The care of patients with sepsis and trauma requires the delivery of appropriate definitive care in the early stages of the illness. Hospitals with limited resources, those in rural and underserved areas of South Carolina, may be unable to consistently provide optimal care to these patients. In addition, the shortage of specialists nationally makes it more difficult for these hospitals to recruit and retain the specialists needed. Patients in these areas continue to pay the rural penalty of poorer outcomes. This study provides specialists' level care through telemedicine consults to rural emergency departments in rural areas of SC to improve outcomes for these patients.

The CREST study is a project that specifically addresses the need to bring health care to rural communities in SC, as well as evaluates methods and tests technology to implement this care in rural communities. The CREST study uses telemedicine remotely from MUSC to rural community hospitals to provide rural community physicians care from specialists for trauma and sepsis, which are both high acuity, difficult to treat conditions.

CREST is a multi-site trial of telemedicine services to meet rural patients' and providers' need for expert evaluation and management of sepsis and trauma. The specific aims of CREST are:

1. To test the hypothesis that a telemedicine program including education and clinical consultation between a tertiary care academic medical center and rural, local hospitals will significantly improve key treatment decisions and outcome measures in sepsis and trauma.
2. To test the hypothesis that the differences in ISS and time to antibiotics for trauma and sepsis patients exposed to telemedicine intervention and those without the intervention matched on propensity scores are not due to unmeasured confounders.

CREST seeks new solutions to rural health disparities, to advance technology, create and retain jobs and address important research opportunities by combining implementation of a novel, trans disciplinary clinical program with rigorous, mixed methods scientific evaluation including clinical, process, and economic outcome measures. The impact on both science and quality healthcare outcomes is broad and CREST has far reaching implications for addressing rural health disparities for acute, life-threatening illnesses.

The impetus for CREST derives from the observation that rural patients are unlikely to acquire on-site access to appropriate specialist care for high stakes, high acuity, crisis-mode conditions such as sepsis and trauma. Novel, collaborative approaches are necessary to provide rural patients with access to appropriate specialist care for sepsis and trauma. CREST is an innovative, technologically levering mechanism that provides the opportunity to lessen the inferior outcomes and "rural penalty" paid by patients in rural communities.

Rural patients are less likely to have onsite access to specialist care for sepsis and trauma for several reasons:

• Personnel: there is a nation-wide shortage of appropriate specialty providers including intensivists, trauma surgeons, and other critical care staff 1. Small, rural communities lack the economies of scale to support resource intense services and if a community has these specialties, there is often not around the clock coverage necessary for best quality care for these conditions 1.
• Physical and material resources: Sepsis and trauma patients require immediate availability of resources such as multidisciplinary ICU staff, large blood banks, operating rooms, and advanced radiological facilities.
• Availability of resources such as multidisciplinary ICU staff, large blood banks, operating rooms, and advanced radiological facilities.
• Location and distance: Patients in rural communities have inferior outcomes in sepsis and trauma 2, 3. This is likely multifactorial and includes delays in entry to the Emergency Medical System network, longer travel times to definitive care facilities and delays in definitive care delivery. Inter-hospital transfer may be limited both by severity of patient illness, acute care and ICU bed shortages in tertiary care centers. Additionally, the scarcity of tertiary care hospital beds increases the importance of initiating transfer for patients who will maximally benefit from tertiary care referral 4, 5.

The impact to rural patients with sepsis and trauma, due to lack of access to appropriate specialist care, is substantial. Sepsis is one of the most common serious critical illnesses. It kills 28-50% of patients diagnosed 6, 7 and is the 10th leading cause of death in the US 8. Furthermore, the incidence is increasing in conjunction with the aging of the US population 9. The impact of rurality on sepsis has received modest investigation, but rural patients likely have worse outcomes.3 In SC, patients referred from Emergency Departments (ED's) in smaller, typically rural hospitals have a 22% increase risk of death as compared to those referred from ED's in larger, urban hospitals (CREST preliminary analysis) suggesting the benefits of specialist directed care available to patients at larger facilities. Equally concerning is an apparent race-associated disparity in which African American patients presenting to rural ED's in SC are twice as likely to be discharged from the ED with a diagnosis of sepsis as compared to white patients. Although the rates for discharge from an ED with a diagnosis of sepsis are low at 3.95% for whites and 8.71% for African Americans (CREST preliminary analysis), it is concerning that any patient is discharged from an ED with this diagnosis and the disparate rates may reflect the well-described phenomenon of unequal treatment 10. Critically ill patients with any diagnosis benefit from intensivist-directed care including a 30-40% reduction in mortality.11 Unfortunately, over half of US hospitals have no intensivist coverage at all and rural hospitals are disproportionately affected by this problem 12.

Likewise, trauma is the leading cause of death among Americans up to age 45 and the fourth leading cause of death overall for all ages 8. Deaths from injury occur in a tri-phasic distribution: 50% of deaths from injury occur at the scene, 30% occur in the first 1-2 hours and 20% occur during hospitalization between 1-2 weeks after admission. Trauma care at designated trauma centers improves survival 13. Yet, rural hospitals are rarely trauma centers and in SC patients injured in rural communities have inferior outcomes and higher costs 14. Significant disparities can be identified among SC's elderly and minority populations. 14 An important barrier to facilitating quality care for SC trauma patients relates to triage efficiency between rural hospitals and a trauma center. Outcomes for seriously injured rural patients are negatively affected when patients are taken to non-trauma EDs and then transferred to a trauma center 13 . Conversely, transfer of patients with low risk of death and disability to trauma centers represents over-triage, and creates additional stressors at already busy trauma centers accruing financial burden on the health care system.

CREST is scientifically innovative on multiple levels. It is the first prospective, mixed-methods evaluation of telemedicine; as well as one of the few studies in remote/telemedicine not sponsored by industry. CREST is the first study to demonstrate improved outcomes for sepsis using telemedicine and will augment the sparse literature on trauma and telemedicine. CREST also serves as a demonstration project showing that access to specialist care for rural patients with high stakes, high acuity conditions can be reliably provided through telemedicine. CREST technology is non-proprietary and relatively low cost and thus allows important cost-effectiveness analyses to ascertain the viability of telemedicine in rural settings.

CREST includes two elements - an educational program for rural providers and clinical consultation for patients with sepsis and trauma. Collaborating hospitals receive telemedicine hardware and software, usage instruction, and 24/7 access to MUSC experts in sepsis and trauma for patient consultation in real-time. This project uses a quasi-experimental design, which incorporates a controlled experiment using pre-existing patient groups as the controls. This design minimizes threats to external validity and offers efficiency in the longitudinal aspect of the research. Main dependent measures include time to initiation of appropriate antibiotics for septic patients and reduced transfer rates of less-severe trauma patients as measured by average increase in the Injury Severity Score (ISS). Data collection for the controls in the retrospective groups and the prospective intervention group will be conducted by using structured audit forms by trained staff. The research plan includes investigation of outcomes, including treatment decisions.

Rural hospitals are the primary targets for CREST because these institutions tend to have fewer specialty providers and often cannot provide definitive care for patients with sepsis and trauma. Eligible rural hospitals are strategically identified based on several criteria. First, the hospital is located in a rural area as designated by the US Department of Agriculture. Second, the hospital has substantial resource limitations as indicated by a Level III or IV Medical Facility Status from the Joint Accreditation Committee for Hospital Organizations. Third, the hospital is within MUSC's typical 120 mile catchment area.

CREST investigators are mindful of the importance of collegial and collaborative approaches to recruiting rural hospitals and providers. To establish a framework for the working relationships important to the success of CREST, collaborating hospitals and MUSC sign a Memorandum of Agreement stipulating the roles and expectations. The fundamental duties for the rural hospitals are: 1) work with MUSC research team to facilitate human subjects approval of CREST; 2) coordinate their providers' availability with CREST's educational program; 3) facilitate working relationships between MUSC clinical experts and appropriate clinicians at rural hospitals and 4) allow CREST investigators access to patient records to ascertain outcome measures. MUSC provides: 1) an educational program 2) the technical expertise and equipment; 3) 24/7 availability for clinical evaluations; 4) training and support of chart abstractors to acquire outcome data; and 5) the conducting of analyses and writing of reports for CREST.

CREST allows rural spoke hospitals 24/7 access to MUSC sepsis and trauma experts who provide formal consultation for patients with these conditions including appropriate documentation in the medical record. Rural hospitals can utilize the telemedicine consultants on a one-time basis for initial evaluation and recommendations or request ongoing collaborative care with MUSC clinical experts. If a patient benefits by transfer to MUSC, this is identified and MUSC clinicians are already familiar with the patient's clinical status, representing another advantage of telemedicine services. Alternatively, if patients can be adequately managed at referring spoke sites this is determined, thus improving triage efficiency for tertiary care referrals.

Inclusion Criteria:

Sepsis study patients are adult patients presenting to the rural hospital's EDs that meet the screening criteria for sepsis and two of four SIRS criteria.

The screening criteria for possible sepsis are:

• known or suspected infection AND EITHER
• systolic blood pressure < 90mmHg after fluid challenge of 30mL/kg over 30 minutes OR
• lactate > 4mmol/L

The screening criteria for possible SIRS are:

• abnormal core body temperature (<36.5 C or > 38 C)
• heart rate > 90 bpm
• respiratory insufficiency (respiratory rate > 20 breaths/min or PaCO2 < 32 mmHg or mechanical ventilation)
• abnormal white blood count (>12,000/cumm or < 4000/cumm)

Trauma study patients are adult patients presenting to the rural hospital's EDs with blunt or penetrating injury and one or more of the following criteria:

• Respiratory insufficiency and/or intubation
• Glasgow coma scale < 8 with mechanism attributed to trauma
• Systolic blood pressure of <90mmHg
• Paralysis
• Penetrating injury to the head/neck/torso
• Crush to torso/upper thighs
• Major amputations
• Loss of consciousness > 5 minutes
• Maxillo-facial trauma
• Significant subcutaneous air
• Evidence of pelvic instability
• Two or more long bone deformities
• Major lacerations involving fascia
• Ejection from vehicle
• Pedestrian struck > 15 mph
• Motorcycle Crash > 25 mph
• Motor Vehicle Collision > 35 mph
• Documented falls > 20 feet or 2 stories

Exclusion Criteria:

The exclusion criteria for septic patients are:

-Age < 18 years

The exclusion criteria for trauma patients are:

• Age < 18 years
• Burn injury of greater than 10% total body surface area
• Traumatic asphyxiation
• Lightning strike/electrical shock
• Patients undergoing cardiopulmonary resuscitation (CPR) on arrival

The study population is adult patients presenting to rural community hospital's emergency departments wth possible sepsis or severe sepsis and/or blunt or penetrating trauma.