Archives of Academic Emergency Medicine. 2021; 9(1): e1 https://doi.org/10.22037/aaem.v9i1.891 OR I G I N A L RE S E A RC H Clinical Predictors of Emergency Department Revisits within 48 Hours of Discharge; a Case Control Study Panvilai Tangkulpanich1, Chaiyaporn Yuksen1∗, Wanchalerm Kongchok1, Chestsadakon Jenpanitpong1 1. Department of Emergency Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand. Received: October 2020; Accepted: October 2020; Published online: 5 November 2020 Abstract: Introduction: Emergency department (ED) revisits increase overcrowding and predicting which patients may need to revisit could increase patient safety. This study aimed to identify clinical variables that could be used to predict the probability of revisiting ED within 48 hours of discharge. Methods: A retrospective case-control study was conducted between July 2018 and January 2019 at the Emergency Medicine Department in Ramathi- bodi Hospital, Bangkok, Thailand. Patients who revisited the ED within 48 hours of discharge (case group) and patients who did not (control group) participated. The predictive factors for ED revisit were identified through multivariate logistic regression analysis. Results: The case group consisted of 372 patients, who revisited the ED within 48 hours, and the control group consisted of 1488 patients. The most common reason for revisiting the ED was recurring gastrointestinal illness, in 107 patients (28.76%). According to the multivariate data analysis , five factors influenced the probability of revisiting the ED: age of more than 60 years (p < 0.001, OR = 2.04, 95%CI: 1.51-2.77), initial Emergency Severity Index (ESI) triage level of 2 (p = 0.007, OR = 1.20, 95%CI: 0.93-1.56), ED stay duration of 4 hours or longer (p = 0.013, OR = 1.12, 95%CI: 0.87-1.44), body temperature of ≥37.5◦C on discharge (p = 0.034, OR = 1.34, 95%CI: 1.00-1.80), and pulse rate of less than 60 (OR = 1.55, 95%CI: 0.87-2.77) or more than 100 beats/minute (OR = 1.53, 95%CI: 1.10-2.11) (p = 0.011). Conclusion: According to the findings, the most important and independent predictive factor of ED revisit within 48 hours of discharge were, age ≥ 60 years, ESI triage level 2, ED length of stay ≥ 4 hours, temperature ≥ 37.5 C, and 60 > pulse rate ≥ 100 beats/minute. Keywords: Emergency service, hospital; patient discharge; clinical decision rules; triage; Thailand Cite this article as: Tangkulpanich P, Yuksen C, Kongchok W, Jenpanitpong C. Clinical Predictors of Emergency Department Revisits within 48 Hours of Discharge; a Case Control Study. Arch Acad Emerg Med. 2021; 9(1): e1. 1. Introduction Emergency department (ED) overcrowding is a significant problem in tertiary care hospitals. Many factors contribute to this problem, such as an increase in the number of pa- tients who come to the ED and insufficient beds for inpa- tients. As a result of overcrowding, waiting time for treatment in ED increases, and specific resources are used unnecessar- ily (1). The average time between discharge and revisit to the ED has not yet been defined. Most authors refer to it as 24 to 72 hours (2-5). Revisit within 72 hours is a widely accepted quality assurance tool. The overall probability of revisiting the ED is approximately 3% in most studies and patients who revisit are considered to have a high risk for complications ∗Corresponding Author: Chaiyaporn Yuksen; 270 Rama VI Road, Thung Phaya Thai, Ratchathewi, Bangkok, Thailand, 10400. Tel: (+66)8-9472- 6911, Emails: Chaipool0634@hotmail.com, Panvilai.tan@mahidol.edu, OR- CIDs: http:// 0000-0002-2922-2761 and death (5). Many researchers have analysed the factors involved in revis- iting the ED: those related to the patient’s illness, ED treat- ment, and medical service systems (6-8). Factors such as the nature of the disease, medical error, patient satisfaction (9), and inadequate assessment or treatment in the first ED visit (5) may account for revisits. If the interval is short (3–7 days), the cause may be related to improper treatment, treat- ment errors, or adverse events after treatment (10). A review of these causes is essential for quality assurance of inpatient care. These can be resolved by educating patients about the diagnosis and treatment (11). Predicting which patients are likely to revisit the ED af- ter discharge would help plan and adequately manage re- sources. Clinicians must identify essential variables to assess the probability of readmission after the discharge of a patient (12). Health care organizations have access to comprehen- sive medical histories, convenient for developing and testing tools to assess the probability of revisiting the ED after dis- This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem P. Tangkulpanich et al. 2 charge (13). This study aimed to identify clinical variables that could be used to predict the probability of revisiting the ED within 48 hours of discharge. 2. Methods 2.1. Study design and setting A retrospective case-control study was conducted from July 2018 to January 2019, at the ED of Ramathibodi Hospital, a university-affiliated super tertiary care hospital in Bangkok, Thailand. Patients who revisited the ED within 48 hours of discharge (case group) and patients who did not (control group) were compared and the predictive factors of ED revisit were iden- tified through multivariate logistic regression analysis. The Ethics committee approved this study on Human Rights Related to Research Involving Human Subjects by Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand (Ethics code: MURA2018/377). 2.2. Participants The patients who revisited the ED within 48 hours after dis- charge were recruited to comprise a case group. The patients who did not revisit within 48 hours after discharge were re- cruited to include a control group. We collected the con- trol group from the emergency medical record (EMR) pro- gram. We excluded patients suffering from trauma, patients younger than 15 years of age, and patients who revisited with a different problem. 2.3. Data gathering The following data, including baseline characteristics, clini- cal factors, and possible effective factors of revisit were col- lected: gender, age, initial triage level based on emergency severity index (ESI), chief complaint, first ED visit length of stay (LOS), vital signs before discharge (body temperature, systolic blood pressure, and pulse rate). Patients revisiting ED were defined as patients who returned to the ED with the same related symptom within 48 hours after ED discharge (14). 2.4. Statistical Analysis We used Stata version 12 software (StataCorp, College Sta- tion, TX, USA) to conduct a two-sample comparison of pro- portions with the following assumptions: alpha = 0.05 (one- sided), power = 0.8, and N2/N1 = 4. The sample size that we planned to recruit was approximately 372 patients who revis- ited the ED and 1488 who did not. The possible predictors of revisiting the ED within 48 hours after discharge were compared between the 2 groups to iden- tify differences (p-value) in clinical characteristics using in- dependent t-test and exact probability test. The predictive factors were identified using univariate and multivariate lo- gistic regression analysis and were presented with area under the receiver operating characteristic (AUROC) curve and 95% confidence interval (95% CI) and odds ratio (OR). Calibration of the prediction was presented using Hosmer–Lemeshow goodness-of-fit test. 3. Results 3.1. Baseline characteristics of participants The average number of patients who revisited within 24 and 48 hours in the studied ED was 40 cases per month (0.02% of patients visiting the ED) and 80 cases per month (0.01% of patients visiting the ED), respectively. Of the 1582 patients who revisited the ED during the study period, 1210 were not eligible according to the research criteria (322 were younger than 15 years of age, 218 patients were suffering from trauma, and 670 revisited with a different problem). Therefore, the case group consisted of 372 patients who revisited the ED within 48 hours after being discharged, and the control group consisted of 1488 patients who did not revisit the ED (figure 1). The most common reason for revisiting the ED was recur- ring gastrointestinal illness, in 107 patients (28.76%). Most of the patients revisiting the ED within 48 hours after discharge were in ESI triage level 2 (16%), level 3 (40%), and level 4 (44%), respectively. Table 1 compares the baseline character- istics of patients between revisited and non-revisited cases. A significant difference was observed between groups regard- ing age group (p < 0.001), initial ESI triage level (p = 0.007), first ED visit length of stay (p = 0.013), discharge time body temperature (p = 0.034), and pulse rate (p = 0.011). 3.2. Modelling According to the multivariate regression analysis (table2), five factors influenced the probability of revisiting the ED: age of more than 60 years (p < 0.001, OR = 2.04, 95%CI: 1.51- 2.77), initial ESI triage level of 2 (p = 0.007, OR = 1.20, 95%CI: 0.93-1.56), duration of ED stay of 4 hours or longer (p = 0.013, OR = 1.12, 95%CI: 0.87-1.44), body temperature of ≥37.5◦C on discharge (p = 0.034, OR = 1.34, 95%CI: 1.00-1.80), and pulse rate of less than 60 (OR = 1.55, 95%CI: 0.87-2.77) or more than 100 beats/minutes (OR = 1.53, 95%CI: 1.10-2.11) (p = 0.011). The significant variables were scored considering their coef- ficient measures to divide patients into three groups: those with scores of 0 (low probability of revisiting ED), those with scores of 1 to 7 points (moderate probability of revisiting ED), and those with scores > 7 points (high probability of revisit- ing ED) (table 2). The positive likelihood ratio of score > 7 for returning to the ED within 48 hours after discharge was 1.48 (Table 3). The AUC of the model in predicting the need for ED revisit was 61% (95%CI: 58-67; figure 2). This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem 3 Archives of Academic Emergency Medicine. 2021; 9(1): e1 Table 1: Comparing the baseline characteristics of patients who revisited and those who did not revisit the emergency department (ED) within 48 hours after discharge Variables Revisited (n = 372) Not-revisited (n = 1488) P N % N % Gender Male 166 44.62 592 39.78 0.099 Female 206 55.38 896 60.22 Age (years) ≤40 85 22.85 499 33.53 40-60 87 23.29 382 25.67 <0.001 > 60 200 53.76 607 40.79 Initial triage level Level 2 120 32.26 376 25.27 0.007 Level 3, 4, or 5 252 67.74 1,112 74.73 Underlying diseases Infectious 101 27.15 395 26.55 0.844 Neurologic 23 6.18 170 11.42 0.002 Gastroenterological 107 28.76 360 24.19 0.071 Respiratory 43 11.56 137 9.07 0.167 Cardiovascular 39 10.48 144 9.68 0.627 Endocrinologic 4 1.08 16 1.08 1.000 Haematologic 4 1.08 16 1.08 1.000 Nephrological 28 7.53 72 4.84 0.053 Ob-gynaecologic 13 3.49 38 2.55 0.373 Musculoskeletal 10 2.69 142 9.54 <0.001 Transportation to ED Ambulance 9 2.42 16 1.08 0.072 Walk in 363 97.58 1,472 98.92 First ED visit, length of stay (hours) <4 218 58.60 976 65.59 0.013 ≥4 154 41.40 512 34.41 Discharge body temperature (degrees Celsius) ≥37.5 95 25.54 302 20.30 0.034 Systolic blood pressure (mmHg) <90 0 0 1 0.07 90–140 243 65.32 977 65.66 0.922 ≥140 129 34.68 510 34.27 Pulse rate (bpm) <60 277 74.46 1,211 81.38 60–100 17 4.57 45 3.02 0.011 >100 78 20.97 232 15.59 4. Discussion The risk factors that cause patients to revisit to the ED within 48 hours after discharge were found to be age of more than 60 years, initial ESI triage level of 2 (No patient triage in level 1 revisit because every patient need to admission from ED), 4 hours or longer duration of ED stay in the first visit, body temperature of ≥37.5◦C on discharge, and pulse rate of less than 60 or more than 100 beats/minutes on discharge. The rate of revisit within 48 hours was 0.01% in our study. The result was better than the other studies that reported the 48-hour revisit rate to be 0.19% (4, 7). The patient dis- charge process in Ramathibodi Hospital would begin based on the opinion of the senior emergency resident and emer- gency staff to reduce the revisit rate. Misdiagnosis was the most common cause of ED revisits (14); patient discharge by the experienced emergency staff was the key to lowering the revisit rate. The study by Sri-on (14), which evaluated the risk factors that lead patients to revisiting the ED within 48 hours after dis- charge in Thailand, showed that most patients who revisit were older than 60 years and the majority of causative dis- eases were related to the digestive system; these findings are similar to our results. The first patient screening at triage revealed that patients with an urgent triage level were most likely to revisit within 48 hours (The urgent level was de- fined as ESI level 3 in this study). These findings are different from our results. In our hospital, patients with ESI level 2 are This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem P. Tangkulpanich et al. 4 Table 2: Multivariate regression analysis of predictors of emergency department (ED) revisit within 48 hours after discharge Predictors Odds ratio 95% CI P value Coefficient Score Age (years) <40 Reference – – – 0 40–60 1.39 1.00–1.95 0.052 0.33 3 ≥60 2.04 1.51–2.77 <0.001 0.72 7 ESI Initial triage level Level 2 1.20 0.93–1.56 0.158 0.19 1.5 Level 3, 4 , or 5 Reference – – – 0 First ED visit length of stay (hour) <4 Reference – – – 0 ≥4 1.12 0.87–1.44 0.376 0.11 1 *Temperature ≥37.5◦C No Reference – – – 0 Yes 1.34 1.00–1.80 0.053 0.29 2.5 *Pulse rate (bpm) <60 1.55 0.87–2.77 0.139 0.44 4 60–100 Reference – – – 0 ≥100 1.53 1.10–2.11 0.010 0.42 4 *Vital sign at the time of discharge; ESI: emergeing severity index; OR: Adjusted odds ratio; CI: confidence interval; ED: emergency department. Figure 1: The process of selecting the study population. ED: emergency department. mostly treated in ED, and the length of stay in ED for more than 4 hours because there was no available bed for admis- sion. Some of the patients with ESI level 2 stayed in the ED for more than 4 hours as no bed was available, some of which were discharged by the emergency physician. Many studies used the Identification of Seniors at Risk (ISAR) for identifying elderly patients at risk of adverse outcomes and predicting hospital readmission after ED discharge. However, the ISAR was unable to predict 48 hours revisit (15, 16). This study used the variables that were available be- This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem 5 Archives of Academic Emergency Medicine. 2021; 9(1): e1 Figure 2: Area under the receiver operator characteristic curve of the model for predicting the probability of revisiting the emergency depart- ment within 48 hours after discharge. Table 3: Groups of patients according to scores indicating the probability of returning to emergency department (ED) within 48 hours Probability Score Revisited No revisit PLR 95% CI P N % N % Low 0 26 10.10 230 89.90 0.62 0.42–0.91 0.006 Moderate 1–7 159 17.45 752 82.54 0.85 0.74–0.96 0.004 High >7 187 26.98 506 73.02 1.48 1.31–1.67 <0.001 CI: confidence interval; PLR: Positive likelihood ratio. fore ED discharge, so they can be evaluated by an emergency nurse or doctor to predict ED revisit. Elderly patients, pa- tients who are initially in triage level 2 when they visit the ED, those with a length of stay of more than 4 hours in ED, and those who have had a high body temperature, or abnormal pulse rate before discharge had a high probability of ED re- visits. If a patient’s score is more than 0 points before discharge, special attention should be given; scores of 1 to 7 points in- dicate a moderate probability of revisit to the ED within 48 hours, and scores more than 7 show a high likelihood of re- turning (Table 3). It may be necessary for patients with scores between 1 and 7 to be monitored according to the indica- tions. Patients with high scores should be repeatedly eval- uated before discharge. At the time of discharge, it is nec- essary to make an appointment to regularly monitor the pa- tient within the following 24 to 48 hours to reduce the risk of complications that may result in a return to the ED after discharge. These patients may require more treatments than delivered in the ED. In the ED of our hospital and those of large general hospitals in Thailand, the majority of patients admitted are older than This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem P. Tangkulpanich et al. 6 60 years of age, and during peak periods, patients may spend extra time in the ED, including waiting times. Laboratory examination, investigation of symptoms and treatment may take more than 4 hours; therefore, if the prediction score for probability of revisit to the ED after discharge is used to as- sess a patient before discharge, most patients will have high scores. The generated score may, therefore, be used as part of deciding whether to discharge the patient or not. 5. Limitations There are some limitations to this study. Ramathibodi Hos- pital is a university hospital whose ED treats a large number of patients. The number of inpatient beds is quite limited; therefore, some patients who need to stay in the hospital may have to be referred to other centers for treatment or be dis- charged home. For patients who go to other hospitals, their data may not be available to Ramathibodi Hospital, and so it is impossible to determine certain variables that may be sta- tistically significant. The tool’s predictive accuracy is 61%, which is low. This may be attributable to the fact that our sample was small. A larger population and investigation in multiple educational institu- tions may help increase the predictive accuracy of the tool. 6. Conclusion According to the findings, the most important and indepen- dent predictive factor of ED revisit within 48 hours of dis- charge were, age ≥ 60 years, ESI triage level 2, ED length of stay ≥ 4 hours, temperature ≥ 37.5 C, and 60 > pulse rate ≥ 100 beat/minutes. 7. Declarations 7.1. Acknowledgements None. 7.2. Author contribution PT and CY conceptualized the research idea, collected data, performed the analysis and wrote the first draft of the manuscript. CY and WK provided oversight and supervision of the research. WK and CJ contributed in the data acqui- sition, analysis, or interpretation. All authors read and ap- proved the final manuscript. 7.3. Funding/Support None. 7.4. Conflict of interest None. References 1. Fan J-S, Kao W-F, Yen DH-T, Wang L-M, Huang C-I, Lee C- H. Risk factors and prognostic predictors of unexpected intensive care unit admission within 3 days after ED dis- charge. The American journal of emergency medicine. 2007;25(9):1009-14. 2. Hu S-C. Analysis of patient revisits to the emergency de- partment. The American journal of emergency medicine. 1992;10(4):366-70. 3. Keith KD, Bocka JJ, Kobernick MS, Krome RL, Ross MA. Emergency department revisits. Annals of emergency medicine. 1989;18(9):964-8. 4. Abualenain J, Frohna WJ, Smith M, Pipkin M, Webb C, Milzman D, et al. The prevalence of quality issues and adverse outcomes among 72-hour return admissions in the emergency department. The Journal of Emergency Medicine. 2013;45(2):281-8. 5. Cheng S-Y, Wang H-T, Lee C-W, Tsai T-C, Hung C-W, Wu K-H. The characteristics and prognostic predictors of un- planned hospital admission within 72 hours after ED dis- charge. The American journal of emergency medicine. 2013;31(10):1490-4. 6. Pierce JM, Kellerman AL, Oster C. “Bounces”: an anal- ysis of short-term return visits to a public hospital emergency department. Annals of emergency medicine. 1990;19(7):752-7. 7. Wu C-L, Wang F-T, Chiang Y-C, Chiu Y-F, Lin T-G, Fu L-F, et al. Unplanned emergency department revisits within 72 hours to a secondary teaching referral hos- pital in Taiwan. The Journal of emergency medicine. 2010;38(4):512-7. 8. Kelly A, Chirnside A, Curry C. An analysis of unsched- uled return visits to an urban emergency department. The New Zealand medical journal. 1993;106(961):334-6. 9. Nunez S, Hexdall A, Aguirre-Jaime A. Unscheduled re- turns to the emergency department: an outcome of med- ical errors? BMJ Quality & Safety. 2006;15(2):102-8. 10. Jin B, Zhao Y, Hao S, Shin AY, Wang Y, Zhu C, et al. Prospective stratification of patients at risk for emer- gency department revisit: resource utilization and pop- ulation management strategy implications. BMC emer- gency medicine. 2016;16(1):10. 11. Lerman B, Kobernick MS. Return visits to the emer- gency department. The Journal of emergency medicine. 1987;5(5):359-62. 12. Kansagara D, Englander H, Salanitro A, Kagen D, Theobald C, Freeman M, et al. Risk prediction mod- els for hospital readmission: a systematic review. Jama. 2011;306(15):1688-98. 13. Hao S, Jin B, Shin AY, Zhao Y, Zhu C, Li Z, et al. Risk prediction of emergency department revisit 30 This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem 7 Archives of Academic Emergency Medicine. 2021; 9(1): e1 days post discharge: a prospective study. PloS one. 2014;9(11):e112944. 14. Sri-On J, Nithimathachoke A, Tirrell GP, Surawongwat- tana S, Liu SW. Revisits within 48 hours to a Thai emer- gency department. Emergency medicine international. 2016;2016. 15. Asomaning N, Loftus C. Identification of seniors at risk (ISAR) screening tool in the emergency depart- ment: implementation using the plan-do-study-act model and validation results. Journal of Emergency Nursing. 2014;40(4):357-64. e1. 16. de Almeida Tavares JP, Sá-Couto P, Boltz M, Capezuti E. Identification of Seniors at Risk (ISAR) in the emer- gency room: A prospective study. International emer- gency nursing. 2017;35:19-24. This open-access article distributed under the terms of the Creative Commons Attribution NonCommercial 3.0 License (CC BY-NC 3.0). Downloaded from: http://journals.sbmu.ac.ir/aaem Introduction Methods Results Discussion Limitations Conclusion Declarations References