The Impact of an Acute Care Emergency SurgicalService on Timely Surgical Decision-Making andEmergency Department OvercrowdingAdnan Qureshi, MD, MSc, Andy Smith, MD, FRCSC, Frances Wright, MD, MEd, FRCSC,Fred Brenneman, MD, FRCSC, FACS, Sandro Rizoli, MD, PhD, FRCSC, Taulee Hsieh, MD,COL Homer C Tien, MD, MSc, FRCSC, FACS

BACKGROUND: This study evaluated how implementation of an acute care emergency surgery service(ACCESS) affected key determinants of emergency department (ED) length of stay, and par-ticularly, surgical decision time. Also, we analyzed how ACCESS affected ED overcrowding.

STUDY DESIGN: We conducted a before and after study of all ED patients referred to ACCESS from January 1,2007 to June 30, 2009. ACCESS was implemented on July 1, 2008. The primary outcome wassurgical decision time; the secondary outcome was a measure of overall ED overcrowding:“time-to-stretcher” for all ED patients. The control groups were patients referred to internalmedicine or urology. Patients with appendicitis were studied in order to analyze the impact onpatient outcomes and to determine barriers to efficient ED patient flow.

RESULTS: Of 2,510 patients, 1,448 patients were pre-ACCESS, and 1,062 were after ACCESS imple-mentation. Implementation of ACCESS was associated with a 15% reduction in surgicaldecision time (12.6 hours vs 10.8 hours, p � 0.01). During the same period, there were nosignificant changes in decision time for our control groups. Also, the mean time-to-stretcher forall ED patients decreased by 20%. In patients with appendicitis, we found that patient flowcould be further improved by a timely request for surgical consultation and expedited imaging.Finally, we found that patients with nonperforated appendicitis with a fecalith on CT imagingwere more likely to suffer perforation while waiting for surgery.

CONCLUSIONS: ACCESS reduced surgical decision time for surgical patients. Also, ACCESS improved overallED crowding, as measured by time-to-stretcher for ED patients. Further improvements couldbe made by improving time to imaging. Patients referred for nonperforated appendicitis with afecalith on CT should have expedited surgery. (J Am Coll Surg 2011;213:284–293. © 2011 by

the American College of Surgeons)

aOvirbvs

a(

Emergency department (ED) overcrowding has been de-fined as a situation in which demand for acute care exceedsthe ability of physicians and nurses to provide timely qual-ity care, which threatens patient health and fosters patientdissatisfaction.1-6 ED overcrowding has been identified as awidespread and serious problem with adverse conse-quences, both in the United States and Canada.5-8

Disclosure Information: Nothing to disclose.This study was funded by a grant from the Ministry Of Health for theProvince of Ontario, Phase III AFP Innovation Fund.

Received January 18, 2011; Revised April 20, 2011; Accepted April 20, 2011.From the Department of Surgery, Sunnybrook Health Sciences Centre, Uni-versity of Toronto, Toronto, Ontario, Canada (Qureshi, Smith, Wright, Bren-neman, Rizoli, Hsieh, Tien) and the Canadian Forces Health Services (Tien).Correspondence address: Dr Homer C Tien, Division of General Surgery,

iSunnybrook Health Sciences Centre, 2075 Bayview Ave, Toronto, ON M4N3M5, Canada. email: homer.tien@sunnybrook.ca

284© 2011 by the American College of SurgeonsPublished by Elsevier Inc.

Ontario is Canada’s largest province, with more than 13million residents and almost 40% of Canada’s total popu-lation.9 There has been considerable political and mediattention in Ontario focused on ED patient wait times.2 Inntario, attention to this problem has prompted the pro-

incial Ministry of Health to place ED wait times as one ofts two top health care priorities, and it has reallocatedesources to address this concern.10 A similar concern haseen expressed in the United States: that increasing EDisit rates and overcrowding are jeopardizing patientafety.6,7,11

Acute care surgery has been defined as the urgent assess-ment and treatment of nontrauma general surgical emer-gencies.12,13 Data from US centers have suggested thatcute care surgical services can reduce ED length of stayLOS) for patients with appendicitis and may potentially

mprove patient outcomes.14,15 Our institution established

ISSN 1072-7515/11/$36.00doi:10.1016/j.jamcollsurg.2011.04.020

dmisCmbCm

285Vol. 213, No. 2, August 2011 Qureshi et al Acute Care Surgery and Emergency Department Crowding

an acute care emergency surgery service (ACCESS), on July1, 2008. The purpose of the ACCESS team was to provideprompt and expert surgical consultation to ED patients,and to provide comprehensive care to these same patientsafter admission. The hope was to improve ED length ofstay and outcomes of surgical patients, and to improveoverall ED crowding.

We wondered if delays in obtaining surgical consultationwere contributing to ED length of stay for surgical patientsand thereby contributing to overall ED crowding. There-fore, we performed a before and after study to determinethe effect of ACCESS implementation on some key deter-minants of overall ED length of stay, namely, the timelinessof the surgical team’s response to consultation requests,and the timely surgical decision-making. We hypothesizedthat ACCESS implementation would be associated withimproved consultation response times by the surgical team,more timely surgical decision-making, and reduced overallED crowding. In a cohort of patients diagnosed with ap-pendicitis, we also looked at patient outcomes and barriersto patient flow.

METHODSSettingThe study was performed at Sunnybrook Health SciencesCentre, a large academic hospital located in Toronto, Can-ada. Sunnybrook Health Sciences Centre receives approx-imately 42,000 ED visits annually.16 There are 44 EDstretchers, with 4 of these dedicated to patients awaitingtransfer to their respective wards. The ED has the supportof an ED residency program and features fellows, residents,and medical students. Sunnybrook also has the Tory Re-gional Trauma Centre, a level 1 trauma center.

The staff general surgeons at Sunnybrook Health Sci-ences Centre focus on 2 major specialty areas: trauma andoncologic surgery. There are 5 trauma surgeons, and 9 on-cologic surgeons. The 9 surgical oncologists are subdividedby subspecialty expertise into colorectal, hepatobiliary, andbreast and melanoma surgical oncology. Eleven of 14 staffsurgeons participate in the ACCESS program, based oninterest, expertise, additional administrative and academicresponsibilities, and age threshold for ongoing call

Abbreviations and Acronyms

ACCESS � acute care emergency surgery serviceCTAS � Canadian Triage and Acuity ScaleED � emergency departmentEDIS � Emergency Department Information SystemLOS � length of stay

coverage.

Selection of participantsWe used the Emergency Department Information System(EDIS) database (Eclipsys Corp) to identify all patientsreferred to the general surgery service from January 1, 2007to June 30, 2009. Patients were excluded if they wereyounger than 16 years of age or if they left the ED againstmedical advice. We also identified a subgroup of patientswith appendicitis to analyze for barriers to efficient patientflow through the ED and to study patient outcomes.

Control groupsDuring the study period, ED wait times became a priorityto the provincial health care system, so funding becameavailable for projects that might improve ED wait times.One concurrent project was a randomized controlled triallooking at the effect of establishing an emergencyphysician/registered nurse triage team on ED length ofstay. This trial protocol is registered at ClinicalTrial.gov(NCT00991471). However, study enrollment was limitedto day-time hours (8:00 AM to 2:00 PM), Monday toFriday. Preliminary results from this show that the impacton general surgery patients was low; only 30 general sur-gery patients were enrolled in this randomized controlledtrial (personal communication, Dr Ivy Cheng).

To determine whether institutional changes affected EDLOS over the study period, we analyzed key determinantsof ED LOS (decision time) for all patients requiring gen-eral internal medicine and urology consultation over thesame time period. Decision time was the time from patientregistration in the ED to the time a final decision was madeabout patient disposition (ie, admission or discharge). Westratified the analysis by the Canadian Triage and AcuityScale (CTAS) level.17 In brief, CTAS is a triage tool that

etermines the priority of patients for ED physician assess-ent. CTAS 1 patients need to be seen by a physician

mmediately 98% of the time; CTAS 2 patients need to beeen by a physician within 15 minutes 95% of the time;TAS 3 patients need to be seen by a physician within 30inutes 90% of the time; CTAS 4 patients need to be seen

y a physician within 60 minutes 85% of the time; andTAS 5 patients need to be seen by a physician within 120inutes 80% of the time.

InterventionOn July 1, 2008, ACCESS was implemented at our insti-tution. Patients were dichotomized based on their registra-tion date to pre-ACCESS or ACCESS. In our ACCESSmodel, general surgeons commit to being available for a7-day period, and to lead a team consisting of a seniorsurgical resident, several junior residents, and medical stu-

dents. During that week, the staff surgeon commits to be-

286 Qureshi et al Acute Care Surgery and Emergency Department Crowding J Am Coll Surg

ing on call overnight for a minimum of 3 nights. Anotherstaff surgeon is on call the other nights. Hand-over forthose “off-nights” has not been formalized and remains adhoc. After implementing ACCESS, the complement of se-nior residents at our institution increased from 4 to 5 toaccommodate this new team. The trauma surgery fellowsalso participate as junior attending staff for the ACCESSservice (under the supervision of the ACCESS staff ), andare available to help in the ACCESS operating rooms. Dur-ing their ACCESS week, the ACCESS staff surgeons donot do elective surgery, but devote themselves to beingavailable in a timely fashion to perform any emergencyoperation or provide emergency consultation. A formalhand-over to the next attending staff occurs every Fridaymorning.

ACCESS leadership has also self-imposed the followingtargets on the ACCESS team: response to any ED requestfor consultation within 30 minutes, and final decision re-garding patient disposition within 2 hours of the consultrequest. These ACCESS targets were briefed to ED staffand hospital leadership. ED staff members were encour-

Figure 1. Definitions of overall emergency department (ED) wait timstay; OR, operating room.

aged to call surgical residents and staff directly if these

targets were not being met. ED physician leadership wasvery supportive of ACCESS implementation. During thisstudy period, there was no formal daytime operating roomtime available for ACCESS. However, shortly after the con-clusion of this study, formal operating room time becameavailable to ACCESS during the day.

OutcomesThe primary outcome was surgical decision time. The be-ginning of the surgical decision time was defined as thetime of ED registration. The endpoint for surgical decisiontime was the time that the ACCESS team made a finaldecision regarding patient disposition (ie, admit or dis-charge). Unfortunately, during the study period, our EDISdatabase did not record time that patients actually departedthe ED; therefore, we were not able to find an exact time ofdeparture from the ED for patients. Hence, we do notreport an overall ED LOS. We also examined the timeinterval “consult request to arrival” and “arrival to finaldecision time,” because these were the response times thatthe ACCESS team had committed to improving. Figure 1

d important key determinants of ED length of stay. LOS, length of

e an

summarizes all key time intervals measured.

tsppas

287Vol. 213, No. 2, August 2011 Qureshi et al Acute Care Surgery and Emergency Department Crowding

For the appendicitis subgroup, we analyzed barriers toefficient ED patient flow and patient outcomes. We lookedat “time to ED MD visit,” “time to availability of labora-tory test results,” “time to availability of preliminary reportof diagnostic imaging,” “time-to-surgery,” appendiceal per-foration rate, and hospital LOS (ED registration to hospitaldischarge) (Fig. 1).

As a secondary outcome, we analyzed changes in the“time-to-stretcher” for all ED patients. Although there issome controversy as to the optimal measure of ED over-crowding, time-to-stretcher and ambulance off-loadingtimes have face validity as a measure of ED overcrowdingand have been used in other studies.5,8 It is reasonable to useime-to-stretcher as a measure of ED overcrowding for thistudy because general surgery patients are usually stretcheratients. Therefore, if ACCESS improved ED LOS by im-roving surgical decision time, it would be reasonable tossume that ACCESS would increase the availability of EDtretchers for other patients.

Data collectionFor the pre-ACCESS and ACCESS periods, the EDIS da-tabase was used to obtain surgical decision time and its keytime intervals for the primary analysis. For the ACCESSperiod, we had a second source of data for surgical decisiontime: ACCESS team members were told to prospectivelydocument the time that they received the ED request forsurgical consultation, the time they first arrived to beginthe consult in the ED, and the time to final decision by thegeneral surgery team. As a result, we had 2 sources of datafor 2 key time intervals in the post-ACCESS implementa-tion period: “consult request to arrival” and “arrival to finaldecision.” We therefore compared the 2 sources of data forthese measures.

For the appendicitis subgroup, manual chart review wasalso done by author AQ. Using the nursing record, wedetermined the “time to ED MD visit.” Using the elec-tronic hospital record, we identified the time that labora-tory samples were collected and when results were avail-able. Preliminary images of the diagnostic imaging werereviewed on the hospital IMPAX (Agfa Inc) system to de-termine the time the preliminary reports were available tophysicians. We reviewed the final electronic report by thestaff radiologist to determine whether the appendix wasperforated at the time of imaging (abscess, phlegmon, freeair, focal perforation). We reviewed the operative report todetermine whether the appendix was perforated at surgery,and we reviewed the operation start time. The final dis-charge report was used to identify any complications dur-ing the hospital course. The electronic record was used to

calculate the hospital LOS.

Statistical methodsWe used independent sample t-test to evaluate differencesbetween the 2 cohorts for continuous variables. An inde-pendent sample t-test was also used to compare differencesbetween consultations directed to general internal medi-cine and urology. Dichotomous variables were comparedusing a Pearson’s chi-square test. All p values are 2-tailedmethod. Means and proportions were reported with stan-dard deviations, and all data were analyzed using SAS (SASInstitute) software (version 9.02). We obtained approvalfrom our institutional review ethics board before conduct-ing this study.

RESULTSDuring the study period, 2,510 patients were referred tothe general surgery service: 1,448 were from the pre-ACCESS period, and 1,062 were seen after ACCESS im-plementation. The majority of patients were referred forappendicitis, diverticulitis, bowel obstruction, biliary dis-eases, or postoperative complications. About 66% of thesepatients had a CTAS score of 3 (Table 1). Implementationof ACCESS was associated with a 15% reduction in overallsurgical decision time for all patients referred to generalsurgery (12.6 hours vs 10.8 hours, p � 0.01).

Control groupsDuring the same period, there were 4,599 consult requeststo internal medicine, with a triage category of CTAS 3.After stratifying them by their registration date, we foundthat there was a significant increase in the decision time(12.8 hours vs 13.2 hours, p � 0.02) for patients who wereevaluated after implementation of ACCESS comparedwith pre-ACCESS. There were no statistically significantchanges in the decision time (11.5 hours vs 10.9 hours, p �0.15) for CTAS-3 patients requiring a urology consultationduring the same time period.

Appendicitis subgroupThere were 314 patients with confirmed appendicitis. Ofthese, 177 were from the pre-ACCESS period, and 137were referred after ACCESS implementation. Table 2shows baseline characteristics of patients with confirmedappendicitis. There were no significant differences betweenthe 2 groups. Most patients (both pre-ACCESS andACCESS) underwent a CT scan before surgery (67% vs58%, p � 0.2). One hundred percent of all appendicitispatients had some form of body imaging (ultrasoundand/or CT). In the pre-ACCESS period, more than 13% ofpatients had perforation already evident on initial diagnos-tic imaging, compared with 10% in the ACCESS period,

but the difference was not significant. During the pre-

R

SM

TLCDADBBPO

288 Qureshi et al Acute Care Surgery and Emergency Department Crowding J Am Coll Surg

ACCESS period, 83% of patients had surgery, comparedwith 88% after ACCESS implementation (p � 0.2).

Emergency department length of stayIn the pre-ACCESS period, the mean ED LOS for patientwith appendicitis was 17.0 hours. After ACCESS was im-plemented, we noted a 30% decrease in ED LOS, to 11.8hours (p � 0.01)(Table 3). When examining the determi-nants of ED LOS for these patients, we noticed that almostall of the reductions in ED LOS were from improved time-liness of general surgery consultation response anddecision-making. During the same period, there were nodifferences in the time interval from registration to the EDphysician assessment, nor were there differences in the timeinterval from ED physician assessment to request for gen-eral surgical consultation. On average, even after ACCESSwas implemented, it took 8 hours before the ED physicianrequested a general surgical consultation for appendicitis.There was also a nonsignificant difference from treatmentdecision to surgery for the pre-ACCESS (8.2 hours) andACCESS cohorts (8.2 vs 7.8 hours, p � 0.45).

Patient outcomesDespite significant reductions in surgical decision time, wedid not find any significant differences in patient out-comes. There were no significant differences in overall hos-pital length of stay. Furthermore, no differences were ob-served in the appendiceal perforation rates noted at time ofoperation, in the pre-ACCESS cohort compared with theACCESS cohort (16.5% vs 12.8%, p � 0.40). Finally,

Table 1. Baseline Characteristics of All General Surgeryeferrals

CharacteristicPre-ACCESS(n � 1,448)

ACCESS(n � 1,062)

pValue

Age, y, mean � SD 57 � 21 57 � 21 0.9ex, % men 51 49 0.3ethod of arrival,ambulance vswalk-in 24% ambulance 23% ambulance 0.4

riage CTAS 2, % 26 30evel CTAS 3, % 66 65TAS 4, % 6 3iagnosis, n (%)ppendicitis 186 (13) 147 (14)iverticulitis 72 (5) 67 (6)owel obstruction 258 (18) 196 (18)iliary tract 176 (12) 115 (11)ostoperatively 130 (9) 107 (10)ther

ACCESS, acute care emergency surgery service; CTAS, Canadian Triage andAcuity Scale.

there were no differences in appendiceal perforation rates

for those who presented initially to the ED with nonperfo-rated appendicitis (Table 4). In pre-ACCESS period, 102patients had no perforation evident on their initial diag-nostic imaging; 11% (n � 11) progressed to perforationwhile waiting for surgery. In the ACCESS period, 71 pa-tients had no evidence of perforation on initial imaging;11% (n � 8) progressed to perforation while waiting forsurgery. Although time to surgery (from general surgeryfinal decision) was not associated with the development ofperforation, the presence of a fecalith on the initial CT wasassociated with progression to perforation; 4 of 178 pa-tients who did not progress to perforation were noted tohave fecaliths, compared with 8 of 19 patients who didprogress to perforation (p � 0.01).

Laboratory, diagnostic imaging, and time tooperating roomOn average, a complete blood count result was availableelectronically within 2 hours after registration. Implemen-tation of ACCESS did not affect this time interval. How-ever, time from registration to preliminary CT report didimprove significantly after ACCESS implementation by1.5 hours. Despite this change, however, even after AC-CESS implementation, it took more than 7 hours fromregistration for a preliminary CT result to be available.Finally, ACCESS implementation did not significantly de-

Table 2. Baseline Characteristics of Emergency DepartmentPresenting Patients Diagnosed with Appendicitis

CharacteristicPre-ACCESS(n � 177)

ACCESS(n � 137)

pValue

Age, y, mean � SD 39.1 �16.8 38.0 �17.4 0.6Sex, % men 46 47 0.8Method of arrival,

ambulance vs walk-in 11% Ambulance 12% Ambulance 0.7

Triage: CTAS 2, % 17 17Level CTAS 3, % 77 80CTAS 4, % 5 3CTAS, mean � SD 2.90 � 0.46 2.85 � 0.85 0.5Systolic blood pressure,

mmHg, mean � SD 129.7 � 19.8 126.9 � 28.9 0.4Pulse, bpm,mean �

SD 85.6 � 17.7 85.8 � 20.4 0.9Diagnostic: CT scan,

% 66.7 57.7Imaging: ultrasound

scan, % 23.1 35.9Both, % 0.2 6.5Perforation on

imaging, % 13.6 10.1 0.5Surgery performed, % 83 88 0.2

ACCESS, acute care emergency surgery service; CTAS, Canadian Triage andAcuity Scale.

Acuit

a

N

289Vol. 213, No. 2, August 2011 Qureshi et al Acute Care Surgery and Emergency Department Crowding

crease the wait times for appendectomy for these patients.The mean time from treatment decision to surgery wasabout 8 hours in both groups, as discussed previously. Asummary of the results can be found in Table 3.

Time-to-stretcher: any emergency departmentpatientDuring the same study period, we noted a decrease in EDcongestion, as measured by time-to-stretcher. For all CTAS3 patients, the mean time-to-stretcher in the pre-ACCESSperiod was 1.5 hours, which decreased by 20% to 1.2 hoursafter ACCESS implementation (p � 0.01). For any CTASlevel, the mean time from ED registration to stretcher de-creased from 1.1 hours in the pre-ACCESS group to 0.9hours in the ACCESS cohort (p � 0.01).

Data qualityAfter ACCESS was implemented, we prospectively col-lected data on 74 patients with appendicitis who also hadoverlapping EDIS data. The mean wait time between con-sult request and consultant arrival was significantly longer

Table 3. Determinants of Emergency Department Length of

Key time intervalsPre-ACC

(n �

Registration to ED MD assessment 275ED MD assessment to consult request 224Registration to return of CBC result 169Registration to urine lab results 342Registration to CT preliminary results 555Registration to US preliminary results 477Consult request to decision 338Decision to OR 490Surgical decision time 1,046

Data shown as mean � SD.ACCESS, acute care emergency surgery service; CTAS, Canadian Triage and

Table 4. Outcomes of Emergency Department Patients Di-gnosed with Appendicitis

OutcomePre-ACCESS(n � 169)

ACCESS(n � 136) p Value

Perforation noted atoperation, n 0.40

Yes, n (%) 23 (16.5) 15 (12.8)No, n 116 102o perforation on

CT butperforationnoted atoperation, n 11 8

Chi-square �0.019(p � 0.89)

Length of stay(registration todischarge), h,

mean �SD 49.5 � 28.7 66.9 � 53.2 0.13

on the EDIS database when compared with the generalsurgery prospective database (91 � 32 minutes vs 27 � 21minutes, p � 0.01). Similarly, the time from general sur-gery consult arrival to decision was significantly longerwhen measured using the EDIS database vs the generalsurgery database (111 � 61 minutes vs 43 � 34 minutes,p � 0.01).

DISCUSSIONEmergency departments serve millions annually as a crucialpoint of access to the health care system, and the initialtreatment for broad spectrum of injuries and illnesses.18

ED overcrowding is the situation in which the number ofpatients in the ED awaiting assessment and treatment ex-ceeds the capacity of the ED. Overcrowding in the ED hasresulted in compromised care to patients requiring urgenttherapy because of prolonged time to therapy, medical er-rors, and ambulance redirection.2 Consequently, ED over-crowding has been described as an “emerging threat topatient safety and public health.”6

We performed a single-center pre- and post-study todetermine the effect of implementing ACCESS on key de-terminants of ED LOS, ED overcrowding, and on patientoutcomes. We found that implementing ACCESS reducedsurgical decision time by 15% for all patients requiringgeneral surgery consultation and by 30% for patients withappendicitis. During the same study period, we did notobserve any improvements in decision time for patientsrequiring internal medicine or urology consultation. Fur-thermore, implementation of ACCESS was associated withan overall improvement in ED crowding; we observed an18% reduction in time-to-stretcher for any ED patient,irrespective of CTAS level.

Improving emergency department overcrowdingAccording to the Canadian Association of Emergency Phy-

y for Patients with Appendicitismin ACCESS, min

(n � 136) p Value

290 � 267 0.202 200 � 129 0.20

122 � 39 0.171 280 � 156 0.016 468 � 246 �0.029 420 � 271 0.339 190 � 156 �0.011 472 � 377 0.40

708 � 362 �0.01

y Scale; ED, emergency department; OR, operating room; US, ultrasound.

StaESS,169)

� 94� 20� 67� 21� 22� 35� 23� 40�399

sicians, ED overcrowding is a multifactorial phenomenon

c

sioetfaOsf

ttmdo

290 Qureshi et al Acute Care Surgery and Emergency Department Crowding J Am Coll Surg

that reflects complex systemic problems within the healthcare system.1 In order to better understand the key factorsontributing to ED overcrowding, Asplin and colleagues19

developed a conceptual model of ED overcrowding. Thismodel partitioned ED overcrowding into 3 interdependentcomponents: input, throughput, and output. The inputcomponent refers to factors that contribute to the volumeof care delivered in the ED, which can be categorized asemergency care, unscheduled urgent care, and safety-netcare. The throughput component refers to factors that con-tribute to the amount of time a patient spends in the ED.This part of the model highlights the importance of EDcare processes and the need to modify them to improvetheir efficiency and effectiveness. The output componentreflects the disposition of ED patients. The 2 main optionsare admission to a hospital bed or discharge. The inabilityto move patients from the ED to an inpatient bed is con-sidered one of the major contributing factors to EDovercrowding.

In a recently published systematic review, Cooke andcolleagues20 reviewed interventions that have been demon-trated to reduce ED waiting times. They found that theres surprisingly little high-level evidence on the effectivenessf changes in service delivery and organization factors inmergency care on ED wait times. When relating the in-erventions back to Asplin’s conceptual framework, Cookeound that the majority of studies evaluated strategies thatddressed only factors associated with ED throughput.verall, only a few of these single-faceted interventions,

uch as the use of point-of-care laboratory testing21 and aast-track system for minor illness and injury22 were sup-

ported by evidence from randomized controlled trials toreduce ED overcrowding. Lower quality studies, however,support the concept that changes to ED organizational andstaffing models can reduce ED overcrowding.23,24

Acute care surgery serviceThe acute care surgery service represents a change in theorganization and staffing model of general surgery servicesacross North America. In the past 5 years, there has been agroundswell of support in both Canada and the UnitedStates for establishment of these services, albeit for differentreasons. In the United States, acute care surgery developedpartially in response to the growing difficulty involved withcaring for patients with acute surgical conditions, the de-crease in operative trauma surgical cases, and the decreasingavailability of surgeons to cover emergency call.25 In con-rast, acute care surgery services have developed in Canadao address the difficulties that general surgeons face in si-ultaneously balancing emergency general surgery on-call

uties with the usual demands of scheduled surgery and

utpatient clinics.12,13 Therefore, in Canada, acute care sur-

gery services have focused primarily on developing modelsfor clinical service delivery, but also, secondarily, on im-proving education and research. In contrast, US-basedacute care surgery programs have focused on developing atraining and career model that aims to produce a specialistwith expertise in trauma surgery, surgical critical care, andelective and emergency general surgery.

This study is the first to show that establishment of anacute care surgery service can improve overall ED over-crowding by decreasing surgical decision time for all gen-eral surgery patients. Our data show that our ACCESSservice had its primary effect on ED overcrowding by im-proving ED throughput of surgical patients; ACCESS re-duced the time from consultation request to final decision.However, the ACCESS service may also have improved EDovercrowding by improving output; the ACCESS team isresponsible for caring for in-patients admitted from theED, and may have reduced their overall length of stay.Although our data from our appendicitis subgroup did notshow any significant change in LOS, it is possible that LOSwas improved for surgical patients with other diagnoses.Unfortunately, we did not review data for overall LOS forall surgical patients. This may be a focus for future studies.

Several other different groups have reported their expe-rience with acute care surgery services, but have reportedonly on the effect of their programs on patients with ap-pendicitis. Britt and colleagues14 and Earley and associ-ates15 independently studied patients with appendicitis andfound that their acute care surgical services reduced thetime spent in the emergency room, reduced appendicealperforation rates, and reduced hospital LOS for these pa-tients. In contrast, Ekeh and colleagues26 found no suchdifference in emergency room times or perforation rates ofpatients with appendicitis after implementation of theiracute care surgery program.

Areas for improvementIn our appendicitis subgroup, we found that implementa-tion of ACCESS reduced only the response time of thegeneral surgical team to consult requests and the time forthe surgical team to make a final treatment decision. Thetime from ED registration to ED physician assessment didnot significantly change. Patients with suspected appendi-citis waited longer than 8 hours before an ED physicianrequested general surgery consultation. Perhaps this delaywas due to the wait for diagnostic imaging results becausepatients waited approximately 7 hours from time of regis-tration before CT scan preliminary results were available.Other studies have made similar observations. Both Yoonand coworkers27 and Qureshi and associates28 have re-ported that ED length of stay is linked to triage level

(CTAS 3 and 4), performance of laboratory investigations

esa

tmttga

epimE

291Vol. 213, No. 2, August 2011 Qureshi et al Acute Care Surgery and Emergency Department Crowding

and imaging studies, and the need for consultation. Onestrategy for reducing ED LOS for patients requiring surgi-cal consultation in the future may be expediting laboratoryand diagnostic imaging services.

Outcomes in patients with appendicitisAs discussed previously, there is some controversy in theliterature as to whether or not implementation of an acutecare surgical service improves outcomes in patients withappendicitis.14,15,26 In our study, we did not observe anyimprovement in outcomes in these patients. Most notably,the perforation rate (noted at surgery) did not change sig-nificantly. As well, the hospital LOS was not shortened inthe ACCESS cohort. Previous work suggests that the like-lihood of appendiceal perforation increases with delays inpresenting to the ED from onset of symptoms.29 The risk ofperforation or phlegmon has also been shown to increasefrom 3% to 27% when an appendectomy cannot be donewithin 12 hours from onset of symptoms.30 Furthermore,when appendectomy was delayed for more than 71 hoursfrom onset of symptoms, the odds for additional progres-sive pathology (periappendicular abscess and gangrenousacute appendicitis) increased 13-fold.29 One reason for thelack of improvement in our patient outcomes is that im-plementation of ACCESS did not significantly improveour time to surgery. This may have been due to operatingroom availability. At our hospital, appendectomies arebooked as a “B” priority, meaning they should be donewithin 2 to 6 hours of booking. This did not change withimplementation of ACCESS.

We did find that approximately 11% of patients pre-sented with nonperforated appendicitis on initial imagingand progressed to perforation while waiting for surgery.Because the presence of a fecalith was associated with pro-gression to perforation in this group of patients, we recom-mend that patients with a fecalith and nonperforated ap-pendicitis on initial diagnostic imaging have theiroperation triaged to become an “A” case (within 2 hours).Our finding is supported by other studies in the literaturethat show that the presence of a fecalith is associated withearly appendiceal perforation.31

LimitationsOne limitation of our study is our selection of time-to-stretcher as our measure of ED overcrowding. Althoughemergency health care providers and administrators mayhave an intuitive sense of when an ED is becoming over-crowded, there is no consensus on what overcrowding is orhow it can best be measured.32 Furthermore, many differ-nt indicators may be applicable to different circum-tances.33 Even so, the Hospital Emergency Department

nd Ambulance Effectiveness Working Group responded

o ED overcrowding by recommending that key perfor-ance indicators should include time from ED registration

o stretcher (room time), and subsequent time to disposi-ion to either admission or discharge.34 Drummond8 ar-ued that time-to-stretcher is one of the important factorssociated with ED overcrowding.

Another limitation of this study is the difficulty instablishing a cause-effect linkage between ACCESS im-lementation and improvements in overall ED crowd-ng. ED overcrowding has many causes. Clearly, imple-

entation of acute care surgical services may not affectD overcrowding in all hospitals. In fact, Drummond8

identified the presence of admitted patients in the ED asthe single most significant factor contributing to EDovercrowding, thereby de-emphasizing the role of im-proving efficient ED patient flow for improving EDovercrowding. Implementation of ACCESS, however,may have improved patient flow from admission to dis-charge, improving ED overcrowding by a differentmechanism.

Another limitation of this study is the quality of thedata. Our ACCESS prospective data may have underes-timated ED time intervals because residents may havebeen motivated by the ACCESS attending to hit EDwait-time targets. Conversely, ED data may have over-estimated ED time intervals because ED staff may havebeen unaware that the ACCESS team had arrived tostart the consult, or that the ACCESS team had madetheir final decision and admitted the patient. This lim-itation speaks to a larger problem of communicationbetween services in the ED. ED overcrowding can par-tially be improved by improving certain efficiencies incommunication, such that the ED staff is notified im-mediately that a surgical patient has been admitted, orthat physicians are alerted of pending laboratory andradiologic results to make their decisions regardingtreatment and disposition.

Finally, data from this study were from a single largeacademic institution that may not be reflective of otherinstitutions. This study needs further validation throughreplication in other hospitals to make the results morebroadly generalizable. However, the care processes im-plemented at our center for evaluation and treatment ofgeneral surgical problems are likely to be similar acrossother centers. As a result, the potential improvements wesuggest likely will be similarly effective, if implemented,in other centers.

CONCLUSIONSWe found that implementation of an acute surgical service

at our institution reduced key determinants of ED LOS for

292 Qureshi et al Acute Care Surgery and Emergency Department Crowding J Am Coll Surg

patients requiring general surgery consultation, and im-proved overall ED crowding as measured by time-to-stretcher for ED patients. Additional studies are necessaryto quantify the overall benefits of acute care surgery servicesto all ED patients. Future efforts should focus on improv-ing the accuracy of wait time data, expediting consult re-quests for surgical services by ED physicians, expeditingdiagnostic imaging, and providing more timely access tothe operating room.

Author ContributionsStudy conception and design: Tien, Smith, Wright, Bren-

neman, Rizoli, QureshiAcquisition of data: Qureshi, HsiehAnalysis and interpretation of data: Qureshi, TienDrafting of manuscript: Qureshi, TienCritical revision: Qureshi, Smith, Brenneman, Wright, Ri-

zoli, Tien, Hsieh

REFERENCES

1. Canadian Association of Emergency Physicians (CAEP) (2010)Emergency Department Overcrowding Statement. Available at:http://www.caep.ca/template.asp?id�C66C924CF39543D9AF7E62BAB1A67835. Accessed August 5, 2010.

2. Canadian Association of Emergency Physicians (CAEP). (2000)Position Statement – Emergency Department Overcrowding.Available at: http://www.caep.ca/template.asp?id�1D7C8FEB2A7C4A939E4C2FE16D654E39. Accessed January 9, 2010.

3. Hoot NR, Aronsky D. Systematic review of emergency depart-ment crowding: causes, effects, and solutions. Ann Emerg Med2008;52:126–136.

4. Kulstad EB, Kelley KM. Overcrowding is associated with delaysin percutaneous coronary intervention for acute myocardial in-farction. Int J Emerg Med 2009;2:149–154.

5. Schull MJ, Szalai JP, Schwartz B, Redelmier DA. Emergencydepartment overcrowding following systematic hospital restruc-turing: Trends at twenty hospitals over ten years. Acad EmergMed 2001;8:1037–1043.

6. Trzeciak S, Rivers EP. Emergency department overcrowding inthe United States: an emerging threat to patient safety and pub-lic health. Emerg Med J 2003;20:402–405.

7. Andrulis DP, Kellermann A, Hintz EA, et al. Emergency depart-ments and crowding in United States teaching hospitals. AnnEmerg Med 1991;20:980–986.

8. Drummond AJ. No room at the inn: overcrowding in Ontario’semergency departments. CJEM 2002;4:91–97

9. Statistics Canada. Population by year, by province and territory(2009). Available at: http://www40.statcan.gc.ca/l01/cst01/demo02a-eng.html. Accessed June 12, 2010.

10. Ontario Ministry of Health and Long-Term Care1 (OMH-LTC) (2010). Ontario wait times. Available at: http://www.health.gov.on.ca/en/pro/programs/waittimes/edrs/default.aspx. Accessed June 12, 2010.

11. Tang N, Stein J, Hsia RY, et al. Trends and characteristics of USemergency department visits, 1997-2007. JAMA 2010;304:

2323–2430.

12. Hameed MS, Brenneman FD, Ball CG, et al. General surgery2.0: the emergence of acute care surgery in Canada. Can J Surg2010;53:79–83.

13. Ball CG, Hameed SM, Brenneman FD. Acute care surgery: anew strategy for the general surgery patients left behind. CanJ Surg 2010;53:84–85.

14. Britt RC, Weireter LJ, Britt, LD. Initial implementation of acutecare surgery model: implications for timeliness of care. J AmColl Surg 2009;298:421–424.

15. Earley AS, Pryor JP, Kim PK, et al. An acute care surgery modelimproves outcomes in patients with appendicitis. Ann Surg2006;244:498–504.

16. Accessibility Working Group (2009) Sunnybrook Health Sci-ences Centre Accessibility Plan. Universal access & a culture ofinclusion. Available at: http://www.sunnybrook.ca/uploads/Accessibility_Plan_SB_2009.pdf. Accessed June 12, 2010.

17. Beveridge R, Clarke B, Janes L, et al. Canadian emergency de-partment triage and acuity scale: implementation guidelines.Can J Emerg Med 1999;1: S1–24.

18. Rowe B, Bond K, Ospina M, et al. Data collection on patients inemergency departments in Canada. Ottawa: Canadian Agency forDrugs and Technologies in Health 2006;Technology report no 67.2.

19. Asplin BR, Magid DJ, Rhodes KV, et al. A conceptual model ofemergency department crowding. Ann Emerg Med 2003;42:173–180.

20. Cooke M, Fisher J, Dale J, et al. Reducing attendances and waitsin emergency departments. A systematic review of present inno-vations. A report to the National Co-ordinating Centre for NHSService Delivery and Organisation R & D (NCCSDO); Lon-don, UK; 2004. Available at: http://www.sdo.lshtm.ac.uk/pdf/evalmodels_cooks_final.pdf. Accessed March 17, 2011.

21. Murray RP, Leroux M, Sabga E, et al. Effect of point of caretesting on length of stay in an adult emergency department.J Emerg Med 1999;17:811–814.

22. Cooke MW, Wilson S, Pearson S. The effect of a separate streamfor minor injuries on accident and emergency department wait-ing times. Emerg Med J 2002;19:28–30.

23. Bucheli B, Martina B. Reduced length of stay in medical emer-gency department patients: a prospective controlled study onemergency physician staffing. Eur J Emerg Med 2004;11:29–34.

24. Kelen GD, Scheulen JJ, Hill PM. Effect of an emergency de-partment (ED) managed acute care unit on ED overcrowdingand emergency medical services diversion. Acad Emerg Med2001;8:1095–1100.

25. Davis K, Rozycki G. Acute care surgery in evolution. Crit CareMed 2010;38:S405–410.

26. Ekeh AP, Monson B, Wozniak CJ, et al. Management of acuteappendicitis by an acute care surgery service: is operative inter-vention timely? J Am Coll Surg 2008;207:43–48.

27. Yoon P, Steiner I, Reinhardt G. Analysis of factors influencinglength of stay in the emergency department. Can J Emerg Med2003;5:155–161.

28. Qureshi A, Morreale M, Klisowsky D, Lock J. Presentingwith chest or abdominal pain: evaluation of emergency de-partment wait-time internals and factors influencing lengthof stay. McMaster University Med J (MUMJ) 2010;7:19–25.

29. Temple CL, Huchcroft SA, Temple WJ. The natural history ofappendicitis in adults. A prospective study. Ann Surg 1995;221:

278–281.

3

3

293Vol. 213, No. 2, August 2011 Qureshi et al Acute Care Surgery and Emergency Department Crowding

30. Ditillo MF, Dziura JD, Rabinovici R. Is it safe to delay appen-dectomy in adults with acute appendicitis? Ann Surg 2006;244:656–660.

1. Alaedeen DI, Cook M, Chwals WJ. Appendiceal fecalith is as-sociated with early perforation in pediatric patients. J PediatrSurg 2008;43:889–892.

2. Hwang U, Concato J. Care in the emergency department: how

crowded is overcrowded. Acad Emerg Med 2004;11:1097–1101.

33. Ospina MB, Bond K, Schull M, et al. Key indicators of over-crowding in Canadian emergency departments: a Delphi study.CJEM 2009;9:339–346.

34. Hospital Emergency Department and Ambulance EffectivenessWorking Group (HED&AE) (2005). Available at: http://www.health.gov.on.ca/english/public/pub/ministry_reports/emerg_dept_05/emerg_dept_05.pdf. Accessed March 18,

2011.