Multi-cohort modeling strategies for scalable globally accessible prostate cancer risk tools.

BACKGROUND: Online clinical risk prediction tools built on data from multiple cohorts are increasingly being utilized for contemporary doctor-patient decision-making and validation. This report outlines a comprehensive data science strategy for building such tools with application to the Prostate Biopsy Collaborative Group prostate cancer risk prediction tool. METHODS: We created models for high-grade prostate cancer risk using six established risk factors. The data comprised 8492 prostate biopsies collected from ten institutions, 2 in Europe and 8 across North America. We calculated area under the receiver operating characteristic curve (AUC) for discrimination, the Hosmer-Lemeshow test statistic (HLS) for calibration and the clinical net benefit at risk threshold 15%. We implemented several internal cross-validation schemes to assess the influence of modeling method and individual cohort on validation performance. RESULTS: High-grade disease prevalence ranged from 18% in Zurich (1863 biopsies) to 39% in UT Health San Antonio (899 biopsies). Visualization revealed outliers in terms of risk factors, including San Juan VA (51% abnormal digital rectal exam), Durham VA (63% African American), and Zurich (2.8% family history). Exclusion of any cohort did not significantly affect the AUC or HLS, nor did the choice of prediction model (pooled, random-effects, meta-analysis). Excluding the lowest-prevalence Zurich cohort from training sets did not statistically significantly change the validation metrics for any of the individual cohorts, except for Sunnybrook, where the effect on the AUC was minimal. Therefore the final multivariable logistic model was built by pooling the data from all cohorts using logistic regression. Higher prostate-specific antigen and age, abnormal digital rectal exam, African ancestry and a family history of prostate cancer increased risk of high-grade prostate cancer, while a history of a prior negative prostate biopsy decreased risk (all p-values < 0.004). CONCLUSIONS: We have outlined a multi-cohort model-building internal validation strategy for developing globally accessible and scalable risk prediction tools.

Genetic ancestry and prostate cancer susceptibility SNPs in Puerto Rican and African American men.

BACKGROUND: The Puerto Rican (PR) population is a racially admixed population that has a high prostate cancer (PCa) mortality rate. We hypothesized in this pilot study that West African Ancestry (WAA) was associated with PCa in this heterogeneous (PR) population. METHODS: A case/case and case/control study was performed. Controls, 207 African American (AA) and 133 PR were defined as men with no PCa, a serum PSA < 2.5 ng/mL and a negative rectal examination. Cases were patients with pathological specimens from radical prostatectomies (RP) (291 PR and 200 AA). DNA was extracted from whole blood of controls and from paraffin embedded normal seminal vesicle from the RPs. We assessed the association of PCa and aggressiveness with genetic ancestry using an ancestry informative marker panel (AIMs) and Wilcoxon rank-sum test and the association of PCa and aggressiveness with 15 previously PCa associated SNPs using Chi square test. Gleason Score (GS) and tumor stage (TS) were used to define low risk (GS ≤ 7[3 + 4]), TS ≤ pT2) and high risk (GS≥ 7[4 + 3], TS > pT2) PCa. Statistical analyses were done using SAS. RESULTS: No difference in overall percent WAA was found between PR cases and controls. Among PR or AA cases WAA was not associated with disease severity based upon risk group, Gleason score or stage. Among AA controls WAA was significantly higher than in cases. The SNP rs7824364 (chromosome 8q24) PCa risk allele was significantly increased among cases versus controls for both AA (P < 0.0001) and PR (P = 0.0001) men. PR men with ≥1 risk allele exhibited a higher percent of WAA (39% vs 29%, P = 0.034). CONCLUSION: The SNP rs7824364, a local marker of WAA in the 8q24 region was associated with PCa among both AA and PR men and with increased WAA among PR men. This novel relationship of PCA risk loci, WAA with PCa and its phenotype among PR men deserves further study.

Trends and outcomes of trauma patients positive to marijuana and cocaine.

PURPOSE: Substance misuse has long been recognized as a major predisposing risk factor for traumatic injury. However, there still exists no clear scientific consensus regarding the impact of drug use on patient outcomes. Therefore, this study aims to evaluate the demographic profile, hospital-course factors, and outcomes of trauma patients based on their toxicology. METHODS: This is a non-concurrent cohort study of 3709 patients treated at the Puerto Rico Trauma Hospital during 2002-2018. The sample was divided into four groups according to their toxicology status. Statistical techniques used included Pearson's chi-square test, Spearman correlation, and negative binomial and logistic regressions. RESULTS: Admission rates for marijuana (rho = 0.87) and marijuana and cocaine positive (rho = 0.68) patients increased. Positive toxicology patients underwent surgery more often than negative testing patients (marijuana: 68.7%, cocaine: 65.6%, marijuana & cocaine: 69.8%, negative: 57.0%). Among patients with non-penetrating injuries, a positive toxicology for cocaine or marijuana was linked to a 48% and 42% increased adjusted risk of complications, 37% and 27% longer TICU LOS, and 32% and 18% longer hospital LOS, respectively. CONCLUSION: Our results show an association between positive toxicology for either marijuana, cocaine, or both with higher need for surgery. Additionally, our results show an increase in complications, TICU LOS, and hospital LOS among non-penetrating trauma patients testing positive for marijuana or cocaine. Therefore, this study provides valuable information on the clinical profile of patients with positive toxicology, suggesting they might benefit from more aggressive management.

Recidivism at the puerto rico trauma hospital.

PURPOSE: Although trauma represents a leading cause of morbidity and mortality worldwide, there is limited and heterogeneous evidence regarding trauma recidivism and its outcomes. This analysis determined the rate and independent risk factors of trauma recidivism and compared the first and second injury episode among recidivists. METHODS: An IRB-approved retrospective cohort study was performed with data from the Puerto Rico Trauma Hospital Registry. Bivariate analyses were done using Pearson's Chi squared, Wilcoxon rank-sum, McNemar, Stuart-Maxwell or Wilcoxon signed-rank tests, as appropriate. Independent predictors for recidivism were determined through a logistic regression model. Statistical significance was set at p < 0.05. RESULTS: 24,650 patients were admitted to the hospital during 2000-2017. Recidivism rate was 14 per 1,000 patients discharged alive. Males and individuals aged 15-24 years old were 3.88 (95% CI: 2.21-6.80) and 3.80 (95% CI: 2.24-6.46) times more likely to be recidivists, respectively. Contrariwise, an ISS [Formula: see text] 25 [adjusted odds ratio (AOR) = 0.44; 95% CI: 0.28-0.68] and a GCS [Formula: see text] 8 (AOR = 0.56; 95% CI 0.34-0.92) were protective factors. Furthermore, recidivists exhibited less in-hospital mortality than their non-recidivist counterparts (7.2% vs. 10.7%; p = 0.045). For recidivists, the median (interquartile range) time to reinjury was 42 (59) months; and the second injury episode was more severe than the first one, as the proportion of patients with ISS [Formula: see text] 25 increased (7.9% vs. 14.1%; p = 0.022). CONCLUSION: The independent predictors of trauma recidivism and the median time to reinjury identified in this study provide valuable information to the development of prevention strategies aimed at reducing the burden of injury.

Epidemiology of trauma in patients with mental disorders.

Objectives: We aimed to explore the influence of mental disorders on the risk of developing complications and in-hospital mortality after trauma. Methods: We conducted an institutional review board-approved cohort study of 23 500 adult patients admitted to the Puerto Rico Trauma Hospital from 2002 to 2019. Participants were divided into 2 groups according to the presence or absence of psychiatric illnesses. Logistic regressions were employed to investigate the effect of mental illness on study outcomes. Results: Psychiatric illness was associated with higher risk of complications; this risk increased with age. The pattern was accentuated for those with substance use disorders (SUD) and attenuated for those with non-substance-related diagnoses (NSRD). Psychiatric patients with Glasgow Coma Scale (GCS) scores of 15 had a 42% (95% CI 1.17 to 1.73) higher risk of dying, while the opposite was seen for those with scores <15 (adjusted OR=0.79; 95% CI 0.64 to 0.99). SUD was associated with a 51% (95% CI 1.21 to 1.88) higher risk of death in patients with GCS scores of 15, while NSRD was linked to a 49% (95% CI 0.33 to 0.79) lower odds of death among subjects with scores <15. Conclusions: Our results suggest that trauma patients with SUD are at increased risk of developing complications and those with SUD and GCS scores of 15 are at increased risk of death. Mental health screening is an essential component of the management of trauma patients. Stratifying based on mental health disorders may be helpful during the clinical management of trauma patients, as those with SUD may benefit from more aggressive management. Level of evidence: Level 4, prognostic and epidemiological study. Study type: Original retrospective cohort study.

A Retrospective Cohort Study on Health Insurance: Related Disparities in Trauma Patients After Penetrating Injuries: 2000-2014.

OBJECTIVE: Although the lack of health insurance has been linked to poor health outcomes in several diseases, this relationship is still understudied in trauma. There exist differences between the Puerto Rico health care system and that of the United States. We therefore aimed to assess mortality disparities related to insurance coverage at the Puerto Rico Trauma Hospital (PRTH). METHODS: A retrospective cohort study of patients who sustained penetrating injuries (presenting at the PRTH from 2000 to 2014) was performed. Individuals were classified by their insurance status. Study variables comprised demographics, clinical characteristics and outcomes. A logistic regression analysis was performed to identify the association between health insurance status and risk of dying. RESULTS: Patients with public health insurance experienced more complications than did individuals who had private health insurance (PrHI) or who were uninsured. This group had longer durations of mechanical ventilation and spent more time in the hospital than did patients who had PrHI or who were uninsured. However, uninsured patients with gunshot wounds were 54% (adjusted odds ratio = 1.54; 95% CI: 1.01, 2.36) more likely to die than were their counterparts who had PrHI. CONCLUSION: Our study suggests that having health insurance could reduce a given patient mortality risk in trauma settings. More studies with larger samples are warranted to confirm these findings. If these findings hold true, then providing equitable access to health services for the entire population could prevent patients suffering trauma from having premature, preventable deaths.

Body mass index affects the diagnosis and progression of prostate cancer in Hispanics.

INTRODUCTION: The occurrence of lower prostate-specific antigen (PSA) levels in overweight White and African American men has been studied, but there is no data regarding Hispanics, which have a higher mortality rate from Prostate Cancer (PCa) than Whites. The objective of this study was to investigate if being overweight could affect both the sensitivity of the PSA as a diagnostic tool and the progression of PCa in this group. METHODS: Retrospective study of records from 400 patients that underwent testing for PCa during 2005 and 400 patients under treatment for PCa from 2003-2005 at the urology clinics of the Veterans Administration Caribbean Healthcare System. Accrued data included body mass index (BMI), age, PSA levels, biopsy results, and cancer status after treatment. RESULTS: In men, with normal age adjusted PSA levels, overweight and obese men had 35.38% and 38.13%, respectively, positive biopsies while men with normal BMI had 26.15%. In addition, 73.84% of overweight men over 61 years old with normal PSA were positive for prostate cancer. There is a statistically significant decrease in PSA sensitivity from 71.7 (95% CI: 58.6-82.5) in men with normal BMI to 55.4 (95% CI: 41.5-68.7) in obese men (P = .015). In multivariate analysis, patients with a BMI over 25 kg/m2 had a 2.63 (CI 95%: 1.23-5.64) fold higher risk of metastases than those with normal BMI. CONCLUSIONS: in overweight Hispanic men the PSA level is a less sensitive marker for PCa and those individuals with higher BMI have higher prevalence of metastatic disease.

Unveiling the Natural History of Category 4 Tropical Cyclones: The Puerto Rico Trauma Hospital Experience after Hurricane Maria.

OBJECTIVES: The hazardous environmental conditions hurricanes create might increase injury incidence almost 7 times. Therefore, a cohort study was performed at the Puerto Rico Trauma Hospital to compare morbidity and mortality patterns of patients after Hurricane Maria with a control period. METHODS: Admissions from September 20, 2017, through January 20, 2018, constituted the post-Maria period (473 patients); the corresponding months of the previous year comprised the pre-Maria period (439 patients). Comparisons were done using Pearson's chi-square or Mann-Whitney U-tests, as appropriate. A logistic regression was performed to assess the association between mortality and the study period. RESULTS: Postlandfall admissions among patients aged 40-64 y increased by 6.6%, while among subjects between ages 18 and 39 y dropped by 7.0% (P = 0.03). Falls, gunshots, and burns were the injury mechanisms that varied the most across the exposure period. The median Injury Severity Score (13 vs 12; P = 0.05) and the frequency of Glasgow Coma Scale scores ≤8 (17.1% vs 10.9%; P = 0.03) were higher among poststorm patients. Moreover, a 2-fold (odds ratio = 1.93; 95% CI: 1.07-3.47) increase in mortality was observed after Maria, when adjusting for covariates. CONCLUSIONS: Following a hurricane, trauma centers might expect an older population, with more severe injuries and a 2-fold increased mortality risk.

A Contemporary Prostate Biopsy Risk Calculator Based on Multiple Heterogeneous Cohorts.

BACKGROUND: Prostate cancer prediction tools provide quantitative guidance for doctor-patient decision-making regarding biopsy. The widely used online Prostate Cancer Prevention Trial Risk Calculator (PCPTRC) utilized data from the 1990s based on six-core biopsies and outdated grading systems. OBJECTIVE: We prospectively gathered data from men undergoing prostate biopsy in multiple diverse North American and European institutions participating in the Prostate Biopsy Collaborative Group (PBCG) in order to build a state-of-the-art risk prediction tool. DESIGN, SETTING, AND PARTICIPANTS: We obtained data from 15 611 men undergoing 16 369 prostate biopsies during 2006-2017 at eight North American institutions for model-building and three European institutions for validation. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: We used multinomial logistic regression to estimate the risks of high-grade prostate cancer (Gleason score ≥7) on biopsy based on clinical characteristics, including age, prostate-specific antigen, digital rectal exam, African ancestry, first-degree family history, and prior negative biopsy. We compared the PBCG model to the PCPTRC using internal cross-validation and external validation on the European cohorts. RESULTS AND LIMITATIONS: Cross-validation on the North American cohorts (5992 biopsies) yielded the PBCG model area under the receiver operating characteristic curve (AUC) as 75.5% (95% confidence interval: 74.2-76.8), a small improvement over the AUC of 72.3% (70.9-73.7) for the PCPTRC (p<0.0001). However, calibration and clinical net benefit were far superior for the PBCG model. Using a risk threshold of 10%, clinical use of the PBCG model would lead to the equivalent of 25 fewer biopsies per 1000 patients without missing any high-grade cancers. Results were similar on external validation on 10 377 European biopsies. CONCLUSIONS: The PBCG model should be used in place of the PCPTRC for prediction of prostate biopsy outcome. PATIENT SUMMARY: A contemporary risk tool for outcomes on prostate biopsy based on the routine clinical risk factors is now available for informed decision-making.