Joseph R Keebler1,2,1,3, Elizabeth H Lazzara4, Brady S Patzer4,5,4,6, Evan M Palmer4,5,4,6, John P Plummer4,5,4,6, Dustin C Smith5, Victoria Lew4, Sarah Fouquet4,5,4,6, Y Raymond Chan4,5,4,6, Robert Riss6. 1. Embry-Riddle Aeronautical University, Daytona Beach, Florida Joseph.keebler@erau.edu. 2. Wichita State University, Kansas Joseph.keebler@erau.edu. 3. Children's Mercy Hospital, Kansas City, Missouri Joseph.keebler@erau.edu. 4. Embry-Riddle Aeronautical University, Daytona Beach, Florida. 5. Wichita State University, Kansas. 6. Children's Mercy Hospital, Kansas City, Missouri.
Abstract
OBJECTIVE: The overall purpose was to understand the effects of handoff protocols using meta-analytic approaches. BACKGROUND: Standardized protocols have been required by the Joint Commission, but meta-analytic integration of handoff protocol research has not been conducted. METHOD: The primary outcomes investigated were handoff information passed during transitions of care, patient outcomes, provider outcomes, and organizational outcomes. Sources included Medline, SAGE, Embase, PsycINFO, and PubMed, searched from the earliest date available through March 30th, 2015. Initially 4,556 articles were identified, with 4,520 removed. This process left a final set of 36 articles, all which included pre-/postintervention designs implemented in live clinical/hospital settings. We also conducted a moderation analysis based on the number of items contained in each protocol to understand if the length of a protocol led to systematic changes in effect sizes of the outcome variables. RESULTS: Meta-analyses were conducted on 34,527 pre- and 30,072 postintervention data points. Results indicate positive effects on all four outcomes: handoff information (g = .71, 95% confidence interval [CI] [.63, .79]), patient outcomes (g = .53, 95% CI [.41, .65]), provider outcomes (g = .51, 95% CI [.41, .60]), and organizational outcomes (g = .29, 95% CI [.23, .35]). We found protocols to be effective, but there is significant publication bias and heterogeneity in the literature. Due to publication bias, we further searched the gray literature through greylit.org and found another 347 articles, although none were relevant to this research. Our moderation analysis demonstrates that for handoff information, protocols using 12 or more items led to a significantly higher proportion of information passed compared with protocols using 11 or fewer items. Further, there were numerous negative outcomes found throughout this meta-analysis, with trends demonstrating that protocols can increase the time for handover and the rate of errors of omission. CONCLUSIONS: These results demonstrate that handoff protocols tend to improve results on multiple levels, including handoff information passed and patient, provider, and organizational outcomes. These findings come with the caveat that publication bias exists in the literature on handoffs. Instances where protocols can lead to negative outcomes are also discussed. APPLICATION: Significant effects were found for protocols across provider types, regardless of expertise or area of clinical focus. It also appears that more thorough protocols lead to more information being passed, especially when those protocols consist of 12 or more items. Given these findings, publication bias is an apparent feature of this literature base. Recommendations to reduce the apparent publication bias in the field include changing the way articles are screened and published.
OBJECTIVE: The overall purpose was to understand the effects of handoff protocols using meta-analytic approaches. BACKGROUND: Standardized protocols have been required by the Joint Commission, but meta-analytic integration of handoff protocol research has not been conducted. METHOD: The primary outcomes investigated were handoff information passed during transitions of care, patient outcomes, provider outcomes, and organizational outcomes. Sources included Medline, SAGE, Embase, PsycINFO, and PubMed, searched from the earliest date available through March 30th, 2015. Initially 4,556 articles were identified, with 4,520 removed. This process left a final set of 36 articles, all which included pre-/postintervention designs implemented in live clinical/hospital settings. We also conducted a moderation analysis based on the number of items contained in each protocol to understand if the length of a protocol led to systematic changes in effect sizes of the outcome variables. RESULTS: Meta-analyses were conducted on 34,527 pre- and 30,072 postintervention data points. Results indicate positive effects on all four outcomes: handoff information (g = .71, 95% confidence interval [CI] [.63, .79]), patient outcomes (g = .53, 95% CI [.41, .65]), provider outcomes (g = .51, 95% CI [.41, .60]), and organizational outcomes (g = .29, 95% CI [.23, .35]). We found protocols to be effective, but there is significant publication bias and heterogeneity in the literature. Due to publication bias, we further searched the gray literature through greylit.org and found another 347 articles, although none were relevant to this research. Our moderation analysis demonstrates that for handoff information, protocols using 12 or more items led to a significantly higher proportion of information passed compared with protocols using 11 or fewer items. Further, there were numerous negative outcomes found throughout this meta-analysis, with trends demonstrating that protocols can increase the time for handover and the rate of errors of omission. CONCLUSIONS: These results demonstrate that handoff protocols tend to improve results on multiple levels, including handoff information passed and patient, provider, and organizational outcomes. These findings come with the caveat that publication bias exists in the literature on handoffs. Instances where protocols can lead to negative outcomes are also discussed. APPLICATION: Significant effects were found for protocols across provider types, regardless of expertise or area of clinical focus. It also appears that more thorough protocols lead to more information being passed, especially when those protocols consist of 12 or more items. Given these findings, publication bias is an apparent feature of this literature base. Recommendations to reduce the apparent publication bias in the field include changing the way articles are screened and published.
Authors: Nicholas A Rattray; Patricia Ebright; Mindy E Flanagan; Laura G Militello; Paul Barach; Zamal Franks; Shakaib U Rehman; Howard S Gordon; Richard M Frankel Journal: BMC Med Educ Date: 2018-11-03 Impact factor: 2.463
Authors: David P Stonko; Bradley M Dennis; Rachael A Callcut; Richard D Betzold; Michael C Smith; Andrew J Medvecz; Oscar D Guillamondegui Journal: PLoS One Date: 2018-12-03 Impact factor: 3.240