Risk of alcohol-associated liver disease in the offspring of parents with alcohol-associated liver disease: A nationwide cohort study.

BACKGROUND AND AIMS: Offspring of patients with alcohol-associated liver disease (ALD) may have a higher risk of ALD. We examined their risk of ALD and survival with ALD. APPROACH AND RESULTS: We used Danish nationwide registries to identify the offspring of patients diagnosed with ALD in 1996-2018 and 20:1 matched comparators from the general population. They were followed for ALD diagnosis through 2018. We used landmark competing risk analysis to estimate the age-specific absolute and relative 10-year risks of ALD. ALD was diagnosed in 385 of 60,707 offspring and 2842 of 1,213,357 comparators during 0.7 and 14.0 million person-years of follow-up, respectively, yielding an incidence rate ratio of 2.73 (95% CI: 2.44-3.03). The risk of being diagnosed with ALD within the next 10 years peaked at age 55 years for offspring and age 57 years for comparators with 10-year risks of 1.66% (95% CI: 1.16-2.30) in offspring and 0.81% (95% CI: 0.68-0.97) in comparators at these ages. Offspring were younger at ALD diagnosis than comparators (median age of 47.4 vs. 48.9 years), yet slightly more of them had developed cirrhosis (60.3% vs. 58.7%). Survival after ALD diagnosis was similar in offspring and comparators, adjusted hazard ratio=1.03 (95% CI: 0.88-1.21), so on average offspring died younger due to their younger age at diagnosis. CONCLUSIONS: Offspring of patients with ALD had a low but increased risk of ALD. Screening offspring for chronic liver disease may be unnecessary, but other interventions to mitigate alcohol-associated harm should be considered.

Lipidomic risk scores are independent of polygenic risk scores and can predict incidence of diabetes and cardiovascular disease in a large population cohort.

Type 2 diabetes (T2D) and cardiovascular disease (CVD) represent significant disease burdens for most societies and susceptibility to these diseases is strongly influenced by diet and lifestyle. Physiological changes associated with T2D or CVD, such has high blood pressure and cholesterol and glucose levels in the blood, are often apparent prior to disease incidence. Here we integrated genetics, lipidomics, and standard clinical diagnostics to assess future T2D and CVD risk for 4,067 participants from a large prospective population-based cohort, the Malmö Diet and Cancer-Cardiovascular Cohort. By training Ridge regression-based machine learning models on the measurements obtained at baseline when the individuals were healthy, we computed several risk scores for T2D and CVD incidence during up to 23 years of follow-up. We used these scores to stratify the participants into risk groups and found that a lipidomics risk score based on the quantification of 184 plasma lipid concentrations resulted in a 168% and 84% increase of the incidence rate in the highest risk group and a 77% and 53% decrease of the incidence rate in lowest risk group for T2D and CVD, respectively, compared to the average case rates of 13.8% and 22.0%. Notably, lipidomic risk correlated only marginally with polygenic risk, indicating that the lipidome and genetic variants may constitute largely independent risk factors for T2D and CVD. Risk stratification was further improved by adding standard clinical variables to the model, resulting in a case rate of 51.0% and 53.3% in the highest risk group for T2D and CVD, respectively. The participants in the highest risk group showed significantly altered lipidome compositions affecting 167 and 157 lipid species for T2D and CVD, respectively. Our results demonstrated that a subset of individuals at high risk for developing T2D or CVD can be identified years before disease incidence. The lipidomic risk, which is derived from only one single mass spectrometric measurement that is cheap and fast, is informative and could extend traditional risk assessment based on clinical assays.

Feasibility of risk assessment for breast cancer molecular subtypes.

PURPOSE: Few breast cancer risk assessment models account for the risk profiles of different tumor subtypes. This study evaluated whether a subtype-specific approach improves discrimination. METHODS: Among 3389 women who had a screening mammogram and were later diagnosed with invasive breast cancer we performed multinomial logistic regression with tumor subtype as the outcome and known breast cancer risk factors as predictors. Tumor subtypes were defined by expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) based on immunohistochemistry. Discrimination was assessed with the area under the receiver operating curve (AUC). Absolute risk of each subtype was estimated by proportioning Gail absolute risk estimates by the predicted probabilities for each subtype. We then compared risk factor distributions for women in the highest deciles of risk for each subtype. RESULTS: There were 3,073 ER/PR+ HER2 - , 340 ER/PR +HER2 + , 126 ER/PR-ER2+, and 300 triple-negative breast cancers (TNBC). Discrimination differed by subtype; ER/PR-HER2+ (AUC: 0.64, 95% CI 0.59, 0.69) and TNBC (AUC: 0.64, 95% CI 0.61, 0.68) had better discrimination than ER/PR+HER2+ (AUC: 0.61, 95% CI 0.58, 0.64). Compared to other subtypes, patients at high absolute risk of TNBC were younger, mostly Black, had no family history of breast cancer, and higher BMI. Those at high absolute risk of HER2+ cancers were younger and had lower BMI. CONCLUSION: Our study provides proof of concept that stratifying risk prediction for breast cancer subtypes may enable identification of patients with unique profiles conferring increased risk for tumor subtypes.

Clinically relevant cut-points for changes in the Liver Frailty Index are associated with waitlist mortality in patients with cirrhosis.

Physical frailty is a critical determinant of mortality in patients with cirrhosis and can be objectively measured using the Liver Frailty Index (LFI), which is potentially modifiable. We aimed to identify LFI cut-points associated with waitlist mortality. Ambulatory adults with cirrhosis without HCC awaiting liver transplantation from 9 centers from 2012 to 2021 for ≥3 months with ≥2 pre-liver transplantation LFI assessments were included. The primary explanatory variable was the change in LFI from first to second assessments per 3 months (∆LFI); we evaluated clinically relevant ∆LFI cut-points at 0.1, 0.2, 0.3, and 0.5. The primary outcome was waitlist mortality (death or delisting for being too sick), with transplant considered as a competing event. Among 1029 patients, the median (IQR) age was 58 (51-63) years; 42% were female; and the median lab Model for End-Stage Liver Disease-Sodium at first assessment was 18 (15-22). For each 0.1 improvement in ∆LFI, the risk of overall mortality decreased by 6% (cause-specific hazard ratio: 0.94, 95% CI: 0.92-0.97, p < 0.001). ∆LFI was associated with waitlist mortality at cut-points as low as 0.1 (cause-specific hazard ratio: 0.63, 95% CI: 0.46-0.87) and 0.2 (HR: 0.61, 95% CI: 0.42-0.87). An improvement in LFI per 3 months as small as 0.1 in the pre-liver transplantation period is associated with a clinically meaningful reduction in waitlist mortality. These data provide estimates of the reduction in mortality risk associated with improvements in LFI that can be used to assess the effectiveness of interventions targeting physical frailty in patients with cirrhosis.

Performance of risk prediction models for post-liver transplant patient and graft survival over time.

Given liver transplantation organ scarcity, selection of recipients and donors to maximize post-transplant benefit is paramount. Several scores predict post-transplant outcomes by isolating elements of donor and recipient risk, including the donor risk index, Balance of Risk, pre-allocation score to predict survival outcomes following liver transplantation/survival outcomes following liver transplantation (SOFT), improved donor-to-recipient allocation score for deceased donors only/improved donor-to-recipient allocation score for both deceased and living donors (ID2EAL-D/-DR), and survival benefit (SB) models. No studies have examined the performance of these models over time, which is critical in an ever-evolving transplant landscape. This was a retrospective cohort study of liver transplantation events in the UNOS database from 2002 to 2021. We used Cox regression to evaluate model discrimination (Harrell's C) and calibration (testing of calibration curves) for post-transplant patient and graft survival at specified post-transplant timepoints. Sub-analyses were performed in the modern transplant era (post-2014) and for key donor-recipient characteristics. A total of 112,357 transplants were included. The SB and SOFT scores had the highest discrimination for short-term patient and graft survival, including in the modern transplant era, where only the SB model had good discrimination (C ≥ 0.60) for all patient and graft outcome timepoints. However, these models had evidence of poor calibration at 3- and 5-year patient survival timepoints. The ID2EAL-DR score had lower discrimination but adequate calibration at all patient survival timepoints. In stratified analyses, SB and SOFT scores performed better in younger (< 40 y) and higher Model for End-Stage Liver Disease (≥ 25) patients. All prediction scores had declining discrimination over time, and scores relying on donor factors alone had poor performance. Although the SB and SOFT scores had the best overall performance, all models demonstrated declining performance over time. This underscores the importance of periodically updating and/or developing new prediction models to reflect the evolving transplant field. Scores relying on donor factors alone do not meaningfully inform post-transplant risk.

Rejoinder to the discussion on "Bayesian meta-analysis of penetrance for cancer risk".

The five discussions of our paper provide several modeling alternatives, extensions, and generalizations that can potentially guide future research in meta-analysis. In this rejoinder, we briefly summarize and comment on some of those points.

Absolute risk from double nested case-control designs: cause-specific proportional hazards models with and without augmented estimating equations.

We estimate relative hazards and absolute risks (or cumulative incidence or crude risk) under cause-specific proportional hazards models for competing risks from double nested case-control (DNCC) data. In the DNCC design, controls are time-matched not only to cases from the cause of primary interest, but also to cases from competing risks (the phase-two sample). Complete covariate data are available in the phase-two sample, but other cohort members only have information on survival outcomes and some covariates. Design-weighted estimators use inverse sampling probabilities computed from Samuelsen-type calculations for DNCC. To take advantage of additional information available on all cohort members, we augment the estimating equations with a term that is unbiased for zero but improves the efficiency of estimates from the cause-specific proportional hazards model. We establish the asymptotic properties of the proposed estimators, including the estimator of absolute risk, and derive consistent variance estimators. We show that augmented design-weighted estimators are more efficient than design-weighted estimators. Through simulations, we show that the proposed asymptotic methods yield nominal operating characteristics in practical sample sizes. We illustrate the methods using prostate cancer mortality data from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Study of the National Cancer Institute.

Bayesian meta-analysis of penetrance for cancer risk.

Multi-gene panel testing allows many cancer susceptibility genes to be tested quickly at a lower cost making such testing accessible to a broader population. Thus, more patients carrying pathogenic germline mutations in various cancer-susceptibility genes are being identified. This creates a great opportunity, as well as an urgent need, to counsel these patients about appropriate risk-reducing management strategies. Counseling hinges on accurate estimates of age-specific risks of developing various cancers associated with mutations in a specific gene, ie, penetrance estimation. We propose a meta-analysis approach based on a Bayesian hierarchical random-effects model to obtain penetrance estimates by integrating studies reporting different types of risk measures (eg, penetrance, relative risk, odds ratio) while accounting for the associated uncertainties. After estimating posterior distributions of the parameters via a Markov chain Monte Carlo algorithm, we estimate penetrance and credible intervals. We investigate the proposed method and compare with an existing approach via simulations based on studies reporting risks for two moderate-risk breast cancer susceptibility genes, ATM and PALB2. Our proposed method is far superior in terms of coverage probability of credible intervals and mean square error of estimates. Finally, we apply our method to estimate the penetrance of breast cancer among carriers of pathogenic mutations in the ATM gene.

Calibration plots for multistate risk predictions models.

INTRODUCTION: There is currently no guidance on how to assess the calibration of multistate models used for risk prediction. We introduce several techniques that can be used to produce calibration plots for the transition probabilities of a multistate model, before assessing their performance in the presence of random and independent censoring through a simulation. METHODS: We studied pseudo-values based on the Aalen-Johansen estimator, binary logistic regression with inverse probability of censoring weights (BLR-IPCW), and multinomial logistic regression with inverse probability of censoring weights (MLR-IPCW). The MLR-IPCW approach results in a calibration scatter plot, providing extra insight about the calibration. We simulated data with varying levels of censoring and evaluated the ability of each method to estimate the calibration curve for a set of predicted transition probabilities. We also developed evaluated the calibration of a model predicting the incidence of cardiovascular disease, type 2 diabetes and chronic kidney disease among a cohort of patients derived from linked primary and secondary healthcare records. RESULTS: The pseudo-value, BLR-IPCW, and MLR-IPCW approaches give unbiased estimates of the calibration curves under random censoring. These methods remained predominately unbiased in the presence of independent censoring, even if the censoring mechanism was strongly associated with the outcome, with bias concentrated in low-density regions of predicted transition probability. CONCLUSIONS: We recommend implementing either the pseudo-value or BLR-IPCW approaches to produce a calibration curve, combined with the MLR-IPCW approach to produce a calibration scatter plot. The methods have been incorporated into the "calibmsm" R package available on CRAN.

Validation of Two Scoring Tools to Predict Risk of Augmented Renal Clearance in Trauma Patients.

INTRODUCTION: Augmented renal clearance (ARC) is prevalent in trauma populations. Identification is underrecognized by calculated creatinine clearance or estimated glomerular filtration rate equations. Predictive scores may assist with ARC identification. The goal of this study was to evaluate validity of the ARCTIC score and ARC Predictor to predict ARC in critically ill trauma patients. METHODS: This single center, retrospective study was performed at an academic level 1 trauma center. Critically ill adult trauma patients undergoing 24-h urine-collection were included. Patients with serum creatinine >1.5 mg/dL, kidney replacement therapy, suspected rhabdomyolysis, chronic kidney disease, or inaccurate urine collection were excluded. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for ARCTIC Score and ARC Predictor were calculated. Receiver operating characteristic curves were created for ARCTIC score and ARC Predictor models. RESULTS: One-hundred and twenty-two patients with ARC and 78 patients without ARC were included. The ARCTIC score sensitivity, specificity, PPV, and NPV were 89%, 54%, 75%, and 75%, respectively. The ARC Predictor demonstrated sensitivity, specificity, PPV, and NPV of 77%, 88%, 91%, and 71%, respectively. Regression analyses revealed both ARCTIC score ≥6 and ARC Predictor threshold >0.5 as significant risk factors for ARC in presence of traumatic brain injury, obesity, injury severity score, and negative nitrogen balance (ARCTIC ≥6: odds ratio 8.59 [95% confidence interval 3.90-18.92], P < 0.001; ARC Predictor >0.5: odds ratio 20.07 [95% confidence interval 8.53-47.19], P < 0.001). CONCLUSIONS: These findings corroborate validity of two pragmatic prediction tools to identify patients at high risk of ARC. Future studies evaluating correlations between ARCTIC score, ARC Predictor, and clinical outcomes are warranted.

Impact of Operative Approach on Acute Kidney Injury Risk Prediction Models for Colectomy.

INTRODUCTION: Acute kidney injury (AKI) is a serious postoperative complication associated with increased morbidity and mortality. Identifying patients at risk for AKI is important for risk stratification and management. This study aimed to develop an AKI risk prediction model for colectomy and determine if the operative approach (laparoscopic versus open) alters the influence of predictive factors through an interaction term analysis. METHODS: The American College of Surgeons National Surgical Quality Improvement Program database was analyzed from 2005 to 2019. Patients undergoing laparoscopic and open colectomy were identified and propensity score matched. Multivariable logistic regression identified significant preoperative demographic, comorbidity, and laboratory value predictors of AKI. The predictive ability of a baseline model consisting of these variables was compared to a proposed model incorporating interaction terms between operative approach and predictor variables using the likelihood ratio test, c-statistic, and Brier score. Shapley Additive Explanations values assessed relative importance of significant predictors. RESULTS: 252,372 patients were included in the analysis. Significant AKI predictors were hypertension, age, sex, race, body mass index, smoking, diabetes, preoperative sepsis, Congestive heart failure, preoperative creatinine, preoperative albumin, and operative approach (P < 0.001). The proposed model with interaction terms had improved predictive ability per the likelihood ratio test (P < 0.05) but had no statistically significant interaction terms. C-statistic and Brier scores did not improve. Shapley Additive Explanations analysis showed hypertension had the highest importance. The importance of age and diabetes showed some variation between operative approaches. CONCLUSIONS: While the inclusion of interaction terms collectively improved AKI prediction, no individual operative approach interaction terms were significant. Including operative approach interactions may enhance predictive ability of AKI risk models for colectomy.

Geriatric Nutritional Risk Index and 30-Day Postoperative Mortality in Geriatric Burn Patients.

INTRODUCTION: The geriatric nutritional risk index (GNRI) can easily identify malnutrition-associated morbidity and mortality. We investigated the association between preoperative GNRI and 30-d mortality in geriatric burn patients who underwent surgery. METHODS: The study involved geriatric burn patients (aged ≥ 65 y) who underwent burn surgery between 2012 and 2022. The GNRI was computed using the following formula: 1.489 × serum albumin concentration (mg/L) + 41.7 × patient body weight/ideal body weight. Patients were dichotomized into the high GNRI (≥ 82) and low GNRI (< 82) groups. GNRI was evaluated as an independent predictor of 30-d postoperative mortality. The study also evaluated the association between GNRI and sepsis, the need for continuous renal replacement therapy (CRRT), major adverse cardiac events (MACE), and pneumonia. RESULTS: Out of 270 patients, 128 (47.4%) had low GNRI (< 82). Multivariate Cox regression analysis revealed that low GNRI was significantly associated with 30-d postoperative mortality (hazard ratio: 1.874, 95% confidence interval [CI]: 1.146-3.066, P = 0.001). Kaplan-Meier analysis revealed that the 30-day mortality rate differed significantly between the low and high GNRI groups (log-rank test, P < 0.001). The 30-d postoperative mortality (hazard ratio: 2.677, 95% CI: 1.536-4.667, P < 0.001) and the incidence of sepsis (odds ratio [OR]: 2.137, 95% CI: 1.307-3.494, P = 0.004), need for CRRT (OR: 1.919, 95% CI: 1.101-3.344, P = 0.025), MACE (OR: 1.680, 95% CI: 1.018-2.773, P = 0.043), and pneumonia (OR: 1.678, 95% CI: 1.019-2.764, P = 0.044), were significantly higher in the low GNRI group than in the high GNRI group. CONCLUSIONS: Preoperative low GNRI was associated with increased 30-d postoperative mortality, sepsis, need for CRRT, MACE, and pneumonia in geriatric burn patients.

Non-White Patients Have a Higher Risk of Stroke Following Transcarotid Artery Revascularization.

INTRODUCTION: Carotid artery revascularization has traditionally been performed by either a carotid endarterectomy or carotid artery stent. Large data analysis has suggested there are differences in perioperative outcomes with regards to race, with non-White patients (NWP) having worse outcomes of stroke, restenosis and return to the operating room (RTOR). The introduction of transcarotid artery revascularization (TCAR) has started to shift the paradigm of carotid disease treatment. However, to date, there have been no studies assessing the difference in postoperative outcomes after TCAR between racial groups. METHODS: All patients from 2016 to 2021 in the Vascular Quality Initiative who underwent TCAR were included in our analysis. Patients were split into two groups based on race: individuals who identified as White and a second group that comprised all other races. Demographic and clinical variables were compared using Student's t-Test and chi-square test of independence. Logistic regression analysis was performed to determine the impact of race on perioperative outcomes of stroke, myocardial infarction (MI), death, restenosis, RTOR, and transient ischemic attack (TIA). RESULTS: The cohort consisted of 22,609 patients: 20,424 (90.3%) White patients and 2185 (9.7%) NWP. After adjusting for sex, diabetes, hypertension, coronary artery disease, history of prior stroke or TIA, symptomatic status, and high-risk criteria at time of TCAR, there was a significant difference in postoperative stroke, with 63% increased risk in NWP (odds ratio = 1.63, 95% confidence interval: 1.11-2.40, P = 0.014). However, we found no significant difference in the odds of MI, death, postoperative TIA, restenosis, or RTOR when comparing NWP to White patients. CONCLUSIONS: This study demonstrates that NWP have increased risk of stroke but similar outcomes of death, MI, RTOR and restenosis following TCAR. Future studies are needed to elucidate and address the underlying causes of racial disparity in carotid revascularization.

Shifting Perspectives: Equal Blunt Cerebrovascular Risk in Low-Versus High-Energy Cervical Fracture.

INTRODUCTION: Clinical implications of screening for blunt cerebrovascular injury (BCVI) after low-energy mechanisms of injury (LEMI) remain unclear. We assessed BCVI incidence and outcomes in LEMI versus high-energy mechanisms of injury (HEMI) patients. METHODS: In this retrospective cohort study, blunt trauma adults admitted between July 2015 and June 2021 with cervical spine fractures, excluding single spinous process, osteophyte, and chronic fractures were included. Demographics, comorbidities, injuries, screening and treatment data, iatrogenic complications, and mortality were collected. Our primary end point was to compare BCVI rates between LEMI and HEMI patients. RESULTS: Eight hundred sixty patients (78%) were screened for BCVI; 120 were positive for BCVI. LEMI and HEMI groups presented similar BCVI rates (12.6% versus 14.4%; P = 0.640). Compared to HEMI patients (n = 95), LEMI patients (n = 25) were significantly older (79 ± 14.9 versus 54.3 ± 17.4, P < 0.001), more likely to be on anticoagulants before admission (64% versus 23.2%, P < 0.001), and less severely injured (LEMI injury severity score 10.9 ± 6.6 versus HEMI injury severity score 18.7 ± 11.4, P = 0.001). All but one LEMI and 90.5% of the HEMI patients had vertebral artery injuries with no significant difference in BCVI grades. One HEMI patient developed acute kidney injury because of BCVI screening. Eleven HEMI patients developed BCVI-related stroke with two related mortalities. One LEMI patient died of a BCVI-related stroke. CONCLUSIONS: BCVI rates were similar between HEMI and LEMI groups when screening based on cervical spine fractures. The LEMI group exhibited no screening or treatment complications, suggesting that benefits may outweigh the risks of screening and potential bleeding complications from treatment.

An evaluation of sample size requirements for developing risk prediction models with binary outcomes.

BACKGROUND: Risk prediction models are routinely used to assist in clinical decision making. A small sample size for model development can compromise model performance when the model is applied to new patients. For binary outcomes, the calibration slope (CS) and the mean absolute prediction error (MAPE) are two key measures on which sample size calculations for the development of risk models have been based. CS quantifies the degree of model overfitting while MAPE assesses the accuracy of individual predictions. METHODS: Recently, two formulae were proposed to calculate the sample size required, given anticipated features of the development data such as the outcome prevalence and c-statistic, to ensure that the expectation of the CS and MAPE (over repeated samples) in models fitted using MLE will meet prespecified target values. In this article, we use a simulation study to evaluate the performance of these formulae. RESULTS: We found that both formulae work reasonably well when the anticipated model strength is not too high (c-statistic < 0.8), regardless of the outcome prevalence. However, for higher model strengths the CS formula underestimates the sample size substantially. For example, for c-statistic = 0.85 and 0.9, the sample size needed to be increased by at least 50% and 100%, respectively, to meet the target expected CS. On the other hand, the MAPE formula tends to overestimate the sample size for high model strengths. These conclusions were more pronounced for higher prevalence than for lower prevalence. Similar results were drawn when the outcome was time to event with censoring. Given these findings, we propose a simulation-based approach, implemented in the new R package 'samplesizedev', to correctly estimate the sample size even for high model strengths. The software can also calculate the variability in CS and MAPE, thus allowing for assessment of model stability. CONCLUSIONS: The calibration and MAPE formulae suggest sample sizes that are generally appropriate for use when the model strength is not too high. However, they tend to be biased for higher model strengths, which are not uncommon in clinical risk prediction studies. On those occasions, our proposed adjustments to the sample size calculations will be relevant.

Assessing treatment effects with adjusted restricted mean time lost in observational competing risks data.

BACKGROUND: According to long-term follow-up data of malignant tumor patients, assessing treatment effects requires careful consideration of competing risks. The commonly used cause-specific hazard ratio (CHR) and sub-distribution hazard ratio (SHR) are relative indicators and may present challenges in terms of proportional hazards assumption and clinical interpretation. Recently, the restricted mean time lost (RMTL) has been recommended as a supplementary measure for better clinical interpretation. Moreover, for observational study data in epidemiological and clinical settings, due to the influence of confounding factors, covariate adjustment is crucial for determining the causal effect of treatment. METHODS: We construct an RMTL estimator after adjusting for covariates based on the inverse probability weighting method, and derive the variance to construct interval estimates based on the large sample properties. We use simulation studies to study the statistical performance of this estimator in various scenarios. In addition, we further consider the changes in treatment effects over time, constructing a dynamic RMTL difference curve and corresponding confidence bands for the curve. RESULTS: The simulation results demonstrate that the adjusted RMTL estimator exhibits smaller biases compared with unadjusted RMTL and provides robust interval estimates in all scenarios. This method was applied to a real-world cervical cancer patient data, revealing improvements in the prognosis of patients with small cell carcinoma of the cervix. The results showed that the protective effect of surgery was significant only in the first 20 months, but the long-term effect was not obvious. Radiotherapy significantly improved patient outcomes during the follow-up period from 17 to 57 months, while radiotherapy combined with chemotherapy significantly improved patient outcomes throughout the entire period. CONCLUSIONS: We propose the approach that is easy to interpret and implement for assessing treatment effects in observational competing risk data.

jmBIG: enhancing dynamic risk prediction and personalized medicine through joint modeling of longitudinal and survival data in big routinely collected data.

We have introduced the R package jmBIG to facilitate the analysis of large healthcare datasets and the development of predictive models. This package provides a comprehensive set of tools and functions specifically designed for the joint modelling of longitudinal and survival data in the context of big data analytics. The jmBIG package offers efficient and scalable implementations of joint modelling algorithms, allowing for integrating large-scale healthcare datasets.By utilizing the capabilities of jmBIG, researchers and analysts can effectively handle the challenges associated with big healthcare data, such as high dimensionality and complex relationships between multiple outcomes.With the support of jmBIG, analysts can seamlessly fit Bayesian joint models, generate predictions, and evaluate the performance of the models. The package incorporates cutting-edge methodologies and harnesses the computational capabilities of parallel computing to accelerate the analysis of large-scale healthcare datasets significantly. In summary, jmBIG empowers researchers to gain deeper insights into disease progression and treatment response, fostering evidence-based decision-making and paving the way for personalized healthcare interventions that can positively impact patient outcomes on a larger scale.

Outcome risk model development for heterogeneity of treatment effect analyses: a comparison of non-parametric machine learning methods and semi-parametric statistical methods.

BACKGROUND: In randomized clinical trials, treatment effects may vary, and this possibility is referred to as heterogeneity of treatment effect (HTE). One way to quantify HTE is to partition participants into subgroups based on individual's risk of experiencing an outcome, then measuring treatment effect by subgroup. Given the limited availability of externally validated outcome risk prediction models, internal models (created using the same dataset in which heterogeneity of treatment analyses also will be performed) are commonly developed for subgroup identification. We aim to compare different methods for generating internally developed outcome risk prediction models for subject partitioning in HTE analysis. METHODS: Three approaches were selected for generating subgroups for the 2,441 participants from the United States enrolled in the ASPirin in Reducing Events in the Elderly (ASPREE) randomized controlled trial. An extant proportional hazards-based outcomes predictive risk model developed on the overall ASPREE cohort of 19,114 participants was identified and was used to partition United States' participants by risk of experiencing a composite outcome of death, dementia, or persistent physical disability. Next, two supervised non-parametric machine learning outcome classifiers, decision trees and random forests, were used to develop multivariable risk prediction models and partition participants into subgroups with varied risks of experiencing the composite outcome. Then, we assessed how the partitioning from the proportional hazard model compared to those generated by the machine learning models in an HTE analysis of the 5-year absolute risk reduction (ARR) and hazard ratio for aspirin vs. placebo in each subgroup. Cochran's Q test was used to detect if ARR varied significantly by subgroup. RESULTS: The proportional hazard model was used to generate 5 subgroups using the quintiles of the estimated risk scores; the decision tree model was used to generate 6 subgroups (6 automatically determined tree leaves); and the random forest model was used to generate 5 subgroups using the quintiles of the prediction probability as risk scores. Using the semi-parametric proportional hazards model, the ARR at 5 years was 15.1% (95% CI 4.0-26.3%) for participants with the highest 20% of predicted risk. Using the random forest model, the ARR at 5 years was 13.7% (95% CI 3.1-24.4%) for participants with the highest 20% of predicted risk. The highest outcome risk group in the decision tree model also exhibited a risk reduction, but the confidence interval was wider (5-year ARR = 17.0%, 95% CI= -5.4-39.4%). Cochran's Q test indicated ARR varied significantly only by subgroups created using the proportional hazards model. The hazard ratio for aspirin vs. placebo therapy did not significantly vary by subgroup in any of the models. The highest risk groups for the proportional hazards model and random forest model contained 230 participants each, while the highest risk group in the decision tree model contained 41 participants. CONCLUSIONS: The choice of technique for internally developed models for outcome risk subgroups influences HTE analyses. The rationale for the use of a particular subgroup determination model in HTE analyses needs to be explicitly defined based on desired levels of explainability (with features importance), uncertainty of prediction, chances of overfitting, and assumptions regarding the underlying data structure. Replication of these analyses using data from other mid-size clinical trials may help to establish guidance for selecting an outcomes risk prediction modelling technique for HTE analyses.

Handling missing data and measurement error for early-onset myopia risk prediction models.

BACKGROUND: Early identification of children at high risk of developing myopia is essential to prevent myopia progression by introducing timely interventions. However, missing data and measurement error (ME) are common challenges in risk prediction modelling that can introduce bias in myopia prediction. METHODS: We explore four imputation methods to address missing data and ME: single imputation (SI), multiple imputation under missing at random (MI-MAR), multiple imputation with calibration procedure (MI-ME), and multiple imputation under missing not at random (MI-MNAR). We compare four machine-learning models (Decision Tree, Naive Bayes, Random Forest, and Xgboost) and three statistical models (logistic regression, stepwise logistic regression, and least absolute shrinkage and selection operator logistic regression) in myopia risk prediction. We apply these models to the Shanghai Jinshan Myopia Cohort Study and also conduct a simulation study to investigate the impact of missing mechanisms, the degree of ME, and the importance of predictors on model performance. Model performance is evaluated using the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). RESULTS: Our findings indicate that in scenarios with missing data and ME, using MI-ME in combination with logistic regression yields the best prediction results. In scenarios without ME, employing MI-MAR to handle missing data outperforms SI regardless of the missing mechanisms. When ME has a greater impact on prediction than missing data, the relative advantage of MI-MAR diminishes, and MI-ME becomes more superior. Furthermore, our results demonstrate that statistical models exhibit better prediction performance than machine-learning models. CONCLUSION: MI-ME emerges as a reliable method for handling missing data and ME in important predictors for early-onset myopia risk prediction.

Dupilumab therapy for atopic dermatitis is associated with increased risk of cutaneous T cell lymphoma: A retrospective cohort study.

BACKGROUND: Dupilumab, a human monoclonal antibody targeting the interleukin 4 alpha receptor, is used for treatment of moderate to severe atopic dermatitis (AD). Previous studies have reported diagnoses of cutaneous T cell lymphoma (CTCL) after dupilumab use. OBJECTIVE: Investigate the risk of CTCL after dupilumab use in patients with AD. METHODS: Using the TrinetX database, incidence of cutaneous and lymphoid malignancies including CTCL was compared between a cohort of patients with AD who used dupilumab and a cohort of patients with AD who never used dupilumab. A second analysis excluding prior disease-modifying antirheumatic drug use was performed. Propensity score matching was performed to control for covariates. RESULTS: An increased risk of CTCL was found in the cohort of AD patients who used dupilumab (odds ratio 4.1003, 95% confidence interval 2.055-8.192). The increased risk persisted after exclusion of prior disease-modifying antirheumatic drug use. Risk was not increased for other cutaneous or lymphoid malignancies. Most (27/41) cases of CTCL were diagnosed more than 1 year after dupilumab use. LIMITATIONS: There is potential for misclassification in the database. Severity of AD could not be assessed. Association between dupilumab and CTCL does not prove causality. CONCLUSION: Dupilumab use is associated with an increased risk of CTCL in patients with AD in this cohort.