Abstract
Background:
Provider handoffs have been identified as sources of deficiencies in care. We sought to determine whether intra-department attending-provider handoffs during a pediatric emergency department (ED) encounter were associated with return ED visits.
Methods:
We analyzed ED encounters for children < 21 years of age discharged from a single ED from 1/2013 through 2/2017. We classified an encounter as having a handoff when the initial attending and discharging attending differed. Our primary outcome was a revisit within 72 hours resulting in hospitalization. Our secondary outcomes were any revisit within 72 hours and revisits within 72 hours resulting in hospitalization with potential deficiencies in care. We compared outcome rates for ED encounters with and without provider handoffs, both with and without adjustment for demographic, clinical and visit characteristics including ED length of stay. We utilized quality improvement data to identify return visits with potential deficiencies in care.
Results:
Of the 177,350 eligible ED encounters, 1,961 (1.1%) had a revisit resulting in hospitalization and 6,821 (3.9%) had any revisit. In unadjusted analyses, handoffs were associated with an increased likelihood of a return ED visit resulting in hospitalization [Odds ratio (OR): 1.54, 95% confidence interval (CI) 1.33, 1.79], or any return ED visit (OR 1.21, 95% CI 1.11, 1.32). However, after adjustment, provider handoffs were not associated with return ED visits resulting in hospitalization (OR 0.96, 95% CI 0.8, 1.13) or any return ED visits (OR 1.03, 95% CI 0.88, 1.20). Among return ED visits resulting in hospitalization, potential deficiencies in care occurred no more commonly among ED encounters with handoffs (OR 1.10, 95% CI 0.52, 2.34).
Conclusions:
Provider handoffs in a pediatric ED did not increase the risk of return ED visits or return ED visits with deficiencies in care after adjustment for demographic, clinical and visit factors.
INTRODUCTION
Handoffs occur when one provider transitions the responsibility for a patient’s care to another provider.1 The Department of Defense and the Joint Commission on Accreditation of Healthcare Organizations identify communication and patient handoffs as potential threats to patient safety.2 In other settings, patient handoffs have been associated with uncertainty in the care plan, deficiencies in care, and serious medical errors.3,4 Strategies to improve patient handoffs have resulted in significant reductions in medical errors.5 While it is known that handoffs in the emergency department (ED) occur more frequently than in other parts of the hospital,6 the impact of handoffs on pediatric care in the ED is uncertain.1
Because handoffs are common in the ED, it is important for clinicians and policy makers to understand how provider handoffs impact the quality and safety of care provided to children in the ED. Return pediatric ED visits represent one quality benchmark for the ED care of children.7–9 Return ED visits may occur due to progression of a child’s underlying disease, an inadequate outpatient management plan, lack of a clear follow-up plan, unclear instructions for outpatient management, or unclear reasons to return to the ED for care.10 In the fast-paced, high-risk setting of the pediatric ED, handoffs between providers could magnify these factors. Through these mechanisms, we hypothesized that handoffs could contribute to unscheduled return visits in children discharged from the ED.
Therefore, we evaluated the impact of intra-departmental, attending-level handoffs on return ED visits in a cohort of children evaluated in a large, urban pediatric ED.
PATIENTS AND METHODS:
Study Design Setting
We performed a retrospective cohort analysis in a single ED between January 2013 and February 2017 following the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.11 The study period was selected to maximize the number of ED encounters where the exposure could be accurately abstracted from the data warehouse. The study center is a quaternary care pediatric hospital and a Level 1 Pediatric Trauma Center with an annual volume of approximately 60,000 patients. This study was approved by the Institutional Review Board with a waiver of informed consent.
At the study ED, handoffs occur independently for attending and trainee level providers because the shift start and end times are asynchronous. Prior to July, 2016, the study ED had no formal written or verbal handoff process. On a rolling basis from July-December 2016, attending physician and nursing providers underwent training for a formalized verbal handoff using the iPASSSM system.5 Training consisted of a 90 minute training session followed by structured feedback on observed handoffs by clinician champions of the handoff effort. At no point during the study period did the clinicians utilize a formal written handoff tool beyond the attending ED clinical note, which is expected to be up to date at the time of handoff.
At the study ED, only the attending physician can make the disposition decision. Discharge teaching with the patient and family is performed by either the trainee or attending physician. There is no formal process for arranging follow-up with outpatient providers after ED discharge, but clinicians can communicate with primary care providers by phone as needed. Primary care providers receive an electronic copy of the ED record once completed (required within 48 hours of ED discharge).
Selection of Participants
We included ED encounters for patients < 21 years of age at the time of ED arrival who were discharged to home from the ED. Encounters were excluded for missing disposition, missing diagnosis, or missing attending assignment. We excluded ED encounters for primary mental health diagnoses because prior quality improvement work within our ED has shown that disposition to home or mental health facility often is not accurately documented for these children. We excluded encounters with an ED length of stay longer than 24 hours, as these patients are frequently boarders receiving inpatient level of care due to inpatient capacity issues.
Interventions
The primary exposure in this study was a handoff between attending level providers during the initial ED encounter. We defined a handoff for a visit if the patient’s initial attending provider was different than the attending provider at the time of discharge (as documented in the hospital data warehouse). A change in attending providers is documented in the data warehouse when a new attending assigns themselves to a patient. It is policy in our division that the attending provider assigned to a patient must be present in the ED at the time of ED departure. We examined only attending-level handoff in these analyses, as attending providers supervise trainees and mid-level providers, examine all patients, and make final patient disposition decisions.
Measurements
We abstracted the following covariates from administrative data: patient age at time of ED encounter, sex (male or female), self/parent reported race (white, black, Asian, other) and ethnicity (Hispanic or non-Hispanic), primary language spoken (English or non-English), insurance (public or private), time of check-in, day of week, Emergency Severity Index (ESI) triage level, ED length of stay, ED census at the time of check-in, trainee involvement (defined as present if a student, resident, fellow or physician assistant evaluated the patient), performance of labs or imaging, attending provider and the primary diagnosis for the encounter. ED length of stay was defined from the time of ED check-in to the time of ED departure. ED census was defined as the total number of ED patients in the ED at the time of the patient’s check-in, including those in the waiting room. Laboratory testing included any blood, urine, or other specimen sent to the hospital clinical laboratory or analyzed in the ED at the point of care. Imaging included any x-ray, ultrasound, computed tomography, magnetic resonance imaging, or fluoroscopy. Point-of-care ultrasound performed by ED providers was not included as an imaging study as these examinations did not routinely guide care during the earlier portion of the study period.
Patients who had both public and private insurance (for example with secondary insurance) were categorized as having private insurance. Time of the initial ED encounter was categorized into the time periods: arrival from 8:00 AM - 3:59 PM (day shift), 4:00 PM -11:59 PM (evening shift) and midnight - 7:59 AM (overnight shift). We categorized day of week of check-in as weekend (Saturday-Sunday) or weekday (Monday-Friday). Diagnoses for encounters were grouped according to the top level of the multi-level Clinical Classifications Software grouping system.12–14
Outcomes
Our primary outcome was a return ED visit resulting in hospitalization within 72 hours of check-out from the index ED visit. Our two secondary outcomes were: 1) any return ED visit within 72 hours of the index ED visit; and 2) any return ED visit resulting in hospitalization in which a clinician identified a potential deficiency in care. We identified primary and secondary outcomes when another BCH ED triage occurred within 72 hours of the documented discharge time from the index encounter. For ED encounters with a return visit within 72 hours, we abstracted the ED disposition for the second encounter (discharge or admit). Patients who were transferred on the return ED visit were excluded from this analysis.
To identify potential deficiencies in care, we leveraged data from an ongoing departmental quality improvement (QI) initiative. As part of this QI initiative, all return ED encounters resulting in hospital admission are reviewed by both the treating attending provider as well as a randomly assigned peer attending physician. For each eligible encounter, clinicians are asked to choose one of the following reasons for the return ED visit resulting in hospital admission: progression of disease, patient/family non-compliance, lack of an alternative follow-up plans, potential deficiency in care, or other. If either reviewer selects a potential deficiency in care, the case is reviewed by a third senior physician. We defined a return ED visit resulting in hospitalization to have been due to a potential deficiency in care if any two reviewers felt a deficiency occurred. If a senior review was not available and the primary physician and peer-reviewer differed in their analysis, we classified a return visit as a potential deficiency in care if identified as such by the peer reviewer.
Analysis
For our primary analysis, we selected encounters with an ED length of stay < 8 hours in order to remove encounters for children who were boarding in the ED when inpatient beds were not available, and whose visits may therefore not be representative of usual ED care. However, because the effect of handoffs may be magnified for prolonged visits, we performed a secondary analysis of our primary and secondary outcomes, among encounters with ED lengths of stay up to 24 hours.
Our unit of analysis was the ED encounter. Individual patients could be included more than once if they had more than one eligible ED encounter. For continuous variables, we reported medians and interquartile ranges (IQRs) and compared difference between groups with the Wilcoxon Rank-Sum test. For ordinal and nominal variables, we calculated proportions and 95% confidence intervals and compared differences in proportions between groups with chi-square tests. Differences were reported as mean differences and proportional differences.
We calculated rates of our primary and secondary outcomes using bivariate analyses for encounters with handoffs compared to encounters without handoffs. Given the high likelihood that there are common factors leading to both handoffs and return visits, we created two mixed-effects models adjusting for potential confounders. Based on our conceptual framework of what factors might impact both handoffs and return ED visit rates, we decided a priori that ED length of stay would likely be strongly associated with both handoffs and return visits. We first evaluated the relationship between ED length of stay and both handoff status and return ED visits resulting in hospitalization. We next performed two adjusted models to account for potential confounders. In model 1 we adjusted only for ED length of stay. In model 2 we adjusted for all clinically relevant variables of interest (age, gender, race, ethnicity, primary language spoken, insurance type, shift of check-in, year of visit, ESI triage level, ED length of stay, ED volume, trainee involvement, performance of labs or imaging and final diagnosis). These variables were selected as we felt they might be associated with the likelihood of a patient handoff and/or a return ED visit resulting in hospitalization. Adjustment for the presence of a formal handoff process was accounted for in our models by adjusting for year of ED visit as the formal handoff process was not completely rolled out until the last included year of our analysis. Both models included a random intercept term for attending physician to account for within-physician correlation in care patterns. To assess for potential effect modification by ED length of stay we performed a post-hoc analysis of return visits stratified by hour of ED length of stay for the initial ED encounter. Finally, we assessed whether potential deficiencies in care were more common among ED revisits resulting in hospital admission where a handoff occurred, compared to those where it did not.
Outputs and analyses were generated using SAS Version 9.0 (SAS Institute Inc. Cary, NC).
RESULTS
Of the 252,419 ED encounters, 177,350 met all necessary inclusion and exclusion criteria for our primary analysis (Figure 1). Among included encounters with an ED length of stay < 8 hours, median patient age was 5.3 years (IQR 1.9-11.5 years) and 94,011 (53.0%) of visits were by males. Median ED length of stay was 160.1 minutes (IQR 103.2-238.0). Handoffs between attending providers occurred in 12,098 (6.8%, 95% CI 6.7, 6.9%) of ED encounters. Among the 177,350 ED encounters resulting in ED discharge, 6,821 (3.9%, 95%CI 3.8, 3.9%) resulted in a return ED visit, of which 1,961 encounters (1.1%, 95% CI 1.1, 1.2%) resulted in a hospitalization including 84 intensive care unit admissions (0.05%, 95% CI 0.04, 0.06%).
Figure 1:
Study Enrollment of Patient ED Encounters and ED Revisits
Abbreviation: ED = Emergency Department.
Encounters with handoffs between attending providers were more likely to occur for ED encounters with any of the following: older age, male sex, white race, non-Hispanic ethnicity, English-speaking, private insurance, higher triage scores (ESI), weekday presentation, longer ED length of stay, and trainee participation in care (Table 1). The most common discharge diagnosis categories associated with attending provider handoffs included: 1) diseases of the digestive system; 2) symptoms, signs and ill-defined conditions; and 3) endocrine, nutritional, metabolic diseases and immunity disorders. ED length of stay was found to be a strongly associated with both handoffs [odds ratio (OR) 1.61, 95%CI 1.59, 1.63 for each additional hour of ED stay], as well as return ED visits resulting in hospital admission (OR 1.38, 95%CI 1.34, 1.42 for each additional hour of ED stay).
Table 1:
ED Encounters Characteristics By Handoff Status
| ED Visits Where Handoffs Occurred N = 12,098 (6.8%) | ED Visits Where Handoffs Did Not Occur N = 165,252 (93.2%) | Difference (95% CI) | |
|---|---|---|---|
| Patient Demographic Factors at Visit | |||
| Age in months, median (IQR) | 75.7 (27.5-153.8) | 62.7 (22.4-136.5) | 13.0 (10.6, 15.5) |
| Male, N (%) | 6,276/12,098 (51.9%) | 87,735/165,240 (53.1%) | −1.2 % (−2.1, 0.3%) |
| Race, N (%) | |||
| White | 4,797/11,364 (42.2%) | 54,748/155,977 (35.1%) | 7.1% (6.2, 8.0%) |
| Black | 2,086 (18.4%) | 33,203 (21.3%) | −2.9% (−3.6, −2.1%) |
| Asian | 413 (3.6%) | 6,081 (3.9%) | −0.3% (−0.6, −0.0) |
| Other | 4,068 (35.8%) | 61,945 (39.7%) | −3.9 (4.8, −3.0) |
| Hispanic Ethnicity, N (%) | 3,246/10,910 (29.8%) | 49,813/151,882 (32.8%) | −3.0 (−3.9, −2.1%) |
| Non-English Primary Language, N (%) | 2,199/11,616 (18.9%) | 33,032/159,607 (20.7%) | −1.8% (−2.5, −1.0%) |
| Insurance Type, N (%) | |||
| Public | 5,537/11,939 (46.4%) | 84,714/163,351 (51.9%) | −5.5% (−6.4, −4.6) |
| Private/Commercial | 6,402 (53.6%) | 78,637 (48.1%) | 5.5 (4.6, 6.4%) |
| Encounter-Level Factors | |||
| Time of Check-In, N (%) | |||
| 8:00 AM – 3: 59 PM | 4,838/12,098 (40.0%) | 63,538/165,252 (38.5%) | 1.5% (0.6, 2.4%) |
| 4:00 PM – 11:59PM | 5,578 (46.1%) | 82,610 (50.0%) | −3.9% (−4.8, −3.0%) |
| 12:00 AM – 7:59 AM | 1,682 (13.9%) | 19,104 (11.6%) | 2.4% (1.7, 2.9%) |
| Day of Visit, N (%) | |||
| Weekday | 8,785/12,098 (72.6%) | 112,908/165,252 (68.3%) | 4.3% (3.5, 5.2%) |
| Weekend | 3,313 (27.4%) | 52,344 (31.7%) | −4.3% (−5.2, −3.5) |
| ESI Level, N (%) | |||
| 5 | 309/12,031 (2.6%) | 7,716/164,127 (4.7%) | −2.1% (−2.4, −1.8%) |
| 4 | 2,875 (23.9%) | 58,630 (35.7%) | −11.8% (−12.6, −11.0%) |
| 3 | 7,041 (58.5%) | 83,755 (51.0%) | 7.5% (6.6, 8.4%) |
| 2 | 1,787 (14.9%) | 13,947 (8.5%) | 6.4% (5.8, 7.1%) |
| 1 | 19 (0.2%) | 79 (0.1%) | 0.1% (0.03, 0.20%) |
| ED Length of Stay in Minutes, median (IQR) | 258.9 (171.0-350.8) | 155.4 (101.0-229.0) | 103.5 (100.8, 106.1) |
| Trainee Involved in Care, N (%) | 7,062/12,098 (58.4%) | 72,946/165,252 (44.1%) | 14.2% (13.3, 15.1%) |
| ED Disposition Diagnosis, N (%) | |||
| Infectious and parasitic diseases | 744/12,098 (6.2%) | 13,456/165,252 (8.1%) | −1.9% (−2.3, −1.4%) |
| Neoplasms | 174 (1.4%) | 1,387 (0.8%) | 0.6% (0.4, 0.8%) |
| Endocrine, nutritional metabolic disease and immunity disorders | 451 (3.7%) | 4,099 (2.5%) | 1.2% (0.9, 1.6%) |
| Diseases of the nervous system and sense organs | 1,256 (10.4%) | 18,812 (11.4%) | −1.0% (−1.6, −0.4%) |
| Diseases of the circulatory system | 152 (1.3%) | 1,731 (1.1%) | 0.2% (0.0, 0.4%) |
| Diseases of the respiratory system | 2,042 (16.9%) | 32,399 (19.6%) | −2.7% (−3.4, −2.0%) |
| Diseases of the digestive system | 1,347 (11.1%) | 16,319 (9.9%) | 1.2% (0.6, 1.8%) |
| Diseases of genitourinary system | 401 (3.3%) | 4,657 (2.8%) | 0.5% (0.2, 0.8%) |
| Diseases of the skin and subcutaneous tissue | 296 (2.5%) | 5,448 (3.3%) | −0.8% (−1.0, −0.5%) |
| Diseases of the musculoskeletal system and connective connective tissue | 389 (3.2%) | 4,720 (2.9%) | 0.3% (−0.02, 0.6%) |
| Symptoms, signs and ill-defined conditions | 1,586 (13.1%) | 18,795 (11.4%) | 1.7% (1.1, 2.3%) |
| Injury and poisonings | 2,817 (23.3%) | 38,323 (23.2%) | 0.1% (−0.7, 0.9%) |
| Residual Codes | 75 (0.6%) | 992 (0.6%) | 0% (−0.2, 0.1%) |
| Diseases of the Blood and Blood Forming Organs | 164 (1.4%) | 1,897 (1.2%) | 0.2% (0.0, 0.4%) |
| Congenital Anomalies | 125 (1.0%) | 1,159 (0.7%) | 0.3% (0.1, 0.5% |
| Complications of Pregnancy and Childbirth | 9 (0.1%) | 67 (0.0%) | 0.1% (0.1, 0.2%) |
| Certain Conditions Originating in Perinatal Period | 70 (0.6%) | 991 (0.6%) | 0% (−0.2, 0.1%) |
| Imaging Performed, N (%) | 1,734/12,098 (14.3%) | 14,175/165,252 (8.6%) | 5.8% (5.1, 6.4%) |
| Labs Performed, N (%) | 4,576/12,098 (37.8%) | 43,312/165,252 (26.2%) | 11.6% (10.7, 12.5%) |
| Patients in the ED and Waiting Room at Time of Patient Check In, median (IQR) | 33 (23-42) | 33 (23-43) | 0 (0, 0) |
In unadjusted analyses, encounters with handoffs were associated with return ED visits resulting in hospitalization (Table 2). After adjusting for ED length of stay, encounters with a handoff were no longer associated with return ED visits resulting in admission (Model 1). A similar result was seen after adjustment for all of the demographic, clinical and visit level factors (Model 2). For the secondary outcome of any return visits, ED encounters with handoffs were associated with any ED return visits, but these differences again were no longer significant after adjustment (Models 1 and 2).
Table 2:
Unadjusted and adjusted odds of return Emergency Department visits by handoff status
| Unadjusted* (95%CI) | Adjusted Model 1* (95%CI) | Adjusted Model 2* (95%CI) | |
|---|---|---|---|
| Primary Cohort with ED Length of Stay < 8 Hours | |||
| Return ED Visits Resulting in Hospital Admission | 1.54 (1.33, 1.79) | 0.94 (0.81, 1.10) | 0.96 (0.81, 1.13) |
| Any Return ED Visit | 1.21 (1.11, 1.32) | 0.97 (0.89, 1.07) | 1.00 (0.90, 1.10) |
| Secondary Cohort with ED Length of Stay < 24 Hours | |||
| Return ED Visit Resulting in Hospital Admission | 1.79 (1.58, 2.02) | 1.08 (0.93, 1.25) | 1.03 (0.88, 1.20) |
| Any Return ED Visit | 1.43 (1.32, 1.53) | 1.05 (0.96, 1.14) | 1.02 (0.93, 1.12) |
Odds ratios represent the odds of a return ED visit for a patient whose care was handed off between attending providers in the Emergency Department relative to a patient whose care was not handed off. An odds ratio > 1 represents and increased odds of a return Emergency Department visit.
Model 1: Adjusted only for ED length of stay and clustered by attending provider
Model 1: Adjusted for age, race, ethnicity, year of visit, weekend vs. weekday, ESI category, year of visit, shift-type, insurance status, sex, whether labs or imaging were performed, co-management with trainee or diagnosis, ED length of stay, and clustered by attending provider
Data on potential deficiencies in care were available for 1,811/1,961 (92.4%) of the return ED visits resulting in hospitalization. Potential deficiencies in care were identified in 74/1,811 (4.1%) of these visits. Among return ED visits resulting in hospitalization, potential deficiencies in care occurred no more commonly among ED encounters with handoffs 8/180 (4.4%, 95% CI 2.0, 8.1) than ED encounters without handoffs 66/1631 (4.1%, 95% CI 3.2, 5.1; OR 1.10, 95% CI 0.52, 2.34).
At no stratum of ED length of stay was the association of handoffs with return ED visits resulting in hospitalization statistically significant, suggesting that ED length of stay was acting as a confounder and not an effect modifier (Table 3).
Table 3:
Risk of Return ED Visits Stratified by ED Length of Stay
| ED Length of Stay at Index ED Encounter, Minutes | Rate of Return ED Visits Resulting Hospital Admission for Encounters with Handoff, Proportion (%) | Rate of Return ED Visits Resulting in Hospital Admission for Encounters with No Handoff, Proportion (%) | Difference, Percent (95%CI) | Odds Ratio (95%CI) |
|---|---|---|---|---|
| 0-59 | 0/277 (0.0%) | 24/12,630 (0.2%) | −0.2% (−0.3, −0.1%) | ** |
| 60-119 | 8/1,285 (0.6%) | 207/43,562 (0.5%) | 0.2% (−0.3, 0.6%) | 1.31 (0.64-2.66) |
| 120-179 | 10/1,738 (0.6%) | 353/42,249 (0.8%) | −0.3% (−0.6, 0.1%) | 0.68 (0.37-1.29) |
| 180-239 | 36/2,080 (1.7%) | 414/30,059 (1.4%) | 0.4% (−0.2, 0.9%) | 1.26 (0.89-1.78) |
| 240-299 | 41/2,103 (2.0%) | 328/18,501 (1.8%) | 0.2% (−0.4, 0.8%) | 1.10 (0.79-1.53) |
| 300-359 | 44/1,862 (2.4%) | 234/10,366 (2.3%) | 0.1% (−0.6, 0.9%) | 1.05 (0.76-1.45) |
| 360-419 | 33/1,544 (2.1%) | 136/5,285 (2.6%) | −0.4% (−1.3, 0.4%) | 0.83 (0.56-1.21) |
| 420-480 | 26/1,208 (2.2%) | 67/2,591 (2.6%) | −0.4% (−1.5, 0.6%) | 0.83 (0.52-1.31) |
Odds ratio could not be calculated as zero outcomes in one cell
DISCUSSION
In this single-center study of all ED encounters over a four year period, we found return ED visits resulting in hospitalization and overall return ED visits within 72 hours occurred more frequently in ED encounters where an attending-provider handoff occurred. However, after adjustment for ED length of stay, and other important demographic, clinical and visit characteristics the effect of handoffs was no longer significant. Potential deficiencies in care were identified in only a minority of ED return visits resulting in hospitalization and were not associated with provider handoffs. Taken together, our findings suggest that handoffs are a marker of risk for return ED visits and subsequent hospital admission. However handoffs are not independently associated with return ED visits, return ED visits resulting in hospital admission, or return ED visits due to deficiencies in care. Rather, factors associated with ED visits where handoffs occur, particularly ED length of stay, increase both the patient’s risk of their care being handed off and of having a return ED visit.
Our study is the first to evaluate the association between handoffs and clinical outcomes for pediatric ED visits. Previous work has demonstrated that clinicians are frequently interrupted during ED handoffs and as a result, may omit critical information.15–18 Handoffs in the pediatric ED have also been noted to be an important cause of near-miss errors in a large 23 center study.19 Calls for improvement in handoffs have occurred across health-care settings.20,21 Both the American Academy of Pediatrics and American College of Emergency Physicians Section of Quality Improvement and Patient Safety released statements emphasizing the importance of standardized handoffs for children cared for in the ED.1,22 This sentiment has been echoed by various studies demonstrating the lack of standardized handoff processes in the ED setting.16,23 In clinical settings outside of the ED, standardized handoff processes have been shown to reduce medical errors and care failures.5,24–27
We selected ED visits resulting in admission as our primary outcome as an important, measureable marker of the quality of care that could be effected by handoffs. Important information lost during handoffs could result in misdiagnosis, inadequate outpatient follow-up plan, prescription errors, or parental discomfort with the outpatient follow-up plan. Therefore, even if patients admitted during return ED visits are not as ill as those admitted on their index visit,28 these return visits represent potential opportunities for improvement in patient care. Our secondary outcome, all return ED visits, captures both patients who may have returned unnecessarily to the ED and could have been avoided with better anticipatory guidance, as well as those who needed to return, but did not require hospital admission. Importantly, we saw no difference in rates of potential deficiencies in care between ED return visits resulting in admission after an initial encounter with or without an attending handoff.
Our data suggest that factors that increase the risk of a handoff also increase the likelihood of a return ED visit with or without hospitalization. Contributing factors may include diagnostic uncertainty, medical complexity, prolonged evaluation, and unusual presentations or conditions. These factors can lead to longer ED length of stay and an increased likelihood of a handoff between ED providers. Our study, limited to ED visits that resulted in discharge to home, evaluated only handoffs between ED attending providers with limited clinical outcomes (return ED visits with and without hospitalization). The effect of handoffs on clinical care in other patient populations (such as those admitted to the hospital during their index ED visit), between other provider types (such as between the ED and inpatient teams) and on other clinical outcomes (such as change in diagnoses or medication errors) may show important differences in clinical outcomes. Furthermore, in unadjusted analyses, ED encounters where handoffs occured were associated with higher rates of return ED visits resulting in hospitalization. Therefore, while handoffs were not independently associated with higher risk of return ED visits resulting in hospitalization after adjustment, structured handoffs and communication should remain an important priority in the pediatric ED. Providers who receive handoffs of care for ED patients must (or should?) ensure they convey the working diagnosis, clear follow-up plan, outpatient management plan, and reasons to return for care, while addressing all patient and parental concerns prior to discharge.
LIMITATIONS
Our study must be interpreted in the context of its limitations. First, this was a retrospective study. However, both our exposure and outcome were objectively captured in our dataset, and are not subject to recall bias. Second, this was a single-center study. Our results may not be generalizable to other EDs with different handoff and discharge practices or different revisit patterns. Furthermore, we could not determine if patients returned to other EDs for care, possibly resulting in underestimation of revisits.9,28 Third, handoffs were not standardized in our ED until the very end of the study period, and the quality of handoffs may have varied between attending providers, diluting the potential impact of high-quality structured ED handoff process. However, despite not having a standardized handoff for most of the study period, we still failed to see an effect on return ED visit rates. Fourth, we could not account for the impact of handoffs between other providers on ED revisit rates including trainees and nurses. However, in the study ED, attendings providers make all patient disposition decisions, and therefore the impact of other provider handoffs on this outcome is likely limited. Fifth, we were only able to categorize certain covariates as performed or not performed, including the performance of laboratory testing or diagnostic imaging. This categorization is unlikely to capture the entire complexity of the patient’s work-up. Last, ED revisits are only one marker of quality of ED care delivered. More comprehensive and studies of ED provider handoffs across multiple domains of outcome measures are needed to measure the full impact on other important clinical outcomes.
CONCLUSIONS
Pediatric ED visits in which handoffs occurred had higher rates of return ED visits. However, after adjustment for important clinical and demographic factors, the impact of attending-provider handoffs was no longer significant. Our findings suggest that ED handoffs do not result in increased rates of return ED visits resulting in hospitalization, but are instead a marker of other factors that predict ED return visits requiring hospital admission. Providers should be aware of the heightened risk of return visits in handed-off patients, especially those with long ED lengths of stay.
Financial Support:
Dr. Lyons was supported by a training grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (National Institutes of Health) Childhood (T32HD040128).
Abbreviations:
- CI
Confidence Interval
- ED
Emergency Department
- ESI
Emergency Severity Index
- IQR
Interquartile Range
- SD
Standard Deviation
Footnotes
Potential Conflicts of Interest: The authors have no conflicts of interest relevant to this article.
Conflicts of Interest Disclosure: Authors TL, KM, CP, LN and AF report no conflicts of interest.
Presentations: This work has been accepted for presentation as a poster at the 2018 Pediatric Academic Societies (PAS) Meeting May 5, 2018 in Toronto, Canada.
REFERENCES
- 1.AMERICAN ACADEMY OF PEDIATRICS Committee on Pediatric Emergency Medicine, AMERICAN COLLEGE OF EMERGENCY PHYSICIANS Pediatric Emergency Medicine Committee, EMERGENCY NURSES ASSOCIATION Pediatric Committee. Handoffs: Transitions of Care for Children in the Emergency Department. Pediatrics. 2016;138(5). [DOI] [PubMed] [Google Scholar]
- 2.The Joint Commission. Sentinel Event Alert: Inadequate Hand-Off Communication. Oakbrook Terrace, IL; 2017. https://www.jointcommission.org/assets/1/18/SEA_58_Hand_off_Comms_9_6_17_FINAL_(1).pdf. [PubMed] [Google Scholar]
- 3.Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. Consequences of inadequate sign-out for patient care. Arch Intern Med. 2008;168(16):1755–1760. [DOI] [PubMed] [Google Scholar]
- 4.Arora S, Patel P, Lahewala S, et al. Etiologies, Trends, and Predictors of 30-Day Readmission in Patients With Heart Failure. Am J Cardiol. 2016;73(5):407–409. [DOI] [PubMed] [Google Scholar]
- 5.Starmer AJ, Spector ND, Srivastava R, et al. Changes in medical errors after implementation of a handoff program. N Engl J Med. 2014;371(19):1803–1812. [DOI] [PubMed] [Google Scholar]
- 6.Pruitt CM, Liebelt EL. Enhancing patient safety in the pediatric emergency department: teams, communication, and lessons from crew resource management. Pediatr Emerg Care. 2010;26(12):942–948; quiz 949-951. [DOI] [PubMed] [Google Scholar]
- 7.Hung GR, Chalut D. A consensus-established set of important indicators of pediatric emergency department performance. Pediatr Emerg Care. 2008;24(1):9–15. [DOI] [PubMed] [Google Scholar]
- 8.Shaw KN, Ruddy RM, Gorelick MH. Pediatric emergency department directors’ benchmarking survey: fiscal year 2001. Pediatr Emerg Care. 2003;19(3):143–147. [DOI] [PubMed] [Google Scholar]
- 9.Michelson KA, Lyons TW, Bachur RG, Monuteaux MC, Finkelstein JA. Timing and Location of Emergency Department Revisits. Pediatrics. 2018;141(5). [DOI] [PubMed] [Google Scholar]
- 10.Rising KL, Padrez KA, O’Brien M, Hollander JE, Carr BG, Shea JA. Return visits to the emergency department: the patient perspective. Ann Emerg Med. 2015;65(4):377–386.e3. [DOI] [PubMed] [Google Scholar]
- 11.von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344–349. [DOI] [PubMed] [Google Scholar]
- 12.Duffy S, Elixhauser A, Sommers J. Diagnosis and Procedure Combinations in Hospital Inpatient Data. Healthc Cost Util Proj (HCUP 3) Res Note 5. 1996;96(0017). [Google Scholar]
- 13.Kindermann DR, Mutter RL, Houchens RL, Barrett ML, Pines JM. Emergency department transfers and transfer relationships in United States hospitals. Acad Emerg Med. 2015;22(2):157–165. [DOI] [PubMed] [Google Scholar]
- 14.Nandan AR, Bohnen JD, Chang DC, et al. The impact of major intraoperative adverse events on hospital readmissions. Am J Surg. 2017;213(1):10–17. [DOI] [PubMed] [Google Scholar]
- 15.Peterson SM, Gurses AP, Regan L. Resident to resident handoffs in the emergency department: an observational study. J Emerg Med. 2014;47(5):573–579. [DOI] [PubMed] [Google Scholar]
- 16.Sinha M, Shriki J, Salness R, Blackburn PA. Need for standardized sign-out in the emergency department: a survey of emergency medicine residency and pediatric emergency medicine fellowship program directors. Acad Emerg Med. 2007;14(2):192–196. [DOI] [PubMed] [Google Scholar]
- 17.Venkatesh AK, Curley D, Chang Y, Liu SW. Communication of Vital Signs at Emergency Department Handoff: Opportunities for Improvement. Ann Emerg Med. 2015;66(2):125–130. [DOI] [PubMed] [Google Scholar]
- 18.Maughan BC, Lei L, Cydulka RK. ED handoffs: observed practices and communication errors. Am J Emerg Med. 2011;29(5):502–511. [DOI] [PubMed] [Google Scholar]
- 19.Ruddy RM, Chamberlain JM, Mahajan PV, et al. Near misses and unsafe conditions reported in a Pediatric Emergency Research Network. BMJ Open. 2015;5(9):e007541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sehgal N Annual Perspective 2014: Handoffs and Transitions. AHRQ Patient Saf Netw. 2015. http://psnet.ahrq.gov/perspective.aspx?perspectiveID=170. Accessed December 9, 2016.
- 21.Dhingra KR, Elms A, Hobgood C. Reducing error in the emergency department: a call for standardization of the sign-out process. Ann Emerg Med. 2010;56(6):637–642. [DOI] [PubMed] [Google Scholar]
- 22.Cheung DS, Kelly JJ, Beach C, et al. Improving handoffs in the emergency department. Ann Emerg Med. 2010;55(2):171–180. [DOI] [PubMed] [Google Scholar]
- 23.Dubosh NM, Carney D, Fisher J, Tibbles CD. Implementation of an emergency department sign-out checklist improves transfer of information at shift change. J Emerg Med. 2014;47(5):580–585. [DOI] [PubMed] [Google Scholar]
- 24.Bigham MT, Logsdon TR, Manicone PE, et al. Decreasing handoff-related care failures in children’s hospitals. Pediatrics. 2014;134(2):e572–e579. [DOI] [PubMed] [Google Scholar]
- 25.Heilman JA, Flanigan M, Nelson A, Johnson T, Yarris LM. Adapting the I-PASS Handoff Program for Emergency Department Inter-Shift Handoffs. West J Emerg Med. 2016;17(6):756–761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gopwani PR, Brown KM, Quinn MJ, Dorosz EJ, Chamberlain JM. SOUND: a structured handoff tool improves patient handoffs in a pediatric emergency department. Pediatr Emerg Care. 2015;31(2):83–87. [DOI] [PubMed] [Google Scholar]
- 27.Starmer AJ, Spector ND, West DC, et al. Integrating Research, Quality Improvement, and Medical Education for Better Handoffs and Safer Care: Disseminating, Adapting, and Implementing the I-PASS Program. Jt Comm J Qual patient Saf. 2017;43(7):319–329. [DOI] [PubMed] [Google Scholar]
- 28.Sills MR, Macy ML, Kocher KE, Sabbatini AK. Return Visit Admissions May Not Indicate Quality of Emergency Department Care for Children. Acad Emerg Med. September 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]