A process evaluation of an active surveillance system for hospitalized 2009-2010 H1N1 influenza cases.

OBJECTIVES: A process evaluation was conducted to evaluate a newly established active influenza surveillance program that utilized 6 sentinel hospitals to collect epidemiologic information for influenza-like illness admissions. Objectives were to determine whether the new system was implemented successfully and met surveillance objectives, including determination of the proportion of patients with 2009 H1N1 influenza, extent of disease severity, and identifying high-risk groups. METHODS: Timeliness and data quality were assessed through analysis of electronic case report form completion and timing of specimen collection and submission to public health laboratories for influenza testing. Simplicity and accessibility of the surveillance system were assessed through a survey of hospital-based surveillance staff. RESULTS: The median number of days from admission to initial reporting was 5 days. The completeness of core variables was more than 98%, 96.2% for complications, and 92.6% for underlying medical conditions. Among influenza-like illness admissions, 77.8% had a specimen submitted for confirmatory testing. Eighty-nine percent of survey respondents found guidance provided by New York State Department of Health to be helpful and case report forms easy to use. CONCLUSIONS: This project was implemented within a context of limited time and resources. Certain aspects of planning, such as securing necessary staffing at some hospitals, could not be carried out prior to implementation. Resource limitations necessitated controls on the numbers of specimens submitted each week. Some reporting lags were noted because of delays in data entry. Reporting timeframes allowed for timely data summarization for internal decision making. Maintaining frequent contact with sentinel sites promoted report completeness, timeliness, and consistency across sites. These results highlight the value of sentinel surveillance methodology and challenges of rapidly deploying a new active surveillance system for an emergent disease.

Two distinct surveillance methods to track hospitalized influenza patients in New York State during the 2009-2010 influenza season.

OBJECTIVE: To better understand the severity of 2009 H1N1 influenza disease, enhanced surveillance of patients hospitalized with influenza was conducted during the 2009-2010 influenza season in New York State through existing Emerging Infections Program surveillance and a newly established sentinel hospital surveillance program. The 2 surveillance systems were compared to determine consistency across surveillance modalities and reveal the strengths and weaknesses of each to accomplish comprehensive influenza surveillance. DESIGN: Similar variables from the aggregate data collected from each system were compared and differences were analyzed in detail. SETTING: New York State. PARTICIPANTS: Hospitalized adult and pediatric patients detected through 2 influenza surveillance programs. MAIN OUTCOME MEASURES: Significant differences in age distribution, timing of illness onset, illness complications, underlying medical conditions, critical care admissions, use of mechanical ventilation, and illness outcomes. RESULTS: Both surveillance systems saw the highest numbers of confirmed influenza infection among patients hospitalized in early fall 2009, with sharp declines thereafter. Sentinel hospital surveillance continued to detect hospitalizations for influenza-like illness that were not due to 2009 H1N1 influenza well into March 2010. Compared to influenza surveillance conducted through the Emerging Infections Program, the sentinel hospital influenza surveillance program tended to detect a sicker population of children and adults, including a higher rate of critical illness and mechanical ventilation, and among adults, higher rates of some underlying medical conditions. There were no differences in disease outcomes detected between the 2 systems. CONCLUSIONS: Although the 2 surveillance systems were complementary, inherent methodologic variations revealed important differences at season conclusion. The lessons learned should be used to determine the best way to allocate resources to meet the needs of future state and national influenza surveillance efforts.