Development and validation of an imageless machine-learning algorithm for the initial screening of prostate cancer.

PURPOSE: Prostate specific antigen (PSA) testing is a low-cost screening method for prostate cancer (PCa). However, its accuracy is limited. While progress is being made using medical imaging for PCa screening, PSA testing can still be improved as an easily accessible first step in the screening process. We aimed to develop and validate a new model by further personalizing the analysis of PSA with demographic, medical history, lifestyle parameters, and digital rectal examination (DRE) results. METHODS: Using data from 34,224 patients in the screening arm of the PLCO trial (22,188 for the training set and 12,036 for the validation set), we applied a gradient-boosting model whose features (Model 1) were one PSA value and the personal variables available in the PLCO trial except those that signaled an ex-ante assumption of PCa. A second algorithm (Model 2) included a DRE result. The primary outcome was the occurrence of PCa, while the aggressiveness of PCa was a secondary outcome. ROC analyses were used to compare both models to other initial screening tests. RESULTS: The areas under the curve (AUC) for Model 2 was 0.894 overall and 0.908 for patients with a suspicious DRE, compared to 0.808 for PSA for patients with a suspicious DRE. The AUC for Model 1 was 0.814 compared to 0.821 for PSA. Model 2 predicted 58% more high-risk PCa than PSA ≥4 combined with an abnormal DRE and had a positive predictive value of 74.7% (vs. 50.6%). CONCLUSION: Personalizing the interpretation of PSA values and DRE results with a gradient-boosting model showed promising results as a potential novel, low-cost method for the initial screening of PCa. The importance of DRE, when included in such a model, was also highlighted.

Uptake and Discontinuation of Pre-Exposure Prophylaxis Among Uninsured Transgender and Cisgender Women: A Public-Private Partnership Model in North Carolina.

HIV pre-exposure prophylaxis (PrEP) remains underutilized among cis and trans women. The PrEP Initiative Program (PIP) is a novel public-private partnership implemented at 12 local clinics in North Carolina. PIP provides HIV/sexually transmitted infection (STI) testing and clinical and laboratory monitoring for PrEP to uninsured/underinsured clients. We sought to understand service-related differences among both cis and trans women enrolled in PIP, including STIs diagnoses, clinic type, sources of referral, services needed, and reasons for PrEP discontinuation. The Kaplan-Meier curves display retention on PrEP over the duration of the program. Since 2018, 142 women (cis n = 113; trans n = 29) enrolled, and 136 started PrEP. The majority were ages 25-34 years (31.7%) or 18-24 years (29.6%), Black (57.8%) or Latinx (24.7%). Approximately 20.6% of recipients reported at least one STI while enrolled. Overall, trans women requested fewer services than cis women. After accounting for the amount of time each patient was taking PrEP, there were higher rates of trans women diagnosed with syphilis than cis women. Rates of persons with other STIs were not notably different between trans and cis women. Clinic access varied by gender: 69% of trans women were enrolled at only one site. Trans women were retained significantly longer: The Kaplan-Meier adjusted median time to discontinuation was 560 and 238 days for trans and cis women, respectively. PIP successfully reached historically marginalized and uninsured cis and trans women who may benefit from PrEP. Further investigations into factors contributing to recruitment and retention of women in HIV prevention programs are needed.

PrEP awareness, willingness, and likelihood to use future HIV prevention methods among undergraduate college students in an ending the HIV epidemic jurisdiction.

OBJECTIVE: Identify factors associated with PrEP awareness, willingness, and future prevention modalities among undergraduate college students. PARTICIPANTS: Undergraduates (N = 701) were recruited from a private university, a public research university, and a private historically Black college and university for an online survey. METHODS: Upon multiple imputations, a multivariate logistic model, a multivariate multinomial model, and independent multivariate ordinal logistic models were used to calculate Rubin's rules-pooled adjusted odds ratios for PrEP awareness, willingness, and future HIV prevention methods. RESULTS: Only 33.4% of students had heard of and 32.4% were willing to take PrEP. PrEP willingness was higher among sexual minority students compared to heterosexual/straight students (OR = 1.65; 95% CI: 1.03-2.63); p = .036). The likelihood to take a future vaccine or antibody prophylaxis treatment was higher than the likelihood to take injectable PrEP or implants. CONCLUSIONS: Interventions to increase PrEP uptake and willingness among undergraduates should emphasize equity in HIV education and include future prevention modalities.

Lost in translation: College students' knowledge of HIV and PrEP in relation to their sexual health behaviors.

OBJECTIVE: Human immunodeficiency virus (HIV) is a notable public health problem among young adults. The present study examined college students' knowledge of HIV and pre-exposure prophylaxis (PrEP) in relation to their sexual health behaviors. PARTICIPANTS AND METHOD: Participants included 1516 students who completed questionnaires on actual and perceived HIV knowledge, perceived PrEP knowledge, and sexual health behaviors. RESULTS: While knowledge of HIV was high, knowledge of PrEP was low. Approximately 73% of the sample reported not using condoms at all times, and 41% reported never having been tested for HIV. Women, students at public schools, upper-year students, and students with higher HIV and PrEP knowledge were more likely to get tested for HIV than their counterparts. Knowledge of HIV and PrEP did not relate to condom use. CONCLUSION: These results suggest the need for increased education about PrEP and strategies to help students translate knowledge about HIV to recommended sexual health behaviors.

Applying a machine learning modelling framework to predict delayed linkage to care in patients newly diagnosed with HIV in Mecklenburg County, North Carolina, USA.

BACKGROUND: Machine learning has the potential to help researchers better understand and close the gap in HIV care delivery in large metropolitan regions such as Mecklenburg County, North Carolina, USA. OBJECTIVES: We aim to identify important risk factors associated with delayed linkage to care for HIV patients with novel machine learning models and identify high-risk regions of the delay. METHODS: Deidentified 2013-2017 Mecklenburg County surveillance data in eHARS format were requested. Both univariate analyses and machine learning random forest model (developed in R 3.5.0) were applied to quantify associations between delayed linkage to care (>30 days after diagnosis) and various risk factors for individual HIV patients. We also aggregated linkage to care by zip codes to identify high-risk communities within the county. RESULTS: Types of HIV-diagnosing facility significantly influenced time to linkage; first diagnosis in hospital was associated with the shortest time to linkage. HIV patients with lower CD4+ cell counts (<200/ml) were twice as likely to link to care within 30 days than those with higher CD4+ cell count. Random forest model achieved high accuracy (>80% without CD4+ cell count data and >95% with CD4+ cell count data) to predict risk of delay in linkage to care. In addition, we also identified top high-risk zip codes of delayed linkage. CONCLUSION: The findings helped public health teams identify high-risk communities of delayed HIV care continuum across Mecklenburg County. The methodology framework can be applied to other regions with HIV epidemic and challenge of delayed linkage to care.

Blood DNA Yield but Not Integrity or Methylation Is Impacted After Long-Term Storage.

Collection of human whole blood for genomic DNA extraction is part of numerous clinical studies. Since DNA extraction cannot always be performed at the time of sample collection, whole blood samples may be stored for years before being processed. The use of appropriate storage conditions is then critical to obtain DNA in sufficient quantity and of adequate quality in order to obtain reliable results from the subsequent molecular biological analyses. In this study, EDTA whole blood samples were collected from 8 healthy volunteers, and different durations (up to 1 year) and temperatures (room temperature, 4°C, -20°C, and -80°C) of storage were compared. The effect of the addition of a DNA preservative agent was also assessed before and after storage. DNA concentrations measured by UV spectrophotometry and spectrofluorometry were used to calculate DNA extraction yields and double-strand DNA ratios. DNA integrity was controlled by agarose gel electrophoresis and long-range polymerase chain reaction. The impact of storage conditions on DNA methylation was also evaluated. Results showed that certain storage conditions have a significant impact on the DNA extraction yield but little or no effect on DNA integrity and methylation. Storage of EDTA blood at -80°C guarantees high-quality DNA with a good yield. Higher DNA extraction yields were obtained with the addition of a DNA preservative agent before thawing EDTA blood stored at -20°C or -80°C. Long-term storage at room temperature in the presence of a DNA preservative agent also appeared to be a reliable procedure.

Method optimization for fecal sample collection and fecal DNA extraction.

BACKGROUND: This is the third in a series of publications presenting formal method validation for biospecimen processing in the context of accreditation in laboratories and biobanks. We report here optimization of a stool processing protocol validated for fitness-for-purpose in terms of downstream DNA-based analyses. METHODS: Stool collection was initially optimized in terms of sample input quantity and supernatant volume using canine stool. Three DNA extraction methods (PerkinElmer MSM I®, Norgen Biotek All-In-One®, MoBio PowerMag®) and six collection container types were evaluated with human stool in terms of DNA quantity and quality, DNA yield, and its reproducibility by spectrophotometry, spectrofluorometry, and quantitative PCR, DNA purity, SPUD assay, and 16S rRNA gene sequence-based taxonomic signatures. RESULTS: The optimal MSM I protocol involves a 0.2 g stool sample and 1000 µL supernatant. The MSM I extraction was superior in terms of DNA quantity and quality when compared to the other two methods tested. Optimal results were obtained with plain Sarstedt tubes (without stabilizer, requiring immediate freezing and storage at -20°C or -80°C) and Genotek tubes (with stabilizer and RT storage) in terms of DNA yields (total, human, bacterial, and double-stranded) according to spectrophotometry and spectrofluorometry, with low yield variability and good DNA purity. No inhibitors were identified at 25 ng/µL. The protocol was reproducible in terms of DNA yield among different stool aliquots. CONCLUSIONS: We validated a stool collection method suitable for downstream DNA metagenomic analysis. DNA extraction with the MSM I method using Genotek tubes was considered optimal, with simple logistics in terms of collection and shipment and offers the possibility of automation. Laboratories and biobanks should ensure protocol conditions are systematically recorded in the scope of accreditation.