March 29, 2015
Recent research published in the Journal of Hospital Medicine in March, 2015 found that ‘face-to-face’ handoffs between internal medicine providers at a tertiary care academic medical center had no impact on the rate of medical codes, deaths or hospital readmissions. However, other researchers point out the study was i) not adequately powered to detect a difference if one existed; and ii) retrospective and thus cannot establish a true cause and effect relationship. So are ‘face-to-face’ handoffs safer than those not done in-person?
The common characteristics of handoffs across several industries with high consequences for failure were well described by Patterson et al in their seminal work which examined the common characteristics of handoffs in a NASA space center, nuclear power generation plants, a railroad dispatch center, and an ambulance dispatch center. The focus of the Patterson et al paper was learnings that medicine could take from how handoffs are conducted in these other settings where the consequences of failures to communicate can be catastrophic. One of the commonalities found across all the settings was the employment of a verbal ‘face-to-face’ in-person update at the time of the handover. In the scholarly publication, Patterson et al discusses some of the potential benefits of ‘face-to-face’ handoffs which included interactive questioning and non-verbal communication and cues.
Face to face handoffs have long been the gold standard for handoffs between providers in medical settings. Moreover, well designed, prospective, multi-center trials of handoff interventions, such as the IPASS study, have demonstrated that standardized handoffs between medical providers that are structured and include a verbal handoff can reduce medical errors and adverse events significantly.
For these reasons, the reported findings of Schouten et al in their March 2015 retrospective analysis have provided more data for discussion and discourse on this important patient safety topic. The reported findings were that ‘face-to-face’ inpatient internal medicine handoffs at the Mayo Clinic, a major tertiary academic medical center in the US, had no impact on the rate of rapid response team calls, medical codes, patient transfers to higher acuity settings, in hospital mortality, 30-day readmission rates, nor adverse events. While the investigators hypothesized that the transfers of care where ‘face-to-face’ handoffs were done would be correlated with fewer adverse events and better outcomes, the data instead confirmed the null hypothesis.
The Schouten et al piece entailed a retrospective analysis of 805 total patients and found that adverse events in the ‘face-to-face’ handoff group were 2.6 percent compared to 3.2 percent in the group without ‘face-to-face’ handoffs, but the difference was not significant. With a baseline adverse event rate of 3 percent, Jones et al, in an editorial piece which also appeared in the same edition of the Journal of Hospital Medicine, point out that conclusions that can be drawn from the Schouten et al analysis are limited. The reason? Jones et al report that the Schouten et al analysis was underpowered and would not be able to detect a difference if one existed in the first place. Based on Jones et al’s power calculations, they estimate that some 10,000 patients, 5,000 each in the ‘face-to-face’ and the ‘non-face-to-face’ arms of the study, would be required in order to detect a 30 percent difference in the adverse event rate.
Both pieces are recommended reads and undoubtedly needed contributions to the literature and the discussion on patient handoffs. What both Schouten et al and Jones et al agree on is that more research is needed in order to draw further conclusions, so it’s likely we will see the academy continue to probe and generate more data and insights in this important area of patient safety.