INTRODUCTION

Ambulance diversion (AD) is defined as the redirection of ambulance to an emergency department (ED) different from the initially intended ED¹. Lagoe and Jastremski² initially used AD in patients with minor injuries with the purpose of both appropriate use of hospital sources for selected patients and relief of overcrowding in EDs. Actually, AD is supposed to be used in cases of disasters or overcrowding in EDs for ED surge capacity and redistribution of patients among regional EDs in the health service network^2,3. Although initial utilization of ADs targeted improved health care to overcome conditions such as insufficient hospital resources [computed tomography scanner, blood bank, operating room, intensive care unit (ICU)], staff shortage (trauma team, on-call specialists etc.) and ED overcrowding, inappropriate overuse of AD turned out to be a daily practice and led to increased patient transport time, delays in emergency medical care and negative outcomes in patient health care^1,3-6.

Determination of frequency, causes and consequences of AD might provide useful information for relevant arrangement and regulations. AD has not been subject to investigation in Turkey so far. The present study aimed to perform a comprehensive and comparative analysis on causes, prevalence, disposition and outcomes of ADs in Ankara, the capital city of Turkey.

MATERIALS AND METHODS

Management of ED and Emergency Medical Service (EMS) in Turkey is executed according to Emergency Medical Services Regulations (EMSR)⁷. EMSR dictates that all hospitals with emergency healthcare services have to admit and perform necessary interventions to patients transferred by ambulance regardless of the ability to pay or presence of insurance. Temporary interruption of emergency healthcare or AD for patients transferred by ambulance are prohibited. However, if emergency treatment of the patient is not possible in the first admitting hospital and requires transfer to another hospital, transfer to another healthcare facility is permitted after the establishment of required transferred conditions via 112 Ambulance Command Control Center (ACCC) coordination⁷. In practical aspect, EMSR was not largely followed in the management of patients transferred by ambulance to ED and their transfer to another hospital after AD.

The study was conducted in the assigned work area of Ankara Provincial Directorate of Health (PDH). EMS in Ankara is provided by the Directorate of Emergency Medical Services Department located in the structure of PDH. Stations of 112 Ambulance Services were settled in previously assigned critical points. A crew consisting of paramedic, emergency medicine technician and ambulance driver are on duty basis of 24/7. A few numbers of ambulances include doctor employment and are reserved for complicated cases. Ambulance crew have been trained for Basic Life Support and Advanced Life Support for pediatric and adult population. When 112 emergency line is dialed for emergency medical help, 112 ACCC assigns the nearest ambulance crew to the scene. Patients collected from the scene are transferred to the nearest hospital ED by the discretion of the ambulance crew. AD practice of EDs in Ankara is quite different from the one defined in the literature. After the transfer of the patient collected from the scene to the nearest hospital ED by the discretion of the ambulance crew, patient is evaluated by ED physician without being unloaded from ambulance stretcher and diverted to another hospital after the declaration of reason for not admitting the patient. Following decision for AD, the selection of the next hospital is decided by ambulance crew, in some occasions, two or more hospitals are visited for admission to hospital. The most convenient destination is generally found after the ambulation of EDs with the patient.

Ambulance run-sheets include patient name, surname, gender, initial diagnosis, treatment content during transport, the name of the receiving hospital and staff. If AD is decided for the case from the first ED, the name of the physician deciding AD and the reason for AD are recorded.

Ambulance run-sheets are collected towards the end of each month in the Committee of Emergency Health Services Coordination (CEHSC) office embodied in Ankara PDH. The data are extracted from ambulance run-sheets as a summary including demographics of the patient, date, the name of the hospital and the physician deciding AD, the reason for AD and the name of the receiving hospital after AD. A written form with the information of extracted data is sent to hospitals involved in ADs of the last month by CEHSC. Evaluation Committee of Ambulance Diversion (ECAD) consists of three physicians one of whom is ED staff and evaluates AD by collecting information from the hospital deciding AD. ECAD of the receiving hospital evaluates the characteristics, final diagnosis and the disposition of the admitted AD patients. Reports from both hospitals involved in the AD process are sent to CEHSC in hardcopy and digital format. A commission in PDH is gathered each month to interprete ADs performed in the past month.

The present study comprised patients loaded to EMS ground ambulances from scene and diverted by EDs of the first hospital they were transported to another hospital in Ankara in the year 2014 (01 January-31 December 2014). Required data were obtained from CEHSC embodied in Ankara PDH and analyzed in a retrospective fashion. The prevalence and causes of AD, patient demographics, final diagnoses, characteristics of the diverting and receiving hospitals, disposition features in the receiving hospital and deaths were investigated.

This study was approved by the Ethical Board of University of Health Sciences Turkey, Ankara Numune Health Training and Research Hospital (Ankara, Turkey; 18.4.2019, decision number: Ethics Committee-19-2669).

Statistical Analysis

Descriptive statistics were used to summarize clinical characteristics. Median and range were used for expressing continuous variables which had abnormal distribution. Normal distribution was tested by the Kolmogorov-Smirnov test. Statistical Package for the Social Sciences (SPSS version 11.0, SPSS Inc., Chicago, IL, USA) was used for analysis.

RESULTS

Ankara had a population of 5.150.072 in year 2014⁸. A total of 174,699 patients were transferred to ED of hospitals in Ankara. A total of 1300 (0.74%) ADs took place. AD was most frequently used in weekends and out of working hours (974, 74.9%).

Six-hundred ninety-two patients (53.2%) were female and the number of patients younger than 18 years was 178 (13.6%). The number of patients between the ages of 18 and 65 years was 662 (50.9%). The median age of the study population was 49 years (100-1).

ADs due to internal medical conditions (974 ADs, 74.9%) were greater than ADs due to surgical conditions (326 ADs, 25%). The most common causes of medical conditions included shortness of breath (95 ADs, 7.3%), chest pain (61 ADs, 4.6%) and cerebrovascular diseases (57 ADs, 4.3%) (Table 1). Furthermore, the most common surgical conditions included road accidents (151 ADs, 11.6%), fall-related injury (95 ADs, 7.3%) and fight injury (22 ADs, 1.6%) (Table 2).

The most common cause of AD was unavailability of beds in EDs, ICUs, and inpatient services (622 ADs, 47.8%) (Table 3).

Unavailability of on-call specialist or a certain clinic with relevant specialization led to 290 (22.3%) ADs. Neurology (30 ADs 2.3%), neurosurgery (28 ADs, 2.1%) and physchiatry (16 ADs, 1.2%) were the main branches causing AD, respectively (Table 3).

Unavailability of imaging devices, laboratory tests or other hospital resources led to 176 (13.5%) ADs. The most common reason among those was broken status or absence of computed tomography (69 ADs, 5.3%). The number of ADs with undetermined cause or missing data was 114 (8.7%) (Table 3).

Sociological factors led to 62 ADs (4.7%). The most common sociological factor leading to AD was patient’s will to be treated in the ED of hospital in which his/her previous treatment was followed (25 ADs, 1.9%).

The presence of multiple factors such as unavailability of ICU bed, relevant specialist, laboratory tests or other hospital resources was detected in 26 ADs (2%). Employment of the required surgeon in another surgical operation caused 10 ADs (0.7%).

A total of 15 (1.1%) AD patients died. Three of them died during ambulance transport, and four died right after arrival to the receiving hospital ED (Table 4). One of the patients died after road accident while the others died from medical conditions.

Five hundred three patients (38.6%) were discharged from the receiving hospital ED and 187 patients (14.3%) were admitted to inpatient clinics of the receiving hospital. Admissions to ICUs were made in 76 patients (5.8%); 54 of those (4.1%) were admitted to ICUs of the receiving hospital while 22 of those (1.6%) were admitted to ICUs of another hospital due to unavailability of ICU bed in the receiving hospital. The disposition of 496 patients (38.1%) in the receiving hospital could not be determined (Table 4). A small portion of ambulances (52 ADs, 4%) were diverted for more than once by hospital EDs.

The most common institutions using AD were Community-General Hospitals (475 ADs, 36.5%) and University Hospitals (371 ADs, 28.5%), respectively. The most common institutions receiving ADs were education and research hospitals (858 ADs, 66%) and university hospitals (207 ADs, 15.9%), respectively (Table 5).

When reports of CEHSC committee assessing appropriate use of ADs were taken into account, 507 ADs (39%) were deemed appropriate while 365 ADs (28%) were deemed inappropriate. Appropriate use could not be determined in 428 ADs (32.9%) due to missing response from hospitals.

DISCUSSION

The number of patients transferred by ambulance in the study period was greater than the number reported in previous studies; however, AD rates were less than those in the literature^1,3,4,9. Prohibition of AD in Turkey is the main reason for this finding. Although AD is forbidden in Turkey, inadequate hospital resources mandated the use of AD. Furthermore, this appears to reflect serious problems within the hospital system and a failure on the part of local health regulatory agencies to assure access to care as described by law or regulation.

The present study showed that AD was most commonly used in weekends and out of working hours. Similarly, previous studies reported increased use of AD in the evening hours^1,3,4. The leading cause might be unavailability of many routine health services and on-call specialists in the off-hour time periods.

In consistence with previous reports, the number of patients older than 18 years involved in ADs were greater than the number of pediatric patients^1,4. Thus, reorganization of EDs providing healthcare service to adult patients might be useful to prevent future ADs.

In countries where AD is legal, eligible conditions and patients for AD are prespecified by AD policies^4,7,10,11. Since AD is prohibited in Turkey and there is no health policy for AD, patients with conditions such as road accident or pregnancy which should not be diverted (even in countries where AD is legal) underwent AD. It was also noticed that this situation turned out to be a variable practice of AD between hospitals regardless of the magnitude or severity of the injury. An AD policy with a better and clear definition of cases which should not be diverted might be developed.

In consistence with previous AD reports, unavailability of beds in EDs, ICUs and inpatient clinics were detected as the most common cause for AD^3-5,9,12. Patients have no restrictions for ED admissions in Turkey and each patient in the ED can seek medical care regardless of the ability to pay. Therefore, there is a sheer amount of admissions to EDs resulting in overcrowding. So, adjustments to increase the number of beds in ICUs and inpatient services are needed to prevent overcrowding of ED and also AD.

Death rates were quite lower than the rates reported in previous studies^1,3. The main underlying reason of this result might be less diversion of unstable patients and lower AD rates when compared to those in the literature.

In this study, it was noticed that a portion of ambulances encountered multiple diversions in hospital EDs. Previous studies also reported the involvement of more than one hospital in AD period in the same time frame⁴. Selection of the most suitable destination hospital by ambulance crew is a result of lack of a healthy communication among 112 ACCC, ambulance crew and hospital ED directors. Previous studies in the literature pointed out limited coordination among hospital and EMS providers as a contributing factor to diversion problem, thus reconstitution of an effective communication and monitoring infrastructure would establish an effective transport and accurate ED destination of the patient^3,13.

The majority of ADs were made by community-general hospitals in consistence with previous reports⁶. Unavailability of on-call specialists for 24 hours and limited hospital resources in those hospitals might have played a key role in this setting.

When reports of CEHSC committee assessing appropriate use of ADs were taken into account, only 39% of ADs were deemed appropriate. Meanwhile, previous reports defined AD as being supposed to be used in cases of disasters or overcrowding in EDs for ED surge capacity and redistribution of patients among regional EDs in the health service network^2,3. Besides, it is clear that hospitals receiving AD have the capacity to overcome this extra patient burden. However, the main problem might be related to lack of contact (radio or telephone) between the prehospital personnel and hospital EDs. This is not within the standard for almost all EMS systems. Therefore, we believe that constitution of an AD policy with an effective and controllable coordination among 112 ACCC, EMS crew and hospitals is crucial in patient transportation from scene to the most appropriate destination.

Another specific finding reported in our study is the discrepancy of patients supposed to be admitted in ICUs. A total of 205 ADs were done due to unavailability of ICU beds; however, only 76 patients of those were admitted to ICUs in the receiving hospitals. This finding suggests decreased accuracy rate of triage made by the ED physician who diverted the patient or lower decision threshold for ICU admission ending with AD.

Study Limitations

Although sample size is one of the largest data in this particular topic, retrospective analysis is a limitation. Some of the data including causes of AD, patient outcomes and disposition were missing. Similarly, patient demographics of patients transported by private ambulance services were missing. Besides, limitation of AD interpretation with 1 particular year might have diminished the yielding effect of data.

CONCLUSION

Unavailability of patient beds in EDs, ICUs or inpatient clinics are the most common causes of AD. Therefore, there is an obvious need for more hospital beds. Furthermore, improvement of communication among ACCC, EMS crew and EDs should be targeted for better patient outcomes and optimal management of resources.

Acknowledgements

We thank Dr. Gamze Babur Güler for her valuable contributions in statistical section.

Ethics

Ethics Committee Approval: This study was approved by the Ethical Board of University of Health Sciences Turkey, Ankara Numune Health Training and Research Hospital (Ankara, Turkey; 18.4.2019, decision number: Ethics Committee-19-2669).

Informed Consent: Retrospective study.

Peer-review: Externally peer-reviewed.

Authorship Contributions

Surgical and Medical Practices: M.A.C., G.G.D., Concept: M.A.C., G.G.D., Design: M.A.C., G.G.D., Data Collection or Processing: M.A.C., G.G.D., Analysis or Interpretation: M.A.C., G.G.D., Literature Search: M.A.C., G.G.D., Writing: M.A.C., G.G.D.

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.