Big Data and Adverse Drug Reaction Detection.

Big Data holds the promise of fundamentally transforming the manner in which adverse drug reactions can be identified and evaluated. This commentary discusses new data sources that are envisioned to form a Big Data-enabled pharmacovigilance system and the role of these data in powering the future of adverse drug reactions detection.

Failing softly: a fracture theory of highly-deformable materials.

Highly-deformable materials, from synthetic hydrogels to biological tissues, are becoming increasingly important from both fundamental and practical perspectives. Their mechanical behaviors, in particular the dynamics of crack propagation during failure, are not yet fully understood. Here we propose a theoretical framework for the dynamic fracture of highly-deformable materials, in which the effects of a dynamic crack are treated with respect to the nonlinearly deformed (pre-stressed/strained), non-cracked, state of the material. Within this framework, we derive analytical and semi-analytical solutions for the near-tip deformation fields and energy release rates of dynamic cracks propagating in incompressible neo-Hookean solids under biaxial and uniaxial loading. We show that moderately large pre-stressing has a marked effect on the stress fields surrounding a crack's tip. We verify these predictions by performing extensive experiments on the fracture of soft brittle elastomers over a range of loading levels and propagation velocities, showing that the newly developed framework offers significantly better approximations to the measurements than standard approaches at moderately large levels of external loadings and high propagation velocities. This framework should be relevant to the failure analysis of soft and tough, yet brittle, materials.

Toward enhanced pharmacovigilance using patient-generated data on the internet.

The promise of augmenting pharmacovigilance with patient-generated data drawn from the Internet was called out by a scientific committee charged with conducting a review of the current and planned pharmacovigilance practices of the US Food and Drug Administration (FDA). To this end, we present a study on harnessing behavioral data drawn from Internet search logs to detect adverse drug reactions (ADRs). By analyzing search queries collected from 80 million consenting users and by using a widely recognized benchmark of ADRs, we found that the performance of ADR detection via search logs is comparable and complementary to detection based on the FDA's adverse event reporting system (AERS). We show that by jointly leveraging data from the AERS and search logs, the accuracy of ADR detection can be improved by 19% relative to the use of each data source independently. The results suggest that leveraging nontraditional sources such as online search logs could supplement existing pharmacovigilance approaches.

Performance of pharmacovigilance signal-detection algorithms for the FDA adverse event reporting system.

Signal-detection algorithms (SDAs) are recognized as vital tools in pharmacovigilance. However, their performance characteristics are generally unknown. By leveraging a unique gold standard recently made public by the Observational Medical Outcomes Partnership (OMOP) and by conducting a unique systematic evaluation, we provide new insights into the diagnostic potential and characteristics of SDAs that are routinely applied to the US Food and Drug Administration (FDA) Adverse Event Reporting System (AERS). We find that SDAs can attain reasonable predictive accuracy in signaling adverse events. Two performance classes emerge, indicating that the class of approaches that address confounding and masking effects benefits safety surveillance. Our study shows that not all events are equally detectable, suggesting that specific events might be monitored more effectively using other data sources. We provide performance guidelines for several operating scenarios to inform the trade-off between sensitivity and specificity for specific use cases. We also propose an approach and demonstrate its application in identifying optimal signaling thresholds, given specific misclassification tolerances.

Pharmacovigilance using clinical notes.

With increasing adoption of electronic health records (EHRs), there is an opportunity to use the free-text portion of EHRs for pharmacovigilance. We present novel methods that annotate the unstructured clinical notes and transform them into a deidentified patient-feature matrix encoded using medical terminologies. We demonstrate the use of the resulting high-throughput data for detecting drug-adverse event associations and adverse events associated with drug-drug interactions. We show that these methods flag adverse events early (in most cases before an official alert), allow filtering of spurious signals by adjusting for potential confounding, and compile prevalence information. We argue that analyzing large volumes of free-text clinical notes enables drug safety surveillance using a yet untapped data source. Such data mining can be used for hypothesis generation and for rapid analysis of suspected adverse event risk.

Novel data-mining methodologies for adverse drug event discovery and analysis.

An important goal of the health system is to identify new adverse drug events (ADEs) in the postapproval period. Datamining methods that can transform data into meaningful knowledge to inform patient safety have proven essential for this purpose. New opportunities have emerged to harness data sources that have not been used within the traditional framework. This article provides an overview of recent methodological innovations and data sources used to support ADE discovery and analysis.

Biclustering of adverse drug events in the FDA's spontaneous reporting system.

In this article, we present a new pharmacovigilance data mining technique based on the biclustering paradigm, which is designed to identify drug groups that share a common set of adverse events (AEs) in the spontaneous reporting system (SRS) of the US Food and Drug Administration (FDA). A taxonomy of biclusters is developed, revealing that a significant number of bona fide adverse drug event (ADE) biclusters have been identified. Statistical tests indicate that it is extremely unlikely that the bicluster structures thus discovered, as well as their content, could have arisen by mere chance. Some of the biclusters classified as indeterminate provide support for previously unrecognized and potentially novel ADEs. In addition, we demonstrate the potential importance of the proposed methodology in several important aspects of pharmacovigilance such as providing insight into the etiology of ADEs, facilitating the identification of novel ADEs, suggesting methods and a rationale for aggregating terminologies, highlighting areas of focus, and providing an exploratory tool for data mining.

Development and experience with an algorithm to evaluate suspected smallpox cases in the United States, 2002-2004.

Concerns that smallpox, an eradicated disease, might reappear because of a bioterror attack and limited experience with smallpox diagnosis in the United States prompted us to design a clinical algorithm. We used clinical features of classic smallpox to classify persons presenting with suspected smallpox rashes into 3 categories: those with high, those with moderate, and those with low risk of having smallpox. The classification guides subsequent diagnostic strategies, limiting smallpox laboratory testing to high-risk persons to minimize the number of false-positive test results. From January 2002 through June 2004, the Centers for Disease Control and Prevention (CDC) received 43 consultations regarding suspected smallpox cases. No patient was at high risk for having smallpox. One patient was tested for the presence of variola virus. Varicella was the diagnosis for 23 cases (53%). The algorithm worked well to guide clinical and public health responses to suspected smallpox cases. The poster is available from CDC, and an interactive version and laboratory protocol are available at http://www.bt.cdc.gov/agent/smallpox/diagnosis/riskalgorithm/index.asp. We recommend use of the algorithm in the United States and elsewhere.

Elimination of new chronic hepatitis B virus infections: results of the Alaska immunization program.

An immunization assessment and a serologic survey were conducted to evaluate the effectiveness of a hepatitis B immunization program in eliminating hepatitis B virus (HBV) transmission among Alaska Natives in a region in which HBV infection is endemic. Hepatitis B vaccine coverage was 93% among 567 children

Injections given in healthcare settings as a major source of acute hepatitis B in Moldova.

BACKGROUND: Reported rates of acute hepatitis B are high in many former Soviet Union republics and modes of transmission are not well defined. METHODS: Two case-control studies were undertaken in Moldova to identify risk factors for acute hepatitis B in people aged 2-15 years (children) and > or =15 years (adults). Serologically confirmed acute hepatitis B cases occurring between 1 January 1994 and 30 August 30 1995, were matched on age, sex, and district of residence to three potential controls who were tested for hepatitis B markers to exclude the immune. Stratified odds ratios (SOR) were calculated using bivariate and multivariate methods. RESULTS: In multivariate analysis, compared with the 175 controls, the 70 adult cases (mean age 25 years, 66% male) were more likely to report receiving injections in the 6 months before illness during a dental visit (SOR = 21; 95% CI: 3.7-120), a hospital visit (SOR = 35; 95% CI: 7.2-170), or a visit to the polyclinic (SOR = 13; 95% CI: 2.4-74). Among children, receiving injections during a hospital visit (SOR = 5.2; 95% CI: 1.2-23) was the only exposure reported significantly more often by the 19 cases (mean age 8 years, 68% male) compared with the 81 controls. CONCLUSION: These results, along with reported unsafe injection practices in Moldova, suggest that injections are a major source of hepatitis B virus transmission and highlight the importance of proper infection-control procedures in preventing transmission of blood-borne infections.

PIP: Two case-control studies were conducted between January 1994 and August 1995 to determine the relative importance of injections and other exposures as a source of acute hepatitis B in Moldova among adults (aged 15 years) and children (aged 2-15 years). Results showed that injections in various health care settings were associated with acute hepatitis B and showed a higher proportion among adults compared with children. Contact with an HBsAg-positive person was also associated with illnesses; however, there was no statistically significant association between acute hepatitis B and other exposures. The risk of HBV transmission following percutaneous exposure is high (at least 30%). Calculation of the population attributable to risk suggests that injections associated with acute hepatitis B cases occurred in adults (52%) and children (21%). Adverse effects of injections may not be apparent in causing chronic infections. Transmission of blood-borne pathogens through unsafe injection practices is a problem increasingly identified worldwide.

Prevalence of hepatitis B, D and C virus infections among children and pregnant women in Moldova: additional evidence supporting the need for routine hepatitis B vaccination of infants.

Rates of acute hepatitis B are high in Moldova, but the prevalence of chronic infection is unknown. In 1994, we surveyed children and pregnant women, collected demographic information, and drew blood for laboratory testing. Among the 439 children (mean age, 5 years), the prevalence of antibody to hepatitis B core antigen (anti-HBc) and hepatitis B surface antigen (HBsAg) were 17.1 and 6.8%, respectively. Among the 1098 pregnant women (mean age, 26 years), 52.4% were anti-HBc-positive and 9.7% were HBsAg-positive. Of the HBsAg-positive pregnant women, 35.6% were hepatitis B e antigen (HBeAg) positive and 18.3% had antibodies to hepatitis D virus. The prevalence of antibody to hepatitis C virus was 1.4% in children and 2.3% in pregnant women. The high HBeAg prevalence among HBsAg-positive pregnant women and the high anti-HBc prevalence among children indicate that both perinatal and early childhood transmission contribute to the high hepatitis B virus endemicity in Moldova.

Transmission of hepatitis B virus to multiple patients from a surgeon without evidence of inadequate infection control.

BACKGROUND: Although about 1 percent of surgeons are infected with hepatitis B virus (HBV), transmission from surgeons to patients is thought to be uncommon. In July 1992, a 47-year-old woman became ill with acute hepatitis B after undergoing a thymectomy in which a thoracic-surgery resident who had had acute hepatitis B six months earlier assisted. METHODS: To determine whether the surgeon transmitted HBV to this patient and others, we conducted chart reviews, interviews, and serologic testing of thoracic-surgery patients at the two hospitals where the surgeon worked from July 1991 to July 1992. Hepatitis B surface antigen (HBsAg) subtypes and DNA sequences from the surgeon and from infected patients were determined. RESULTS: Of 144 susceptible patients in whose surgery the infected surgeon participated, 19 had evidence of recent HBV infection (13 percent). One of the hospitals was selected for additional study, and none of the 124 susceptible patients of the other thoracic surgeons at this hospital had evidence of recent HBV infection (relative risk, infinity; 95 percent confidence interval, 4.7 to infinity). No evidence was found for any common source of HBV other than the infected surgeon. The HBsAg subtype and the partial HBV DNA sequences from the surgeon were identical to those in the infected patients. Transmission of the infection was associated with cardiac transplantation (relative risk, 4.9; 95 percent confidence interval, 1.5 to 15.5) but not with other surgical procedures. The surgeon was positive for hepatitis B e antigen and had a high serum HBV DNA concentration (15 ng per milliliter). Our investigations identified no deficiencies in the surgeon's infection-control practices. CONCLUSIONS: In this outbreak there was surgeon-to-patient HBV transmission despite apparent compliance with recommended infection-control practices. We could not identify any specific events that led to transmission.

Serum cytokine profiles in experimental human malaria. Relationship to protection and disease course after challenge.

Serum cytokine profiles were evaluated in immunized and nonimmunized human volunteers after challenge with infectious Plasmodium falciparum sporozoites. Three volunteers had been immunized with x-irradiated sporozoites and were fully protected from infection. Four nonimmune volunteers all developed symptomatic infection at which time they were treated. Sera from all volunteers were collected at approximately 20 time points during the 28-d challenge period; levels of IL-1 alpha, IL-1 beta, IL-2, IFN-gamma, tumor necrosis factor-alpha, IL-4, IL-6, granulocyte macrophage-colony-stimulating factor, and soluble CD4, CD8, and IL-2 receptor (sCD4, sCD8, and sIL-2R, respectively) were determined by ELISA. C-reactive protein (CRP) was assayed by radial immunodiffusion. Parasitemic subjects developed increases in CRP and IFN-gamma, with less marked increases in sIL-2R and sCD8; the other cytokines tested did not change. CRP increases were abrupt and occurred at the onset of fever (day 14 after challenge). IFN-gamma increases were also abrupt, preceding those of fever and CRP by one day. Increases in sIL-2R and sCD8 were more gradual. Increases in fever, CRP, IFN-gamma, and sCD8 were concordant in each volunteer. Early IL-6 increases were noted in the protected vaccinees. Thus, after challenge with virulent P. falciparum, unique systemic cytokine profiles were detectable both in immunized, nonparasitemic volunteers and in unvaccinated, parasitemic subjects. The contrasting cytokine profiles in the two groups may relate to mechanisms of protection and immunopathology in experimental human malaria.

DNA modification in rat lungs following intratracheal or subcutaneous administration of 4-nitroquinoline-1-oxide, benzo[a]pyrene or 2-aminoanthracene.

Benzo[a]pyrene (BP)-, 2-aminoanthracene (2AA)- and 4-nitroquinoline-1-oxide (4NQO)-mediated DNA modification were investigated in rat lungs by using alkaline sucrose gradient sedimentation. The exposure-route, the physicochemical nature of the administered compound and the number of treatments were all important in determining the extent of DNA modification. 4NQO produced qualitatively similar modification whether instilled intratracheally (i.t.) as a suspension or injected subcutaneously (s.c.) in a soluble form. BP and 2AA produced no DNA alteration when injected s.c; they did, however, modify DNA sedimentation when instilled as a suspension, but not until 24 h after treatment. Furthermore, BP caused no DNA modification at any sampling time when instilled in a lipid solvent. In contrast to the DNA modification observed at 24 h after a single i.t. treatment with a BP suspension, no such alteration was detected 12 or 24 h after the last of 5 similar daily treatments. These results are discussed with respect to mechanisms of differential transport, clearance and metabolism of administered carcinogens.