Evolution of HIV virulence in response to disease-modifying vaccines: A modeling study.

Pathogens face a tradeoff with respect to virulence; while more virulent strains often have higher per-contact transmission rates, they are also more likely to kill their hosts earlier. Because virulence is a heritable trait, there is concern that a disease-modifying vaccine, which reduces the disease severity of an infected vaccinee without changing the underlying pathogen genotype, may result in the evolution of higher pathogen virulence. We explored the potential for such virulence evolution with a disease-modifying HIV-1 vaccine in an agent-based stochastic epidemic model of HIV in United States men who have sex with men (MSM). In the model, vaccinated agents received no protection against infection, but experienced lower viral loads and slower disease progression. We compared the genotypic set point viral load (SPVL), a measure of HIV virulence, in populations given vaccines that varied in the degree of SPVL reduction they induce. Sensitivity analyses were conducted under varying vaccine coverage scenarios. With continual vaccination rollout under ideal circumstances of 90 % coverage over thirty years, the genotypic SPVL of vaccinated individuals evolved to become greater than the genotypic SPVL of unvaccinated individuals. This virulence evolution in turn diminished the public health benefit of the vaccine, and in some scenarios resulted in an accelerated epidemic. These findings demonstrate the complexity of viral evolution and have important implications for the design and development of HIV vaccines.

Why does age at HIV infection correlate with set point viral load? An evolutionary hypothesis.

BACKGROUND: Set-point viral load (SPVL) correlates with the age at which people acquire HIV. Although immunosenescence may seem like a parsimonious explanation for this, it does not easily explain the observation that the relationship between age and SPVL attenuates when accounting for source partner SPVL. Here we propose an alternative explanation that encompasses this latter finding: that decreasing risk of acquisition with older age generates a selection bottleneck that selects for more virulent strains with age. METHODS: We adapted a previously published model of HIV transmission and evolution (EvoNetHIV), parameterized here for men who have sex with men (MSM). We conducted a series of simulation experiments that vary seven behavioral or clinical parameters that affect exposure risk as people age. We conducted regressions to determine the mean increase in SPVL per 10-year increase in seroconversion age, with and without source SPVL in the model. RESULTS: All runs generated significant relationships between seroconversion age and SPVL when not including source SPVL. All saw attenuated relationships, most to near 0, with source SPVL included. Four of our behavioral measures (relational duration, age-related homophily, coital frequency, and mean age at relationship formation) had clear effects on this relationship, all in the hypothesized direction. Combining multiple forms of behavioral heterogeneity yielded an increase of 0.056 log10 copies/mL SPVL per 10-year increase in seroconversion age, nearly as large as that seen in two empirical studies of age-SPVL correlations in MSM. CONCLUSION: The higher virulence of HIV among those infected later in life may be partly explained by a combination of selective bottlenecks and behavioral heterogeneity by age. Variation in the strength of this effect across populations may be in part due to different behavioral, epidemiological and clinical conditions, and not require assumptions about differences in patterns of immunosenescence among populations.

Evaluation of individual and ensemble probabilistic forecasts of COVID-19 mortality in the United States.

Short-term probabilistic forecasts of the trajectory of the COVID-19 pandemic in the United States have served as a visible and important communication channel between the scientific modeling community and both the general public and decision-makers. Forecasting models provide specific, quantitative, and evaluable predictions that inform short-term decisions such as healthcare staffing needs, school closures, and allocation of medical supplies. Starting in April 2020, the US COVID-19 Forecast Hub (https://covid19forecasthub.org/) collected, disseminated, and synthesized tens of millions of specific predictions from more than 90 different academic, industry, and independent research groups. A multimodel ensemble forecast that combined predictions from dozens of groups every week provided the most consistently accurate probabilistic forecasts of incident deaths due to COVID-19 at the state and national level from April 2020 through October 2021. The performance of 27 individual models that submitted complete forecasts of COVID-19 deaths consistently throughout this year showed high variability in forecast skill across time, geospatial units, and forecast horizons. Two-thirds of the models evaluated showed better accuracy than a naïve baseline model. Forecast accuracy degraded as models made predictions further into the future, with probabilistic error at a 20-wk horizon three to five times larger than when predicting at a 1-wk horizon. This project underscores the role that collaboration and active coordination between governmental public-health agencies, academic modeling teams, and industry partners can play in developing modern modeling capabilities to support local, state, and federal response to outbreaks.

Risk compensation after HIV-1 vaccination may accelerate viral adaptation and reduce cost-effectiveness: a modeling study.

Pathogen populations can evolve in response to selective pressure from vaccine-induced immune responses. For HIV, models predict that viral adaptation, either via strain replacement or selection on de novo mutation, may rapidly reduce the effectiveness of an HIV vaccine. We hypothesized that behavioral risk compensation after vaccination may accelerate the transmission of vaccine resistant strains, increasing the rate of viral adaptation and leading to a more rapid decline in vaccine effectiveness. To test our hypothesis, we modeled: (a) the impact of risk compensation on rates of HIV adaptation via strain replacement in response to a partially effective vaccine; and (b) the combined impact of risk compensation and viral adaptation on vaccine-mediated epidemic control. We used an agent-based epidemic model that was calibrated to HIV-1 trends in South Africa, and includes demographics, sexual network structure and behavior, and within-host disease dynamics. Our model predicts that risk compensation can increase the rate of HIV viral adaptation in response to a vaccine. In combination, risk compensation and viral adaptation can, under certain scenarios, reverse initial declines in prevalence due to vaccination, and result in HIV prevalence at 15 years equal to or greater than prevalence without a vaccine.

Test-and-treat coverage and HIV virulence evolution among men who have sex with men.

HIV set point viral load (SPVL), the viral load established shortly after initial infection, is a proxy for HIV virulence: higher SPVLs lead to higher risk of transmission and faster disease progression. Three models of test-and-treat scenarios, mainly in heterosexual populations, found that increasing treatment coverage selected for more virulent viruses. We modeled virulence evolution in a population of men who have sex with men (MSM) with increasing test-and-treat coverage. We extended a stochastic, dynamic network model (EvoNetHIV). We varied relationship patterns (MSM vs. heterosexual), HIV transmission models (increasing vs. plateauing probability of transmission at very high viral loads), and treatment roll-out (with explicit testing or fixed intervals between infection and treatment). In scenarios most similar to previous models (longer relational durations and the plateauing transmission function), we replicated trends previously found: increasing treatment coverage led to increased virulence (0.12 log10 increase in mean population SPVL between 20% and 100% treatment coverage). In scenarios reflecting MSM behavioral data using the increasing transmission function, increasing treatment coverage selected for viruses with lower virulence (0.16 log10 decrease in mean population SPVL between 20% and 100% treatment coverage). These findings emphasize the impact of sexual network conditions and transmission function details on predicted epidemiological and evolutionary outcomes. Varying these features creates very different evolutionary environments, which in turn lead to opposite effects in mean population SPVL evolution. Our results suggest that, under some realistic conditions, effective test-and-treat strategies may not face the previously reported tradeoff in which increasing coverage leads to evolution of greater virulence. This suggests instead that a virtuous cycle of increasing treatment coverage and diminishing virulence is possible.

Large benefits to youth-focused HIV treatment-as-prevention efforts in generalized heterosexual populations: An agent-based simulation model.

Predominantly heterosexual HIV-1 epidemics like those in sub-Saharan Africa continue to have high HIV incidence in young people. We used a stochastic, agent-based model for age-disparate networks to test the hypothesis that focusing uptake and retention of ART among youth could enhance the efficiency of treatment as prevention (TasP) campaigns. We used the model to identify strategies that reduce incidence to negligible levels (i.e., < 0.1 cases/100 person-years) 20-25 years after initiation of a targeted TasP campaign. The model was parameterized using behavioral, demographic, and clinical data from published papers and national reports. To keep a focus on the underlying age effects we model a generalized heterosexual population with average risks (i.e., no MSM, no PWIDs, no sex workers) and no entry of HIV+ people from other regions. The model assumes that most people (default 95%, range in variant simulations 60-95%) are "linkable"; i.e., could get linked to effective care given sufficient resources. To simplify the accounting, we assume a rapid jump in the number of people receiving treatment at the start of the TasP campaign, followed by a 2% annual increase that continues until all linkable HIV+ people have been treated. Under historical scenarios of CD4-based targeted ART allocation and current policies of untargeted (random) ART allocation, our model predicts that viral replication would need to be suppressed in 60-85% of infected people at the start of the TasP campaign to drive incidence to negligible levels. Under age-based strategies, by contrast, this percentage dropped by 18-54%, depending on the strength of the epidemic and the age target. For our baseline model, targeting those under age 30 halved the number of people who need to be treated. Age-based targeting also minimized total and time-discounted AIDS deaths over 25 years. Age-based targeting yielded benefits without being highly exclusive; in a model in which 60% of infected people were treated, ~87% and ~58% of those initiating therapy during a campaign targeting those <25 and <30 years, respectively, fell outside the target group. Sensitivity analyses revealed that youth-focused TasP is beneficial due to age-related risk factors (e.g. shorter relationship durations), and an age-specific herd immunity (ASHI) effect that protects uninfected adolescents entering the sexually active population. As testing rates increase in response to UNAIDS 90-90-90 goals, efforts to link all young people to care and treatment could contribute enormously to ending the HIV epidemic.

Models to predict the public health impact of vaccine resistance: A systematic review.

Pathogen evolution is a potential threat to the long-term benefits provided by public health vaccination campaigns. Mathematical modeling can be a powerful tool to examine the forces responsible for the development of vaccine resistance and to predict its public health implications. We conducted a systematic review of existing literature to understand the construction and application of vaccine resistance models. We identified 26 studies that modeled the public health impact of vaccine resistance for 12 different pathogens. Most models predicted that vaccines would reduce overall disease burden in spite of evolution of vaccine resistance. Relatively few pathogens and populations for which vaccine resistance may be problematic were covered in the reviewed studies, with low- and middle-income countries particularly under-represented. We discuss the key components of model design, as well as patterns of model predictions.

Relational concurrency, stages of infection, and the evolution of HIV set point viral load.

HIV viral load (VL) predicts both transmission potential and rate of disease progression. For reasons that are still not fully understood, the set point viral load (SPVL) established after acute infection varies across individuals and populations. Previous studies have suggested that population mean SPVL (MSPVL) has evolved near an optimum that reflects a trade-off between transmissibility and host survival. Sexual network structures affect rates of potential exposure during different within-host phases of infection marked by different transmission probabilities, and thus affect the number and timing of transmission events. These structures include relational concurrency, which has been argued to explain key differences in HIV burden across populations. We hypothesize that concurrency will alter the fitness landscape for SPVL in ways that differ from other network features whose impacts accrue at other times during infection. To quantitatively test this hypothesis, we developed a dynamic, stochastic, data-driven network model of HIV transmission, and evolution to assess the impact of key sexual network phenomena on MSPVL evolution. Experiments were repeated in sensitivity runs that made different assumptions about transmissibility during acute infection, SPVL heritability, and the functional form of the relationship between VL and transmissibility. For our main transmission model, scenarios yielded MSPVLs ranging from 4.4 to 4.75 log10 copies/ml, covering much of the observed empirical range. MSPVL evolved to be higher in populations with high concurrency and shorter relational durations, with values varying over a clinically significant range. In linear regression analyses on these and other predictors, main effects were significant (P < 0.05), as were interaction terms, indicating that effects are interdependent. We also noted a strong correlation between two key emergent properties measured at the end of the simulations-MSPVL and HIV prevalence-most clearly for phenomena that affect transmission networks early in infection. Controlling for prevalence, high concurrency yielded higher MSPVL than other network phenomena. Interestingly, we observed lower prevalence in runs in which SPVL heritability was zero, indicating the potential for viral evolution to exacerbate disease burden over time. Future efforts to understand empirical variation in MSPVL should consider local HIV burden and basic sexual behavioral and network structure.

HIV population-level adaptation can rapidly diminish the impact of a partially effective vaccine.

BACKGROUND: Development of an HIV vaccine might be essential to ending the HIV/AIDS pandemic. However, vaccines can result in the emergence and spread of vaccine-resistant strains. Indeed, analyses of breakthrough infections in the HIV phase 3 vaccine trial RV144 identified HIV genotypes with differential rates of transmission in vaccine and placebo recipients. We hypothesized that, for HIV vaccination programs based on partially effective vaccines similar to RV144, HIV adaptation will rapidly diminish the expected vaccine impact. METHODS AND FINDINGS: Using two HIV epidemic models, we simulated large-scale vaccination programs and, critically, included HIV strain diversity with respect to the vaccine response. We show here that rapid population-level viral adaptation can lead to decreased overall vaccine efficacy and substantially fewer infections averted by vaccination, when comparing scenarios with and without viral evolution (with outcomes depending on vaccination coverage, vaccine efficacy against the sensitive allele, and the initial resistant allele frequency). Translating this to the epidemic in South Africa, a scenario with 70% vaccination coverage may result in 250,000 infections (non-averted by vaccination) within 10 years of vaccine rollout that are due solely to HIV adaptation, all else being equal. CONCLUSIONS: These findings suggest that approaches to HIV vaccine development, program implementation, and epidemic modeling may require attention to viral adaptation in response to vaccination.

Making pharmacogenomic-based prescribing alerts more effective: A scenario-based pilot study with physicians.

To facilitate personalized drug dosing (PDD), this pilot study explored the communication effectiveness and clinical impact of using a prototype clinical decision support (CDS) system embedded in an electronic health record (EHR) to deliver pharmacogenomic (PGx) information to physicians. We employed a conceptual framework and measurement model to access the impact of physician characteristics (previous experience, awareness, relative advantage, perceived usefulness), technology characteristics (methods of implementation-semi-active/active, actionability-low/high) and a task characteristic (drug prescribed) on communication effectiveness (usefulness, confidence in prescribing decision), and clinical impact (uptake, prescribing intent, change in drug dosing). Physicians performed prescribing tasks using five simulated clinical case scenarios, presented in random order within the prototype PGx-CDS system. Twenty-two physicians completed the study. The proportion of physicians that saw a relative advantage to using PGx-CDS was 83% at the start and 94% at the conclusion of our study. Physicians used semi-active alerts 74-88% of the time. There was no association between previous experience with, awareness of, and belief in a relative advantage of using PGx-CDS and improved uptake. The proportion of physicians reporting confidence in their prescribing decisions decreased significantly after using the prototype PGx-CDS system (p=0.02). Despite decreases in confidence, physicians perceived a relative advantage to using PGx-CDS, viewed semi-active alerts on most occasions, and more frequently changed doses toward doses supported by published evidence. Specifically, sixty-five percent of physicians reduced their dosing, significantly for capecitabine (p=0.002) and mercaptopurine/thioguanine (p=0.03). These findings suggest a need to improve our prototype such that PGx CDS content is more useful and delivered in a way that improves physician's confidence in their prescribing decisions. The greatest increases in communication effectiveness and clinical impact of PGx-CDS are likely to be realized through continued focus on content, content delivery, and tailoring to physician characteristics.

Usability evaluation of pharmacogenomics clinical decision support aids and clinical knowledge resources in a computerized provider order entry system: a mixed methods approach.

BACKGROUND: Pharmacogenomics (PGx) is positioned to have a widespread impact on the practice of medicine, yet physician acceptance is low. The presentation of context-specific PGx information, in the form of clinical decision support (CDS) alerts embedded in a computerized provider order entry (CPOE) system, can aid uptake. Usability evaluations can inform optimal design, which, in turn, can spur adoption. OBJECTIVES: The study objectives were to: (1) evaluate an early prototype, commercial CPOE system with PGx-CDS alerts in a simulated environment, (2) identify potential improvements to the system user interface, and (3) understand the contexts under which PGx knowledge embedded in an electronic health record is useful to prescribers. METHODS: Using a mixed methods approach, we presented seven cardiologists and three oncologists with five hypothetical clinical case scenarios. Each scenario featured a drug for which a gene encoding drug metabolizing enzyme required consideration of dosage adjustment. We used Morae(®) to capture comments and on-screen movements as participants prescribed each drug. In addition to PGx-CDS alerts, 'Infobutton(®)' and 'Evidence' icons provided participants with clinical knowledge resources to aid decision-making. RESULTS: Nine themes emerged. Five suggested minor improvements to the CPOE user interface; two suggested presenting PGx information through PGx-CDS alerts using an 'Infobutton' or 'Evidence' icon. The remaining themes were strong recommendations to provide succinct, relevant guidelines and dosing recommendations of phenotypic information from credible and trustworthy sources; any more information was overwhelming. Participants' median rating of PGx-CDS system usability was 2 on a Likert scale ranging from 1 (strongly agree) to 7 (strongly disagree). CONCLUSIONS: Usability evaluation results suggest that participants considered PGx information important for improving prescribing decisions; and that they would incorporate PGx-CDS when information is presented in relevant and useful ways.

Visualization and analytics tools for infectious disease epidemiology: a systematic review.

BACKGROUND: A myriad of new tools and algorithms have been developed to help public health professionals analyze and visualize the complex data used in infectious disease control. To better understand approaches to meet these users' information needs, we conducted a systematic literature review focused on the landscape of infectious disease visualization tools for public health professionals, with a special emphasis on geographic information systems (GIS), molecular epidemiology, and social network analysis. The objectives of this review are to: (1) identify public health user needs and preferences for infectious disease information visualization tools; (2) identify existing infectious disease information visualization tools and characterize their architecture and features; (3) identify commonalities among approaches applied to different data types; and (4) describe tool usability evaluation efforts and barriers to the adoption of such tools. METHODS: We identified articles published in English from January 1, 1980 to June 30, 2013 from five bibliographic databases. Articles with a primary focus on infectious disease visualization tools, needs of public health users, or usability of information visualizations were included in the review. RESULTS: A total of 88 articles met our inclusion criteria. Users were found to have diverse needs, preferences and uses for infectious disease visualization tools, and the existing tools are correspondingly diverse. The architecture of the tools was inconsistently described, and few tools in the review discussed the incorporation of usability studies or plans for dissemination. Many studies identified concerns regarding data sharing, confidentiality and quality. Existing tools offer a range of features and functions that allow users to explore, analyze, and visualize their data, but the tools are often for siloed applications. Commonly cited barriers to widespread adoption included lack of organizational support, access issues, and misconceptions about tool use. DISCUSSION AND CONCLUSION: As the volume and complexity of infectious disease data increases, public health professionals must synthesize highly disparate data to facilitate communication with the public and inform decisions regarding measures to protect the public's health. Our review identified several themes: consideration of users' needs, preferences, and computer literacy; integration of tools into routine workflow; complications associated with understanding and use of visualizations; and the role of user trust and organizational support in the adoption of these tools. Interoperability also emerged as a prominent theme, highlighting challenges associated with the increasingly collaborative and interdisciplinary nature of infectious disease control and prevention. Future work should address methods for representing uncertainty and missing data to avoid misleading users as well as strategies to minimize cognitive overload.

Expanding assessments of translational research programs: supplementing metrics with value judgments.

The University of Washington (UW) Institute for Translational Health Sciences (ITHS), funded by a Clinical and Translational Sciences Award program, has supplemented its initial Kellogg Logic Model-based program evaluation with the eight judgment-based evaluative elements of the World Health Organization's (WHO) Health Services Assessment Model. This article describes the relationship between the two models, the rationale for the decision to supplement the evaluation with WHO evaluative elements, the value-added results of the WHO evaluative elements, and plans for further developing the WHO assessments.

Using random walks to identify cancer-associated modules in expression data.

BACKGROUND: The etiology of cancer involves a complex series of genetic and environmental conditions. To better represent and study the intricate genetics of cancer onset and progression, we construct a network of biological interactions to search for groups of genes that compose cancer-related modules. Three cancer expression datasets are investigated to prioritize genes and interactions associated with cancer outcomes. Using a graph-based approach to search for communities of phenotype-related genes in microarray data, we find modules of genes associated with cancer phenotypes in a weighted interaction network. RESULTS: We implement Walktrap, a random-walk-based community detection algorithm, to identify biological modules predisposing to tumor growth in 22 hepatocellular carcinoma samples (GSE14520), adenoma development in 32 colorectal cancer samples (GSE8671), and prognosis in 198 breast cancer patients (GSE7390). For each study, we find the best scoring partitions under a maximum cluster size of 200 nodes. Significant modules highlight groups of genes that are functionally related to cancer and show promise as therapeutic targets; these include interactions among transcription factors (SPIB, RPS6KA2 and RPS6KA6), cell-cycle regulatory genes (BRSK1, WEE1 and CDC25C), modulators of the cell-cycle and proliferation (CBLC and IRS2) and genes that regulate and participate in the map-kinase pathway (MAPK9, DUSP1, DUSP9, RIPK2). To assess the performance of Walktrap to find genomic modules (Walktrap-GM), we evaluate our results against other tools recently developed to discover disease modules in biological networks. Compared with other highly cited module-finding tools, jActiveModules and Matisse, Walktrap-GM shows strong performance in the discovery of modules enriched with known cancer genes. CONCLUSIONS: These results demonstrate that the Walktrap-GM algorithm identifies modules significantly enriched with cancer genes, their joint effects and promising candidate genes. The approach performs well when evaluated against similar tools and smaller overall module size allows for more specific functional annotation and facilitates the interpretation of these modules.

Technical desiderata for the integration of genomic data into Electronic Health Records.

The era of "Personalized Medicine," guided by individual molecular variation in DNA, RNA, expressed proteins and other forms of high volume molecular data brings new requirements and challenges to the design and implementation of Electronic Health Records (EHRs). In this article we describe the characteristics of biomolecular data that differentiate it from other classes of data commonly found in EHRs, enumerate a set of technical desiderata for its management in healthcare settings, and offer a candidate technical approach to its compact and efficient representation in operational systems.

Timeliness and data element completeness of immunization data in Washington State in 2010: a comparison of data exchange methods.

Health information systems receive data through various methods. These data exchange methods have the potential to influence data quality. We assessed a de-identified 2010 dataset including 757,476 demographic records and 2,634,101 vaccination records from Washington State's Immunization Information System (IIS) to describe timeliness and completeness of IIS data across several data exchange methods: manual entry, HL7, and flat file upload. Overall, manually-entered data and HL7 records were more timely than records imported as flat files. Completeness, though very high overall, was slightly higher for records arriving via flat file. Washington State IIS users, including clinicians and public health, rely on its data to inform patient care and determine population coverage of immunizations. Our results suggest that although data element completeness in systems like Washington's IIS will likely not be immediately or significantly impacted by provider's migration to HL7 connections with IISs, timeliness could be substantially improved when using HL7 connections.

Methods to identify standard data elements in clinical and public health forms.

The fragmentation of clinical and public health systems results in divergent information collection practices, presenting challenges to standardization and EHR certification efforts. Data forms employed in public health jurisdictions nationwide reflect these differences in patient treatment, monitoring and evaluation, and follow-up, presenting challenges for data integration. To study these variations, we surveyed tuberculosis contact investigation forms from all fifty states, three municipalities and two countries. We apply statistics and cluster analysis to analyze the divergent content of contact investigation forms with the goal of characterizing normative practices and identifying a common core of data fields. We found widespread variation in data elements between states in the study, with the "Name" field being the only ubiquitous data element. Our method reveals distinct groupings of data fields employed in certain regions, allowing the simultaneous identification of core standard data fields as well as variations in practice.

Comparison of electronic data capture (EDC) with the standard data capture method for clinical trial data.

BACKGROUND: Traditionally, clinical research studies rely on collecting data with case report forms, which are subsequently entered into a database to create electronic records. Although well established, this method is time-consuming and error-prone. This study compares four electronic data capture (EDC) methods with the conventional approach with respect to duration of data capture and accuracy. It was performed in a West African setting, where clinical trials involve data collection from urban, rural and often remote locations. METHODOLOGY/PRINCIPAL FINDINGS: Three types of commonly available EDC tools were assessed in face-to-face interviews; netbook, PDA, and tablet PC. EDC performance during telephone interviews via mobile phone was evaluated as a fourth method. The Graeco Latin square study design allowed comparison of all four methods to standard paper-based recording followed by data double entry while controlling simultaneously for possible confounding factors such as interview order, interviewer and interviewee. Over a study period of three weeks the error rates decreased considerably for all EDC methods. In the last week of the study the data accuracy for the netbook (5.1%, CI95%: 3.5-7.2%) and the tablet PC (5.2%, CI95%: 3.7-7.4%) was not significantly different from the accuracy of the conventional paper-based method (3.6%, CI95%: 2.2-5.5%), but error rates for the PDA (7.9%, CI95%: 6.0-10.5%) and telephone (6.3%, CI95% 4.6-8.6%) remained significantly higher. While EDC-interviews take slightly longer, data become readily available after download, making EDC more time effective. Free text and date fields were associated with higher error rates than numerical, single select and skip fields. CONCLUSIONS: EDC solutions have the potential to produce similar data accuracy compared to paper-based methods. Given the considerable reduction in the time from data collection to database lock, EDC holds the promise to reduce research-associated costs. However, the successful implementation of EDC requires adjustment of work processes and reallocation of resources.

Supervised learning for the automated transcription of spacer classification from spoligotype films.

BACKGROUND: Molecular genotyping of bacteria has revolutionized the study of tuberculosis epidemiology, yet these established laboratory techniques typically require subjective and laborious interpretation by trained professionals. In the context of a Tuberculosis Case Contact study in The Gambia we used a reverse hybridization laboratory assay called spoligotype analysis. To facilitate processing of spoligotype images we have developed tools and algorithms to automate the classification and transcription of these data directly to a database while allowing for manual editing. RESULTS: Features extracted from each of the 1849 spots on a spoligo film were classified using two supervised learning algorithms. A graphical user interface allows manual editing of the classification, before export to a database. The application was tested on ten films of differing quality and the results of the best classifier were compared to expert manual classification, giving a median correct classification rate of 98.1% (inter quartile range: 97.1% to 99.2%), with an automated processing time of less than 1 minute per film. CONCLUSION: The software implementation offers considerable time savings over manual processing whilst allowing expert editing of the automated classification. The automatic upload of the classification to a database reduces the chances of transcription errors.