Study design
This is a retrospective registry-based study with data extracted from a prospective cohort study called Towards Improved Trauma Care Outcomes in India (TITCO). TITCO study is a multi-centre research consortium of university hospital formed to develop a trauma registry in India.
Setting
The study was conducted in four public university hospitals in India between October 2013 and December 2015. The hospitals included in the study are from three metropolitan cities, namely Mumbai, Delhi and Kolkata. The hospitals were King Edward Memorial Hospital (KEMH) and Lokmanya Tilak Municipal General Hospital (LTMGH) in Mumbai, Jai Prakash Narayan Apex Trauma Centre (JPNATC) in New Delhi and the Institute of Post-Graduate Medical Education and Research and Seth Sukhlal Karnani Memorial Hospital (SSKM) in Kolkata.
The urban referral trauma centres are situated in Kolkata, Mumbai (2-centers) and Delhi, cities with populations of more than 10 million. Except for the JPNATC, which is a standalone trauma centre, the others are trauma units providing trauma care as a part of a general hospital. The user fees are nominal and classified as free-to-public. The hospitals mainly serve the lower socioeconomic strata of the population in their respective area. Each of these hospitals receive 40 to 100 major trauma patients per week. They have round the clock emergency services, imaging, operating theatres and sub-speciality available.
Source and method of participant selection
All admitted patients that presented with history of trauma on arrival to any of the study hospitals were included in the TITCO registry. Data of patients with liver trauma either isolated or concomitant with other injuries was extracted using the ICD-10 code, S36.1 for liver injury.
Data Collection
Project officers included those with a master in science, who were then trained in the methods of data selection for the study in a workshop format, for a period of one week. These trained project officers at each hospital, worked eight-hour shifts with a rotating schedule between day, evening and night shifts through all days of the week. Data from patients admitted outside of the shift hours was collected retrospectively from the hospital medical records. The patients were followed up until discharge, death or to a maximum of 30-days. If discharged before 30 days, the patients were considered to be alive at 30 days. There was no follow-up after patient discharge or after the 30 days.
Study Variables
The primary outcome was 30-day in-hospital mortality following liver injury. Patients who died during their hospital stay up to 30 days was recorded. Those discharged before 30 days were considered to be alive at 30 days. The data set was analysed for patients’ demographic profile, mechanism of injury, severity, management and outcome.
The data also included serially recorded parameters like pulse, systolic blood pressure (SBP), Glasgow Coma Score (GCS) and interventions done, if any. Those patients with a systolic blood pressure of ≤90 mmHg were considered as hemodynamically unstable having hypotension.
The severity of injury has been graded based on the World Society of Emergency Surgery (WSES) guidelines. WSES grading of liver injuries has graded based on the American Association of Surgery for Trauma (AAST) scale (anatomical classification of liver injuries) and the hemodynamic stability (physiological parameter) for grading liver injuries from I-IV [8]. The classification has been added as additional file. (see additional file A-1).
Patients management was divided and labelled as operative management (OM) in those who underwent laparotomy and NOM in those who were conservatively managed without a laparotomy. Those patients who survived NOM were labelled as successfully managed. The patients who died after NOM were labelled as NOM failure. The overall management of these patients along with the treatment for other associated injuries was recorded.
Quantitative variables
All continuous variables were represented as mean with their standard deviation and categorical variables as counts and proportions. ISS was represented as median with inter-quantile range.