Taipei's use of a multi-channel mass risk communication program to rapidly reverse an epidemic of highly communicable disease

Abstract

Background: In September 2007, an outbreak of acute hemorrhagic conjunctivitis (AHC) occurred in Keelung City and spread to Taipei City. In response to the epidemic, a new crisis management program was implemented and tested in Taipei.

Methodology and principal findings: Having noticed that transmission surged on weekends during the Keelung epidemic, Taipei City launched a multi-channel mass risk communications program that included short message service (SMS) messages sent directly to approximately 2.2 million Taipei residents on Friday, October 12th, 2007. The public was told to keep symptomatic students from schools and was provided guidelines for preventing the spread of the disease at home. Epidemiological characteristics of Taipei's outbreak were analyzed from 461 sampled AHC cases. Median time from exposure to onset of the disease was 1 day. This was significantly shorter for cases occurring in family clusters than in class clusters (mean+/-SD: 2.6+/-3.2 vs. 4.39+/-4.82 days, p = 0.03), as well as for cases occurring in larger family clusters as opposed to smaller ones (1.2+/-1.7 days vs. 3.9+/-4.0 days, p<0.01). Taipei's program had a significant impact on patient compliance. Home confinement of symptomatic children increased from 10% to 60% (p<0.05) and helped curb the spread of AHC. Taipei experienced a rapid decrease in AHC cases between the Friday of the SMS announcement and the following Monday, October 15, (0.70% vs. 0.36%). By October 26, AHC cases reduced to 0.01%. The success of this risk communication program in Taipei (as compared to Keelung) is further reflected through rapid improvements in three epidemic indicators: (1) significantly lower crude attack rates (1.95% vs. 14.92%, p<0.001), (2) a short epidemic period of AHC (13 vs. 34 days), and (3) a quick drop in risk level (1 approximately 2 weeks) in Taipei districts that border Keelung (the original domestic epicenter).

Conclusions and significance: The timely launch of this systematic, communication-based intervention proved effective at preventing a dangerous spike in AHC and was able to bring this high-risk disease under control. We recommend that public health officials incorporate similar methods into existing guidelines for preventing pandemic influenza and other emerging infectious diseases.

Figure 1. Spatial Distribution of the 2007 AHC Attack Rates in Taipei City and Keelung City.

Geographical location of Taipei City and Keelung City and their spatial relationship was shown in panel A. Spatial and temporal changes of AHC attack rates (AR) between Taipei City and Keelung City were plotted according to their place in the outbreak timeline (panel B). Day 0 is used to indicate the days that cluster cases were first identified in Taipei or in Keelung cities. Darker colored areas indicate regions with higher AR.

Figure 2. Incidence Rate (IR) and Cumulative Number of AHC Cases among School Children in Taipei City and Keelung City, October - November 2007.

The figure's arrow indicates the weekend in which children stayed at home instead of attending school. The MRCP was launched on October 12th in Taipei City, causing the incidence rate (IR) of AHC to decline more rapidly in Taipei City than in Keelung City, where AHC cases continued to increase.

Figure 3. Spatio-temporal Diffusion Patterns of AHC Outbreaks from Keelung City to Taipei City, 2007.

The X-axis reflects the temporal scale, while the Y-axis displays district names ordered by geographical correlation from North to South, East to West. The horizontal purple line marks the border between Keelung City and Taipei City.

Figure 4. Distribution of AHC Illness Duration as Related to Family Cluster Size.

Onset duration tended to be shorter when cluster size was more than 3 family members. (1.2±1.7 days, range 0 to 6 days, vs. smaller cluster size 3.9±4.0 days, range 0 to 16 days, p<0.01).