Combined Probability Test Research Articles

Exercising prediction validation methods for analysis of quiet supersonic overflight

En route noise certification procedures for quiet supersonic aircraft are in development under the International Civil Aviation Organization for a proposed standard that could enable civil supersonic overland flight. While these quiet supersonic aircraft do not exist yet, methods are being developed and evaluated using existing N-wave empirical datasets and low-boom simulation datasets. One potential aspect of future certification methods is the ability to use a validated prediction model to predict the certification noise level under reference day weather conditions. Methods to validate predictions, such as the two-sample t-test, Fisher's combined probability test, two-sample Kolmogorov–Smirnov test, and equivalence testing, have been exercised and modified for this application. While some methods compare mean values, others compare values across the distribution, and a combination of methods can be used. The methods were applied to a subset of an N-wave dataset to demonstrate the applicability and identify potential areas for improvement. Atmospheric turbulence variability was included as a factor in these analyses through KZKFourier propagation simulations with parameters matching actual flights and weather conditions. Results are presented for six noise metrics (PL, ASEL, BSEL, DSEL, ESEL, and ISBAP), and plans to expand the analyses are discussed.

A-49 Predictors of sleep quality among adolescents following concussion.

Abstract Purpose To identify predictors of sleep quality following concussion in adolescents. Methods Ninety-four (47%-F) participants aged 11-17 (14.31±2.8) years within 7 days of injury were recruited from a specialty concussion clinic and emergency department. Participants completed demographics, mTBI and medical history, Adolescent Sleep-Wake Scale (ASWS), Concussion Clinical Profiles Screening (CP Screen), and Post-Concussion Symptom Scale (PCSS). Stepwise logistic regressions (LR) were used to identify predictors of ASWS scores. Results of the LR were used to create a probability test variable for receiver operating characteristic (ROC) curve analyses. Results Results of the LR for ASWS total sleep score were significant (p<.001) and accurately identified 83.1% of adolescents with sensitivity of 76.7% and specificity of 89.1%, with history of anxiety (Adj OR=9.52, 95%CI=1.71-52.63, p=.01) and CP Screen anxiety domain total score (Adj OR=1.92, 95% CI=1.41-2.61, p<.001) as significant predictors. ROC analysis of the combined probability test variable for total sleep score was significant with area under the curve (AUC)=.88 (95%CI=.80-.94, p<.001). Similar findings were supported for history of anxiety and CP Screen anxiety domain scores on individual sleep domain scores. Conclusions History of anxiety and higher CP Screen anxiety score were predictive of poor sleep among adolescents following concussion. Surprisingly, age, sex, mTBI and migraine history, as well as total higher symptoms burden were not associated with poor sleep following concussion. The findings suggest that clinicians should focus on interventions to reduce anxiety to enhance sleep among adolescents following concussion.

Mechanical behavior and its variability analysis of fiber reinforced polymer rebars after high temperatures

This study investigated the influence of high temperatures on the mechanical properties of glass and basalt fiber reinforced polymer (GFRP and BFRP) rebars. These rebars with a wide range of diameter from 10 to 25 mm were heated up to 400 ℃ and tested under shear and compression. Based on the data from the extensive experiments, the variability of shear and compressive strength was assessed. Results indicate that the reduction of both shear and compressive strength with high temperatures did not show notable dependence on the rebar diameter and type. High temperatures, especially 350 and 400 ℃, would induce more reduction of compressive strength than that of shear strength. Using the combined probability test and the Bayesian information criterion (BIC) test, this study proposed an approach to establishing appropriate probabilistic strength models for FPR rebars after high temperatures. The Gamma distribution outstood all chosen models for capturing the variability of residual shear strength of FRP rebars, while both Gamma and Weibull distributions were needed for compressive strength. Results herein can facilitate a better understanding of high-temperature induced change in properties of GFRP and BFRP rebars, and provide useful probabilistic strength models for probability-based evaluation of FRP rebars reinforced concrete structures after a fire.

Sleep disturbance is associated with perturbations in immune-inflammatory pathways in oncology outpatients undergoing chemotherapy

Objective/backgroundSleep disturbance is a common problem in patients receiving chemotherapy. Purpose was to evaluate for perturbations in immune-inflammatory pathways between oncology patients with low versus very high levels of sleep disturbance. Patients/methodsSleep disturbance was evaluated using the General Sleep Disturbance Scale six times over two cycles of chemotherapy. Latent profile analysis was used to identify subgroups of patients with distinct sleep disturbance profiles. Pathway impact analyses were performed in two independent samples using gene expression data obtained from RNA sequencing (n = 198) and microarray (n = 162) technologies. Fisher's combined probability test was used to identify significantly perturbed pathways between Low versus Very High sleep disturbance classes. ResultsIn the RNA sequencing and microarray samples, 59.1% and 51.9% of patients were in the Very High sleep disturbance class, respectively. Thirteen perturbed pathways were related to immune-inflammatory mechanisms (i.e., endocytosis, phagosome, antigen processing and presentation, natural killer cell mediated cytotoxicity, cytokine-cytokine receptor interaction, apoptosis, neutrophil extracellular trap formation, nucleotide-binding and oligomerization domain-like receptor signaling, Th17 cell differentiation, intestinal immune network for immunoglobulin A production, T-cell receptor signaling, complement and coagulation cascades, and tumor necrosis factor signaling). ConclusionsFirst study to identify perturbations in immune-inflammatory pathways associated with very high levels of sleep disturbance in oncology outpatients. Findings suggest that complex immune-inflammatory interactions underlie sleep disturbance.

Fisher’s Combined Probability Test for High-Dimensional Covariance Matrices

Testing large covariance matrices is of fundamental importance in statistical analysis with high-dimensional data. In the past decade, three types of test statistics have been studied in the literature: quadratic form statistics, maximum form statistics, and their weighted combination. It is known that quadratic form statistics would suffer from low power against sparse alternatives and maximum form statistics would suffer from low power against dense alternatives. The weighted combination methods were introduced to enhance the power of quadratic form statistics or maximum form statistics when the weights are appropriately chosen. In this article, we provide a new perspective to exploit the full potential of quadratic form statistics and maximum form statistics for testing high-dimensional covariance matrices. We propose a scale-invariant power-enhanced test based on Fisher’s method to combine the p-values of quadratic form statistics and maximum form statistics. After carefully studying the asymptotic joint distribution of quadratic form statistics and maximum form statistics, we first prove that the proposed combination method retains the correct asymptotic size under the Gaussian assumption, and we also derive a new Lyapunov-type bound for the joint distribution and prove the correct asymptotic size of the proposed method without requiring the Gaussian assumption. Moreover, we show that the proposed method boosts the asymptotic power against more general alternatives. Finally, we demonstrate the finite-sample performance in simulation studies and a real application. Supplementary materials for this article are available online.

Perturbations in Neuroinflammatory Pathways Are Associated With a Worst Pain Profile in Oncology Patients Receiving Chemotherapy

Unrelieved pain occurs in 55% of cancer patients. Identification of molecular mechanisms for pain may provide insights into therapeutic targets. Purpose was to evaluate for perturbations in neuroinflammatory pathways between oncology patients with and without severe pain. Worst pain severity was rated using a 0 to 10 numeric rating scale six times over two cycles of chemotherapy. Latent profile analysis was used to identify subgroups of patients with distinct pain profiles. Pathway impact analyses were performed in two independent samples using gene expression data obtained from RNA sequencing (n = 192) and microarray (n = 197) technologies. Fisher's combined probability test was used to identify significantly perturbed pathways between None versus the Severe pain classes. In the RNA sequencing and microarray samples, 62.5% and 56.3% of patients were in the Severe pain class, respectively. Nine perturbed pathways were related to neuroinflammatory mechanisms (i.e., retrograde endocannabinoid signaling, gamma-aminobutyric acid synapse, glutamatergic synapse, Janus kinase-signal transducer and activator of transcription signaling, phagosome, complement and coagulation cascades, cytokine-cytokine receptor interaction, chemokine signaling, calcium signaling). First study to identify perturbations in neuroinflammatory pathways associated with severe pain in oncology outpatients. Findings suggest that complex neuroimmune interactions are involved in the maintenance of chronic pain conditions. PerspectiveIn this study that compared oncology patients with none versus severe pain, nine perturbed neuroinflammatory pathways were identified. Findings suggest that complex neuroimmune interactions are involved in the maintenance of persistent pain conditions.

RapidET: a MEMS-based platform for label-free and rapid demarcation of tumors from normal breast biopsy tissues

The rapid and label-free diagnosis of malignancies in ex vivo breast biopsy tissues has significant utility in pathology laboratories and operating rooms. We report a MEMS-based platform integrated with microchips that performs phenotyping of breast biopsy tissues using electrothermal sensing. The microchip, fabricated on a silicon substrate, incorporates a platinum microheater, interdigitated electrodes (IDEs), and resistance temperature detectors (RTDs) as on-chip sensing elements. The microchips are integrated onto the platform using a slide-fit contact enabling quick replacement for biological measurements. The bulk resistivity (ρB), surface resistivity (ρS), and thermal conductivity (k) of deparaffinized and formalin-fixed paired tumor and adjacent normal breast biopsy samples from N = 8 patients were measured. For formalin-fixed samples, the mean ρB for tumors showed a statistically significant fold change of 4.42 (P = 0.014) when the tissue was heated from 25 °C to 37 °C compared to the adjacent normal tissue, which showed a fold change of 3.47. The mean ρS measurements also showed a similar trend. The mean k of the formalin-fixed tumor tissues was 0.309 ± 0.02 W m−1 K−1 compared to a significantly higher k of 0.563 ± 0.028 W m−1 K−1 for the adjacent normal tissues. A similar trend was observed in ρB,ρS, and k for the deparaffinized tissue samples. An analysis of a combination of ρB, ρS, and k using Fisher’s combined probability test and linear regression suggests the advantage of using all three parameters simultaneously for distinguishing tumors from adjacent normal tissues with higher statistical significance.

Solar-type periodicities in the climate variability of Northern Fennoscandia during the last three centuries: Real influence of solar activity or natural instability in the climate system

Fifteen proxy records of summer temperature in Fennoscandia, Northern Europe and in Yamal and Taymir Peninsulas (Western Siberia) were analyzed for the AD 1700–2000 period. Century-long (70–100 year) and quasi bi-decadal periodicities were found from proxy records representing different parts of Fennoscandia. Decadal variation was revealed in a smaller number of records. Statistically significant correlations were revealed between the timescale-dependent components of temperature variability and solar cycles of Schwabe (~11 year), Hale (~22 year), and Gleissberg (сentury-long) as recorded in solar activity data. Combining the results from our correlation analysis with the evidence of solar-climatic linkages over the Northern Fennoscandia obtained over the past 20 years suggest that there are two possible explanations for the obtained solar-proxy relations: (a) the Sun’s activity actually influences the climate variability in Northern Fennoscandia and in some regions of the Northern Hemisphere albeit the mechanism of such solar-climatic linkages are yet to be detailed; (b) the revealed solar-type periodicities result from natural instability of climate system and, in such a case, the correlations may appear purely by chance. Multiple lines of evidence support the first assumption but we note that the second one cannot be yet rejected. Guidelines for further research to elucidate this question are proposed including the Fisher’s combined probability test in the presence of solar signal in multiple proxy records.

A method to fuse multiphysics waveforms and improve predictive explosion detection: theory, experiment and performance

SUMMARYNatural and human-made sources of transient energy often emit multiple geophysical signatures that include mechanical and electromagnetic waveforms. We present a constructive method to fuse and evaluate statistics that we derive from such multiphysics waveforms that improves our capability to detect small, near-ground explosions over similar methods that consume single signature waveforms. Our method advances Fisher's Combined Probability Test (Fisher's Method) to operate under both hypotheses of a binary test on noisy data and provide researchers with the density functions required to forecast the ability of Fisher's Method to screen fused explosion signatures from noise. We apply this method against 12 d, multisignature explosion and noise records to show (1) that a fused multiphysics waveform statistic that combines radio, acoustic and seismic waveform data can identify explosions roughly 0.8 magnitude units lower than an acoustic emission, STA/LTA detector for the same detection probability and (2) that we can quantitatively predict how this fused, multiphysics statistic performs with Fisher's Method. Our work thereby offers a baseline method for predictive waveform fusion that supports multiphenomenological explosion monitoring (multiPEM) and is applicable to any binary testing problem in observational geophysics.

Systematic Review and Meta-Analysis of Cricoid Pressure Training and Education Efficacy.

We examined the evidence on the effectiveness of education and training on cricoid pressure (ie, Sellick maneuver) application. Cricoid pressure is used during rapid sequence induction and intubation to reduce the risk of pulmonary aspiration. We searched seven databases for studies on the education and training of health care personnel expected to apply cricoid pressure as a part of their regular clinical responsibilities. We appraised the included studies using standardized tools validated for this purpose and we performed a meta-analysis of the eight eligible studies. The success rate for intubation increased after training according to the random effect model (95% CI = 0.157 to 0.452). Fisher's method combined probability test rejected the null hypothesis for patients (P = 4.93e-6), indicating that the application of cricoid pressure significantly improved after educational interventions. Regularly scheduled training in the application of cricoid pressure could sustain health care personnel's ability to apply effective cricoid pressure.

Cardiac MR elastography using reduced-FOV, single-shot, spin-echo EPI.

PurposeTo implement a reduced field of view (rFOV) technique for cardiac MR elastography (MRE) and to demonstrate the improvement in image quality of both magnitude images and post‐processed MRE stiffness maps compared to the conventional full field of view (full‐FOV) acquisition.MethodsWith Institutional Review Board approval, 17 healthy volunteers underwent both full‐FOV and rFOV cardiac MRE scans using 140‐Hz vibrations. Two cardiac radiologists blindly compared the magnitude images and stiffness maps and graded the images based on several image quality attributes using a 5‐point ordinal scale. Fisher's combined probability test was performed to assess the overall evaluation. The octahedral shear strain‐based signal‐to‐noise ratio (OSS‐SNR) and median stiffness over the left ventricular myocardium were also compared.ResultsOne volunteer was excluded because of an inconsistent imaging resolution during the exam. In the remaining 16 volunteers (9 males, 7 females), the rFOV scans outperformed the full‐FOV scans in terms of subjective image quality and ghosting artifacts in the magnitude images and stiffness maps, as well as the overall preference. The quantitative measurements showed that rFOV had significantly higher OSS‐SNR (median: 1.4 [95% confidence interval (CI): 1.2–1.5] vs. 2.1 [95% CI: 1.8–2.4]), P < 0.05) compared to full‐FOV. Although no significant change was found in the median myocardial stiffness between the 2 scans, we observed a decrease in the stiffness variation within the myocardium from 2.1 kPa (95% CI: [1.9, 2.3]) to 1.9 kPa (95% CI: [1.7, 2.0]) for full‐FOV and rFOV, respectively (P < 0.05) in a subgroup of 7 subjects with ghosting present in the myocardium.ConclusionThis pilot volunteer study demonstrated that rFOV cardiac MRE has the capability to reduce ghosting and to improve image quality in both MRE magnitude images and stiffness maps. Magn Reson Med 80:231–238, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

FLOWER: Fusing global and local associations towards personalized social recommendation

In big data era people are dependent on a variety of social media to manage their social circles. Many online social networks employ social recommendation as an increasingly important component. Although global and local recommendation methods have achieved remarkable success, current studies seldom consider to play advantages of both in social networks. To demonstrate the effectiveness of incorporating local methods to global ones, we first investigated associations between triangular motifs and existing social relations and found that potential links are related with common relations. Further, we analyzed correlations of all methods and clustered them and found obvious strong correlations among methods with same type, which demonstrated the necessity of taking advantages of both. Consequently, we proposed a novel method FLOWER which resorts to Fisher’s combined probability test to systematically calibrate statistical significance of global and local associations. FLOWER utilizes information of social relations in both local and global scopes, which are less correlated with each other, and therefore imply possibilities of different aspects for a candidate link. We demonstrated the effectiveness of FLOWER by considering each possible pairwise combination of six global approaches with two local methods and performing 10-fold cross-validation experiments on five real social network datasets (Facebook, Last.fm, Epinions, HEP-PH and Delicious). Results show that FLOWER-based methods significantly outperform either their global or local components in accuracy and retrieval performance.

IFORM: Incorporating Find Occurrence of Regulatory Motifs.

Accurately identifying the binding sites of transcription factors (TFs) is crucial to understanding the mechanisms of transcriptional regulation and human disease. We present incorporating Find Occurrence of Regulatory Motifs (iFORM), an easy-to-use and efficient tool for scanning DNA sequences with TF motifs described as position weight matrices (PWMs). Both performance assessment with a receiver operating characteristic (ROC) curve and a correlation-based approach demonstrated that iFORM achieves higher accuracy and sensitivity by integrating five classical motif discovery programs using Fisher’s combined probability test. We have used iFORM to provide accurate results on a variety of data in the ENCODE Project and the NIH Roadmap Epigenomics Project, and the tool has demonstrated its utility in further elucidating individual roles of functional elements. Both the source and binary codes for iFORM can be freely accessed at https://github.com/wenjiegroup/iFORM. The identified TF binding sites across human cell and tissue types using iFORM have been deposited in the Gene Expression Omnibus under the accession ID GSE53962.

Fusing geophysical signatures of locally recorded surface explosions to improve blast detection

AbstractWe recorded acoustic, seismic and radio-frequency signatures of 70 solid charge (∼2–12 kg) surface explosions (shots) at local distances (0.1–1.5 km) to determine if such signals could be fused for blast monitoring. We observed that each geophysical signature was sufficiently repeatable between similar shots to be identifiable with multichannel correlation detectors. Using template signals from a large explosion, we then processed heavily contaminated data recording a smaller shot with these detectors, and missed or marginally detected the resultant target signals. By then fusing the p-values of these statistics through Fisher's combined probability test, we clearly identified the same explosion signals at thresholds consistent with the false alarm on noise rates of the correlation detectors. This resulting Fisher test thereby provided high-probability detections, zero false alarms and higher theoretical detection capability.

Statistical Method for Integrative Platform Analysis: Application to Integration of Proteomic and Microarray Data.

To perform integrative analysis on multiple genomic data sources, we propose to use Fisher's combined probability test for consolidated inference. The method combines the individual p-values from different data sources and constructs a chi-square test statistics for the overall significance. This method is valid to combine results across independent data sources. We further improve the method to accommodate the scenario that the data sources are dependent or the data samples are too small to obtain valid p-values through exact distributions. The proposed method is convenient to use in practice and is robust to distributional assumptions and small sample sizes.

How to Assess Whether Qualified Evaluations of e-Government Projects Are Conducted? The Case of Slovenia

Countries are investing significant resources in the development of e-government, but evaluations of e-government projects are rarely conducted in a comprehensive way. In order to take the right measures and to be effective in fostering the use and raising the quality of evaluations of e-government projects it is important to reveal the state of the practice. The paper presents the results of the research based on identification of parameters required to be measured in order to conduct qualified evaluations of e-government projects. Thirteen such paramaters were identified via the focus group, with participants comprising (Slovenian) e-government decision-makers and researchers. Using these 13 parameters and 7 additional questions about the knowledge and use of evalution methods a questionnaire was designed and applied in a survey on Slovenian public administration. Employing a binomial test and Fisher’s combined probability test it was established that in Slovenian public administration qualified evaluations of e-government projects were not being conducted. The method has demonstrated an eligible usability both for analysing the general situation and for exposing opportunities for improvement in various specific situations.

TAFFY: incorporating tag information into a diffusion process for personalized recommendations

The last few years have witnessed an explosion of information caused by the exponential growth of the Internet and World Wide Web, which confronted us with information overload and brought about an era of big data, appealing for efficient personalized recommender systems to assist the screening of useful information from various sources. As for a recommender system with more than the fundamental object-user rating information, such accessorial information as tags can be exploited and integrated into final ranking lists to improve recommendation performance. However, although existing studies have demonstrated that tags, as the additional yet useful resource, can be designed to improve recommendation performance, most network-based approaches take users, objects and tags as two bipartite graphs, or a tripartite graph, and therefore overlook either the important information among homogeneous nodes in each sub-graph, or the bipartite relations between users, objects or tags. Moreover, recent studies have suggested that the filtration of weak relationships in networks may reasonably enhance recommendation performance of collaborative filtering methods, and it has also been demonstrated that approaches based on the diffusion processes could more effectively capture relationships between objects and users, hence exhibiting higher performance than a typical collaborative filtering method. Based on these understandings, we propose a data fusion approach that integrates historical and tag data towards personalized recommendations. Our method coverts historical and tag data into complex networks, resorts to a diffusion kernel to measure the strength of associations between users and objects, and adopts Fisher's combined probability test to obtain the statistical significance of such associations for personalized recommendations. We validate our approach via 10-fold cross-validation experiments. Results show that our method outperforms existing methods in not only the recommendation accuracy and diversity, but also retrieval performance. We further show the robustness of our method to related parameters.

Meta-analysis based on weighted ordered P-values for genomic data with heterogeneity.

BackgroundMeta-analysis has become increasingly popular in recent years, especially in genomic data analysis, due to the fast growth of available data and studies that target the same questions. Many methods have been developed, including classical ones such as Fisher’s combined probability test and Stouffer’s Z-test. However, not all meta-analyses have the same goal in mind. Some aim at combining information to find signals in at least one of the studies, while others hope to find more consistent signals across the studies. While many classical meta-analysis methods are developed with the former goal in mind, the latter goal has much more practicality for genomic data analysis.ResultsIn this paper, we propose a class of meta-analysis methods based on summaries of weighted ordered p-values (WOP) that aim at detecting significance in a majority of studies. We consider weighted versions of classical procedures such as Fisher’s method and Stouffer’s method where the weight for each p-value is based on its order among the studies. In particular, we consider weights based on the binomial distribution, where the median of the p-values are weighted highest and the outlying p-values are down-weighted. We investigate the properties of our methods and demonstrate their strengths through simulations studies, comparing to existing procedures. In addition, we illustrate application of the proposed methodology by several meta-analysis of gene expression data.ConclusionsOur proposed weighted ordered p-value (WOP) methods displayed better performance compared to existing methods for testing the hypothesis that there is signal in the majority of studies. They also appeared to be much more robust in applications compared to the rth ordered p-value (rOP) method (Song and Tseng, Ann. Appl. Stat. 2014, 8(2):777–800). With the flexibility of incorporating different p-value combination methods and different weighting schemes, the weighted ordered p-values (WOP) methods have great potential in detecting consistent signal in meta-analysis with heterogeneity.

Proteomic Analysis of Eggs from Mytilus edulis Females Differing in Mitochondrial DNA Transmission Mode

Many bivalves have an unusual mechanism of mitochondrial DNA (mtDNA) inheritance called doubly uniparental inheritance (DUI) in which distinctly different genomes are inherited through the female (F genome) and male (M genome) lineages. In fertilized eggs that will develop into male embryos, the sperm mitochondria remain in an aggregation, which is believed to be delivered to the primordial germ cells and passed to the next generation through the sperm. In fertilized eggs that will develop into female embryos, the sperm mitochondria are dispersed throughout the developing embryo and make little if any contribution to the next generation. The frequency of embryos with the aggregated or dispersed mitochondrial type varies among females. Previous models of DUI have predicted that maternal nuclear factors cause molecular differences among unfertilized eggs from females producing embryos with predominantly dispersed or aggregated mitochondria. We test this hypothesis using females of each of the two types from a natural population. We have found small, yet detectable, differences of the predicted type at the proteome level. We also provide evidence that eggs of females giving the dispersed pattern have consistently lower expression for different proteasome subunits than eggs of females giving the aggregated pattern. These results, combined with those of an earlier study in which we used hatchery lines of Mytilus, and with a transcriptomic study in a clam that has the DUI system of mtDNA transmission, reinforce the hypothesis that the ubiquitin-proteasome system plays a key role in the mechanism of DUI and sex determination in bivalves. We also report that eggs of females giving the dispersed pattern have higher expression for arginine kinase and enolase, enzymes involved in energy production, whereas ferritin, which is involved in iron homeostasis, has lower expression. We discuss these results in the context of genetic models for DUI and suggest experimental methods for further understanding the role of these proteins in DUI.

Risk of Cardiovascular Disease from Antiretroviral Therapy for HIV: A Systematic Review

BackgroundRecent studies suggest certain antiretroviral therapy (ART) drugs are associated with increases in cardiovascular disease.PurposeWe performed a systematic review and meta-analysis to summarize the available evidence, with the goal of elucidating whether specific ART drugs are associated with an increased risk of myocardial infarction (MI).Data SourcesWe searched Medline, Web of Science, the Cochrane Library, and abstract archives from the Conference on Retroviruses and Opportunistic Infections and International AIDS Society up to June 2011 to identify published articles and abstracts.Study SelectionEligible studies were comparative and included MI, strokes, or other cardiovascular events as outcomes.Data ExtractionEligibility screening, data extraction, and quality assessment were performed independently by two investigators.Data SynthesisRandom effects methods and Fisher’s combined probability test were used to summarize evidence.FindingsTwenty-seven studies met inclusion criteria, with 8 contributing to a formal meta-analysis. Findings based on two observational studies indicated an increase in risk of MI for patients recently exposed (usually defined as within last 6 months) to abacavir (RR 1.92, 95% CI 1.51–2.42) and protease inhibitors (PI) (RR 2.13, 95% CI 1.06–4.28). Our analysis also suggested an increased risk associated with each additional year of exposure to indinavir (RR 1.11, 95% CI 1.05–1.17) and lopinavir (RR 1.22, 95% CI 1.01–1.47). Our findings of increased cardiovascular risk from abacavir and PIs were in contrast to four published meta-analyses based on secondary analyses of randomized controlled trials, which found no increased risk from cardiovascular disease.ConclusionAlthough observational studies implicated specific drugs, the evidence is mixed. Further, meta-analyses of randomized trials did not find increased risk from abacavir and PIs. Our findings that implicate specific ARTs in the observational setting provide sufficient evidence to warrant further investigation of this relationship in studies designed for that purpose.