594879
research-article2015
AJMXXX10.1177/1062860615594879American Journal of Medical QualityCheng et al
Article
Emergency Department Return Visits Resulting in Admission: Do They Reflect Quality of Care?
American Journal of Medical Quality 1–11 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1062860615594879 ajmq.sagepub.com
John Cheng, MD1,2, Amita Shroff, MD3, Naghma Khan, MD4,5, and Shabnam Jain, MD, MPH4,5
Abstract Prior studies have suggested that emergency department (ED) return visits resulting in admission may be a more robust quality indicator than all 72-hour returns. The objective was to evaluate factors that contribute to admission within 72 hours of ED discharge. Each return visit resulting in admission was independently reviewed by 3 physicians. Analysis was by descriptive statistics. Of 45 071 ED discharges, 4.1% returned within 72 hours; 0.96% returned for related reasons and were admitted to wards (91.2%), intensive care units (6.5%), or operating rooms (1.2%). Management was acceptable in 92.6%, suboptimal in 7.4%. Admissions were illness (94.9%), patient (1.6%), and physician related (3.5%). Almost all admissions within 72 hours after ED discharge are illness related, including all intensive care unit admissions and the majority of operating room admissions. Deficiencies in ED care are rarely the reason for admission on return. ED return visits resulting in admission may not be reflective of ED quality of care. Keywords emergency department, quality of care, return visit, pediatrics
Multiple studies have been published over the past 3 decades on unplanned returns to the emergency department (ED).1-11 It has been suggested to monitor 48- or 72-hour returns to the ED as a quality metric with the underlying premise that this cohort may identify potential deficiencies in initial medical management.3,4,12,13 Many previous studies have reported that the majority of unplanned returns to the ED are related to noncompliance with care (patient related) or progression of illness (illness related) rather than deficiencies in the initial medical management (physician related).14-17 Most studies thus far also have found that the majority of patients with return visits were discharged home on the subsequent visit. A recent study suggested that monitoring only returns that resulted in admission might have more value in identifying deficiencies in initial medical management.18 A few studies outside the United States also have reported that physicianrelated factors contribute to return visits that result in admission.5,6 The goal of this study is to evaluate factors that contribute to hospital admission within 72 hours of discharge from a pediatric ED in a US health care system, with a focus on physician-related parameters of care at the first visit.
Methods Setting The study ED is an academic, tertiary care, level one pediatric trauma center and a referral center for subspecialty care for the state and surrounding region. It serves an urban population and has an annual volume of more than 50 000, with an admission rate of about 15%. The ED is part of a large pediatric health care system that has 2 tertiary care sites and several urgent care sites that share an electronic medical record (EMR) and account for the vast majority of pediatric admissions in the metro area. A
1
Pediatric Emergency Medicine Associates, LLC, Atlanta, GA Children’s Healthcare of Atlanta at Scottish Rite Children’s Hospital, Atlanta, GA 3 Children’s Emergency Center at Gwinnett Medical Center, Lawrenceville, GA 4 Emory University, Atlanta, GA 5 Children’s Healthcare of Atlanta, Atlanta, GA 2
Corresponding Author: John Cheng, MD, Emergency Department, Children’s Healthcare of Atlanta at Scottish Rite Children’s Hospital, 1001 Johnson Ferry Road NE, Atlanta, GA 30342. Email: [email protected]
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
2
American Journal of Medical Quality
quality improvement (QI) initiative was implemented in which physicians reviewed the initial care received by patients who returned within 72 hours after discharge from the ED and were admitted to the hospital. This study was conducted after completion of the QI project. The physicians were blinded to the purpose of the study when completing their reviews.
Inclusion/Exclusion Criteria All patients presenting to the study ED from January 1 to December 31, 2011, who returned to any facility in the health care system within 72 hours (visit 2) after discharge from the ED (visit 1) and were admitted (including direct admissions) were included. Patients were considered admissions if they were not discharged from the ED. Dispositions at visit 2 included admission to the general ward, intensive care unit (ICU), or operating room (OR), transfer to another facility, or to the morgue. Patients who left without being seen (LWBS) or against medical advice (AMA) at visit 1 were excluded. Patients who returned after being admitted at visit 1 also were excluded to avoid any confounding factors from inpatient management.
Case Review Three physicians who are board certified in pediatric emergency medicine (PEM) or in pediatrics reviewed both ED visits 1 and 2 and, when pertinent, the inpatient record of visit 2. Reviewer 1 was the treating physician at visit 1. Reviewers 2 and 3 were physicians who had no contact with the patient at either visit 1 or 2. All 3 physicians reviewed the cases independently and completed a standardized Web-based form (Figure 1) regarding (1) the relatedness of the 2 visits, (2) the adequacy of care at visit 1 (overall medical management, written discharge instructions [DCI], and follow-up plans), (3) the transfer of care from one physician to another, and (4) the details at visit 2 (patient compliance with the original discharge plan, reason for return). All 3 reviewers had to select a reason for admission from a drop-down menu based on their review. An “other” category with a free-text option was included. The reasons for admission were categorized into 3 groups: illness related (progression of disease or callback for abnormal laboratory results), patient related (noncompliance), and physician related (deficiencies in medical management, written DCI, or follow-up plans, or missed significant radiologic findings). Majority opinion was used when responses were not unanimous. However, if even one reviewer classified the reason for admission as physician related, the authors reviewed the visit and reclassified it when necessary. This was done to ensure that no physician-related reasons for
admission were inadvertently overlooked during the initial review. Similarly, cases that had no majority opinion as to the reason for admission were reviewed and reclassified into one of the aforementioned 3 categories.
Statistics Proportions, means, and standard deviations were used to describe the results. The study institutional review board exempted this study as it was a QI initiative that did not affect patient management in real time.
Results In 2011, a total of 53 328 patients presented to the study ED; 38 patients LWBS, 3 left AMA, and 45 071 (84.5%) were discharged from the ED. Of these, 1829 (4.1%) returned within 72 hours: 8 LWBS, 474 were admitted, and the balance were discharged again. Of the 8 who LWBS at visit 2, 3 returned within 24 hours and were discharged; the rest did not return within 72 hours after visit 2. Of the 474 admitted at visit 2, a total of 40 of the visits were deemed unrelated to visit 1 (28 by unanimous and 12 by majority agreement) and were excluded from further analysis. In all 12 instances where one reviewer thought the visits were related, the reviewer felt that the admission was for illness-related, not physician-related, factors. Of the 45 071 ED patients who were discharged at the first visit, 434 (0.96%) returned and were admitted for a related condition. There were no deaths in this group. The average time to return was 34.7 (±18.4) hours (Figure 2). Agreement between reviewers regarding the reason for admission was unanimous in 291 (67.1%) cases; almost all of these were progression of illness. Of the remaining 143 cases, 134 (30.9%) had majority consensus and 9 had no consensus. In total, 98% of reviews had unanimous or majority agreement regarding the reason for admission. All 9 cases with no consensus among the 3 reviewers were reviewed by the authors and recategorized: 6 as illness related, 3 as physician related. Of the cases with majority agreement, 19 were categorized as illness related by 2 reviewers and physician related by the third reviewer. All of these were reviewed by the authors; 4 of the 19 were recategorized as physician related. Figure 3 shows details of the physician reviews. Of the 434 patients with returns for a related reason, medical management at visit 1 was acceptable in 402 (92.6%). Reviewers deemed the care suboptimal in 32/434 (7.4%) of cases. The main reasons for admissions for the 434 patients are shown in Figure 3. Of note, all abnormal laboratory results were positive cultures; almost all were blood cultures. Written DCI were acceptable in 424 (97.7%); follow-up plans were acceptable in 425 (97.9%).
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
3
Cheng et al
Reviewer Name
physician reviewer
Case Number
de-identified case number
Are the chief complaints and final diagnosis(es) from the both visits related?
Yes No
VISIT 1 Was there transfer of care?
Yes No
Was transfer of care documented?
Yes No
Was the final diagnosis related to the chief complaint?
Yes No
In your opinion, was the MANAGEMENT acceptable?
Yes No (comment below)
The management was suboptimal because:
free text box
In your opinion were the DISCHARGE INSTRUCTIONS acceptable?
Yes No (comment below)
The discharge instructions were suboptimal because:
free text box
In your opinion was the FOLLOW UP PLAN acceptable?
Yes No (comment below)
The follow up plan was suboptimal because:
free text box
VISIT 2 Was there any documented non-compliance with the plan from the first visit?
Yes No (comment below)
Details of non-compliance:
free text box
Reason for admission:
Progression of illness. Non-compliant. Planned return. Callback for abnormal lab results. Callback for X-ray over-read. Deficiency in initial management. Deficiency in discharge instructions and/or follow up plan. Re-referred in by outside provider. Other: specify below.
Other reason for admission:
free text box
Other factors that contributed to the return visit:
free text box
LESSONS LEARNED Learning points:
None. Pay closer attention to vital signs. Pay closer attention to other caregivers’ notes (RT/RN/tech). Document reassessments after interventions. Follow up on ALL lab and radiology results. Give sufficient detail during transfer of care. Stay objective, especially when patients come “pre-diagnosed.” Consultants are only consultants: you supersede them. Supervise residents completely in ALL aspects of patient care. Other: comment below.
Other learning points:
free text box
Should this be presented as a case conference?
Yes No
Figure 1. Case review questionnaire. Shown above is the Web-based form used by physicians to review the charts for visits 1 and 2. The answers on the right-hand side were drop-down boxes that have been expanded for this figure.
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
4
American Journal of Medical Quality
Figure 2. Flow diagram of emergency department (ED) patients admitted within 72 hours of initial discharge. ICU, intensive care unit. OR, operating room.
Overall, 412/434 (94.9%) of patients were admitted for illness-related factors at visit 2, and 7/434 (1.6%) for patient-related factors. Physician-related factors accounted for only 15/434 (3.5%) of the admissions. Two scenarios that may seem contradictory are discussed in the following sections: patients with suboptimal management at visit 1 who had illness-related reasons for admission and patients with acceptable management at visit 1 who had physician-related reasons for admission.
Patients With Suboptimal Management at Visit 1 and an Illness-Related Reason for Admission at Visit 2 There were 32 patients classified as having suboptimal management at visit 1; 20 were admitted for illnessrelated reasons. One patient’s care was deemed suboptimal with regard to the DCI and follow-up plan; however, the patient was admitted for positive cultures. Of the remaining 19, a total of 12 had suboptimal care related to DCI (when to return, home medications, or follow-up plans). For one patient, suboptimal care was related to insufficient physician documentation about the patient’s status and response to treatment
(although improvement in patient condition was documented by ED staff). The remaining 6 patients had a “deficiency” in their workup related to either a longstanding primary complaint that did not warrant admission at visit 1 or a secondary complaint that may not have warranted evaluation at visit 1 but did progress and was the reason for admission at visit 2. In other words, for these 19 patients, while there was room for improvement in the management at visit 1 (ie, suboptimal management), the reviewers felt that the reason for admission at visit 2 was illness related and not physician related.
Patients With Acceptable Management at Visit 1 and a Physician-Related Reason for Admission at Visit 2 All of these were missed radiological findings. Out of these 6 patients, 2 cases were deemed physician related because the radiologist’s reading was available to the ED physician prior to discharge. In the other 4 cases, medical management was deemed acceptable because the official reading was available only after the patient was discharged.
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
5
Cheng et al
Admissions at 72-hr Return Visit: 434
Visit 1
Acceptable Medical Management: 402 (92.6%)
Suboptimal Medical Management: 32 (7.4%)
Visit 2
Visit 2
Reason for Admission
Reason for Admission
IllnessRelated: 392
PhysicianRelated: 4
PatientRelated: 6
IllnessRelated: 20
Progression of Illness: 362
Progression of Illness:19
Abnormal Lab Results: 30
Abnormal Lab Results: 1
Non-Compliance: 6
PhysicianRelated: 11
PatientRelated: 1
Non-Compliance: 1 Deficiency in Management: 8
Missed Significant Radiology Findings: 4
TOTALS of Reasons for Admissions: 434 Illness-Related: 412 (94.9%)
Deficiency in Discharge Instruction/Follow Up Plan: 1 Missed Significant Radiology Findings: 2
Patient-Related: 7 (1.6%) Physician-Related: 15 (3.5%)
Figure 3. Reasons for admission within 72 hours after initial discharge from emergency department.
Admissions Because of Physician-Related Factors at Visit 1 Table 1 lists all 15 admissions that were because of physician-related factors, including the 6 with missed radiologic findings, and details the specific deficiency identified at visit 1. Only 8 patients were admitted specifically because of a deficiency in medical management at visit 1; this constitutes 1.8% of the study population who were admitted for
a related condition after discharge at visit 1 and 0.4% of all 1829 return visits to the ED within 72 hours.
Admissions to the ICU/OR at Visit 2 Both visits were reviewed in detail for all return visits resulting in ICU admissions (28 patients), transfer to the OR (5 patients), or transfer to another facility (5 patients; Table 2).
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
6
American Journal of Medical Quality
Table 1. Details of Cases With Physician-Related Reasons for Admission. Diagnosis at Visit 1
Disposition at Visit 2
Final Diagnosis at Visit 2
Review of Management at Visit 1
Type of Deficiency at Visit 1
Acceptable Acceptable
Missed radiologic findinga Missed radiologic findingb
Acceptable
Missed radiologic findingc
General ward
Acute appendicitis Acute chest syndrome, fever Pneumothorax, rib fracture Intussusception
Acceptable
Missed radiologic findingd
OR
Esophageal foreign body
Suboptimal
Medical management
General ward
Headache, double vision, vomiting Croup TGA with VSD, cardiomyopathy
Suboptimal
Medical management
Suboptimal Suboptimal
Medical management Medical management
General ward
Status asthmaticus
Suboptimal
Chest pain, pneumonia
General ward
Suboptimal
Sickle-cell disease with fever NGT replacement GERD Urinary tract infection, site not specified Vomiting alone, diarrhea, diaper rash
General ward
Pneumonia, pleural effusion Acute chest syndrome
Written discharge instructions/follow-up plans Missed radiologic findinge
Suboptimal
Missed radiologic findingf
General ward General ward General ward
Renal colic Pyloric stenosis Fever
Suboptimal Suboptimal Suboptimal
Medical management Medical management Medical management
OR
Intussusception
Suboptimal
Medical management
RLQ abdominal pain Fever, cough, sickle-cell disease Rib fracture
General ward General ward
Diarrhea of presumed infectious origin Exam for suspected condition, drooling Classic migraine ASOM End-stage cardiomyopathy, near syncope Asthma attack
General ward General ward
General ward
Abbreviations: ASOM, acute suppurative otitis media; GERD, gastroesophageal reflux disease; NGT, nasogastric tube; OR, operating room; RLQ, right lower quadrant; TGA, transposition of great arteries; VSD, ventricular septal defect. a Missed fecalith on abdominal X-ray. b Missed infiltrate on chest X-ray. c Missed pneumothorax on chest X-ray. d Missed intussusception on abdominal X-ray. e Missed pleural effusion on chest X-ray; final radiology reading available prior to discharge. f Missed infiltrate on chest X-ray; final radiology reading available prior to discharge.
ICU Admissions. Twenty-six of 28 (92.9%) patients who were admitted to the ICU at visit 2 had acceptable care during visit 1. The care of the remaining 2 patients was deemed suboptimal based on the physician’s failure to document postintervention improvement in one patient and lack of clear written DCI specifying a time frame to follow up with the primary care pediatrician in the other. Although management was considered suboptimal in these 2 cases, the main reason for admission for all 28 patients was deemed to be illness related (progression of illness). OR Admissions. All 5 patients who were transferred to the OR had gastrointestinal diagnoses. Two had acceptable care at visit 1. The reason for admission for these 2 patients was illness related (progression of disease) in
one patient and patient related (noncompliance) in the other. Of the 3 with suboptimal care at visit 1 who went to the OR at visit 2, the reason for admission in 2 patients was progression of illness. The deficiencies at visit 1 pertained to written DCI in one (a child who returned with large bowel obstruction secondary to constipation) and to follow-up plans in the other (a child with abdominal pain who was given no specified time frame to follow-up and subsequently was diagnosed with intussusception). In these 2 patients, there did not appear to be any indication of emergent surgical intervention at the time of discharge; however, the reviewers felt that these diagnoses may have been detected sooner if close follow-up after discharge was recommended in the DCI and follow-up plans. The third patient with suboptimal care at visit 1 went to the OR because of physician-related reasons (no
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
7
Cheng et al Table 2. Diagnoses at Visit 1 and Visit 2 and Reason for Admission for Patients Who Were Admitted to the ICU (n = 28) or OR (n = 5) at Visit 2. Disposition at Visit 2
Final Diagnosis Visit 2
Asthma exacerbation
ICU
Bleeding from GT site
ICU
Fever in chronic illness Asthma exacerbation
ICU ICU
Pneumonia
ICU
Viral exanthem
ICU
Pneumonia Herpetic gingivostomatitis Seizures
ICU ICU ICU
Atypical pneumonia
ICU
Seizure, cardiac surgery, trisomy 21 Pneumonia, asthma exacerbation Posttraumatic HA, scalp swelling/bruise
ICU
Pneumonia, hypoxia, status asthmaticus Shock, respiratory failure, post tonsillectomy bleeding, GERD, congenital heart disease RSV bronchiolitis Status asthmaticus, hypoxia, tachypnea Pulmonary insufficiency, hypoxia, wheezing Erythema multiforme major, rash Pneumonia Empyema, hyponatremia Altered mental status, subdural hematoma, seizure Hypoxia, status asthmaticus, pneumonia Status epilepticus
Fever, exam for hypoxia, chronic hypoxia, ex preemie, pulmonary HTN, pulmonary insufficiency Acute upper respiratory infections of unspecified site Cough, acute upper respiratory infections of unspecified site Acute upper respiratory infections of unspecified site Bronchiolitis Langerhan’s cell histiocytosis, fever, viral infection
ICU
Asthma, unspecified, with (acute) exacerbation Acute bronchiolitis due to RSV Acute suppurative otitis media without spontaneous rupture of eardrum
ICU
Diagnosis at Visit 1
ICU ICU
ICU ICU ICU ICU ICU
ICU ICU
Review of Management at Visit 1
Reason for Admission
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable Acceptable
Progression of illness Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable Acceptable Acceptable
Progression of illness Progression of illness Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Bronchiolitis, respiratory Acceptable distress, hypoxia Subdural hematoma, HA, fall, Acceptable cerebral ventriculomegaly, posterior fossa extradural hemorrhage Respiratory failure, Acceptable pulmonary HTN, BPD, GERD
Progression of illness
Myasthenia gravis, dysphagia, Acceptable GT dependent Acute renal failure, HUS, Acceptable hypertension, anemia Septic shock Acceptable
Progression of illness
Bronchiolitis Respiratory distress, wheezing, fever, Langerhan’s cell histiocytosis Status asthmaticus, hypoxia, hypokalemia Feeding difficulty, respiratory failure, apnea, RSV RSV, pneumonia, acute suppurative otitis media, eczema
Acceptable Acceptable
Progression of illness Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Progression of illness
Progression of illness
Progression of illness Progression of illness
(continued)
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
8
American Journal of Medical Quality
Table 2. (continued) Diagnosis at Visit 1
Disposition at Visit 2
Seizure, other convulsions
ICU
Acute nasopharyngitis (common cold) Extrinsic asthma, with (acute) exacerbation Bronchiolitis
ICU
Wheezing
ICU
Extrinsic asthma with status asthmaticus Constipation Abdominal pain
ICU
Constipation Exam for suspected condition, drooling Vomiting, diarrhea
Final Diagnosis Visit 2
Review of Management at Visit 1
Reason for Admission
Status epilepticus, altered mental status RSV bronchiolitis, respiratory distress Status asthmaticus, respiratory distress Moraxella catarrhalis pneumonia, RSV bronchiolitis, acute respiratory failure Bronchiolitis, hypoxia, respiratory distress Asthma exacerbation
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Suboptimal
Progression of illness
Suboptimal
Progression of illness
Acceptable Acceptable
Progression of illness Noncompliance
OR OR
Sigmoid volvulus Acute abdomen, septic shock Large bowel obstruction Esophageal foreign body
Suboptimal Suboptimal
OR
Intussusception
Suboptimal
Progression of illness Deficiency in managementa Progression of illness
ICU ICU
OR OR
Abbreviations: BPD, bronchopulmonary dysplasia; GERD, gastroesophogeal reflux disease; GT, gastrostomy tube; HA, headache; HTN, hypertension; HUS, hemolytic uremic syndrome; ICU, intensive care unit; OR, operating room; RSV, respiratory syncytial virus. a No X-ray obtained in a 1-year-old with new-onset persistent drooling.
imaging at visit 1 in a one-year-old with persistent drooling, subsequently diagnosed with an esophageal foreign body). Transfers. Of the 5 patients who were transferred at visit 2, all were thought to have had acceptable care at visit 1 and transfer for admission was for illness-related reasons. Two patients were transferred to another hospital at the request of their pediatrician; another 2 were transferred to psychiatric facilities for inpatient care; and one was transferred for further management at an adult facility (24 years of age).
Discussion Monitoring unplanned returns 48 to 72 hours after an ED visit is a commonly used metric of ED quality. The present study focused on a subset of return visits: patients who were admitted to the hospital within 72 hours of a prior ED discharge. The goal of this study was to identify opportunities for improvement in care at the first visit for this potentially higher risk population. The study found that the vast majority of admissions at visit 2 were because of illness-related factors (specifically progression of the disease process). In this study, patient-related
or physician-related factors were rarely the reason for admission at visit 2, even for ICU and OR admissions. Previous studies have suggested that patients admitted at the return ED visit might be a particularly high-risk group for errors or omissions in care at the first visit and for bad outcomes.19 Ovens and Goldman reported on a small ED audit of 31 patients who were admitted at the return visit and found no pattern of problems in any one physician, time of day, or specific diagnostic category or age group.19 Since this small audit, a few have studied pediatric admissions at the return visit.15,18 In a more recent study, Pham et al examined all returns to the ED within 72 hours and concluded that the use of 72-hour returns as a quality or safety indicator was not supported by their findings. They suggested that a more refined variation such as 72-hour returns resulting in admission might have more value.18 Traditionally, the return visits metric reports all returns to the ED, regardless of disposition at either visit. The present study focused on patients who were discharged from the ED at the initial visit, returned within 72 hours, and were then admitted, as a more specific indicator of ED quality of care. In this study, less than 1% of patients discharged from the pediatric ED returned within 72 hours and were admitted. A small subset of these patients were high-risk admissions (to ICU or OR). Because few
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
9
Cheng et al hospitals currently monitor this group of patients, there are no available benchmarks. The vast majority of admissions at the 72-hour return visit were because of illness-related factors, specifically progression of illness. Rarely was inadequacy of care or physician-related factors during visit 1 the reason for admission at visit 2. This also was true for high-risk admissions to the ICU or OR, where nearly all admissions were the result of progression of illness. Pierce et al found that the most common reason for return visits to a general ED was patient-related factors; these authors studied all returns, not just those who were admitted at the second visit.17 Goldman et al reported a higher acuity of illness, likely representing progression of illness, as the sole factor associated with admission in children who returned to the ED.15 DePiero et al, focusing on reasons for admission at the return visit in a pediatric ED, reported similar findings to those of the present study: 90% of 261 patients admitted at the return visit were because of progression of illness.20 More than a decade later, the present study corroborates these findings using a more objective method of review by 3 independent reviewers using a standardized methodology for chart review. Similarly, in a recent report by Abualenain et al, in a very large number of patients in a general ED, quality problems including either substandard care or adverse outcomes among 72-hour return admissions were relatively rare. These authors concluded that 72-hour return rates should not be reported as an indicator of quality of ED care without a concurrent chart review.21 Easter and Bachur focused attention on who should perform chart reviews of admissions at the return visit and concluded that EDs should not rely exclusively on the treating physician to identify the reason for return.13 To avoid such bias, the present study had not only the treating physician but also 2 other physicians not involved in the care of the patient reviewing both visits of each case independently, with all physicians blinded to the others’ reviews. Finally, to further reduce bias, the authors additionally reviewed any cases with disagreement between the reviewers. Of note are 2 interesting, and seemingly contradictory, findings on review of care provided at visit 1 and reason for admission at visit 2: suboptimal management at visit 1, but the reason for admission was not considered physician related, and the converse, acceptable management at visit 1, but the reason for admission was thought to be physician related. The former situation usually occurred when there were minor issues in care noted in visit 1, such as documentation of reassessment, and incomplete written DCI, among others. Although some may feel that DCI and follow-up plans should be considered a physician-related deficiency, the reviewers in this study may have made a distinction between these deficiencies as being the reason for return to the ED but not the reason
for admission at visit 2. The latter situation, where despite the classification of management as being acceptable, the reason for admission was physician related, occurred primarily in patients for whom an ED physician missed a radiologic finding. In this situation, although the reason for admission was clearly because of physician-related factors, the reviewers recognized that the care provided at visit 1 met accepted standards, given that ED physicians are not expected to be able to accurately detect radiologic findings 100% of the time. Studies have shown an expected rate of discrepancies between radiologic interpretations by ED physicians and radiologists.22,23 Even with these discrepancies, very rarely do the changes in the readings affect clinical care. Although missed radiologic findings was the major factor for physician-related reasons for admission at visit 2, overall this was a very small proportion of all return visits; all of these patients returned for persistent symptoms and ultimately received appropriate care. The subset of most interest is the 8 patients who had suboptimal management at visit 1 and were admitted at visit 2 for physician-related deficiencies in management that are within the expertise of a PEM physician. They represented about 2% of return visits to the ED resulting in admission and only 0.4% of all return visits. The present study found that nearly one in 4 patients who returned within 72 hours of a prior ED discharge was admitted. Although this rate is higher than what has been reported for community EDs that care for children, presumably because of a selection bias for sicker and medically complex children, it is also somewhat higher than reported rates for tertiary care pediatric centers in other studies.1,13,15 Although there may be several factors contributing to this, one trend noted in the present study was that many patients admitted at the return visit had complex chronic medical conditions. Sometimes a subspecialist would recommend discharge at visit 1, but would then request admission at visit 2. For these patients with chronic illnesses, a medical home that could provide more comprehensive follow-up after an ED visit for acute problems may have prevented repeated ED visits leading to an admission.
Limitations This study has some limitations. It is a retrospective chart review; this type of study needs to have a retrospective design by nature, occurring after a patient has returned and been admitted. In this study methodology, biases may have been introduced during the review process. The reviewers were unblinded to the disposition at visit 2; however, it was felt that this allowed them to review the care at visit 1 even more critically. Reviewers also were unblinded to the identity of the treating physician at visit 1;
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
10
American Journal of Medical Quality
and one of them was the treating physician at visit 1, reviewing his or her own care. However, including the treating physician as one of the reviewers allowed consideration of specific circumstances that the other reviewers, who had not been involved in the care of the patient, may not have been privy to. Using majority opinion for reason for admission may have overlooked cases with physician-related reasons for admission; however, the authors reviewed all of these cases. In addition, the survey-based design may have limited the option choices that reviewers had. Although this study included direct admissions, as well as presentations to the rest of the facilities in the system that use the same EMR, patients who returned to an ED outside the health care system may have been missed. However, the 2 EDs and corresponding hospitals in this study account for the vast majority of pediatrics admissions in the local metro area; the authors believe this study captured the majority of admissions on return visits.
Conclusion In conclusion, admission to the hospital after returning to the ED within 72 hours of a prior ED discharge is almost always because of illness-related factors; physicianrelated factors rarely contribute to these admissions. This is true even for admissions to a higher level of care such as the ICU or OR. Return visits resulting in admission may not be reflective of the quality of care provided in the pediatric ED. Readmissions after hospital discharge have recently come under scrutiny as the Affordable Care Act requires the Centers for Medicare & Medicaid Services to consider this metric in its payments to hospitals.24 However, a recent commentary argues that readmission to the hospital is “clearly multi-factorial and likely outside the influence of hospital care,” and questions the validity of this as a quality metric.25 Multiple studies show that the majority of ED return visits also are related to factors outside the influence of ED care. The findings of the present study extend this to ED return visits resulting in admission, and further question the utility of routinely monitoring ED return visits in lieu of other quality measures that are perhaps more tightly associated with actual quality of care. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The authors received no financial support for the research, authorship, and/or publication of this article.
References 1. Alessandrini EA, Lavelle JM, Grenfell SM, Jacobstein CR, Shaw KN. Return visits to a pediatric emergency department. Pediatr Emerg Care. 2004;20:166-171. 2. Cho CS, Shapiro DJ, Cabana MD, Maselli JH, Hersh AL. A national depiction of children with return visits to the emergency department within 72 hours, 2001-2007. Pediatr Emerg Care. 2012;28:606-610. 3. Costabel S, Piccotti E, Sartini M, Magnani M, Di Pietro P. Return visits to the paediatric emergency department: first analysis in Italy. J Prev Med Hyg. 2008;49:142-147. 4. Di Giuseppe G, Abbate R, Albano L, Marinelli P, Angelillo IF. Characteristics of patients returning to emergency departments in Naples, Italy. BMC Health Serv Res. 2008;8:97. 5. Hu KW, Lu YH, Lin HJ, Guo HR, Foo NP. Unscheduled return visits with and without admission post emergency department discharge. J Emerg Med. 2012;43:1110-1118. 6. Imsuwan I. Characteristics of unscheduled emergency department return visit patients within 48 hours in Thammasat University Hospital. J Med Assoc Thai. 2011;94(suppl 7):S73-S80. 7. Keith KD, Bocka JJ, Kobernick MS, Krome RL, Ross MA. Emergency department revisits. Ann Emerg Med. 1989;18:964-968. 8. Khan NU, Razzak JA, Saleem AF, Khan UR, Mir MU, Aashiq B. Unplanned return visit to emergency department: a descriptive study from a tertiary care hospital in a low-income country. Eur J Emerg Med. 2011;18: 276-278. 9. Kirby SE, Dennis SM, Jayasinghe UW, Harris MF. Unplanned return visits to emergency in a regional hospital. Aust Health Rev. 2012;36:336-341. 10. Kuan WS, Mahadevan M. Emergency unscheduled returns: can we do better? Singapore Med J. 2009;50:1068-1071. 11. Minnee D, Wilkinson J. Return visits to the emergency department and related hospital admissions by people aged 65 and over. N Z Med J. 2011;124:67-74. 12. Joint policy statement—guidelines for care of children in the emergency department. Pediatrics. 2009;124: 1233-1243. 13. Easter JS, Bachur R. Physicians’ assessment of pediatric returns to the emergency department. J Emerg Med. 2013;44:682-688. 14. Ali AB, Place R, Howell J, Malubay SM. Early pediatric emergency department return visits: a prospective patientcentric assessment. Clin Pediatr. 2012;51:651-658. 15. Goldman RD, Kapoor A, Mehta S. Children admitted to the hospital after returning to the emergency department within 72 hours. Pediatr Emerg Care. 2011;27:808-811. 16. Kelly AM, Chirnside AM, Curry CH. An analysis of unscheduled return visits to an urban emergency department. N Z Med J. 1993;106:334-336. 17. Pierce JM, Kellerman AL, Oster C. “Bounces”: an analysis of short-term return visits to a public hospital emergency department. Ann Emerg Med. 1990;19:752-757. 18. Pham JC, Kirsch TD, Hill PM, DeRuggerio K, Hoffmann B. Seventy-two-hour returns may not be a good indicator of
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
11
Cheng et al safety in the emergency department: a national study. Acad Emerg Med. 2011;18:390-397. 19. Ovens H, Goldman B. Analysis of return visits. Ann Emerg Med. 1990;19:339. 20. Depiero AD, Ochsenschlager DW, Chamberlain JM. Analysis of pediatric hospitalizations after emergency department release as a quality improvement tool. Ann Emerg Med. 2002;39:159-163. 21. Abualenain J, Frohna WJ, Smith M, et al. The prevalence of quality issues and adverse outcomes among 72-hour return admissions in the emergency department. J Emerg Med. 2013;45:281-288. 22. Klein EJ, Koenig M, Diekema DS, Winters W. Discordant radiograph interpretation between emergency physicians
and radiologists in a pediatric emergency department. Pediatr Emerg Care. 1999;15:245-248. 23. Simon HK, Khan NS, Nordenberg DF, Wright JA. Pediatric emergency physician interpretation of plain radiographs: is routine review by a radiologist necessary and cost-effective? Ann Emerg Med. 1996;27: 295-298. 24. Centers for Medicare & Medicaid Services. Readmissions Reduction Program. http://www.cms.gov/Medicare/MedicareFee-for-Service-Payment/AcuteInpatientPPS/ReadmissionsReduction-Program.html. Accessed September 3, 2013. 25. Alverson BK, O’Callaghan J. Hospital readmission: quality indicator or statistical inevitability? Pediatrics. 2013;132:569-570.
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
research-article2015
AJMXXX10.1177/1062860615594879American Journal of Medical QualityCheng et al
Article
Emergency Department Return Visits Resulting in Admission: Do They Reflect Quality of Care?
American Journal of Medical Quality 1–11 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1062860615594879 ajmq.sagepub.com
John Cheng, MD1,2, Amita Shroff, MD3, Naghma Khan, MD4,5, and Shabnam Jain, MD, MPH4,5
Abstract Prior studies have suggested that emergency department (ED) return visits resulting in admission may be a more robust quality indicator than all 72-hour returns. The objective was to evaluate factors that contribute to admission within 72 hours of ED discharge. Each return visit resulting in admission was independently reviewed by 3 physicians. Analysis was by descriptive statistics. Of 45 071 ED discharges, 4.1% returned within 72 hours; 0.96% returned for related reasons and were admitted to wards (91.2%), intensive care units (6.5%), or operating rooms (1.2%). Management was acceptable in 92.6%, suboptimal in 7.4%. Admissions were illness (94.9%), patient (1.6%), and physician related (3.5%). Almost all admissions within 72 hours after ED discharge are illness related, including all intensive care unit admissions and the majority of operating room admissions. Deficiencies in ED care are rarely the reason for admission on return. ED return visits resulting in admission may not be reflective of ED quality of care. Keywords emergency department, quality of care, return visit, pediatrics
Multiple studies have been published over the past 3 decades on unplanned returns to the emergency department (ED).1-11 It has been suggested to monitor 48- or 72-hour returns to the ED as a quality metric with the underlying premise that this cohort may identify potential deficiencies in initial medical management.3,4,12,13 Many previous studies have reported that the majority of unplanned returns to the ED are related to noncompliance with care (patient related) or progression of illness (illness related) rather than deficiencies in the initial medical management (physician related).14-17 Most studies thus far also have found that the majority of patients with return visits were discharged home on the subsequent visit. A recent study suggested that monitoring only returns that resulted in admission might have more value in identifying deficiencies in initial medical management.18 A few studies outside the United States also have reported that physicianrelated factors contribute to return visits that result in admission.5,6 The goal of this study is to evaluate factors that contribute to hospital admission within 72 hours of discharge from a pediatric ED in a US health care system, with a focus on physician-related parameters of care at the first visit.
Methods Setting The study ED is an academic, tertiary care, level one pediatric trauma center and a referral center for subspecialty care for the state and surrounding region. It serves an urban population and has an annual volume of more than 50 000, with an admission rate of about 15%. The ED is part of a large pediatric health care system that has 2 tertiary care sites and several urgent care sites that share an electronic medical record (EMR) and account for the vast majority of pediatric admissions in the metro area. A
1
Pediatric Emergency Medicine Associates, LLC, Atlanta, GA Children’s Healthcare of Atlanta at Scottish Rite Children’s Hospital, Atlanta, GA 3 Children’s Emergency Center at Gwinnett Medical Center, Lawrenceville, GA 4 Emory University, Atlanta, GA 5 Children’s Healthcare of Atlanta, Atlanta, GA 2
Corresponding Author: John Cheng, MD, Emergency Department, Children’s Healthcare of Atlanta at Scottish Rite Children’s Hospital, 1001 Johnson Ferry Road NE, Atlanta, GA 30342. Email: [email protected]
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
2
American Journal of Medical Quality
quality improvement (QI) initiative was implemented in which physicians reviewed the initial care received by patients who returned within 72 hours after discharge from the ED and were admitted to the hospital. This study was conducted after completion of the QI project. The physicians were blinded to the purpose of the study when completing their reviews.
Inclusion/Exclusion Criteria All patients presenting to the study ED from January 1 to December 31, 2011, who returned to any facility in the health care system within 72 hours (visit 2) after discharge from the ED (visit 1) and were admitted (including direct admissions) were included. Patients were considered admissions if they were not discharged from the ED. Dispositions at visit 2 included admission to the general ward, intensive care unit (ICU), or operating room (OR), transfer to another facility, or to the morgue. Patients who left without being seen (LWBS) or against medical advice (AMA) at visit 1 were excluded. Patients who returned after being admitted at visit 1 also were excluded to avoid any confounding factors from inpatient management.
Case Review Three physicians who are board certified in pediatric emergency medicine (PEM) or in pediatrics reviewed both ED visits 1 and 2 and, when pertinent, the inpatient record of visit 2. Reviewer 1 was the treating physician at visit 1. Reviewers 2 and 3 were physicians who had no contact with the patient at either visit 1 or 2. All 3 physicians reviewed the cases independently and completed a standardized Web-based form (Figure 1) regarding (1) the relatedness of the 2 visits, (2) the adequacy of care at visit 1 (overall medical management, written discharge instructions [DCI], and follow-up plans), (3) the transfer of care from one physician to another, and (4) the details at visit 2 (patient compliance with the original discharge plan, reason for return). All 3 reviewers had to select a reason for admission from a drop-down menu based on their review. An “other” category with a free-text option was included. The reasons for admission were categorized into 3 groups: illness related (progression of disease or callback for abnormal laboratory results), patient related (noncompliance), and physician related (deficiencies in medical management, written DCI, or follow-up plans, or missed significant radiologic findings). Majority opinion was used when responses were not unanimous. However, if even one reviewer classified the reason for admission as physician related, the authors reviewed the visit and reclassified it when necessary. This was done to ensure that no physician-related reasons for
admission were inadvertently overlooked during the initial review. Similarly, cases that had no majority opinion as to the reason for admission were reviewed and reclassified into one of the aforementioned 3 categories.
Statistics Proportions, means, and standard deviations were used to describe the results. The study institutional review board exempted this study as it was a QI initiative that did not affect patient management in real time.
Results In 2011, a total of 53 328 patients presented to the study ED; 38 patients LWBS, 3 left AMA, and 45 071 (84.5%) were discharged from the ED. Of these, 1829 (4.1%) returned within 72 hours: 8 LWBS, 474 were admitted, and the balance were discharged again. Of the 8 who LWBS at visit 2, 3 returned within 24 hours and were discharged; the rest did not return within 72 hours after visit 2. Of the 474 admitted at visit 2, a total of 40 of the visits were deemed unrelated to visit 1 (28 by unanimous and 12 by majority agreement) and were excluded from further analysis. In all 12 instances where one reviewer thought the visits were related, the reviewer felt that the admission was for illness-related, not physician-related, factors. Of the 45 071 ED patients who were discharged at the first visit, 434 (0.96%) returned and were admitted for a related condition. There were no deaths in this group. The average time to return was 34.7 (±18.4) hours (Figure 2). Agreement between reviewers regarding the reason for admission was unanimous in 291 (67.1%) cases; almost all of these were progression of illness. Of the remaining 143 cases, 134 (30.9%) had majority consensus and 9 had no consensus. In total, 98% of reviews had unanimous or majority agreement regarding the reason for admission. All 9 cases with no consensus among the 3 reviewers were reviewed by the authors and recategorized: 6 as illness related, 3 as physician related. Of the cases with majority agreement, 19 were categorized as illness related by 2 reviewers and physician related by the third reviewer. All of these were reviewed by the authors; 4 of the 19 were recategorized as physician related. Figure 3 shows details of the physician reviews. Of the 434 patients with returns for a related reason, medical management at visit 1 was acceptable in 402 (92.6%). Reviewers deemed the care suboptimal in 32/434 (7.4%) of cases. The main reasons for admissions for the 434 patients are shown in Figure 3. Of note, all abnormal laboratory results were positive cultures; almost all were blood cultures. Written DCI were acceptable in 424 (97.7%); follow-up plans were acceptable in 425 (97.9%).
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
3
Cheng et al
Reviewer Name
physician reviewer
Case Number
de-identified case number
Are the chief complaints and final diagnosis(es) from the both visits related?
Yes No
VISIT 1 Was there transfer of care?
Yes No
Was transfer of care documented?
Yes No
Was the final diagnosis related to the chief complaint?
Yes No
In your opinion, was the MANAGEMENT acceptable?
Yes No (comment below)
The management was suboptimal because:
free text box
In your opinion were the DISCHARGE INSTRUCTIONS acceptable?
Yes No (comment below)
The discharge instructions were suboptimal because:
free text box
In your opinion was the FOLLOW UP PLAN acceptable?
Yes No (comment below)
The follow up plan was suboptimal because:
free text box
VISIT 2 Was there any documented non-compliance with the plan from the first visit?
Yes No (comment below)
Details of non-compliance:
free text box
Reason for admission:
Progression of illness. Non-compliant. Planned return. Callback for abnormal lab results. Callback for X-ray over-read. Deficiency in initial management. Deficiency in discharge instructions and/or follow up plan. Re-referred in by outside provider. Other: specify below.
Other reason for admission:
free text box
Other factors that contributed to the return visit:
free text box
LESSONS LEARNED Learning points:
None. Pay closer attention to vital signs. Pay closer attention to other caregivers’ notes (RT/RN/tech). Document reassessments after interventions. Follow up on ALL lab and radiology results. Give sufficient detail during transfer of care. Stay objective, especially when patients come “pre-diagnosed.” Consultants are only consultants: you supersede them. Supervise residents completely in ALL aspects of patient care. Other: comment below.
Other learning points:
free text box
Should this be presented as a case conference?
Yes No
Figure 1. Case review questionnaire. Shown above is the Web-based form used by physicians to review the charts for visits 1 and 2. The answers on the right-hand side were drop-down boxes that have been expanded for this figure.
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
4
American Journal of Medical Quality
Figure 2. Flow diagram of emergency department (ED) patients admitted within 72 hours of initial discharge. ICU, intensive care unit. OR, operating room.
Overall, 412/434 (94.9%) of patients were admitted for illness-related factors at visit 2, and 7/434 (1.6%) for patient-related factors. Physician-related factors accounted for only 15/434 (3.5%) of the admissions. Two scenarios that may seem contradictory are discussed in the following sections: patients with suboptimal management at visit 1 who had illness-related reasons for admission and patients with acceptable management at visit 1 who had physician-related reasons for admission.
Patients With Suboptimal Management at Visit 1 and an Illness-Related Reason for Admission at Visit 2 There were 32 patients classified as having suboptimal management at visit 1; 20 were admitted for illnessrelated reasons. One patient’s care was deemed suboptimal with regard to the DCI and follow-up plan; however, the patient was admitted for positive cultures. Of the remaining 19, a total of 12 had suboptimal care related to DCI (when to return, home medications, or follow-up plans). For one patient, suboptimal care was related to insufficient physician documentation about the patient’s status and response to treatment
(although improvement in patient condition was documented by ED staff). The remaining 6 patients had a “deficiency” in their workup related to either a longstanding primary complaint that did not warrant admission at visit 1 or a secondary complaint that may not have warranted evaluation at visit 1 but did progress and was the reason for admission at visit 2. In other words, for these 19 patients, while there was room for improvement in the management at visit 1 (ie, suboptimal management), the reviewers felt that the reason for admission at visit 2 was illness related and not physician related.
Patients With Acceptable Management at Visit 1 and a Physician-Related Reason for Admission at Visit 2 All of these were missed radiological findings. Out of these 6 patients, 2 cases were deemed physician related because the radiologist’s reading was available to the ED physician prior to discharge. In the other 4 cases, medical management was deemed acceptable because the official reading was available only after the patient was discharged.
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
5
Cheng et al
Admissions at 72-hr Return Visit: 434
Visit 1
Acceptable Medical Management: 402 (92.6%)
Suboptimal Medical Management: 32 (7.4%)
Visit 2
Visit 2
Reason for Admission
Reason for Admission
IllnessRelated: 392
PhysicianRelated: 4
PatientRelated: 6
IllnessRelated: 20
Progression of Illness: 362
Progression of Illness:19
Abnormal Lab Results: 30
Abnormal Lab Results: 1
Non-Compliance: 6
PhysicianRelated: 11
PatientRelated: 1
Non-Compliance: 1 Deficiency in Management: 8
Missed Significant Radiology Findings: 4
TOTALS of Reasons for Admissions: 434 Illness-Related: 412 (94.9%)
Deficiency in Discharge Instruction/Follow Up Plan: 1 Missed Significant Radiology Findings: 2
Patient-Related: 7 (1.6%) Physician-Related: 15 (3.5%)
Figure 3. Reasons for admission within 72 hours after initial discharge from emergency department.
Admissions Because of Physician-Related Factors at Visit 1 Table 1 lists all 15 admissions that were because of physician-related factors, including the 6 with missed radiologic findings, and details the specific deficiency identified at visit 1. Only 8 patients were admitted specifically because of a deficiency in medical management at visit 1; this constitutes 1.8% of the study population who were admitted for
a related condition after discharge at visit 1 and 0.4% of all 1829 return visits to the ED within 72 hours.
Admissions to the ICU/OR at Visit 2 Both visits were reviewed in detail for all return visits resulting in ICU admissions (28 patients), transfer to the OR (5 patients), or transfer to another facility (5 patients; Table 2).
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
6
American Journal of Medical Quality
Table 1. Details of Cases With Physician-Related Reasons for Admission. Diagnosis at Visit 1
Disposition at Visit 2
Final Diagnosis at Visit 2
Review of Management at Visit 1
Type of Deficiency at Visit 1
Acceptable Acceptable
Missed radiologic findinga Missed radiologic findingb
Acceptable
Missed radiologic findingc
General ward
Acute appendicitis Acute chest syndrome, fever Pneumothorax, rib fracture Intussusception
Acceptable
Missed radiologic findingd
OR
Esophageal foreign body
Suboptimal
Medical management
General ward
Headache, double vision, vomiting Croup TGA with VSD, cardiomyopathy
Suboptimal
Medical management
Suboptimal Suboptimal
Medical management Medical management
General ward
Status asthmaticus
Suboptimal
Chest pain, pneumonia
General ward
Suboptimal
Sickle-cell disease with fever NGT replacement GERD Urinary tract infection, site not specified Vomiting alone, diarrhea, diaper rash
General ward
Pneumonia, pleural effusion Acute chest syndrome
Written discharge instructions/follow-up plans Missed radiologic findinge
Suboptimal
Missed radiologic findingf
General ward General ward General ward
Renal colic Pyloric stenosis Fever
Suboptimal Suboptimal Suboptimal
Medical management Medical management Medical management
OR
Intussusception
Suboptimal
Medical management
RLQ abdominal pain Fever, cough, sickle-cell disease Rib fracture
General ward General ward
Diarrhea of presumed infectious origin Exam for suspected condition, drooling Classic migraine ASOM End-stage cardiomyopathy, near syncope Asthma attack
General ward General ward
General ward
Abbreviations: ASOM, acute suppurative otitis media; GERD, gastroesophageal reflux disease; NGT, nasogastric tube; OR, operating room; RLQ, right lower quadrant; TGA, transposition of great arteries; VSD, ventricular septal defect. a Missed fecalith on abdominal X-ray. b Missed infiltrate on chest X-ray. c Missed pneumothorax on chest X-ray. d Missed intussusception on abdominal X-ray. e Missed pleural effusion on chest X-ray; final radiology reading available prior to discharge. f Missed infiltrate on chest X-ray; final radiology reading available prior to discharge.
ICU Admissions. Twenty-six of 28 (92.9%) patients who were admitted to the ICU at visit 2 had acceptable care during visit 1. The care of the remaining 2 patients was deemed suboptimal based on the physician’s failure to document postintervention improvement in one patient and lack of clear written DCI specifying a time frame to follow up with the primary care pediatrician in the other. Although management was considered suboptimal in these 2 cases, the main reason for admission for all 28 patients was deemed to be illness related (progression of illness). OR Admissions. All 5 patients who were transferred to the OR had gastrointestinal diagnoses. Two had acceptable care at visit 1. The reason for admission for these 2 patients was illness related (progression of disease) in
one patient and patient related (noncompliance) in the other. Of the 3 with suboptimal care at visit 1 who went to the OR at visit 2, the reason for admission in 2 patients was progression of illness. The deficiencies at visit 1 pertained to written DCI in one (a child who returned with large bowel obstruction secondary to constipation) and to follow-up plans in the other (a child with abdominal pain who was given no specified time frame to follow-up and subsequently was diagnosed with intussusception). In these 2 patients, there did not appear to be any indication of emergent surgical intervention at the time of discharge; however, the reviewers felt that these diagnoses may have been detected sooner if close follow-up after discharge was recommended in the DCI and follow-up plans. The third patient with suboptimal care at visit 1 went to the OR because of physician-related reasons (no
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
7
Cheng et al Table 2. Diagnoses at Visit 1 and Visit 2 and Reason for Admission for Patients Who Were Admitted to the ICU (n = 28) or OR (n = 5) at Visit 2. Disposition at Visit 2
Final Diagnosis Visit 2
Asthma exacerbation
ICU
Bleeding from GT site
ICU
Fever in chronic illness Asthma exacerbation
ICU ICU
Pneumonia
ICU
Viral exanthem
ICU
Pneumonia Herpetic gingivostomatitis Seizures
ICU ICU ICU
Atypical pneumonia
ICU
Seizure, cardiac surgery, trisomy 21 Pneumonia, asthma exacerbation Posttraumatic HA, scalp swelling/bruise
ICU
Pneumonia, hypoxia, status asthmaticus Shock, respiratory failure, post tonsillectomy bleeding, GERD, congenital heart disease RSV bronchiolitis Status asthmaticus, hypoxia, tachypnea Pulmonary insufficiency, hypoxia, wheezing Erythema multiforme major, rash Pneumonia Empyema, hyponatremia Altered mental status, subdural hematoma, seizure Hypoxia, status asthmaticus, pneumonia Status epilepticus
Fever, exam for hypoxia, chronic hypoxia, ex preemie, pulmonary HTN, pulmonary insufficiency Acute upper respiratory infections of unspecified site Cough, acute upper respiratory infections of unspecified site Acute upper respiratory infections of unspecified site Bronchiolitis Langerhan’s cell histiocytosis, fever, viral infection
ICU
Asthma, unspecified, with (acute) exacerbation Acute bronchiolitis due to RSV Acute suppurative otitis media without spontaneous rupture of eardrum
ICU
Diagnosis at Visit 1
ICU ICU
ICU ICU ICU ICU ICU
ICU ICU
Review of Management at Visit 1
Reason for Admission
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable Acceptable
Progression of illness Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable Acceptable Acceptable
Progression of illness Progression of illness Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Bronchiolitis, respiratory Acceptable distress, hypoxia Subdural hematoma, HA, fall, Acceptable cerebral ventriculomegaly, posterior fossa extradural hemorrhage Respiratory failure, Acceptable pulmonary HTN, BPD, GERD
Progression of illness
Myasthenia gravis, dysphagia, Acceptable GT dependent Acute renal failure, HUS, Acceptable hypertension, anemia Septic shock Acceptable
Progression of illness
Bronchiolitis Respiratory distress, wheezing, fever, Langerhan’s cell histiocytosis Status asthmaticus, hypoxia, hypokalemia Feeding difficulty, respiratory failure, apnea, RSV RSV, pneumonia, acute suppurative otitis media, eczema
Acceptable Acceptable
Progression of illness Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Progression of illness
Progression of illness
Progression of illness Progression of illness
(continued)
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
8
American Journal of Medical Quality
Table 2. (continued) Diagnosis at Visit 1
Disposition at Visit 2
Seizure, other convulsions
ICU
Acute nasopharyngitis (common cold) Extrinsic asthma, with (acute) exacerbation Bronchiolitis
ICU
Wheezing
ICU
Extrinsic asthma with status asthmaticus Constipation Abdominal pain
ICU
Constipation Exam for suspected condition, drooling Vomiting, diarrhea
Final Diagnosis Visit 2
Review of Management at Visit 1
Reason for Admission
Status epilepticus, altered mental status RSV bronchiolitis, respiratory distress Status asthmaticus, respiratory distress Moraxella catarrhalis pneumonia, RSV bronchiolitis, acute respiratory failure Bronchiolitis, hypoxia, respiratory distress Asthma exacerbation
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Acceptable
Progression of illness
Suboptimal
Progression of illness
Suboptimal
Progression of illness
Acceptable Acceptable
Progression of illness Noncompliance
OR OR
Sigmoid volvulus Acute abdomen, septic shock Large bowel obstruction Esophageal foreign body
Suboptimal Suboptimal
OR
Intussusception
Suboptimal
Progression of illness Deficiency in managementa Progression of illness
ICU ICU
OR OR
Abbreviations: BPD, bronchopulmonary dysplasia; GERD, gastroesophogeal reflux disease; GT, gastrostomy tube; HA, headache; HTN, hypertension; HUS, hemolytic uremic syndrome; ICU, intensive care unit; OR, operating room; RSV, respiratory syncytial virus. a No X-ray obtained in a 1-year-old with new-onset persistent drooling.
imaging at visit 1 in a one-year-old with persistent drooling, subsequently diagnosed with an esophageal foreign body). Transfers. Of the 5 patients who were transferred at visit 2, all were thought to have had acceptable care at visit 1 and transfer for admission was for illness-related reasons. Two patients were transferred to another hospital at the request of their pediatrician; another 2 were transferred to psychiatric facilities for inpatient care; and one was transferred for further management at an adult facility (24 years of age).
Discussion Monitoring unplanned returns 48 to 72 hours after an ED visit is a commonly used metric of ED quality. The present study focused on a subset of return visits: patients who were admitted to the hospital within 72 hours of a prior ED discharge. The goal of this study was to identify opportunities for improvement in care at the first visit for this potentially higher risk population. The study found that the vast majority of admissions at visit 2 were because of illness-related factors (specifically progression of the disease process). In this study, patient-related
or physician-related factors were rarely the reason for admission at visit 2, even for ICU and OR admissions. Previous studies have suggested that patients admitted at the return ED visit might be a particularly high-risk group for errors or omissions in care at the first visit and for bad outcomes.19 Ovens and Goldman reported on a small ED audit of 31 patients who were admitted at the return visit and found no pattern of problems in any one physician, time of day, or specific diagnostic category or age group.19 Since this small audit, a few have studied pediatric admissions at the return visit.15,18 In a more recent study, Pham et al examined all returns to the ED within 72 hours and concluded that the use of 72-hour returns as a quality or safety indicator was not supported by their findings. They suggested that a more refined variation such as 72-hour returns resulting in admission might have more value.18 Traditionally, the return visits metric reports all returns to the ED, regardless of disposition at either visit. The present study focused on patients who were discharged from the ED at the initial visit, returned within 72 hours, and were then admitted, as a more specific indicator of ED quality of care. In this study, less than 1% of patients discharged from the pediatric ED returned within 72 hours and were admitted. A small subset of these patients were high-risk admissions (to ICU or OR). Because few
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
9
Cheng et al hospitals currently monitor this group of patients, there are no available benchmarks. The vast majority of admissions at the 72-hour return visit were because of illness-related factors, specifically progression of illness. Rarely was inadequacy of care or physician-related factors during visit 1 the reason for admission at visit 2. This also was true for high-risk admissions to the ICU or OR, where nearly all admissions were the result of progression of illness. Pierce et al found that the most common reason for return visits to a general ED was patient-related factors; these authors studied all returns, not just those who were admitted at the second visit.17 Goldman et al reported a higher acuity of illness, likely representing progression of illness, as the sole factor associated with admission in children who returned to the ED.15 DePiero et al, focusing on reasons for admission at the return visit in a pediatric ED, reported similar findings to those of the present study: 90% of 261 patients admitted at the return visit were because of progression of illness.20 More than a decade later, the present study corroborates these findings using a more objective method of review by 3 independent reviewers using a standardized methodology for chart review. Similarly, in a recent report by Abualenain et al, in a very large number of patients in a general ED, quality problems including either substandard care or adverse outcomes among 72-hour return admissions were relatively rare. These authors concluded that 72-hour return rates should not be reported as an indicator of quality of ED care without a concurrent chart review.21 Easter and Bachur focused attention on who should perform chart reviews of admissions at the return visit and concluded that EDs should not rely exclusively on the treating physician to identify the reason for return.13 To avoid such bias, the present study had not only the treating physician but also 2 other physicians not involved in the care of the patient reviewing both visits of each case independently, with all physicians blinded to the others’ reviews. Finally, to further reduce bias, the authors additionally reviewed any cases with disagreement between the reviewers. Of note are 2 interesting, and seemingly contradictory, findings on review of care provided at visit 1 and reason for admission at visit 2: suboptimal management at visit 1, but the reason for admission was not considered physician related, and the converse, acceptable management at visit 1, but the reason for admission was thought to be physician related. The former situation usually occurred when there were minor issues in care noted in visit 1, such as documentation of reassessment, and incomplete written DCI, among others. Although some may feel that DCI and follow-up plans should be considered a physician-related deficiency, the reviewers in this study may have made a distinction between these deficiencies as being the reason for return to the ED but not the reason
for admission at visit 2. The latter situation, where despite the classification of management as being acceptable, the reason for admission was physician related, occurred primarily in patients for whom an ED physician missed a radiologic finding. In this situation, although the reason for admission was clearly because of physician-related factors, the reviewers recognized that the care provided at visit 1 met accepted standards, given that ED physicians are not expected to be able to accurately detect radiologic findings 100% of the time. Studies have shown an expected rate of discrepancies between radiologic interpretations by ED physicians and radiologists.22,23 Even with these discrepancies, very rarely do the changes in the readings affect clinical care. Although missed radiologic findings was the major factor for physician-related reasons for admission at visit 2, overall this was a very small proportion of all return visits; all of these patients returned for persistent symptoms and ultimately received appropriate care. The subset of most interest is the 8 patients who had suboptimal management at visit 1 and were admitted at visit 2 for physician-related deficiencies in management that are within the expertise of a PEM physician. They represented about 2% of return visits to the ED resulting in admission and only 0.4% of all return visits. The present study found that nearly one in 4 patients who returned within 72 hours of a prior ED discharge was admitted. Although this rate is higher than what has been reported for community EDs that care for children, presumably because of a selection bias for sicker and medically complex children, it is also somewhat higher than reported rates for tertiary care pediatric centers in other studies.1,13,15 Although there may be several factors contributing to this, one trend noted in the present study was that many patients admitted at the return visit had complex chronic medical conditions. Sometimes a subspecialist would recommend discharge at visit 1, but would then request admission at visit 2. For these patients with chronic illnesses, a medical home that could provide more comprehensive follow-up after an ED visit for acute problems may have prevented repeated ED visits leading to an admission.
Limitations This study has some limitations. It is a retrospective chart review; this type of study needs to have a retrospective design by nature, occurring after a patient has returned and been admitted. In this study methodology, biases may have been introduced during the review process. The reviewers were unblinded to the disposition at visit 2; however, it was felt that this allowed them to review the care at visit 1 even more critically. Reviewers also were unblinded to the identity of the treating physician at visit 1;
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
10
American Journal of Medical Quality
and one of them was the treating physician at visit 1, reviewing his or her own care. However, including the treating physician as one of the reviewers allowed consideration of specific circumstances that the other reviewers, who had not been involved in the care of the patient, may not have been privy to. Using majority opinion for reason for admission may have overlooked cases with physician-related reasons for admission; however, the authors reviewed all of these cases. In addition, the survey-based design may have limited the option choices that reviewers had. Although this study included direct admissions, as well as presentations to the rest of the facilities in the system that use the same EMR, patients who returned to an ED outside the health care system may have been missed. However, the 2 EDs and corresponding hospitals in this study account for the vast majority of pediatrics admissions in the local metro area; the authors believe this study captured the majority of admissions on return visits.
Conclusion In conclusion, admission to the hospital after returning to the ED within 72 hours of a prior ED discharge is almost always because of illness-related factors; physicianrelated factors rarely contribute to these admissions. This is true even for admissions to a higher level of care such as the ICU or OR. Return visits resulting in admission may not be reflective of the quality of care provided in the pediatric ED. Readmissions after hospital discharge have recently come under scrutiny as the Affordable Care Act requires the Centers for Medicare & Medicaid Services to consider this metric in its payments to hospitals.24 However, a recent commentary argues that readmission to the hospital is “clearly multi-factorial and likely outside the influence of hospital care,” and questions the validity of this as a quality metric.25 Multiple studies show that the majority of ED return visits also are related to factors outside the influence of ED care. The findings of the present study extend this to ED return visits resulting in admission, and further question the utility of routinely monitoring ED return visits in lieu of other quality measures that are perhaps more tightly associated with actual quality of care. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The authors received no financial support for the research, authorship, and/or publication of this article.
References 1. Alessandrini EA, Lavelle JM, Grenfell SM, Jacobstein CR, Shaw KN. Return visits to a pediatric emergency department. Pediatr Emerg Care. 2004;20:166-171. 2. Cho CS, Shapiro DJ, Cabana MD, Maselli JH, Hersh AL. A national depiction of children with return visits to the emergency department within 72 hours, 2001-2007. Pediatr Emerg Care. 2012;28:606-610. 3. Costabel S, Piccotti E, Sartini M, Magnani M, Di Pietro P. Return visits to the paediatric emergency department: first analysis in Italy. J Prev Med Hyg. 2008;49:142-147. 4. Di Giuseppe G, Abbate R, Albano L, Marinelli P, Angelillo IF. Characteristics of patients returning to emergency departments in Naples, Italy. BMC Health Serv Res. 2008;8:97. 5. Hu KW, Lu YH, Lin HJ, Guo HR, Foo NP. Unscheduled return visits with and without admission post emergency department discharge. J Emerg Med. 2012;43:1110-1118. 6. Imsuwan I. Characteristics of unscheduled emergency department return visit patients within 48 hours in Thammasat University Hospital. J Med Assoc Thai. 2011;94(suppl 7):S73-S80. 7. Keith KD, Bocka JJ, Kobernick MS, Krome RL, Ross MA. Emergency department revisits. Ann Emerg Med. 1989;18:964-968. 8. Khan NU, Razzak JA, Saleem AF, Khan UR, Mir MU, Aashiq B. Unplanned return visit to emergency department: a descriptive study from a tertiary care hospital in a low-income country. Eur J Emerg Med. 2011;18: 276-278. 9. Kirby SE, Dennis SM, Jayasinghe UW, Harris MF. Unplanned return visits to emergency in a regional hospital. Aust Health Rev. 2012;36:336-341. 10. Kuan WS, Mahadevan M. Emergency unscheduled returns: can we do better? Singapore Med J. 2009;50:1068-1071. 11. Minnee D, Wilkinson J. Return visits to the emergency department and related hospital admissions by people aged 65 and over. N Z Med J. 2011;124:67-74. 12. Joint policy statement—guidelines for care of children in the emergency department. Pediatrics. 2009;124: 1233-1243. 13. Easter JS, Bachur R. Physicians’ assessment of pediatric returns to the emergency department. J Emerg Med. 2013;44:682-688. 14. Ali AB, Place R, Howell J, Malubay SM. Early pediatric emergency department return visits: a prospective patientcentric assessment. Clin Pediatr. 2012;51:651-658. 15. Goldman RD, Kapoor A, Mehta S. Children admitted to the hospital after returning to the emergency department within 72 hours. Pediatr Emerg Care. 2011;27:808-811. 16. Kelly AM, Chirnside AM, Curry CH. An analysis of unscheduled return visits to an urban emergency department. N Z Med J. 1993;106:334-336. 17. Pierce JM, Kellerman AL, Oster C. “Bounces”: an analysis of short-term return visits to a public hospital emergency department. Ann Emerg Med. 1990;19:752-757. 18. Pham JC, Kirsch TD, Hill PM, DeRuggerio K, Hoffmann B. Seventy-two-hour returns may not be a good indicator of
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
11
Cheng et al safety in the emergency department: a national study. Acad Emerg Med. 2011;18:390-397. 19. Ovens H, Goldman B. Analysis of return visits. Ann Emerg Med. 1990;19:339. 20. Depiero AD, Ochsenschlager DW, Chamberlain JM. Analysis of pediatric hospitalizations after emergency department release as a quality improvement tool. Ann Emerg Med. 2002;39:159-163. 21. Abualenain J, Frohna WJ, Smith M, et al. The prevalence of quality issues and adverse outcomes among 72-hour return admissions in the emergency department. J Emerg Med. 2013;45:281-288. 22. Klein EJ, Koenig M, Diekema DS, Winters W. Discordant radiograph interpretation between emergency physicians
and radiologists in a pediatric emergency department. Pediatr Emerg Care. 1999;15:245-248. 23. Simon HK, Khan NS, Nordenberg DF, Wright JA. Pediatric emergency physician interpretation of plain radiographs: is routine review by a radiologist necessary and cost-effective? Ann Emerg Med. 1996;27: 295-298. 24. Centers for Medicare & Medicaid Services. Readmissions Reduction Program. http://www.cms.gov/Medicare/MedicareFee-for-Service-Payment/AcuteInpatientPPS/ReadmissionsReduction-Program.html. Accessed September 3, 2013. 25. Alverson BK, O’Callaghan J. Hospital readmission: quality indicator or statistical inevitability? Pediatrics. 2013;132:569-570.
Downloaded from ajm.sagepub.com at The University of Melbourne Libraries on June 5, 2016
