North Carolina's direct involvement in the 2001 anthrax investigations, and the state's previous experience assessing medical consequences resulting from Hurricane Floyd in 1999, required close collaboration between the North Carolina Division of Public Health (NCDPH), local health departments (LHDs), and many of the state's hospitals. In both events, state and local public health officials worked closely with hospital staff to conduct syndrome-based surveillance in response to a public health emergency. Several factors hindered the efforts of investigators. These included the limited relationships among organizations (hospitals and public health agencies) which resulted in less than ideal familiarity with each other's capacity, inefficient communication and information sharing, and, in the case of anthrax, the need to educate healthcare providers about the epidemiology of inhalational anthrax [3].

Recognizing the need for more efficient communication and collaboration between hospitals and public health agencies to better respond to future public health emergencies, NCDPH established an innovative program in 2003 that placed public health epidemiologists (PHEs) in hospitals around the state. Funded by the state's Public Health Emergency Preparedness cooperative agreement with CDC, PHEs are employees of their respective hospitals, serving NCDPH and LHDs as preferential partners.

As of 2010, eleven PHEs were based in the state's largest hospitals (including a Veterans Affairs Medical Center), covering approximately 39% of general/acute care beds and 30% of emergency department visits. The PHE program serves to "develop a communications infrastructure to facilitate and ensure the timely dissemination and transfer of information between the healthcare and public health sectors" as described in the HHS Pandemic Influenza Plan [4]. PHEs are responsible for: 1) surveillance, detection, and monitoring of community-acquired infections and potential bioterrorism events; 2) assisting LHDs with public health investigations; 3) educating clinicians regarding diseases of public health importance; 4) enhancing communication among clinicians, hospitals, and the public health system; and, 5) conducting special studies. In April 2010, the PHE program was recognized as a "promising practice" by the Center for Infectious Disease Research and Policy for aiding North Carolina's H1N1 response by shortening emergency response time [5].

As PHEs' contact with LHDs occurs primarily through communicable disease and tuberculosis (TB) control nurses, the lead individuals in these positions were surveyed regarding the importance of services provided by PHEs to LHDs, length of time for PHEs to respond to LHD requests, and impact of the PHE program on LHDs' capacity to detect and investigate communicable disease outbreaks and potential bioterrorism events. A link to an electronic (SurveyMonkey^®) survey was sent to 150 lead communicable disease and TB control nurses based in the state's 85 LHDs; 6 surveys were returned due to non-deliverable email addresses. The final sample size (N) for this survey was 144.

Semi-structured interviews were conducted with key informants at NCDPH who work closely with PHEs. Interviews focused on the specific ways NCDPH uses the network of PHEs for ongoing state-based syndromic surveillance and communicable disease management, the perceived value of the PHE program, and how NCDPH utilized the PHEs to respond to the H1N1 pandemic. Semi-structured interviews were also conducted with PHEs' hospital supervisors. Interviews focused on the services PHEs provide to their host hospitals and how valuable these services are to the hospital's ability to prepare for and respond to public health emergencies, such as the H1N1 pandemic.

Of the 144 communicable disease and TB control nurses surveyed, 119 completed the electronic survey (response rate = 82.6%). These 119 nurses represented 74 (87.1%) of the state's 85 LHDs. Respondents indicated they had been in their current positions for an average of 7 years (range, 2 months-42 years). Of the 119 nurses that completed the survey, 88 (73.9%) had interacted with one or more PHEs on a professional basis in the past year (August 2009-July 2010).

Four NCDPH key informants were interviewed (2 program managers and 2 epidemiologists). All 11 hospital supervisors were interviewed (response rate = 100%). Hospital supervisors held positions as the director, assistant director, or manager of infection control (8), infectious disease physician (1), hospital epidemiologist (1), and chief medical officer (1).

For syndromic surveillance systems to be effective in detecting outbreaks and potential bioterrorism events in their earliest stages, signals must be investigated in real-time [9]. Because PHEs are hospital employees focused on diseases of public health significance, they are uniquely positioned to quickly and accurately investigate signals from North Carolina's syndromic surveillance system for their hospitals, which account for 30% of emergency department visits in North Carolina. To confirm or rule out potential cases, PHEs have the ability to immediately access confidential patient medical records and can also consult with emergency department and infectious disease physicians, infection control practitioners, and lab technicians, as well as NCDPH staff.

PHEs further enhance syndromic surveillance efforts in the state by actively monitoring admissions, lab, and death reports for their hospital systems, which account for 39% of general/acute care beds in North Carolina. Although PHEs do not perform routine infection control activities, they are able, using hospital data systems, to detect and investigate unusual cases or clusters of communicable disease not only among emergency department patients, but among inpatients as well.

The findings of the study indicate that PHEs enhance communicable disease management in North Carolina by improving the completeness of communicable disease reporting. Underreporting of communicable diseases is well documented [10]. PHEs likely reduce underreporting (for the state's 11 largest hospital systems) by directly reporting cases to LHDs and NCDPH (or assisting physicians with reporting) and educating medical students, residents, and physicians on the importance of reporting. Active surveillance has been shown to enhance completeness of reporting for certain communicable diseases, including hepatitis and salmonellosis [11]. PHEs conduct surveillance on a daily basis by monitoring North Carolina's syndromic surveillance system and investigating signals and also by monitoring hospital admissions, lab, and death reports. In addition, PHEs conduct active case-finding based on alerts from LHDs and NCDPH, and during known outbreaks. Through these efforts, and by facilitating communication between the NCDPH meaningful use coordinator and appropriate hospital staff, it is likely that PHEs further reduce underreporting of communicable diseases.

Study results indicate that PHEs enhance the timeliness of public health investigations that involve cases seen at their hospital. As hospital employees serving NCDPH and LHDs as preferential partners, PHEs prioritize LHD and NCDPH requests for information needed from patient medical records. LHD respondents reported that PHEs are significantly more likely to respond "immediately" to their requests compared to staff at hospitals without PHEs. NCDPH key informants similarly highlighted PHEs ability to quickly access patient medical records and share information.

During the H1N1 pandemic, PHEs played an indispensible role in the surveillance "information loop." This loop involves healthcare facilities reporting cases to public health officials who in turn analyse case data in aggregate and develop recommendations for prevention and control. Public health officials then disseminate these recommendations back to healthcare facilities and others who are in a position to act on them, thereby completing the loop [12]. Because PHEs are based at the state's 11 largest, tertiary care/referral hospitals, where the most serious cases of H1N1 were likely to be seen, NCDPH was able to utilize these hospitals as informal sentinel surveillance sites during the pandemic. PHEs provided NCDPH with timely data on overall numbers of cases, as well as detailed information on the most severe cases. PHEs were able to quickly adapt the focus of their surveillance data to provide NCDPH with detailed case information on specific populations most severely affected by H1N1. NCDPH then used the information provided by PHEs, in conjunction with federal guidelines, to develop guidance (e.g., case definitions, testing guidelines, isolation and quarantine protocols) for healthcare providers. This guidance was disseminated back to hospitals through the PHEs, who distributed it to appropriate hospital staff and assisted in development of hospital protocols and policies related to H1N1.