Search

The tree-based scan statistic (TreeScan; Martin Kulldorff, Harvard Medical School, Boston, Massachusetts) is a data-mining method that adjusts for multiple testing of correlated hypotheses when screening thousands of potential adverse events for signal identification. Simulation has demonstrated the promise of TreeScan with a propensity score (PS)-matched cohort design. However, it is unclear which variables to include in a PS for applied signal identification studies to simultaneously adjust for confounding across potential outcomes. We selected 4 pairs of medications with well-understood safety profiles. For each pair, we evaluated 5 candidate PSs with different combinations of 1) predefined general covariates (comorbidity, frailty, utilization), 2) empirically selected (data-driven) covariates, and 3) covariates tailored to the drug pair. For each pair, statistical alerting patterns were similar with alternative PSs (≤11 alerts in 7,996 outcomes scanned). Inclusion of covariates tailored to exposure did not appreciably affect screening results. Inclusion of empirically selected covariates can provide better proxy coverage for confounders but can also decrease statistical power. Unlike tailored covariates, empirical and predefined general covariates can be applied "out of the box" for signal identification. The choice of PS depends on the level of concern about residual confounding versus loss of power. Potential signals should be followed by pharmacoepidemiologic assessment where confounding control is tailored to the specific outcome(s) under investigation. [ABSTRACT FROM AUTHOR]

The tree-based scan statistic is a statistical data mining tool that has been used for signal detection with a self-controlled design in vaccine safety studies. This disproportionality statistic adjusts for multiple testing in evaluation of thousands of potential adverse events. However, many drug safety questions are not well suited for self-controlled analysis. We propose a method that combines tree-based scan statistics with propensity score-matched analysis of new initiator cohorts, a robust design for investigations of drug safety. We conducted plasmode simulations to evaluate performance. In multiple realistic scenarios, tree-based scan statistics in cohorts that were propensity score matched to adjust for confounding outperformed tree-based scan statistics in unmatched cohorts. In scenarios where confounding moved point estimates away from the null, adjusted analyses recovered the prespecified type 1 error while unadjusted analyses inflated type 1 error. In scenarios where confounding moved point estimates toward the null, adjusted analyses preserved power, whereas unadjusted analyses greatly reduced power. Although complete adjustment of true confounders had the best performance, matching on a moderately mis-specified propensity score substantially improved type 1 error and power compared with no adjustment. When there was true elevation in risk of an adverse event, there were often co-occurring signals for clinically related concepts. TreeScan with propensity score matching shows promise as a method for screening and prioritization of potential adverse events. It should be followed by clinical review and safety studies specifically designed to quantify the magnitude of effect, with confounding control targeted to the outcome of interest. [ABSTRACT FROM AUTHOR]

The self-controlled tree-temporal scan statistic--a new signal-detection method--can evaluate whether any of a wide variety of health outcomes are temporally associated with receipt of a specific vaccine, while adjusting for multiple testing. Neither health outcomes nor postvaccination potential periods of increased risk need be prespecified. Using US medical claims data in the Food and Drug Administration's Sentinel system, we employed the method to evaluate adverse events occurring after receipt of quadrivalent human papillomavirus vaccine (4vHPV). Incident outcomes recorded in emergency department or inpatient settings within 56 days after first doses of 4vHPV received by 9-through 26.9-year-olds in 2006-2014 were identified using International Classification of Diseases, Ninth Revision, diagnosis codes and analyzed by pairing the new method with a standard hierarchical classification of diagnoses. On scanning diagnoses of 1.9 million 4vHPV recipients, 2 statistically significant categories of adverse events were found: cellulitis on days 2-3 after vaccination and "other complications of surgical and medical procedures" on days 1-3 after vaccination. Cellulitis is a known adverse event. Clinically informed investigation of electronic claims records of the patients with "other complications" did not suggest any previously unknown vaccine safety problem. Considering that thousands of potential short-term adverse events and hundreds of potential risk intervals were evaluated, these findings add significantly to the growing safety record of 4vHPV. [ABSTRACT FROM AUTHOR]

Each day, the New York City Department of Health and Mental Hygiene uses the free SaTScan software to apply prospective space-time permutation scan statistics to strengthen early outbreak detection for 35 reportable diseases. This method prompted early detection of outbreaks of community-acquired legionellosis and shigellosis. [ABSTRACT FROM AUTHOR]

Methods for the assessment of spatial variations in temporal trends (SVTT) are important tools for disease surveillance, which can help governments to formulate programs to prevent diseases, and measure the progress, impact, and efficacy of preventive efforts already in operation. The linear SVTT method is designed to detect areas with unusual different disease linear trends. In some situations, however, its estimation trend procedure can lead to wrong conclusions. In this article, the quadratic SVTT method is proposed as alternative of the linear SVTT method. The quadratic method provides better estimates of the real trends, and increases the power of detection in situations where the linear SVTT method fails. A performance comparison between the linear and quadratic methods is provided to help illustrate their respective properties. The quadratic method is applied to detect unusual different cervical cancer trends in white women in the United States, over the period 1969 to 1995. [ABSTRACT FROM AUTHOR]

Objectives. We compared an electronic health record-based influenza-like illness (ILI) surveillance system with manual sentinel surveillance and virologic data to evaluate the utility of the automated system for routine ILI surveillance. Methods. We obtained weekly aggregate ILI reports from the Electronic medical record Support for Public Health (ESP) disease-detection and reporting system, which used an automated algorithm to identify ILI visits among a patient population of about 700,000 in Eastern Massachusetts. The percentage of total visits for ILI ("percent ILI") in ESP, percent ILI in the Massachusetts Department of Public Health's sentinel surveillance system, and percentage of laboratory specimens submitted to participating Massachusetts laboratories that tested positive for influenza were compared for the period October 2007-September 2011. We calculated Spearman's correlation coefficients and compared ESP and sentinel surveillance systems qualitatively, in terms of simplicity, flexibility, data quality, acceptability, timeliness, and usefulness. Results. ESP and sentinel surveillance percent ILI always peaked within one week of each other. There was 80% correlation between the two and 71%-73% correlation with laboratory data. Sentinel surveillance percent ILI was higher than ESP percent ILI during influenza seasons. The amplitude of variation in ESP percent ILI was greatest for 5- to 49-year-olds and typically peaked for the 5- to 24-year-old age group before the others. Conclusions. The ESP system produces percent ILI data of similar quality to sentinel surveillance and offers the advantages of shifting disease reporting burden from clinicians to information systems, allowing tracking of disease by age group, facilitating efficient surveillance for very large populations, and producing consistent and timely reports. [ABSTRACT FROM AUTHOR]

National infectious disease incidence rates are often estimated by standardizing locally derived rates using national-level age and race distributions. Data on other factors potentially associated with incidence are often not available in the form of patient-level covariates. Including characteristics of patients’ area of residence may improve the accuracy of national estimates. The authors used data from the Centers for Disease Control and Prevention's Active Bacterial Core Surveillance program (2004–2005), adjusted for census-based variables, to estimate the national incidence of invasive pneumococcal disease (IPD). The authors tested Poisson and negative binomial models in a cross-validation procedure to select variables best predicting the incidence of IPD in each county. Including census-level information on race and educational attainment improved the fit of both Poisson and negative binomial models beyond that achieved by adjusting for other census variables or by adjusting for an individual's race and age alone. The Poisson model with census-based predictors led to a national estimate of IPD of 16.0 cases per 100,000 persons as compared with 13.5 per 100,000 persons using an individual's age and race alone. Accuracy of, and confidence intervals for, these estimates can only be determined by obtaining data from other randomly selected US counties. However, incorporating census-derived characteristics should be considered when estimating national incidence of IPD and other diseases. [ABSTRACT FROM PUBLISHER]

OBJECTIVE: To describe the Vaccine Safety Datalink (VSD) project's experience with population-based, active surveillance for vaccine safety and draw lessons that may be useful for similar efforts. PATIENTS AND METHODS: The VSD comprises a population of 9.2 million people annually in 8 geographically diverse US health care organizations. Data on vaccinations and diagnoses are updated and extracted weekly. The safety of 5 vaccines was monitored, each with 5 to 7 prespecified outcomes. With sequential analytic methods, the number of cases of each outcome was compared with the number of cases observed in a comparison group or the number expected on the basis of background rates. If the test statistic exceeded a threshold, it was a signal of a possible vaccine-safety problem. Signals were investigated by using temporal scan statistics and analyses such as logistic regression. RESULTS: Ten signals appeared over 3 years of surveillance: 1 signal was reported to external stakeholders and ultimately led to a change in national vaccination policy, and 9 signals were found to be spurious after rigorous internal investigation. Causes of spurious signals included imprecision in estimated background rates, changes in true incidence or coding overtime, other confounding, inappropriate comparison groups, miscoding of outcomes in electronic medical records, and chance. In the absence of signals, estimates of adverse-event rates, relative risks, and attributable risks from up-to-date VSD data have provided rapid assessment of vaccine safety to policy-makers when concerns about a specific vaccine have arisen elsewhere. CONCLUSIONS: Care with data quality, outcome definitions, comparison groups, and length of surveillance are required to enable detection of true safety problems while minimizing false signals. Some causes of false signals in the VSD system were preventable and have been corrected, whereas others will be unavoidable in any active surveillance system. Temporal scan statistics, analyses to control for confounding, and chart review are indispensable tools in signal investigation. The VSD's experience may inform new systems for active safety surveillance. [ABSTRACT FROM AUTHOR]

Spatial scan statistics with Bernoulli and Poisson models are commonly used for geographical disease surveillance and cluster detection. These models, suitable for count data, were not designed for data with continuous outcomes. We propose a spatial scan statistic based on an exponential model to handle either uncensored or censored continuous survival data. The power and sensitivity of the developed model are investigated through intensive simulations. The method performs well for different survival distribution functions including the exponential, gamma, and log-normal distributions. We also present a method to adjust the analysis for covariates. The cluster detection method is illustrated using survival data for men diagnosed with prostate cancer in Connecticut from 1984 to 1995. [ABSTRACT FROM AUTHOR]

High breast cancer mortality rates have been reported in the northeastern part of the United States, with recent attention focused on Long Island, New York. In this study, the authors investigate whether the high breast cancer mortality is evenly spread over the Northeast, in the sense that any observed clusters of deaths can be explained by chance alone, or whether there are clusters of statistical significance. Demographic data and age-specific breast cancer mortality rates for women were obtained for all 244 counties in 11 northeastern states and for the District of Columbia for 1988–1992. A recently developed spatial scan statistic is used, which searches for clusters of cases without specifying their size or location ahead of time, and which tests for their statistical significance while adjusting for the multiple testing inherent in such a procedure. The basic analysis is adjusted for age, with further analyses examining how the results are affected by incorporating race, urbanicity, and parity as confounding variables. There is a statistically significant and geographically broad cluster of breast cancer deaths in the New York City-Philadelphia, Pennsylvania, metropolitan area (p = 0.0001), which has a 7.4% higher mortality rate than the rest of the Northeast. The cluster remains significant when race, urbanicity, and/or parity are included as confounding variables. Four smaller subclusters within this area are also significant on their own strength: Philadelphia with suburbs (p = 0.0001), Long Island (p = 0.0001), central New Jersey (p = 0.0001), and northeastern New Jersey (p = 0.0001). The elevated breast cancer mortality on Long Island might be viewed less as a unique local phenomenon and more as part of a more general situation involving large parts of the New York City-Philadelphia metropolitan area. The several known and hypothesized risk factors for which we could not adjust and that may explain the detected cluster are most notably age at first birth, age at menarche, age at menopause, breastfeeding, genetic mutations, and environmental factors. Am J Epidemiol 1997;146:161-70. [ABSTRACT FROM AUTHOR]

Background: Utilizing highly precise spatial resolutions within disease outbreak detection, such as the patients’ address, is most desirable as this provides the actual residential location of the infected individual(s). However, this level of precision is not always readily available or only available for purchase, and when utilized, increases the risk of exposing protected health information. Aggregating data to less precise scales (e.g., ZIP code or county centroids) may mitigate this risk but at the expense of potentially masking smaller isolated high risk areas. Methods: To experimentally examine the effect of spatial data resolution on space-time cluster detection, we extracted administrative medical claims data for 122500 viral lung episodes occurring during 2007–2010 in Tennessee. We generated 10000 spatial datasets with varying cluster location, size and intensity at the address-level. To represent spatial data aggregation (i.e., reduced resolution), we then created 10000 corresponding datasets both at the ZIP code and county level for a total of 30000 datasets. Using the space-time permutation scan statistic and the SaTScan™ cluster software, we evaluated statistical power, sensitivity and positive predictive values of outbreak detection when using exact address locations compared to ZIP code and county level aggregations. Results: The power to detect disease outbreaks did not largely diminish when using spatially aggregated data compared to more precise address information. However, aggregations negatively impacted the ability to more accurately determine the exact spatial location of the outbreak, particularly in smaller clusters (<800 km2). Conclusions: Spatial aggregations do not necessitate a loss of power or sensitivity; rather, the relationship is more complex and involves simultaneously considering relative risk within the cluster and cluster size. The likelihood of spatially over-estimating outbreaks by including geographical areas outside the actual disease cluster increases with aggregated data.

Background: The ability to detect disease outbreaks early is important in order to minimize morbidity and mortality through timely implementation of disease prevention and control measures. Many national, state, and local health departments are launching disease surveillance systems with daily analyses of hospital emergency department visits, ambulance dispatch calls, or pharmacy sales for which population-at-risk information is unavailable or irrelevant.Methods and Findings: We propose a prospective space-time permutation scan statistic for the early detection of disease outbreaks that uses only case numbers, with no need for population-at-risk data. It makes minimal assumptions about the time, geographical location, or size of the outbreak, and it adjusts for natural purely spatial and purely temporal variation. The new method was evaluated using daily analyses of hospital emergency department visits in New York City. Four of the five strongest signals were likely local precursors to citywide outbreaks due to rotavirus, norovirus, and influenza. The number of false signals was at most modest.Conclusion: If such results hold up over longer study times and in other locations, the space-time permutation scan statistic will be an important tool for local and national health departments that are setting up early disease detection surveillance systems. [ABSTRACT FROM AUTHOR]