Persons using assistive technology might not be able to fully access information in this file. For assistance, please send e-mail to: mmwrq@cdc.gov. Type 508 Accommodation and the title of the report in the subject line of e-mail.
Abstracts
Monitoring System for Detecting Starts
and Declines of
Influenza
Epidemics
Eva Andersson
Goteborg University, Goteborg, Sweden
Corresponding author: Eva Andersson, Statistical Research Unit, Goteborg University, P.O. Box 660, SE 405 30, Goteborg,
Sweden. Telephone: 0046 (0)31 773 12 64; Fax: 0046 (0)31 773 12 74; E-mail:
eva.andersson@statistics.gu.se.
Abstract
Introduction: A nonparametric surveillance system was constructed for early detection of influenza outbreaks. The system uses weekly data on the number of influenza cases.
Objectives: For this analysis, a nonparametric method of surveillance was compared with the likelihood
ratio method, which is optimal because it yields a minimal expected delay for a fixed false-alert probability.
The evaluation was conducted by using probability of successful detection within a specified time and predictive value at different time points. The optimal surveillance method requires knowledge of the parametric model for the
given process (i.e., the influenza cycle). Influenza cycles differ in shape and amplitude from one season to the next. Therefore, finding a parametric model based on influenza data from previous seasons is difficult. Also, using data from previous seasons might lead to misspecification of the cycles.
Methods: In the nonparametric method, the influenza cycles were estimated under monotonicity restrictions
(i.e., monotonically increasing during the outbreak and monotonically decreasing during the outbreak's decline). The surveillance system was evaluated in a theoretical simulation study. The performance of the nonparametric method was compared with that of the optimal method. The effect of a misspecification of the parametric model was also studied.
Results: For most surveillance methods, the probability of successful detection of an influenza outbreak within 1 week depends on when the outbreak began relative to the start of the surveillance. The predictive value depends on when the alert is generated (Table).
Conclusions: The nonparametric method has lower detection probability then the optimal method when
the outbreak begins immediately after surveillance is started. However, the nonparametric method
avoids misspecifications. A parametric method with a misspecification results in poor detection probability for early outbreaks and low predictive value for late alerts.
Use of trade names and commercial sources is for identification only and does not imply endorsement by the U.S. Department of
Health and Human Services.References to non-CDC sites on the Internet are
provided as a service to MMWR readers and do not constitute or imply
endorsement of these organizations or their programs by CDC or the U.S.
Department of Health and Human Services. CDC is not responsible for the content
of pages found at these sites. URL addresses listed in MMWR were current as of
the date of publication.
Disclaimer
All MMWR HTML versions of articles are electronic conversions from ASCII text
into HTML. This conversion may have resulted in character translation or format errors in the HTML version.
Users should not rely on this HTML document, but are referred to the electronic PDF version and/or
the original MMWR paper copy for the official text, figures, and tables.
An original paper copy of this issue can be obtained from the Superintendent of Documents,
U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800.
Contact GPO for current prices.
**Questions or messages regarding errors in formatting should be addressed to
mmwrq@cdc.gov.
