The Trauma Triage Study

A Comparison of Prediction Models for Triage of Adult Trauma Patients Presenting to Emergency Departments in Urban India

Summary

Introduction

In India a substantial share of trauma deaths happens in urban areas. Here, prehospital care is largely lacking or poorly organised. Ambulances are mainly used for transfer of patients between hospitals. Hence, even the most severe patients arrive to hospital without being assessed in the prehospital setting and then mix with the general patient population that presents to the emergency department. In such scenarios, one of the key challenges in trauma care is how to prioritise patients so that the patients in greatest need of care are seen first, i.e. triage. Our aim is to validate and compare published prediction models for use in early trauma care.

Design

Prospective multicentre cohort

Setting

Three hospitals in urban India

Participants

Any person aged > 18 years presenting to the emergency department (ED) of participating sites with history of trauma.

Primary outcome

Mortality within 30 days of arrival to hospital. Extracted by project officer from patient record as date and time of death or discharge. If the patient is discharged alive before 30 days of arrival the research officer will call the patient or patient relative 30 days after arrival to hospital to establish the patient’s status.

Secondary outcomes

• Mortality within 24 hours of presentation to hospital. Extracted by project officer from patient record as date and time of death or discharge. If the patient is discharged alive before 24 hours of arrival the research officer will call the patient or patient relatives 24 hours after arrival to hospital to establish the patient’s status.
• Mortality within six months of presentation to hospital. Extracted by project officer from patient record as date and time of death or discharge. If the patient is discharged alive before six months of arrival the research officer will call the patient or patient relatives six months after arrival to hospital to establish the patient’s status.
• Mortality in participating hospital. Extracted by project officer from patient record as date and time of death or discharge.
• Admission to ward/hospital admission. Extracted by project officer from patient record and/or ascertained during follow up calls with patient or relative.
• Admission to intensive care unit. Extracted by project officer from patient record and/or ascertained during follow up calls with patient or relative.

Covariates

Demographics, vital signs and injury data.

Statistical methods

We will compare models using measures of discrimination, calibration, and net benefit.

More