Evaluation of PFGE Cluster Investigations
E. coli O157:H7
We recently published a paper in Epidemiology and Infection in which we evaluated E. coli O157:H7 PFGE cluster data from 2000 through 2008: Dynamics of Escherichia coli O157:H7 outbreak detection and investigation, Minnesota 2000-2008 (see abstract below). In this analysis, clusters of >3 cases and clusters in which the first 2 case isolates were received at a public health laboratory within 7 days were associated with a cluster being solved. Utilizing the same methodology, 4 additional years of data (2009-2012) were added to this study, and the cumulative results are given below.
Seventy E. coli O157:H7 outbreaks were identified in Minnesota during 2000-2012 (Table 3). Of these, about half were foodborne outbreaks and half were due to other routes of transmission. Pulsed-field gel electrophoresis subtyping of isolates submitted as part of routine laboratory-based surveillance was the most common method of outbreak detection, accounting for 45 (64%) of the 70 outbreaks.
The median number of E. coli O157:H7 isolates subtyped per year was 123.5 (range, 99-136) (Figure 4). The median number of clusters per year was 14 (range, 11-16). The median number of confirmed E. coli O157:H7 outbreaks per year was 4.5 (range, 4-7). There were no significant temporal trends in the proportion of clusters that resulted in the identification of a confirmed outbreak (p=0.98).
There was seasonal variation by number of isolates, clusters, and solved clusters (confirmed outbreaks) (Figure 5). Sixty-nine percent of all isolates, 71% of all clusters, and 60% of all solved clusters occurred during June-September (peak season). Clusters occurring during October-May (low season) were more than twice as likely to be solved as clusters occurring during peak season, but this difference was not statistically significant.
Analyses of the cumulative PFGE data (2000-2012) provided results consistent with the previous analyses; no new predictors of a cluster being solved were identified. Clusters of the uncommon E. coli O157:H7 PFGE subtypes were not significantly more likely to be solved than common E. coli O157:H7 PFGE subtypes (Table 4). In general the probability of a cluster being solved increased significantly as the number of cluster cases increased. Clusters of 3 and ≥5 cases were significantly more likely to be solved than clusters of 2 cases. Clusters of 4 cases were twice as likely to be solved compared to clusters of 2 cases; this difference was not statistically significant, but this could have been a function of the small number of clusters with 4 cases.
Clusters in which the first two cluster case isolates were received at the public health laboratory on the same day were significantly more likely to be solved than clusters in which the first two cluster case isolates were received during a period of 8 – 14 days.
Conclusion
The cumulative data indicate that success rates were similar for investigations of clusters of common vs. uncommon PFGE subtypes. Clusters of >3 case isolates were significantly more likely to be solved than clusters of 2 case isolates. Greater success in solving clusters also was observed with clusters in which the first 2 case isolates were received at the public health laboratory on the same day.
Recommendation
We strongly recommend interviewing all E. coli O157:H7 cases and investigating all PFGE subtype. If resources do not permit investigation of all PFGE clusters, our data suggest that investigating all clusters with >3 cases should be done. One caveat, however, is that these data are based on the MDH system, which includes interviewing of every case with a detailed questionnaire as soon as it is reported. It is not known how these data would (or would not) extrapolate to other systems. Another caveat is that by not interviewing all cases, or by limiting investigation to clusters with >3 cases, some outbreaks will not be detected.
Contact us if you would like to discuss E. coli O157:H7 cluster investigations.
Abstract
We determined characteristics of Escherichia coli O157:H7 pulsed-field gel electrophoresis clusters that predict their being solved (i.e. that result in identification of a confirmed outbreak). Clusters were investigated by the Minnesota Department of Health (MDH) using a dynamic iterative model. During 2000–2008, 19 (23%) of 84 clusters were solved. Clusters of ≥3 isolates were more likely to be solved than clusters of two isolates. Clusters in which the first two case isolates were received at MDH on the same day were more likely to be solved than were clusters in which the first two case isolates were received over ≥8 days. Investigation of clusters of ≥3 E. coli O157:H7 cases increased the success of cluster investigations.