Unobserved Risk Research Articles

The risk screening effect of digital insurance distribution

AbstractExtant research holds that digital sales channels advance competition and service or goods availability but rarely details the attendant information asymmetry. Leveraging a large unique dataset, this study examines a specific case in which consumers have a choice between offline and digital channels for insurance purchases. We find that digital channels screen in consumers with lower unobserved risk. For term life, endowment and disease insurance products, the average risks of the policies purchased through digital channels were significantly lower than those purchased offline after controlling for all observed risk characteristics. This risk screening effect mainly comes from the inclusion of new low‐risk enrollees. As a consequence, digital channels exhibit lower information asymmetry and greater profitability compared to offline channels.

Owner Guarantees, Observed and Unobserved Risks, and Bank Lending Spreads

AbstractIn this paper, we examine the influence of owner guarantees on loan rates. Other studies on the influence of business collateral have used survey and credit register data to establish this link. They support it with the pre-lending information theory and post-lending incentive theory and find that the latter predominates the former. Thus, we measure observed and unobserved risks on bank lending spreads negotiated over the counter to show the co-existence of both theories. The main findings and robustness tests show that owner guarantees reduce bank lending spreads for safe but not risky firms, supporting the pre-lending information theory rather than the post-lending incentive theory. These findings underscore the role of owner guarantees in allaying the information problems the bank confronts prior to lending.

Effect of risk and incentives on contracts: The case of farm operators in Iowa

AbstractLongitudinal data on Iowa landowner–operator contracts are used to examine how the mean, variance, and skewness of expected farmland returns affected contract terms from 2008 to 2019. To control for unobserved operator risk preferences and ability as a fixed effect, the sample includes operators with multiple contracts. Our empirical work shows that farmland with lower variance is more likely to be rented for cash, whereas land with highly variable returns is more likely to be custom contracted. [EconLit Citations: L14, Q15]

COMPOSITE SCORES FOR TRANSPLANT CENTER EVALUATION: A NEW INDIVIDUALIZED EMPIRICAL NULL METHOD.

Risk-adjusted quality measures are used to evaluate healthcare providers with respect to national norms while controlling for factors beyond their control. Existing healthcare provider profiling approaches typically assume that the between-provider variation in these measures is entirely due to meaningful differences in quality of care. However, in practice, much of the between-provider variation will be due to trivial fluctuations in healthcare quality, or unobservable confounding risk factors. If these additional sources of variation are not accounted for, conventional methods will disproportionately identify larger providers as outliers, even though their departures from the national norms may not be "extreme" or clinically meaningful. Motivated by efforts to evaluate the quality of care provided by transplant centers, we develop a composite evaluation score based on a novel individualized empirical null method, which robustly accounts for overdispersion due to unobserved risk factors, models the marginal variance of standardized scores as a function of the effective sample size, and only requires the use of publicly-available center-level statistics. The evaluations of United States kidney transplant centers based on the proposed composite score are substantially different from those based on conventional methods. Simulations show that the proposed empirical null approach more accurately classifies centers in terms of quality of care, compared to existing methods.

Mental health and abortions among young women: time-varying unobserved heterogeneity, health behaviors, and risky decisions

In this paper, we provide causal evidence on abortions and risky health behaviors as determinants of mental health development among young women. Using administrative in- and outpatient records from Sweden, we apply a novel grouped fixed-effects estimator proposed by Bonhomme and Manresa (2015) to allow for time-varying unobserved heterogeneity. We show that the positive association obtained from standard estimators shrinks to zero once we control for grouped unobserved heterogeneity that varies across ages. We estimate the group profiles of unobserved heterogeneity, which reflect differences in unobserved risk to be diagnosed with a mental health condition. In addition, we analyze mental health development and risky health behaviors other than unintended pregnancies across groups. Our results suggest that these are determined by the same type of unobserved heterogeneity, which we attribute to the same unobserved process of decision-making. We develop and estimate a theoretical model of risky choices and mental health that is consistent with our empirical results. In the model, mental health disparity across groups is generated by different degrees of self-control problems. Our findings imply that mental health concerns cannot be used to justify restrictive abortion policies.

Food Insecurity, Disability and Age

Abstract Background Food insecurity is a growing public health concern. Disabled people are at higher risk of food insecurity than non-disabled people. Evidence suggests this diminishes for older people. We examine how the relationship between disability and food insecurity varies at different ages and explore the extent to which socio-economic factors account for and modify the relationship. Methods Data came from the fourth and fifth waves of the UK's Food Standards Agency's Food and You survey (n = 6,187), which contain data on disability and household food insecurity. We explored predicted probabilities of food insecurity from logistic regression models between disabled and non-disabled people for 10-year age cohorts. Stratified analyses, further explored differences in vulnerability to food insecurity at working age and 65+. Results Predicted probability of food insecurity for disabled people declined with age, as did the gap in risk between disabled and non-disabled people. The predicted probability of food insecurity among disabled people under 65 was 23% (95% CI: 19%-27%), whereas for disabled people 65+, it was 5% (95% CI: 3%-7%). Among people of working age, socio-economic factors explained some, but not all, of the increased risk of food insecurity associated with disability. Disabled people in full-time work were still significantly more likely to be food insecure than non-disabled people in work: 15% (95% CI: 11%-20%) vs 8 % (95% CI: 6%-9%) respectively. Among older adults, disabled people with savings were not at significantly higher risk of food insecurity, whereas those without savings were: 0.4% (95% CI: 0.4%-1%) vs 5% (95% CI: 3%-7%), respectively. Conclusions Findings indicate the increased risk of food insecurity for disabled people is partially explained by socio-economic factors. However, there is a need for further research to explain unobserved risk factors for food insecurity among disabled people of working age especially. Key messages • Socioeconomic factors play an important role in disabled people's increased risk of food insecurity. • Observed socioeconomic factors alone do not fully explain increased risk of food insecurity among working age disabled people. More research is needed.

Optimization method of PMU placements base on costs and risk assessments of state estimation

In the context of vigorous development of new energy sources, optimizing the arrangement of phase measurement units (PMUs) is crucial to realize real-time data monitoring for distribution network (DN) condition assessment. Firstly, we propose a multi-objective model for optimal PMU placements. Secondly, we classified the buses based on the Jacobi matrix and the grid structure, and calculated the unobservable risk probability of the system by estimating the state of the PMUs installed on the buses, so as to obtain the observability of the system. Finally, two simulations of PMU placements in the DN implementing the IEEE39 bus are designed and the placement schemes are determined through solving the non-dominated Pareto solution sets by NSGA-II. The verifications show that the model reduces the placement number of PMUs, resulting in a cost reduction of 35.4% and an unobservable risk indicator δ reduction of 13.8%, and achieves an optimal placement of PMUs.

A lean additive frailty model: With an application to clustering of melanoma in Norwegian families.

Additive frailty models are used to model correlated survival data. However, the complexity of the models increases with cluster size to the extent that practical usage becomes increasingly challenging. We present a modification of the additive genetic gamma frailty (AGGF) model, the lean AGGF (L-AGGF) model, which alleviates some of these challenges by using a leaner additive decomposition of the frailty. The performances of the models were compared and evaluated in a simulation study. The L-AGGF model was used to analyze population-wide data on clustering of melanoma in 2391125 two-generational Norwegian families, 1960-2015. Using this model, we could analyze the complete data set, while the original model limited the analysis to a restricted data set (with cluster sizes ). We found a substantial clustering of melanoma in Norwegian families and large heterogeneity in melanoma risk across the population, where 52% of the frailty was attributed to the 10% of the population at highest unobserved risk. Due to the improved scalability, the L-AGGF model enables a wider range of analyses of population-wide data compared to the AGGF model. Moreover, the methods outlined here make it possible to perform these analyses in a computationally efficient manner.

Potential Benefits in Remapping the Special Flood Hazard Area: Evidence from the U.S. Housing Market

A typical U.S. homebuyer's understanding of whether a property faces flood risk is based on whether the property is located inside the National Flood Insurance Program's (NFIP) Special Flood Hazard Area (SFHA). The SFHA boundary, however, may bias homebuyers’ perceptions of flood risk relative to unobserved true risk because the SFHA is an incomplete, and sometimes inaccurate, representation of flood hazards. Using a national dataset of property transactions, flood hazard data from the First Street Foundation, state-level measures of flood disclosure laws, and a spatially restricted triple-difference design, we distinguish capitalization effects of policy-driven (SFHA) versus quasi-objective (First Street) indicators of flood hazard. We identify these effects by assessing thousands of local spatial interactions between property SFHA designations, measures of quasi-objective flood hazard, and being in a state that mandates flood history disclosures. Being inside the SFHA generates a risk discount, but the signal is muted relative to underlying hazard exposure. Further, the SFHA signal can result in inefficient discounts for properties erroneously mapped in the SFHA. Disclosure requirements about flood history accentuate price effects for hazardous properties and result in more adequate risk internalization. In the absence of disclosures, however, we find either no or a weak risk signal for houses outside the SFHA that face flood hazard. Our results highlight potential benefits of updating SFHA boundaries to include all houses that may experience flooding, and argue in favor of requiring flood-related disclosures from sellers to improve the market's ability to internalize flood risk.

Exploring the differentiation of behavioural and emotional problems across childhood: A prospective longitudinal cohort study.

An individual's overall burden of behavioural and emotional problems across childhood is associated with increased likelihood of later mental health conditions. However, the relative extent of behavioural versus emotional problems - that is, the extent to which the domains are differentiated from one another - may provide additional information about who is at risk of developing a mental health condition. Here, we seek to validate differentiation as an independent predictor of later mental health conditions, and to explore its aetiology. We analysed data from ~79,000 children in the population-based Norwegian Mother, Father, and Child Cohort Study, and linked health-care registries. In preregistered analyses, we modelled the extent and rate of differentiation of behavioural and emotional problems between ages 1.5-5years, and estimated associations with later symptoms (age 8) and diagnoses (after age 8). We also explored the aetiology of differentiation by estimating associations with early life exposures and, in a subset of 23,945 full siblings, assessing the impact of accounting for unobserved familial confounding. Differentiation of behavioural and emotional problems was associated with later symptoms and diagnoses of mental health conditions, independent of total problems. Maternal at-risk drinking (β=0.04 [0.02, 0.06]) and parental relationship problems (β=0.04 [0.02, 0.05]) were associated with higher behavioural relative to emotional problems at age 5. Maternal prenatal distress (|β|=0.04 [0.03, 0.06]), concurrent distress (|β|=0.04 [0.02, 0.06]) and parental education (|β|=0.05 [0.04, 0.07]) predicted higher emotional relative to behavioural problems at age 5. Estimates for maternal prenatal distress and at-risk drinking were consistent across both unadjusted and adjusted analyses accounting for unobserved familial risk. Differentiation of behavioural and emotional problems in early childhood represents a valid source of inter-individual variability linked to the later emergence of psychopathology and may be relevant for early detection and prevention strategies for mental health.

Known Unknowns: The Inevitability of Cyber Attacks

As described by Former U.S. Secretary of Defense, Donald Rumsfeld in his 2011 book, Known and Unknown, “there are many things of which we are completely unaware—in fact, there are things of which we are so unaware, we don’t even know we are unaware of them. Throughout history the world has faced numerous catastrophic events that were not foreseen but in hindsight were discoverable including the devastating effects of Pearl Harbor, and the September 11 terrorist attacks. More recently, the potential for catastrophic loss has been magnified in the 2020 Solar Winds and 2021 Colonial Pipeline cyber-attacks. We may not know when or how these events will occur or how much damage or destruction will occur, but we do know that these events are possible. The literature differentiates between events that occur totally by surprise, and outcomes or events that actors have identified as possibly existing but do not know whether they will take place or not. The aim of this paper is to provide insight, based on an empirical review of selected attacks both within and outside the cyber space literature to uncover the underlying risk, uncertainty, and complexity that may have been known but not seriously considered by those who had the knowledge and capability to investigate the warning signs. Based on the case study analysis, this paper will present the reasons for inaction and how we can learn from these experiences. The following two theories – institutionalization and rationalization have been found to provide some reasons for the occurrence of behaviors which increase the possibility of unobserved risks. In this paper we explore these theories through case study analysis and propose a framework consisting of four concepts for increasing awareness of these situations.

Improving risk classification and ratemaking using mixture‐of‐experts models with random effects

AbstractIn the underwriting and pricing of nonlife insurance products, it is essential for the insurer to utilize both policyholder information and claim history to ensure profitability and proper risk management. In this paper, we apply a flexible regression model with random effects, called theMixed Logit‐weighted Reduced Mixture‐of‐Experts, which leverages both policyholder information and their claim history, to categorize policyholders into groups with similar risk profiles, and to determine a premium that accurately captures the unobserved risks. Estimates of model parameters and the posterior distribution of random effects can be obtained by a stochastic variational algorithm, which is numerically efficient and scalable to large insurance portfolios. Our proposed framework is shown to outperform the classical benchmark models (Logistic and Lognormal GL(M)M) in terms of goodness‐of‐fit to data, while offering intuitive and interpretable characterization of policyholders' risk profiles to adequately reflect their claim history.

Prophylactic Anti-Seizure Medication Use in Newly Diagnosed Brain Tumor Patients (P3-13.002)

<h3>Objective:</h3> To evaluate the use of prophylactic anti-seizure medications (ASM) in seizure-naïve patients with newly diagnosed brain tumors <h3>Background:</h3> Many physicians continue to prescribe ASM to seizure-naïve brain tumor patients, even though treatment guidelines recommend against such practice. <h3>Design/Methods:</h3> In this retrospective observational study, electronic medical records from the TriNetX Dataworks – USA Network were used to identify patients with a record of brain tumor diagnosis and no evidence of seizure diagnosis from 2012 to 2017. Patients who were prescribed a prophylactic ASM were propensity-score (PS) matched with those who were not at a 1:1 ratio on age, sex, race, ethnicity, tumor types, comorbidities, prescription medications, and procedures. Prophylactic ASM use was defined as the prescription of ≥1 ASM from 30 days before the first brain tumor diagnosis and up to 30 days afterwards. Hazard ratios for seizure development and mortality were assessed using Cox proportional hazards regression models. The first brain tumor diagnosis date was considered the index date, and patients were followed up to 5 years from their first prescription record, death, or whichever occurred first. <h3>Results:</h3> Of the 117,834 seizure-naïve patients who were newly diagnosed with brain tumor, 18,963 (16.1%) were prescribed a prophylactic ASM. Of the 14,238 patients in each PS-matched cohort, 778 of those on prophylaxis (5.5%) and 452 of those not on prophylaxis (3.0%) developed a seizure during the follow-up period (Hazard Ratio: 2.3 [95% CI: 2.0, 2.6]). Mortality hazard was 1.1 times (95% CI: 1.1, 1.2) higher in those on prophylactic ASM. <h3>Conclusions:</h3> Seizure-naïve patients with brain tumor who were prescribed a prophylactic ASM had a higher hazard of developing a seizure compared with those who were not. Unobserved risk factors that were not controlled for may explain the apparent analysis results. Clinical trials are warranted to further examine this controversial disease management approach. <b>Disclosure:</b> Miss Jang has received personal compensation for serving as an employee of Eisai Inc.. Miss Jang has received research support from Eisai Inc.. Dr. Peters has received personal compensation in the range of $0-$499 for serving on a Scientific Advisory or Data Safety Monitoring board for Vivacitas Oncology. Dr. Peters has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Servier. Dr. Peters has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Sapience. Dr. Peters has received personal compensation in the range of $0-$499 for serving on a Scientific Advisory or Data Safety Monitoring board for NuVox Therapeutics. Dr. Peters has received personal compensation in the range of $0-$499 for serving on a Scientific Advisory or Data Safety Monitoring board for Cordance Medical. Dr. Peters has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for ONO Pharmaceutical. Dr. Peters has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Blue Earth Diagnostics. The institution of Dr. Peters has received research support from Novocure. The institution of Dr. Peters has received research support from Biomimetix. The institution of Dr. Peters has received research support from Servier. The institution of Dr. Peters has received research support from Varian. The institution of Dr. Peters has received research support from Sapience. Dr. Peters has received personal compensation in the range of $500-$4,999 for serving as a Educational Lecturer with Eisai. Dr. Ngo has received personal compensation for serving as an employee of Eisai Inc.. An immediate family member of Dr. Morita-Sherman has received personal compensation for serving as an employee of Forrester Research. Dr. Morita-Sherman has received personal compensation for serving as an employee of Eisai Inc.. An immediate family member of Dr. Morita-Sherman has stock in Forrester Research. The institution of Dr. Morita-Sherman has received research support from Cleveland Clinic.

OP10 Proctocolectomy with permanent ileostomy is associated with better transplant-free survival in patients with primary sclerosing cholangitis: a retrospective cohort study

Abstract Background Inflammatory bowel disease (IBD) is the strongest known risk factor for primary sclerosing cholangitis (PSC) and seems associated with a worse prognosis. After liver transplantation, colectomy is associated with decreased rates of recurrent PSC and better graft survival.1 At the same time, the effect of colectomy on the disease course of PSC remains unclear.2 Available evidence depends either on small patient numbers, may be confounded by unobserved risk factors or colectomy was not considered as time-depending covariate. The aim of our study is to estimate the effect of colectomy on the disease course of PSC by taking time dependency and known risk factors into account. Methods We conducted a retrospective cohort study between January 1st 2008 and December 31st 2021 based on the Dutch population-based EpiPSC2 cohort. Transplant-free survival was defined as liver transplantation or PSC-related death (excluding colorectal carcinoma). The effect of colectomy was determined as time depending variable via Cox proportional hazards regression and was corrected for the following known risk factors; sex, age at diagnosis, large or small duct PSC, features of auto-immune hepatitis, time depending IBD status (i.e. no IBD, ulcerative colitis (UC), Crohn’s disease (CD) or IBD Unspecified (IBD-U)), and tertiary centre inclusion. Effect of colectomy was stratified for extent (i.e. hemi-, subtotal, or proctocolectomy with pouch or ileostomy). Indication of colectomy (i.e. active disease, dysplasia and colorectal carcinoma or other) was also considered as predictor variable. Events in the first year after colectomy were censored due to a presumed delayed effect of colectomy on disease course. Results Of the 1341 patients included, 912(68%) were diagnosed with IBD and 187(14%) had undergone a colectomy. During a total follow-up of 14,393 patient years, 370(28%) patients underwent liver transplantation and 318(24%) patients died due to a PSC-related cause. Median transplant-free survival time was 20 years. Hazard ratio (HR) was significantly decreased in patients who had undergone proctocolectomy with ileostomy (HR 0.47(0.24-0.93)) compared to patients without colectomy. This effect was absent in the group with any other type of colectomy (HR 0.90(0.67-1.19)). Indication of colectomy had no apparent effect. Presence of UC, CD or IBD-U was not associated with a worse prognosis compared to absence of IBD (HR 0.99(0.81-1.21), HR 0.85(0.64-1.13), HR 0.75(0.43-1.29)). Conclusion Our data suggest that proctocolectomy with ileostomy confers protection for liver transplantation and PSC-related death. This points towards the putative pathophysiologic role of the colonic microbiota in PSC and may open avenues to a better understanding of the aetiology of PSC.

Bayesian joint spatial modelling of anemia and malaria in Guinea

&lt;abstract&gt;&lt;p&gt;In sub-Saharan Africa, anemia and malaria are the leading causes of morbidity and mortality among children under the age of five years. Guinea is one of the countries where the two diseases have devastating effects. Both of these diseases have been studied separately, but the two diseases exhibit inherent dependence between them, therefore, modelling them in isolation negates practical reality. This study aims at jointly estimating the spatial linear correlation between anemia and malaria, as well as to investigate the differences in contextual, socioeconomic and demographic factors affecting morbidity among children under five years in Guinea. Statistical approaches are used to handle modelling of binary outcomes with allowance for spatial components and joint responses. In particular, a latent model approach is proposed in the methodology to investigate the linear correlation between anemia and malaria allowing for spatial and non-spatial effects. All the parameters are estimated using Bayesian approach based on Markov chain Monte Carlo (MCMC) technique. According to the findings, 76.15% of children under the age of five years in Guinea were anemic, and 14.31% had malaria. Furthermore, the results showed that the child's malaria status is significantly associated with the place of residence, his/her age and ownership of television as an indicator of well being. In terms of anemia in children, there was a significant association with age, mother's education level and ethnicity group of the household head. The Nzerekore region, had both high malaria and anemia prevalences in children under five years. The latent model results showed that there was weak positive correlation between anemia and malaria in Nzerekore and Boke regions. Based on the shared component model, there was a significant unobserved risk factor that both diseases share.&lt;/p&gt;&lt;/abstract&gt;

Exchange rate predictability, risk premiums, and predictive system

Uncovered interest rate parity (UIP), a basic assumption in many theoretical models, is known to perform poorly in forecasting exchange rate movements, especially in the short run. One possible reason for this failure is the existence of unobservable risk premium. We estimate the unobservable risk premium with a Bayesian approach having the risk premium as a latent variable and the implied volatility of at-the-money currency options as an imperfect predictor. We find in most cases that expected exchange rate changes, constructed from forward-spot differentials and estimated risk premiums, track actual exchange rate changes more closely than do the fitted values of the predictive regression (i.e. the Fama regression). An out-of-sample analysis reveals that adding the estimated risk premium greatly improves the short-run predictability of exchange rates in general. These findings strongly suggest that the risk premium is important in understanding the dynamics of exchange rate and the UIP puzzle.

Risk of recurrent stillbirth and neonatal mortality: mother-specific random effects analysis using longitudinal panel data from Indonesia (2000 – 2014)

BackgroundDespite significant government investments to improve birth outcomes in low and middle-income countries over the past several decades, stillbirth and neonatal mortality continue to be persistent public health problems. While they are different outcomes, there is little evidence regarding their shared and unique population-level risk factors over a mother’s reproductive lifespan. Data gaps and measurement challenges have left several areas in this field unexplored, especially assessing the risk of stillbirth or neonatal mortality over successive pregnancies to the same woman. This study aimed to assess the risk of stillbirth and neonatal mortality in Indonesia during 2000–2014, using maternal birth histories from the Indonesia Family Life Survey panel data.MethodsData from three panels were combined to create right-censored birth histories. There were 5,002 unique multiparous mothers with at least two singleton births in the sample. They reported 12,761 total births and 12,507 live births. Random effects (RE) models, which address the dependency of variance in births to the same mother, were fitted assuming births to the same mother shared unobserved risk factors unique to the mother.ResultsThe main finding was that there having had a stillbirth increased the odds of another stillbirth nearly seven-fold and that of subsequent neonatal mortality by over two-fold. Having had a neonatal death was not associated with a future neonatal death. Mothers who were not educated and nullipara were much more likely to experience a neonatal death while mothers who had a prior neonatal death had no risk of another neonatal death due to unmeasured factors unique to the mother.ConclusionsThe results suggest that for stillbirths, maternal heterogeneity, as explained by a prior stillbirth, could capture underlying pathology while the relationship between observed risk factors and neonatal mortality could be much more dependent on context. Establishing previous adverse outcomes such as neonatal deaths and stillbirth could help identify high-risk pregnancies during prenatal care, inform interventions, and improve health policy.

First Episodes of Norovirus and Sapovirus Gastroenteritis Protect Against Subsequent Episodes in a Nicaraguan Birth Cohort.

Norovirus and sapovirus cause a large burden of acute gastroenteritis (AGE) in young children. We assessed protection conferred by norovirus and sapovirus AGE episodes against future episodes. Between June 2017 and July 2018, we recruited 444 newborns in León, Nicaragua. Weekly household surveys identified AGE episodes over 36 months, and AGE stools were tested by reverse transcriptase-quantitative polymerase chain reaction (RT-qPCR) for norovirus genogroup (G)I/GII and sapovirus. We used recurrent-event Cox models and negative control methods to estimate protection conferred by first episodes, controlling for observed and unobserved risk factors, respectively. Sapovirus episodes conferred a 69% reduced hazard of subsequent episodes using the negative control method. Norovirus GI (hazard ratio [HR] = 0.67; 95% confidence interval [CI] = 0.31, 1.3) and GII (HR = 0.20; 95% CI = 0.04, 0.44) episodes also appeared highly protective. Protection against norovirus GII was enhanced following two episodes. Evidence of natural immunity in early childhood provides optimism for the future success of pediatric norovirus and sapovirus vaccines.

Examining infant and child death clustering among families in the cross-sectional and nationally representative Bangladesh Demographic and Health Survey 2017–2018

ObjectivesWe aim to examine the phenomenon of infant and child death clustering while considering the unobserved heterogeneity (frailty) at the family level.Design, setting, and participantsWe analysed Bangladesh Demographic and Health...

Predicting Psychosis: Field Moves From Assessment of Clinical Risk to Search for Biomarkers

Predicting Psychosis: Field Moves From Assessment of Clinical Risk to Search for Biomarkers