Current Trends Proposed Changes in Format for Presentation of Notifiable Disease Report Data

This article introduces a proposed graphic format for displaying national notifiable disease data in the MMWR. The proposed format is designed to facilitate interpretation of these data and enable timely public health responses to changes in disease patterns.

National notifiable disease reporting is a basic component of public health surveillance in the United States (1). Disease data are reported weekly to CDC by state health departments and are published as Tables I through III in the MMWR. To enhance interpretation of these data, a bar graph (Figure 1) is proposed to replace Table I. This new format compares provisional reports over time and indicates whether the number of reported cases of a disease for a specific reporting period differs from that of a previous period. In addition, line graphs would appear quarterly for four diseases (acquired immunodeficiency syndrome (AIDS), gonorrhea, syphilis, and tuberculosis) that may have secular trends but do not generally have substantial month-to-month changes in the reported number of cases (Figures 2-5). Proposed specific changes are described below. Figure 1

The current Table I ("Summary--cases of specified notifiable diseases, United States") would be replaced by a bar graph (Figure 1) that compares, for each disease, the number of cases reported in a 4-week period with the mean of 15 4-week totals (from comparable, previous, and subsequent 4-week periods for the past 5 years). For example, Figure 1 compares the number of reports for the 4 weeks ending November 25, 1989 (MMWR weeks 44-47), with the 5-year average for weeks 40-43, 44-47, and 48-51 of 1984-1988. For each disease, a horizontal bar indicates the ratio of the current value to the 5-year average. Bars to the right and left of the vertical axis at "1" indicate increases and decreases, respectively, in the number of reported cases.

Striping in the bars in Figure 1 indicates whether the number of reported cases during the most recent 4-week period are higher or lower than historical limits. The limits show typical variability in the ratios and are computed as 1 plus or minus 2(SD/X), where SD=standard deviation and X=mean of the 15 4-week totals. When the current ratio is outside the limits, the elevated (or diminished) portion of the ratio is striped. If no striping is present, the current ratio is within historical limits.

A change in disease occurrence identified by this approach should be regarded as an indication for more detailed examination of the data and monitoring of succeeding reports. For example, a recent increase in measles incidence in February and March 1989 would have been readily apparent if presented in the proposed graph format. However, the graph alone should not be the basis for conclusions. Figures 2-5

For diseases in which long-term variations in numbers of reported cases are more important than month-to-month variations (AIDS, gonorrhea, syphilis, and tuberculosis), line graphs (Figures 2-5) would appear quarterly. These graphs would reflect the provisional number of cases by 4-week periods since 1982. Other Changes

• Selected diseases that appear in Table I, such as leprosy and toxic shock syndrome, would be listed in an expanded version of current Table II ("Notifiable diseases of low frequency, United States"); this table would be renumbered Table I and re named "Summary--cases of specified notifiable diseases, United States" (Table 1).

• The monthly number of reported AIDS cases would be provided quarterly (Figure 2) rather than in the weekly MMWR tables.* AIDS reports are received at CDC monthly rather than weekly as is the case for the other notifiable diseases; thus, a plot of the 4-week (28 days) totals shown for AIDS (Figure 2) may differ from a plot of the monthly (28-31 days) surveillance data.

• Tables III and IV would be renumbered II and III, respectively, but otherwise would remain unchanged, except for the deletion of AIDS reports. The annual MMWR Summary of Notifiable Diseases would also remain unchanged and would continue to provide yearly state-specific disease report data in tables and graphs. Reported by: Statistics and Surveillance Br, Div of Surveillance and Epidemiologic Studies, Epidemiology Program Office, CDC.

  Editorial Note

Editorial Note: Several caveats may influence the interpretation of notifiable disease surveillance data presented in the MMWR tables and figures. For example, the data are provisional and subject to change because of late reports or corrections in case classification. Additionally, variations in reporting may result from differences in transmission of public health surveillance information (e.g., batch reporting of cases at the end of a month vs. weekly reports) or from changes or differences in case definitions. Also, surveillance data are generated by a process that may result in incomplete reports or underreporting (1); nevertheless, these data are useful indicators of trends in disease incidence.

The method illustrated by Figure 1 will not detect all epidemics for at least three reasons. First, differences in the number of case reports from the 5-year baseline value do not incorporate statistical theory, i.e., the limits are not confidence or prediction intervals and should not be interpreted as such. Rather, the limits represent an analytic framework for identifying aberrations in the number of reports during a specific time period. Second, use of the 5-year average as the baseline for comparison potentially could affect interpretation, particularly if knowledge about a disease is rapidly evolving or if large variations occurred during the baseline period. Third, regular seasonal fluctuations in disease occurrence will not be detected by this approach since a 4-week period is compared with the same season in previous years.

CDC is examining diverse statistical techniques for detecting aberrations in public health surveillance data (2). Techniques that might be useful are various parametric approaches (including the scan statistic (3,4) and a normal theory confidence interval calculated similarly to the historical limits as described above) and a nonparametric bootstrap approach (5). Other methods under consideration are the ratio of two Poisson random variables for low-frequency diseases, a Box-Jenkins time series approach incorporating the cusum statistic, and Bayesian and nonlinear time series methods. In addition, CDC is evaluating potential mapping and graphic changes in current Tables III and IV to improve interpretation of these data.

Comments and suggestions on the proposed new format or on statistical techniques for detecting aberrations in public health surveillance data are welcome and should be provided by January 12, 1989, to G. David Williamson, Ph.D., Statistics and Surveillance Branch, Division of Surveillance and Epidemiologic Studies, Epidemiology Program Office, CDC, Mailstop C08, Atlanta, GA 30333.

References

1. Thacker SB, Berkelman RL. Public health surveillance in the United States. Epidemiol Rev 1988;10:164-90.

2. Stroup DF, Williamson GD, Herndon JL, Karon JM. Detection of aberrations in the occurrence of notifiable diseases surveillance data. Stat Med 1989;8:323-9.

3. Naus JI. The distribution of the size of the maximum cluster of points on a line. J Am Stat Assoc 1965;60:532-8.

4. Wallenstein S. A test for detection of clustering over time. Am J Epidemiol 1980;111:367-72.

5. Efron B. The 1977 Rietz Lecture: bootstrap methods--another look at the jacknife. Ann Stat 1979;7:1-26. *AIDS data are published each month in the HIV/AIDS Surveillance Report; single copies are available free from the National AIDS Information Clearinghouse, P.O. Box 6003, Rockville, MD 20850.

Disclaimer   All MMWR HTML documents published before January 1993 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 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.

Page converted: 08/05/98

HOME  |  ABOUT MMWR  |  MMWR SEARCH  |  DOWNLOADS  |  RSS |  CONTACT POLICY  |  DISCLAIMER  |  ACCESSIBILITY

Morbidity and Mortality Weekly Report Centers for Disease Control and Prevention 1600 Clifton Rd, MailStop E-90, Atlanta, GA 30333, U.S.A Department of Health and Human Services

This page last reviewed 5/2/01