Aberration Detection Updates 16th Semiannual Meeting of the TBESC Aberration

TB Notes Newsletter
No. 1, 2010
SURVEILLANCE, EPIDEMIOLOGY, AND OUTBREAK INVESTIGATIONS BRANCH UPDATE
On this Page
Aberration Detection Updates 16th Semiannual Meeting of the TBESC Aberration Detection Updates

What is aberration detection?
Aberration detection is the use of statistical analysis to identify deviations from expected patterns. When applied to disease, aberration detection methods can be integrated into different types of surveillance systems, from traditional case-based surveillance (including some forms of laboratory-based surveillance) to surveillance based on clusters of signs and symptoms (i.e., syndromic surveillance). The objective of aberration detection systems is to identify unusual events (e.g., clusters of illness) and direct public health resources toward investigating and intervening to prevent additional illnesses.

The nature of TB transmission, in combination with the advent of the U.S. National Tuberculosis Genotyping Service (NTGS), lends itself well to the use of the spatio-temporal approach to the design and development of an aberration detection system. An aberration is defined as an excessive geospatial concentration of a given genotype compared to the national average. Various statistical methods such as CUSUM and log likelihood ratio (LLR) can be applied to geographic regions. At present, CDC is primarily using the LLR method applied to individual counties to detect aberrant genotype clusters of TB in those counties. The higher the LLR, the greater the chance that the local genotype cluster within the county represents an unexpected geospatial concentration; unexpected geospatial concentrations might indicate recent transmission of TB. Examining genotype clusters prioritized by high LLRs might help TB programs focus attention for public health response activities.

One limitation of the county-based method is that cases in neighboring areas will not be included in the initially identified genotype cluster. To address this limitation, CDC is evaluating the use of SaTScan (Spatial and Temporal Scanning) software, which uses spatial scanning techniques to identify significant genotype clusters within circles of a given radius and then calculates the LLR for those clusters.

Validation of Aberration Detection Methods
CDC has conferred with TB control experts to compare genotype cluster prioritization based on LLR results to priorities established by expert opinion. We found good correlation of cluster priority rankings between LLR and expert opinion. In addition, Tuberculosis Epidemiologic Studies Consortium (TBESC) Task Order 26, “Improving the Utilization and Integration of TB Genotyping into Routine TB Program Practice: Analyzing the Impact through Public Health Interventions,” has been initiated to further validate the use of aberration detection methods to identify TB clusters of public health importance. Four study sites (Georgia, Maryland, Massachusetts, and the city of Houston, Texas) will systematically assess and investigate TB genotype clusters in their jurisdictions. In addition, a panel of TB control experts will evaluate and prioritize (e.g., high, medium, low) TB clusters selected for investigation. The expert panel will be blinded to the LLR calculated for each cluster, and the study investigators will compare the experts’ conclusions and the LLRs determined at study onset to the results of the cluster investigations at the conclusion of the study. Task Order 26 has just gotten underway and is expected to contribute greatly to the validation and improvement of existing aberration detection models. Tools developed through these studies will also be used for cluster assessment and investigation.

Enhanced access to genotyping and aberration detection data
The TB Genotyping Information Management System (TB GIMS) was rolled out to state and local TB programs on March 8, 2010. TB GIMS allows public health partners to access genotype results of TB cases and view genotype clusters (including demographic and risk factor information) in their jurisdictions. These reports include county prioritization lists with descending LLR results to assist TB control programs in identifying clusters for public health attention. Easy access to these data allows state and local programs to incorporate genotyping and aberration detection methods into their routine TB control activities and direct resources toward the greatest problems. Fact sheets and training materials related to understanding and application are under development. For questions about TB GIMS, contact your CDC program consultant or tbgims@cdc.gov.

New staff to develop aberration detection system methods and assist state and local programs
Juliana Grant, MD, MPH, and Adam Langer, DVM, MPH, recently joined the staff of the Surveillance, Epidemiology, and Outbreak Investigations Branch to focus on developing methods and procedures for responding to genotype clusters of TB identified through the NTGS and CDC’s aberration detection activities. Drs. Grant and Langer will work with state and local public health partners to assess genotype clusters and develop methods for ascertaining if recent transmission has occurred, and what steps might be taken to reduce ongoing transmission. Several states have already invited Outbreak Investigations Team staff to collaborate in assessing potentially aberrant genotype clusters. State and local TB programs that would like consultation on epidemiologic interpretation of genotyping data, including assessment and investigation of potentially aberrant genotype clusters, may contact their CDC program consultants for assistance.

—Reported by Adam Langer, DVM, MPH, Juliana Grant, MD, MPH,
Smita Chatterjee, MS, Roque Miramontes, PA-C, MPH,
and Paul Tribble, MA, Division of Tuberculosis Elimination

16th Semiannual Meeting of the Tuberculosis Epidemiologic Studies Consortium (TBESC)
Approximately 100 principal investigators, project coordinators, and other Tuberculosis Epidemiologic Studies Consortium (TBESC) personnel attended the meeting January 20–21, 2010, at the Crowne Plaza Ravinia Hotel in Atlanta, GA. The purpose of the meeting was to discuss recent tuberculosis (TB) research and to plan next steps needed for current research via breakout sessions. Attendees were welcomed by Phil Talboy from DTBE and Jane Tapia of Emory University.

One highlight of the first day of the meeting was a scientific presentation on “Assessing QuantiFERON-TB Gold In-Tube (QFT-G) as an initial screening tool for U.S.-bound applicants for immigration and feasibility of follow-up in U.S. immigrants,” given by Randall Reves, MD. The study was performed to evaluate the sensitivity and specificity of QFT-G and TST in screening for TB in the immigrant population.

Dr. Reves, the study’s principal investigator, found that QFT-G had approximately 88% sensitivity in screening for culture-positive TB. QFT-G sensitivity was similar in the study immigrant population to the sensitivity of a tuberculin skin test (TST) reading of 5 mm. Concordance between the two tests was approximately 80% in the two populations, Dr. Reves said.

Enrolled participants for the study were immigrants in Vietnam applying for residency in the United States. Two thirds of enrolled immigrants had an abnormal chest x-ray (CXR). Participants were tested with TST, QFT-G, and TB culture.

Study investigators also evaluated the sensitivity of QFT-G and TST in predicting culture-positive TB among persons with an abnormal CXR. Results indicated sensitivities similar to those found when screening for all persons with culture-positive TB, Dr. Reves said (88% QFT-G sensitivity; 86% TST sensitivity using a cut-point of 5 mm).

Denise Garrett, MD, project officer for TBESC, also gave an update on the TBESC recompetition. Her presentation included a timeline for the recompetition, goals for the next TBESC, and an overview of future steps needed to establish the new TBESC.

Highlights of the second day of the meeting included an overview of economic analysis of TB in the presentation, “Estimating the Cost of TB in the United States,” by Dr. Thad Miller. Dr. Miller discussed the means by which one’s perspective of cost affects the total cost estimate, and reported that the estimated cost to society for one TB case was over $300,000.

Bob Belknap, MD, also presented data from Task Order 18, “Evaluation of new interferon-gamma release assays in the diagnosis of LTBI in health care workers.” Preliminary results indicated a 0.4% TST conversion in health care workers from 0 to 6 months, but 3% test conversion of both QuantiFERON-TB Gold In-Tube and T-Spot.

The 17th Semiannual TBESC meeting will be held in San Francisco, CA, July 28–29, 2010.

—Reported by Suzanne Beavers, MD
Div of TB Elimination
http://www.cdc.gov/tb/publications/newsletters/notes/TBN_1_10/seoib.htm

(PDF - 928K - please see below - 30 pages)
http://www.cdc.gov/tb/publications/newsletters/notes/TBN_1_10/images/tbn110.pdf