Lessons learned from an evaluation of referrals to the emergency department

Input factors reflect patient flow into the ED [5‐7]. Factors that have been studied include nonurgent ED visits. Reasons for nonurgent ED referrals include a lack of community medical services beyond regular working hours (evenings, nights, and weekends), physician referrals, pain complaints, minor trauma, age, convenience, seasonality, and belief in the superior treatment capabilities of ED physicians. Certain characteristics of primary care are also associated with an increased likelihood of visiting the ED, including lack of a regular physician, unmet healthcare needs, poor continuity of care, and perceived lack of rapid access to care [8].

Throughput factors reflect workflow and bottlenecks within the ED [6, 7]. ED resources include staff, beds, registration, documentation, and laboratory testing. Additionally, inefficient work processes can lead to delays in treatment. These include a lack of continuity of patient care due to shift changes, communication problems between the treating teams, and lack of access to important medical information. Extra-ED factors include delays caused by waiting for laboratory tests, imaging studies, and consultants.

Output factors reflect bottlenecks in the discharge process or hospitalization stage [6, 7], and can be measured based on length of stay (LOS) [6]. LOS can be prolonged due to “access block,” when a patient is forced to wait in the ED for over 8 h due to lack of proper access to an inpatient hospital bed [2].

In Israel, health insurance is mandated by the National Health Insurance Law (1994), and coverage for every citizen is ensured through four Health Maintenance Organizations (HMOs) [9]. Services include hospitalizations, primary care, specialty consultations, prescription drugs, certain preventive services, mental health care, and dental care for children [10]. Only soldiers, who receive health care directly from the military, are exempt from the system.

An integral part of Israel’s health care system is urgent care. Urgent care centers (UCCs) may be private or operated through an HMO. Each of the four HMOs maintains UCCs throughout Israel, with the Meuhedet HMO listing the most, at 100 sites (https://​www.​meuhedet.​co.​il). However, many of these centers are open only during certain hours of the day or evening. In the United States, there are over 9000 UCCs [11]. Israel’s major private UCC operator is known as TEREM (acronym in Hebrew for Immediate Medical Treatment), which currently operates 24 clinics throughout the country, many for 24 h a day (https://​terem.​com/​Clinics.​aspx?​lang=​1). The centers are staffed by physicians and paramedical staff, including nurses, emergency medical technicians, or paramedics. Although not all physicians at TEREM UCCs are board-certified in a specialty, the organization does arrange numerous training sessions and asynchronous learning programs to teach advanced urgent care skills and concepts. They also have an on-call specialist for phone consultations. The centers can perform advanced blood tests, including troponin and D-dimer; this means that in a subset of low-risk patients, they have the capability to rule out both a myocardial infarction and a pulmonary embolus or deep venous thrombosis. During 2012, a total of 200,000 patients visited TEREM UCCs in Jerusalem, similar to the number of visits to hospitals in Jerusalem in the same year (D. Zimmerman, personal communication, April 23, 2017). The Ministry of Health reports that cities with a TEREM UCC have 35 to 45% fewer ED visits [12].

Unnecessary ED referrals made by medical personnel may contribute to the input factors that cause crowding [6]. In general, patients who use a usual source of primary care receive more appropriate preventive care and experience fewer ED visits and hospitalizations [8]. In Israel, the percentage of ED visits with a medical referral is increasing, with 69% in 2015 compared to 66% in 2010 and 63% in 2006 (excluding maternity visits) [13].

Referrals to the ED can be made by the patient’s primary care physician (PCP) or any number of physicians affiliated with one of Israel’s many UCCs. Obtaining a referral is important both to reduce ED overcrowding and for financial reasons, as patients must pay 800 New Israel Shekels (approximately 218 US dollars) to receive care in an ED without a referral. It should be noted that there are several exemption cases based on the Israel National Insurance Law for which a referral is not needed. Examples include a visit that results in hospitalization, involvement in a traffic accident with a letter from the police, confirmed school injury, new fracture, shoulder dislocation, foreign body to the eye, complication of cancer, and active labor (https://​www.​health.​gov.​il/​English/​Services/​Citizen_​Services/​Pages/​ER.​aspx).

There is a paucity of literature that evaluates why physicians decide to refer patients to the ED and whether these referrals are necessary. A recent study of ED referrals from a number of UCCs local to Las Vegas, Nevada found that 35.9% (28.8% in adults) were not justified [14].

Interestingly, there is no consensus on the definition of “unnecessary” or “unjustified” referrals. Zitek and colleagues assume they refer to visits by patients who have no advanced imaging studies, procedures, or specialty consultations done in the ED and are not admitted to the hospital [14]. In the pediatric literature, “essential ED interventions” include simple procedures such as incision and drainage, catheterization, and laceration repair [15]. Other studies imply that “unnecessary” visits are those not requiring urgent treatment or specialist input [16]. This is problematic because concerns or health issues that initially seem urgent or emergent may, in the end, be benign, but evaluation in the emergency department may still be justified. A classic example is epigastric pain in a patient over 50, which may include coronary disease in the differential diagnosis, even if the patient is ultimately discharged with reflux or gastritis. This study objectively defines an “unnecessary” ED visit as one where the referring center has the same resources as the ED, and no further studies are carried out prior to the patient’s discharge.

Finally, being referred to the hospital is not always better. Interestingly, patients diagnosed with hypertensive urgency who were treated outside of a hospital had better outcomes than those sent to the ED for the same diagnosis [17].

This study aims to determine whether patterns of potentially unnecessary referrals from either PCPs or UCCs can be identified, which could be used to reduce unnecessary inflow to the ED. The results may lead to process improvement in the healthcare system.

Subjects’ ages ranged from 2 to 99 years, with the most visits made by those between 20 and 39 (37.1%). In this group, 46.6% of visits were for obstetric/gynecological (OB/GYN) issues. Although most patients below the age of 18 were sent directly to the SZMC pediatric ED, a small number of children were seen in the general ED for orthopedic problems. This was due to convenience because at that time the general ED had a room equipped with orthopedic supplies and staffed by an orthopedic resident at all times. The majority of visitors (55.8%) were women (Table 1).

Emergency medicine is a relatively new specialty in Israel. During the study period, triage was based on both the Canadian Triage and Acuity Scale and various specialty services. Most patients were seen by either a resident or an attending physician in emergency or internal medicine (45.1%), but many were seen directly by a specialist. In this ED, discharge privileges are given only to residents who have finished a one-year rotating internship, have a minimum of approximately 2 years of experience in their specialty, and have passed their written boards. Therefore, even the most junior physicians entitled to discharge patients have at least 3 years of experience, the same as board-certified internists or emergency physicians in the United States. The most common specialty referrals were for general surgeons, obstetricians/gynecologists, or orthopedists. Other referrals were seen by ophthalmologists, otolaryngologists (ENTs), and urologists (Table 2).

Significantly more referrals were made across the board by PCPs compared to UCCs (1712 vs. 280, p < 0.001). This was consistent for all specialties, with statistical significance for orthopedics, general surgery, and OB/GYN cases (p = 0.039, p < 0.001, and p = 0.003, respectively) (Table 2).

A higher percentage of patients referred by PCPs had potentially unnecessary visits than patients referred by UCCs (13.9% vs. 7.9%, p = 0.005) (Table 3).

The city of Jerusalem officially includes 58 neighborhoods, making it difficult to calculate referrals by neighborhood. For statistical calculation and clear data segmentation, neighborhoods were grouped into districts based on geographical proximity according to a map of the Jerusalem municipality used by the Central Bureau of Statistics (https://​jerusaleminstitu​te.​org.​il/​wp-content/​uploads/​2019/​05/​shnaton_​SA_​Area_​Map.​png). Twenty districts were included in the analysis: 18 in the Jerusalem metropolitan area, 14 in Jerusalem itself (A-M), and 4 outside of Jerusalem. Each included several adjacent neighborhoods. One group (“Other”) was located outside of the Jerusalem metropolitan area and one (“Unknown”) represented referrals from an undetected address. District E had the highest percentage of potentially unnecessary referrals (19.70%) compared to the rest of the districts (p = 0.029), and a total of five districts had 15% or more potentially unnecessary referrals)Table 4).

Because this study involved a single center in an urban center of Israel, the results may not be generalizable to other hospitals in Israel or around the world. The UCCs surrounding the target hospital in this study may have both more resources and higher quality staff than those in other cities or countries. Additionally, although more unnecessary visits were referred by PCPs than UCCs, the study did not examine how many of the patients who were referred to the ED from the UCC were previously sent to the UCC by their PCP.

The definition of a potentially unnecessary visit is not standardized in the medical literature. Although this study assumes that the emergency department is the “gold standard” of care, there was no outcome data for any of the patients who were discharged. However, the SZMC does have a rigorous auditing process, and it is expected that there would have been feedback in the case of a negative outcome. Further analysis of the breakdown of cases by surgical subspecialty is also necessary to determine whether further training of UCC physicians could reduce unnecessary referrals.