HIM 550 Module 5 Final Project Milestone Two Example

Final Project Milestone Two: Does Leaving Before Noon Shorten Stays? Discharge Timing and Bed Flow in a Louisiana Hospital

[Student Name]

Southern New Hampshire University

HIM 550: Data Management and Data Quality

Final Project Milestone Two

[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.

Final Project Milestone Two: Does Leaving Before Noon Shorten Stays? Discharge Timing and Bed Flow in a Louisiana Hospital

Cypress Hollow Medical Center's leaders intend to raise the share of patients discharged before noon from about a third to 40%, expecting shorter stays and fewer admitted patients held on emergency stretchers until a bed opens. The first milestone found the data usable only after replacing the registration discharge time with the nursing departure time and applying a unit crosswalk. This milestone uses the corrected data to ask whether earlier discharge is associated with shorter length of stay, whether it is associated with shorter emergency boarding and what leaders should expect from the new target.

Analytic Sample and Measures

The analysis began with the 9,840 discharges assessed in Milestone One. Following the written rules from that assessment, 186 deaths and 212 transfers to other acute hospitals were excluded, 131 patients who left against medical advice were set aside for separate review and 671 routine discharges with no nursing departure time were excluded, leaving 8,640 routine discharges.

Discharge before noon was defined as a nursing departure time before 12:00. Length of stay was measured in days from admission time to departure time. Emergency boarding was measured as hours from the admission order to the patient's arrival on an inpatient unit, drawn from the bed management system for 4,915 patients admitted through the emergency department. Every measure and exclusion is recorded in the project's data dictionary so the analysis can be repeated.

Describing the Data

Length of stay is strongly right-skewed: most patients stay a few days, and a small number stay weeks. The mean was 4.9 days, pulled upward by long stays, while the median was 3.7 days and half of all stays fell between 2.1 and 6.0 days. Because of this skew, the analysis reports medians and ranges rather than means and uses comparisons that do not assume a normal distribution. Overall, 3,180 of 8,640 patients, or 36.8%, left before noon. Table 1 shows that the rate differed sharply between medical and surgical units.

Table 1. Discharge Timing and Length of Stay by Unit Type, January to June

Note. Prepared by the author from the corrected discharge data set; deaths, transfers and patients leaving against advice excluded.

Weekly Pattern

A run chart of the weekly before-noon rate across 26 weeks showed a median of 36.5% with no shifts, trends or unusual runs. Perla et al. (2011) explain that such rules separate ordinary variation from signals of change, and by those rules the process was stable: nothing leaders tried during the six months, including a February reminder email, changed it. That matters for the target, because a stable process will not reach 40% without a change in how discharges are prepared.

Discharge Timing and Length of Stay

Patients who left before noon did not have shorter stays. Their median was 3.8 days compared with 3.6 days for afternoon departures, and the pattern held within both medical and surgical units. The likely explanation is that many morning discharges are patients held over an extra night, whose departure the following morning counts as early.

Published findings point the same way. In the program Wertheimer et al. (2014) described, the before-noon share climbed from 11% to 38% while the hospital's observed-to-expected length of stay fell, but the same effort also changed rounding and teamwork, so the timing itself cannot be credited with the drop. Rajkomar et al. (2016), studying medical and surgical discharges at an academic center, reported that morning discharge went along with slightly longer, not shorter, stays across medical and surgical patients. Together with local data, these studies suggest the target should not be sold as a way to shorten length of stay.

Discharge Timing and Emergency Boarding

The picture changes when the question is bed flow. On the 38 weekdays when at least 40% of patients left before noon, median boarding for emergency admissions was 3.2 hours; on the 41 weekdays below 30%, it was 4.6 hours. Emergency admission requests peak between late morning and midafternoon, while most departures occur between 2 and 5 p.m., so beds freed earlier line up better with the demand. This is an association across days, not proof of cause. Busy days may have fewer morning discharges and longer boarding for reasons unrelated to timing, such as high census or staffing, and the analysis could not adjust for all of them.

What the Old Timestamp Would Have Shown

To show why the quality assessment mattered, the analysis was repeated with the registration discharge time the dashboard has always used. On that basis only 29% of patients left before noon, the medical units appeared to discharge fewer than one patient in five in the morning and the link between morning discharges and shorter boarding almost disappeared, because clerk closing times bear little relation to when beds actually empty. Leaders reading that version would have concluded that the hospital was far from its target and that earlier discharge made no difference to the emergency department. Both conclusions would have been wrong. The comparison makes a practical case for the data management recommendations in the next milestone: the measure must be built on the event it claims to describe.

Limitations

The analysis covers only six months at one hospital. Nursing departure times were missing for 671 routine discharges, and if those patients differ from the rest, the rates may be slightly off. Length of stay was not adjusted for severity, although comparing within unit types reduces that concern. Finally, the before-noon and boarding results describe patterns, not the effect of an intervention.

What Leaders Should Expect

The data support pursuing earlier discharge for bed flow, not for shorter stays. If the goal is described as reducing emergency boarding, leaders can track the right outcome, set realistic expectations and avoid keeping patients an extra night to leave in the morning, which would lengthen stays. The medical units, at 28%, have the most room to improve. The next milestone will recommend how the hospital should manage these data so the measure stays accurate as the program begins.

References

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

Rajkomar, A., Valencia, V., Novelero, M., Mourad, M., & Auerbach, A. (2016). The association between discharge before noon and length of stay in medical and surgical patients. Journal of Hospital Medicine, 11(12), 859-861. https://doi.org/10.1002/jhm.2529

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