Risk of retinopathy in women with pregnancy-induced hypertension: a nationwide population-based cohort study of 9-year follow-up after delivery.

BACKGROUND: The retina is potentially associated with several physiological, hormonal, and metabolic changes during pregnancy. The few available epidemiologic studies of ocular changes in pregnancy have mainly concerned retinopathies. Pregnancy-induced hypertension, which leads to ocular manifestations including blurred vision, photopsia, scotoma, and diplopia, might induce reactive changes in the retinal vessels. Although several studies have suggested the existence of pregnancy-induced hypertension-related retinal ocular disease, there are few large cohort studies on this topic. OBJECTIVE: This study aimed to investigate the risk of major retinal diseases including central serous chorioretinopathy, diabetic retinopathy, retinal vein occlusion, retinal artery occlusion, and hypertensive retinopathy in the long-term postpartum stage according to the presence of previous pregnancy-induced hypertension in a large cohort based on the Korean National Health Insurance Database. STUDY DESIGN: On the basis of Korean health data, 909,520 patients who delivered from 2012 to 2013 were analyzed. Among them, patients who had previous ocular diseases or hypertension and multiple births were excluded. Finally, 858,057 mothers were assessed for central serous chorioretinopathy (ICD-10: H35.70), diabetic retinopathy (ICD-10: H36.0, E10.31, E10.32, E11.31, E11.32, E12.31, E13.31, E13.32, E14.31, E14.32), retinal vein occlusion (ICD-10: H34.8), retinal artery occlusion (ICD-10: H34.2), and hypertensive retinopathy (ICD-10: H35.02) for 9 years after delivery. Enrolled patients were divided into 2 groups: 10,808 patients with and 847,249 without pregnancy-induced hypertension. The primary outcomes were the incidence of central serous chorioretinopathy, diabetic retinopathy, retinal vein occlusion, retinal artery occlusion, and hypertensive retinopathy 9 years after delivery. Clinical variables were age, parity, cesarean delivery, gestational diabetes mellitus, and postpartum hemorrhage. In addition, pregestational diabetes mellitus, kidney diseases, cerebrovascular diseases, and cardiovascular diseases were adjusted. RESULTS: Postpartum retinal disease during the 9 years after delivery and total retinal diseases showed higher rates in patients with pregnancy-induced hypertension. In detail, the rates of central serous chorioretinopathy (0.3% vs 0.1%), diabetic retinopathy (1.79% vs 0.5%), retinal vein occlusion (0.19% vs 0.1%), and hypertensive retinopathy (0.62% vs 0.05%) were higher than those found in patients without pregnancy-induced hypertension. After adjusting for confounding factors, pregnancy-induced hypertension was associated with development of postpartum retinopathy, with a >2-fold increase (hazard ratio, 2.845; 95% confidence interval, 2.54-3.188). Furthermore, pregnancy-induced hypertension affected the development of central serous chorioretinopathy (hazard ratio, 3.681; 95% confidence interval, 2.667-5.082), diabetic retinopathy (hazard ratio, 2.326; 95% confidence interval, 2.013-2.688), retinal vein occlusion (hazard ratio, 2.241; 95% confidence interval, 1.491-3.368), and hypertensive retinopathy (hazard ratio, 11.392; 95% confidence interval, 8.771-14.796) after delivery. CONCLUSION: A history of pregnancy-induced hypertension increases the risk of central serous chorioretinopathy, diabetic retinopathy, retinal vein occlusion, and hypertensive retinopathy according to 9-year long-term ophthalmologic follow-up.

Accountable survival contrast-learning for optimal dynamic treatment regimes.

Dynamic treatment regime (DTR) is an emerging paradigm in recent medical studies, which searches a series of decision rules to assign optimal treatments to each patient by taking into account individual features such as genetic, environmental, and social factors. Although there is a large and growing literature on statistical methods to estimate optimal treatment regimes, most methodologies focused on complete data. In this article, we propose an accountable contrast-learning algorithm for optimal dynamic treatment regime with survival endpoints. Our estimating procedure is originated from a doubly-robust weighted classification scheme, which is a model-based contrast-learning method that directly characterizes the interaction terms between predictors and treatments without main effects. To reflect the censorship, we adopt the pseudo-value approach that replaces survival quantities with pseudo-observations for the time-to-event outcome. Unlike many existing approaches, mostly based on complicated outcome regression modeling or inverse-probability weighting schemes, the pseudo-value approach greatly simplifies the estimating procedure for optimal treatment regime by allowing investigators to conveniently apply standard machine learning techniques to censored survival data without losing much efficiency. We further explore a SCAD-penalization to find informative clinical variables and modified algorithms to handle multiple treatment options by searching upper and lower bounds of the objective function. We demonstrate the utility of our proposal via extensive simulations and application to AIDS data.

Association of long-term exposure to PM2.5 and survival following ischemic heart disease.

BACKGROUND: Numerous studies have suggested that long-term exposure to particulate matter ≤2.5 µm (PM2.5) may cause cardiovascular morbidity and mortality. However, susceptibility among those with a history of ischemic heart disease is less clearly understood. We aimed to evaluate whether long-term PM2.5 exposure is related to mortality among patients with ischemic heart disease. METHODS: We followed up 306,418 patients hospitalized with ischemic heart disease in seven major cities in South Korea between 2008 and 2016 using the National Health Insurance Database. We linked the modeled PM2.5 data corresponding to each patient's administrative districts and estimated hazard ratios (HRs) of cause-specific mortality associated with the long-term exposure to PM2.5 in time-varying Cox proportional hazard models after adjusting for individual- and area-level characteristics. We also estimated HRs by sex, age group (65-74 vs. ≥75 years), and household income. RESULTS: Of the patients with ischemic heart disease, mean age at the discharge was 76.8 years, and 105,913 died during a mean follow-up duration of 21.4 months. The HR of all-cause mortality was 1.10 [95% confidence intervals (CI): 1.07, 1.14] per 10 µg/m3 increase in a 12-month moving average PM2.5. The HRs of cardiovascular, stroke, and ischemic heart disease were 1.17 (95% CI: 1.11, 1.24), 1.17 (95% CI: 1.06, 1.30), and 1.25 (95% CI: 1.15, 1.35), respectively. The subgroup analyses showed that participants aged 65-74 years were more susceptible to adverse effects of PM2.5 exposure. We did not observe any differences in the risk by sex and household income. CONCLUSION: Mortality from all-cause and cardiovascular disease following hospitalization due to ischemic heart disease was higher among individuals with greater PM2.5 exposure in seven major cities in South Korea. The result supports the association of long-term exposure to air pollution with poor prognosis among patients with ischemic heart disease.

Doubly-robust methods for differences in restricted mean lifetimes using pseudo-observations.

In clinical studies or trials comparing survival times between two treatment groups, the restricted mean lifetime (RML), defined as the expectation of the survival from time 0 to a prespecified time-point, is often the quantity of interest that is readily interpretable to clinicians without any modeling restrictions. It is well known that if the treatments are not randomized (as in observational studies), covariate adjustment is necessary to account for treatment imbalances due to confounding factors. In this article, we propose a simple doubly-robust pseudo-value approach to effectively estimate the difference in the RML between two groups (akin to a metric for estimating average causal effects), while accounting for confounders. The proposed method combines two general approaches: (a) group-specific regression models for the time-to-event and covariate information, and (b) inverse probability of treatment assignment weights, where the RMLs are replaced by the corresponding pseudo-observations for survival outcomes, thereby mitigating the estimation complexities in presence of censoring. The proposed estimator is double-robust, in the sense that it is consistent if at least one of the two working models remains correct. In addition, we explore the potential of available machine learning algorithms in causal inference to reduce possible bias of the causal estimates in presence of a complex association between the survival outcome and covariates. We conduct extensive simulation studies to assess the finite-sample performance of the pseudo-value causal effect estimators. Furthermore, we illustrate our methodology via application to a dataset from a breast cancer cohort study. The proposed method is implementable using the R package drRML, available in GitHub.

Weighted least-squares regression with competing risks data.

The semiparametric accelerated failure time (AFT) model linearly relates the logarithm of the failure time to a set of covariates, while leaving the error distribution unspecified. This model has been widely investigated in survival literature due to its simple interpretation and relationship with linear models. However, there has been much less focus on developing AFT-type linear regression methods for analyzing competing risks data, in which patients can potentially experience one of multiple failure causes. In this article, we propose a simple least-squares (LS) linear regression model for a cause-specific subdistribution function, where the conventional LS equation is modified to account for data incompleteness under competing risks. The proposed estimators are shown to be consistent and asymptotically normal with consistent estimation of the variance-covariance matrix. We further extend the proposed methodology to risk prediction and analysis under clustered competing risks scenario. Simulation studies suggest that the proposed method provides rapid and valid statistical inferences and predictions. Application of our method to two oncology datasets demonstrate its utility in routine clinical data analysis.

Hospitalization for acute coronary syndrome increases the long-term risk of pneumonia: a population-based cohort study.

It is well established that the risk of acute coronary syndrome (ACS) increases after respiratory infection. However, the reverse association has not been evaluated. We tested the hypothesis that the long-term risk of pneumonia is increased after a new ACS event. A matched-cohort study was conducted using a nationally representative dataset. We identified patients with admission for ACS between 2004 and 2014, without a previous history of ACS or pneumonia. Incidence density sampling was used to match patients, on the basis of age and sex, to 3 controls who were also free from both ACS and pneumonia. We examined the incidence of pneumonia after ACS until the end of the cohort observation (Dec 31, 2014). The analysis cohort consisted of 5469 ACS cases and 16,392 controls (median age, 64 years; 68.3% men). The incidence rate ratios of the first and the total pneumonia episodes in the ACS group relative to the control group was 1.25 (95% confidence interval [CI], 1.11-1.41) and 1.23(95% CI 1.11-1.36), respectively. A significant ACS-related increase in the incidence of pneumonia was observed in the Cox-regression, shared frailty, and joint frailty model analyses, with hazard ratios of 1.25 (95% CI 1.09-1.42), 1.35 (95% CI 1.15-1.58), and 1.24 (95% CI 1.10-1.39), respectively. In this population-based cohort of patients who were initially free from both ACS and pneumonia, we found that hospitalization for ACS substantially increased the long term risk of pneumonia. This should be considered when formulating post-discharge care plans and preventive vaccination strategies in patients with ACS.

Sleep duration and mortality in Korean adults: a population-based prospective cohort study.

BACKGROUND: Increasing evidence suggests that sleep duration is associated with risks of various diseases including type 2 diabetes, cardiovascular disease (CVD), and certain types of cancer. However, the relationship with mortality is not clear, particularly in non-European populations. In this study, we investigated the association between sleep duration and mortality in a population-based prospective cohort of Korean adults. METHODS: This analysis included 34,264 participants (14,704 men and 19,560 women) of the Korea National Health and Nutrition Examination Survey (KNHANES) 2007-2013 who agreed to mortality follow-up through December 31, 2016. Sleep duration was self-reported at baseline and was categorized into four groups: ≤4, 5-6, 7-8, and ≥ 9 h/day. Cox proportional hazards models were performed to estimate hazard ratios (HR) and 95% confidence intervals (CI) for the associations with mortality (all-cause as well as CVD- and cancer-specific), adjusting for potential confounders. RESULTS: During up to 9.5 years of follow-up, we identified a total of 1028 deaths. We observed the lowest mortality at 5-6 h/day sleep. Compared with 7-8 h/day of sleep, short (≤4 h/day) and long (≥9 h/day) sleep were associated with a 1.05-fold (95% CI = 0.79-1.39) and 1.47-fold (95% CI = 1.15-1.87) higher all-cause mortality, respectively. After additional adjustment for self-rated health, the positive association with short sleep disappeared (HR = 0.99, 95% CI = 0.75-1.32) and the association with long sleep was slightly attenuated (HR = 1.38, 95% CI = 1.08-1.76). Long sleep was also nonsignificantly positively associated with both cancer-mortality (HR = 1.30, 95% CI = 0.86-1.98) and CVD-mortality (HR = 1.27, 95% CI = 0.73-2.21). There was no statistically significant evidence for nonlinearity in the relationships between sleep duration and mortality (all-cause as well as CVD- and cancer-specific). Effect modification by age, sex, education, and occupation were not statistically significant. CONCLUSIONS: Our findings suggest that long sleep duration is associated with an increased all-cause mortality in Korean adults.

Statistical analysis of clustered mixed recurrent-event data with application to a cancer survivor study.

In long-term follow-up studies on recurrent events, the observation patterns may not be consistent over time. During some observation periods, subjects may be monitored continuously so that each event occurence time is known. While during the other observation periods, subjects may be monitored discretely so that only the number of events in each period is known. This results in mixed recurrent-event and panel-count data. In these data, there is dependence among within-subject events. Furthermore, if the data are collected from multiple centers, then there is another level of dependence among within-center subjects. Literature exists for clustered recurrent-event data, but not for clustered mixed recurrent-event and panel-count data. Ignoring the cluster effect may lead to less efficient analysis. In this paper, we present a marginal modeling approach to take into account the cluster effect and provide asymptotic distributions of the resulting regression parameters. Our simulation study demonstrates that this approach works well for practical situations. It was applied to a study comparing the hospitalization rates between childhood cancer survivors and healthy controls, with data collected from 26 medical institutions across North America during more than 20 years of follow-up.

Secular trends in postpartum weight retention from 2003 to 2012: a nationwide, population-based, retrospective, longitudinal study in South Korea.

OBJECTIVE: To assess the secular trends in postpartum weight retention (PWR) over a decade with the population-based risk factors. DESIGN: Retrospective cohort study. SETTING: A national health screening examination data provided by the National Health Insurance Service in South Korea. PARTICIPANTS: 130 551 women who delivered babies between 1 January 2003 and 31 December 2012 and who underwent a national health screening examination 1 to 2 years prior to delivery and within 1 year after delivery. METHODS: Their PWR were determined during the study period of 2003-2012. We fitted logistic regression and linear mixed models to assess the independent contribution of PWR to obesity after adjusting for potential confounders. PRIMARY AND SECONDARY OUTCOME MEASURES: Prepregnancy and postpartum weight and body mass index (BMI). RESULTS: The adjusted PWR increased from mean value of 2.02 kg in 2003 (95% CI 1.88 to 2.15) to 2.79 kg in 2012 (95% CI 2.73 to 2.84) (p value for trend <0.01), after adjusting potential confounders including age, prepregnancy time, postpartum time, prepregnancy BMI, income and smoking status. The risk for a PWR of more than 5 kg also increased over the study period. CONCLUSIONS: Secular increases in PWR have been significantly observed between 2003 and 2012 for childbearing women. It is necessary to identify risk factors contributing to the observed increase and develop effective strategies to address the heightened risk for PWR.

Smoothed quantile regression analysis of competing risks.

Censored quantile regression models, which offer great flexibility in assessing covariate effects on event times, have attracted considerable research interest. In this study, we consider flexible estimation and inference procedures for competing risks quantile regression, which not only provides meaningful interpretations by using cumulative incidence quantiles but also extends the conventional accelerated failure time model by relaxing some of the stringent model assumptions, such as global linearity and unconditional independence. Current method for censored quantile regressions often involves the minimization of the L1 -type convex function or solving the nonsmoothed estimating equations. This approach could lead to multiple roots in practical settings, particularly with multiple covariates. Moreover, variance estimation involves an unknown error distribution and most methods rely on computationally intensive resampling techniques such as bootstrapping. We consider the induced smoothing procedure for censored quantile regressions to the competing risks setting. The proposed procedure permits the fast and accurate computation of quantile regression parameter estimates and standard variances by using conventional numerical methods such as the Newton-Raphson algorithm. Numerical studies show that the proposed estimators perform well and the resulting inference is reliable in practical settings. The method is finally applied to data from a soft tissue sarcoma study.

Nonparametric estimation of median survival times with applications to multi-site or multi-center studies.

We propose a nonparametric shrinkage estimator for the median survival times from several independent samples of right-censored data, which combines the samples and hypothesis information to improve the efficiency. We compare efficiency of the proposed shrinkage estimation procedure to unrestricted estimator and combined estimator through extensive simulation studies. Our results indicate that performance of these estimators depends on the strength of homogeneity of the medians. When homogeneity holds, the combined estimator is the most efficient estimator. However, it becomes inconsistent when homogeneity fails. On the other hand, the proposed shrinkage estimator remains efficient. Its efficiency decreases as the equality of the survival medians is deviated, but is expected to be as good as or equal to the unrestricted estimator. Our simulation studies also indicate that the proposed shrinkage estimator is robust to moderate levels of censoring. We demonstrate application of these methods to estimating median time for trauma patients to receive red blood cells in the Prospective Observational Multi-center Major Trauma Transfusion (PROMMTT) study.

Semiparametric accelerated failure time cure rate mixture models with competing risks.

Modern medical treatments have substantially improved survival rates for many chronic diseases and have generated considerable interest in developing cure fraction models for survival data with a non-ignorable cured proportion. Statistical analysis of such data may be further complicated by competing risks that involve multiple types of endpoints. Regression analysis of competing risks is typically undertaken via a proportional hazards model adapted on cause-specific hazard or subdistribution hazard. In this article, we propose an alternative approach that treats competing events as distinct outcomes in a mixture. We consider semiparametric accelerated failure time models for the cause-conditional survival function that are combined through a multinomial logistic model within the cure-mixture modeling framework. The cure-mixture approach to competing risks provides a means to determine the overall effect of a treatment and insights into how this treatment modifies the components of the mixture in the presence of a cure fraction. The regression and nonparametric parameters are estimated by a nonparametric kernel-based maximum likelihood estimation method. Variance estimation is achieved through resampling methods for the kernel-smoothed likelihood function. Simulation studies show that the procedures work well in practical settings. Application to a sarcoma study demonstrates the use of the proposed method for competing risk data with a cure fraction.

Early Fibrinolysis Associated with Hemorrhagic Progression Following Traumatic Brain Injury.

BACKGROUND: Progressive hemorrhagic injury (PHI) is common in patients with severe traumatic brain injury (TBI) and is associated with worse outcomes. PHI pathophysiology remains poorly understood and difficult to predict. We performed an exploratory analysis aimed at identifying markers in need of further investigation to establish their predictive value in PHI following TBI. METHODS: We performed a retrospective chart review of prospectively collected data from 424 highest-level activation trauma patients from January 2012 through December 2013. Patients with severe TBI, defined as head acute injury scale (AIS) score ≥3 and intracranial hemorrhage (ICH) on initial CT, were included. Stable hemorrhage (SH) and PHI was determined by measuring ICH expansion on repeat CT within 6 h. Of 424 patients evaluated, 72 met inclusion criteria. Twenty-five patients had repeated samples available and were dichotomized into SH (n = 6, 24%) and PHI (n = 19, 76%). Levels of plasminogen, urokinase and tissue plasminogen activators (uPA, tPA), plasminogen activator inhibitor-1, α2-antiplasmin (α2AP), and D-Dimers (DD) were measured upon admission and 2, 4, and 6 h later. RESULTS: Longitudinal models identified tPA and DD as positively associated and α2AP inversely associated with PHI. High DD levels are strongly associated with developing PHI over time. Using the full TBI cohort of N = 72, receiver operating curve analysis provided a cutoff of 3.04 µg/mL admission DD to distinguish SH from PHI patients. CONCLUSION: Our findings support a relationship between markers of fibrinolysis in polytrauma patients with severe TBI and PHI, warranting further investigation into the potential for novel, predictive biomarkers.

Estimating the ratio of multivariate recurrent event rates with application to a blood transfusion study.

In comparative effectiveness studies of multicomponent, sequential interventions like blood product transfusion (plasma, platelets, red blood cells) for trauma and critical care patients, the timing and dynamics of treatment relative to the fragility of a patient's condition is often overlooked and underappreciated. While many hospitals have established massive transfusion protocols to ensure that physiologically optimal combinations of blood products are rapidly available, the period of time required to achieve a specified massive transfusion standard (e.g. a 1:1 or 1:2 ratio of plasma or platelets:red blood cells) has been ignored. To account for the time-varying characteristics of transfusions, we use semiparametric rate models for multivariate recurrent events to estimate blood product ratios. We use latent variables to account for multiple sources of informative censoring (early surgical or endovascular hemorrhage control procedures or death). The major advantage is that the distributions of latent variables and the dependence structure between the multivariate recurrent events and informative censoring need not be specified. Thus, our approach is robust to complex model assumptions. We establish asymptotic properties and evaluate finite sample performance through simulations, and apply the method to data from the PRospective Observational Multicenter Major Trauma Transfusion study.

Safety and efficacy of coil embolization of the septal perforator for septal ablation in patients with hypertrophic obstructive cardiomyopathy.

OBJECTIVE: The objective of this study was to evaluate safety, efficacy, and durability of coil embolization of the major septal perforator of the left anterior descending coronary artery in patients with hypertrophic obstructive cardiomyopathy (HOCM). BACKGROUND: The long-term effect of coil embolization therapy in HOCM patients is not well defined. METHODS: We evaluated 24 symptomatic HOCM patients in a single center who underwent coil embolization of the septal perforator artery(ies). RESULTS: Twenty-four patients on optimal medical therapy presented with NYHA functional class III (75%) or IV (25%) underwent the procedure. The procedure was successful in 22 patients, with significant reduction in left ventricular outflow tract (LVOT) gradient. The functional class significantly improved to class I (54.2%) or II (41.7%) (P < = 0.01). The LVOT gradient was significantly lower during follow up echocardiography (21.3 ± 19 vs. 81.3 ± 41 mm Hg; P < = 0.01). Interventricular septal thickness decreased over time (16.3 ± 3 vs. 18.5 ± 2 mm, P< = 0.01). The procedure was aborted in one of the patients after the third coil prolapsed from the septal perforator in to the left anterior descending artery. The coil was effectively snared out. Three patients required additional coil placement in the second major septal perforator. New permanent pacemaker placement was required in one patient. However, three patients underwent ICD implantation at follow up due to ventricular arrhythmias. CONCLUSIONS: The results of this study suggest that the use of coil embolization for septal ablation is safe, effective, and durable in patients with symptomatic HOCM. © 2016 Wiley Periodicals, Inc.

Recurrent event frailty models reduced time-varying and other biases in evaluating transfusion protocols for traumatic hemorrhage.

OBJECTIVE: Transfusion research seeks to improve survival for severely injured and hemorrhaging patients using optimal plasma and platelet ratios over red blood cells (RBCs). However, most published studies comparing different ratios are plagued with serious bias and ignore time-varying effects. We applied joint recurrent event frailty models to increase validity and clinical utility. STUDY DESIGN AND SETTING: Using the PRospective Observational Multicenter Major Trauma Transfusion study data, our joint random-effects models estimated the association of (1) clinical covariates with transfusion rate intensities and (2) varying plasma:RBC and platelet:RBC ratios with survival over the 24 hours after hospital admission. Along with survival time, baseline patient vital signs, laboratory values, and longitudinal data on types and volumes of transfusions were included. RESULTS: Baseline systolic blood pressure, heart rate, pH, and hemoglobin were significantly associated with RBC transfusion rates. Increased transfusion rates (per hour) of plasma (P = 0.05), platelets (P < 0.001), or RBCs were associated with increased 24-hour mortality. Higher ratios of plasma:RBC (P = 0.107) and platelet:RBC (P < 0.001) were associated with reduced mortality in a time-varying pattern (P < 0.001). CONCLUSIONS: The proposed joint analysis of transfusion rates and ratios offers a more valid statistical approach to evaluate survival effects in the presence of informative censoring by early death.

A two-stage approach for dynamic prediction of time-to-event distributions.

Dynamic prediction uses longitudinal biomarkers for real-time prediction of an individual patient's prognosis. This is critical for patients with an incurable disease such as cancer. Biomarker trajectories are usually not linear, nor even monotone, and vary greatly across individuals. Therefore, it is difficult to fit them with parametric models. With this consideration, we propose an approach for dynamic prediction that does not need to model the biomarker trajectories. Instead, as a trade-off, we assume that the biomarker effects on the risk of disease recurrence are smooth functions over time. This approach turns out to be computationally easier. Simulation studies show that the proposed approach achieves stable estimation of biomarker effects over time, has good predictive performance, and is robust against model misspecification. It is a good compromise between two major approaches, namely, (i) joint modeling of longitudinal and survival data and (ii) landmark analysis. The proposed method is applied to patients with chronic myeloid leukemia. At any time following their treatment with tyrosine kinase inhibitors, longitudinally measured BCR-ABL gene expression levels are used to predict the risk of disease progression. Copyright © 2016 John Wiley & Sons, Ltd.

A joint latent class analysis for adjusting survival bias with application to a trauma transfusion study.

There is no clear classification rule to rapidly identify trauma patients who are severely hemorrhaging and may need substantial blood transfusions. Massive transfusion (MT), defined as the transfusion of at least 10 units of red blood cells within 24 h of hospital admission, has served as a conventional surrogate that has been used to develop early predictive algorithms and establish criteria for ordering an MT protocol from the blood bank. However, the conventional MT rule is a poor proxy, because it is likely to misclassify many severely hemorrhaging trauma patients as they could die before receiving the 10th red blood cells transfusion. In this article, we propose to use a latent class model to obtain a more accurate and complete metric in the presence of early death. Our new approach incorporates baseline patient information from the time of hospital admission, by combining respective models for survival time and usage of blood products transfused within the framework of latent class analysis. To account for statistical challenges, caused by induced dependent censoring inherent in 24-h sums of transfusions, we propose to estimate an improved standard via a pseudo-likelihood function using an expectation-maximization algorithm with the inverse weighting principle. We evaluated the performance of our new standard in simulation studies and compared with the conventional MT definition using actual patient data from the Prospective Observational Multicenter Major Trauma Transfusion study. Copyright © 2015 John Wiley & Sons, Ltd.

A joint latent class model for classifying severely hemorrhaging trauma patients.

BACKGROUND: In trauma research, "massive transfusion" (MT), historically defined as receiving ≥10 units of red blood cells (RBCs) within 24 h of admission, has been routinely used as a "gold standard" for quantifying bleeding severity. Due to early in-hospital mortality, however, MT is subject to survivor bias and thus a poorly defined criterion to classify bleeding trauma patients. METHODS: Using the data from a retrospective trauma transfusion study, we applied a latent-class (LC) mixture model to identify severely hemorrhaging (SH) patients. Based on the joint distribution of cumulative units of RBCs and binary survival outcome at 24 h of admission, we applied an expectation-maximization (EM) algorithm to obtain model parameters. Estimated posterior probabilities were used for patients' classification and compared with the MT rule. To evaluate predictive performance of the LC-based classification, we examined the role of six clinical variables as predictors using two separate logistic regression models. RESULTS: Out of 471 trauma patients, 211 (45 %) were MT, while our latent SH classifier identified only 127 (27 %) of patients as SH. The agreement between the two classification methods was 73 %. A non-ignorable portion of patients (17 out of 68, 25 %) who died within 24 h were not classified as MT but the SH group included 62 patients (91 %) who died during the same period. Our comparison of the predictive models based on MT and SH revealed significant differences between the coefficients of potential predictors of patients who may be in need of activation of the massive transfusion protocol. CONCLUSIONS: The traditional MT classification does not adequately reflect transfusion practices and outcomes during the trauma reception and initial resuscitation phase. Although we have demonstrated that joint latent class modeling could be used to correct for potential bias caused by misclassification of severely bleeding patients, improvement in this approach could be made in the presence of time to event data from prospective studies.