Article

Impact of wait time during a first pediatric emergency room visit on likelihood of revisit in the next year

a b s t r a c t

Objective: Pediatric emergency departments (ED) develop strategies to decrease wait time. Yet, lowering wait times may incite patients to come back, and increase patient volume. We aim to determine if wait time in a first visit influenced the likelihood of a revisit to the same setting.

Methods: We performed a retrospective cohort study of children with a first visit to a single pediatric ED between November 1st 2016, and October 31st 2017. First visit was defined as no visit in the same ED in the previous 12 months. The primary outcome was the occurrence of a revisit at the same ED within 12 months of the first visit. Our main predictor was the wait time at the first visit for primary evaluation by a physician. We used mul- tivariable logistic regression models to adjust for potential risk factors (age, triage level, day of visit and disposi- tion).

Results: Among 85,844 ED visits during the study period, 36,844 were first visits and fulfilled Inclusion/exclusion criteria. Median wait time was 101 min (interquartile range: 56-177 min). Among those with a first visit, 11,351 (30.8%) had a revisit within 12 months. In multivariable analysis, each one hour increase in wait time was asso- ciated with a lower probability of revisit (OR: 0.92; 95% CI: 0.91-0.94).

Conclusions: Shorter wait time was associated with higher likelihood of a revisit to the same ED in the following 12 months. Strategies to reduce wait times should take into consideration possible concomitant increase in pa- tient volume.

(C) 2019

Introduction

Emergency room crowding is associated with poorer quality of care [1,2]. Wait time is an important predictor of crowding, and increasing waiting time has been associated with an increase in the proportion of patients leaving the ED without being seen by a physician (LWBS) [3]. There is very little information regarding the outcome of patients who LWBS. pediatric studies suggest that approximately 50% of these pa- tients visit a physician within the following days [4] and approximately 2% have an unfavorable outcome [5].

Abbreviations: ED, emergency department; IQR, interquartile range; LWBS, left without being seen; OR, odds ratio.

? Preliminary data will be presented at the Pediatric Academic Societies conference in

Baltimore (MD) in April 2019, the Canadian Association for Health Services and Policy Research in Halifax (NS), May 2019 and at the Society of Academic Emergency Medicine conference in Las Vegas (NV) in May 2019.

* Corresponding author at: Division of General Pediatrics, CHU Sainte-Justine, 3175 Chemin de la Cote-Sainte Catherine, Suite 7939, Montreal, Quebec H3T 1C5, Canada.

E-mail address: [email protected] (O. Drouin).

Consequently, reducing wait time is often a major objective in emer- gency department (ED) reorganization and an important metric tracked by payers and policy-makers [6,7]. Initiatives such as dedicated rapid as- sessment zones and fast track have the explicit goal of reducing wait times, especially for lower acuity patients [8-11]. While this is a laudable goal, it has proven very difficult to achieve with frequent concomitant increase in patient volume; begging the question as to whether those two metrics are related [7,12].

Dedicated pediatric emergency departments (PED) historically have had lower wait times than general emergency departments, and, at least anecdotally, are perceived by parents as being associated with better quality care [12]. Shorter wait times in ED are associated with higher satisfaction [13-17], which in turn is linked to increased willingness to return to the same ED [18-20]. As such, it is possible that once patients and families have experienced lower wait times in PED, they may be more likely to consult the same PED again in the future.

The main objective of this study was to determine if wait time during a first visit to a pediatric ED is associated with the likelihood that the pa- tients return to the same ED within the following year. Our secondary

https://doi.org/10.1016/j.ajem.2019.07.005

0735-6757/(C) 2019

objective was to determine if children who leave the ED without being seen by a physician are at lower probability of returning to the same ED during the next year.

Methods

Study design and setting

We conducted a retrospective cohort study using the computerized database of a single, university-affiliated, tertiary care pediatric ED with an annual census of approximately 80,000 patient-visits. This pediatric ED is one of two dedicated pediatric ED in a city of approximately 2 mil- lion people and operates in a single (public) payer system. Included par- ticipants visited the ED between November 1st 2016 and October 31st 2017. However, to assess eligibility of participants and have a follow- up of 12 months, all visits that occurred between November 1st 2015 and October 31st 2018 were included in the analysis.

Participants

All children aged 18 years or less who consulted the ED for a first (index) visit in the ED during the study period was eligible. First visit was defined as a visit without another ED visit in the previous 12 months. Patient triaged level 1 (resuscitation) and 2 (emergent) ac- cording to the Canadian Triage and Acuity Scale [21] were excluded be- cause they are sicker and have minimal waiting time. In addition, many of them have underlying chronic disease (ex: cancer or sickle cell dis- ease) that put them at very high risk of revisit to the ED.

Patients who registered in the ED but left before being seen by a phy- sician were excluded from the primary analysis as we could not evalu- ate their wait time to be seen (the independent variable). However, those patients were included in a secondary analysis.

Primary outcome

The primary outcome was the occurrence of a revisit to the same ED during the 12 months following the index visit, but excluding visits that occurred in the 48 h following the index visit. This criterion was added to eliminate bounce-back and planned follow-up in the ED.

Exposures

The main exposure of interest was the wait time before being seen by a physician at the index visit. This was defined as the duration of time between the patient’s arrival in the ED and the first evaluation by a physician. Our secondary exposure was whether the patient had Left without being seen by a physician.

Covariates

Covariates available in the database and included in the analysis in- cluded age of the patient, Canadian Triage and Acuity Scale (CTAS) [21] triage level at the index visit, chief complaint (divided into 35 broad cat- egories in our computerized database), final disposition, family living in proximity to the hospital (b5 km), and day of the week of the index visit (weekday vs. weekend/holiday).

Analysis

First, we evaluated whether wait time was correlated with the prob- ability of a revisit to the same ED. For this analysis, we used unadjusted logistic regression using the presence of a revisit as the outcome of in- terest and wait time as the main (continuous) predictor. We also ex- plored whether the use of wait time as an ordinal variable influenced the relationship between the exposure and the outcome, and whether stratification by triage level changed the relationship between wait time and the likelihood of a revisit. We then created a multivariable lo- gistic regression model using wait time as the main (continuous) pre- dictor, to further adjust for potential confounding variables (see above). This model allowed exploring whether any of the covariates was significantly associated with the outcome. We also conducted a stratified analysis to determine whether wait time had a differential im- pact on the odds of a revisit by triage level. Finally, we explored the probability of revisit for children who left without being seen in com- parison to children who were evaluated by a physician using logistic re- gression. All analyses were conducted using SPSS. For all analyses, we used the standard threshold of p b 0.05 for statistical significance.

This study was approved by the ethics review board of the CHU Sainte-Justine on December 27th 2018. Consent was waved by the institution’s review board since we used a de-identified database.

Fig. 1. Flow of participants.

Results

Between November 1st 2016 and October 31st 2017, there were 85,136 visits to the emergency department. Among them, 36,844 pa- tients fulfilled the inclusion/exclusion criteria (Fig. 1). The most com- mon reasons for exclusion were not a first visit (i.e. the patient had another visit in the previous 12 months) (N = 43,560), triage level 1 or 2 at index visit, (n = 7861) and patient LWBS (n = 6807). Baseline demographics of all patients, and participants included in the analysis are shown in Table 1. There was no missing data for the children in- cluded in the primary analysis. Briefly, 54% of the participants were aged less than five years old, and the most common chief complaints were fever, respiratory problems and trauma. Five percent of the chil- dren were hospitalized and 4% children returned to the ED in the follow- ing 48 h. The median wait time was 101 min (interquartile range (IQR) 56-177 min). Wait time differed by triage level (median of 83, 124, and

Table 1

Baseline demographics.a

95 min for triage level 3, 4 and 5 respectively). Median length of stay in the emergency department was 211 min (IQR 130-322 min).

Of the 36,844 participants included in the study, 11,352 (30.8%) returned to the ED in the following 12 months (excluding the first 48 h). Using univariate analysis, wait time was found to be a significant predictor of revisit, with an odds ratio (OR) of 0.91 (95% CI: 0.90-0.92) for each additional hour of wait time (Table 2). In other words, the pro- portion of patients with a revisit increased as wait times decreased; from 23% for children who waited for more than 4 h, up to 33% for chil- dren who waited less than 1 h (Appendix 1).

When analyzed independently of wait time, younger age was also associated with a higher likelihood of revisit with an odds ratio of 4.3 (95% CI: 3.8-4.7) for children b 3 months compared to children aged 12-18 years old (Table 2). Patients who were transferred to another clinic or hospital had a lower probability of revisit compared to all other dispositions while triage level and type of day of first visit were not statistically associated with a risk of revisit (Table 2).

In multivariable logistic regression, each additional hour of waiting time was associated with a 8% lower odds of having a revisit (OR: 0.92; 95% CI: 0.91-0.94) after adjusting for age, day of visit, triage

All patients (N = 85,136)

Included participants (N = 36,844)

level, disposition, and whether the family lived in proximity to the hos- pital. In addition to shorter wait times, patients who had greater odds of having a revisit were younger, had a higher triage level, lived in proxim-

n % n %

ity to the hospital, and had a different disposition than transfer to an-

Age group

b3 months

5155

(6.1)

2345

(6.4)

other institution (Table 2).

In a stratified analysis by triage level, wait time remained a signifi-

3-12 months

11,506

(13.5)

4441

(12.1)

cant predictor of the odds of having a revisit for each triage level evalu-

ated (3, 4 and 5) (Appendix 2).

1-3 years

24,991

(29.4)

8083

(21.9)

3-5 years

12,127

(14.2)

5027

(13.6)

5-12 years

21,275

(25.0)

11,438

(31.0)

12-18 years

Chief complaint

10,082

(11.8)

5510

(15.0)

Fever and respiratory

18,482

(21.7)

6346

(17.2)

Trauma

10,565

(12.4)

7380

(20.0)

Fever

10,044

(11.8)

3622

(9.8)

Respiratory

8277

(9.7)

3178

(8.6)

Vomiting, diarrhea

6156

(7.2)

2630

(7.1)

Other

31,612

(37.3)

13,688

(37.2)

Live in proximity to the hospital

22,651

(26.6)

9139

(24.8)

Arrival by ambulance

4530

(5.3)

1871

(5.1)

Day of visit

Holiday

2992

(3.5)

1272

(3.5)

LWBS and probability of return

We also performed a secondary analysis to evaluate whether chil- dren who LWBS had different likelihood of revisit than children who were evaluated by a physician. Table 3 shows that eligible children who LWBS were at lower risk of revisit (OR: 0.63; 95% CI: 0.58-0.69) when compared to other children.

Discussion

Week end

22,921

(26.9)

9679

(26.3)

This retrospective database study clearly demonstrates that wait

Week day Season

59,223

(69.6)

25,983

(70.3)

time to see a physician in a pediatric emergency department is a signif-

icant predictor of a revisit, within 12 months, to the same setting. Also

Table 2

Winter

22,335

(26.2)

9313

(25.3)

Spring

21,438

(25.2)

9292

(25.2)

Summer

19,332

(22.7)

8692

(23.6)

Fall Triage level

Level 1

22,031

434

(25.9)

(0.5)

9547

(25.9)

Association between risk factors and return to the same ED.

Level 2

7427

(8.7)

Bivariate analysis

Multivariate analysis

Level 3

38,955

(45.8) 17,841

(48.4)

Waiting time (in h)

0.91 (0.90-0.92)

0.923 (0.91-0.94)

Level 4

33,677

(39.6) 17,603

(47.8)

Age group

Level 5

4154

(4.9) 1400

(3.8)

b3 months

4.26 (3.84-4.73)

4.05 (3.64-4.51)

N/A

489

(0.6)

3-11.99 months

3.55 (3.25-3.88)

3.63 (3.30-3.98)

Wait time, in min

1-2.99 years

2.36 (2.17-2.55)

2.21 (2.01-2.42)

Median (IQR)

89

(46;164) 101

(56;177)

3-4.99 years

1.53 (1.40-1.68)

1.39 (1.26-1.54)

Total length of stay in the ED, in min

Median (IQR) 206 (122;320) 211 (130;322)

Final disposition

5-11.99 years 1.13 (1.04-1.22) 1.11 (1.02-1.21)

12-18 years Ref Ref Triage level

ED: emergency department, IQR: interquartile range.

a Unless indicated, data are expressed as number (percentage) of patients. Percentages have been rounded and may not sum to 100.

Day hospital 2.03 (1.66-2.48) 1.92 (1.46-2.54)

Transfer Ref Ref

Home

68,020

(79.9)

32,983

(89.2)

Level 3

1.06 (0.94-1.19)

0.83 (0.72-0.95)

Hospitalization

6169

(7.2)

1926

(5.2)

Level 4

0.82 (0.73-0.93)

0.90 (0.79-1.04)

Medical day

2280

(2.7)

888

(2.4)

Level 5

Ref

Ref

Transfer to other hospital

1855

(2.2)

1046

(2.8)

Day of visit

LWBS

6807

(8.0)

0

(0.0)

Holiday

1.00 (0.89-1.13)

0.99 (0.86-1.14)

Death

5

(b0.1)

1

(b0.1)

Week end

1.02 (0.97-1.07)

1.01 (0.95-1.07)

Return to the ED 0-2 days

4388

(5.2)

1463

Week day Ref Ref

(4.0) Disposition of first visit

3-7 days

2428

(2.9)

2269

(6.2)

Home

1.66 (1.43-1.93)

1.73 (1.38-2.18)

8-365 days

33,966

(39.9)

10,083

(27.4)

Hospitalization

1.55 (1.30-1.85)

1.54 (1.20-1.98)

Live in proximity to the hospital 1.27 (1.21-1.34) 1.32 (1.25-1.40)

Table 3

Association between patients having Left without being seen and probability of a return visit.

Status

Total number of patients

Patients with a return visit (%)

OR (95% CI)

OR (95% CI)

Multivariatea

LWBS

3,432b

756 (22.0)

0.63 (0.58-0.69)

0.59 (0.54-0.65)

Others

36,844

11,351 (30.8)

Ref

Ref

LWBS: left without being seen.

a Multivariate analysis adjusted for age, triage level, proximity of living and type of day of visit.

b Number of patients who LWBS but met other inclusion criteria (first visit and CATS 3, 4 or 5).

younger age, and living close to the hospital were associated with higher likelihood of having a revisit to the same ED.

Numerous previous studies evaluated the risk of acute return visit (bounce back) [22-25]. Our study differs from them in that we examined medium-term (12 months) revisit, rather than short-term (48-, 72-hour) return visits. Indeed, short-term return visits are related to disease pro- gression, quality of care received during the initial visit, or quality of the healthcare system [22-26]. On the other hand, a second visit beyond this immediate time window may be a sign of good quality of care, as pa- tients are more likely to come back to the same setting if they were satis- fied the first time [20]. Others found that younger age was associated with higher likelihood of short-term return visit, a finding also true in our study, although possibly for different reason [26-29]. Indeed, beyond the first 72 h, the increase in revisit may have more to do with the fact that children in pre-school years tend to have more infectious illnesses that may warrant medical assessment. Living closer to the hospital was also previously linked to higher likelihood of having a revisit [26].

While we found no study linking wait time and the risk of future visit

to the same ED, there are many studies showing that a shorter wait-time is associated to higher satisfaction among parents and patients visiting the ED [13-17]. In addition, other studies demonstrated that higher satisfac- tion was associated to higher willingness to return to the same ED [18-20]. There are limitations to our study. First, as with any study using ad- ministrative data, it is possible that there are errors in recordED wait times or other covariates. We expect the number of such cases to be low, and given the size of our sample, unlikely to influence significantly the direction and magnitude of the effects observed. Also, such random error should bias the results toward an absence of difference between the groups. We chose our index visits as being a visit without a preceding visit in the previous 12 months. It is however possible that among those patients, some had previous experience with our hospital, for example by having a visit 2 or 3 years prior. As such, there might be some hetero- geneity in our sample. However, we expect the number of such cases to be relatively small, and the reason for having a revisit (or not) in those pa- tients to be similar as for patients with a “true” lifetime first visit. Our anal- ysis is also limited to a single center, and as such, it is possible that patients had revisits to other institutions, which would not be captured in our analysis. It is therefore impossible to determine whether patients who did not return within 12 months were simply not sick enough to warrant an emergency room visit, or whether they consulted elsewhere. A future study using province-wide health services data would be necessary to dis- criminate between those two possibilities. Finally, other factors may be important in predicting revisits. Families’ satisfaction and experience

with care for example, whether the patient has access to a primary care physician, or whether the patient has medical comorbidities warranting being seen in a tertiary care center, are all potential contributors not avail- able in our database [26,30]. A follow-up study to evaluate the relative contribution of those factors is planned.

Our study has implications for the organization of care in pediatric emergency medicine departments. While efforts to decrease wait times are warranted in an effort to reduce crowding, which is associated with poorer outcomes for patients, we believe it is warranted to con- sider the effect those initiatives may have on the volume of patients being seen in pediatric ED. Rapid assessment zones (RAZ) and fast track of lower acuity patients may help reduce the number of patients with low acuity problems waiting for long periods [31,32], but it is pos- sible that they also send a message that it is appropriate to consult the ED for those problems. This may lead to more patients consulting the ED. We believe that initiatives to increase access to alternate locations of care (such as regular primary care providers office or urgent care) are warranted, especially if wait times at those locations are similar or better than in the ED.

Conclusion

In summary, we found that among patients with a first visit to a pe- diatric emergency department, lower wait times to see a physician was associated with increased likelihood of a revisit in the same setting in the following 12 months. This finding has implications for the organiza- tion of both the pediatric emergency department as well as pediatric ur- gent care in general.

Funding source

This research did not receive any specific grant from funding agen- cies in the public, commercial, or not-for-profit sectors.

Financial disclosure statement

The authors have no financial relationships relevant to this article to disclose.

Declaration of Competing Interest

The authors have no conflict of interest relevant to this article to disclose.

Appendix 1. Association between wait time (ordinal variable) and probability of return visit

Wait time N of patients Patients with a return visit OR (95% CI) OR (95% CI)

n

%

Univariate

Multivariatea

0-60 min

8700

2893

(33.3)

1.71 (1.57-1.85)

1.57 (1.44-1.72)

61-120 min

9411

2961

(31.5)

1.57 (1.45-1.71)

1.44 (1.32-1.57)

121-180 min

5470

1645

(30.1)

1.47 (1.34-1.61)

1.38 (1.26-1.52)

181-240 min

3108

875

(28.2)

1.34 (1.21-1.49)

1.31 (1.17-1.46)

N240 min

4420

999

(22.6)

1.00 ref

1.00 ref

OR: odds ratio.

a Multivariate analysis adjusted for age, triage level, final disposition of first visit, type of day of visit and proximity of living.

Appendix 2. Results of the stratified analysis by triage level, on the association between waiting time (in h) and odds of having a return probability

Triage level

OR (95% CI)

Univariate

OR (95% CI)

Multivariatea

Level 3

0.97 (0.94-0.99)

0.93 (0.91-0.95)

Level 4

0.89 (0.88-0.91)

0.91 (0.90-0.94)

Level 5

0.89 (0.84-0.96)

0.94 (0.87-1.01)

OR: odds ratio.

a Multivariate analysis adjusted for age, final disposition of first visit, type of day of visit and proximity of living.

References

  1. Hoot NR, Aronsky D. Systematic review of emergency department crowding: causes, effects, and solutions. Ann Emerg Med 2008;52(2):126-36.
  2. Chan M, Meckler G, Doan Q. Paediatric emergency department overcrowding and adverse patient outcomes. Paediatr Child Health 2017;22(7):377-81.
  3. Gaucher N, Bailey B, Gravel J. Who are the children leaving the emergency depart- ment without being seen by a physician? Acad Emerg Med 2011;18(2):152-7.
  4. Goldman RD, et al. Patients who leave the pediatric emergency department without being seen: a case-control study. Can Med Assoc J 2005;172(1):39-43.
  5. Gravel J, et al. Unfavourable outcome for children leaving the emergency depart- ment without being seen by a physician. Can J Emerg Med 2013;15(5):289-99.
  6. Wiler JL, et al. Optimizing emergency department front-end operations. Ann Emerg Med 2010;55(2):142-60 [e1].
  7. Cardin S, et al. Intervention to decrease emergency department crowding: does it have an effect on return visits and Hospital readmissions? Ann Emerg Med 2003; 41(2):173-85.
  8. Hampers LC, et al. Fast track and the pediatric emergency department: resource uti- lization and patients outcomes. Acad Emerg Med 1999;6(11):1153-9.
  9. Simon HK, Ledbetter DA, Wright J. Societal savings by “fast tracking” lower acuity patients in an urban pediatric emergency department. Am J Emerg Med 1997;15 (6):551-4.
  10. Simon HK, et al. “Fast tracking” patients in an urban pediatric emergency depart- ment. Am J Emerg Med 1996;14(3):242-4.
  11. Bullard MJ, et al. The role of a rapid assessment zone/pod on reducing overcrowding in emergency departments: a systematic review. Emerg Med J 2012;29(5):372-8.
  12. Doan Q, Genuis ED, Yu A. Trends in use in a Canadian pediatric emergency depart- ment. CJEM 2014;16(5):405-10.
  13. Byczkowski TL, et al. A comprehensive view of parental satisfaction with pediatric emergency department visits. Ann Emerg Med 2013;62(4):340-50.
  14. Sonis JD, et al. Emergency department patient experience: a systematic review of the literature. J Patient Exp 2018;5(2):101-6.
  15. Welch SJ. Twenty years of patient satisfaction research applied to the emergency de- partment: a qualitative review. Am J Med Qual 2010;25(1):64-72.
  16. Handel DA, et al. Associations between patient and emergency department opera- tional characteristics and patient satisfaction scores in an adult population. Ann Emerg Med 2014;64(6):604-8.
  17. Nichol JR, et al. Association between patient and emergency department operational characteristics and patient satisfaction scores in a pediatric population. Pediatr Emerg Care 2016;32(3):139-41.
  18. Lee KJ. A practical method of predicting client revisit intention in a hospital setting. Health Care Manage Rev 2005;30(2):157-67.
  19. Sun BC, Adams JG, Burstin HR. Validating a model of patient satisfaction with emer- gency care. Ann Emerg Med 2001;38(5):527-32.
  20. Sun BC, et al. Determinants of patient satisfaction and willingness to return with emergency care. Ann Emerg Med 2000;35(5):426-34.
  21. Warren DW, et al. Revisions to the Canadian triage and acuity scale paediatric guide- lines (PaedCTAS). Can J Emerg Med 2008;10(3):224-32.
  22. Truong M, Meckler G, Doan QH. Emergency department return visits within a large geographic area. J Emerg Med 2017;52(6):801-8.
  23. Rising KL, et al. Return visits to the emergency department: the patient perspective. Ann Emerg Med 2015;65(4):377-86 [e3].
  24. Augustine EM, et al. A survey of return visits to the pediatric emergency department. the caretakers’ perspective Pediatr Emerg Care 2018;34(1):27-32.
  25. Pittsenbarger Z, et al. Hospital-level factors associated with pediatric emergency de- partment return visits. J Hosp Med 2017;12(7):536-43.
  26. Akenroye AT, et al. Prevalence and predictors of return visits to pediatric emergency departments. J Hosp Med 2014;9(12):779-87.
  27. Goldman RD, Ong M, Macpherson A. UnschedulED return visits to the pediatric emergency department-one-year experience. Pediatr Emerg Care 2006;22(8): 545-9.
  28. LeDuc K, et al. Pediatric emergency department recidivism: demographic character- istics and diagnostic predictors. J Emerg Nurs 2006;32(2):131-8.
  29. Zimmerman DR, et al. Repeat pediatric visits to a general emergency department. Ann Emerg Med 1996;28(5):467-73.
  30. Jacobstein CR, et al. UnschedulED revisits to a pediatric emergency department: risk factors for children with fever or infection-related complaints. Pediatr Emerg Care 2005;21(12):816-21.
  31. Aksel G, et al. Effects of fast-track in a university emergency department through the National Emergency Department Overcrowding Study. J Pak Med Assoc 2014;64(7): 791-7.
  32. Yarmohammadian MH, et al. Overcrowding in emergency departments: a review of strategies to decrease future challenges. J Res Med Sci 2017:22.

Leave a Reply

Your email address will not be published. Required fields are marked *