Study design, population, and setting
This was a retrospective observational study using the ORION database [8]. The study period spanned 1 year from January 2016 to December 2016. Osaka Prefecture is located in the western area of Japan, covers an area of 1905 km2, and has a population of 8.8 million. The proportion of male inhabitants was 48.1% and that of elderly people (aged ≥65 years) was 26.1% in 2015 [8]. There are 519 hospitals (106,273 beds) in Osaka Prefecture, of which 288 are emergency hospitals including 16 critical care centers that are designated to accept patients with life-threatening emergency conditions such as severe trauma and sepsis [8].
We included all emergency patients registered in the ORION database. We selected hospitalized patients with sepsis or suspected sepsis using the ICD (International Statistical Classification of Diseases and Related Health Problems) 10 codes shown in Supplementary Table 1. Each ICD 10 code was evaluated after hospital admission by physicians. Next, we excluded the following patients: those with i) missing data (outcome, Japan Coma Scale [JCS], respiratory rate, blood pressure), ii) respiratory rate ≥60/min, and iii) blood pressure ≥250 mmHg. These were evaluated by ambulance personnel when they first contacted the patient in the prehospital setting.
This study was approved by the Ethics Committee of the Osaka University Graduate School of Medicine (No. 15003). Personal identifiers were removed beforehand from the ORION database, and thus the patients’ right to informed consent was waived. This study was conducted based on the ORION database under the present researchers’ responsibility, and it differs from the statistics published by Osaka Prefecture. This research was not conducted by Osaka Prefecture. This study was written based on the STROBE statement to assess the reporting of cohort and cross-sectional studies [9].
ORION
In January 2013, the Osaka Prefecture Government first developed and introduced an information system for emergency patients (the Osaka Emergency Information Research Intelligent Operation Network [ORION] system) that uses a smartphone app for hospital selection by on-scene EMS personnel and since then, it has been accumulating all ambulance records. Furthermore, since January 2015, medical institutions have registered information on the diagnosis and outcome of emergency patients transported to medical institutions, and the ORION system has merged these data with the respective ambulance records and smartphone app data. This report describes the ORION system and its profile of hospital information, EMS characteristics, and in-hospital diagnoses and outcomes.
qSOFA score
The qSOFA was introduced with the Sepsis-3 criteria. The score ranges from 0–3 with 1 point assigned for each of the following criteria met by the patient: systolic arterial blood pressure ≤ 100 mmHg; respiratory rate > 21 breaths/min; or altered mental status [1]. For the prehospital evaluation of mental status, Japanese EMS providers have adopted the JCS instead of the Glasgow Coma Scale since its introduction in 1974 [10]. The JCS has four main grades (grade 0: alert; grade 1: possible verbal response without any stimulation, not lucid; grade 2: possible eye-opening, verbal and motor response upon stimulation; and grade 3: no eye-opening and coma upon stimulation). Therefore, in this study, an evaluation other than JCS 0 (alert) was defined as ‘altered mental status’ and was considered equal to a Glasgow Coma Scale score of ≤14.
Endpoint
The primary endpoint was discharge to death.
Statistical analysis
Patient characteristics and outcomes were evaluated between two groups using the Wilcoxon rank-sum test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. One-way analysis of variance (ANOVA) was used to evaluate the differences in mortality according to the qSOFA score. Multivariable analysis of the eligible patients was used to assess factors associated with the outcomes by using logistic regression models, and adjusted odds ratios (AORs) and their 95% confidence intervals (CIs) were calculated. Potential confounding factors (age [continuous value] and sex [male, female]) based on biological plausibility and previous studies were included in the multivariable analysis. To analyze the effectiveness of qSOFA positive/negative for predicting hospital mortality, we created a receiver operating characteristic (ROC) curve. A p value of ≤ 0.05 was considered significant. All statistical analyses were performed using JMP Pro 13 (SAS Institute Inc., Cary, NC, USA).