BACKGROUND: Campylobacteriosis is one of the most common causes of diarrheal illness in the United States, and Hawaii historically has higher rates of disease than seen on the mainland. The Hawaii Department of Health (HDOH) receives physical and electronic laboratory reports of Campylobacter culture positive stool specimens (non-culture diagnostics are not routinely used in Hawaii). We evaluated the quality of our passive campylobacteriosis surveillance system and its usefulness to inform the state on the epidemiology of campylobacteriosis in Hawaii
METHODS: We reviewed data from all campylobacteriosis cases reported through Electronic Laboratory Reporting (ELR) and other reporting methods in Maven, Hawaii’s comprehensive disease case management system, from November 2012 to September 2015. Data were de-duplicated before starting our evaluation. We assessed records for data quality, timeliness, representativeness, and predictive value positive (PVP). PVP was calculated as the proportion of confirmed cases (isolation of Campylobacter from a clinical specimen) among all cases disclosed in the electronic and physical reports. Data quality was assessed by measuring data completeness and representativeness of reports for the population. Timeliness was defined as the difference between the specimen collection date and the creation date of the event in Maven.
RESULTS: We identified 2,321 reported campylobacteriosis cases in Maven during the period of interest, of which 220 were found to be duplicate records. The 2,101 distinct cases in Maven were pulled from all major clinical laboratories in the state. Data fields included demographic information (age, birth date, gender, race, address) and laboratory information (facility, serotype, collection date). No information on risk history or exposure factors was available. Data completion had a median of 93%, with a range from 2% (serotype) to 100% (age). The PVP was 97%, taking into consideration all cases reported into Maven. Timeliness ranged from 0–131 days (median 3 days). We also identified inconsistencies with case definition use, with the inclusion of non-standard case terms.
CONCLUSIONS: Maven provides timely, representative, and reasonably complete data for passive campylobacteriosis surveillance. Additionally, our evaluation uncovered inconsistencies in the case definition usage that will be addressed to improve data going forward. Although risk factor evaluation is not possible with the data currently available, the data in Maven provide a good starting point in the understanding of Campylobacter epidemiology in Hawaii and allow more in-depth investigations to be undertaken.