Trust but Verify: Lessons Learned for the Application of AI to Case-Based Clinical Decision-Making From Postmarketing Drug Safety Assessment at the US Food and Drug Administration.

Adverse drug reactions are a common cause of morbidity in health care. The US Food and Drug Administration (FDA) evaluates individual case safety reports of adverse events (AEs) after submission to the FDA Adverse Event Reporting System as part of its surveillance activities. Over the past decade, the FDA has explored the application of artificial intelligence (AI) to evaluate these reports to improve the efficiency and scientific rigor of the process. However, a gap remains between AI algorithm development and deployment. This viewpoint aims to describe the lessons learned from our experience and research needed to address both general issues in case-based reasoning using AI and specific needs for individual case safety report assessment. Beginning with the recognition that the trustworthiness of the AI algorithm is the main determinant of its acceptance by human experts, we apply the Diffusion of Innovations theory to help explain why certain algorithms for evaluating AEs at the FDA were accepted by safety reviewers and others were not. This analysis reveals that the process by which clinicians decide from case reports whether a drug is likely to cause an AE is not well defined beyond general principles. This makes the development of high performing, transparent, and explainable AI algorithms challenging, leading to a lack of trust by the safety reviewers. Even accounting for the introduction of large language models, the pharmacovigilance community needs an improved understanding of causal inference and of the cognitive framework for determining the causal relationship between a drug and an AE. We describe specific future research directions that underpin facilitating implementation and trust in AI for drug safety applications, including improved methods for measuring and controlling of algorithmic uncertainty, computational reproducibility, and clear articulation of a cognitive framework for causal inference in case-based reasoning.

Integrated Hepatitis C-Opioid Use Disorder Care Through Facilitated Telemedicine: A Randomized Trial.

Importance: Facilitated telemedicine may promote hepatitis C virus elimination by mitigating geographic and temporal barriers. Objective: To compare sustained virologic responses for hepatitis C virus among persons with opioid use disorder treated through facilitated telemedicine integrated into opioid treatment programs compared with off-site hepatitis specialist referral. Design, Setting, and Participants: Prospective, cluster randomized clinical trial using a stepped wedge design. Twelve programs throughout New York State included hepatitis C-infected participants (n = 602) enrolled between March 1, 2017, and February 29, 2020. Data were analyzed from December 1, 2022, through September 1, 2023. Intervention: Hepatitis C treatment with direct-acting antivirals through comanagement with a hepatitis specialist either through facilitated telemedicine integrated into opioid treatment programs (n = 290) or standard-of-care off-site referral (n = 312). Main Outcomes and Measures: The primary outcome was hepatitis C virus cure. Twelve programs began with off-site referral, and every 9 months, 4 randomly selected sites transitioned to facilitated telemedicine during 3 steps without participant crossover. Participants completed 2-year follow-up for reinfection assessment. Inclusion criteria required 6-month enrollment in opioid treatment and insurance coverage of hepatitis C medications. Generalized linear mixed-effects models were used to test for the intervention effect, adjusted for time, clustering, and effect modification in individual-based intention-to-treat analysis. Results: Among 602 participants, 369 were male (61.3%); 296 (49.2%) were American Indian or Alaska Native, Asian, Black or African American, multiracial, or other (ie, no race category was selected, with race data collected according to the 5 standard National Institutes of Health categories); and 306 (50.8%) were White. The mean (SD) age of the enrolled participants in the telemedicine group was 47.1 (13.1) years; that of the referral group was 48.9 (12.8) years. In telemedicine, 268 of 290 participants (92.4%) initiated treatment compared with 126 of 312 participants (40.4%) in referral. Intention-to-treat cure percentages were 90.3% (262 of 290) in telemedicine and 39.4% (123 of 312) in referral, with an estimated logarithmic odds ratio of the study group effect of 2.9 (95% CI, 2.0-3.5; P < .001) with no effect modification. Observed cure percentages were 246 of 290 participants (84.8%) in telemedicine vs 106 of 312 participants (34.0%) in referral. Subgroup effects were not significant, including fibrosis stage, urban or rural participant residence location, or mental health (anxiety or depression) comorbid conditions. Illicit drug use decreased significantly (referral: 95% CI, 1.2-4.8; P = .001; telemedicine: 95% CI, 0.3-1.0; P < .001) among cured participants. Minimal reinfections (n = 13) occurred, with hepatitis C virus reinfection incidence of 2.5 per 100 person-years. Participants in both groups rated health care delivery satisfaction as high or very high. Conclusions and Relevance: Opioid treatment program-integrated facilitated telemedicine resulted in significantly higher hepatitis C virus cure rates compared with off-site referral, with high participant satisfaction. Illicit drug use declined significantly among cured participants with minimal reinfections. Trial Registration: ClinicalTrials.gov Identifier: NCT02933970.

Social determinants of health derived from people with opioid use disorder: Improving data collection, integration and use with cross-domain collaboration and reproducible, data-centric, notebook-style workflows.

Deriving social determinants of health from underserved populations is an important step in the process of improving the well-being of these populations and in driving policy improvements to facilitate positive change in health outcomes. Collection, integration, and effective use of clinical data for this purpose presents a variety of specific challenges. We assert that combining expertise from three distinct domains, specifically, medical, statistical, and computer and data science can be applied along with provenance-aware, self-documenting workflow tools. This combination permits data integration and facilitates the creation of reproducible workflows and usable (reproducible) results from the sensitive and disparate sources of clinical data that exist for underserved populations.

Multivariate semiparametric control charts for mixed-type data.

A useful tool that has gained popularity in the Quality Control area is the control chart which monitors a process over time, identifies potential changes, understands variations, and eventually improves the quality and performance of the process. This article introduces a new class of multivariate semiparametric control charts for monitoring multivariate mixed-type data, which comprise both continuous and discrete random variables (rvs). Our methodology leverages ideas from clustering and Statistical Process Control to develop control charts for MIxed-type data. We propose four control chart schemes based on modified versions of the KAy-means for MIxed LArge KAMILA data clustering algorithm, where we assume that the two existing clusters represent the reference and the test sample. The charts are semiparametric, the continuous rvs follow a distribution that belongs in the class of elliptical distributions. Categorical scale rvs follow a multinomial distribution. We present the algorithmic procedures and study the characteristics of the new control charts. The performance of the proposed schemes is evaluated on the basis of the False Alarm Rate and in-control Average Run Length. Finally, we demonstrate the effectiveness and applicability of our proposed methods utilizing real-world data.

High Satisfaction with Patient-Centered Telemedicine for Hepatitis C Virus Delivered to Substance Users: A Mixed-Methods Study.

Background: While telemedicine may increase health care access for vulnerable populations, data are limited on whether people with opioid use disorder (PWOUD) are satisfied with telemedicine. We assessed PWOUD satisfaction with telemedicine and identified factors that increase telemedicine satisfaction. Methods: We conducted a mixed-methods study among hepatitis C virus (HCV)-infected persons enrolled at 12 opioid treatment programs (OTPs) throughout New York State. Participants successfully completed HCV treatment either through telemedicine integrated into OTPs (N = 238) or through offsite referral (N = 106). We evaluated Patient Satisfaction Questionnaire (PSQ) response scores at the initial and final health care encounters and subsequently interviewed telemedicine study participants (N = 25) to assess their experiences with telemedicine. Results: All participants (N = 344) successfully completed HCV treatment. We observed no differences in PSQ scores between telemedicine and in-person encounters (98.3% and 98.7% of telemedicine participants provided PSQ scores of satisfied or highly satisfied at each timepoint, respectively). Study participants indicated that attributes associated with high telemedicine encounter satisfaction included: (1) communicating study information, (2) gaining trust, and (3) delivering patient-centered care. Participants weighted "General Satisfaction" and "Time Spent with Doctor" higher than "Accessibility and Convenience," and female participants were significantly more satisfied than males. Satisfaction with health care delivery among all participants increased significantly comparing timepoints. Conclusions: Participants were highly satisfied with HCV telemedicine encounters equivalent to in-person encounters. Communication augments trust facilitating delivery of patient-centered care through telemedicine. Participants value empathy and trust with providers over accessibility and convenience. In summary, PWOUD are highly satisfied with the facilitated telemedicine model and value empathetic and trusting providers. ClinicalTrials.gov Identifier: NCT02933970.

On the use of the likelihood ratio test methodology in pharmacovigilance.

The safety of medical products due to adverse events (AE) from drugs, therapeutic biologics, and medical devices is a major public health concern worldwide. Likelihood ratio test (LRT) approaches to pharmacovigilance constitute a class of rigorous statistical tools that permit objective identification of AEs of a specific drug and/or a class of drugs cataloged in spontaneous reporting system databases. However, the existing LRT approaches encounter certain theoretical and computational challenges when an underlying Poisson model assumption is violated, including in cases of zero-inflated data. We briefly review existing LRT approaches and propose a novel class of (pseudo-) LRT methods to address these challenges. Our approach uses an alternative parametrization to formulate a unified framework with a common test statistic that can handle both Poisson and zero-inflated Poisson (ZIP) models. The proposed framework is computationally efficient, and it reveals deeper insights into the comparative behaviors of the Poisson and the ZIP models for handling AE data. Our extensive simulation studies document notably superior performances of the proposed methods over existing approaches particularly under zero-inflation, both in terms of statistical (eg, much better control of the nominal level and false discovery rate with substantially enhanced power) and computational ( ∼ $$ \sim $$ 100-500-fold gains in average running times) performance metrics. An application of our method on the statin drug class from the FDA FAERS database reveals interesting insights on potential AEs. An R package, pvLRT, implementing our methods has been released in the public domain.

Innovations in education: A prospective study of storytelling narratives to enhance hepatitis C virus knowledge among substance users.

BACKGROUND: Even though substance users have the highest hepatitis C virus (HCV) burden, many lack knowledge about the infection. Lack of knowledge is an important obstacle to pursuing HCV care. Although printed materials are conventionally utilized to disseminate HCV-related knowledge, narrative story-telling videos may be an alternative. Data are extremely limited, however, in the ability of storytelling videos to increase HCV knowledge among substance users. In this study, we hypothesized that a story-telling narrative video would increase substance user's immediate and 1-month HCV-related knowledge compared to a printed format. AIM: To assess immediate and 1-month HCV-related knowledge retention among substance users comparing education delivered via a storytelling narrative video compared to a printed format. METHODS: We conducted a prospective matched, case-control study among substance users actively prescribed buprenorphine enrolled from two sites. The intervention site received the video and the control site, the brochure. Participants (n = 176) were matched on age, gender, and race. We obtained extensive patient and stakeholder input on the video's design, validated the video's content, and developed a recruitment plan to guide participant enrollment. Knowledge was assessed by administration of a 25-item instrument immediately before, immediately after, or one month after the intervention. Data were analyzed using nonparametric and generalized linear mixed-effects models. RESULTS: We recruited a total of 176 substance users, 90 and 86 individuals, from each site, respectively. One-month follow up occurred in 92% and 94% of enrollees in the control and intervention groups, respectively. In comparison with the pre-intervention scores, immediate knowledge recall increased significantly for both the intervention (P < 0.0001) and control (P < 0.0001) groups. Multivariate modeling revealed a significant improvement in HCV-related knowledge and retention (P = 0.033) among participants who viewed the storytelling video. CONCLUSION: Storytelling narratives emphasizing HCV education appear to be an effective method to increase HCV-related knowledge among substance users. They should become an educational cornerstone to promote HCV management among this population.

Patient-centered HCV care via telemedicine for individuals on medication for opioid use disorder: Telemedicine for Evaluation, Adherence and Medication for Hepatitis C (TEAM-C).

BACKGROUND: Telemedicine has the potential to increase healthcare access especially for vulnerable populations. Telemedicine for Evaluation, Adherence, and Medication for Hepatitis C (TEAM-C) is comparing telemedicine access to specialty medical care to usual care for management of hepatitis C virus (HCV) infection among persons with opioid use disorder (PWOUD). PWOUD have the highest hepatitis C virus (HCV) prevalence and incidence, yet they infrequently receive HCV care. The study objectives are to compare access to specialty care via telemedicine to offsite specialty referral (usual care) on 1) treatment initiation, completion, and sustained virological response, 2) patient satisfaction with health care delivery, and 3) HCV reinfection after successful HCV cure. METHODS: TEAM-C is a multi-site, non-blinded, randomized pragmatic clinical trial conducted at 12 opioid treatment programs (OTP) throughout New York State that utilizes the stepped-wedge design. The unit of randomization is the OTP with a total sample size of 624 participants. HCV-infected PWOUD were treated via telemedicine or referral. Telemedicine encounters are conducted onsite in the OTP with co-administration of direct acting antivirals for HCV with medications for opioid use disorder. The primary outcome is undetectable HCV RNA obtained 12 weeks post-treatment cessation. We also follow participants for two years to assess for reinfection. CONCLUSIONS: The study utilizes a rigorous study design to evaluate the effectiveness and implementation of virtual treatment for HCV integrated into behavioral treatment. We demonstrate the feasibility, engagement principles and lessons learned from the initial prospective randomized trial of telemedicine targeted to a vulnerable population.

Distance-Based Estimation Methods for Models for Discrete and Mixed-Scale Data.

Pearson residuals aid the task of identifying model misspecification because they compare the estimated, using data, model with the model assumed under the null hypothesis. We present different formulations of the Pearson residual system that account for the measurement scale of the data and study their properties. We further concentrate on the case of mixed-scale data, that is, data measured in both categorical and interval scale. We study the asymptotic properties and the robustness of minimum disparity estimators obtained in the case of mixed-scale data and exemplify the performance of the methods via simulation.

A framework for patient-centered telemedicine: Application and lessons learned from vulnerable populations.

Virtual technologies can facilitate clinical monitoring, clinician-patient interactions, and enhance patient-centered approaches to healthcare delivery. Telemedicine, two-way communication between a healthcare provider and a patient not in the same physical location, emphasizes patient preference and convenience by substituting the transportation of patients with information transfer. We present a framework for implementation of a comprehensive, dynamic, patient-centered telemedicine network deployed in 12 opioid treatment programs (OTP) located throughout New York State (NYS). The program aims to effectively manage hepatitis C virus (HCV) infection via telemedicine with co-administration of HCV and substance use medications. We have found that the Sociotechnical System model with emphasis on patient-centered factors provides a framework for telemedicine deployment and implementation to a vulnerable population. The issue of interoperability between the telemedicine platform and the electronic health record (EHR) system as well as clinical information retrieval for medical decision-making are challenges with implementation of a comprehensive, dynamic telemedicine system. Targeting telemedicine to a vulnerable population requires additional consideration of trust in the security and confidentiality of the telemedicine system. Our contribution is the valuable lessons learned from implementing a comprehensive, dynamic, patient-centered telemedicine system among an OTP network throughout NYS as applied to a vulnerable population that can be generalized to other difficult-to-reach populations.

Toward precision prescribing for methadone: Determinants of methadone deposition.

BACKGROUND: Despite the World Health Organization listing methadone as an essential medication, effective dose selection is challenging, especially in racial and ethnic minority populations. Subtherapeutic doses can result in withdrawal symptoms while supratherapeutic doses can result in overdose and death. Although CYP3A4 was conventionally considered the principal methadone metabolizing enzyme, more recent data have identified CYP2B6 as the principal enzyme. CYP2B6 has ethnically-associated polymorphisms that affect the metabolic rate. Our objective was to investigate the effects of genetic and nongenetic factors on methadone metabolism. METHODS: We measured trough plasma methadone levels in 100 participants with opioid use disorder. We assessed methadone metabolism by calculating the metabolite ratio (major metabolite: 2-ethylidene-1,5-dimethyl-3,3-diphenylpyrrolidine [EDDP] divided by methadone concentration). We assessed hepatic fibrosis and steatosis by transient elastography and CYP2B6 alleles, principally responsible for methadone metabolism. Mixed effects models modeled the data in 97 participants. RESULTS: Participants were largely male (58%), minority (61% African American) and non-Hispanic (68%). Forty percent were HCV mono-infected, 40% were uninfected, and 20% were HCV/HIV co-infected. Female sex had significant effects on (R)- and (S)-methadone metabolism (p = 0.016 and p = 0.044, respectively). CYP2B6 loss of function (LOF) alleles significantly affected (S)-methadone metabolism (p = 0.012). Body mass index (BMI) significantly affected (R)-methadone metabolism (p = 0.034). Methadone metabolism appeared to be lower in males, in individuals with LOF alleles, and elevated BMI. CONCLUSIONS: Genetic analysis, especially in minority populations, is essential to delivering individualized treatments. Although the principal methadone metabolizing enzyme remains controversial, our results suggest that sex, CYP2B6 genotype, and BMI should be incorporated into multivariate models to create methadone dosing algorithms. Methadone dosing algorithms should facilitate medication delivery, improve patient satisfaction, and diminish overdose potential.

An evaluation of statistical approaches to postmarketing surveillance.

Safety of medical products presents a serious concern worldwide. Surveillance systems of postmarket medical products have been established for continual monitoring of adverse events (AEs) in many countries, and the proliferation of electronic health record systems further facilitates continual monitoring for AEs. We review existing statistical methods for signal detection that are mostly in use in postmarketing safety surveillance of spontaneously reported AEs and we study their performance characteristics by simulation. We compare those with the likelihood ratio test (LRT) method (appropriately modified for use in pharmacovigilance) and use three different methods to generate data (AE based, drug based, and a modification of the method of Ahmed et al). Performance metrics include type I error, power, sensitivity, and false discovery rate, among others. The results show superior performance of the LRT method in almost all simulation experiments. An application to the FDA Adverse Event Reporting System database is illustrated using rhabdomyolysis-related preferred terms reported to FDA during the third-quarter of 2014 to the first-quarter of 2017 for statin drugs. We present a critical discussion and recommendations for use of these methods.

Influence of social determinants of health and substance use characteristics on persons who use drugs pursuit of care for hepatitis C virus infection.

BACKGROUND: Social determinants of health (SDOH), i.e., social, behavioral and environmental factors, are increasingly recognized for their important influence on health outcomes. Data are limited on the influence of SDOH, substance use characteristics, and their interactions on pursuit of hepatitis C virus (HCV) care among individuals with opioid use disorder (OUD). Linkage to HCV care remains low in this population despite high HCV prevalence and incidence. AIMS: To investigate the influence of SDOH, substance use factors, and their interactions on HCV treatment uptake among OUD patients in a methadone treatment program. METHODS: Information on patient demographics, SDOH, substance use characteristics, and co-morbid medical conditions were obtained from the paper and electronic medical records of OUD patients on methadone. We applied multiple correspondence analysis, k-means algorithm, and logistic regression with least absolute shrinkage and selection operator penalty to identify variables and clusters associated with pursuit of HCV care. RESULTS: Data from 161 patients (57% male, 60% Caucasian, mean age 45 years) were evaluated. Being employed, the absence of support systems, and a history of foster care were the strongest positive predictors of treatment pursuit. The use of crack/cocaine as the initial illicit substance, criminal activity without incarceration, and the absence of a family history of chemical dependency were the strongest negative predictors. We identified clusters among persons with OUD based upon their likelihood to pursue HCV management. CONCLUSION: Utilizing data from the medical record, we were able to identify factors positively and negatively associated with linkage-to-care for HCV. We were also able to divide patients into clusters of factors associated with linkage-to-care for HCV. These results could be used to identify individuals with OUD based upon their readiness for HCV care.

Integrated, Co-located, Telemedicine-based Treatment Approaches for Hepatitis C Virus Management in Opioid Use Disorder Patients on Methadone.

BACKGROUND: Despite high hepatitis C virus (HCV) prevalence, opioid use disorder (OUD) patients on methadone rarely engage in HCV treatment. We investigated the effectiveness of HCV management via telemedicine in an opioid substitution therapy (OST) program. METHODS: OUD patients on methadone underwent biweekly telemedicine sessions between a hepatologist and physician assistant during the entire HCV treatment course. All pretreatment labs (HCV RNA, genotype, and noninvasive fibrosis assessments) were obtained onsite and direct-acting antivirals were coadministered with methadone using modified directly observed therapy. We used multiple correspondence analysis, least absolute shrinkage and selection operator, and logistic regression to identify variables associated with pursuit of HCV care. RESULTS: Sixty-two HCV RNA-positive patients (24% human immunodeficiency virus [HIV] infected, 61% male, 61% African American, 25.8% Hispanic) were evaluated. All patients were stabilized on methadone and all except 4 were HCV genotype 1 infected. Advanced fibrosis/cirrhosis was present in 34.5% of patients. Of the 45 treated patients, 42 (93.3%) achieved viral eradication. Of 17 evaluated patients who were not treated, 5 were discontinued from the drug treatment program or did not follow up after the evaluation, 2 had HIV adherence issues, and 10 had insurance authorization issues. Marriage and a mental health diagnosis other than depression were the strongest positive predictors of treatment pursuit, whereas being divorced, separated, or widowed was the strongest negative predictor. CONCLUSIONS: HCV management via telemedicine integrated into an OST program is a feasible model with excellent virologic effectiveness. Psychosocial and demographic variables can assist in identification of subgroups with a propensity or aversion to pursue HCV treatment.

Patient Reaction to Telemedicine for Clinical Management of Hepatitis C Virus Integrated into an Opioid Treatment Program.

Background and Introduction: Virtual integration of hepatitis C virus (HCV) infection management within the opioid treatment program (OTP) through telemedicine may overcome limited treatment uptake encountered when patients are referred offsite. To evaluate the diffusion of telemedicine within the OTP, we conducted a pilot study to assess acceptance of and satisfaction with telemedicine among 45 HCV-infected opioid use disorder (OUD) patients on methadone.Materials and Methods: We administered a modified 11-item telemedicine satisfaction questionnaire after the initial HCV telemedicine evaluation, when initiating HCV treatment, and 3 months post-HCV treatment completion. Among a patient subset, a semistructured interview further assessed issues of participant referral to the telemedicine program as well as convenience and confidentiality with the telemedicine encounters.Results: Patients demonstrated their acceptance of telemedicine-based encounters by referral of additional participants. They highlighted the convenience of on-site treatment with a liver specialist through recognition of the benefit of "one-stop shopping." They also expressed confidence in the privacy and confidentiality of telemedicine encounters.Discussion: In this pilot study, telemedicine appears to be well accepted as a modality for HCV management among OUD patients on methadone. Virtual integration of medical and behavioral therapy through telemedicine warrants further investigation for its use in this population.Conclusions: In this pilot study, we found that a largely racial minority population of substance users grew to accept telemedicine over time with diminished privacy and confidentiality concerns. Telemedicine was well accepted within the OTP community as reflected by participant referral to the program.

Hepatitis C virus knowledge improves hepatitis C virus screening practices among primary care physicians.

AIM: To understand the role of knowledge as a promoter of hepatitis C virus (HCV) screening among primary care physicians (PCP). METHODS: A 45-item online questionnaire assessing knowledge of HCV natural history, risk factors, and treatment was distributed to 163 PCP. Logistic regression, adjusted for survey responses, assessed associations between PCP knowledge of HCV natural history and treatment and birth cohort (i.e., birth between 1945 and 1965) screening. Response stratification and weighting were used to account for nonresponse and to permit extension of responses to the entire survey population. Associations between various predictors including demographic characteristics, level of training, and HCV treatment experience and HCV knowledge were assessed. RESULTS: Ninety-one individuals (55.8%) responded. Abnormal liver enzymes (49.4%), assessment of HCV-related risk factors (30.6%), and birth cohort membership (20%) were the leading HCV screening indications. Most PCP (64.7%) felt that the combination of risk-factor and birth cohort screening utilizing a self-administered survey while awaiting the physician (55.3%) were the most efficient screening practices. Implementation of birth cohort screening was associated with awareness of the recommendations (P-value = 0.01), knowledge of HCV natural history (P-value < 0.01), and prior management of HCV patients (P-value < 0.01). PCP with knowledge of HCV treatment was also knowledgeable about HCV natural history (P-value < 0.01). Similarly, awareness of age-based screening recommendations was associated with HCV treatment knowledge (P-value = 0.03). CONCLUSION: Comprehensive knowledge of HCV is critical to motivate HCV screening. PCP-targeted educational interventions are required to expand the HCV workforce and linkage-to-care opportunities as we seek global HCV eradication.

Non-Quadratic Distances in Model Assessment.

One natural way to measure model adequacy is by using statistical distances as loss functions. A related fundamental question is how to construct loss functions that are scientifically and statistically meaningful. In this paper, we investigate non-quadratic distances and their role in assessing the adequacy of a model and/or ability to perform model selection. We first present the definition of a statistical distance and its associated properties. Three popular distances, total variation, the mixture index of fit and the Kullback-Leibler distance, are studied in detail, with the aim of understanding their properties and potential interpretations that can offer insight into their performance as measures of model misspecification. A small simulation study exemplifies the performance of these measures and their application to different scientific fields is briefly discussed.

A Comparison of High Dimensional Variable Selection Methods with Missing Covariates in a Prostate Cancer Study.

Prostate cancer is the most common cancer in American men. Dozens of specific genes have been shown to be correlated to prostate cancer, to benign and non-benign cancer cases, from a biology perspective. In this paper, we apply a penalized logistic regression model with different penalty functions to select genes that contribute to benign and non-benign cases, based on the data from a prostate cancer study. The tuning parameter is determined by cross validation. In order to take into account some specific genes that have been classified as prostate cancer genes through biology research but with missing values, multiple imputation is adopted to create complete data sets. We analyze the prostate cancer data by comparing the selection results with completely observed data only, and the results with imputed data. We also conduct a simulation study to validate our proposed method.

A comparative study of subgroup identification methods for differential treatment effect: Performance metrics and recommendations.

Subgroup identification with differential treatment effects serves as an important step towards precision medicine, as it provides evidence regarding how individuals with specific characteristics respond to a given treatment. This knowledge not only supports the tailoring of treatment strategies but also prompts the development of new treatments. This manuscript provides a brief overview of the issues associated with the methodologies aimed at identifying subgroups with differential treatment effects, and studies in depth the operational characteristics of five data-driven methods that have appeared recently in the literature. The performance of the methods under study to identify correctly the covariates affecting treatment effects is evaluated via simulation and under various conditions. Two clinical trial data sets are also used to illustrate the application of these methods. Discussion and recommendations pertaining to the use of these methods are provided, with emphasis on the relative performance of the methods under the conditions studied.

Hepatitis C virus core antigen: A potential alternative to HCV RNA testing among persons with substance use disorders.

BACKGROUND: The hepatitis C virus (HCV) core antigen (HCVcAg) may be an alternative diagnostic method to HCV RNA especially in populations such as substance users, the homeless or in resource-limited settings. AIMS: To evaluate performance of HCVcAg test in patients with opioid use disorder (OUD) on methadone in order to document its performance characteristics in the target population and to ensure that its specificity remains consistent across different populations. METHODS: HCVcAg levels from 109 methadone-maintained patients were compared to HCV RNA levels. RESULTS: Mean age was 53.8±7.8years, 59.6% were male, 68.8% African American, and 44% HCV-infected. HCVcAg was detectable in 47 of 48 HCV-infected, and undetectable in all HCV RNA negative patients. The HCVcAg assay had sensitivity of 97.9% and specificity of 100%. Correlation with HCV RNA levels was excellent (r=0.88, 95% CI 0.76; 0.95, p<0.01). CONCLUSION: HCVcAg has excellent performance for the diagnosis of HCV infection in patients with OUD on methadone.