Censoring Of Data Research Articles

Distal Repair After Total Aortic Arch Replacement With Frozen Elephant Trunk in Patients With Chronic Multilevel Thoracic Aortic Disease

To examine the management of distal aortic disease after total arch replacement with the frozen elephant trunk (TAR+ FET) in patients with chronic thoracic aortic disease. Two centre retrospective study of consecutive patients treated between January 2010 and December 2019. The primary endpoint was 30 day or in hospital death. The secondary endpoint was midterm survival. Data are presented as median (interquartile range [IQR]). The χ2 or Fisher's exact test was used as appropriate. Estimated survival (standard error) was assessed by calculating the Kaplan-Meier product limit estimator with right censoring of survival data. A p value of < .050 was considered statistically significant. STROBE guidelines were followed. A total of 158 patients (72 men; median age 70 years, IQR 64, 75; median distal aortic diameter 58 mm, IQR 46, 68; 127 aneurysmal disease, 31 chronic dissection) underwent TAR+ FET. The peri-operative mortality rate was 10.1% (9/107 elective, 7/51 non-elective). Of 74 (46.8%) patients with a primary distal seal, seven (9.5%) died peri-operatively, the distal seal was maintained during follow up in 51, nine underwent late distal repair (two planned, seven unplanned; one open, eight endovascular; one peri-operative death) with a median interval to unplanned repair of 777 days (IQR 462, 1480), and seven with loss of seal had no intervention. Distal seal failed in 2/28 (7%) patients with a distal seal length > 30 mm and device oversizing > 10%, compared with 12/39 (31%) patients who did not meet these criteria (p= .031). In 84 patients without a primary distal seal, nine (10.7%) died peri-operatively, the distal aorta remained below the size threshold for repair during follow up in 12 patients, 44 had distal repair (median aortic diameter 64 mm, IQR 60, 75; eight open, one hybrid, 35 endovascular repairs; no deaths) at a median of 256 days (IQR 135, 740), and 19 did not have distal repair at the end of the follow up period: six died before planned repair at a median interval of 115 days (IQR 85, 120); eight were considered unfit; one was assessed as fit but declined; and four patients were awaiting assessment. Median follow up was 46 months (IQR 26, 75): no patients were lost to follow up. Estimated ± standard error five year survival was 61.5 ± 4.1%: elective 70.6 ± 4.7%, non-elective 43.2 ± 7.2%. TAR+ FET achieved primary distal seal in 47% of patients, but late failure occurred in 21%. Distal repair was ultimately indicated in 84% of survivors without a primary distal seal and of these 70% underwent repair, almost 10% died before planned repair, and 13% were considered unfit. Earlier distal endovascular repair and better assessment of patient fitness may improve midterm outcomes.

Genetic Susceptibility Modifies Relationships Between Air Pollutants and Stroke Risk: A Large Cohort Study.

The extent to which genetic susceptibility modifies the associations between air pollutants and the risk of incident stroke is still unclear. This study was designed to investigate the separate and joint associations of long-term exposure to air pollutants and genetic susceptibility on stroke risk. The participants of this study were recruited by the UK Biobank between 2006 and 2010. These participants were followed up from the enrollment until the occurrence of stroke events or censoring of data. Hazard ratios (HRs) and 95% CIs for stroke events associated with long-term exposure to air pollutants were estimated by fitting both crude and adjusted Cox proportional hazards models. Additionally, the polygenic risk score was calculated to estimate whether the polygenic risk score modifies the associations between exposure to air pollutants and incident stroke. A total of 502 480 subjects were included in this study. After exclusion, 452 196 participants were taken into the final analysis. During a median follow-up time of 11.7 years, 11 334 stroke events were observed, with a mean age of 61.60 years, and men accounted for 56.2% of the total cases. Long-term exposures to particulate matter with an aerodynamic diameter smaller than 2.5 µm (adjusted HR, 1.70 [95% CI, 1.43-2.03]) or particulate matter with an aerodynamic diameter smaller than 10 µm (adjusted HR, 1.50 [95% CI, 1.36-1.66]), nitrogen dioxide (adjusted HR, 1.10 [95% CI, 1.07-1.12]), and nitrogen oxide (adjusted HR, 1.04 [95% CI, 1.02-1.05]) were pronouncedly associated with increased risk of stroke. Meanwhile, participants with high genetic risk and exposure to high air pollutants had ≈45% (31%, 61%; particulate matter with an aerodynamic diameter smaller than 2.5 µm), 48% (33%, 65%; particulate matter with an aerodynamic diameter smaller than 10 µm), 51% (35%, 69%; nitrogen dioxide), and 39% (25%, 55%; nitrogen oxide) higher risk of stroke compared with those with low genetic risk and exposure to low air pollutants, respectively. Of note, we observed additive and multiplicative interactions between genetic susceptibility and air pollutants on stroke events. Chronic exposure to air pollutants was associated with an increased risk of stroke, especially in populations at high genetic risk.

The variances of non-parametric estimates of the cross-sectional distribution of durations

This paper focuses on the link between non-parametric survival analysis and three distributions. The delta method is applied to derive the variances of the non-parametric estimators of three distributions: the distribution of durations (DD), the cross-sectional distribution of ages (CSA) and the cross-sectional distribution of (completed) durations (CSD). The non-parametric estimator of the the cross-sectional distribution of durations (CSD) has been defined and derived by Dixon (2012) and used in the generalized Taylor price model (GTE) by Dixon and Le Bihan (2012). The Monte Carlo method is applied to evaluate the variances of the estimators of DD and CSD and how their performance varies with sample size and the censoring of data. We apply those estimators to two data sets: the UK CPI micro-price data and waiting-time data from UK hospitals. Both the estimates of the distributions and their variances are calculated. Depending on the empirical results, the estimated variances indicate that the DD and CSD estimators are all significant.

Kaplan-Meyer Survival Curves: Simulation Technique

The right censoring of survival data, being the most conventional method of research, is analyzed. The patient survival is explored in a time span that is shorter in fact than the actual survival time. However, when the actual survival time is unknown, the proxy of the observable survival time will be used for estimating the actual survival time. &#x0D; The algorithm for estimation of survival probabilities is demonstrated by data on 20 patients during six months, with visualizing the technique of simulating Kaplan – Meyer curves by categorical variables (method of treatment and gender) using GraphPad Prism software for statistical data processing. It is argued that Kaplan – Meyer curves could provide an effective tool in simulating the patient survival in case of COVID-19 by various criteria of grouping: gender (male and female); treatment method; associated diseases (diabetes and others); age group; vaccinated or not vaccinated patients etc. &#x0D; The significance of differences between survival curves of patienst in various groups can be found using Log-Rank test, Gehan – Wilcoxon test, Mantel – Cox test and others. The results of tests produced on the basis of data on 42 patients ill with leukemia show significant differences in the survival between two groups of patients. This confirms the assumption that the new method of treatment is more effective than the conventional one. The main deficiency of the nonparametric method of Kaplan – Meyer is that it is incapable to build curves by several categorical variables. The main advantages of Cox regression based on the Cox proportional hazards model are demonstrated.

A Phase II Trial of Sirolimus with Standard Induction Chemotherapy in Patients with De Novo Acute Myeloid Leukemia

Background: The initial treatment of FLT3 wild type acute myeloid leukemia (AML) has not significantly changed since induction therapy with Ara-C and anthracyclines was first developed. Preclinical data suggests constitutive activation of the AKT3/mammalian target of rapamycin (mTOR) pathway may play a role in pathogenesis of this disease in a subset of AML patients. Previous data from our group has shown that the presence of phosphorylated ribosomal S6 (pS6) in AML blasts as detected by flow cytometry may predict response to the combination of sirolimus and induction chemotherapy. Here we report the clinical and pharmacodynamics results of a phase II study of the combination of these drugs.Methods: Subjects had newly diagnosed AML based on WHO criteria. Subjects received oral sirolimus starting on day 1 (12 mg loading dose) then 4 mg daily on days 2-10 with idarubicin 12mg/m2 days 4-7 and cytarabine 100mg/m2 days 4-10. Clinical response was assessed at hematologic recovery or by day 42 using IWG for response. Samples for pharmacodynamic studies of blasts were drawn prior to treatment with sirolimus on day 1, and then again prior to dosing of idarubicin and cytarabine on day 4. Assessment of pS6 on gated blasts, a biomarker of mTORC1 activation, was assessed by flow cytometry using previously described methods.Results: 55 patients enrolled on this trial received sirolimus with an average age of 58 years. Toxicity was similar to published 7+3 data and prolonged aplasia without recovery was not observed. Twenty-nine patients (53%) had high risk cytogenetic or molecular mutations, 15 (27%) had intermediate risk disease, and 11 (20%) patients had lower risk disease. As a whole, 35 patients (64%) entered into CR or CRp, 1 had a PR, 17 (27%) were non-responders. Two (4%) patients (ages 64 and 71) died during induction. Thirteen (45%) patients with high risk disease had a CR or CRp, 14 (48%) were non-responders, and 2 died in induction. Of the 15 subjects with intermediate risk disease, 11 (79%) patients had a CR or CRp, 1 had a partial response, and 3 were non-responders. All low risk patients had a CR, and 9 of 11 are alive and in remission at the time of censoring of data with a median time on study of 674 days. One patient with low risk disease subsequently relapsed, went to stem cell transplant, and is currently in CR. Two patients with low risk disease refused to proceed with standard of care consolidation; of these, 1 is alive and in remission 946 days from study entry, and 1 died in relapse. Fifteen patients (27%) subsequently underwent allogeneic stem cell transplantation in CR, and 8 (53%) are still alive an average of 358 days after transplant.Thirty-seven of 55 patients had adequate samples for pharmacodynamic analysis. Surprisingly, all patients demonstrated activation of ribosomal S6 prior to therapy in contrast to 67% positivity seen in previous studies that focused on patients with relapsed AML. mTORC1 was inhibited greater than 40% in 28/37 (76%) of patients. Inhibition did not correlate with either overall survival or age of patients.Conclusions: Sirolimus and 7+3 is a well-tolerated regimen. mTORC1 appears to be activated in almost all patients at the time of diagnosis of AML as assessed by flow cytometric assessment of ribosomal S6. Inhibition of mTORC1 did not correlate with response, suggesting that AML cells may have redundant signaling pathways that regulate chemosensitivity when mTORC1 is inhibited. Given the large percentage of samples with mTORC1 activation in this upfront AML study, future research into dual mTORC inhibitors may prove valuable. DisclosuresCarroll:Astellas Pharmaceuticals: Research Funding; Incyte Pharmaceuticals: Research Funding. Porcu:Innate Pharma: Research Funding; Kura: Research Funding; Miragen: Research Funding; Kiowa: Research Funding; Galderma: Research Funding; Tetralogic: Research Funding; Cell Medica: Research Funding; Celgene: Research Funding. Perl:Arog Pharmaceuticals: Consultancy; Novartis: Other: Advisory Board; Pfizer: Other: Advisory Board; Seattle Genetics: Other: Advisory board; Asana Biosciences: Other: Scientific advisory board; Astellas: Consultancy; Daiichi Sankyo: Consultancy; Actinium Pharmaceuticals: Other: Scientific Advisory Board.

Monitoring for clean air spaces during wildfire events

Background: Many communities are exposed to extended periods of smoke from wildland fires. A common recommendation to reduce exposure during smoke events is to go indoors. However, building characteristics that translate to clean air spaces during such events are poorly understood. The purpose of this study was to compare outdoor and indoor fine particulate (PM2.5) at commercial buildings during smoke events.Methods: Purple Air Sensors (PA-II-SD) were co-located with regulatory grade reference monitors to generate individual sensor correction equations. Outdoor and indoor sensors were co-located at 18 buildings in Missoula, MT from July to September 2019. Outdoor to indoor daily PM2.5 averages were compared for assessment of building air handling systems. Spikes in indoor PM2.5, identified as any hour where the standard deviation of the 2-minute data was over 10% of the hourly average, were flagged and removed from the comparisons.Results: Individual correction equations and strength of agreement were similar across the set of sensors (slope: 0.36 -0.47 and R2: 0.91 to 0.95). Censoring of data flagged as indoor generated PM2.5 accounted for &lt;5% to 14% of observations across buildings. Fourteen buildings had MERV8 filter systems, two buildings had higher MERV rating filters, and two buildings had no air handling systems. All buildings with air handling systems had reductions in indoor PM2.5 compared to outdoor concentrations but building-to-building observations were highly variable (6% to 44% reduction for MERV8 buildings). Outdoor-to-indoor reductions were similar during smoke impacted periods compared to non-smoke periods.Conclusions: This study demonstrates the utility of low-cost PM2.5 sensors for identifying poor indoor air quality. The variable findings for buildings with similar air handling filter ratings indicate that other building features and operation behaviors are important factors and need to be understood to inform exposure mitigation strategies during smoke events.

Censored quantile regression model with time-varying covariates under length-biased sampling.

Quantile regression is a flexible and effective tool for modeling survival data and its relationship with important covariates, which often vary over time. Informative right censoring of data from the prevalent cohort within the population often results in length-biased observations. We propose an estimating equation-based approach to obtain consistent estimators of the regression coefficients of interest based on length-biased observations with time-dependent covariates. In addition, inspired by Zeng and Lin 2008, we also develop a more numerically stable procedure for variance estimation. Large sample properties including consistency and asymptotic normality of the proposed estimator are established. Numerical studies presented demonstrate convincing performance of the proposed estimator under various settings. The application of the proposed method is demonstrated using the Oscardataset.

A population-based heritability estimate of bipolar disorder – In a Swedish twin sample

Twin- and family studies have shown variations in the heritability estimates of bipolar disorder (BPD). The current study uses an updated statistical methodology for heritability estimation in BPD by taking available time of follow-up into account while controlling for co-variates. We identified monozygotic and dizygotic same and different sex twins with BPD (n = 804) or unaffected from BPD (n = 91,604) from the Swedish Twin Register and the National Patient Register. We applied structural equational modeling with inversed probability weighting to estimate the heritability, taking into account censoring and truncation of data. Sex-limitation models were constructed to analyze qualitative or quantitative sex-differences in BPD. Heritability for BPD was 60.4% (95% Confidence Interval: 50.3–70.5) after age, sex, left-hand truncation and censoring of the data was taken into account. A larger proportion of females were affected from BPD (females 62.2%; males 37.8%, p < 0.001), but no sex-difference in BPD heritability was found, nor any sex-specific genetic effects. We demonstrated a robust 60% heritability for BPD with no evidence of sex-specific genetic effects on disease liability.

The influence of pharmacologically-induced affective states on attention bias in sheep.

When an individual attends to certain types of information more than others, the behavior is termed an attention bias. The occurrence of attention biases in humans and animals can depend on their affective states. Based on evidence from the human literature and prior studies in sheep, we hypothesized that an attention bias test could discriminate between pharmacologically-induced positive and negative affective states in sheep. The test measured allocation of attention between a threat and a positive stimulus using key measures of looking time and vigilance. Eighty 7-year-old Merino ewes were allocated to one of four treatment groups; Anxious (m-chlorophenylpiperazine), Calm (diazepam), Happy (morphine) and Control (saline). Drugs were administered 30 min prior to attention bias testing. The test was conducted in a 4 × 4.2 m arena with high opaque walls. An approximately life-size photograph of a sheep was positioned on one wall of the arena (positive stimulus). A small window with a retractable opaque cover was positioned on the opposite wall, behind which a dog was standing quietly (threat). The dog was visible for 3 s after a single sheep entered the arena, then the window was covered and the dog was removed. Sheep then remained in the arena for 3 min while behaviors were recorded. Key behaviors included time looking toward the dog wall or photo wall, duration of vigilance behavior and latency to become non-vigilant. In contrast with our hypothesis, no significant differences were found between treatment groups for duration of vigilance or looking behaviors, although Anxious sheep tended to be more vigilant than Control animals (P < 0.1) and had a longer latency to become non-vigilant (P < 0.001). A total of 24 of 80 animals were vigilant for the entire test duration. This censoring of data may explain why no differences were detected between groups for vigilance duration. Overall, a lack of difference between groups may suggest the test cannot discriminate positive and negative states in sheep. We suggest that the censoring of vigilance duration data, the use of insufficient drug doses, the potential influence of background noise and the age of the sheep may explain a lack of difference between groups. Due to these potential effects, it remains unclear whether the attention bias test can detect positive states in sheep.

Composite estimator of mean residual lifetime with length-biased and right-censored data

Length-biased data are very common in prevalent cases by cross-sectional sampling when studying time-to-event data, the distribution of length bias data changes from the population distribution and the censoring is informative censoring. But it is difficult to estimate lifetime exactly from such biased observational data with censoring. Bias and the informative censoring in the length-biased data make many classical statistical methods invalid.In this paper, we study a mean residual lifetime model when sample is subject to length-biased and right-censored data. Nonparametric estimation methods are proposed based on estimating equations. To improve the efficiency of the estimator, we take account into the informative censoring and auxiliary information of length-bias data. Although the maximum likelihood estimation is efficient, its computation has so many iterations. Thus we propose a simple way to add the auxiliary information, where the new estimator needs less calculation times. At the same time, asymptotic properties of nonparametric estimators for the mean residual lifetime are established. In simulation studies, it can be seen that our estimator is more effective and simpler.

Assessing the timeliness of vaccine administration in children under five years in India, 2013

Morbidity from vaccine-preventable diseases is high in India, but precise estimates of vaccination timeliness are difficult to compute because many children lack records of vaccination dates. This study assessed vaccination timeliness after accounting for right and left censoring of data. This cross-sectional study used the 2012–2013 District Level Household and Facility Survey in India. The outcome was vaccination timeliness for 9 vaccine doses: 1 dose Bacillus Calmette-Guérin (BCG), 4 doses oral polio vaccine, 3 doses diphtheria-pertussis-tetanus vaccine (DPT), and 1 dose measles-containing vaccine. Age-specific probabilities of vaccination were calculated using Turnbull estimators: children not yet vaccinated were right censored, and children vaccinated but without a recorded date were left censored. Data from 108,783 children under 5 years were available. For children 25–60 months, maternal recall was a more common source of information than a vaccination record with dates. At one month past the recommended vaccination age, estimated coverage ranged from 35% for DPT-3 to 55% for BCG. Accounting for censored data improved vaccination timeliness measures, and demonstrated little increase in vaccination coverage after age one. Efforts to reduce morbidity from vaccine-preventable diseases in India should focus on eliminating missed opportunities for vaccination and instituting special vaccination programs for older children.

A Phase II Trial of Sirolimus with Standard Induction Chemotherapy in Patients with De Novo Acute Myeloid Leukemia

Abstract Background: The initial treatment of FLT3 wild type acute myeloid leukemia (AML) has not significantly changed since induction therapy with Ara-C and anthracyclines was first developed. Preclinical data suggests constitutive activation of the AKT3/mammalian target of rapamycin (mTOR) pathway may play a role in pathogenesis of this disease in a subset of AML patients. Previous data from our group has shown that the presence of phosphorylated ribosomal S6 (pS6) in AML blasts as detected by flow cytometry may predict response to the combination of sirolimus and induction chemotherapy. Here we report the clinical and pharmacodynamics results of a phase II study of the combination of these drugs. Methods: Subjects had newly diagnosed AML based on WHO criteria. Subjects received oral sirolimus starting on day 1 (12 mg loading dose) then 4 mg daily on days 2-10 with idarubicin 12mg/m2 days 4-7 and cytarabine 100mg/m2 days 4-10. Clinical response was assessed at hematologic recovery or by day 42 using IWG for response. Samples for pharmacodynamic studies of blasts were drawn prior to treatment with sirolimus on day 1, and then again prior to dosing of idarubicin and cytarabine on day 4. Assessment of pS6 on gated blasts, a biomarker of mTORC1 activation, was assessed by flow cytometry using previously described methods. Results: 55 patients enrolled on this trial received sirolimus with an average age of 58 years. Toxicity was similar to published 7+3 data and prolonged aplasia without recovery was not observed. Twenty-nine patients (53%) had high risk cytogenetic or molecular mutations, 15 (27%) had intermediate risk disease, and 11 (20%) patients had lower risk disease. As a whole, 35 patients (64%) entered into CR or CRp, 1 had a PR, 17 (27%) were non-responders. Two (4%) patients (ages 64 and 71) died during induction. Thirteen (45%) patients with high risk disease had a CR or CRp, 14 (48%) were non-responders, and 2 died in induction. Of the 15 subjects with intermediate risk disease, 11 (79%) patients had a CR or CRp, 1 had a partial response, and 3 were non-responders. All low risk patients had a CR, and 9 of 11 are alive and in remission at the time of censoring of data with a median time on study of 674 days. One patient with low risk disease subsequently relapsed, went to stem cell transplant, and is currently in CR. Two patients with low risk disease refused to proceed with standard of care consolidation; of these, 1 is alive and in remission 946 days from study entry, and 1 died in relapse. Fifteen patients (27%) subsequently underwent allogeneic stem cell transplantation in CR, and 8 (53%) are still alive an average of 358 days after transplant. Thirty-seven of 55 patients had adequate samples for pharmacodynamic analysis. Surprisingly, all patients demonstrated activation of ribosomal S6 prior to therapy in contrast to 67% positivity seen in previous studies that focused on patients with relapsed AML. mTORC1 was inhibited greater than 40% in 28/37 (76%) of patients. Inhibition did not correlate with either overall survival or age of patients. Conclusions: Sirolimus and 7+3 is a well-tolerated regimen. mTORC1 appears to be activated in almost all patients at the time of diagnosis of AML as assessed by flow cytometric assessment of ribosomal S6. Inhibition of mTORC1 did not correlate with response, suggesting that AML cells may have redundant signaling pathways that regulate chemosensitivity when mTORC1 is inhibited. Given the large percentage of samples with mTORC1 activation in this upfront AML study, future research into dual mTORC inhibitors may prove valuable. Disclosures Carroll: Astellas Pharmaceuticals: Research Funding; Incyte Pharmaceuticals: Research Funding. Porcu: Innate Pharma: Research Funding; Kura: Research Funding; Miragen: Research Funding; Kiowa: Research Funding; Galderma: Research Funding; Tetralogic: Research Funding; Cell Medica: Research Funding; Celgene: Research Funding. Perl: Arog Pharmaceuticals: Consultancy; Novartis: Other: Advisory Board; Pfizer: Other: Advisory Board; Seattle Genetics: Other: Advisory board; Asana Biosciences: Other: Scientific advisory board; Astellas: Consultancy; Daiichi Sankyo: Consultancy; Actinium Pharmaceuticals: Other: Scientific Advisory Board.

Exploring the Effects of Early Censoring and Analysis of Clinical Trial Survival Data on Effectiveness and Cost-effectiveness Estimation through a Case Study in Advanced Breast Cancer.

Interim analyses of clinical trial data are frequently used to provide evidence to obtain marketing authorization for new drugs. However, results from such analyses may not reflect true estimates of relative effectiveness when trial follow-up is complete. Survival results, available at 2 time points from a breast cancer clinical trial, were compared to test the hypothesis that using immature data and a widely used right-censoring rule leads to biased survival estimates. Kaplan-Meier progression-free and overall survival data from 2 published CLEOPATRA trial reports (2012 and 2014) were digitized. Overlaying these results highlighted divergent trends. Parametric functions were fitted to both data sets but did not indicate consistent patterns that could be used as a basis for long-term extrapolation. Heavy censoring of patients in the early data cut coincides with sudden changes in hazard trends and survival patterns, supporting the hypothesis of censoring bias. This challenges the validity of estimates of clinical benefit (progression-free survival and overall survival) based on extrapolation of results from interim analyses of trial data, using a commonly employed censoring rule.

High-sensitivity cardiac troponin: do think twice, it’s not all right

A questionable scientific approach to measuring at low concentrations and inappropriate censoring of results below certain cut-offs have resulted in the dichotomous classification of troponin assays based on their so-called analytical sensitivity. The definition of "high-sensitivity" cardiac troponin is flawed. Evidence suggests that its apparent diagnostic superiority may be explained by the censoring of data. In the evaluation of the detection and quantification capabilities of analytical methods we recommend alignment with International Union of Pure and Applied Chemistry (IUPAC) guidelines, including reporting of all results. This will allow the objective evaluation of the diagnostic performance of troponin assays and will render the current troponin assay classification and nomenclature obsolete.

Anomaly detection and other practical considerations for estimating acoustical metrics from time series data

The scope of environmental acoustic monitoring and rate of data collection are growing rapidly. These increases in the quantity of information have elevated the necessity of detecting anomalous data and the difficulty of doing so. Analysis of contaminated data leads to incorrect results, including biased parameter estimation and flawed model selection. Censoring of data requires strong justification, as the loss of information can lead to these same problems. For acoustics, this issue is compounded by the widespread use of time-average sound levels (Leq), which are especially sensitive to anomalous measurements. This talk will discuss applications of anomaly detection to acoustic pressure time series data for the calculation of long-term metrics. There is no single universally applicable or generic approach to data cleaning and preparation, which often consumes a disproportionate amount of time and effort relative to interpretation of results. Examples drawn from terrestrial and aquatic acoustical monitoring data sets of varying resolution and associated metadata will provide context for issues that compromise data quality, anomaly detection methods, and mitigation efforts.

Introduction of a new continuous time and state space stochastic process in condition prediction

Periodic condition assessments of pavements together with condition predictions are the basis for investment decisions in every pavement management system (PMS). Typical approaches include surveys of distress types every 3–6 years with analysis rating and calculation of condition indices for road safety and/or structural health. Furthermore, advanced PMS prediction models allow a comparison of maintenance alternatives and an optimisation of investment strategies. This paper presents an overview of current survey and rating approaches in Germany, Switzerland and Austria, together with an impact analysis of different methods, utilised deterministic performance functions and condition threshold (trigger) values for all major distress types. The core of this paper is a comparison of common deterministic condition prediction models with discrete stochastic approaches and prediction models based on advanced regression techniques mainly from scientific literature and an innovative stochastic continuous time and continuous state space process (HOFFMANN – Process). All prediction models are applied to real-world data from condition surveys in Austria and the long-term pavement performance Database (USA) at single-section and network level. The paper provides evidence why deterministic prediction approaches are leading to substantial bias in condition distribution and remaining service life as they do not account for the stochastic nature of pavements. Classic Markov-chain approaches do not account for censoring of survey data and neglect changes in transition probabilities with increasing age. Applying common bivariate and multiple regression techniques may also lead to certain bias due to collinearity effects and specification bias. The paper provides mathematical evidence on ways to avoid these shortcomings based on the presented innovative stochastic process leading to a higher reliability in condition assessment, rating and accuracy of condition predictions. The aspects of censoring, distress-specific assignment and optimisation of treatments with this new HOFFMANN-process will be covered in forthcoming papers.

Ustekinumab dosing, persistence, and discontinuation patterns in patients with moderate-to-severe psoriasis

Objectives: Ustekinumab is the most recently approved biologic for the treatment of moderate-to-severe psoriasis. Real-world dosing patterns of ustekinumab are yet to be fully characterized. Methods: A retrospective, observational study was conducted using MarketScan Commercial and Medicare databases. A cohort of psoriasis patients treated with ustekinumab between 25 September 2009 and 31 October 2010 was evaluated. Main outcomes included ustekinumab dosing and treatment patterns. Kaplan–Meier estimates were calculated to adjust for the censoring of data. Subgroup analysis was conducted for biologic-experienced patients and biologic-naïve patients. Results: One thousand ustekinumab patients were included, of whom 60% were biologic-experienced. The average age was 49.0 and 53.9% were male. 63.3% of patients initiated ustekinumab with a 45 mg dose and 34.5% initiated with a 90 mg dose. Mean (median) days from initial dose to second dose was 31.1 (28.0). During maintenance therapy, dose intervals spanned from 80.6 to 81.2 (84.0) days. About 81.4% of patients were persistent during the variable-length follow-up period. Conclusions: The majority of patients received the 45 mg ustekinumab dose. The mean dosing intervals were consistent with the US prescribing guidelines. Biologic-naïve and biologic-experienced patients had similar dosing patterns. Ustekinumab treatment achieved a persistency rate as high as 81.4% over an average of 186.5 (SD 114.2) days of follow-up.

Improving estimates of critical lower percentiles by induced censoring

In this article, we present an approach based on induced censoring for improving the estimation of critical lower percentiles. We validate this technique via simulation results and practical industrial insights. Data from product components that have at least two aging periods (e.g., bathtub failure rate) is investigated. When such data are improperly fit by certain reliability distributions, estimates of lower percentiles are impacted by longer-lasting failures, resulting in larger root mean square errors (RMSE) and bias. In lieu of utilizing a more complex bathtub model, we propose induced right censoring of data at various points to substantially reduce RMSE and bias of lower percentile estimates. A technique for finding optimal or near optimal censoring points is discussed and two real world examples illustrate how this works in practice.

O20.6 Sexual Partnership Patterns Among Young People in Rural Tanzania

BackgroundSexual partnership patterns influence risk of STI transmission. We describe the pattern of partnerships reported by youth in a survey in rural Tanzania and calculate the UNAIDS recommended measure of...

Generalized Redistribute-to-the-Right Algorithm: Application to the Analysis of Censored Cost Data

Medical cost estimation is a challenging task when censoring of data is present. Although researchers have proposed methods for estimating mean costs, these are often derived from theory and are not always easy to understand. We provide an alternative method, based on a replace-from-the-right algorithm, for estimating mean costs more efficiently. We show that our estimator is equivalent to an existing one that is based on the inverse probability weighting principle and semiparametric efficiency theory. We also propose an alternative method for estimating the survival function of costs, based on the redistribute-to-the-right algorithm, that was originally used for explaining the Kaplan-Meier estimator. We show that this second proposed estimator is equivalent to a simple weighted survival estimator of costs. Finally, we develop a more efficient survival estimator of costs, using the same redistribute-to-the-right principle. This estimator is naturally monotone, more efficient than some existing survival estimators, and has a quite small bias in many realistic settings. We conduct numerical studies to examine the finite sample property of the survival estimators for costs, and show that our new estimator has small mean squared errors when the sample size is not too large. We apply both existing and new estimators to a data example from a randomized cardiovascular clinical trial.