Study Design
This single-centre prospective cohort study is part of the Trauma Triage Study in India (TTRIS) which compares prediction models for triage in adult trauma patients presenting to various emergency departments across India. Data was collected in the study site from July 2016 to November 2019.
Study Setting
The study site was the ED of Khurshedji Behramji Bhabha Hospital (KBBH), a 436 bedded regional secondary healthcare centre located in Mumbai, India, catering to approximately 350 patients each day in the ED. It is a secondary-care public hospital with free or nominal fees, providing access to low socio-economic groups and receives patients from across the city. At KBBH, trauma care is imparted as a subspeciality along with medical, surgical, and obstetric care. The hospital has an intensive care facility but there is no neurosurgery department, so patients in need of neurosurgical management are referred to tertiary care centres. Plain radiography and ultrasonography are available round the clock; however computerized tomography (CT) is only available in-house from 7am to 6pm. The patients arriving at the ED are first seen by a casualty medical officer (CMO) largely on a first-come, first-served basis without a formalised system of triaging patients at the ED.
Clinician Triage
As part of data collection of TTRIS, the triage-naive clinicians were informed about the trauma triage categories, without provision to any formal tool or training about the same. The clinicians involved have a minimum of 2 years clinical experience, however, they are neither trained in trauma care as a speciality nor are they necessarily trained in trauma management courses like ATLS. Medical training of the clinicians include aspects such as triage systems in theory however not put in practise due to lack of formal triage protocols. After their initial on-arrival assessment of each patient, clinicians were asked by the research officers to categorise the patient as per their understanding of how urgently the patient requires treatment into the aforementioned colour-coded triage groups; red, orange, yellow, green, with red and green denoting the most and least urgent patient status respectively,(15) henceforth referred to as the triage levels. The clinicians were allowed to use all available information that was extracted by them during initial routine assessment (such as wound assessment) along with patient vitals that were collected by the research officer for determining the urgency of treatment required and thereby the triage level. The triage levels were not used to determine treatment decisions in the ED as there was no formalised tool or protocol in place for assigning the triage and coupling it with patient management. The clinicians were individually informed about the aim and methodology used for the TTRIS study at the start of their respective posting at the ED, however, the clinicians were neither involved in the conception nor were they part of the research team analysing the results.
Participants
Inclusion Criteria
We included all the patients aged 18 years and above presenting to the KBBH ED with a history of trauma and triaged green by clinical triage on initial evaluation irrespective of their injury severity. A history of trauma was defined here as having any of the external causes of morbidity and mortality listed in block V01-Y36, chapter XX of the International Classification of Disease version 10 (ICD-10) online codebook, with some exclusions (see online supplementary material).(20) Only patients with above mentioned causes of trauma as their primary complaint were included. For example, patients with a history of fall due to dizziness were not included in the study.
Exclusion Criteria
Patients with one or more vital parameters missing among the variables used for analysis or who did not consent to follow-up were excluded from the analysis.
Source and methods of selection of participants and follow-up
The research officer at KBBH observed morning, evening and night shifts (6-hour observational shifts). These shifts were not aligned with the working hours of the clinical staff to reduce bias and accounting for shift fatigue of the clinicians. Data were collected from the first 10 consecutive patients only, irrespective of their triage, during each shift. Due to the large patient load and time and budgetary constraints of the project, data collection of only the first 10 patients was considered feasible for follow-up.
Follow-up was performed by the research officer at 24-hours, 30-days and 6-months after arrival at the ED. The time frame of the study was chosen to ensure that all included patients had completed 6 months of follow-up to minimise the loss to follow-up. The follow-up was completed in-person or by telephone, depending on whether the patient was still hospitalised or if the patient had been discharged. The phone numbers of one or more contact persons, mostly relatives, were collected on enrolment and those people were contacted if the participant did not reply to the follow-up telephone calls. The outcome was recorded as missing if neither the patient nor the relative were available for follow-up at the specified time-points.
Variables
The outcome measures were mortality within 24-hours, 30-days and 6-months.
Additionally, for each participant, age, sex, transfer status, time of injury, mechanism of injury, injury-related details, number of serious injuries, and the assigned informal triage category were collected. Physiological measures including systolic blood pressure (SBP), respiratory rate (RR), heart rate (HR), peripheral capillary oxygen saturation (SpO2), Glasgow Coma Scale (GCS) and Alert Verbal Pain unresponsive scale (AVPU) were recorded.
A serious injury was defined as an injury that warrants hospitalisation.(21) GCS was categorised into no or mild traumatic brain injury (TBI) (13–15), moderate TBI (9–12), severe TBI (3–8).(22) ED-length of stay (ED-LoS) was calculated using the date and time of arrival and the date and time of discharge from the ED, either to be sent home or admitted in the hospital. Length of stay in the hospital (LoS) was calculated using the data and time of admission in the hospital to the data and time of discharge alive from the hospital, mortality, Leave against medical advice (LAMA) or abscond. Injury severity score (ISS) was allocated retrospectively with ‘mild’ (3–8), ‘moderate’ (9–15), ‘severe’ (16–25) and ‘profound’ (> 25) categories. Patients for whom ISS could not be coded, for example when there were no recorded injuries, were assigned ‘no defined ISS’.(23) The revised trauma score (RTS) was computed and categorised as RTS < 4 and RTS > 4.(23, 24)
Injuries were recorded and coded using ICD-10 in the TTRIS dataset. Patients were divided into categories with respect to the most critical injury namely, crush injury, injury to internal organs, blood vessel injury, amputation, fracture, dislocation, burn, multiple injury, unspecified injury, open wound, superficial injury.(20) Injury characteristics of patients that presented to the ED with no injuries were categorized as ‘no defined injury’. For patients with multiple injuries, the more critical one was considered for categorising patients as per injury.
Triage as per Cape Triage Score (CTS) System
The CTS has three versions, the adult version (those over 12 years of age or 150 cm in height), the child version (those between 3–12 years old or 95–150 cm) and the infant version (those less than 3 years of age or under 95 cm).(18) The adult CTS was used retrospectively to check for the appropriateness of the informal triage performed by the clinicians.(18) The physiological parameters were each scored against the adult Triage Early Warning Score (TEWS) scoring sheet, to calculate a total TEWS (Table 1). Each patient was assigned a triage level that corresponds to the TEWS score as follows: 0–2 green, 3–4 yellow, 5–6 orange, 7 or more red as seen in Table 2. Each patient was further categorised into the four triage levels using the Cape Triage Group (CTG) list of discriminators, also known as the South African Triage Scale (SATS) colour code. If a patient was categorised into a higher level by the SATS colour code than the TEWS score, then the higher level was considered as the correct triage level. For example, if the SATS color code categorised a patient with closed fracture into yellow, the triage level assigned to the patient would be yellow even if the TEWS score categorised the patient into green. The physiological parameters are considered for assigning tirage and the discriminators are used as a safety net in case patients do not present with abnormal physiology.
Table 1
The Cape Triage Score depicting the TEWS (18)
| Adult Triage Score |
| 3 | 2 | 1 | 0 | 1 | 2 | 3 |
Mobility | | | | Walking | With help | Stretcher/ Immobile | |
RR | | Less than 9 | | 9–14 | 15–20 | 21–29 | More than 29 |
HR | | Less than 41 | 41–50 | 51–100 | 101–110 | 111–129 | More than 129 |
SBP | Less than 71 | 71–80 | 81–100 | 101–199 | | More than 199 | |
Temp. | | Less than 35 | | 35-38.4 | | 38.5 or more | |
AVPU | | | | Alert | Reacts to Voice | Reacts to Pain | Unresponsive |
Trauma | | | | No | Yes | | |
Over 12 years / taller than 150 cm |
Table 2
The Cape Triage Score depicting the SATS color code (18)
Colour | Red | Orange | Yellow | Green |
TEWS | 7 or more | 5–6 | 3–4 | 0–2 |
Target time to treat | Immediate | Less than 10 min | Less than 60 min | Less than 240 min |
Mechanism of injury | | High energy transfer | | |
Presentation | | Shortness of breath - acute | | All other patients |
| Coughing blood | |
| Chest pain | |
| Haemorrhage uncontrolled | Haemorrhage - controlled |
Seizure - current | Seizure-post ictal | |
| Focal neurology - acute | |
| Level of consciousness reduced | |
| Psychosis/aggression | |
| Threatened limb | |
| Dislocation - other joint | Dislocation - finger or toe |
| Fracture - compound | Fracture - closed |
Burn-face/ inhalation | Burn over 20% | Burn - other |
Burn - electrical | |
Burn-circumferential | |
Burn-chemical | |
| Poisoning/overdose | Abdominal pain |
Hypoglycemia glucose less than 3 | Diabetic - glucose over 11 & ketonuria | Diabetic - glucose over 17 (no ketonuria) |
| Vomiting - fresh blood | Vomiting - persistent |
| Pregnancy and abdominal trauma or pain | Pregnancy and trauma |
| | Pregnancy and PV bleed |
Pain | | Severe | Moderate | Mild |
Senior health care professional's discretion |
Quality assurance
There were three layers of quality control. First, data was entered using a dedicated electronic data collection instrument with extensive logical checks and prompts for unlikely but possible values. Second, the collected data were reviewed on a weekly basis and discussed during weekly online conferences with all project officers and the project leads throughout the duration of the data collection period. Third, on-site quality control sessions were conducted every 3–4 months. During these sessions, a second research officer collected data alongside the research officer who worked at the ED. The quality-controlled data was then compared with the standard data.
Data Analysis
Data analysis was performed using R version 4.04 statistical software.(25) Complete case analysis was performed to exclude all patients with missing data among the variables included for analysis. Descriptive statistics were generated for all variables. The results are presented as frequencies and percentages for categorical variables, mean and standard deviation (SD) for continuous variables and median and inter quartile range (IQR) for variables with abnormal distribution, for example LoS. The patients triaged green as per the clinician triage were further triaged as per the CTS, using the TEWS and SATS colour code into red, orange, yellow and green. The number of patients triaged green by the CTS was divided by the number of patients triaged green as per clinician triage (4135), the resultant proportion minus one was considered as the proportion of patients mis-triaged.