Aman Verma

The BCG World Atlas: A Database of Global BCG Vaccination Policies and Practices

Madhu Pai and colleagues introduce the BCG World Atlas, an open access, user friendly Web site for TB clinicians to discern global BCG vaccination policies and practices and improve the care of their patients.

Risk of Injury Associated with Opioid Use in Older Adults

OBJECTIVES: To estimate the dose‐related risk of injuries in older adults associated with the use of low‐, medium‐, and high‐potency opioids.

Association of Off-label Drug Use and Adverse Drug Events in an Adult Population

IMPORTANCE Off-label use of prescription drugs has been identified as an important contributor to preventable adverse drug events (ADEs) in children. Despite concerns regarding adverse outcomes, to date, no systematic investigation of the effects of off-label drug use in adult populations has been performed. OBJECTIVE To monitor and evaluate off-label use of prescription drugs and its effect on ADEs in an adult population. DESIGN, SETTING, AND PARTICIPANTS A cohort of 46,021 patients who received 151,305 incident prescribed drugs was assembled from primary care clinics in Quebec, Canada, using the Medical Office of the XXIst Century electronic health record, which supports documentation of treatment indications and treatment outcomes. Prescriptions dispensed from January 1, 2005, through December 30, 2009, were followed up from the date of the prescription to the date the drug use was discontinued, the end of treatment, or the end of follow-up (December 30, 2010). Data were analyzed from January 5, 2012, to March 15, 2015. EXPOSURES Off-label prescription drug use with and without strong scientific evidence. MAIN OUTCOMES AND MEASURES Adverse drug events in off-label use with and without strong scientific evidence. Analysis used multivariate marginal Cox proportional hazards regression for clustered data with the drug as the unit of analysis. RESULTS A total of 3484 ADEs were found in the 46,021 study patients, with an incidence rate of 13.2 per 10,000 person-months. The rate of ADEs for off-label use (19.7 per 10,000 person-months) was higher than that for on-label use (12.5 per 10,000 person-months) (adjusted hazard ratio [AHR], 1.44; 95% CI, 1.30-1.60). Off-label use lacking strong scientific evidence had a higher ADE rate (21.7 per 10,000 person-months) compared with on-label use (AHR, 1.54; 95% CI, 1.37-1.72). However, off-label use with strong scientific evidence had the same risk for ADEs as on-label use (AHR, 1.10; 95% CI, 0.88-1.38). The risks for ADEs were higher for drugs approved from 1981 to 1995 (14.4 per 10,000 person-months; AHR, 1.62; 95% CI, 1.45-1.80) and for those used by women (14.3 per 10,000 person-months; AHR, 1.17; 95% CI, 1.06-1.28), patients receiving 5 to 7 drugs (12.1 per 10,000 person-months; AHR, 3.23; 95% CI, 2.66-3.92), and patients receiving cardiovascular drugs (15.9 per 10,000 person-months; AHR, 3.30; 95% CI, 2.67-4.08) and anti-infectives (66.2 per 10,000 person-months; AHR, 6.33; 95% CI, 4.58-8.76). Patients with a 1-unit increase in the continuity of care index had a 19% increase in ADEs (AHR, 1.19; 95% CI, 1.12-1.26). CONCLUSIONS AND RELEVANCE Off-label use of prescription drugs is associated with ADEs. Caution should be exercised in prescribing drugs for off-label uses that lack strong scientific evidence. Future electronic health records should be designed to enable postmarket surveillance of treatment indications and treatment outcomes to monitor the safety of on- and off-label uses of drugs.

A novel method of adverse event detection can accurately identify venous thromboembolisms (VTEs) from narrative electronic health record data

Background Venous thromboembolisms (VTEs), which include deep vein thrombosis (DVT) and pulmonary embolism (PE), are associated with significant mortality, morbidity, and cost in hospitalized patients. To evaluate the success of preventive measures, accurate and efficient methods for monitoring VTE rates are needed. Therefore, we sought to determine the accuracy of statistical natural language processing (NLP) for identifying DVT and PE from electronic health record data. Methods We randomly sampled 2000 narrative radiology reports from patients with a suspected DVT/PE in Montreal (Canada) between 2008 and 2012. We manually identified DVT/PE within each report, which served as our reference standard. Using a bag-of-words approach, we trained 10 alternative support vector machine (SVM) models predicting DVT, and 10 predicting PE. SVM training and testing was performed with nested 10-fold cross-validation, and the average accuracy of each model was measured and compared. Results On manual review, 324 (16.2%) reports were DVT-positive and 154 (7.7%) were PE-positive. The best DVT model achieved an average sensitivity of 0.80 (95% CI 0.76 to 0.85), specificity of 0.98 (98% CI 0.97 to 0.99), positive predictive value (PPV) of 0.89 (95% CI 0.85 to 0.93), and an area under the curve (AUC) of 0.98 (95% CI 0.97 to 0.99). The best PE model achieved sensitivity of 0.79 (95% CI 0.73 to 0.85), specificity of 0.99 (95% CI 0.98 to 0.99), PPV of 0.84 (95% CI 0.75 to 0.92), and AUC of 0.99 (95% CI 0.98 to 1.00). Conclusions Statistical NLP can accurately identify VTE from narrative radiology reports.

The re-identification risk of Canadians from longitudinal demographics

BackgroundThe public is less willing to allow their personal health information to be disclosed for research purposes if they do not trust researchers and how researchers manage their data. However, the public is more comfortable with their data being used for research if the risk of re-identification is low. There are few studies on the risk of re-identification of Canadians from their basic demographics, and no studies on their risk from their longitudinal data. Our objective was to estimate the risk of re-identification from the basic cross-sectional and longitudinal demographics of Canadians.MethodsUniqueness is a common measure of re-identification risk. Demographic data on a 25% random sample of the population of Montreal were analyzed to estimate population uniqueness on postal code, date of birth, and gender as well as their generalizations, for periods ranging from 1 year to 11 years.ResultsAlmost 98% of the population was unique on full postal code, date of birth and gender: these three variables are effectively a unique identifier for Montrealers. Uniqueness increased for longitudinal data. Considerable generalization was required to reach acceptably low uniqueness levels, especially for longitudinal data. Detailed guidelines and disclosure policies on how to ensure that the re-identification risk is low are provided.ConclusionsA large percentage of Montreal residents are unique on basic demographics. For non-longitudinal data sets, the three character postal code, gender, and month/year of birth represent sufficiently low re-identification risk. Data custodians need to generalize their demographic information further for longitudinal data sets.

A scoping review of malaria forecasting: past work and future directions

Objectives There is a growing body of literature on malaria forecasting methods and the objective of our review is to identify and assess methods, including predictors, used to forecast malaria. Design Scoping review. Two independent reviewers searched information sources, assessed studies for inclusion and extracted data from each study. Information sources Search strategies were developed and the following databases were searched: CAB Abstracts, EMBASE, Global Health, MEDLINE, ProQuest Dissertations & Theses and Web of Science. Key journals and websites were also manually searched. Eligibility criteria for included studies We included studies that forecasted incidence, prevalence or epidemics of malaria over time. A description of the forecasting model and an assessment of the forecast accuracy of the model were requirements for inclusion. Studies were restricted to human populations and to autochthonous transmission settings. Results We identified 29 different studies that met our inclusion criteria for this review. The forecasting approaches included statistical modelling, mathematical modelling and machine learning methods. Climate-related predictors were used consistently in forecasting models, with the most common predictors being rainfall, relative humidity, temperature and the normalised difference vegetation index. Model evaluation was typically based on a reserved portion of data and accuracy was measured in a variety of ways including mean-squared error and correlation coefficients. We could not compare the forecast accuracy of models from the different studies as the evaluation measures differed across the studies. Conclusions Applying different forecasting methods to the same data, exploring the predictive ability of non-environmental variables, including transmission reducing interventions and using common forecast accuracy measures will allow malaria researchers to compare and improve models and methods, which should improve the quality of malaria forecasting.

Residential address errors in public health surveillance data: A description and analysis of the impact on geocoding

The residential addresses of persons with reportable communicable diseases are used increasingly for spatial monitoring and cluster detection, and public health may direct interventions based upon the results of routine spatial surveillance. There has been little assessment, however, of the quality of address data in reportable disease notifications and of the corresponding impact of these errors on geocoding and routine public health practices. The objectives of this study were to examine address errors for a selected reportable disease in a large urban center in Canada and to assess the impact of identified errors on geocoding and the estimated spatial distribution of the disease. We extracted data for all notifications of campylobacteriosis from the Montreal public health department from 1995 to 2008 and used an address verification algorithm to determine the validity of the residential address for each case and to suggest corrections for invalid addresses. We assessed the types of address errors as well as the resulting positional errors, calculating the distance between the original address and the correct address as well as changes in disease density. Address errors and missing addresses were prevalent in the public health records (10% and 5%, respectively) and they influenced the observed distribution of campylobacteriosis in Montreal, with address correction changing case location by a median of 1.1 km. Further examination of the extent of address errors in public health data is essential, as is the investigation of how these errors impact routine public health functions.

Socio-Economic Disparities in the Burden of Seasonal Influenza: The Effect of Social and Material Deprivation on Rates of Influenza Infection

Background There is little empirical evidence in support of a relationship between rates of influenza infection and level of material deprivation (i.e., lack of access to goods and services) and social deprivation (i.e. lack of social cohesion and support). Method Using validated population-level indices of material and social deprivation and medical billing claims for outpatient clinic and emergency department visits for influenza from 1996 to 2006, we assessed the relationship between neighbourhood rates of influenza and neighbourhood levels of deprivation using Bayesian ecological regression models. Then, by pooling data from neighbourhoods in the top decile (i.e., most deprived) and the bottom decile, we compared rates in the most deprived populations to the least deprived populations using age- and sex-standardized rate ratios. Results Deprivation scores ranged from one to five with five representing the highest level of deprivation. We found a 21% reduction in rates for every 1 unit increase in social deprivation score (rate ratio [RR] 0.79, 95% Credible Interval [CrI] 0.66, 0.97). There was little evidence of a meaningful linear relationship with material deprivation (RR 1.06, 95% CrI 0.93, 1.24). However, relative to neighbourhoods with deprivation scores in the bottom decile, those in the top decile (i.e., most materially deprived) had substantially higher rates (RR 2.02, 95% Confidence Interval 1.99, 2.05). Conclusion Though it is hypothesized that social and material deprivation increase risk of acute respiratory infection, we found decreasing healthcare utilization rates for influenza with increasing social deprivation. This finding may be explained by the fewer social contacts and, thus, fewer influenza exposure opportunities of the socially deprived. Though there was no evidence of a linear relationship with material deprivation, when comparing the least to the most materially deprived populations, we observed higher rates in the most materially deprived populations.

Electronic Event–based Surveillance for Monitoring Dengue, Latin America

Dengue, a potentially fatal disease, is spreading around the world. An estimated 2.5 billion people in tropical and subtropical regions are at risk. Early detection of outbreaks is crucial to prevention and control of dengue virus and other viruses. Case reporting may often take weeks or months. Therefore, researchers explored whether electronic sources of real-time information (such as Internet news outlets, health expert mailing lists, social media sites, and queries to online search engines) might be faster, and they were. Although information from unofficial sources should be interpreted with caution, when used in conjunction with traditional case reporting, real-time electronic surveillance can help public health authorities allocate resources in time to avert full-blown epidemics.

Neighborhood Determinants of 2009 Pandemic A/H1N1 Influenza Vaccination in Montreal, Quebec, Canada

Neighborhood-level analyses of influenza vaccination can identify the characteristics of vulnerable neighborhoods, which can inform public health strategy for future pandemics. In this study, the authors analyzed rates of 2009 pandemic A/H1N1 influenza vaccination in Montreal, Quebec, Canada, using individual-level vaccination records from a vaccination registry with census, survey, and administrative data to estimate the population at risk. The neighborhood socioeconomic and demographic determinants of vaccination were identified using Bayesian ecologic logistic regression, with random effects to account for spatial autocorrelation. A total of 918,773 (49.9%) Montreal residents were vaccinated against pandemic A/H1N1 influenza from October 22, 2009, through April 8, 2010. Coverage was greatest among females, children under age 5 years, and health-care workers. Neighborhood vaccine coverage ranged from 33.6% to 71.0%. Neighborhoods with high percentages of immigrants (per 5% increase, odds ratio = 0.90, 95% credible interval: 0.86, 0.95) and material deprivation (per 1-unit increase in deprivation score, odds ratio = 0.93, 95% credible interval: 0.88, 0.98) had lower vaccine coverage. Half of the Montreal population was vaccinated; however, considerable heterogeneity in coverage was observed between neighborhoods and subgroups. In future vaccination campaigns, neighborhoods that are materially deprived or have high percentages of immigrants may benefit from focused interventions.