Objective To derive and validate a predictive model and novel emergency medical services (EMS) screening tool for severe sepsis (SS). Design Retrospective cohort study. Setting A single EMS system and an urban, public hospital. Patients Sequential adult, nontrauma, nonarrest, at-risk, EMS-transported patients between January 1, 2011, and December 31, 2012 were included in the study. At-risk patients were defined as having all 3 of the following criteria present in the EMS setting: (1) heart rate greater than 90 beats/min, (2) respiratory rate greater than 20 beats/min, and (3) systolic blood pressure less than 110 mm Hg. Interventions None. Measurements and Main Results Among 66,439 EMS encounters, 555 met the criteria for analysis. Fourteen percent (n = 75) of patients had SS, of which 19% (n = 14) were identified by EMS clinical judgment. In-hospital mortality for patients with SS was 31% (n = 23). Six EMS characteristics were found to be predictors of SS: older age, transport from nursing home, Emergency Medical Dispatch (EMD) 9-1-1 chief concern category of "sick person," hot tactile temperature assessment, low systolic blood pressure, and low oxygen saturation. The final predictive model showed good discrimination in derivation and validation subgroups (areas under curves, 0.843 and 0.820, respectively). Sensitivity of the final model was 91% in the derivation group and 78% in the validation group. At a predefined threshold of 2 or more points, prehospital severe sepsis (PRESS) score sensitivity was 86%. Conclusions The PRESS score is a novel EMS screening tool for SS that demonstrates a sensitivity of 86% and a specificity of 47%. Additional validation is needed before this tool can be recommended for widespread clinical use.

BACKGROUND: Emergency medical services (EMS) systems provide professional prehospital emergency medical care and transportation to help improve outcomes from emergency conditions. Ghana's national ambulance service has relatively low public utilization in comparison with the large burden of acute disease. METHODS: A survey instrument was developed using Pechansky and Thomas's model of access covering 5 dimensions of availability, accessibility, accommodation, affordability, and acceptability. The instrument was used in a cross-sectional survey in 2013 in Accra, Ghana; eligible participants were those 18 years and older who spoke English, French, or Twi. Although the analysis was mainly descriptive, logistic regression was used to identify factors associated with reported intention to call for an ambulance in the case of a medical emergency. RESULTS: 468 participants completed surveys, with a response rate of 78.4%. Few (4.5%) respondents had ever used an ambulance in prior emergency situations. A substantial proportion (43.8%) knew about the public access medical emergency telephone number, but of those only 37.1% knew it was a toll-free call. Most (54.7%) respondents believed EMTs offered high-quality care, but 78.0% believed taxis were faster than ambulances and 69.2% thought the number of ambulances in Accra insufficient. Many (23.4%) thought using ambulances to transport corpses would be appropriate. In two hypothetical emergency scenarios, respondents most commonly reported taxis as the preferred transportation (63.6% if a family member were burned in a house fire, 64.7% if a pedestrian were struck by a vehicle). About 1 in 5 respondents said they would call an ambulance in either scenario (20.7% if a family member were burned in a house fire, 23.3% if a pedestrian were struck by a vehicle) while 15.5% and 10.2%, respectively, would use any available vehicle. Those aged 18-35 years were more likely than older respondents to prefer an ambulance (odds ratio [OR], 2.27; confidence interval [CI], 1.47 to 3.68), as were those with prior ambulance experience (OR, 1.75; CI, 0.98 to 3.09) (compared with those with no prior experience) and those who believed ambulances were safer than taxis (OR, 2.17; CI, 1.1 to 4.2) (compared with those who did not hold such beliefs). CONCLUSIONS: Perceptions of public ambulance services in Accra, Ghana, are generally favorable, although use is low. Public health education to improve awareness of the toll-free medical emergency number and about appropriate use of ambulances while simultaneously improving the capacity of ambulance agencies to receive increased caseload could improve use of the EMS system.