Bayesian joint modeling of multivariate longitudinal and survival outcomes using Gaussian copulas.

There is an increasing interest in the use of joint models for the analysis of longitudinal and survival data. While random effects models have been extensively studied, these models can be hard to implement and the fixed effect regression parameters must be interpreted conditional on the random effects. Copulas provide a useful alternative framework for joint modeling. One advantage of using copulas is that practitioners can directly specify marginal models for the outcomes of interest. We develop a joint model using a Gaussian copula to characterize the association between multivariate longitudinal and survival outcomes. Rather than using an unstructured correlation matrix in the copula model to characterize dependence structure as is common, we propose a novel decomposition that allows practitioners to impose structure (e.g., auto-regressive) which provides efficiency gains in small to moderate sample sizes and reduces computational complexity. We develop a Markov chain Monte Carlo model fitting procedure for estimation. We illustrate the method's value using a simulation study and present a real data analysis of longitudinal quality of life and disease-free survival data from an International Breast Cancer Study Group trial.

Adaptive risk-aware sharable and individual subspace learning for cancer survival analysis with multi-modality data.

Biomedical multi-modality data (also named multi-omics data) refer to data that span different types and derive from multiple sources in clinical practices (e.g. gene sequences, proteomics and histopathological images), which can provide comprehensive perspectives for cancers and generally improve the performance of survival models. However, the performance improvement of multi-modality survival models may be hindered by two key issues as follows: (1) how to learn and fuse modality-sharable and modality-individual representations from multi-modality data; (2) how to explore the potential risk-aware characteristics in each risk subgroup, which is beneficial to risk stratification and prognosis evaluation. Additionally, learning-based survival models generally refer to numerous hyper-parameters, which requires time-consuming parameter setting and might result in a suboptimal solution. In this paper, we propose an adaptive risk-aware sharable and individual subspace learning method for cancer survival analysis. The proposed method jointly learns sharable and individual subspaces from multi-modality data, whereas two auxiliary terms (i.e. intra-modality complementarity and inter-modality incoherence) are developed to preserve the complementary and distinctive properties of each modality. Moreover, it equips with a grouping co-expression constraint for obtaining risk-aware representation and preserving local consistency. Furthermore, an adaptive-weighted strategy is employed to efficiently estimate crucial parameters during the training stage. Experimental results on three public datasets demonstrate the superiority of our proposed model.

Development and validation of novel interpretable survival prediction models based on drug exposures for severe heart failure during vulnerable period.

BACKGROUND: Severe heart failure (HF) has a higher mortality during vulnerable period while targeted predictive tools, especially based on drug exposures, to accurately assess its prognoses remain largely unexplored. Therefore, this study aimed to utilize drug information as the main predictor to develop and validate survival models for severe HF patients during this period. METHODS: We extracted severe HF patients from the MIMIC-IV database (as training and internal validation cohorts) as well as from the MIMIC-III database and local hospital (as external validation cohorts). Three algorithms, including Cox proportional hazards model (CoxPH), random survival forest (RSF), and deep learning survival prediction (DeepSurv), were applied to incorporate the parameters (partial hospitalization information and exposure durations of drugs) for constructing survival prediction models. The model performance was assessed mainly using area under the receiver operator characteristic curve (AUC), brier score (BS), and decision curve analysis (DCA). The model interpretability was determined by the permutation importance and Shapley additive explanations values. RESULTS: A total of 11,590 patients were included in this study. Among the 3 models, the CoxPH model ultimately included 10 variables, while RSF and DeepSurv models incorporated 24 variables, respectively. All of the 3 models achieved respectable performance metrics while the DeepSurv model exhibited the highest AUC values and relatively lower BS among these models. The DCA also verified that the DeepSurv model had the best clinical practicality. CONCLUSIONS: The survival prediction tools established in this study can be applied to severe HF patients during vulnerable period by mainly inputting drug treatment duration, thus contributing to optimal clinical decisions prospectively.

Upper gastrointestinal haemorrhage patients' survival: A causal inference and prediction study.

BACKGROUND: Upper gastrointestinal (GI) bleeding is a common medical emergency. This study aimed to develop models to predict critically ill patients with upper GI bleeding in-hospital and 30-day survival, identify the correlation factor and infer the causality. METHODS: A total of 2898 patients with upper GI bleeding were included from the Medical Information Mart for Intensive Care-IV and eICU-Collaborative Research Database, respectively. To identify the most critical factors contributing to the prognostic model, we used SHAP (SHapley Additive exPlanations) for machine learning interpretability. We performed causal inference using inverse probability weighting for survival-associated prognostic factors. RESULTS: The optimal model using the light GBM (gradient boosting algorithm) algorithm achieved an AUC of .93 for in-hospital survival, .81 for 30-day survival in internal testing and .87 for in-hospital survival in external testing. Important factors for in-hospital survival, according to SHAP, were SOFA (Sequential organ failure assessment score), GCS (Glasgow coma scale) motor score and length of stay in ICU (Intensive critical care). In contrast, essential factors for 30-day survival were SOFA, length of stay in ICU, total bilirubin and GCS verbal score. Our model showed improved performance compared to SOFA alone. CONCLUSIONS: Our interpretable machine learning model for predicting in-hospital and 30-day mortality in critically ill patients with upper gastrointestinal bleeding showed excellent accuracy and high generalizability. This model can assist clinicians in managing these patients to improve the discrimination of high-risk patients.

Comparison of partitioned survival modeling with state transition modeling approaches with or without consideration of brain metastasis: a case study of Osimertinib versus pemetrexed-platinum.

BACKGROUND: The partitioned survival model (PSM) and the state transition model (STM) are widely used in cost-effectiveness analyses of anticancer drugs. Using different modeling approaches with or without consideration of brain metastasis, we compared the quality-adjusted life-year (QALY) estimates of Osimertinib and pemetrexed-platinum in advanced non-small cell lung cancer with epidermal growth factor receptor mutations. METHODS: We constructed three economic models using parametric curves fitted to patient-level data from the National Health Insurance Review and Assessment claims database from 2009 to 2020. PSM and 3-health state transition model (3-STM) consist of three health states: progression-free, post-progression, and death. The 5-health state transition model (5-STM) has two additional health states (brain metastasis with continuing initial therapy, and with subsequent therapy). Time-dependent transition probabilities were calculated in the state transition models. The incremental life-year (LY) and QALY between the Osimertinib and pemetrexed-platinum cohorts for each modeling approach were estimated over seven years. RESULTS: The PSM and 3-STM produced similar incremental LY (0.889 and 0.899, respectively) and QALY (0.827 and 0.840, respectively). However, 5-STM, which considered brain metastasis as separate health states, yielded a slightly higher incremental LY (0.910) but lower incremental QALY (0.695) than PSM and 3-STM. CONCLUSIONS: Our findings indicate that incorporating additional health states such as brain metastases into economic models can have a considerable impact on incremental QALY estimates. To ensure appropriate health technology assessment decisions, comparison and justification of different modeling approaches are recommended in the economic evaluation of anticancer drugs.

Fertility-sparing surgery vs standard surgery for early-stage cervical cancer: difference in 5-year life expectancy by tumor size.

BACKGROUND: Cervical cancer incidence among premenopausal women is rising, and fertility-sparing surgery serves as an important option for this young population. There is a lack of evidence on what tumor size cutoff should be used to define candidacy for fertility-sparing surgery. OBJECTIVE: We sought to describe how the association between fertility-sparing surgery (compared with standard surgery) and life expectancy varies by tumor size among patients with cervical cancers measuring ≤4 cm in largest diameter. Our secondary objective was to quantify the probability of undergoing adjuvant radiotherapy among patients who underwent fertility-sparing surgery as a function of tumor size. STUDY DESIGN: We identified patients in the National Cancer Database aged ≤45 years, diagnosed with stage I cervical cancer with tumors ≤4 cm between 2006 and 2018, who received no preoperative radiation or chemotherapy, and who underwent either fertility-sparing surgery (cone or trachelectomy, either simple or radical) or standard surgery (simple or radical hysterectomy) as their primary treatment. Propensity-score matching was performed to compare patients who underwent fertility-sparing surgery with those who underwent standard surgery. A flexible parametric model was employed to quantify the difference in life expectancy within 5 years of diagnosis (restricted mean survival time) based on tumor size among patients who underwent fertility-sparing and those who underwent standard surgery. In addition, among those who underwent fertility-sparing surgery, a logistic regression model was used to explore the relationship between tumor size and the probability of receiving adjuvant radiation. RESULTS: A total of 11,946 patients met the inclusion criteria of whom 904 (7.6%) underwent fertility-sparing surgery. After propensity-score matching, 897 patients who underwent fertility-sparing surgery were matched 1:1 with those who underwent standard surgery. Although the 5-year life expectancy was similar among patients who had fertility sparing surgery and those who had standard surgery regardless of tumor sizes, the estimates of life-expectancy differences associated with fertility-sparing surgery were more precise among patients with smaller tumors (1-cm tumor: restricted mean survival time difference, -0.10 months; 95% confidence interval, -0.67 to 0.47) than among those with larger tumors (4-cm tumor: restricted mean survival time difference, -0.11 months; 95% confidence interval, -3.79 to 3.57). The probability of receiving adjuvant radiation increased with tumor size, ranging from 5.6% (95% confidence interval, 3.9-7.9) for a 1-cm tumor to 37% (95% confidence interval, 24.3-51.8) for a 4-cm tumor. CONCLUSION: Within 5 years of diagnosis, young patients with stage I cancers measuring ≤4 cm had similar survival outcomes after either fertility-sparing surgery or standard surgery. However, because few patients with tumors >2 cm underwent fertility-sparing surgery, a clinically important survival difference could not be excluded in this population.

Maternal vascular indices at 36 weeks' gestation in the prediction of preeclampsia.

BACKGROUND: Epidemiological studies have shown that women with preeclampsia (PE) are at increased long term cardiovascular risk. This risk might be associated with accelerated vascular ageing process but data on vascular abnormalities in women with PE are scarce. OBJECTIVE: This study aimed to identify the most discriminatory maternal vascular index in the prediction of PE at 35 to 37 weeks' gestation and to examine the performance of screening for PE by combinations of maternal risk factors and biophysical and biochemical markers at 35 to 37 weeks' gestation. STUDY DESIGN: This was a prospective observational nonintervention study in women attending a routine hospital visit at 35 0/7 to 36 6/7 weeks' gestation. The visit included recording of maternal demographic characteristics and medical history, vascular indices, and hemodynamic parameters obtained by a noninvasive operator-independent device (pulse wave velocity, augmentation index, cardiac output, stroke volume, central systolic and diastolic blood pressures, total peripheral resistance, and fetal heart rate), mean arterial pressure, uterine artery pulsatility index, and serum concentration of placental growth factor and soluble fms-like tyrosine kinase-1. The performance of screening for delivery with PE at any time and at <3 weeks from assessment using a combination of maternal risk factors and various combinations of biomarkers was determined. RESULTS: The study population consisted of 6746 women with singleton pregnancies, including 176 women (2.6%) who subsequently developed PE. There were 3 main findings. First, in women who developed PE, compared with those who did not, there were higher central systolic and diastolic blood pressures, pulse wave velocity, peripheral vascular resistance, and augmentation index. Second, the most discriminatory indices were systolic and diastolic blood pressures and pulse wave velocity, with poor prediction from the other indices. However, the performance of screening by a combination of maternal risk factors plus mean arterial pressure was at least as high as that of a combination of maternal risk factors plus central systolic and diastolic blood pressures; consequently, in screening for PE, pulse wave velocity, mean arterial pressure, uterine artery pulsatility index, placental growth factor, and soluble fms-like tyrosine kinase-1 were used. Third, in screening for both PE within 3 weeks and PE at any time from assessment, the detection rate at a false-positive rate of 10% of a biophysical test consisting of maternal risk factors plus mean arterial pressure, uterine artery pulsatility index, and pulse wave velocity (PE within 3 weeks: 85.2%; 95% confidence interval, 75.6%-92.1%; PE at any time: 69.9%; 95% confidence interval, 62.5%-76.6%) was not significantly different from a biochemical test using the competing risks model to combine maternal risk factors with placental growth factor and soluble fms-like tyrosine kinase-1 (PE within 3 weeks: 80.2%; 95% confidence interval, 69.9%-88.3%; PE at any time: 64.2%; 95% confidence interval, 56.6%-71.3%), and they were both superior to screening by low placental growth factor concentration (PE within 3 weeks: 53.1%; 95% confidence interval, 41.7%-64.3%; PE at any time: 44.3; 95% confidence interval, 36.8%-52.0%) or high soluble fms-like tyrosine kinase-1-to-placental growth factor concentration ratio (PE within 3 weeks: 65.4%; 95% confidence interval, 54.0%-75.7%; PE at any time: 53.4%; 95% confidence interval, 45.8%-60.9%). CONCLUSION: First, increased maternal arterial stiffness preceded the clinical onset of PE. Second, maternal pulse wave velocity at 35 to 37 weeks' gestation in combination with mean arterial pressure and uterine artery pulsatility index provided effective prediction of subsequent development of preeclampsia.

Weibull parametric model for survival analysis in women with endometrial cancer using clinical and T2-weighted MRI radiomic features.

BACKGROUND: Semiparametric survival analysis such as the Cox proportional hazards (CPH) regression model is commonly employed in endometrial cancer (EC) study. Although this method does not need to know the baseline hazard function, it cannot estimate event time ratio (ETR) which measures relative increase or decrease in survival time. To estimate ETR, the Weibull parametric model needs to be applied. The objective of this study is to develop and evaluate the Weibull parametric model for EC patients' survival analysis. METHODS: Training (n = 411) and testing (n = 80) datasets from EC patients were retrospectively collected to investigate this problem. To determine the optimal CPH model from the training dataset, a bi-level model selection with minimax concave penalty was applied to select clinical and radiomic features which were obtained from T2-weighted MRI images. After the CPH model was built, model diagnostic was carried out to evaluate the proportional hazard assumption with Schoenfeld test. Survival data were fitted into a Weibull model and hazard ratio (HR) and ETR were calculated from the model. Brier score and time-dependent area under the receiver operating characteristic curve (AUC) were compared between CPH and Weibull models. Goodness of the fit was measured with Kolmogorov-Smirnov (KS) statistic. RESULTS: Although the proportional hazard assumption holds for fitting EC survival data, the linearity of the model assumption is suspicious as there are trends in the age and cancer grade predictors. The result also showed that there was a significant relation between the EC survival data and the Weibull distribution. Finally, it showed that Weibull model has a larger AUC value than CPH model in general, and it also has smaller Brier score value for EC survival prediction using both training and testing datasets, suggesting that it is more accurate to use the Weibull model for EC survival analysis. CONCLUSIONS: The Weibull parametric model for EC survival analysis allows simultaneous characterization of the treatment effect in terms of the hazard ratio and the event time ratio (ETR), which is likely to be better understood. This method can be extended to study progression free survival and disease specific survival. TRIAL REGISTRATION: ClinicalTrials.gov NCT03543215, https://clinicaltrials.gov/ , date of registration: 30th June 2017.

Multi-state Markov model for time to treatment changes for HIV/AIDS patients: a retrospective cohort national datasets, Ethiopia.

BACKGROUND: Virological failure, drug resistance, toxicities, and other issues make it difficult for ART to maintain long-term sustainability. These issues would force a modification in the patient's treatment plan. The aim of this research was to determine whether first-line antiretroviral therapy is durable and to identify the factors that lead to patients on HAART changing their first highly active antiretroviral therapy regimen. METHODS: A retrospective cohort study was conducted from October, 2019-March, 2020 across all regional states including Addis Ababa and Dire Dawa administrative cities. The target population is from all health facilities that have been providing ART service for at least the past 6 months as of October 2019. Multi-stage clustered sampling method was used to select study facilities and participants. Simple random selected ART medical records of patients ever enrolled in ART treatment services. We adopted a multi-state survival modelling (msm) approach assuming each treatment regimen as state. We estimate the transition probability of patients to move from one regimen to another for time to treatment change/switch. We estimated the transition probability, prediction probabilities and length of stay and factor associated with treatment modification of patients to move from one regimen to another. RESULTS: Any of the six therapy combinations (14.4%) altered their treatment at least once during the follow-up period for a variety of reasons. Of the patients, 4,834 (13.26%) changed their treatments just once, while 371 (1.1%) changed it more than once. For 38.6% of the time, a treatment change was undertaken due to toxicity, another infection or comorbidity, or another factor, followed by New drugs were then made accessible and other factors 18.3% of the time, a drug was out of supply; 2.6% of those instances involved pregnancy; and 43.1% involved something else. Highly active anti-retroviral therapy (HAART) combinations TDF + 3TC + NVP, d4T + 3TC + NVP, and TDF + 3TC + EFV were high to treatment alterations in all reasons of treatment modifications, with 29.74%, 26.52%, and 19.52% treatment changes, respectively. Early treatment modification or regime change is one of the treatment combinations that include the d4T medication that creates major concern. The likelihood of staying and moving at the the start of s = 0 and 30-month transitions increased, but the likelihood of staying were declined. For this cohort dataset, the presence of opportunistic disease, low body weight, baseline CD4 count, and baseline TB positive were risk factors for therapy adjustment. CONCLUSION: Given that the current study took into account a national dataset, it provides a solid basis for ART drug status and management. The patient had a higher likelihood of adjusting their treatment at some point during the follow-up period due to drug toxicity, comorbidity, drug not being available, and other factors, according to the prediction probability once more. Baseline TB positivity, low CD4 count, opportunistic disease, and low body weight were risk factors for therapy adjustment in this cohort dataset.

Screening for pre-eclampsia by maternal serum glycosylated fibronectin and angiogenic markers at 36 weeks' gestation.

OBJECTIVES: First, to examine the predictive performance of maternal serum glycosylated fibronectin (GlyFn) at 35 + 0 to 36 + 6 weeks' gestation in screening for delivery with pre-eclampsia (PE) and delivery with gestational hypertension (GH) at ≥ 37 weeks' gestation, both within 3 weeks and at any time after the examination. Second, to compare the predictive performance for delivery with PE and delivery with GH of various combinations of biomarkers, including GlyFn, mean arterial pressure (MAP), uterine artery pulsatility index (UtA-PI), serum placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1). Third, to compare the predictive performance for delivery with PE and delivery with GH by serum PlGF concentration, sFlt-1/PlGF concentration ratio and the competing-risks model with different combinations of biomarkers as above. Fourth, to compare the predictive performance of screening at 11 + 0 to 13 + 6 weeks vs 35 + 0 to 36 + 6 weeks for delivery with PE and delivery with GH at ≥ 37 weeks' gestation. METHODS: This was a case-control study in which maternal serum GlyFn was measured in stored samples from a non-intervention screening study in singleton pregnancies at 35 + 0 to 36 + 6 weeks' gestation using a point-of-care device. We used samples from women who delivered at ≥ 37 weeks' gestation, including 100 who developed PE, 100 who developed GH and 600 controls who did not develop PE or GH. In all cases, MAP, UtA-PI, PlGF and sFlt-1 were measured during the routine visit at 35 + 0 to 36 + 6 weeks. We used samples from patients that had been examined previously at 11 + 0 to 13 + 6 weeks' gestation. Levels of GlyFn were transformed to multiples of the expected median (MoM) values after adjusting for maternal demographic characteristics and elements from the medical history. Similarly, the measured values of MAP, UtA-PI, PlGF and sFlt-1 were converted to MoM. The competing-risks model was used to combine the prior distribution of the gestational age at delivery with PE, obtained from maternal risk factors, with various combinations of biomarker MoM values to derive the patient-specific risks of delivery with PE. The performance of screening of different strategies was estimated by examining the detection rate (DR) at a 10% fixed false-positive rate (FPR) and McNemar's test was used to compare the DRs between the different methods of screening. RESULTS: The DR, at 10% FPR, of screening by the triple test (maternal risk factors plus MAP, PlGF and sFlt-1) was 83.7% (95% CI, 70.3-92.7%) for delivery with PE within 3 weeks of screening and 80.0% (95% CI, 70.8-87.3%) for delivery with PE at any time after screening, and this performance was not improved by the addition of GlyFn. The performance of screening by a combination of maternal risk factors, MAP, PlGF and GlyFn was similar to that of the triple test, both for delivery with PE within 3 weeks and at any time after screening. The performance of screening by a combination of maternal risk factors, MAP, UtA-PI and GlyFn was similar to that of the triple test, and they were both superior to screening by low PlGF concentration (PE within 3 weeks: DR, 65.3% (95% CI, 50.4-78.3%); PE at any time: DR, 56.0% (95% CI, 45.7-65.9%)) or high sFlt-1/PlGF concentration ratio (PE within 3 weeks: DR, 73.5% (95% CI, 58.9-85.1%); PE at any time: DR, 63.0% (95% CI, 52.8-72.4%)). The predictive performance of screening at 35 + 0 to 36 + 6 weeks' gestation for delivery with PE and delivery with GH at ≥ 37 weeks' gestation was by far superior to screening at 11 + 0 to 13 + 6 weeks. CONCLUSION: GlyFn is a potentially useful biomarker in third-trimester screening for term PE and term GH, but the findings of this case-control study need to be validated by prospective screening studies. © 2023 International Society of Ultrasound in Obstetrics and Gynecology.

Ophthalmic artery Doppler at 36 weeks' gestation in prediction of pre-eclampsia: validation and update of previous model.

OBJECTIVE: To validate and extend a model incorporating maternal ophthalmic artery Doppler at 35-37 weeks' gestation in the prediction of subsequent development of pre-eclampsia (PE). METHODS: This was a prospective validation study of screening for PE (defined according to the 2019 American College of Obstetricians and Gynecologists criteria) by maternal ophthalmic artery peak systolic velocity (PSV) ratio in 6746 singleton pregnancies undergoing routine care at 35 + 0 to 36 + 6 weeks' gestation (validation dataset). Additionally, the data from the validation dataset were combined with those of 2287 pregnancies that were previously used for development of the model (training dataset), and the combined data were used to update the original model parameters. The competing-risks model was used to estimate the individual patient-specific risk of delivery with PE at any time and within 3 weeks from assessment by a combination of maternal demographic characteristics and medical history with PSV ratio alone and in combination with the established PE biomarkers of mean arterial pressure (MAP), uterine artery pulsatility index (UtA-PI), serum placental growth factor (PlGF) and serum soluble fms-like tyrosine kinase-1 (sFlt-1). We evaluated the predictive performance of the model by examining, first, the ability to discriminate between the PE and non-PE groups using the area under the receiver-operating-characteristics curve and the detection rate (DR) at fixed screen-positive (SPR) and false-positive rates of 10% and, second, calibration by measuring the calibration slope and calibration-in-the-large. McNemar's test was used to compare the performance of screening by a biophysical test (maternal factors, MAP, UtA-PI and PSV ratio) vs a biochemical test (maternal factors, PlGF and sFlt-1), low PlGF concentration (< 10th percentile) or high sFlt-1/PlGF concentration ratio (> 90th percentile). RESULTS: In the validation dataset, the performance of screening by maternal factors and PSV ratio for delivery with PE within 3 weeks and at any time after assessment was consistent with that in the training dataset, and there was good agreement between the predicted and observed incidence of PE. In the combined data from the training and validation datasets, good prediction for PE was achieved in screening by a combination of maternal factors, MAP, UtA-PI, PlGF, sFlt-1 and PSV ratio, with a DR, at a 10% SPR, of 85.0% (95% CI, 76.5-91.4%) for delivery with PE within 3 weeks and 65.7% (95% CI, 59.2-71.7%) for delivery with PE at any time after assessment. The performance of a biophysical test was superior to that of screening by low PlGF concentration or high sFlt-1/PlGF concentration ratio but not significantly different from the performance of a biochemical test combining maternal factors with PlGF and sFlt-1 for both PE within 3 weeks and PE at any time after assessment. CONCLUSION: Maternal ophthalmic artery PSV ratio at 35-37 weeks' gestation in combination with other biomarkers provides effective prediction of subsequent development of PE. © 2023 International Society of Ultrasound in Obstetrics and Gynecology.

Cost-utility analysis of second-line axicabtagene ciloleucel versus standard of care in Japan based on the ZUMA-7 trial.

Aim: To perform a cost-effectiveness analysis comparing axicabtagene ciloleucel (axi-cel) with standard of care (SoC; salvage chemoimmunotherapy, followed by high-dose therapy with autologous stem cell rescue for responders) for second-line (2L) treatment of adults with relapsed or refractory large B-cell lymphoma (r/r LBCL) in the pivotal ZUMA-7 trial data from a Japanese payer perspective. Materials & methods: A three-state partitioned survival model was utilized using population and clinical inputs from the ZUMA-7 trial data over a lifetime horizon. Results: Axi-cel was associated with greater incremental quality-adjusted life-years (2.06) and higher incremental total costs ($48,685.59/¥6.9 million) leading to an incremental cost-effectiveness ratio of $23,590.34/¥3.3 million per quality-adjusted life-years compared with SoC. Conclusion: Axi-cel is a cost-effective treatment alternative to SoC for 2L treatment of adults with r/r LBCL.

[Box: see text].

Improving pressure injury risk assessment using real-world data from skilled nursing facilities: A cohort study.

This study aimed to improve the predictive accuracy of the Braden assessment for pressure injury risk in skilled nursing facilities (SNFs) by incorporating real-world data and training a survival model. A comprehensive analysis of 126 384 SNF stays and 62 253 in-house pressure injuries was conducted using a large calibrated wound database. This study employed a time-varying Cox Proportional Hazards model, focusing on variations in Braden scores, demographic data and the history of pressure injuries. Feature selection was executed through a forward-backward process to identify significant predictive factors. The study found that sensory and moisture Braden subscores were minimally contributive and were consequently discarded. The most significant predictors of increased pressure injury risk were identified as a recent (within 21 days) decrease in Braden score, low subscores in nutrition, friction and activity, and a history of pressure injuries. The model demonstrated a 10.4% increase in predictive accuracy compared with traditional Braden scores, indicating a significant improvement. The study suggests that disaggregating Braden scores and incorporating detailed wound histories and demographic data can substantially enhance the accuracy of pressure injury risk assessments in SNFs. This approach aligns with the evolving trend towards more personalized and detailed patient care. These findings propose a new direction in pressure injury risk assessment, potentially leading to more effective and individualized care strategies in SNFs. The study highlights the value of large-scale data in wound care, suggesting its potential to enhance quantitative approaches for pressure injury risk assessment and supporting more accurate, data-driven clinical decision-making.

Cost-effectiveness analysis of pembrolizumab in combination with chemotherapy compared with chemotherapy alone as first-line treatment for patients with advanced biliary tract cancer in China.

OBJECTIVE: The objective of this study is to evaluate the cost-effectiveness of adding pembrolizumab to the standard first-line therapy of advanced biliary tract cancer (BTC) with gemcitabine and cisplatin from the perspective of the Chinese healthcare system. METHODS: The partitioned survival model developed from clinical data obtained in The KEYNOTE-966 trial served as the basis for a simulation in the TreeAge Pro 2011 software. The objective of the research was to estimate the 10-year life expectancy and total healthcare costs of patients with BTC, utilizing primary outcomes that evaluated costs, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratio (ICER). To establish the willingness-to-pay (WTP) threshold, the 2022 Chinese per capita gross domestic product (GDP) of $37304.346/QALY was adopted. Furthermore, sensitivity analysis was conducted to ascertain the study's results under varying levels of uncertainty. RESULTS: Compared to chemotherapy alone, the addition of pembrolizumab to chemotherapy has been shown to yield an incremental gain of 0.184 quality-adjusted life years (QALY) at an additional cost of $103940.706. This translates into an incremental cost-effectiveness ratio (ICER) of $564895.141/QALY, which exceeds the willingness-to-pay (WTP) threshold in China. One-way sensitivity analyses performed on the model recognize the utility of PD, subsequent cost, and the cost of Pembrolizumab 100 mg had a major influence on the outcomes. However, no parameter elicited an ICER lower than the willingness-to-pay (WTP) threshold. CONCLUSIONS: Based on the perspective of the Chinese healthcare system, the utilization of pembrolizumab in combination with chemotherapy as an first-line treatment option for BTC does not appear to be a cost-effective approach compared to chemotherapy as a standalone therapy.

Cost-effectiveness of first-line immunotherapy combinations with or without chemotherapy for advanced non-small cell lung cancer: a modelling approach.

BACKGROUND: Many studies have explored the cost-effectiveness of immunotherapy versus chemotherapy alone. However, there is paucity of evidence on direct pharmacoeconomic studies related to immunotherapy combinations. Thus, we aimed at assessing the economic outcomes of first-line immunotherapy combinations in the treatment of advanced non-small cell lung cancer (NSCLC) from the Chinese health care perspective. METHODS: The mutual hazard ratios (HRs) of ten immunotherapy combinations and one chemotherapy regimen for the overall survival (OS) and progression-free survival (PFS) were obtained from a network meta-analysis. Based on proportional hazard (PH) assumption, adjusted OS and PFS curves were established to make the effects comparable. With the parameters of cost and utility, and of scale and shape from the fit of adjusted OS and PFS curves obtained from previous studies, a partitioned survival model was designed to estimate the cost-effectiveness of immunotherapy combinations versus chemotherapy alone. Parameter uncertainty in model inputs was assessed using one-way deterministic and probabilistic sensitivity analyses. RESULTS: The incremental cost of camrelizumab plus chemotherapy versus chemotherapy alone was $13,180.65, the lowest among all the other immunotherapy combinations. Furthermore, sintilimab plus chemotherapy (sint-chemo) provided the highest quality-adjusted life-year (QALY) benefit versus chemotherapy alone (incremental QALYs = 0.45). Sint-chemo yielded the best incremental cost-effectiveness ratio (ICER) versus chemotherapy alone (ICER = $34,912.09/QALY), at the current price. The cost-effectiveness probabilities were 32.01% and 93.91% for pembrolizumab plus chemotherapy, and atezolizumab plus bevacizumab plus chemotherapy, respectively (if the original price of the pembrolizumab, atezolizumab, and bevacizumab were decreased by 90%). CONCLUSIONS: Based on the fact that there is fierce competition in the PD-1/PD-L1 market, pharmaceutical enterprises should strive for greater efficacy, and optimal pricing strategy for therapies.

Divorce and risk of suicide attempt: a Swedish national study.

BACKGROUND: Prior research has reported an association between divorce and suicide attempt. We aimed to clarify this complex relationship, considering sex differences, temporal factors, and underlying etiologic pathways. METHODS: We used Swedish longitudinal national registry data for a cohort born 1960-1990 that was registered as married between 1978 and 2018 (N = 1 601 075). We used Cox proportional hazards models to estimate the association between divorce and suicide attempt. To assess whether observed associations were attributable to familial confounders or potentially causal in nature, we conducted co-relative analyses. RESULTS: In the overall sample and in sex-stratified analyses, divorce was associated with increased risk of suicide attempt (adjusted hazard ratios [HRs] 1.66-1.77). Risk was highest in the year immediately following divorce (HRs 2.20-2.91) and declined thereafter, but remained elevated 5 or more years later (HRs 1.41-1.51). Divorcees from shorter marriages were at higher risk for suicide attempt than those from longer marriages (HRs 3.33-3.40 and 1.20-1.36, respectively). In general, HRs were higher for divorced females than for divorced males. Co-relative analyses suggested that familial confounders and a causal pathway contribute to the observed associations. CONCLUSIONS: The association between divorce and risk of suicide attempt is complex, varying as a function of sex and time-related variables. Given evidence that the observed association is due in part to a causal pathway from divorce to suicide attempt, intervention or prevention efforts, such as behavioral therapy, could be most effective early in the divorce process, and in particular among females and those whose marriages were of short duration.

Personalized stratification of pregnancy care for small for gestational age neonates from biophysical markers at midgestation.

BACKGROUND: Antenatal identification of pregnancies at high risk of delivering small for gestational age neonates may improve the management of the condition and reduce the associated adverse perinatal outcomes. In a series of publications, we have developed a new competing-risks model for small for gestational age prediction, and we demonstrated that the new approach has a superior performance to that of the traditional methods. The next step in shaping the appropriate management of small for gestational age is the timely assessment of these high-risk pregnancies according to an antenatal stratification plan. OBJECTIVE: This study aimed to demonstrate the stratification of pregnancy care based on individual patient risk derived from the application of the competing-risks model for small for gestational age that combines maternal factors with sonographic estimated fetal weight and uterine artery pulsatility index at midgestation. STUDY DESIGN: This was a prospective observational study of 96,678 singleton pregnancies undergoing routine ultrasound examination at 19 to 24 weeks of gestation, which included recording of estimated fetal weight and measurement of uterine artery pulsatility index. The competing-risks model for small for gestational age was used to create a patient-specific stratification curve capable to define a specific timing for a repeated ultrasound examination after 24 weeks. We examined different stratification plans with the intention of detecting approximately 80%, 85%, 90%, and 95% of small for gestational age neonates with birthweight <3rd and <10th percentiles at any gestational age at delivery until 36 weeks; all pregnancies would be offered a routine ultrasound examination at 36 weeks. RESULTS: The stratification of pregnancy care for small for gestational age can be based on a patient-specific stratification curve. Factors from maternal history, low estimated fetal weight, and increased uterine artery pulsatility index shift the personalized risk curve toward higher risks. The degree of shifting defines the timing for assessment for each pregnancy. If the objective of our antenatal plan was to detect 80%, 85%, 90%, and 95% of small for gestational age neonates at any gestational age at delivery until 36 weeks, the median (range) proportions (percentages) of population examined per week would be 3.15 (1.9-3.7), 3.85 (2.7-4.5), 4.75 (4.0-5.4), and 6.45 (3.7-8.0) for small for gestational age <3rd percentile and 3.8 (2.5-4.6), 4.6 (3.6-5.4), 5.7 (3.8-6.4), and 7.35 (3.3-9.8) for small for gestational age <10th percentile, respectively. CONCLUSION: The competing-risks model provides an effective personalized continuous stratification of pregnancy care for small for gestational age which is based on individual characteristics and biophysical marker levels recorded at the midgestation scan.

Screening for pre-eclampsia by maternal serum glycosylated fibronectin at 11-13 weeks' gestation.

OBJECTIVE: To examine the performance of screening for preterm and term pre-eclampsia (PE) at 11-13 weeks' gestation by maternal factors and combinations of maternal serum glycosylated fibronectin (GlyFn), mean arterial pressure (MAP), uterine artery pulsatility index (UtA-PI) and serum placental growth factor (PlGF). METHODS: This was a case-control study in which maternal serum GlyFn was measured using a point-of-care device in stored samples from a non-intervention screening study of singleton pregnancies at 11 + 0 to 13 + 6 weeks' gestation. In the same samples, PlGF was measured by time-resolved fluorometry. We used samples from women who delivered with PE at < 37 weeks' gestation (n = 100), PE at ≥ 37 weeks (n = 100), gestational hypertension (GH) at < 37 weeks (n = 100), GH at ≥ 37 weeks (n = 100) and 1000 normotensive controls with no pregnancy complications. In all cases, MAP and UtA-PI had been measured during the routine 11-13-week visit. Levels of GlyFn were transformed to multiples of the expected median (MoM) values after adjusting for maternal demographic characteristics and elements of medical history. Similarly, the measured values of MAP, UtA-PI and PlGF were converted to MoMs. The competing-risks model was used to combine the prior distribution of gestational age at delivery with PE, obtained from maternal characteristics, with various combinations of biomarker MoM values to derive the patient-specific risks of delivery with PE or GH at < 37 and ≥ 37 weeks' gestation. Screening performance was estimated by examining the area under the receiver-operating-characteristics curve (AUC) and detection rate (DR) at 10% fixed false-positive rate (FPR). RESULTS: The maternal characteristics and elements of medical history with a significant effect on the measurement of GlyFn were maternal age, weight, height, race, smoking status and history of PE. In pregnancies that developed PE, GlyFn MoM was increased and the deviation from normal decreased with increasing gestational age at delivery. The DR and AUC of screening for delivery with PE at < 37 weeks' gestation by maternal factors alone were 50% and 0.834, respectively, and these increased to 80% and 0.949, respectively, when maternal risk factors were combined with MAP, UtA-PI and PlGF (triple test). The performance of the triple test was similar to that of screening by a combination of maternal factors, MAP, UtA-PI and GlyFn (DR, 79%; AUC, 0.946) and that of screening by a combination of maternal factors, MAP, PlGF and GlyFn (DR, 81%; AUC, 0.932). The performance of screening for delivery with PE at ≥ 37 weeks' gestation was poor; the DR for screening by maternal factors alone was 35% and increased to only 39% with use of the triple test. Similar results were obtained when GlyFn replaced PlGF or UtA-PI in the triple test. The DR of screening for GH with delivery at < 37 and ≥ 37 weeks' gestation by maternal factors alone was 34% and 25%, respectively, and increased to 54% and 31%, respectively, with use of the triple test. Similar results were obtained when GlyFn replaced PlGF or UtA-PI in the triple test. CONCLUSIONS: GlyFn is a potentially useful biomarker in first-trimester screening for preterm PE, but the findings of this case-control study need to be validated by prospective screening studies. The performance of screening for term PE or GH at 11 + 0 to 13 + 6 weeks' gestation by any combination of biomarkers is poor. © 2023 International Society of Ultrasound in Obstetrics and Gynecology.

Prediction of hypertensive disorders after screening at 36 weeks' gestation: comparison of angiogenic markers with competing-risks model.

OBJECTIVE: To compare the performance at 35 + 0 to 36 + 6 weeks' gestation of screening for delivery with pre-eclampsia (PE) at various timepoints, using one of three approaches: placental growth factor (PlGF) concentration, soluble fms-like tyrosine kinase-1 (sFlt-1) to PlGF concentration ratio, or the competing-risks model, which combines maternal risk factors with biomarkers to estimate patient-specific risk. METHODS: This was a prospective observational study of women attending for a routine hospital visit at 35 + 0 to 36 + 6 weeks' gestation at one of two maternity hospitals in England between 2016 and 2022. During the visit, maternal demographic characteristics and medical history were recorded and serum PlGF, serum sFlt-1 and mean arterial pressure (MAP) were measured. Detection rates (DRs) were evaluated for delivery with PE (defined as per American College of Obstetricians and Gynecologists 2019 criteria) within 1 week, within 2 weeks or at any time after screening, using the following strategies: (i) low PlGF (< 10th percentile); (ii) high sFlt-1/PlGF ratio (> 90th percentile); or (iii) the competing-risks model, in which maternal factors were combined with multiples of the median values of PlGF ('single test'), PlGF and sFlt-1 ('double test') or PlGF, sFlt-1 and MAP ('triple test'). Risk cut-offs corresponded to a screen-positive rate of 10%. DRs were compared between tests. RESULTS: Of 34 782 pregnancies, 831 (2.4%) developed PE. In screening for delivery with PE at any time from assessment, the DR at 10% screen-positive rate was 47% by low PlGF alone, 54% by the single test, 55% by high sFlt-1/PlGF ratio, 61% by the double test and 68% by the triple test. In screening for delivery with PE within 2 weeks from assessment, the respective values were 67%, 74%, 74%, 80% and 87%. In screening for delivery with PE within 1 week from assessment, the respective values were 77%, 81%, 85%, 88% and 91%. For prediction of PE at any time, the DR was significantly higher with the triple test compared to PlGF alone or the sFlt-1/PlGF ratio, with a DR difference (95% CI) of 20.1% (16.7-23.0%) and 12.4% (9.7-15.3%), respectively. Similar results were seen for prediction of PE within 2 weeks (20.6% (14.9-26.8%) and 12.9% (7.7-17.5%), respectively) and prediction of PE within 1 week (13.5% (5.4-21.6%) and 5.4% (0.0-10.8%), respectively). The double test was superior to the sFlt-1/PlGF ratio and the single test was superior to PlGF alone in the prediction of PE within 2 weeks and at any time from assessment, but not within 1 week of assessment. CONCLUSION: At 35 + 0 to 36 + 6 weeks' gestation, the performance of screening for PE by the competing-risks model triple test is superior to that of PlGF alone or the sFlt-1/PlGF ratio for the development of disease within 1 week, within 2 weeks and at any time from screening. © 2023 International Society of Ultrasound in Obstetrics and Gynecology.

INES: Interactive tool for construction and extrapolation of partitioned survival models.

BACKGROUND: INES (INteractive model for Extrapolation of Survival and cost) provides an open-access tool powered by R that implements three-state partitioned survival models (PSM). This article describes the properties of the tool, and the situations where INES may or may not be suitable. METHODS: INES is designed to be used by investigators or healthcare professionals who have a good grasp of the principles of economic evaluation and understand the strengths and weaknesses of partitioned survival models, but are not sufficiently familiar with a statistical package such as Excel or R to be able to construct and test a de-novo PSM themselves. INES is delivered to the user via a batch file. Once downloaded to the user's hard drive, it interacts with the user via a portable version of R with web interactivity built in Shiny. INES requires absolutely no knowledge of R and the user does not need to have R or any of its dependences installed. Hence the user will deal with a standalone Shiny app. Inputs (digitalized survival curves, unit costs, posology, hazard ratios, discount rate) can be uploaded from a template spreadsheet. RESULTS: The INES application provides a seamlessly integrated package for estimating a set of parametric hazard functions for progression free and overall survival, selecting an appropriate function from this menu, and applying this as an input to a PSM to calculate mean costs and quality-adjusted life years. Examples are given that may serve as a tutorial. CONCLUSION: INES offers a rapid, flexible, robust and transparent tool for parametric survival analysis and calculating a PSM that can be used in many different contexts.