The city of Yokohama in Japan recently introduced a new emergency medical service system designed to dispatch appropriate and sufficient emergency medical service staff to patients in a critical condition. Utilizing a computer algorithm, the risk to life of a patient requiring ambulance transport is assessed using information obtained from the emergency call.
by Dr Kenji Ohshige
Emergency medical services in Japan
In Japan, local governments provide prehospital emergency medical services as a public service. Anyone can request the use of an ambulance free of charge by phoning 119. The number of ambulances dispatched in Japan has risen rapidly over the last decade. The increased demand for ambulance services has resulted in a gradual lengthening of the time it takes for an ambulance crew to respond and arrive at the scene. As delayed response time reduces the number of patients who survive, priority dispatch of ambulances to patients in a critical condition has become a matter of importance for the Japanese prehospital emergency
medical services system.
Yokohama
Demand for the ambulance service has also increased in Yokohama, Japan. The number of ambulances dispatched is forecast to increase dramatically since the population of elderly people, who need emergency services more often than younger people, is itself increasing rapidly [1].
Yokohama’s prehospital emergency medical service is unified and managed by the Yokohama Safety Management Bureau (formerly the Fire Bureau). There are two advisory committees on the city’s pre-hospital emergency medical service; the Regional Emergency Medical Service System (REMS) committee, and the Medical Control System (MCS) committee.
At one stage, the Yokohama REMS committee had considered introducing a program to reduce ambulance overuse, in which ambulance users share the cost of the ambulance service. Demand-side cost sharing has already been used as a means of decreasing the extra demand for medical services [2,3]. However, the committee abandoned this approach because it was considered that a cost-sharing or cost-containment policy for ambulance services was likely to not only decrease medically un-necessary ambulance use, but also medically necessary use [4].
Yokohama city implemented a city-wide public awareness campaign on the subject of appropriate ambulance use in 2005, with the aim of reducing medically unnecessary ambulance usage. During the actual campaign period the number of ambulance call outs declined, which seemed to indicate campaign success. However, deeper study revealed a phenomenon whereby it was shown that such campaigns were actually likely to reduce ambulance calls for both non-serious and serious conditions [5]. As several other reports have shown [6,7], the Yokohama experience indicated that it is not easy for the average person or member of the public to assess risk correctly.
On October 1st, 2008, the city of Yokohama started a new emergency medical service system in which a computer system is used to assess patient risk. Utilizing a computer algorithm, the risk to a patient is evaluated from information obtained from the ambulance call at the time of the call. The system is designed to dispatch ample and appropriate emergency service medical staff to patients facing a critical situation.
Yokohama computer triage system
In Yokohama’s new system, emergency call workers systematically ask ambulance callers for information using an interview protocol; they ask for the type of caller, i.e., the patient him/herself, a family member, nursing home staff, or other (not the patient himself/herself, nor a family member, nor a nursing home staff member), and patient’s characteristics such as age and sex. Call workers also interview callers to obtain essential information that is used to assess the patient’s risk such as consciousness level, breathing status, walking ability, position, complexion and symptoms. This information is entered into a computer-based triage form during the phone call [Figure 1]. Based on previous data (see below), the triage form classifies patients into three categories: category A (there is a probability that the patient faces an imminent risk of dying); category B (there is a possibility that the patient faces a risk of dying); or category C (the probability that the patient faces a risk of dying is very low).
The triage form also quantitatively estimates the patient’s life threat risk. A logistic model was used in the program. If the estimated life threat risk is higher than 10%, the triage form categorizes patients into A+ (the probability is very high that the patient faces a risk of dying). For example, when a call was made by a family member who was in panic, if the patient’s age was 70 years, consciousness not clear and breathing status abnormal, if the patient was lying down and unable to walk, the patient’s face cyanotic, and sweating was unable to be confirmed, then the life threat risk was estimated to be 19.2% by the model. Such a patient is categorized as A+. The algorithm had originally been constructed based on data collected previously from 4,301 cases, prior to the start of the new system. The accuracy of attributing category A+ has been evaluated and published elsewhere [8].
Evaluation of the algorithm for estimating a patient’s life threat risk
The data used in the evaluation study [8] were collected during operation of the Yokohama New Emergency System from October 1st, 2008 to March 31st, 2009. The study compared the estimated life threat risk occurring at the time of the emergency call to the patient’s state or severity of condition: i.e. death confirmed at the scene by ambulance crews; resulted in death at emergency departments (EDs); life-threatening condition with occurrence of cardiac and/or pulmonary arrest (CPA); life-threatening condition without CPA; serious but not life-threatening condition; moderate condition; and mild condition.
The life threat risk as estimated from the algorithm is shown in Figure 2 by the status of the patient. Out of the 61,027 cases during October, 2008 to March, 2009, there were 4,423 cases that were categorized as A+. The status of these 4423 patients classified as A+ is shown in Table 1. he sensitivity, specificity, positive predictive value and negative predictive value of categorizing patients as A+ whose condition resulted in death or CPA were 80.2% (95% CI: 78.6% - 81.8%), 96.0% (95% CI: 95.8% - 96.1%), 42.6% (95% CI: 41.1% - 44.0%), and 99.2% (95% CI: 99.2% - 99.3%), respectively.
Future work
In the Yokohama system, if patients are categorized as A+, an ambulance, a fast response car, and a fire engine are dispatched. The Emergency Medical Division of the Yokohama Safety Management Bureau have reported that for patients categorized as A+ using the new algorithm, the mean arrival time of the first responder to the scene was approximately one minute shorter than for other patients. Whether the new system improved survival rate from CPA is being evaluated in further studies.
The computer triage form also assesses patient’s risk as A, B or C. Category A targets patients who are in a life-threatening condition at EDs without CPA, category B targets patients who are classed as having a serious but not life-threatening condition at EDs; category C targets patients who are classed as having a moderate or mild condition at EDs. The methods for categorizing A, B and C are different from the algorithm for categorizing A+. The accuracy of attributing categories A, B and C will be analysed and reported in the near future. The algorithm to assess a patient’s risk can be improved with the data obtained under the new emergency medical services system, in which information obtained during emergency calls is recorded as digital data. Data of more than 120,000 triaged cases per year should contribute to the development of an accurate triage algorithm.
Conclusion
A triage algorithm has been constructed based on real data in Yokohama; the accuracy of the algorithm will be improved as additional data are collected. The Emergency Medical Division of Yokohama periodically reports the results of call-triage [9] and the Medical Control System Committee of Yokohama checks the triage algorithm regularly and encourages improvements to the algorithm.
References
1. Ohshige K et al. A descriptive study on the trend of ambulance utilization in an aging society, Yokohama, Japan. Yokohama Med Bull 2003;50:15-23.
2. Ellis RP et al. Supply-side and demand-side cost sharing in health care. J Econ Perspect 1993; 7: 135-151.
3. Anderson GM et al. A comparison of cost-sharing versus free care in children: effects on the demand for office-based medical care. Med Care 1991; 29: 890-898.
4. Ohshige K et al. A contingent valuation study of the appropriate user price for ambulance service. Acad Emerg Med 2005; 12: 932-940.
5. Ohshige K. Reduction in ambulance transports during a public awareness campaign for appropriate ambulance use. Acad Emerg Med 2008; 15: 289-293.
6. Pattenden J et al. Decision making processes in people with symptoms of acute myocardial infarction: qualitative study. BMJ 2002; 324: 1006-1009.
7. Gibler WB et al. Persistence of delays in presentation and treatment for patients with acute myocardial infarction: The GUSTO-I and GUSTO-III experience. Ann Emerg Med 2002; 39: 123-130.
8. Ohshige K et al. Evaluation of an algorithm for estimating a patient’s life threat risk from an ambulance call. BMC Emerg Med 2009; 9: 21.
9. Yokohama City Safety Management Bureau [Internet homepage]. Press release [updated 2009 November 13; cited 2010 January 25]. www.city.yokohama.jp/ne/news/press/200911/images/phpQ3Je6I.pdf (in Japanese).
The author
Kenji Ohshige, MD, PhD,
Department of Public Health,
Yokohama City University School of Medicine,
3-9 Fukuura,
Kanazawa-ku,
Yokohama 236-0004,
Japan.
Tel: +81-(0)45-787-7610,
Fax: +81-(0)45-787-7609,
E-mail: kenoh@med.yokohama-cu.ac.jp
