Background:

Failures in communication among healthcare personnel during intrahospital handoffs in care are known threats to patient safety. In August, 2009, our healthcare system held a multistakeholder summit on handoffs, developed consensus around the need for a system–wide electronic handoff tool, and recommended a pilot study to develop and evaluate this technology.

Methods:

We adapted a web–based handoff tool used by a single residency program. Enhancements to the existing tool included: (1) ability to implement the tool at a second hospital in our system; (2) support for simultaneous handoffs by nurses, residents/PAs, and attendings with shared information among the different roles; (3) custom structured templates for each user group; and (4) the ability to create progress notes and multiple sign–out forms from the same core data. The tool was refined and tested on a general medicine teaching service at one hospital and a hematologic malignancy PA service at the other. For 3 months preintervention and 4 months postimplementation, we surveyed receivers of handoffs regarding continuity of care and evaluated signout content using explicit criteria. We also conducted formal usability testing using simulated cases. We conducted principal components analysis to derive categories from the survey questions and create composite scores for each category.

Results:

We received survey responses from 315 clinicians (66% response rate). In a pre–post analysis, two of five composite scores improved: perceived negative impact of handoff on clinical information and decision–making (composite score 14.7 pre, 10.2 post, p = 0.01), and negative subjective rating of handoff quality and accuracy (28.4 vs 25.8, p = 0.01). Among survey questions to nurses, 10 improved, including an increase in how well handoffs prepared them for things that might go wrong (47.3 vs 65.2, p = 0.01). In the explicit review of written sign–outs, inclusion of five data elements (e.g., % tasks with if/then statements) increased, but decreases were noted in other data elements. Usability testing revealed a tension between desire for a clinical narrative and the use of structured template fields.

Conclusions:

A multidisciplinary, web–based sign–out tool was able to increase subjective measures of sign–out quality and impact on clinical decision–making, particularly among nurses. Much of the improvement may have come from the ability to produce both a progress note and sign–out with one tool, which led to more frequent updating of sign–outs and greater faith in their accuracy. The use of customized “templated” fields was inconsistent and suggests that these should be minimized to those most necessary for continuity of care. Greater improvements in care may require further enhancements in usability of the tool, training in use of the tool, and education in best practices in handoffs in care.