Abstract
Background
Unplanned re-attendance at the emergency department (ED) of patients with systemic inflammatory response syndrome (SIRS) remains unclear, and studies exploring health service factors related to unplanned re-attendance are limited. This study aimed to analyze the 30-day incidence of unplanned re-attendance in the ED following discharge and determine factors affecting unplanned re-attendance, including sociodemographic and clinical characteristics, agent, and the health service delivery system in patients with SIRS.
Methods
This study employed a cross-sectional and prospective cohort study design. The sample comprised 900 patients and 14 ED supervisors. This study was conducted between February 1, 2021 and July 30, 2022 at 14 hospitals in Thailand. Data were collected using a standardized questionnaire and information from medical records.
Results
The majority of the sample met two SIRS criteria (76.4%), of which respiratory rate (85.1%) and heart rate (74.5%) were the most common criteria. Most of the re-attendances were one-time (90.3%), and 30% developed sepsis and septic shock and required inpatient and critical care (45.8%). The unplanned re-attendance incidence was 16%, with an incidence rate of 6 persons per 1,000 persons/day. Middle-level hospitals had a higher re-attendance incidence (24%) than high-level (14.3%) and first-level (12.7%) hospitals. Factors affecting unplanned re-attendance to the ED within 30 days included comorbidities (hazard ratio [HR] = 5.0, 95% confidence interval [CI] = 3.056–8.413, p < 0.001), alcohol use (HR = 6.2, 95% CI = 3.555–10.854, p < 0.001), the model of care in the ED and discharge (HR = 11.1, 95% CI = 2.619–47.499, p < 0.001), and care in the ED and observed symptoms and discharge (HR = 13.8, 95% CI = 3.401–56.167, p < 0.001), which was the highest risk factor for unplanned re-attendance.
Conclusions
The findings of this study indicate a high re-attendance occurrence among patients with SIRS. Both individual characteristics and health service delivery system efficiency play significant roles in influencing unplanned re-attendance. These factors can serve as valuable inputs for crafting policies aimed at enhancing the standard of care provided in EDs and guiding future research endeavors.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12913-025-12371-y.
Keywords: Unplanned Re-Attendance, Emergency Department, SIRS, Situational Analysis
Background
Patients with systemic inflammatory response syndrome (SIRS) are frequently present to the emergency department (ED); in developed countries, 25—35% of adults presenting to EDs have SIRS [1–3]. Moreover, more than 70% of patients are categorized under levels III and IV of the emergency medical index (ESI), which are considered urgent and semi-urgent. Patients are initially diagnosed and treated, and approximately 70% are discharged [2]. Conversely, compared with patients who are categorized under other levels of the ESI, 80% of those with levels III and IV of the ESI are likely to re-attend to the ED [4]. The majority of patients have respiratory infections, pain, and fever, which are similar to the symptoms of SIRS. SIRS or suspected infections is a group of clinical symptoms characterized by two of the following four symptoms: body temperature > 38 °C or < 36 °C; heart rate > 90 beats/min; respiratory rate > 20 cycles/min or partial pressure of carbon dioxide (PaCO2) < 32 mmHg, along with white blood cell count > 12,000 or < 4,000 cells/mm3 or immature (band) forms > 10% [5–7].
Unplanned re-attendance refers to the return of patients to the ED owing to similar symptoms or more severe symptoms within 30 days following discharge. Generally, most of the patients who re-attend exhibit more severe symptoms [8, 9], including sepsis and septic shock, which are symptoms that develop following SIRS. Sepsis and septic shock are severe life-threatening infections [10]. Sepsis and septic shock have global mortality rates of 25%–30%: in developed countries the mortality rates are approximately 30%–33% [11, 12] whilst they can be as high as 60% in underdeveloped and developing countries [13].
In Thailand, sepsis is one of the five leading causes of deaths, with mortality rates of > 30% [14]. In addition, one in three patients with sepsis and septic shock seek treatment and care at the ED [15]: most are afflicted with community-acquired sepsis [16]. Patients with SIRS tend to be diagnose based on the clinical system that are affected (e.g. respiratory, urinary tract), rather than SIRS. Patients with SIRS or suspected infection who have been discharged and return to the hospital with more severe symptoms require more healthcare resources and are more susceptible to death [17]. The healthcare costs tend to be two to threefold higher than those of patients without re-attendance [18].
Re-attendance is a healthcare quality indicator of emergency care; it reflects the quality and safety of the emergency medical services system [19]. For this reason, EDs worldwide should develop their care provision to achieve their goals [20, 21]. Therefore, to more clearly identify the severity of the problem and analyze factors leading to re-attendance at the ED, a situation analysis of the re-attendance of patients with SIRS at the ED is deemed significant. The findings will be beneficial to the healthcare team members to better manage primary prevention and subsequently minimize the prognosis and severity of the infections and reduce re-attendance rates [22].
This study aimed to describe patients with SIRS presenting to ED to determine the 30-day incidence of unplanned re-attendance and identify factors affecting unplanned re-attendance survival time, including sociodemographic and clinical characteristics, agent, and health service delivery system. Furthermore, this study aimed to explore factors (host/person factors, agent factors, and health service delivery system factors) associated with 30-day unplanned re-attendance among patients with SIRS. The intention of this situational analysis was to lead to the development of health service delivery and nursing systems as well as recommendations for policy-making to prevent unplanned re-attendance, which, in turn, helps reduce mortality rates and healthcare costs.
Methods
Study design and setting
This study employed a cross-sectional and prospective cohort study design. As a cross-sectional study, it aimed to identify the sociodemographic and clinical characteristics, agent, and health service delivery system among patients with SIRS. As a prospective cohort study, it aimed to explore the 30-day incidence of unplanned re-attendance and factors associated with 30-day unplanned re-attendance. Owing to the policy of the Ministry of Public Health (MOPH) of Thailand, the study settings included hospitals within area health regions (AHs) 3, 4, and 5 located in the central part of Thailand, comprising seven provinces. In each province, there are hospitals with different capacity levels according to the MOPH criteria as follows: (1) first-level hospitals, (2) middle-level hospitals, and (3) standard-level and advanced-level hospitals (Ministry of Public Health, B.E. 2555). The health service setting was selected in accordance with each of the health areas. Fourteen healthcare settings were selected, encompassing three, three, and eight first-, middle-, and standard- and advanced-level hospitals, respectively. In Thailand, various hospital levels differ in their emergency services in terms of competencies and resources for providing patient care. In emergency cases, patients can go to the nearest hospital to alleviate life-threatening events. However, in non-emergency cases, patients should visit the designated hospital related to their insurance. Checking patient data in another hospital was not possible; therefore, researchers had to draft an official letter and contact the hospital case by case.
Participants
All patients presenting during the study period were considered for inclusion.
The following were the inclusion criteria: (1) male and female patients aged ≥ 18 years; (2) can communicate in the Thai language; (3) received treatment in one of the study EDs between February and July 2021; (4) if cannot provide the needed data or respond to the questionnaires, family members or caregivers were present who can provide data instead of them; (5) discharged to home or observed in the observation unit before discharge to home; and (6) if aged ≥ 65 years and had cognitive impairment, as determined using the Mimi-Cog Thai version 3-item recall with the interpretation of 1–2 items and cannot perform clock drawing or 3-item recall [23], data able to be elicited from their family members or caregivers.
The following were the exclusion criteria: (1) those experiencing trauma-induced SIRS, (2) patients with a revisit to the ED but with no history of the first visit at the research setting or no record of an initial visit, and (3) patients admitted to the study site or transferred to another hospital.
Patients with SIRS
SIRS was defined as presence of at least two of the following criteria: (1) Body temperature > 38 °C or < 36 °C; (2) Heart rate > 90 beats/min; (3) Respiratory rate > 20 cycles/min or PaCO2 < 32 mmHg; and (4) white blood cell count > 12,000 or < 4,000 cells/mm3 or immature forms > 10% [5, 6]. In addition, the doctor had to suspect an infection, as evidenced by the prescription of antibiotics or order of a culture.
Sample size
Patients with SIRS. The sample size was calculated on the basis of the research objectives, including the 30-day incidence rate of unplanned re-attendance among patients with SIRS and survival analysis. The sample size was calculated on the basis of the incidence of unplanned re-attendance when the finite population correction was determined before multiplying it with the design effect. Based on an extensive review of studies, the incidence rate used in the sample size calculation in the present study was 15% of individuals with unplanned re-attendance [24, 25]. Therefore, the minimum sample size was 815 participants. The sample size for the survival analysis multiplied with the design effect, based on a review of literature, predictors and outcomes of re-attendance in older adults discharged from the ED [26]. Subsequently, using the nQuery Advisor program, the hazard ratios (HRs) of the participants with and without factors influencing unplanned re-attendance were calculated, and the sample size was determined to be approximately 800 participants and a 10% attrition rate was add. Ultimately, the sample size was determined to be 900 participants.
Data collection
Following Institutional Review Board (IRB) approval, the researchers recruited a research assistant (RA) from each hospital. who was a full-time nurse working at the ED. The principal investigator provided training on the standardized process to collect data using each instrument.
The researchers or RAs collected data from the participants’ medical records, including demographic and clinical data, diagnoses, medical devices (nasogastric tube, Foley catheter, and surgical drains), smoking habits, alcohol consumption and laboratory examination results. At all sites, vital signs were collected on a form upon arrival of the patients, and any patient with one or more SIRS criteria received a prescription for a blood sample with white cell counts. Illness severity was assessed using the Rapid Emergency Medicine Score (REMS) [27], Chronic health disorders and comorbidities were scored using the Charlson Comorbidity Index (CCI) [28, 29], Subsequently, the researchers collected data regarding the health service delivery system, including access to the ED, time of visit, triage, waiting time, treatment duration, duration at the ED, treatment at the ED, discharge planning, and discharge status (Supplementary File 1). A 30-day prospective follow-up was conducted by consulting data recorded in the medical records during each shift on a daily basis.
Data collection instruments and definitions
Microbiological analysis and report
Microbiological analysis and report on microbiological examination results were included when these examinations were performed by medical technicians within the hospital laboratory.
Centers for Disease Control and Prevention/National Healthcare and Safety Network (CDC/NHSN) surveillance definitions for specific types of infections
A report on the sources of infection identified using the CDC/NHSN surveillance definition for specific types of infections was used to determine the sources of infection [30] that led to SIRS or sepsis in accordance with the nine major bodily systems. The report is standardized and used worldwide for reporting the causes of infection to ensure mutual understanding (Supplementary File 1).
Health service delivery system questionnaire
Data on health service delivery characteristics were collected through interviews with the head of the ED, acting head of the ED or the head of the shift, who were working between January and June, 2021, using the health service delivery system structure interview. This tool, is designed for the situational analysis of ED management and was developed by the World Health Organization [31]; to better suit the contexts of the target groups, the National Institute for Emergency Medicine Thailand was adapted and revised by the researchers (Supplementary File 2). It is a standardized instrument that focuses on the health service delivery system factor of patients with SIRS in the ED.
Mini-Cog Thai version
Cognitive impairment was measured using the Mini-Cog Thai version developed by Borson, Scanlan [32] and was subsequently translated into the Thai language by Trongsakul [23]. This instrument was used for assessing cognitive impairment to conduct dementia screening tests in older adults aged ≥ 65 years. A score of 3 or higher, which indicates a lower likelihood of cognitive impairment, was needed for inclusion.
Re-attendance record form
The 30-day unplanned re-attendance was measured using the re-attendance record form, which was used to record and follow-up on the 30-day unplanned re-attendance, including the date and time of the first visit, date and time of the re-visit, duration between the first and return visits, and final disposition of the patients during the revisit (Supplementary File 3).
A variable called Model of Care was developed. Patients either received the Model of Care or did not. Patient charts that included evidence of the use of SIRS clinical practice guidelines, clinical practice nursing guidelines, and/or evidence-based care management plans were considered as having received the Model of Care.
Patients were further categorized into three groups: as Model of Care with discharge home, Model of Care, discharge to observation unit, then discharge home, or Model of Care, observation in ED, then discharge home. and patients who received the model of care in ED including three model, ED and discharged to home (ED → D/C), ED and observed in the observation unit and discharged to home (ED → OU → D/C), and ED and observed in the ED and subsequently discharged to home (ED → OED → D/C).
Statistical analysis
All variables were reported as frequencies and percentages or means, medians, and standard deviations (SDs). Sociodemographic and clinical characteristics, pathogen characteristics, and the health service delivery systems were compared between the non–re-attendance and re-attendance groups using Chi-square test, non-parametric test, independent t-test, and Fisher’s exact test. Epidemiological data, including the incidence of unplanned re-attendance of patients with SIRS, were analyzed as cumulative incidence (CI) and incidence rate (IR) [33].
Factors related to 30-day unplanned re-attendance involving survival function and median unplanned re-attendance survival free time, with the confidence level of 95% CI using the Kaplan–Meier method and the HR, with the confidence level of 95% CI using the Cox regression analysis.
HR refers to the proportion of the hazard rate of unplanned re-attendance of patients with SIRS seeking services at the ED in the study and comparison groups (with and without factors). Data were analyzed using IBM Statistical Package for the Social Sciences Statistics for Windows, Version 27.0.
Ethical consideration
Ethical certificate approval was received from the Multicenter Research of Mahidol University (project number: IRB-NS2020/48.0211). Moreover, this study received certificate approval and permissions from the director of each hospital.
Results
Study selection
Between February and July 2021, a total of 7,580 ED presentations were noted. Overall, 6,050 patients were excluded owing to age < 18 years, trauma-related injury, lack of suspected infection; or no due to a revisit with no record of an initial visit. Of the remaining 1,530 patients, 630 were excluded because they were admitted to the study site or transferred to other hospitals. Ultimately, 900 patients satisfied the inclusion criteria, including 688, 193, and 19 who met two, three, and four SIRS criteria, respectively (Fig. 1).
Fig. 1.
Flow chart of the study
Sociodemographic characteristics
Of the 900 patients with SIRS, 410 and 490 were male and female patients, respectively. The mean age of the patients was 52.8 (range, 18–95, SD = 21.9) years, > 50% of them had been to primary school and high school, 350 of them had an income of ≤ 5,000 Thailand Baht (137 USD; 1 USD = 36.32 Thailand Baht), and 320 had an income of 5,001–10,000 Thailand Baht (138–275 USD). Approximately 51.9% of the patients were married. One-fourth of the patients were students and worked as company employees. Most of the patients’ health insurance was universal coverage scheme. The sociodemographic characteristics of the study participants are presented in Table 1.
Table 1.
Sociodemographic characteristics
| Sociodemographic characteristics | No. (%) | p-value | ||
|---|---|---|---|---|
|
Overall (n = 900) |
Non–re-attendance (n = 756) |
Re-attendance (n = 144) |
||
| Gendeª | 0.243 | |||
| Male | 410 (45.6) | 338 (44.7) | 72 (50) | |
| Female | 490 (54.4) | 418 (55.3) | 72 (50) | |
| Ageb (years) | < 0.001* | |||
|
Mean (SD) Range (min–max) |
52.8 (21.9) 77 (18–95) |
50.95 (22.2) 77 (18–95) |
62.33 (17.3) 71 (19–90) |
|
| Age group (years) | ||||
| 18–40 | 297 (33.0) | 282 (37.3) | 15 (10.4) | |
| 41–60 | 236 (26.2) | 194 (25.7) | 42 (29.2) | |
| 61–80 | 247 (27.5) | 189 (25.0) | 58 (40.3) | |
| ≥ 81 | 120 (13.3) | 91 (12.0) | 29 (20.1) | |
| Educational levelª | < 0.001* | |||
| Illiterate | 107 (11.9) | 72 (9.5) | 35 (24.3) | |
| Primary school | 323 (35.9) | 266 (35.2) | 57 (39.6) | |
| High school | 255 (28.3) | 227 (30) | 28 (19.4) | |
| Vocational school or Diploma Degree | 104 (11.6) | 80 (10.6) | 24 (16.7) | |
| Incomeª (Thailand Baht) | 0.196 | |||
| ≤ 5,000 | 350 (38.9) | 281 (37.2) | 69 (47.9) | |
| 5,001–10,000 | 320 (25.6) | 198 (26.2) | 32 (22.2) | |
| 10,001–15,000 | 200 (22.2) | 174 (23.0) | 26 (18.1) | |
| 15,001–30,000 | 108 (12.0) | 93 (12.3) | 15 (10.4) | |
| > 30,000 | 12 (1.3) | 10 (1.3) | 2 (1.4) | |
| Marital statusª | < 0.001* | |||
| Married | 467 (51.9) | 372 (49.2) | 95 (66.0) | |
| Single | 278 (30.9) | 247 (32.7) | 31 (21.5) | |
| Divorced/Separated/Widowed | 155 (17.2) | 137 (18.1) | 18 (12.5) | |
| Occupationª | < 0.001* | |||
| Student | 230 (25.6) | 187 (24.7) | 43 (30.0) | |
| Company employee | 183 (20.3) | 139 (18.4) | 44 (30.6) | |
| Self-employed | 158 (17.6) | 138 (18.3) | 20 (13.9) | |
| Retired | 105 (11.7) | 87 (11.5) | 18 (12.5) | |
| Agriculturist | 94 (10.4) | 78 (10.3) | 16 (11.0) | |
| Government employee | 85 (9.4) | 82 (10.8) | 3 (2.0) | |
| Government official | 45 (5.0) | 45 (6.0) | 0 (0.0) | |
| Health insurancesª | 0.005* | |||
| Universal Coverage Scheme | 623 (69.2) | 519 (68.7) | 104 (72.1) | |
| Social Security Scheme | 124 (13.8) | 117 (15.5) | 7 (4.9) | |
| Government | 104 (11.6) | 80 (10.5) | 24 (16.7) | |
| Self-pay | 49 (5.4) | 40 (5.3) | 9 (6.3) | |
ªCategories: comparisons using Chi-square test
bMedian: comparisons using the Mann–Whitney U rank exact test
*p-value is significant
Of the 900 patients presenting with SIRS, 144 (16.0%) of these were representations with a previous presentation for SIRS in the 30 days prior. Older people, those with low education, married people and company employees (compared to government employees) were statistically significantly more likely to reattend.
Clinical characteristics
Two and three SIRS criteria were met by 76.4% and 21.4% of the patients, respectively. Respiratory and heart rates were the most common criteria for SIRS. Fever, fatigue, and dyspnea were the most common chief complaints of patients. The top three signs and symptoms included fever, fatigue, and cough. Most of the patients had symptom onset at 1–2 days (69.8%), whereas others at < 24 h (17.0%).
Body temperatures at the first ED visit in all patients were between 37.8 °C and 39.4 °C. The majority of patients had heart rates of > 100 beats/min, and approximately 50% of them had respiratory rates of 21–28 cycles/min. Moreover, of the patients, 50% had systolic blood pressure between < 120 and 120–139 mmHg, and 50% had a diastolic blood pressure of < 80 mmHg. Most of the patients had a mean arterial pressure of ≥ 70 mmHg. Additionally, most of them had an oxygen saturation of ≥ 95%. Most of the patients had a Glasgow coma scale score of 13–15. Furthermore, most, 35.3%, 26.1%, and 30.4% of the patients had REMSs of 0–9, 0–2, 3–5, and 6–9, respectively.
Of the participants, 52.6% had comorbidities, and 90.6% had a CCI of levels 0–2. In all patients, hypertension, diabetes without end-organ damage, and chronic obstructive pulmonary disease (COPD)/asthma were the top three comorbidities, respectively. Based on the International Classification of Diseases, Tenth Revision (ICD-10), the top three diagnoses were R50.9 (fever), A49.9 (bacterial infection), and N39.0 (urinary tract infection), respectively. A few patients (4.3%) had medical device insertion; Foley catheters were the most frequently used medical devices (55.3%).
Regarding smoking habits, 760 (84.4%) patients never smoked. Moreover, 87 (9.7%) patients were current non-smokers, whereas 53 (5.9%) patients were current smokers. Additionally, 750 (83.3%) patients were lifetime abstainers. Of the participants, 96 (10.7%) were former drinkers, whereas 54 (6%) were current drinkers (Table 2).
Table 2.
Clinical characteristics
| Clinical characteristics | No (%) | p-value | ||
|---|---|---|---|---|
|
Overall (n = 900) |
Non–re-attendance (n = 756) |
Re-attendance (n = 144) |
||
| SIRS Criteriaª | < 0.001* | |||
| 2 Criteria | 688 (76.4) | 560 (74.1) | 128 (88.9) | |
| 3 Criteria | 193 (21.4) | 177 (23.4) | 16 (11.1) | |
| 4 Criteria | 19 (2.2) | 19 (2.5) | 0 (0.0) | |
| Respiratory rate | 766 (85.1) | 639 (84.5) | 127 (88.2) | |
| Heart rate | 671 (74.5) | 544 (80.0) | 127 (88.2) | |
| Body temperature | 482 (53.5) | 442 (58.5) | 40 (27.8) | |
|
White blood cell count > 12,000 or < 4,000 cells/mm3 |
256 (28.4) | 246 (32.5) | 10 (6.9) | |
| Chief Complaintª ** | ||||
| Fever | 517 (57.4) | 479 (63.4) | 38 (26.4) | < 0.001* |
| Dyspnea | 147 (16.3) | 90 (11.9) | 57 (39.6) | < 0.001* |
| Vomit | 88 (9.8) | 82 (10.8) | 6 (4.2) | 0.020* |
| Abdominal pain | 85 (9.4) | 61 (8.1) | 24 (16.7) | 0.003* |
| Diarrhea | 82 (9.1) | 77 (10.2) | 5 (3.5) | 0.016* |
| Sore throat | 57 (6.3) | 54 (7.1) | 3 (3.0) | 0.023* |
| Chest pain | 32 (3.6) | 19 (2.5) | 12 (8.3) | < 0.001* |
| Edema | 17 (1.9) | 10 (1.3) | 7 (4.9) | 0.012* |
| Headache | 99 (1.0) | 91 (12.0) | 8 (5.6) | 0.033* |
| Signs and Symptomsª ** | ||||
| Fever | 567 (63.0) | 523 (69.2) | 44 (30.6) | < 0.001* |
| Dyspnea | 168 (18.7) | 107 (14.1) | 61 (42.4) | < 0.001* |
| Vomit | 157 (17.4) | 151 (20.0) | 6 (4.2) | < 0.001* |
| Chill | 114 (12.7) | 107 (14.1) | 7 (4.9) | 0.002* |
| Diarrhea | 97 (10.8) | 89 (11.8) | 8 (5.6) | 0.027* |
| Chest pain | 34 (3.8) | 22 (2.9) | 12 (8.3) | 0.004* |
| Onsetb (min) | < 0.001* | |||
| Median | 1,440 | 2,840 | 1,440 | |
| Range (min–max) | 10,070 (10–10,080) | 8,630 (10–8,640) | 10,060 (20–10,080) | |
| < 1 day | 153 (17.0) | 116 (15.3) | 37 (25.7) | |
| 1–2 days | 628 (69.8) | 534 (70.7) | 94 (65.2) | |
| 2–3 days | 90 (10.0) | 81 (10.7) | 9 (6.3) | |
| 3–7 days | 29 (3.2) | 25 (3.3) | 4 (2.8) | |
| Top 5 Comorbidities** | ||||
| Hypertension | 289 (32.1) | 225 (29.8) | 64 (44.4) | |
| Diabetes without end-organ damage | 156 (17.3) | 126 (16.7) | 30 (20.8) | |
| COPD/Asthma | 102 (11.3) | 51 (6.7) | 51 (35.4) | |
| Cerebrovascular disease | 42 (4.7) | 34 (4.5) | 8 (5.6) | |
| Moderate or severe renal disease | 35 (3.9) | 26 (3.4) | 9 (6.3) | |
ªCategories: comparisons using Chi-square test
bMedian: comparisons using the Mann–Whitney U rank exact test
cMean: comparisons using independent t-test
dPercentage: comparisons using Fisher’s exact test
*p-value is significant
**One participant may have chief complaint, sign and symptom, comorbidities more than one sign/symptom
Amongst the 144 representations, patients whose initial SIRS visit involved chest pain, dyspnea, and/or abdominal pain were more likely to reattend compared to patients with other symptoms (Table 2). Patients with hypertension, COPD/asthma or renal disease were also more likely to reattend with SIRS.
Health service delivery system characteristics
Most of the patients arrived to the ED by walk-in (90.2%). At the first ED visit, 421 (46.8%) patients arrived in the evening (4 PM–midnight). Regarding triage level, most of the patients were categorized as ESI III and II. Most of the patients had waiting times of ≤ 10 and 11–30 min. Moreover, most of them had a treatment duration of ≤ 30 min. Most of the patients had ED LOS of ≤ 120 and 121–240 min.
Of the participants, 40.6% received antibiotics. Most of the patients received the following model of care: ED → Observed in the ED → Discharged (62.8%). Furthermore, patients received primary care and treatment and needed observation or waiting for investigation in some areas in the ED (no observation unit specifically), and when the clinical symptoms improved, they were discharged from the ED.
Of the participants, approximately 24.1% received the recommended discharge planning practice. Most of the patients were discharged from the ED by GP and EP doctors. Eighteen patients were discharged by ED nurses; EN and ENP nurses reported no unplanned re-attendances within 30 days. More than 50% of the patients were followed up with a healthcare service.
Upon representation, of the patients with two SIRS criteria, 57% were discharged, 41.4% were admitted to the general ward, and two (2) individuals died in the ED. Of those with three SIRS criteria, 56.3%, 31.3%, and 12.4% were admitted to the general ward, discharged, and admitted to critical care, respectively. No patients with four SIRS criteria had unplanned re-attendance (Table 3).
Table 3.
Health service delivery system characteristics
| Health service delivery system characteristics | n (%) or Median (min–max), n = 900 | p-value | ||
|---|---|---|---|---|
| Overall |
Non– re-attendance (n = 756) |
Re-attendance (n = 144) |
||
| Mode of arrival to the EDª | 0.148 | |||
| Walk-in | 812 (90.2) | 666 (90.7) | 126 (87.5) | |
| Emergency medical service | 88 (9.8) | 70 (9.3) | 18 (12.5) | |
| First-time shiftª | < 0.001* | |||
| Morning (8 AM–4 PM) | 304 (33.8) | 233 (30.8) | 71 (49.3) | |
| Evening (4 PM–midnight) | 421 (46.8) | 379 (50.1) | 42 (29.2) | |
| Night (midnight–8 AM) | 175 (19.4) | 144 (19.0) | 31 (21.5) | |
| Triage levelª | < 0.001* | |||
| ESI I | 13 (1.4) | 11 (1.5) | 2 (1.4) | |
| ESI II | 218 (24.2) | 168 (22.2) | 50 (34.7) | |
| ESI III | 544 (60.5) | 453 (59.9) | 91 (63.2) | |
| ESI IV | 110 (12.2) | 109 (14.4) | 1 (0.7) | |
| ESI V | 15 (1.7) | 15 (2.0) | 0 (0) | |
| Waiting timeb (min) | 9.0 (0, 180) | 9.5 (0, 180) | 6.0 (1, 156) | 0.001* |
| ≤ 10 | 356 (59.6) | 441 (58.3) | 85 (66) | |
| 11–30 | 271 (30.1) | 233 (30.8) | 38 (26.4) | |
| 31–60 | 74 (8.2) | 64 (8.4) | 10 (6.9) | |
| 61–180 | 19 (2.1) | 18 (2.4) | 1 (0.7) | |
| > 180 | 9 (0–180) | 9.5 (0–180) | 6 (1–156) | |
| Treatment durationb (min) | 17.0 (0, 555) | 19.0 (0, 555) | 15.0 (2, 156) | 0.022* |
| ≤ 30 | 706 (78.5) | 589 (77.9) | 117 (81.3) | |
| 31–60 | 144 (16.0) | 118 (15.6) | 26 (18.1) | |
| 61–120 | 37 (4.1) | 37 (4.9) | 0 (0.0) | |
| 121–180 | 11 (1.2) | 10 (1.3) | 1 (0.7) | |
| > 180 | 2 (0.2) | 2 (0.3) | 0 (0.0) | |
| ED length of stayb (min) | 79.5 (2, 630) | 68.5 (2, 630) | 106.0 (8, 297) | < 0.001* |
| ≤ 120 | 623 (69.3) | 545 (72.0) | 78 (54.2) | |
| 121–240 | 220 (24.4) | 164 (21.7) | 56 (38.9) | |
| 241–360 | 49 (5.4) | 39 (5.2) | 10 (6.9) | |
| > 360 | 8 (0.9) | 8 (1.1) | 0 (0.0) | |
| Treatmentª | < 0.001* | |||
| No antibiotics | 535 (59.4) | 425 (56.2) | 110 (76.4) | |
| Antibiotics | 365 (40.6) | 331(43.8) | 34 (23.6) | |
| Model of careª | < 0.001* | |||
| ED → Observed → Discharged | 565 (62.8) | 448 (59.3) | 117 (81.3) | |
| ED → Discharged | 239 (26.5) | 214 (28.3) | 25 (17.4) | |
| ED → Observed Unit → Discharged | 96 (10.7) | 94 (12.4) | 2 (1.4) | |
| Discharge planning qualityª | 0.133 | |||
| No quality discharge planning | 683 (75.9) | 588 (75.1) | 115 (80.0) | |
| Quality discharge planning | 217 (24.1) | 188 (24.9) | 29 (20.0) | |
| Disposition by healthcare providerª | 0.090 | |||
| GP doctors | 730 (81.1) | 607 (80.3) | 123 (85.4) | |
| EP doctors | 140 (15.6) | 123 (16.3) | 17 (11.8) | |
| Other specialist doctors | 12 (1.3) | 8 (1.1) | 4 (2.8) | |
| ED nurses | 12 (1.3) | 12 (1.6) | 0 (0.0) | |
| EN and ENP nurses | 6 (0.7) | 6 (0.8) | 0 (0.0) | |
| Follow-upª | 0.012* | |||
| No | 424 (47.1) | 369 (48.8) | 55 (38.2) | |
| Yes | 476 (59.2) | 387 (51.2) | 99 (61.8) | |
| Symptoms | 277 (58.2) | 210 (54.3) | 67 (75.3) | |
| Laboratory | 152 (32.0) | 135 (34.9) | 17 (19.1) | |
| Specific clinic | 47 (9.8) | 42 (10.8) | 5 (5.6) | |
ESI Emergency Severity Index
ªCategories: comparisons using Chi-square test
bMedian: comparisons using the Mann–Whitney U rank exact test
*p-value is significant
Compared to patients who did not reattend, patients who reattended were more likely to have presented in the morning on the first presentation, to have been triaged in category II or III, and to not have received antibiotics (Table 3).
Thirty-day incidence of unplanned re-attendance
Following ED discharge, 144 patients were unplanned re-attendances within 30 days. The 30-day incidence of unplanned re-attendance was measured in two ways as follows: CI and IR. The CI of 30-day unplanned re-attendance at the ED among patients with SIRS was 16%. Comparing the CI on the basis of levels of hospitals showed that middle-level hospitals had a higher CI (24%) than high- (14.3%) and first-level hospitals (12.7%). Comparing the CI on the basis of AHs, AH 3 had a higher CI (18.8%) than AHs 5 (15.4%) and 4 (14.4%).
The IR of 30-day unplanned re-attendance at the ED among patients with SIRS was 6 persons/1,000 persons/day. Majority of the re-attendances were one-time (90.3%), followed by two-time (8.3%) and three-time (1.4%) re-attendances. The re-attendance duration was approximately one-third within 3 and 4–7 days (18.8%) and 8–14 days (20.8%).
Identified pathogen characteristics
Reattendances were more common amongst patients who had not had a microbiology test on the first presentation, 55.9% underwent a microbiology test, with 64.4% of them had a blood specimen collected. Moreover, approximately 50% of the patients had positive microbiology reports, and 78.1% of the pathogens were identified as bacteria.
Of the participants, 44.6%, 25%, and 17% had urinary system infection, lower respiratory system infection other than pneumonia, and gastrointestinal system infection, respectively (Table 4).
Table 4.
Agent characteristics
| Agent characteristics | No. (%), n = 900 | p-value | ||
|---|---|---|---|---|
| Overall |
Non–re-attendance (n = 756) |
Re-attendance (n = 144) |
||
| Microbiology testª | < 0.001* | |||
| No | 397 (44.1) | 294 (38.9) | 103 (71.5) | |
| Yes | 503 (55.9) | 462 (61.1) | 41 (28.5) | |
| Blood | 324 (64.4) | 299 (64.7) | 25 (61) | |
| Urine | 70 (13.9) | 62 (13.4) | 8 (19.5) | |
| Stool | 1 (0.2) | 1 (0.2) | 0 (0) | |
| Sputum | 5 (1) | 5 (1.1) | 0 (0) | |
| Blood and urine | 100 (19.9) | 92 (19.9) | 8 (19.5) | |
| Blood and sputum | 3 (0.6) | 3 (0.6) | 0 (0) | |
| Microbiology reportª | 0.319 | |||
| Negative | 252 (50.1) | 229 (49.7) | 23 (54.8) | |
| Positive | 251 (49.9) | 232 (50.3) | 19 (45.2) | |
| Pathogensª | 0.067 | |||
| Bacterial | 196 (78.1) | 177 (76.3) | 16 (84.2) | |
|
Virus None identified |
55 (21.9) 0 (0.0) |
55 (23.7) 0 (0.0) |
3 (15.8) 0 (0.0) |
|
| Sources of infection categories by the CDCª | n = 220 | n = 207 | n = 22 | 0.380 |
| USI – urinary system infection | 102 (44.6) | 92 (44.4) | 10 (45.5) | |
| LRI – lower respiratory system infection other than pneumonia | 57 (25) | 53 (25.6) | 4 (18.2) | |
| GI – gastrointestinal system infection | 39 (17) | 37 (17.9) | 2 (9.1) | |
| SST – skin and soft tissue infection | 12 (5.2) | 9 (4.3) | 3 (13.6) | |
| CVS – cardiovascular system infection | 9 (3.9) | 8 (3.9) | 1 (4.5) | |
| EENT – eye, ear, nose, throat, or mouth infection | 9 (3.9) | 7 (3.4) | 2 (9.1) | |
| BJ – bone and joint infection | 1 (0.4) | 1 (0.5) | 0 (0) | |
ªCategories: comparisons using Chi-square test
*p-value is significant
Thirty-day unplanned re-attendance free survival time
Unplanned re-attendance free survival time curves were calculated using the Kaplan–Meier method, and the log-rank test was used for comparing the curves. No median unplanned re-attendance free survival time was noted as > 50% of the patients remained no unplanned re-attendance event. The numbers of patients who were under observation at 3, 7, 14, 21, and 30 days were 857 (95.3%), 832 (92.5%), 792 (88.0%), 772 (86.4%), and 407 (83.8%), respectively. We used censor at 30 days following discharge to deal with participants without the outcome of interest as shown in Fig. 2.
Fig. 2.
Kaplan–Meier estimate of overall 30-day unplanned re-attendance free survival rates of the study participants
Multivariable analyses of the association between 30-day unplanned re-attendance and sociodemographic and clinical characteristics, agent, health service delivery system using multiple Cox proportional hazard model
Univariable analysis revealed 23 factors that were statistically significantly associated with 30-day unplanned re-attendance free survival time (p-value < 0.05), including age, educational level, income, health insurances, chief complaints, signs and symptoms, illness severity, comorbidities, CCI, smoking habits, alcohol dirking, beds observed in the ED, room observed in the ED, EP doctor, EN/ENP nurse, nurses’ workload, model of care, type of guidelines, follow-up, first-time shift, waiting time, treatment duration, and ED LOS. All of these variables with p-value < 0.05 in the simple Cox proportional hazard model including educational level, comorbidity, vomit symptom, fever symptom, alcohol drinking, and model of care were included in the multiple Cox proportional hazard model using the forward stepwise (Wald) method.
Patients who had primary (6 years) or high (12 years) school educational attainment had a lower risk of 30-day unplanned re-attendance at 0.50-fold (95% CI = 0.340–0.758, p < 0.001) and 0.45-fold (95% CI = 0.261–0.775, p = 0.004), respectively, than those who were illiterate. Patients with comorbid conditions had a 5.0-fold (95% CI = 3.056–8.413, p < 0.001) higher risk of 30-day unplanned re-attendance than those with no comorbidities. Regarding the fever symptom, patients who had a fever on the first visit had a 0.308-fold (95% CI = 0.112–0.578, p < 0.001) lower risk of 30-day unplanned re-attendance than those with other symptoms. Regarding alcohol drinking, the hazard of 30-day unplanned re-attendance in patients who were ever alcohol drinkers was 1.54-fold (95% CI = 0.975–2.453, p = 0.064), but not significantly, whereas that of current alcohol drinkers was 6.2-fold (95% CI = 3.555–10.854, p < 0.001) than those with no alcohol drinking history. Regarding the model of care, patients who received the model of care of ED and discharged (ED → D/C) had an 11.1-fold (95% CI = 2.619–47.499, p < 0.001) higher risk of 30-day unplanned re-attendance than those who received the model of care of ED and observed in the observation unit and discharged (ED → OU → D/C); moreover, the model of care of ED, observed in the ED, and subsequently discharged (ED → OED → D/C) had the highest risk of 30-day unplanned re-attendance at 13.8-fold (95% CI = 3.401–56.167, p < 0.001). Details are shown in Table 5.
Table 5.
Multivariable Cox regression analyses of factors and unplanned re-attendance at the emergency department among patients with SIRS
| Factors | Adjusted HR (95% CI) |
p-value |
|---|---|---|
| Education level | ||
| No | Reference | |
| Primary to high school | 0.50 (0.340–0.758) | < 0.001* |
| Certificated to graduate | 0.45 (0.261–0.775) | 0 .004* |
| Comorbidity | ||
| No | 1.0 (reference) | |
| Yes | 5.0 (3.056–8.413) | < 0.001* |
| Vomit symptom | ||
| No | 1.0 (reference) | |
| Yes | 0.254 (0.112–0.578) | < 0.001* |
| Fever symptom | ||
| No | 1.0 (reference) | |
| Yes | 0.308 (0.213–0.446) | < 0.001* |
| Alcohol drinking | ||
| No | 1.0 (reference) | |
| Ever drunk | 1.54 (0.975–2.453) | 0.064 |
| Current drinker | 6.2 (3.555–10.854) | < 0.001* |
| Model of care | ||
|
ED → Observed- Unit → Discharged |
1.0 (reference) | |
| ED → Discharged | 11.1 (2.619–47.499) | < 0.001* |
|
ED → Observed → Discharged |
13.8 (3.401–56.167) | < 0.001* |
CI confidence interval, HR hazard ratio
*p-value is significant
Discussion
The results revealed that the CI of 30-day unplanned re-attendance was 16%, and the IR of 30-day unplanned re-attendance was 6 persons/1,000 persons/day. Factors affecting 30-day unplanned re-attendance to the ED were comorbidities, alcohol use, and the model of care in the ED and discharge; care in the ED, observed in the ED, and subsequently discharged was the highest risk factor for unplanned re-attendance. At 30 days following ED discharge, > 50% of the participants exhibited no unplanned re-attendance events (no median survival time) because of most of them were no unplanned re-attendance and some of patients re-attendance more than 30-day.
The incidence of re-attendance in the ED varies depending on demographic characteristics and health care and nursing system factors. The incidence rates of unplanned re-attendance at the ED of patients with all kinds of conditions range from 0.07—43.9% [34–36]. In this study, patients with SIRS who have symptoms similar to those of patients with fever or infection who come to the ED had a re-attendance rate of 16%, which may be higher than the study in some samples, including urinary tract infection, and lower than in some samples, including pneumonia and asthma. This finding is likely because this study included more patients with SIRS who had comorbidities, had a history of alcohol use and/or smoking, and were observed in the ED or observation unit than those who did not re-attend, which may be factors affecting the re-attendance.
The closest incidence of re-attendance to this study was in the group with asthma, with approximately 14% of re-attendance within 14 days [37]. Among patients with coronavirus disease 2019 and pneumonia, 18.3% had re-attendance within 30 days [38]. Approximately 17% of patients presenting with fever had re-attendance within 30 days [39]. However, other studies have reported a lower re-attendance incidence than that reported in the present study; for example, among patients with urinary tract infection, only 6% had re-attendance within 3 days and only 2% within 30 days [24].
Different levels of hospitals were associated with different re-attendance incidences. This study observed that middle-level hospitals had higher CI than high- and first-level hospitals, respectively. This finding is likely because annual ED visits can be a factor associated with re-attendance, which is consistent with the findings by Lu et al. [40], wherein hospital level and annual ED visits were associated with re-attendance incidence within 72 h. Middle-level hospitals with the highest number of annual ED visits had the highest re-attendance rate.
Regarding area health, the results showed that the CI of 30-day unplanned re-attendance at the ED of patients with SIRS in AH 3 had higher CI than those in AHs 5 and 4, respectively. This finding is likely because the number of total beds in the ED, beds observed in the ED, and rooms observed in the ED can correlate with re-attendance incidence. Beds and rooms observed in the ED are services provided to patients in the ED while they need to be observed, awaiting laboratory tests, or receiving ongoing medications or continual treatment. When high numbers of total beds in the ED, beds observed in the ED, and rooms observed in the ED are noted, a low re-attendance incidence may be observed. Furthermore, regarding the human resources factor, areas with fewer healthcare personnel have a higher re-attendance incidence than those with more human resources. This finding is likely because having an adequate number of healthcare personnel will affect the quality of care and patient safety in health and nursing service systems. AH 3 tended to have fewer human resources than AHs 4 and 5, that is, GPs, EPs, EMTs, paramedics, RNs, and EN/ENP physicians.
At 30 days following ED discharge, the unplanned re-attendance free survival time remained > 50%. This finding is likely because patients with SIRS did not have severe symptoms; however, the re-attendance rate gradually increased with increasing duration in the ER. The results of this study showed that most patients were triaged to ESI level III (urgent) (60.5%) and ESI level II (emergent) (24.2%) likely because they had mild symptoms. The results of this study are consistent with those by Gao et al. [41], who reported that the 30-day unplanned revisit from ED was approximately 22% and that demographic, diagnosis, hospital, and healthcare system were factors affecting 30-day re-attendance.
A higher education level was a factor associated with decreased re-attendance rates as evidenced by the sample group with education levels of primary to high school and certificated to graduate within 30 days. The non–re-attendance group had a higher education level than the re-attendance group. This finding is likely because according to the fundamental cause theory, an idea exists that education level is a social factor in the fundamentals of health and disease. According to the human capital theory, an individual’s education level will affect the development of knowledge, skills, reasoning, effectiveness, and a broad range of other abilities [42]. Moreover, education level is a health determinant, with low education levels being associated with poor health, stress, and a lack of confidence [43]. Several previous studies reported that individuals with lower education levels tend to be associated with more health problems and lower self-care behaviors than those with higher education levels [44, 45]. The results of this study are consistent with those by Naseer, Agerholm [46], who investigated 30-day re-attendance in Sweden and showed that a low education level (primary to lower secondary school) was a risk factor for re-attendance at 1.08-fold. Middle-level education attainment (upper secondary to post-secondary < 2 years) is a risk factor for re-attendance at 1.03-fold compared with individuals with a high education level (post-secondary to ≥ 2 years).
The HR of 30-day unplanned re-attendance of patients with SIRS who had comorbidities was significantly higher at approximately 5.0-fold than those without comorbidities. The re-attendance group had more comorbidities than those without re-attendance, and the top three most common comorbidities in the re-attendance group were hypertension (44%), diabetes without end-organ damage (20.8%), and COPD/asthma (35.4%). This finding is likely because patients with two or more comorbid conditions have negative health outcomes, complex caring, management, and have greater demands on health resources and costs [47]. The results of this study are consistent with those of other studies that reported that comorbidities were associated with unscheduled re-attendance in patients who come to EDs; for example, Jorgensen, Zurayk [24] reported that patients with three or more comorbidities have a 2.85-fold higher risk of re-attendance than those with less than three comorbidities. This finding is also consistent with that of Gabayan, Asch [48], who showed that patients with comorbidities had a risk of revisit within 7 days, and the top three most common diseases were chronic renal disease (not end-stage renal disease) at approximately 3.3-fold, followed by end-stage renal disease at approximately 2.9-fold, and congestive heart failure at approximately 2.5-fold higher risks than asthma [49]. Additionally, the CCI was a factor affecting 72-h re-attendance; particularly, a CCI score of ≥ 2 had a 1.31-fold higher risk of re-attendance than a CCI score of < 2. This study revealed that 80% of patients with unplanned re-attendance had comorbidities, and most CCI scores were in the range of 0–2.
This study showed that patients who had vomit symptoms on the first visit had a 0.254-fold lower risk of 30-day unplanned re-attendance than those with other symptoms. Furthermore, this study noted that the non–re-attendance group had more vomit symptoms than the re-attendance group, indicating that patients with vomiting are less likely to re-attend than those with other symptoms. This finding is likely because vomiting can be a sign or symptom of a disease or condition that is not severe and can be easily managed.
Patients who had fever symptoms on the first visit had a 0.308-fold lower risk of 30-day unplanned re-attendance than those with other symptoms. Additionally, this study noted that fever was more common in the non–re-attendance group than that in the re-attendance group, suggesting that individuals with fever are less likely to re-attend than those with other conditions. This finding is likely because patients with fever who receive a good service system and care in the ED can have a lower likelihood of re-attendance than those with other symptoms. This study also observed that the non–re-attendance group received more antibiotics than the re-attendance group (43.8% vs. 23.6%). This finding may be a factor that causes fewer recurrences in patients with fever. Moreover, Moura, Oliveira [50] investigated the signs and symptoms of patients with critical illness visiting the ED and reported that those with signs and symptoms of ventilation disturbances, neurological dysfunction, and pain have symptoms associated with critical illness that are not observed in the fever syndrome.
Patients with a history of drinking alcohol had a 6.2-fold higher risk of 30-day re-attendance than those in the group without an alcohol history. Of the participants, 12.5% who had re-attendance were identified as current drinkers, compared with non-alcohol drinkers at 4.8%. This finding is likely because alcohol consumption has short- and long-term effects on the body, depending on the amount and duration of drinking. In this study, the majority of alcohol drinkers drank higher concentration of alcoholic beverages (e.g., rice whiskey) up to 40% almost every day, consuming approximately half a glass to 1 bottle, and for a long period at approximately 1–10 years. Long-term alcohol use can have physical, psychological, and social effects on individuals, such as hypertension, learning and memory difficulties, and mental and social issues [51]. The impact of this health condition may be one of the factors associated with re-attendance at the ED. The results of this study are consistent with those by Pham, Kirsch [52], who investigated 72-h unplanned re-attendance rates and reported that patients with habitual alcohol use had a 1.39-fold higher risk than those without an alcohol history [53].
Patients who received care in the ED and subsequently discharged (ED → D/C) had a 11.1-fold higher risk of 30-day unplanned re-attendance; moreover, the model of care of ED, observed in the ED, and subsequently discharged (ED → OED → D/C) had the highest risk (13.8-fold) compared with the group receiving ED and observed in the observation unit and subsequently discharged (ED → OU → D/C). This finding is likely because patients who were observed in the ED more likely required observation. Awaiting laboratory results, administering intravenous fluids or medications, waiting to see a doctor or consult a specialist doctor, or waiting to evaluate the results of treatment causes patients to have a longer waiting time or LOS in the ED, which may be a factor correlated with re-attendance. SIRS is a syndrome, and doctors will not diagnose patients with SIRS immediately; however, doctors should diagnose the disease according to ICD-10 principles. Therefore, it may be necessary for patients to be observed in the ED. This scenario is likely because patients who were observed in specific observation units tend to receive more care and monitoring from health care providers than those in the ED. The results of this study are consistent with those by Pereira, Choquet [54], who demonstrated that patients who were observed in the ED were more likely to re-attend than those who were not observed in the ED. The care model of ED and subsequently discharged (ED → D/C) had a 2.32-fold re-attendance risk.
Therefore, observing in the ED or discharged when no good and appropriate service system is available in this part is organized in terms of arranging observation areas, critical care bed observation, arranging manpower, assigning work, and clearly defining roles and responsibilities. Furthermore, the organization of various care and support systems may be factors affecting patients’ re-attendance.
However, in this study, the incidence and factors associated with 30-day unplanned re-attendance of patients with SIRS may have been influenced by other factors affecting unplanned re-attendance, including patient-related factors (age, income, and health insurances), illness-related factors (illness severity), and organization-related factors (EP doctor, EN/ENP nurse, nurses’ workload, guideline, follow-up, first-time shift, waiting time, treatment duration, and ED LOS). All these variables were associated with patients with SIRS in the univariable analysis, and further observation and exploration in other situations is warranted.
From this study aimed to explore factors associated with 30-day unplanned re-attendance among patients with SIRS using survival analysis. Although randomization is the gold standard for causal inference, there are many situations in which randomization of patients with SIRS is unfeasible. However, survival analysis is a robust approach to infer causality or several factors from observational data and provide powerful tool to facilitate clinical studies.
Strengths and limitations
In this study, data collection was confined to 3 of 12 AHs in Thailand, revealing the highest mortality rates from community-acquired severe sepsis/septic shock in these areas compared with others. Consequently, although this study provides detailed insights into the burden in these specific areas, it acknowledges limitations in generalizability. Variations may exist in other regions, including differences in unplanned re-attendance incidence, patient demographics, clinical presentations, health service delivery, and factors influencing unplanned re-attendance. Additionally, 756 patients did not return for follow-up, which may be because of a lack of detailed understanding of the reasons behind this non–re-attendance. It is hypothesized that some patients did not return because they received satisfactory emergency care that negated the need for further follow-up. Alternatively, other patients may have sought care at different hospitals or healthcare facilities for their follow-up possibly because of convenience, preference, or perceived quality of care. Lastly, this study lacked control of patient mortality owing to some patients had more severe illness before ED arrival. This limitation underscores a potential gap in obtaining comprehensive patient outcomes and understanding the full spectrum of patient behavior post-discharge, which could inform improvements in patient care and follow-up strategies.
Despite these limitations, this study had significant areas of strengths. First, the combination of cross-sectional and prospective cohort study design was a strength of this study. Therefore, the results of this study were reliable and provided better data quality. A prospective cohort study reduces the possibility that the results will be biased by selecting participants for the comparison group. Second, no missing data for the analyzed variables were noted. The multivariable model included 900 participants, which is a comparably large sample size for this study. Lastly, the sample reflected data from 14 different hospitals, which increases statistical power, greater rigor, and external reliability, and increases the likelihood of impacting policy and clinical practice for these regions.
Conclusions
The results of this study show that the 30-day incidence of unplanned re-attendance of patients with SIRS is at a high level. Patients with more than three SIRS criteria should be admitted as they will receive primary care until improved and subsequently discharged from the ED. Therefore, they remain under observation and re-attendance. Individual and health service delivery systems are factors affecting unplanned re-attendance, which can be used for providing insights for developing policies and health service delivery and improving the quality of care in the ED and can inform future studies.
Supplementary Information
Additional file 1. Demographic and clinical characteristic questionnaire
Additional file 2. Health service delivery system interview
Additional file 3. Re-attendance record form
Acknowledgements
We would like to express our gratitude to all collaborating participants, including hospital directors, emergency department supervisors, major advisors, co-advisors, statisticians, and the Srimahasarakham Nursing College and the National Research Council of Thailand (NRCT). We forever grateful for your support in this project.
Abbreviations
- ED
Emergency department
- SIRS
Systemic inflammatory response syndrome
- HR
Hazard ratio
- CI
Cumulative incidence
- ESI
Emergency severity index
- MOPH
Ministry of Public Health
- PaCO2
Partial pressure of carbon dioxide
- REMS
Rapid Emergency Medicine Score
- CCI
Charlson Comorbidity Index
- CDC/NHSN
Centers for Disease Control and Prevention/National Healthcare and Safety NetworkRA, research assistant
- SD
Standard deviation
- IR
Incidence rate
- COPD
Chronic obstructive pulmonary disease
- ICD-10
International Classification of Diseases, Tenth Revision
- LOS
Length of stay
- Ahs
Area health regions
Authors’ contributions
S.S. Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing—original draft, Writing—review & editing. K.U. Conceptualization, Methodology, Validation, Formal analysis, Investigation, Resources, Data curation, Writing—review & editing, Supervision, Project administration. W.T. Conceptualization, Methodology, Writing—review & editing. S.S. Conceptualization, Methodology, Writing—review & editing. C.V. Methodology, Statistical analysis. T.T. Conceptualization, Methodology, Writing—review & editing.
Funding
The National Research Council of Thailand (NRCT) provided support for this study in 2021, No. N41D640036.
Data availability
Data cannot be shared openly but are available on request from corresponding author.
Declarations
Ethics approval and consent to participate
This study received certificate approval from the Multicenter Research of Mahidol University on January 15, 2021 (project number IRB-NS2020/48.0211). Moreover, this study received certificate approval and permissions from the director of hospitals for data collection. All methods were performed in accordance with relevant guidelines and regulations. Informed written consent was obtained from all participants before interviews, and anonymity, confidentiality, and privacy were ensured during and after data collection.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Comstedt P, Storgaard M, Lassen AT. The Systemic Inflammatory Response Syndrome (SIRS) in acutely hospitalised medical patients: a cohort study. Scand J Trauma Resusc Emerg Med. 2009;17(1):67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Horeczko T, Green JP, Panacek EA. Epidemiology of the systemic inflammatory response syndrome (SIRS) in the emergency department. West J Emerg Med. 2014;15(3):329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tuttle A, Bissonette A, Grzybowski M, M Nowak R, S Ander D, C. Morris D, et al. Systemic Inflammatory Response Syndrome (SIRS) in the Emergency Department: An Original Investigation – A Look Back in Time. Emerg Med Crit Care. 2018;1(2).
- 4.Sri-On J, Nithimathachoke A, Tirrell GP, Surawongwattana S, Liu SW. Revisits within 48 Hours to a Thai Emergency Department. Emerg Med Int. 2016;2016:8983573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Burdette SD, Parilo M, Kaplan L, Bailey H. Systemic inflammatory response syndrome [updated 2018 May 7; cited 2020 January 16]. Available from Medscape, eMedicine database.
- 6.Balk RA. Systemic inflammatory response syndrome (SIRS) Where did it come from and is it still relevant today? Virulence. 2014;5(1):20–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest. 1992;101(6):1644–55. [DOI] [PubMed] [Google Scholar]
- 8.Ng C, Chung C. An analysis of unscheduled return visits to the accident and emergency department of a general public hospital. Hong Kong Journal of Emergency Medicine. 2003;10(3):153–61. [Google Scholar]
- 9.Khan NU, Razzak JA, Saleem AF, Khan UR, Mir MU, Aashiq B. Unplanned return visit to emergency department: a descriptive study from a tertiary care hospital in a low-income country. Eur J Emerg Med. 2011;18(5):276–8. [DOI] [PubMed] [Google Scholar]
- 10.Hotchkiss RS, Moldawer LL, Opal SM, Reinhart K, Turnbull IR, Vincent JL. Sepsis and septic shock. Nat Rev Dis Primers. 2016;2:16045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Husak L, Marcuzzi A, Herring J, Wen E, Yin L, Capan DD, et al. National analysis of sepsis hospitalizations and factors contributing to sepsis in-hospital mortality in Canada. Age. 2010;37:39.7. [DOI] [PubMed] [Google Scholar]
- 12.Smyth MA, Brace-McDonnell SJ, Perkins GD. Impact of prehospital care on outcomes in sepsis: a systematic review. West J Emerg Med. 2016;17(4):427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Becker JU, Theodosis C, Jacob ST, Wira CR, Groce NE. Surviving sepsis in low-income and middle-income countries: new directions for care and research. Lancet Infect Dis. 2009;9(9):577–82. [DOI] [PubMed] [Google Scholar]
- 14.Ministry of Public Health, Thailand. Disease and Cause of Death [cited 2017 August 10]. Available from Health Data Center Web site: https://hdcservice.moph.go.th/hdc/main/index.php.
- 15.Yu C-W, Chang S-S, Lai C-C, Wu J-Y, Yen DW, Lee M-tG, et al. Epidemiology of Emergency Department Sepsis: A National Cohort Study Between 2001 and 2012. 2019;51(5):619-24. [DOI] [PubMed]
- 16.Page DB, Donnelly JP, Wang HE. Community-, Healthcare-and Hospital-Acquired Severe Sepsis Hospitalizations in the University HealthSystem Consortium. Crit Care Med. 2015;43(9):1945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Liu VX, Escobar GJ, Chaudhary R, Prescott HC. Healthcare Utilization and Infection in the Week Prior to Sepsis Hospitalization. Crit Care Med. 2018;46(4):513–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Duseja R, Bardach NS, Lin GA, Yazdany J, Dean ML, Clay TH, et al. Revisit rates and associated costs after an emergency department encounter: a multistate analysis. Ann Intern Med. 2015;162(11):750–6. [DOI] [PubMed] [Google Scholar]
- 19.The College of Emergency Medicine. Emergency Department Clinical Quality Indicators: A CEM guide to implementation; 2008 [cited 2020 Agust 25]. Available from Web site HRH The Princess Royal.
- 20.Calder L, Pozgay A, Riff S, Rothwell D, Youngson E, Mojaverian N, et al. Adverse events in patients with return emergency department visits. BMJ Qual Saf. 2015;24(2):142–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Graff L, Stevens C, Spaite D, Foody J. Measuring and improving quality in emergency medicine. Acad Emerg Med. 2002;9(11):1091–107. [DOI] [PubMed] [Google Scholar]
- 22.Kempker JA, Wang HE, Martin GS. Sepsis is a preventable public health problem. Crit Care. 2018;22(1):116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Trongsakul S, Lambert R, Clark A, Wongpakaran N, Cross J. Development of the T hai version of M ini-C og, a brief cognitive screening test. Geriatr Gerontol Int. 2015;15(5):594–600. [DOI] [PubMed] [Google Scholar]
- 24.Jorgensen S, Zurayk M, Yeung S, Terry J, Dunn M, Nieberg P, et al. Risk factors for early return visits to the emergency department in patients with urinary tract infection. Am J Emerg Med. 2018;36(1):12–7. [DOI] [PubMed] [Google Scholar]
- 25.Sabbatini AK, Kocher KE, Basu A, Hsia RY. In-Hospital Outcomes and Costs Among Patients Hospitalized During a Return Visit to the Emergency Department. JAMA. 2016;315(7):663–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.de Gelder J, Lucke JA, de Groot B, Fogteloo AJ, Anten S, Heringhaus C, et al. Predictors and Outcomes of Revisits in Older Adults Discharged from the Emergency Department. J Am Geriatr Soc. 2018;66(4):735–41. [DOI] [PubMed] [Google Scholar]
- 27.Gilboy N, Tanabe T, Travers D, Rosenau AM. Emergency Severity Index (ESI): A Triage Tool for Emergency Department Care, Version 4. Implementation Handbook 2012 Edition. AHRQ. 2011.
- 28.Urtritaprasit K. The relationship between recovery symptom and functional outcome in Thai CABG patients. [dissertation]. Ohaio: Case Western Reserve University; 2001. [Google Scholar]
- 29.Charlson ME, Charlson RE, Peterson JC, Marinopoulos SS, Briggs WM, Hollenberg JP. The Charlson comorbidity index is adapted to predict costs of chronic disease in primary care patients. J Clin Epidemiol. 2008;61(12):1234–40. [DOI] [PubMed] [Google Scholar]
- 30.The Centers for Disease Control Prevention. CDC/NHSN surveillance definitions for specific types of infections [updated 2014 January 10; cited 2020 January 16]. Available from: https://www.socinorte.com/wp-content/uploads/2014/06/17pscNosInfDef_current.pdf.
- 31.Spiegel DA, Abdullah F, Price RR, Gosselin RA, Bickler SW. World Health Organization global initiative for emergency and essential surgical care: 2011 and beyond. World J Surg. 2013;37(7):1462–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Borson S, Scanlan J, Brush M, Vitaliano P, Dokmak A. The mini-cog: a cognitive “vital signs” measure for dementia screening in multi-lingual elderly. Int J Geriatr Psychiatry. 2000;15(11):1021–7. [DOI] [PubMed] [Google Scholar]
- 33.Dicker R, Coronado F, Koo D, Parrish RG. Principles of Epidemiology in Public Health Practice; an Introduction to Applied Epidemiology and Biostatistics. 2006.
- 34.Hutchinson C, McCloughen A, Curtis K. Incidence, characteristics and outcomes of patients that return to Emergency Departments. An integrative review. Aust Emerg Care. 2019;22. [DOI] [PubMed]
- 35.Trivedy CR, Cooke MW. Unscheduled return visits (URV) in adults to the emergency department (ED): a rapid evidence assessment policy review. Emerg Med J. 2015;32(4):324–9. [DOI] [PubMed] [Google Scholar]
- 36.Han CY, Chen LC, Barnard A, Lin CC, Hsiao YC, Liu HE, et al. Early Revisit to the Emergency Department: An Integrative Review. J Emerg Nurs. 2015;41(4):285–95. [DOI] [PubMed] [Google Scholar]
- 37.Nik MN. Factors Associated with Emergency Department Revisits and Hospitalization Following Discharged Acute Asthma Exarcerbation. Med Health. 2016;11:22–8. [Google Scholar]
- 38.Albert A, Llorens P, Aguirre A, Martín-Sánchez FJ, Mínguez S, Moreno O, et al. Revisit after discharge from the emergency department in a cohort of patients with COVID-19 pneumonia and analysis based on the healthcare resource used for follow-up. RESALSEVID study. J Healthcare Qual Res. 2023;38(1):26–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ortega M, Miró O, Muñoz A, López B, Coll-Vinent B, Sánchez M. Factors associated with revisit and admission in patients complaining of fever that are discharged from the emergency department. Med Clin (Barc). 2007;128(7):251–3. [DOI] [PubMed] [Google Scholar]
- 40.Lu T-C, Ling D-A, Tsai C-L, Shih F-Y, Fang C-C. Emergency department revisits: a nation-wide database analysis on the same and different hospital revisits. 2020;27(2):114–20. [DOI] [PubMed] [Google Scholar]
- 41.Gao K, Pellerin G, Kaminsky L. Predicting 30-day emergency department revisits. Am J Manag Care. 2018;24(11):e358–64. [PubMed] [Google Scholar]
- 42.Zajacova A, Lawrence E. The relationship between education and health: reducing disparities through a contextual approach. 2018;39:273–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.World Health Organization. Regional Office for South-East Asia. Social determinants of health. WHO Regional Office for South-East Asia. 2008. https://iris.who.int/handle/10665/206363.
- 44.Cavicchioli S, MG,Ferraz de Campos TB, da Silva Rosa A, Lopes De Domenico EB, Andrade Frederico G, de Oliveira Monteiro O, et al. Educational program to promote the self-care of people with diabetes mellitus. 2019;37(2):169–79. [Google Scholar]
- 45.Robat Sarpooshi D, Taghipour A, Mahdizadeh M, Peyman N. Enablers of and Barriers to Effective Diabetes Self-Care in Iran: A Qualitative Study. Patient related outcome measures. 2020;11:109–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Naseer M, Agerholm J, Fastbom J, Schön P, Ehrenberg A, Dahlberg L. Factors associated with emergency department revisits among older adults in two Swedish regions: A prospective cohort study. Arch Gerontol Geriatr. 2020;86:103960. [DOI] [PubMed] [Google Scholar]
- 47.Valderas JM, Starfield B, Sibbald B, Salisbury C, Roland M. Defining comorbidity: implications for understanding health and health services. Ann Fam Med. 2009;7(4):357–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Gabayan GZ, Asch SM, Hsia RY, Zingmond D, Liang L-J, Han W, et al. Factors associated with short-term bounce-back admissions after emergency department discharge. Ann Emerg Med. 2013;62(2):136-44. e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Liu SW. Risk factors of admission in 72-h return visits to emergency department. Tzu chi medical journal. 2021;33(2):169–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Moura BRS, Oliveira GN, Medeiros G, Vieira AS, Nogueira LS. Rapid triage performed by nurses: Signs and symptoms associated with identifying critically ill patients in the emergency department. Int J Nurs Pract. 2022;28(1):e13001. [DOI] [PubMed] [Google Scholar]
- 51.Zhong L, Chen W, Wang T, Zeng Q, Lai L, Lai J, et al. Alcohol and Health Outcomes: An Umbrella Review of Meta-Analyses Base on Prospective Cohort Studies. Front Public Health. 2022;10:859947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Pham JC, Kirsch TD, Hill PM, DeRuggerio K, Hoffmann B. Seventy-two-hour returns may not be a good indicator of safety in the emergency department: a national study. Acad Emerg Med. 2011;18(4):390–7. [DOI] [PubMed] [Google Scholar]
- 53.Geniş B, Coşar B, Arikan Z. Readmission Rates and Causes Within The First Six Months After Discharge in Patients with Alcohol Addiction. Noro psikiyatri arsivi. 2021;58(1):57–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Pereira L, Choquet C, Perozziello A, Wargon M, Juillien G, Colosi L, et al. Unscheduled-return-visits after an emergency department (ED) attendance and clinical link between both visits in patients aged 75 years and over: a prospective observational study. PLoS ONE. 2015;10(4):e0123803. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Additional file 1. Demographic and clinical characteristic questionnaire
Additional file 2. Health service delivery system interview
Additional file 3. Re-attendance record form
Data Availability Statement
Data cannot be shared openly but are available on request from corresponding author.