Free
Perioperative Medicine  |   October 2014
Intraoperative Transitions of Anesthesia Care and Postoperative Adverse Outcomes
Author Affiliations & Notes
  • Leif Saager, Dr. med.
    From the Department of Outcomes Research, Anesthesiology Institute, Cleveland Clinic, Cleveland, Ohio (L.S., B.D.H., J.Y., A.T., E.J.M., D.I.S., A.K.) and Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio (J.Y., E.J.M.).
  • This article is featured in “This Month in Anesthesiology,” page 1A. Corresponding article on page 673.
    This article is featured in “This Month in Anesthesiology,” page 1A. Corresponding article on page 673.×
  • Submitted for publication November 13, 2013. Accepted for publication June 25, 2014.
    Submitted for publication November 13, 2013. Accepted for publication June 25, 2014.×
  • Address correspondence to Dr. Saager: Department of Outcomes Research, Cleveland Clinic, 9500 Euclid Avenue, P-77, Cleveland, Ohio 44195. saagerl@ccf.org. This article may be accessed for personal use at no charge through the Journal Web site, www.anesthesiology.org.
Article Information
Perioperative Medicine / Clinical Science / Education / CPD / Patient Safety
Perioperative Medicine   |   October 2014
Intraoperative Transitions of Anesthesia Care and Postoperative Adverse Outcomes
Anesthesiology 10 2014, Vol.121, 695-706. doi:10.1097/ALN.0000000000000401
Anesthesiology 10 2014, Vol.121, 695-706. doi:10.1097/ALN.0000000000000401
Abstract

Background:: Transfers of patient care and responsibility among caregivers, “handovers,” are common. Whether handovers worsen patient outcome remains unclear. The authors tested the hypothesis that intraoperative care transitions among anesthesia providers are associated with postoperative complications.

Methods:: From the records of 138,932 adult Cleveland Clinic (Cleveland, Ohio) surgical patients, the authors assessed the association between total number of anesthesia handovers during a case and an adjusted collapsed composite of in-hospital mortality and major morbidities using multivariable logistic regression.

Results:: Anesthesia care transitions were significantly associated with higher odds of experiencing any major in-hospital mortality/morbidity (incidence of 8.8, 11.6, 14.2, 17.0, and 21.2% for patients with 0, 1, 2, 3, and ≥4 transitions; odds ratio 1.08 [95% CI, 1.05 to 1.10] for an increase of 1 transition category, P < 0.001). Care transitions among attending anesthesiologists and residents or nurse anesthetists were similarly associated with harm (odds ratio 1.07 [98.3% CI, 1.03 to 1.12] for attending [incidence of 9.4, 13.9, 17.4, and 21.5% for patients with 0, 1, 2, and ≥3 transitions] and 1.07 [1.04 to 1.11] for residents or nurses [incidence of 9.4, 13.0, 15.4, and 21.2% for patients with 0, 1, 2, and ≥3 transitions], both P < 0.001). There was no difference between matched resident only (8.5%) and nurse anesthetist only (8.8%) cases on the collapsed composite outcome (odds ratio, 1.00 [98.3%, 0.93 to 1.07]; P = 0.92).

Conclusion:: Intraoperative anesthesia care transitions are strongly associated with worse outcomes, with a similar effect size for attendings, residents, and nurse anesthetists.

Each anesthetic handover increased the risk of any major in-hospital morbidity or mortality by 8%. The adverse effects of handovers were similar for attending anesthesiologists and medically directed residents and certified registered nurse anesthetists. The adverse effects of handovers were virtually identical for residents and certified registered nurse anesthetists.

What We Already Know about This Topic
  • Intraoperative transfers of patient care and responsibilities among anesthesia caregivers, that is, handovers, are relatively frequent

  • Lost critical information during handovers may result in delays, inefficiencies, suboptimal care, or patient harm

What This Article Tells Us That Is New
  • Each anesthetic handover increased the risk of any major in-hospital morbidity or mortality by 8%

  • The adverse effects of handovers were similar for attending anesthesiologists and medically directed residents and certified registered nurse anesthetists

  • The adverse effects of handovers were virtually identical for residents and certified registered nurse anesthetists

TRANSFERS of patient care and responsibility among caregivers, “handovers,” are inevitable as care for individuals often extends over shifts—and sometimes over days or weeks. The number of handovers, at least in academic hospitals, has increased as a result of duty-hour limitations.1–4 
Critical details may be lost during handovers resulting in delays,5  inefficiencies,6,7  suboptimal care,8  or even patient harm.9,10  Consequently, the Joint Commission on Hospital Accreditation declared in 2006 that “implementing a standardized approach to handoff communications including the opportunity to ask and respond to questions” was a national patient safety goal.*01  They also identified “communication failure” to be the root cause of 65% of all sentinel events in 2006.†02  The World Health Organization similarly listed “communication during patient care handover” as one of its “High five” patient safety initiatives.‡03  Numerous studies have identified challenges associated with handovers and evaluated various systems and methods for enhancing communication and information transfer.11–17  There are also studies evaluating anecdotal complications18–20  and malpractice cases.21  But surprisingly, there is little evidence that care transitions worsen patient outcome.
The high-risk perioperative period presents an opportunity to study care transitions and their effect on mortality and serious complications. Typically, a single surgical team provides care throughout an operation. However, handovers among anesthesia providers are common, and may involve attendings, residents, and certified registered nurse anesthetists (CRNAs). Currently, no universally accepted guidelines or recommendations for performing intraoperative handovers exist, and very few studies have investigated anesthesia care transitions.
As with other types of care transition, it remains unknown whether changes in anesthesia providers worsen patient outcome. We, therefore, tested the primary hypothesis that the total number of intraoperative handovers among anesthesia providers is associated with an increase in a composite of postoperative mortality and serious complications. Secondarily, we evaluated independent associations for attending handovers, and for resident and CRNA handovers.
Materials and Methods
With approval from the Cleveland Clinic Institutional Review Board (Cleveland, Ohio), patient information was obtained from the Cleveland Clinic Perioperative Health Documentation System. The registry contains all patients who had noncardiac surgery since 2005 at Cleveland Clinic’s main campus. It integrates preoperative variables (demographics, conditions, etc.), intraoperative variables (via our Anesthesia Record Keeping System), and postoperative outcomes (by linking to the larger Cleveland Clinic billing data systems).
Handovers among anesthesia providers at the Cleveland Clinic do not follow a formal script, and we do not normally use checklists. Although anesthesia providers are trained to convey all-important information to their relief, no formalized training or standardized process has been implemented.
Statistical Analysis
We assessed the association between the total number of anesthesia handovers during a case and a collapsed composite (any vs. none) of in-hospital mortality and six major morbidities including serious cardiac, respiratory, gastrointestinal, urinary, bleeding, and infectious complications (as defined in appendix 1), using multivariable logistic regression. We adjusted for the following prespecified potential confounding variables: age, sex, race, American Society of Anesthesiologists (ASA) physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure.
The total number of anesthesia handovers includes handovers among attending anesthesiologists and handovers among medical-directed anesthesia providers including residents and fellows, CRNAs, and student nurse anesthetists. For medical-directed anesthesia providers, breaks of less than 40 min were not counted as a handover; for example, it was not considered a handover when a provider relieved someone for, say lunch, and then returned within 40 min. The total number of anesthesia handovers was truncated at four because there were more than 4 in only 1,448 (1%) of the patients.
We adjusted for severity of procedure (in terms of risk of outcome) as follows: First, we characterized each patient’s primary procedure using the U.S. Agency for Healthcare Research and Quality’s single-level Clinical Classifications Software for International Classification of Diseases, 9th Revision, Clinical Modification procedure codes. The single-level Clinical Classifications Software is a tool for aggregating the 1,965 individual procedure codes in our dataset into 207 clinically meaningful procedure categories. Because of this large number of categories, we adjusted for severity of procedure as a continuous covariable by using the incidence of the collapsed composite outcome for each Clinical Classifications Software category. Clinical Classifications Software categories with a frequency less than 20 were collapsed into one category. Diagnosis-related risk for the collapsed composite outcome was estimated and adjusted for in the analysis in a similar manner.
We conducted a sensitivity analysis comparing each positive number of handovers (1, 2, 3, and ≥4) with 0 handovers using propensity score matching to adjust for potential confounders (i.e., a total of four propensity score matching analyses).22  This was in contrast to the primary analysis in which confounding was adjusted for by multivariable modeling and the association between number of handovers and outcome was assumed to be linear. First, we estimated the probability (i.e., the propensity score) of having exactly one handover (vs. none) using logistic regression based on age, sex, race, ASA status, start time of surgery, duration of surgery, and severity of principal diagnosis and procedure. We used a 1-to-2 greedy distance-matching algorithm SAS macro: gmatch,§04  which makes the locally optimal choice, employing a maximum propensity score difference of 0.01 units. Specifically, the algorithm tried to match each patient having one handover to a maximum of two patients having no handovers with the smallest propensity score difference (the maximum allowable difference was 0.01). Similarly, we obtained the other three propensity matched sets of patients (i.e., 2 handovers vs. 0, 3 vs. 0, and ≥4 vs. 0). Assessment of covariable balance after matching was performed using standardized differences (i.e., difference in means or proportions divided by the pooled SD). Imbalance was very conservatively defined as a standardized difference greater than (n1, n2, are the number of matched patients in each group) in absolute value; any such covariables would have been entered into our multivariable logistic regression model when comparing the matched groups on outcomes to reduce potential confounding. The significance criterion was P value less than 0.0125 for each comparison to maintain the overall alpha at 0.05 across these four analyses.
We conducted another sensitivity analysis in which we assessed individual associations between the number of handovers (as a continuous variable) and specific components of the composite as well as the common effect “global” odds ratio (OR) of the number of handovers across all the components of the composite using separate distinct effects generalized estimating equation multivariate models with unstructured covariance matrix. A Bonferroni correction for simultaneous comparisons was employed to control the type I error, so that P value less than 0.007 was considered significant for a particular component (i.e., 0.05/7 = 0.007).
Secondary Analyses
Furthermore, for informational purposes, we conducted four exploratory analyses in which we evaluated the relationships between the total number of anesthesia handovers and the collapsed composite of major morbidities in the following subsets of cases: (1) those not started in regular work hours (before 7:00 am and 5:00 pm); (2) those patients with ASA physical status 3 or 4; (3) those cases less than 1 h; and (4) those cases more than 4 h. Each analysis used the same statistical method as the primary analysis.
Also, we evaluated the relationship between total number of anesthesia handovers and length of postoperative hospital stay using Cox proportional hazards regression. The outcome event in the model was “discharged alive.” Patients who died in-hospital were analyzed as never having the event and were assigned a censoring time equal to the observed longest hospitalization among those discharged alive.
Secondarily, we simultaneously assessed the relationship between number of attending anesthesiologist handovers and number of medical-directed provider handovers with the collapsed composite in-hospital mortality/morbidity using a single multivariable logistic regression. For this analysis, the number of attending anesthesiologist and medical-directed handovers were both truncated to three to facilitate modeling.
Anesthesia care at our institution is provided by residents and CRNAs, and sometimes both are involved in a single anesthetic. Residents and CRNAs are always supervised by an attending anesthesiologist. For training and educational purposes, residents are typically assigned to more challenging or complex cases. Furthermore, night calls and weekend calls are mostly covered by residents. We thus conducted an additional analysis comparing patients who were managed by attending anesthesiologist and residents only, or by attending anesthesiologist and CRNAs only on the collapsed composite outcome, using a multivariable logistic regression. To control for potential confounding, we exactly matched on principal procedure and diagnosis, start time of the case, and ASA status for 31,816 patients who were managed exclusively by attending anesthesiologist and residents to 31,816 patients who were managed by exclusively attending anesthesiologist and CRNAs. We also adjusted for age, sex, race, duration of surgery, and number of handovers.
The significance level was maintained at 0.05 within the primary and secondary analyses. Thus, the significance criterion was P value less than 0.006 for each secondary analysis (a total of eight analyses, Bonferroni correction). SAS software version 9.3 (SAS Institute, Cary, NC) was used for all statistical analyses.
Results
We included data from 138,932 adults who had noncardiac surgery with general and/or regional anesthesia at the Cleveland Clinic between January 06, 2005 and December 31, 2012 and had an ASA physical status 4 or less. Patients with any missing values were excluded. Therefore, 135,810 patients were included in our analyses; 82,644 (61%), 27,982 (21%), 15,102 (11%), 6,172 (5%), and 3,910 (3%) patients had 0, 1, 2, 3, and 4 or more handovers, respectively. Table 1 shows baseline and intraoperative characteristics
Table 1.
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers×
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers
Table 1.
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers×
×
The observed incidence of the collapsed composite in-hospital mortality/morbidity was 8.8, 11.6, 14.2, 17.0, and 21.2% for patients with 0, 1, 2, 3, and 4 or more anesthesia handovers, respectively (table 2). More anesthesia handovers during a case were significantly associated with higher odds of experiencing any major in-hospital mortality/morbidity (P < 0.001). The estimated OR was 1.08 (95% CI, 1.05 to 1.10) for an increase of one transition category, after adjusting for age, sex, race, ASA status, principal diagnosis and procedure, duration of surgery, and start time of the case (appendix 2). Consistent results were provided by our propensity score matching sensitivity analysis. Increasing numbers of anesthesia handovers during a case (2, 3, and 4 or more) was significantly associated with higher odds of experiencing in-hospital mortality/morbidity compared to no handover (table 3 and appendix 3).
Table 2.
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers×
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers
Table 2.
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers×
×
Table 3.
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching×
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching
Table 3.
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching×
×
Furthermore, all the evaluated individual associations between number of handovers and specific components included in our composite were in the same direction; more anesthesia handovers during a case was significantly associated with higher odds of experiencing cardiac, gastrointestinal, bleeding, and infectious morbidities (table 2 and fig. 1). The common effect OR of handovers across the individual components of the composite outcome was estimated as 1.15 (95% CI, 1.12 to 1.19) for a difference of one transition category.
Fig. 1.
Odds ratios of having any in-hospital mortality/morbidities (collapsed composite), each specific individual component of the composite (individual component), and common “global” odds ratio across the individual components (common effect) for each increase in the total number of anesthesia handovers. *CIs for the individual components were 99.3%, adjusted using the Bonferroni correction.
Odds ratios of having any in-hospital mortality/morbidities (collapsed composite), each specific individual component of the composite (individual component), and common “global” odds ratio across the individual components (common effect) for each increase in the total number of anesthesia handovers. *CIs for the individual components were 99.3%, adjusted using the Bonferroni correction.
Fig. 1.
Odds ratios of having any in-hospital mortality/morbidities (collapsed composite), each specific individual component of the composite (individual component), and common “global” odds ratio across the individual components (common effect) for each increase in the total number of anesthesia handovers. *CIs for the individual components were 99.3%, adjusted using the Bonferroni correction.
×
In the exploratory analyses, we found that more anesthesia handovers was significantly associated with increased risk of the collapsed composite outcome for those started late (P < 0.001), those patients with ASA physical status 3 or 4 (P < 0.001), and those cases more than 4 h (P < 0.001), but not for cases less than 1 h (P = 0.92) (table 4).
Table 4.
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets×
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets
Table 4.
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets×
×
Length of postoperative hospital stay was 1 [0, 3] (median [Q1, Q3]), 2 [1, 5], 3 [1, 6], 4 [2, 6], and 4 [2, 8] days for patients with 0, 1, 2, 3, and 4 or more handovers, respectively (univariable P < 0.001, log-rank test). However, after controlling for the same set of potential confounding variables as in the primary analysis, the association between number of handovers and length of hospital stay was not significant (hazards ratio [99.4 % CI], 1.00 [0.99 to 1.01] for a difference of one transition category, P = 0.40).
Second, we found that more anesthesia handovers among attending anesthesiologists and among medical-directed anesthesia providers during a case were both significantly associated with higher odds of experiencing any major in-hospital mortality/morbidity (both P < 0.001; fig. 2). The observed incidence of the collapsed composite in-hospital mortality/morbidity was 9.4% (out of 102,516), 13.9% (26,754), 17.4% (5,464), and 21.5% (1,076) for patients with 0, 1, 2, and 3 or more attending handovers, respectively; similarly, the observed incidence was 9.4% (98,412), 13.0 (25,249), 15.4% (9,173), and 21.1% (2,976) for patients with 0, 1, 2, and 3 or more medical-directed handovers, respectively. After adjusting for the potential confounding variables the estimated ORs were 1.07 (99.4% CI, 1.02 to 1.13) for a difference of one in the number of anesthesia attending handovers and 1.07 (99.4% CI, 1.03 to 1.11) for a difference of one in the number of medically directed handovers. Furthermore, there was no interaction between attending anesthesiologist handovers and medical-directed handovers (P = 0.11).
Fig. 2.
Odds ratios of having any in-hospital mortality/morbidities for (A) each level of attending anesthesiologist handovers versus no attending anesthesiologist handovers; and (B) each level of medical-directed provider handovers (including residents, fellows, certified registered nurse anesthetists, and student registered nurse anesthetists) versus no medically directed provider handovers. We adjusted for age, sex, race, American Society of Anesthesiologists physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure. *The significance criterion for each analysis was P < 0.006 (i.e., 0.05/8, a total of eight secondary analyses, Bonferroni correction). Multiple comparisons for each analysis were further adjusted using the Bonferroni correction; thus, 99.8% CIs were estimated (significance criterion: 0.006/3, three comparisons for each analysis).
Odds ratios of having any in-hospital mortality/morbidities for (A) each level of attending anesthesiologist handovers versus no attending anesthesiologist handovers; and (B) each level of medical-directed provider handovers (including residents, fellows, certified registered nurse anesthetists, and student registered nurse anesthetists) versus no medically directed provider handovers. We adjusted for age, sex, race, American Society of Anesthesiologists physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure. *The significance criterion for each analysis was P < 0.006 (i.e., 0.05/8, a total of eight secondary analyses, Bonferroni correction). Multiple comparisons for each analysis were further adjusted using the Bonferroni correction; thus, 99.8% CIs were estimated (significance criterion: 0.006/3, three comparisons for each analysis).
Fig. 2.
Odds ratios of having any in-hospital mortality/morbidities for (A) each level of attending anesthesiologist handovers versus no attending anesthesiologist handovers; and (B) each level of medical-directed provider handovers (including residents, fellows, certified registered nurse anesthetists, and student registered nurse anesthetists) versus no medically directed provider handovers. We adjusted for age, sex, race, American Society of Anesthesiologists physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure. *The significance criterion for each analysis was P < 0.006 (i.e., 0.05/8, a total of eight secondary analyses, Bonferroni correction). Multiple comparisons for each analysis were further adjusted using the Bonferroni correction; thus, 99.8% CIs were estimated (significance criterion: 0.006/3, three comparisons for each analysis).
×
Within the matched subset of resident-only and CRNA-only cases, there was no difference between resident-only cases and CRNA-only cases on the collapsed composite outcome: OR, 1.00 (99.4% CI, 0.93 to 1.09) resident versus CRNA; P = 0.92.
Discussion
Rather than evaluate a surrogate endpoint such as information transfer, we directly evaluated a composite of in-hospital mortality and serious complications—an outcome that is important to patients and to the healthcare system. Our primary result is that each anesthetic handover increased the risk of composite outcome by a statistically significant 8%.
Previous work clearly demonstrates that critical information, including administered medications,12  is often lost during care transitions.11,23,24  Although it is logical to assume that improved information transfer will improve patient outcomes, there is in fact limited previous evidence that handovers actually worsen patient outcomes. Our results strongly suggest that they do; furthermore, the effect is substantial—1.08 times more likely to develop serious complications and mortality during hospital stay per transition, 1.17 (i.e., 1.082) times more likely for two transitions, and so forth.
To illustrate this further, we could expect to have 0.4 to 0.8% more patients experiencing at least one major in-hospital morbidity or mortality per transition of care, based on the observed incidence of 8.8% for patients with no handovers. We conducted a sensitivity analysis, where we assessed the common “global” effect of the handovers across all the individual components of the collapsed composite outcome and found that the common effect OR was 1.15 (1.12 to 1.19) for each increase in the total number of anesthesia handovers. This corresponds to approximately 0.2 to 0.3% increase in the incidence of each component of the composite outcome for each transition, based on the observed average incidence of 1.55% for patients with no handovers; thus, an overall of 1.3 to 2.0% increase in the incidence of all components (5 to 7.5 more complications per week). Given all the factors contributing to perioperative mortality and complications, it is remarkable that a single care transition is so harmful.
The adverse effect of handovers was similar for attending anesthesiologists and medically directed residents and CRNAs. Furthermore, the adverse effects were virtually identical for residents (who are still in training) and CRNAs (most of whom have considerable experience). These data suggest that the adverse effects of handovers are not limited to physicians-in-training; handovers even by experienced attendings and CRNAs comparably worsened patient outcomes.
The Cleveland Clinic does not have a formal handover process for anesthesia. Formal protocols for handovers, including checklists, clearly improve information transfer.25–28  The observed adverse effect of anesthetic turnovers might thus have been ameliorated—or even eliminated—by an enhanced handover process.29,30  Previous work indicated that checklists improve information transfer during care handovers.25,27,28,31,32  One reasonable response to our results might thus be to formalize the handover process.
There are compelling reasons to restrict duty hours since fatigue per se markedly impairs judgment,33  to say nothing of concentration and attention.34  Nonetheless, limits on duty hours for residents have increased the number of handovers in training hospitals.1  A second reasonable response to our results might thus be workflow redesigns that reduce the number of handovers, while keeping residents within duty-hour limits and CRNAs and attending shift durations within safe limits.
We studied the intraoperative period because surgery is a high-risk procedure; furthermore, anesthetic decisions must often be made quickly and on the basis of information already known to the practitioner without recourse to medical records. Handovers were relatively frequent in our patients, whereas care transitions in critical care units and regular nursing floors typically occur only once at the end of a shift. It remains to be determined whether transitions worsen patient outcomes in these and other clinical situations.
Breaks for most anesthesia providers at most U.S. institutions last 15 to 30 min. We excluded these temporary care transitions because providers who are familiar with a patient and return to continue care seem quite different from a provider adopting a complete new case. Previous research supports our notion that short breaks do not affect patient outcomes.29,30,35 
Statistical adjustment for potential confounding factors was key to our analysis. For example, it is obvious that handovers are more likely during longer than shorter cases. Similarly, handovers are more likely when cases start later in the day. We thus fully adjusted for these and many other factors including principal diagnosis and procedure. Furthermore, we conducted a sensitivity analysis using the propensity score matching technique comparing each number of handovers with no handovers, which provides some protection against selection bias and confounding due to measured factors.
Although we adjusted for start time and duration, diagnosis, procedure, and ASA physical status, the dataset utilized for this study includes a total of 5,918 (4.4%) emergency surgeries. If emergency surgery is included into the model, the estimated OR is almost identical with our original finding (1.080 [95% CI, 1.058 to 1.103] vs. 1.076 [1.054 to 1.099] for each increase in the total number of anesthesia handovers).
Our study was conducted at a single academic medical center. Presumably the association between handovers and adverse outcomes differs among institutions. The frequency of handovers also differs among hospitals, depending on structure, case duration, and scheduling priorities. For example, handovers are relatively rare in private settings.
As with all retrospective analyses, it is important to recognize that our results show a strong association between anesthesia care transitions and adverse outcomes, but causality cannot be assumed.
In summary, intraoperative care transitions between anesthesia providers were associated with significantly worsened patient outcomes. The effect size was similar for attendings, residents, and CRNAs. Our results suggest that reducing the number of care transitions has the potential to improve patient care. It is likely that formalizing the handover process will also help.
Acknowledgments
Support was provided solely from institutional and/or departmental sources.
Competing Interests
The authors declare no competing interests.
*2009 National Patient Safety Goals, Jt Comm Perspect, 2008. Available at: http://www.jcrinc.com/common/PDFs/fpdfs/pubs/pdfs/JCReqs/JCP-07-08-S1.pdf. Accessed October 20, 2013.
2009 National Patient Safety Goals, Jt Comm Perspect, 2008. Available at: http://www.jcrinc.com/common/PDFs/fpdfs/pubs/pdfs/JCReqs/JCP-07-08-S1.pdf. Accessed October 20, 2013.×
Joint C: Improving America’s Hospitals. The Joint Commission’s Annual Report on Quality and Safety, 2007. Available at: http://www.jointcommission.org/assets/1/18/2006_Annual_Report.pdf. Accessed October 20, 2013.
Joint C: Improving America’s Hospitals. The Joint Commission’s Annual Report on Quality and Safety, 2007. Available at: http://www.jointcommission.org/assets/1/18/2006_Annual_Report.pdf. Accessed October 20, 2013.×
World Health Organization Collaborating Center for Patient Safety: Communication during Patient Handovers. Geneva, Switzerland, WHO Press; 2007. Available at: http://www.who.int/patientsafety/solutions/patientsafety/PS-Solution3.pdf. Accessed October 20, 2013.
World Health Organization Collaborating Center for Patient Safety: Communication during Patient Handovers. Geneva, Switzerland, WHO Press; 2007. Available at: http://www.who.int/patientsafety/solutions/patientsafety/PS-Solution3.pdf. Accessed October 20, 2013.×
§Bergstralh EKJ. Gmatch SAS Program, Mayo Clinic Division of Biomedical Statistics and Informatics. Rochester, Mayo Clinic (HSR CodeXchange), 2003. Computerized matching of cases to controls using the greedy matching algorithm with a fixed number of controls per case. Available at: http://www.mayo.edu/research/documents/gmatchsas/doc-10027248. Accessed October 20, 2013.
Bergstralh EKJ. Gmatch SAS Program, Mayo Clinic Division of Biomedical Statistics and Informatics. Rochester, Mayo Clinic (HSR CodeXchange), 2003. Computerized matching of cases to controls using the greedy matching algorithm with a fixed number of controls per case. Available at: http://www.mayo.edu/research/documents/gmatchsas/doc-10027248. Accessed October 20, 2013.×
References
Horwitz, LI, Krumholz, HM, Green, ML, Huot, SJ Transfers of patient care between house staff on internal medicine wards: A national survey.. Arch Intern Med. (2006). 166 1173–7 [Article] [PubMed]
Okie, S An elusive balance—Residents’ work hours and the continuity of care.. N Engl J Med. (2007). 356 2665–7 [Article] [PubMed]
Van Eaton, EG, Horvath, KD, Lober, WB, Rossini, AJ, Pellegrini, CA A randomized, controlled trial evaluating the impact of a computerized rounding and sign-out system on continuity of care and resident work hours.. J Am Coll Surg. (2005). 200 538–45 [Article] [PubMed]
Barden, CB, Specht, MC, McCarter, MD, Daly, JM, Fahey, TJIII Effects of limited work hours on surgical training.. J Am Coll Surg. (2002). 195 531–8 [Article] [PubMed]
Beach, C, Croskerry, P, Shapiro, M Center for Safety in Emergency Care, Profiles in patient safety: Emergency care transitions.. Acad Emerg Med. (2003). 10 364–7 [Article] [PubMed]
Lofgren, RP, Gottlieb, D, Williams, RA, Rich, EC Post-call transfer of resident responsibility: Its effect on patient care.. J Gen Intern Med. (1990). 5 501–5 [Article] [PubMed]
Lyons, MN, Standley, TD, Gupta, AK Quality improvement of doctors’ shift-change handover in neuro-critical care.. Qual Saf Health Care. (2010). 19 e62 [Article] [PubMed]
Arora, V, Johnson, J, Lovinger, D, Humphrey, HJ, Meltzer, DO Communication failures in patient sign-out and suggestions for improvement: A critical incident analysis.. Qual Saf Health Care. (2005). 14 401–7 [Article] [PubMed]
Pronovost, PJ, Thompson, DA, Holzmueller, CG, Lubomski, LH, Dorman, T, Dickman, F, Fahey, M, Steinwachs, DM, Engineer, L, Sexton, JB, Wu, AW, Morlock, LL Toward learning from patient safety reporting systems.. J Crit Care. (2006). 21 305–15 [Article] [PubMed]
Kitch, BT, Cooper, JB, Zapol, WM, Marder, JE, Karson, A, Hutter, M, Campbell, EG Handoffs causing patient harm: A survey of medical and surgical house staff.. Jt Comm J Qual Patient Saf. (2008). 34 563–70 [PubMed]
Anderson, J, Shroff, D, Curtis, A, Eldridge, N, Cannon, K, Karnani, R, Abrams, T, Kaboli, P The Veterans Affairs shift change physician-to-physician handoff project.. Jt Comm J Qual Patient Saf. (2010). 36 62–71 [PubMed]
Arora, V, Kao, J, Lovinger, D, Seiden, SC, Meltzer, D Medication discrepancies in resident sign-outs and their potential to harm.. J Gen Intern Med. (2007). 22 1751–5 [Article] [PubMed]
Bhabra, G, Mackeith, S, Monteiro, P, Pothier, DD An experimental comparison of handover methods.. Ann R Coll Surg Engl. (2007). 89 298–300 [Article] [PubMed]
Cheah, LP, Amott, DH, Pollard, J, Watters, DA Electronic medical handover: Towards safer medical care.. Med J Aust. (2005). 183 369–72 [PubMed]
Kemp, CD, Bath, JM, Berger, J, Bergsman, A, Ellison, T, Emery, K, Garonzik-Wang, J, Hui-Chou, HG, Mayo, SC, Serrano, OK, Shridharani, S, Zuberi, K, Lipsett, PA, Freischlag, JA The top 10 list for a safe and effective sign-out.. Arch Surg. (2008). 143 1008–10 [Article] [PubMed]
Pickering, BW, Hurley, K, Marsh, B Identification of patient information corruption in the intensive care unit: Using a scoring tool to direct quality improvements in handover.. Crit Care Med. (2009). 37 2905–12 [Article] [PubMed]
Nagpal, K, Arora, S, Abboudi, M, Vats, A, Wong, HW, Manchanda, C, Vincent, C, Moorthy, K Postoperative handover: Problems, pitfalls, and prevention of error.. Ann Surg. (2010). 252 171–6 [Article] [PubMed]
Beckmann, U, Gillies, DM, Berenholtz, SM, Wu, AW, Pronovost, P Incidents relating to the intra-hospital transfer of critically ill patients. An analysis of the reports submitted to the Australian Incident Monitoring Study in Intensive Care.. Intensive Care Med. (2004). 30 1579–85 [Article] [PubMed]
Petersen, LA, Brennan, TA, O’Neil, AC, Cook, EF, Lee, TH Does housestaff discontinuity of care increase the risk for preventable adverse events?. Ann Intern Med. (1994). 121 866–72 [Article] [PubMed]
Mukherjee, S A precarious exchange.. N Engl J Med. (2004). 351 1822–4 [Article] [PubMed]
Greenberg, CC, Regenbogen, SE, Studdert, DM, Lipsitz, SR, Rogers, SO, Zinner, MJ, Gawande, AA Patterns of communication breakdowns resulting in injury to surgical patients.. J Am Coll Surg. (2007). 204 533–40 [Article] [PubMed]
Rosenbaum, PR, Rubin, DB The central role of the propensity score in observational studies for causal effects.. Biometrika. (1983). 70 41–55 [Article]
Borowitz, SM, Waggoner-Fountain, LA, Bass, EJ, Sledd, RM Adequacy of information transferred at resident sign-out (in-hospital handover of care): A prospective survey.. Qual Saf Health Care. (2008). 17 6–10 [Article] [PubMed]
Stahl, K, Palileo, A, Schulman, CI, Wilson, K, Augenstein, J, Kiffin, C, McKenney, M Enhancing patient safety in the trauma/surgical intensive care unit.. J Trauma. (2009). 67 430–3; discussion 433–5 [Article] [PubMed]
Foster, S, Manser, T The effects of patient handoff characteristics on subsequent care: A systematic review and areas for future research.. Acad Med. (2012). 87 1105–24 [Article] [PubMed]
Haynes, AB, Weiser, TG, Berry, WR, Lipsitz, SR, Breizat, AH, Dellinger, EP, Herbosa, T, Joseph, S, Kibatala, PL, Lapitan, MC, Merry, AF, Moorthy, K, Reznick, RK, Taylor, B, Gawande, AA Safe Surgery Saves Lives Study Group, A surgical safety checklist to reduce morbidity and mortality in a global population.. N Engl J Med. (2009). 360 491–9 [Article] [PubMed]
Kalkman, CJ Handover in the perioperative care process.. Curr Opin Anaesthesiol. (2010). 23 749–53 [Article] [PubMed]
Nagpal, K, Vats, A, Ahmed, K, Vincent, C, Moorthy, K An evaluation of information transfer through the continuum of surgical care: A feasibility study.. Ann Surg. (2010). 252 402–7 [Article] [PubMed]
Cooper, JB, Long, CD, Newbower, RS, Philip, JH Critical incidents associated with intraoperative exchanges of anesthesia personnel.. Anesthesiology. (1982). 56 456–61 [Article] [PubMed]
Cooper, JB Do short breaks increase or decrease anesthetic risk?. J Clin Anesth. (1989). 1 228–31 [Article] [PubMed]
Karakaya, A, Moerman, AT, Peperstraete, H, François, K, Wouters, PF, de Hert, SG Implementation of a structured information transfer checklist improves postoperative data transfer after congenital cardiac surgery.. Eur J Anaesthesiol. (2013). 30 764–9 [Article] [PubMed]
Salzwedel, C, Bartz, HJ, Kühnelt, I, Appel, D, Haupt, O, Maisch, S, Schmidt, GN The effect of a checklist on the quality of post-anaesthesia patient handover: A randomized controlled trial.. Int J Qual Health Care. (2013). 25 176–81 [Article] [PubMed]
Landrigan, CP, Rothschild, JM, Cronin, JW, Kaushal, R, Burdick, E, Katz, JT, Lilly, CM, Stone, PH, Lockley, SW, Bates, DW, Czeisler, CA Effect of reducing interns’ work hours on serious medical errors in intensive care units.. N Engl J Med. (2004). 351 1838–48 [Article] [PubMed]
Lockley, SW, Cronin, JW, Evans, EE, Cade, BE, Lee, CJ, Landrigan, CP, Rothschild, JM, Katz, JT, Lilly, CM, Stone, PH, Aeschbach, D, Czeisler, CA Harvard Work Hours, Health and Safety Group, Effect of reducing interns’ weekly work hours on sleep and attentional failures.. N Engl J Med. (2004). 351 1829–37 [Article] [PubMed]
Epstein, RM Morbidity and mortality from anesthesia: A continuing problem.. Anesthesiology. (1978). 49 388–9 [Article] [PubMed]
Descriptions of Individual In-hospital Surgical Mortality/Morbidities
Multivariable Association between Number of Anesthesia Handovers and the Collapsed Composite of In-hospital Mortality/Morbidity (N = 135,810)
Demographics Baseline and Intraoperative Characteristics for Each Propensity Score-matched Subset
Fig. 1.
Odds ratios of having any in-hospital mortality/morbidities (collapsed composite), each specific individual component of the composite (individual component), and common “global” odds ratio across the individual components (common effect) for each increase in the total number of anesthesia handovers. *CIs for the individual components were 99.3%, adjusted using the Bonferroni correction.
Odds ratios of having any in-hospital mortality/morbidities (collapsed composite), each specific individual component of the composite (individual component), and common “global” odds ratio across the individual components (common effect) for each increase in the total number of anesthesia handovers. *CIs for the individual components were 99.3%, adjusted using the Bonferroni correction.
Fig. 1.
Odds ratios of having any in-hospital mortality/morbidities (collapsed composite), each specific individual component of the composite (individual component), and common “global” odds ratio across the individual components (common effect) for each increase in the total number of anesthesia handovers. *CIs for the individual components were 99.3%, adjusted using the Bonferroni correction.
×
Fig. 2.
Odds ratios of having any in-hospital mortality/morbidities for (A) each level of attending anesthesiologist handovers versus no attending anesthesiologist handovers; and (B) each level of medical-directed provider handovers (including residents, fellows, certified registered nurse anesthetists, and student registered nurse anesthetists) versus no medically directed provider handovers. We adjusted for age, sex, race, American Society of Anesthesiologists physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure. *The significance criterion for each analysis was P < 0.006 (i.e., 0.05/8, a total of eight secondary analyses, Bonferroni correction). Multiple comparisons for each analysis were further adjusted using the Bonferroni correction; thus, 99.8% CIs were estimated (significance criterion: 0.006/3, three comparisons for each analysis).
Odds ratios of having any in-hospital mortality/morbidities for (A) each level of attending anesthesiologist handovers versus no attending anesthesiologist handovers; and (B) each level of medical-directed provider handovers (including residents, fellows, certified registered nurse anesthetists, and student registered nurse anesthetists) versus no medically directed provider handovers. We adjusted for age, sex, race, American Society of Anesthesiologists physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure. *The significance criterion for each analysis was P < 0.006 (i.e., 0.05/8, a total of eight secondary analyses, Bonferroni correction). Multiple comparisons for each analysis were further adjusted using the Bonferroni correction; thus, 99.8% CIs were estimated (significance criterion: 0.006/3, three comparisons for each analysis).
Fig. 2.
Odds ratios of having any in-hospital mortality/morbidities for (A) each level of attending anesthesiologist handovers versus no attending anesthesiologist handovers; and (B) each level of medical-directed provider handovers (including residents, fellows, certified registered nurse anesthetists, and student registered nurse anesthetists) versus no medically directed provider handovers. We adjusted for age, sex, race, American Society of Anesthesiologists physical status, start time of surgery, duration of surgery, and principal diagnosis and procedure. *The significance criterion for each analysis was P < 0.006 (i.e., 0.05/8, a total of eight secondary analyses, Bonferroni correction). Multiple comparisons for each analysis were further adjusted using the Bonferroni correction; thus, 99.8% CIs were estimated (significance criterion: 0.006/3, three comparisons for each analysis).
×
Table 1.
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers×
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers
Table 1.
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers
Demographics Baseline and Intraoperative Characteristics by Total Number of Anesthesia Handovers×
×
Table 2.
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers×
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers
Table 2.
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers
Incidence of the Collapsed Composite In-hospital Surgical Mortality/Morbidities and the Individual Components for Each Number of Handovers×
×
Table 3.
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching×
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching
Table 3.
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching
Sensitivity Analysis 1: Comparisons with No Handovers on Composite In-hospital Surgical Mortality/Morbidities Using Propensity Score Matching×
×
Table 4.
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets×
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets
Table 4.
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets
Exploratory Analysis: Association between Total Number of Intraoperative Anesthesia Handovers and Collapsed Composite In-hospital Surgical Mortality/Morbidities for Patients in Various Subsets×
×