at baseline only a clock window, [1030] to [1200], could be measured. Implementation data will use the same clock window for comparison, and the unit secretaries will record tray arrival times for [two] weeks during implementation to test whether the clock window is a fair proxy.

The second gap is paging data. The paging system keeps logs for only [30] days, so baseline escalation times could not be verified, and the baseline response measure uses scheduled insulin orders instead, which are permanent.

The third gap is that the oncology and emergency admission order sets were not included in the baseline order file because of a different field structure. Their patients, about [18] percent of qualifying doses, are therefore missing from the baseline. The implementation analysis will report results with and without them.

What the Baseline Supports

The baseline supports a comparison of the primary outcome, monitored hyperglycemic days, and of check completion by clock window. It supports a weaker comparison of response, through scheduled insulin orders. It does not support any statement about tray-timed checks. Reporting these limits in advance follows guidance that improvement reports describe how data were collected and what their weaknesses were (Ogrinc et al., 2016), and the decision to display the baseline as a weekly series rather than as a single average follows run chart practice (Perla et al., 2011; Provost & Murray, 2022).

References

Ogrinc, G., Davies, L., Goodman, D., Batalden, P., Davidoff, F., & Stevens, D. (2016). SQUIRE 2.0 (Standards for QUality Improvement Reporting Excellence): Revised publication guidelines from a detailed consensus process. BMJ Quality & Safety, 25(12), 986-992. https://doi.org/10.1136/bmjqs-2015-004411

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. (2022). The health care data guide: Learning from data for improvement (2nd ed.). Jossey-Bass.