DDHA 9902 Week 10 Findings and Recommendations Paper Example

Late Summaries, Traced From Material to Recommendation: The Findings and Recommendations Paper for the Discharge Summary Project, With One Numbering Throughout

Student Name

Walden University

DDHA 9902: Practice-Based Problem Analysis, Findings, and Recommendations

Instructor Name

Month Day, Year

Late Summaries, Traced From Material to Recommendation: The Findings and Recommendations Paper for the Discharge Summary Project, With One Numbering Throughout

The Problem, in Brief

Last term established that summaries written at Larchmont's two hospitals, an invented setting, arrived at outside primary care practices inside the two business days promised to its accountable care partners only about half the time. The question this paper answers is where those summaries fail and what should change.

The Analysis Performed

A plan filed before any data were examined set out five steps, each fixed in advance with its own rule for what counts and its own label, following the logic Nosek et al. (2018) describe for analyses meant to be trusted. The steps covered 1,190 eligible summer discharges. Step S1 sorted the 571 late summaries by where each failed. Step S2 compared practices registered for secure messaging with fax-only practices. Step S3 compared the hospitals. Step S4 compared Friday discharges with those earlier in the week. Step S5, exploratory, compared readmission in a Medicare subset. Two deviations occurred: the wrong-practice category could be checked only where a follow-up visit was recorded, and registration status for secure messaging was available only as of today. Sensitivity checks with a three-day window and with recently registered practices removed did not change any decision.

What the Analysis Found

F1. The fax path was the largest single point of failure: 239 of the 571 late summaries, 42%, went to fax-only practices with no confirmed fax inside the window.

F1a. The other late summaries divided among signing after the window (86, or 15%), secure messages bounced from inactive addresses (63, or 11%), a practice other than the one visited (at least 41, or 7%) and causes that could not be identified (142, or 25%).

F2. Practices registered for secure messaging received 68% of their summaries on time; fax-only practices received 33%.

F3. The two hospitals differed by four points, 53% against 49%, under the threshold the plan set for treating them differently.

F4. Friday discharges reached practices on time 38% of the time, against 56% for discharges from Monday to Thursday.

F5. In the Medicare subset, readmission was 12.8% after timely summaries and 15.9% after late ones, with an interval that included no difference.

Where Each Finding Stops

Each finding has a limit, and the limits are attached to the findings they cap. F1 and F2 describe one summer, because the fax server keeps three months of records. None of F1 to F4 can say whether a delivered summary was read, since the logs record delivery to an address, not opening. The wrong-practice figure in F1a is a floor rather than an estimate, because visit data covered only part of the population, and the undetermined quarter could hide a different mix of causes. F2's gap may partly reflect the kind of practice that relies on fax, often smaller and less staffed, as well as the delivery route. F4 may reflect Friday volume or staffing as well as signing time. F5 was exploratory and inconclusive and supports no claim about clinical effect. Ioannidis (2007) found limitations commonly stated in general terms that constrain nothing; stating each against its finding is meant to avoid that.

What Should Change

Each recommendation names the finding beneath it.

R1, from F1 and F2. Bring every fax-only practice with regular Larchmont discharges onto secure messaging, starting with those receiving the most, and stop faxing them once their address is confirmed. Success would appear as a smaller fax share of late summaries and a rising on-time share for newly registered practices.

R2, from F1a. Have health information management work through each morning's failed transmissions and resend them within a business day, and reconcile practice addresses with the messaging vendor monthly. Success would appear as fewer bounced messages and a smaller undetermined share.

R3, from F4 and the signing portion of F1a. Ask hospitalists to sign before the patient leaves on Fridays and weekends, with the evening or weekend physician signing anything left open. Success would appear as Friday on-time delivery moving toward the midweek figure.

R4, from the wrong-practice portion of F1a, carried forward as weak. Confirm the primary care practice at registration and again during discharge planning, after first widening the visit data so that the true size of the problem is known.

F3 and F5 support no recommendation. Because the hospitals did not differ meaningfully, every recommendation applies to both. Because the readmission comparison was inconclusive, readmission is not used to judge any of them.

How the Recommendations Fit Together

The four recommendations are designed to reinforce one another. R2 comes first in practice because its daily list will reveal causes that are now undetermined and will show whether R1's newly registered practices are receiving messages reliably. R1 attacks the largest failure point directly. R3 addresses a predictable weekly pattern with a change in timing rather than technology. R4 waits for better measurement, because acting on a weak finding could spend effort on a problem smaller than it looks. Together they target every identified cause except the undetermined quarter, which R2 is expected to shrink.

Walking a Recommendation Back

The paper is built so that any recommendation can be traced to its material without leaving the document. Take R3. It cites F4 and part of F1a. F4 comes from Step S4, which compared Friday and midweek discharges using the signature audit trail and the two transmission logs for the summer months. Its limit, that Friday volume or staffing may explain part of the gap, is stated in the section above, which is why R3 is framed as a trial. Each other recommendation can be walked back the same way.

What the Paper Does Not Propose

It does not propose a new transitions platform. Nothing in the findings identifies a failure that new software would correct and the current system would not, since the record system already sends summaries automatically once they are signed and correctly addressed. Kripalani et al. (2007) describe the risks of information gaps after discharge, and those risks argue for fixing the failure points identified here quickly and cheaply.

References

Ioannidis, J. P. A. (2007). Limitations are not properly acknowledged in the scientific literature. Journal of Clinical Epidemiology, 60(4), 324-329. https://doi.org/10.1016/j.jclinepi.2006.09.011

Kripalani, S., LeFevre, F., Phillips, C. O., Williams, M. V., Basaviah, P., & Baker, D. W. (2007). Deficits in communication and information transfer between hospital-based and primary care physicians: Implications for patient safety and continuity of care. JAMA, 297(8), 831-841. https://doi.org/10.1001/jama.297.8.831

Nosek, B. A., Ebersole, C. R., DeHaven, A. C., & Mellor, D. T. (2018). The preregistration revolution. Proceedings of the National Academy of Sciences, 115(11), 2600-2606. https://doi.org/10.1073/pnas.1708274114