Article Text
Abstract
Objective There is lack of scientific evidence regarding the effectiveness of prehospital triage systems. This study compared the two-level Taiwan Prehospital Triage System (TPTS) with the five-level Taiwan Triage and Acuity Scale (TTAS) at ED arrival regarding the prediction of patient outcomes and the utilisation of medical resources.
Design This was a retrospective cohort study. Adult patients transported via the emergency medical service (EMS), who arrived at the ED of a medical centre in northern Taiwan during the study period were enrolled. TTAS acuity levels 1–2 were considered comparable to the designation of ‘emergent’ by the prehospital TPTS system. The outcomes were analysed by comparing TPTS and TTAS by acuity levels.
Results Among 4430 enrolled patients, 25.2% and 74.8% were classified as emergent and non-emergent by TPTS; 44.1% and 55.9% were classified as levels 1–2 and levels 3–5 by TTAS. Of the TPTS emergent patients, 15.2% were classified as TTAS levels 3–5, whereas 30.4% of TPTS non-emergent transports were classified as TTAS levels 1–2 at the ED. TTAS levels 1–2 showed better predictability than TPTS emergent level for hospitalisation rate with a sensitivity of 70.3% (95% CI 68.3% to 72.2%) versus 41.1% (95% CI 39.0% to 43.2%), and a negative predictive value of 74.8% (95% CI 73.4% to 76.0%) versus 62.6% (95% CI 61.7% to 63.5%).
Conclusion The current prehospital triage system is insufficient and inappropriate in classifying patients transported to the ED. The present study offers supporting evidence for the introduction of a five-level triage system to prehospital EMS systems.
- emergency care systems
- pre-hospital
- triage, emergency departments
- emergency ambulance systems, effectiveness
Statistics from Altmetric.com
- emergency care systems
- pre-hospital
- triage, emergency departments
- emergency ambulance systems, effectiveness
Key messages
What is already known on this subject?
An insufficient prehospital triage system results in treatment delays and overcrowded ED. It is necessary to evaluate a triage system to demonstrate its use as clinically beneficial and cost effective.
What this study adds?
In this retrospective study of 4430 transports to a single ED in Taiwan, the two-tier prehospital system was inferior to the ED five-tier system for prediction of admission. Our study suggests that a two-level prehospital triage system is insufficient.
Introduction
Background
Emergency medical services (EMS) provide prehospital medical care in situations of acute illness.1 The use of EMS has recently increased in many countries, with the potential to overload the system.2–5 Determining priority of care is a major factor when optimising response times and appropriateness, transportation organisation, resource allocation and disaster preparedness. A triage and acuity system is a scale for systemically prioritising patient care into different levels according to the severity of medical conditions.6 The quality of patient care may benefit from implementing a standardised prehospital triage scale and acuity categorisation process.4 5
Prehospital triage systems a re widely applied in different countries. In most countries, the triage system used has typically been chosen at the local agency level and shows enormous diversity.5 7 8 One study reported scales of prehospital triage systems ranging from two to five levels of urgency in different provinces in Australia.8 Although a valid and accurate triage system is critical for the EMS system efficiency, there i s lack of scientific evidence regarding the effectiveness of validated prehospital triage systems and of using the same triage system in two or more EMS settings.1 5 Indeed, the optimal prehospital triage scale remains unclear.9
Over the past decade, Taiwan has mandated the use of a two-level Taiwan Prehospital Triage System (TPTS) that classifies patients as emergent versus non-emergent according to their initial vital signs and chief complaints.10 The TPTS was initially derived from triage in a trauma scenario but added specific criteria for other patients who might require advanced healthcare services and transport them to the nearest hospital. The use of EMS and ambulance services has grown enormously in Taiwan. Due to the increasing needs of EMS services and crowding of the ED during the past decade,11 a prehospital triage system became important to identify patient at risk and to direct them to a hospital accordingly. The current TPTS in Taiwan has not been adequately validated and misclassifications occur. Its adequacy to identify those who are at risk, and to divert non-emergent patients to other resources, is a major concern.
Recently we have seen the development of a common triage system which can be used by the paramedics, ambulance communications officers and ED staff to practice an efficient diverting and routing based on patient’s conditions.5 12 13 The Ontario province EMS had applied a five-level Prehospital Canadian Triage and Acuity Scale (pre-CTAS) for an indicator of patient acuity in order to improve the communication between different emergency care providers and to assist EMS providers in determination of the most appropriate destination based on the acuity of the patient.14 This study aims to examine the effectiveness of a five-level triage system over the conventional two-level prehospital triage system in Taiwan.
Methods
Study design and setting
This was a retrospective cohort study. This study was approved by the Hospital Ethics Committee on Human Research of the Chang Gung Medical Foundation. The study protocol was reviewed and qualified as exempt from the requirement to obtain informed consent.
The research ED is at a medical centre in northern Taiwan; annual adult ED visits are approximately 1 20 000. Adult patients transported via the EMS system to the ED during the period from July to December 2012 were eligible for enrolment. A total of 63 155 patients arrived within the study period and 7.0% (4430/63 155) were transported via EMSs. All patients transported by EMSs were included in this study.
Two-level and five-level triage systems
In the prehospital phase, patients were triaged using the two-level TPTS. The criteria for TPTS emergent level patients are those who present with abnormal vital signs, including a GCS score <14, HR >140 or<50 beats/min (bpm), RR >30 or <10/min, systolic BP >220 or <90 mm Hg, body temperature >40°C or<32°C, and oxygen saturation, via pulse oximeter <90%, or high or low blood sugar via finger stick, or critical chief complaints suggestive of acute stroke, ischaemic heart disease, seizures, life-threatening intoxication, precipitate labour, snake bite, drowning or hypo/hyperglycaemia. In trauma events, criteria for emergent level include the presence of a specific physiological or anatomical injury or the evidence of a high-risk mechanism of injury, in addition to abnormal vital signs (table 1). All other patients were considered non-emergent.10
The patients were triaged according to the TTAS on arrival at the ED. The TTAS was modified from the CTAS, with a computerised decision support system (eTTAS) which had also been validated and shown to be reliable for Taiwan ED triage prioritisation.15–17 The TTAS was implemented by the Ministry of Health and Welfare in 2010. The TTAS retains most features of the CTAS but includes some major modifications as follows:
shortened intervals of reassessment;
division of the chief complaint list into two domains, non-trauma and trauma (the non-trauma domain is similar to that of the CTAS, and the trauma chief complaint list is categorised using anatomical region and environmental injury), with 179 total presenting complaints. The non-trauma system includes 14 categories and 132 chief complaint, and the trauma system includes 15 categories and 47 chief complaints;
explicit threshold levels for haemodynamic stability (eg, tachycardia/bradycardia (140 bpm/50 bpm) with or without symptoms of shock or a BP <70 mm Hg);
revised pain severity ratings, omitting chronic pain to accommodate ED conditions specific to Taiwan.
TTAS classifies patients in descending order of acuity: level 1, resuscitation; level 2, emergent; level 3, urgent; level 4, less urgent and level 5, non-urgent.
In Taiwan, the computerised triage decision support tool (eTTAS) links a standardised list of presenting complaints to preferred TTAS acuity level. Such a complaint-specific TTAS-based template with all relevant level 1–5 discriminators (modifiers) assists the ED nurse in assigning the appropriate triage score. Table 1 shows the difference between the TPTS and TTAS.
Data collection and outcome measures
A trained study assistant reviewed the de-identified computer-based registry records and conducted data abstraction using a standard reporting template with clear definitions and codes. Demographic data and prehospital covariates were collected from the EMS records, including the age and gender of the patient, TPTS level, TTAS level at ED arrival, final disposition (death, admit to intensive care unit (ICU), admission to ward and discharge) and ED medical expenses in New Taiwan dollar (NTD). The medical resources utilisations were defined as the sum of medical expenses that incurred at the ED including blood test, image studies, drugs, procedures and physician fees. All items were reimbursed according to a price list by the Taiwan National Health Insurance.
Primary data analysis
All data elements reported on the study data sheet were entered into a Microsoft Excel spreadsheet by a single investigator. Study data were summarised using descriptive statistics. Data were analysed using the triage level of the two systems as the independent variable. Categorical variables are presented as numbers and percentages and were compared using the χ2 test. Continuous variables are presented as means and SDs or as medians and IQRs. TTAS acuity levels 1–2 were considered comparable to the designation of ‘emergent’ by the prehospital TPTS system. Hospitalisation rates (including ICU admission and death) and ED medical expenses were analysed according to the acuity levels. In order to compare the predictability of hospitalisation rate between these two triage systems, we calculated the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), negative likelihood ratio (LR−) in TPTS (emergent vs non-emergent) and TTAS (levels 1–2 vs 3–5).
Multiple linear regression models adjusting for age and gender were used to compare ED medical expenses among triage and acuity levels. ED medical expenses were log transformed before use in the multiple linear regression analysis. The effect size (f = F·df/n) was calculated to reflect the variation in indicators between the two triage systems.15
All data were analysed using the SAS software V. 9.3. A p value <0.05 was considered to indicate statistical significance.
Results
Basic demographics
There were 4430 adult patients who were sent to the ED of study hospital via the EMS system from July to December 2012. Table 2 shows the demographic characteristics and outcomes of the study population. Of the enrolled patients, 25.2% and 74.8% were classified as emergent and non-emergent by TPTS; 44.1% and 55.9% were classified as levels 1–2 and levels 3–5 by TTAS. 54.1% of the enrolled patients were non-trauma cases. Among the enrolled patients, 36.7% were admitted to the hospital, 8.4% were admitted to the ICU and 2.4% were dead.
Distribution of TPTS to TTAS
Table 3 compares the distribution of patients from two-level TPTS to five-level TTAS acuity score. Many of the TPTS emergent and non-emergent patients were widely redistributed using the five-level TTAS. Of the patients deemed emergent in the TPTS, 15.2% (170/1117) were classified into TTAS levels 3–5 at the ED, indicating no life-threatening condition or time-sensitive illness. Contrarily, of the patients deemed non-emergent in the TPTS, 30.4% (1008/3313) were classified into TTAS levels 1–2 at the ED (table 3). The top five chief complaints of the TPTS non-emergent patients who were classified into TTAS levels 1–2 at the ED included shortness of breath (11.9%), syncope (7.7%), fever/chills (5.6%), blunt injury to the face (5.2%) and chest pain (5.1%). The TPTS emergent patients who were classified into TTAS levels 3–5 presented mostly with chief complaints of abdominal pain (9.9%), head injury (9.9%), blunt injury to the lower limbs (7.2%), scalp laceration/abrasion wound (5.9%) and blunt injury to the upper limbs (4.6%).
Patient outcomes in TPTS and TTAS
Tables 4 and 5 show the final disposition status in TPTS emergent and non-emergent patients compared with TTAS level. The mortality rate was 8.4% (94/1117) for TPTS emergent level versus 5.3% (104/1955) for TTAS levels 1–2 and 0.3% (10/3313) for TPTS non-emergent level versus 0% for TTAS levels 3–5. The ICU admission rate was 22.5% (251/1117) for TPTS emergent level versus 16.9% (331/1955) for TTAS levels 1–2 and 3.7% (123/3313) for TPTS non-emergent level versus 1.7% (43/2475) for TTAS levels 3–5.
Relationship between both triage system with hospitalisation rate and ED medical expenses
The hospitalisation rate was 77.4% for TPTS emergent level versus 75.6% (1478/1955) for TTAS levels 1–2 and 37.4% for TPTS non-emergent level versus 25.3% (625/2475) for TTAS levels 3–5. The sensitivity, specificity, PPV, NPV, LR+ and LR− of TPTS emergent level to predict hospitalisation rate of patients were 41.1% (95% CI 39.0% to 43.2%), 89.1% (95% CI 87.8% to 90.4%), 77.4% (95% CI 75.1% to 79.5%), 62.6% (95% CI 61.7% to 63.5%), 3.78 (95% CI 3.33 to 4.29) and 0.66 (95% CI 0.64 to 0.69), respectively (table 6). The sensitivity, specificity, PPV, NPV, LR+ and LR− of TTAS levels 1–2 to predict hospitalisation rate of patients were 70.3% (95% CI 68.3% to 72.2%), 79.5% (95% CI 77.8% to 81.1%), 75.6% (95% CI 74.0% to 77.1%), 74.8% (95% CI 73.4% to 76.0%), 3.43 (95% CI 3.15 to 3.73) and 0.37 (95% CI 0.35 to 0.40), respectively.
The median ED medical expenses were NT$9604 (IQR 5774–16 473) and NT$4180 (IQR 1999–8914) for TPTS emergent and non-emergent level, respectively (table 7). By comparison, the median medical expenses applying TTAS were NT$12 992 (IQR 7887–19 557), NT$8614 (IQR 4874–13 767), NT$3785 (IQR 1946–7787), NT$1899 (IQR 1379–3439) and NT$1669 (IQR 1058–2226) for acuity levels 1–5, respectively. The effect sizes of the medical expenses in TPTS and TTAS were 0.070 and 0.565, respectively. A higher differentiability of medical expenses was found with the five-level TTAS (F=626.6) compared with the two-level TPTS (F=313.3) after adjusting for age and gender using multiple linear regression models (table 7).
Discussion
This study compared a two-level prehospital triage and an ED five-level triage system for 6 months of ED attendances from a medical centre in northern Taiwan. Our study demonstrated that the TTAS led to greater discrimination and predictability among triage patients in terms of measured outcomes, such as hospitalisation rate and ED medical resource utilisation, compared with TPTS. TTAS appeared to be a more robust screening tool for its higher sensitivity and significant better NPV compared with TPTS. The results suggest that the current two-level prehospital triage system is insufficient and inappropriate. These results are comparable to the published reports, which examined the five-level triage system versus fewer level triage systems in the ED16 18–20
A validated triage system should not only be consistent with medical needs but also predict outcomes, such as morbidity, mortality and hospitalisation.18 20–22 On the basis of this study and other published reports, the need for standardised, evidence-based guideline development for EMS triage system is apparent.23 24 As the healthcare systems become more complex, and demand and costs rise, there is a call for using a common language triage system to efficiently allocate resources, improved patient outcome and survival.5 12 13
All major five-level ED triage systems have been well developed and validated.16 18 19 25–27 There are also studies showing that computer decision support further aids reliability.16 21 22 28 29 As clinical medicine embraces clinical information systems, designing triage scales that are easily programmable and assist tier accuracy through clinical decision support is important.16 21 22 28 29 Instead of introducing a newly designed triage system, a five-level triage with higher discrimination and with computer decision support system, such as the TTAS, could be implemented into prehospital setting.
Several aspects of our study results should be discussed further. First, according to the study results, 30.4% of non-emergent patients in TPTS were classified into TTAS levels 1 and 2 at the ED and were considered under triaged. 15.2% of emergent patient were classified into TTAS levels 3–5 and were considered over triaged. Although over-triage rate may exhaust resources prematurely, the over-triage ratio was not higher than the published literature report; 4 30 31 instead, under triage is of major concern as it may endanger patient safety and cause unfavourable outcome. It might not be feasible to divert non-emergent patient to other resources according to the current TPTS.
The current TPTS is primarily trauma based and only recognise eight critical medical conditions yet the majority of ambulance transports were due to medical problems, rather than trauma (54.1% vs 45.9%) in this study. Thus, the TPTS may overlook certain critical complaints, such as shortness of breath, syncope and fever which might be present in patients with life-threatening conditions. Second, a pain scale is not specified in the prehospital triage system. A pain scale is considered a fifth vital sign,32 indicating an emergency condition and the need for immediate relief. Besides, a serial pain scale measurements can reveal the course of the disease and the effects of the treatment. Thus, a pain scale should be considered in the prehospital triage system. Finally, the less structured, lack of formal educational programme and implementation guideline in TPTS system may also lead to problems in inaccurate classification of the patients. For those complaints, conditions or vital signs not listed in the guideline, the EMS paramedics may not recognise the underlying condition or may lack confidence to assign an appropriate triage score for the patient. Several studies have shown under-triage rates varying from 5% to 17.9%.23 24 33–35 Our study had a higher under-triage rate, which could be attributed to the less structure and lack of formal training and implementation guidelines in the current TPTS system. In contrast, there were specific implementation guidelines of the TTAS with standardised training materials. A more structured, five-level triage system might aid decision-making by EMS personnel.
A universal triage system in both the prehospital and ED settings might improve communication between EMS paramedics and ED medical staff, facilitating the continuity and consistency of care. Two recent studies that evaluated usability of CTAS and Emergency Severity Index (ESI) by paramedics in the prehospital setting have shown moderate inter-rater reliability between EMS providers and ED triage nurse,13 36 whereas another study showed that paramedics using CTAS can predict admission comparably to ED nurses using the ESI.35 These studies support the use of a valid five-level triage system in the prehospital setting. Further prospective multicentre studies of a five-level triage system in the prehospital setting to aid in the development of a uniform triage and communication strategy between paramedics and ED staff is warranted.
Limitations
This study should be interpreted in the context of the following limitations. First, there were unmeasured confounding factors, such as transportation duration, prehospital management, EMS personnel and the triage nurses. Patient presentation at the scene might differ from that after arrival at the ED. Transportation duration and prehospital management might have affected the patients’ conditions, resulting in a change in acuity levels among the patients at the ED. In addition, there are subjective modulators in the triage system, such as the EMS personnel and triage nurses. Second, this study was conducted at a medical centre in northern Taiwan, within a limited period, which may have restricted the general applicability of our findings. Further validation studies conducted in different settings and regions would be of interest.
Conclusions
The current prehospital two-level triage system is insufficient and inappropriate. The five-level TTAS provided greater acuity discrimination and more valid predictability of hospitalisation rate and medical expenses than the two-level prehospital system. The present study offers supporting evidence for the introduction of a five-level triage system into prehospital EMS systems.
Acknowledgments
The authors specially thank for the statistical support from the Healthy Aging Research Center at Chang Gung University, Taiwan.
References
Footnotes
Contributors All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. C-JN had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. L-HT and C-HH carried out analysis and interpretation of the data, drafting of the manuscript and acquisition of the data. Y-CS conceived the study, participated in its design and coordination and helped to draft the manuscript. Y-MW and C-HC performed the statistical analysis and given technical or material support. C-WK participated in the material support and obtained funding. W-CL participated in acquisition of the data.
Funding This study is also indebted to grant funding from the Ministry of Science and Technology, Taiwan (Fund No. NMRPG3d0181).
Competing interests None declared.
Ethics approval Ethics approval was provided by the Ethics on Human Research of the Chang Gung Medical Foundation.
Provenance and peer review Not commissioned; externally peer reviewed.