Search

With the availability of massive amounts of data from electronic health records and registry databases, incorporating time-varying patient information to improve risk prediction has attracted great attention. To exploit the growing amount of predictor information over time, we develop a unified framework for landmark prediction using survival tree ensembles, where an updated prediction can be performed when new information becomes available. Compared to conventional landmark prediction with fixed landmark times, our methods allow the landmark times to be subject-specific and triggered by an intermediate clinical event. Moreover, the nonparametric approach circumvents the thorny issue of model incompatibility at different landmark times. In our framework, both the longitudinal predictors and the event time outcome are subject to right censoring, and thus existing tree-based approaches cannot be directly applied. To tackle the analytical challenges, we propose a risk-set-based ensemble procedure by averaging martingale estimating equations from individual trees. Extensive simulation studies are conducted to evaluate the performance of our methods. The methods are applied to the Cystic Fibrosis Foundation Patient Registry (CFFPR) data to perform dynamic prediction of lung disease in cystic fibrosis patients and to identify important prognosis factors.

In many populations, the peak period of incidence of type 1 diabetes (T1D) has been observed to be around 10–14 years of age, coinciding with puberty, but direct evidence of the role of puberty in the development of T1D is limited. We therefore aimed to investigate whether puberty and the timing of its onset are associated with the development of islet autoimmunity (IA) and subsequent progression to T1D. A Finnish population-based cohort of children with HLA-DQB1-conferred susceptibility to T1D was followed from 7 years of age until 15 years of age or until a diagnosis of T1D (n = 6920). T1D-associated autoantibodies and growth were measured at 3- to 12-month intervals, and pubertal onset timing was assessed based on growth. The analyses used a three-state survival model. IA was defined as being either positive for islet cell antibodies plus at least one biochemical autoantibody (ICA + 1) or as being repeatedly positive for at least one biochemical autoantibody (BC1). Depending on the IA definition, either 303 (4.4%, ICA + 1) or 435 (6.3%, BC1) children tested positive for IA by the age of 7 years, and 211 (3.2%, ICA + 1)) or 198 (5.3%, BC1) developed IA during follow-up. A total of 172 (2.5%) individuals developed T1D during follow-up, of whom 169 were positive for IA prior to the clinical diagnosis. Puberty was associated with an increase in the risk of progression to T1D, but only from ICA + 1-defined IA (hazard ratio 1.57; 95% confidence interval 1.14, 2.16), and the timing of pubertal onset did not affect the association. No association between puberty and the risk of IA was detected. In conclusion, puberty may affect the risk of progression but is not a risk factor for IA.

The analysis of recurrent events in the presence of terminal events requires special attention. Several approaches have been suggested for such analyses either using intensity models or marginal models. When analysing treatment effects on recurrent events in controlled trials, special attention should be paid to competing deaths and their impact on interpretation. This paper proposes a method that formulates a marginal model for recurrent events and terminal events simultaneously. Estimation is based on pseudo-observations for both the expected number of events and survival probabilities. Various relevant hypothesis tests in the framework are explored. Theoretical derivations and simulation studies are conducted to investigate the behaviour of the method. The method is applied to two real data examples. The bivariate marginal pseudo-observation model carries the strength of a two-dimensional modelling procedure and performs well in comparison with available models. Finally, an extension to a three-dimensional model, which decomposes the terminal event per death cause, is proposed and exemplified.

Purpose: To identify parameters that can predict the postoperative risk for progression of retinopathy to a vision-threatening stage after cataract surgery. This may optimize the timing of surgery and the postoperative follow-up strategy in diabetic patients. Methods: Multi-state survival analysis with death as competing risk was used to investigate how year of onset and age of onset of diabetes, gender, body mass index, HbA1c and blood pressure had affected the risk for developing diabetic macular oedema (DME) and proliferative diabetic retinopathy (PDR) among 2540 right eyes from 2797 diabetic patients operated for cataract on one or both eyes during 25 years until July 1. 2019. Results: Cataract surgery had been performed in 98.8% of patients reaching 90 years of age. The risk for developing both DME and PDR was increased by cataract surgery. The risk was highest during the first postoperative years and increased by pre-operative variability in HbA1c. The risk after more than 20 years postoperatively increased by increased cumulative HbA1c pre-operatively. The other studied risk factors contributed differently to the development of the two complications. Conclusions: Decision models for the timing of cataract surgery in diabetic patients should consider that the risk for developing vision-threatening retinopathy depends on follow-up time. Differences in the risk profiles for developing DME and PDR after cataract surgery support that the two complications should be regarded as separate late complications.

Multi-state models are frequently used when data come from subjects observed over time and where focus is on the occurrence of events that the subjects may experience. A convenient modeling assumption is that the multi-state stochastic process is Markovian, in which case a number of methods are available when doing inference for both transition intensities and transition probabilities. The Markov assumption, however, is quite strict and may not fit actual data in a satisfactory way. Therefore, inference methods for non-Markov models are needed. In this paper, we review methods for estimating transition probabilities in such models and suggest ways of doing regression analysis based on pseudo observations. In particular, we will compare methods using land-marking with methods using plug-in. The methods are illustrated using simulations and practical examples from medical research.

Developing electric ride-hailing vehicles (ERVs) is an integral part of promoting green transportation. To understand how ERVs operate, and in turn to improve their battery systems, we have developed a comprehensive, methodological framework to construct their utilization patterns and driving cycles. The methodology is based on a multi-state model (MSM). It builds one-day activity patterns of electric vehicles that belong in an ERV fleet. The driving patterns outline when (time of occurrence), for how long (duration), and under what battery conditions (state of charge) a driver typically decides to either operate or charge a vehicle. Two driving patterns, pointing toward two typical types of drivers, have been identified. By utilizing operational data in Xiamen, China, obtained from Didi Chuxing, we propose a novel method to build the first driving cycle series of ERVs, which can delineate how ERV drivers drive under the operating state. The cycle series consists of four driving cycles, synthesized by the micro-trip-based method using operational data from four time slices. Analysis against real data has shown that our method significantly outperforms existing methods, demonstrating the representativeness of the cycles. © 2020 Elsevier Ltd

© 2020. Elsevier, Background. Whether risk-reducing salpingo-oophorectomy (RRSO) in BRCA1/2 carriers reduces the breast cancer (BC) risk is conflicting, potentially due to methodological issues of prior analysis. We analysed the association between premenopausal RRSO and BC risk in BRCA1/2 carriers after adjusting for potential biases. Methods. We analysed data from 444 BRCA1 and 409 BRCA2 carriers under age 51 with no cancer prior to genetic testing or during first 6 months of surveillance (to avoid cancer-induced testing bias and prevalent-cancer bias). Observation started 6 months after genetic testing (to avoid event-free time bias), until BC diagnosis, risk-reducing mastectomy (RRM) or death. A multistate model with four states (non-RRSO, RRSO, RRM and BC) and five transitions was fitted to characterise outcomes and to calculate the BC risk reduction after premenopausal RRSO (before age 51). A systematic review was performed to assess the association between premenopausal RRSO and BC. Results. During a mean follow-up of 4.3 years, 96 women (11.3%) developed BC (54 BRCA1, 42 BRCA2). The risk of BC after premenopausal RRSO decreased significantly in BRCA1 carriers (hazard ratio (HR) = 0.45 [95% confidence interval (CI):0.22–0.92]), but was not conclusive in BRCA2 carriers (HR = 0.77 [95%CI:0.35–1.67]). The systematic review suggested that premenopausal RRSO is associated with a decrease of BC risk in both BRCA1 and BRCA2 carriers. Conclusions. Premenopausal RRSO was associated with BC risk reduction in BRCA1 carriers, which can help guide cancer risk-reducing strategies in this population. Longer follow-up and larger sample size may be needed to estimate the potential benefit in BRCA2 carriers., Peer Reviewed, Article signat per 14 autors/es: Neda Stjepanovic, Guillermo Villacampa, Kevin T. Nead, Sara Torres-Esquius, Guadalupe G. Melis, Katherine L. Nathanson, Alexandre Teule, Joan Brunet, Teresa R. y Cajal, Gemma Llort, Rodrigo Dienstmann, Montserrat Rue, Susan M. Domchek, Judith Balmaña, Postprint (published version)

Multi-state models are considered in the field of survival analysis for modelling illnesses that evolve through several stages over time. Multi-state models can be developed by applying several techniques, such as non-parametric, semi-parametric and stochastic processes, particularly Markov processes. When the development of an illness is being analysed, its progression is tracked periodically. Medical reviews take place at discrete times, and a panel data analysis can be formed. In this paper, a discrete-time piecewise non-homogeneous Markov process is constructed for modelling and analysing a multi-state illness with a general number of states. The model is built, and relevant measures, such as survival function, transition probabilities, mean total times spent in a group of states and the conditional probability of state change, are determined. A likelihood function is built to estimate the parameters and the general number of cut-points included in the model. Time-dependent covariates are introduced, the results are obtained in a matrix algebraic form and the algorithms are shown. The model is applied to analyse the behaviour of breast cancer. A study of the relapse and survival times of 300 breast cancer patients who have undergone mastectomy is developed. The results of this paper are implemented computationally with MATLAB and R.

Whether females participate in reproduction every year or rather skip years is pivotal information for understanding the dynamics of animal populations. In ectotherms this reproduction frequency might depend on factors like temperature, individual size, and prey availability. Here we studied the reproduction rate of the smooth snake Coronella austriaca at two latitudes: in Norway and The Netherlands. Reproductive and morphological data were collected in Norway (34 years) and The Netherlands (5 years). To account for differential resighting probabilities of pregnant and non-pregnant females, we performed multi-state mark-recapture analyses. Reproductive rate, quantified as the probability of participating in reproduction in consecutive years, was calculated for both pregnant and non-pregnant females. The results showed a clear difference in reproductive rate between the two latitudes, with only 15% of pregnant females reproducing again in the following year in Norway versus 93% in The Netherlands. Furthermore, in Norway females grew larger and started reproduction at a higher age. Litter size and juvenile total length and body mass were also significantly higher in Norwegian smooth snakes.

With the incorporation of higher shares of intermittent renewable energies (RES), more flexible resources are required in power systems to keep load balance. Under some extreme circumstances, the flexible demand resources (FDRs) may have the potential to dominate and obtain excess benefits, preventing other FDRs from participating in the electricity markets. Therefore, it is of great significance to identify the key FDR market power locations and implement some corresponding regulations. However, the relevant researches in power systems focused on the supply side, rather than the demand side. In this paper, a novel nodal market power analysis method is proposed to evaluate the potential influence of FDRs on electricity markets. Firstly, a multi-state model is established to present the multiple power system operation states including the random failures of system components. Then, the nodal market power assessment model is established under each specific state and new indices are proposed to evaluate the nodal market power of FDRs quantitatively. Furthermore, the key FDR nodes in demand side with stronger power in capturing excess revenue are identified. The 24-bus IEEE Reliability Test System is modified to demonstrate the feasibility of the proposed method. The numerical results of the proposed method are capable to display the existence of market power in demand side, and provide some valuable guidance for classification and operation of electricity markets., QC 20190325

ObjectiveTo evaluate the risk of Alzheimer's disease-related neuropathology burden at autopsy given older adults' current cognitive state.MethodParticipants included 1,303 individuals who enrolled in the Religious Orders Study (ROS) and 1,789 who enrolled in the Rush Memory and Aging Project (MAP). Cognitive status was evaluated via standardized assessments of global cognition and episodic memory. At the time of analyses, about 50% of participants were deceased with the remaining numbers right censored. Using multi-state Cox proportional hazard models, we compared the cognitive status of all subjects alive at a given age and estimated future risk of dying with different AD-related neuropathologies. Endpoints considered were Braak Stages (0-2, 3-4, 5-6), CERAD (0, 1, 2, 3), and TDP-43 (0, 1, 2, 3) level.ResultsFor all three pathological groupings (Braak, CERAD, TDP-43), we found that a cognitive test score one standard deviation below average put individuals at up to three times the risk for being diagnosed with late stage AD at autopsy according to pathological designations. The effect remained significant after adjusting for sex, APOE-e4 status, smoking status, education level, and vascular health scores.ConclusionApplying multi-state modeling techniques, we were able to identify those at risk of exhibiting specific levels of neuropathology based on current cognitive test performance. This approach presents new and approachable possibilities in clinical settings for diagnosis and treatment development programs.

Objectives: Overdiagnosis is regarded as a harm of screening. We aimed to develop a non-homogeneous multi-state model to consider the age-specific transition rates for estimation of overdiagnosis, to validate the model by a simulation study where the true frequency of overdiagnosis can be calculated, and to compare our estimate with the cumulative incidence method. Methods: We constructed a four-state model to describe the natural history of breast cancer. The latent disease progression and the observed states for each individual were simulated in a trial with biennial screening of women aged 51-69 and a control group of the same size without screening. We performed 100 repetitions of the simulation with one million women to evaluate the performance of estimates. A sensitivity analysis with reduced number of controls was performed to imitate the data from the service screening programme. Results Based on the 100 repetitions, the mean value of the true frequency of overdiagnosis was 12.5% and the average estimates by the cumulative incidence method and the multi-state model were 12.9% (interquartile range: 2.46%) and 13.4% (interquartile range: 2.16%), respectively. The multi-state model had a greater bias of overdiagnosis than the cumulative incidence method, but the variation in the estimates was smaller. When the number of unscreened group was reduced, the variation of multi-state model estimates increased. Conclusions: The multi-state model produces a proper estimate of overdiagnosis and the results are comparable with the cumulative incidence method. The multi-state model can be used in the estimation of overdiagnosis, and might be useful for the ongoing service screening programmes.

Background: Overdiagnosis, defined as the detection of a cancer that would not become clinically apparent in a woman’s lifetime without screening, has become a growing concern. Similar underlying risk of breast cancer in the screened and control groups is a prerequisite for unbiased estimates of overdiagnosis, but a contemporary control group is usually not available in organized screening programs. Methods: We estimated the frequency of overdiagnosis of breast cancer due to screening in women 50–69 years old by using individual screening data from the population-based organized screening program in Stockholm County 1989–2014. A hidden Markov model with four latent states and three observed states was constructed to estimate the natural progression of breast cancer and the test sensitivity. Piecewise transition rates were used to consider the time-varying transition rates. The expected number of detected non-progressive breast cancer cases was calculated. Results: During the study period, 2,333,153 invitations were sent out; on average, the participation rate in the screening program was 72.7% and the average recall rate was 2.48%. In total, 14,648 invasive breast cancer cases were diagnosed; among the 8305 screen-detected cases, the expected number of non-progressive breast cancer cases was 35.9, which is equivalent to 0.43% (95% confidence interval (CI) 0.10%–2.2%) overdiagnosis. The corresponding estimates for the prevalent and subsequent rounds were 15.6 (0.87%, 95% CI 0.20%–4.3%) and 20.3 (0.31%, 95% CI 0.07%–1.6%), respectively. The likelihood ratio test showed that the non-homogeneous model fitted the data better than an age-homogeneous model (P<0.001). Conclusions: Our findings suggest that overdiagnosis in the organized biennial mammographic screening for women 50–69 in Stockholm County is a minor phenomenon. The frequency of overdiagnosis in the prevalent screening round was higher than that in subsequent rounds. The non-homogeneous model performed bett

Background and purpose: To evaluate the quality of life (QoL) of patients with inoperable non-small cell lung cancer treated with conventionally fractionated radiotherapy (CF) vs. continuous hyperfractionated accelerated radiotherapy weekend-less (CHARTWEL). Material and methods: The largest monocentric subgroup of the phase III CHARTWEL trial was analyzed up to three years after randomization. QoL was assessed with the European Organization for Research and Treatment of Cancer QoL Core Questionnaire (QLQ-C30) and lung cancer module (QLQ-LC13) and compared using linear mixed models. QoL interrelations with recurrence, metastasis, and death were explored by multi-state modeling. Results: 160 patients (98%) provided at least one QoL assessment. Average treatment differences of CF vs. CHARTWEL over three years were -5.4 points (95%CI [-13.6,2.8], p = 0.19) in global QoL, 11.9 ([2.8,21.0], p = 0.01) in fatigue, 13.4 ([3.5,23.3], p = 0.009) in pain, 10.5 ([1.3,19.6], p = 0.03) in dyspnea, and 5.2 ([-2.7,13.0], p = 0.19) in dysphagia. At 12 months, the probabilities of being disease-free with good, good or moderate, any global QoL, or alive were 5.1%, 20.3%, 34.2%, 54.4% under CF and 10.4%, 21.0%, 37.5%, 65.3% under CHARTWEL. Conclusions: Over three years, QoL was similar or more favorable under CHARTWEL compared to CF. Modeling QoL together with disease states provided additional insight into treatment comparisons.

Motivated by data from an observational, prospective study of 286 patients with non-mild acute pancreatitis (AP) who entered the Intensive Care Unit (ICU) of the Donostia University Hospital between 2001 and mid-2017, we propose a multi-state modeling approach to describe the evolution of patients with acute pancreatitis and at least one organ failure or local complications. An extension of the illness-death model is used allowing to take into account the disease-related events of interest, that is, entry to ICU, discharge from ICU and death due to acute pancreatitis. To this end, the present study describes the course of AP patients and evaluates the effect of different prognostic factors on the multiple disease-related events.

Utility optimization of households deciding upon consumption and life insurance are studied. In particular, optimality of the life insurance product savings protection is investigated. Distinguishing between private wealth and institutional wealth, partial equilibrium optimal controls are specied on closed form in continuous-time. This is done rstly in the the survival model and then generalized into a multi-state Markov chain. Analytical characterizations of agents who prefer savings protection are found to be unattainable for sophisticated agents with labour income. For various naive agents (and sophisticated agents without labour income), the recurring characteristic for agents demanding savings protection is disclosed to be equal appreciation of consumption regardless of the size of the household. Budget constraints are inferred in a discrete-time Markov model and optimal controls are estimated for two identical agents using numerical dynamic programming. It appears that prohibiting negative institutional wealth has a dampening eect on the demand for life insurance. As a consequence, the marginal utility loss from constraints is so extensive that it overshadows the marginal utility loss of savings protection. However, the utility loss arising from savings protection is still considerable.

Peer Reviewed

Migrationsverket är en myndighet som prövar ansökningar från personer som vill söka skydd, ha medborgarskap, studera eller vill jobba i Sverige. Då det på senare tid varit en stor ökning i dessa ansökningar har tiden för vilket ett beslut tar ökat. Varje typ av ansökning (exempelvis medborgarskap) är en process som består av flera steg. Hur beslutet går igenom dessa steg kallas för flöde. Migrationsverket vill därför öka sin flödeseffektivitet. När beslutet är klart och personen tagit del av det men inte är nöjd kan denne överklaga. Detta är en av de mest komplexa processerna på Migrationsverket. Syftet är analysera hur lång tid denna process tar och vilka steg i processen som påverkar tiden. Ett steg (som senare visar sig ha en stor effekt på tiden) är yttranden. Det är när domstolen begär information om vad personen som överklagar har att säga om varför denne överklagar. För att analysera detta var två metoder relevanta, accelerated failure time (AFT) och \multi-state models (MSM). Den ena kan predicera tid till händelse (AFT) medan den andra kan analysera effekten av tidspåverkan (MSM) i stegen. Yttranden tidigt i processen har stor betydelse för hur snabbt en överklagan får en dom samtidigt som att antal yttranden ökar tiden enormt. Det finns andra faktorer som påverkar tiden men inte i så stor grad som yttranden. Då yttranden tidigt i processen samtidigt som antal yttranden har betydelse kan flödeseffektiviteten ökas med att ta tid på sig att skriva ett informativt yttrande som gör att domstolen inte behöver begära flera yttranden., The Swedish Migration Board is an agency that review applications from individuals who wish to seek shelter, have citizenship, study or want to work in Sweden. In recent time there has been a large increase in applications and the time for which a decision is made has increased. Each type of application (such as citizenship) is a process consisting of several stages. How the decision is going through these steps is called flow. The Swedish Migration Board would therefore like to increase their flow efficiency. When the decision is made and the person has take part of it but is not satisfied, he can appeal. This is one of the most complex processes at the Board. The aim is to analyze how long this process will take and what steps in the process affects the time. One step (which was later found to have a significant effect on time) is opinions. This is when the court requests information on what the person is appealing has to say about why he is appealing. To analyze this, two methods were relevant, accelerated failure time (AFT) and the multi-state models (MSM). One can predict time to event (AFT), the other to analyze the effect of time-manipulation (MSM) in the flow. Opinions early in the process is crucial to how quickly an appeal get judgment while the number of opinions increases the time enormously. There are other factors that affect the time but not so much as opinions. The flow efficiency can be increased by taking time to write an informative opinion which allows the court need not to ask for more opinions.

As one of the most important renewable energy resources, wind power has drawn much attention in recent years. The stochastic characteristics of wind speed lead to generation output uncertainties of wind energy conversion system (WECS) and affect power system reliability, especially at high wind power penetration levels. Therefore, a more comprehensive analysis toward WECS as well as an appropriate reliability assessment model are essential for maintaining the reliable operation of power systems. In this paper, the impact of wind turbine outage probability on system reliability is firstly developed by considering the following factors: running time, operating environment, operating conditions, and wind speed fluctuations. A multi-state model for wind farms is also established. Numerical results illustrate that the proposed model can be well applied to power system reliability assessment as well as solving a series of reliability-centered decision-making problems of power system scheduling and maintenance arrangements.