JMIR Publications

ABSTRACT

Background:

Nurse workload depends not only on staffing levels but on how patients are distributed among the nurses on a shift. Nurse leaders make this decision under time pressure and with information scattered across the electronic health record (EHR), verbal report, and their own knowledge, yet how they construct assignments remains poorly described.

Objective:

To describe how nurse leaders manage staff nurses’ workload and construct nurse-patient assignments, and to translate these findings into design requirements for data-based decision support.

Methods:

We conducted a qualitative descriptive study at one academic health system in the southeastern United States. Eleven nurse leaders responsible for assignments on nine inpatient units each completed an initial and a follow-up semi-structured interview guided by the Systems Engineering Initiative for Patient Safety (SEIPS) framework. Transcripts and shift notes were analyzed using codebook-based thematic analysis.

Results:

Five themes emerged: core assignment heuristics and strategies (central theme), evaluating patient-driven workload, staff characteristics and competency matching, information gathering and sensemaking, and navigating systems and environmental constraints. Leaders shared a common set of principles (workload equity, pursued by dispersion of high-demand patients, geographic proximity, and continuity of care) but weighted them according to unit size, layout, and patient turnover. Psychosocial and behavioral demand was among the most consequential inputs yet was largely absent from the EHR, reaching leaders through verbal report and direct observations; EHR tools served as a starting reference rather than a decision basis. The outgoing charge nurse built the assignment, the oncoming charge nurse revised it throughout the shift, and leaders and experienced nurses absorbed work the assignment could not hold.

Conclusions:

The main burden of nurse-patient assignment lies in retrieving and reconciling information rather than in reasoning about it. Decision support should assemble information across sources rather than allocate patients, use locally calibrated weights, and keep the leader’s judgment at the point of decision. Clinical Trial: N/A