Much ado about nothing: A comparison of missing data methods and software to fit incomplete data regression models

Adjusted Indirect Treatment Comparison of Progression-Free Survival with D-Rd and VRd Based on MAIA and SWOG S0777 Individual Patient-Level Data

INTRODUCTION

Daratumumab plus lenalidomide and dexamethasone (D-Rd) and bortezomib plus lenalidomide and dexamethasone (VRd) are commonly used treatment combinations for transplant-ineligible (TIE) patients with newly diagnosed multiple myeloma (NDMM). D-Rd and VRd demonstrated superior efficacy relative to lenalidomide and dexamethasone (Rd) in the MAIA and SWOG S0777 trials, respectively, but have not been compared directly in a head-to-head trial. Naïve comparisons of efficacy across the two trials may be biased because MAIA enrolled only TIE patients (median age 73 years), whereas SWOG S0777 enrolled both TIE patients and transplant-eligible patients who chose to defer/refuse frontline stem cell transplantation (median age 63 years). The present study compared progression-free survival (PFS) in TIE patients with NDMM treated with D-Rd versus VRd based on an adjusted indirect treatment comparison (ITC) that leveraged individual patient-level data from MAIA and SWOG S0777.

METHODS

Harmonized inclusion/exclusion criteria (including age ≥ 65 years as a proxy for transplant ineligibility) and propensity-score weighting were used to balance the trial populations on measured baseline characteristics. After differences in trial populations were adjusted for, an anchored ITC was performed wherein within-trial PFS hazard ratios (HRs) for D-Rd versus Rd and VRd versus Rd were estimated and used to make indirect inference about PFS for D-Rd versus VRd.

RESULTS

PFS HRs were 0.52 (95% confidence interval [CI] 0.41-0.67) for D-Rd versus Rd based on MAIA data, 0.88 (95% CI 0.63-1.23) for VRd versus Rd based on SWOG S0777 data, and 0.59 (95% CI 0.39-0.90) for the Rd-anchored ITC of D-Rd versus VRd. Sensitivity and subgroup analyses produced results consistent with the primary results.

CONCLUSION

This anchored ITC demonstrated a greater PFS benefit for D-Rd versus VRd in TIE patients with NDMM. In the absence of head-to-head trials comparing D-Rd and VRd, the present trial may help inform treatment selection in this patient population.

[Posttraumatic stress disorder in children and adolescents: results of a cross-sectional study on the effects of the newly formulated PTSD and CPTSD diagnoses in the ICD-11]

BACKGROUND

ICD-11 presents narrowed criteria for posttraumatic stress disorder (PTSD) and introduces complex PTSD (CPTSD) with additional difficulties in self-organization (DSO). These changes can have significant effects on the frequency of the diagnosis. The aim of this study was to investigate which ICD-11 symptom clusters cause children and adolescents to miss the diagnosis and whether caregivers are more likely to attribute changes in DSO to developmental level or to the traumatic event, and how these attributions are in turn related to symptom severity.

METHODS

N = 88 German-speaking children and adolescents (age: 7-17 years) after traumatic events and N = 79 caregivers participated between September 2019 and November 2020 in a survey on PTSD symptom severity (CATS-2) and attribution of DSO symptoms (caregiver questionnaire).

RESULTS

The ICD-11 criteria (CATS‑2 and a developmentally adapted version) showed lower frequency rates for PTSD as compared to DSM‑5 and ICD-10. The ICD-11 clusters re-experiencing and hyperarousal were met the least often. Changes in DSO symptoms were predominantly rated as event-related. This attribution was associated with higher PTSD and DSO symptom severity in caregiver reports. The age-related attribution was associated with higher DSO-symptom severity, but not PTSD symptom severity in caregiver reports.

DISCUSSION

In the context of the diagnostic process and the revision of diagnostic instruments for ICD-11 (C)PTSD, development-specific symptoms should be taken into account. The trauma-related differentiation of DSO symptom changes as compared to development-related fluctuations is challenging and therefore requires several sources of information.

Construction of relatedness matrices in autopolyploid populations using low-depth high-throughput sequencing data

KEY MESSAGE

An improved estimator of genomic relatedness using low-depth high-throughput sequencing data for autopolyploids is developed. Its outputs strongly correlate with SNP array-based estimates and are available in the package GUSrelate. High-throughput sequencing (HTS) methods have reduced sequencing costs and resources compared to array-based tools, facilitating the investigation of many non-model polyploid species. One important quantity that can be computed from HTS data is the genetic relatedness between all individuals in a population. However, HTS data are often messy, with multiple sources of errors (i.e. sequencing errors or missing parental alleles) which, if not accounted for, can lead to bias in genomic relatedness estimates. We derive a new estimator for constructing a genomic relationship matrix (GRM) from HTS data for autopolyploid species that accounts for errors associated with low sequencing depths, implemented in the R package GUSrelate. Simulations revealed that GUSrelate performed similarly to existing GRM methods at high depth but reduced bias in self-relatedness estimates when the sequencing depth was low. Using a panel consisting of 351 tetraploid potato genotypes, we found that GUSrelate produced GRMs from genotyping-by-sequencing (GBS) data that were highly correlated with a GRM computed from SNP array data, and less biased than existing methods when benchmarking against the array-based GRM estimates. GUSrelate provides researchers with a tool to reliably construct GRMs from low-depth HTS data.

Exercise prescription variables predict reductions in pain intensity in adults with chronic low back pain: secondary analysis of a randomised controlled trial

Objectives

The relationship between exercise training variables and clinical outcomes in low back pain (LBP) is unclear. The current study aimed to explore the relationship between exercise training parameters and pain intensity in individuals with chronic LBP.

Methods

This study is a secondary analysis of a previously reported randomised controlled trial comparing the effects of general strength and conditioning to motor control exercises and manual therapy. This secondary analysis includes adults with chronic LBP (n=20) randomised to the general strength and conditioning programme only. Primary outcomes of this analysis were exercise training parameters (time under tension, rating of perceived exertion (RPE), session duration, session-RPE and training frequency) and pain intensity (0-100 mm visual analogue acale) measured every 2 weeks from baseline to 6 months follow-up. Linear mixed models with random effects (participants) and allowance for heterogeneity of variance (study date) were used to determine the association between pain intensity and training parameters over time.

Results

Mean (95% CI) pain intensity decreased over time from baseline to 6 months follow-up by 10.7 (2.8 to 18.7) points (p=0.008). Over the 6-month intervention, lower pain intensity was associated with higher RPE (β (95% CI) -27.168 (-44.265 to -10.071), p=0.002), greater time under tension (-0.029 (-0.056 to -0.001), p=0.040) and shorter session duration (1.938 (0.011 to 3.865), p=0.049).

Conclusion

During 6 months of general strength and conditioning, lower pain intensity was associated with higher participant-reported training intensity, greater volume and shorter session duration. To ensure positive outcomes to exercise training, these variables should be monitored on a short-term basis.

Trial registration number

ACTRN12615001270505.

Improving Regression Analysis with Imputation in a Longitudinal Study of Alzheimer's Disease

Background

Missing data is prevalent in the Alzheimer's Disease Neuroimaging Initiative (ADNI). It is common to deal with missingness by removing subjects with missing entries prior to statistical analysis; however, this can lead to significant efficiency loss and sometimes bias. It has yet to be demonstrated that the imputation approach to handling this issue can be valuable in some longitudinal regression settings.

Objective

The purpose of this study is to demonstrate the importance of imputation and how imputation is correctly done in ADNI by analyzing longitudinal Alzheimer's Disease Assessment Scale -Cognitive Subscale 13 (ADAS-Cog 13) scores and their association with baseline patient characteristics.

Methods

We studied 1,063 subjects in ADNI with mild cognitive impairment. Longitudinal ADAS-Cog 13 scores were modeled with a linear mixed-effects model with baseline clinical and demographic characteristics as predictors. The model estimates obtained without imputation were compared with those obtained after imputation with Multiple Imputation by Chained Equations (MICE). We justify application of MICE by investigating the missing data mechanism and model assumptions. We also assess robustness of the results to the choice of imputation method.

Results

The fixed-effects estimates of the linear mixed-effects model after imputation with MICE yield valid, tighter confidence intervals, thus improving the efficiency of the analysis when compared to the analysis done without imputation.

Conclusions

Our study demonstrates the importance of accounting for missing data in ADNI. When deciding to perform imputation, care should be taken in choosing the approach, as an invalid one can compromise the statistical analyses.

Appropriately estimating the standardized average treatment effect with missing data: A simulation and primer

Reporting standardized effects in randomized treatment studies aids interpretation and facilitates future meta-analyses and policy considerations. However, when outcome data are missing, achieving an unbiased, accurate estimate of the standardized average treatment effect, sATE, can pose challenges even for those with general knowledge of missing data handling, given that the sATE is a ratio of a mean difference to a (within-group) standard deviation. Under both homogeneity and heterogeneity of variance, a Monte Carlo simulation study was conducted to compare missing data handling strategies in terms of bias and accuracy in the sATE, under specific missingness patterns plausible for randomized pretest posttest studies. Within two broad missing data handling approaches, maximum likelihood and multiple imputation, modeling choices were thoroughly investigated including the analysis model, variance estimator, imputation algorithm, and method of pooling results across imputed datasets. Results demonstrated that although the sATE can be estimated with little bias using either maximum likelihood or multiple imputation, particular attention should be paid to the model and variance estimator, especially at smaller sample sizes (i.e., N = 50). Differences in accuracy were driven by differences in bias. To improve estimation of the sATE in practice, recommendations and a software demonstration are provided. Moreover, a pedagogical explanation of the causes of bias, described separately for the numerator and denominator of the sATE is provided, demonstrating visually how and why bias occurs with certain methods.

Assessment of interleukin 1 receptor antagonist (IL-1RA) levels in children with and without community acquired pneumonia: a hospital based case-control study

The primary objective was to compare serum interleukin-1 receptor antagonist (IL-1RA) levels in cases of community acquired pneumonia (CAP) and healthy age-gender-matched controls. The secondary objective was to compare serum IL-1RA levels in cases which were positive or negative for Streptococcus pneumoniae in the blood by real-time-polymerase chain reaction (RT-PCR). Hospitalized children with World Health Organization defined CAP, aged 2-59 months, were included as cases. Healthy controls were recruited from the immunization clinic of the hospital. Enzyme-linked immunosorbent assay (ELISA) test was used to detect serum IL-1RA levels. Identification of S.pneumoniae in blood was done by RT-PCR. From October 2019 to October 2021, 330 cases (123, 37.27% female) and 330 controls (151, 45.75% females) were recruited. Mean serum IL-1RA levels (ng/ml) were 1.36 ± 0.95 in cases and 0.25 ± 0.25 in controls (p < 0.001). Within cases, serum IL-1RA levels were significantly higher in those whose RT-PCR was positive for S.pneumoniae. Thus serum IL-1RA levels may be evaluated as a surrogate marker of S.pneumoniae in future studies.

Predicting risk of suicidal behaviour after initiation of selective serotonin reuptake inhibitors in children, adolescents and young adults: protocol for development and validation of clinical prediction models

INTRODUCTION

There is concern regarding suicidal behaviour risk during selective serotonin reuptake inhibitor (SSRI) treatment among the young. A clinically useful model for predicting suicidal behaviour risk should have high predictive performance in terms of discrimination and calibration; transparency and ease of implementation are desirable.

METHODS AND ANALYSIS

Using Swedish national registers, we will identify individuals initiating an SSRI aged 8-24 years 2007-2020. We will develop: (A) a model based on a broad set of predictors, and (B) a model based on a restricted set of predictors. For the broad predictor model, we will consider an ensemble of four base models: XGBoost (XG), neural net (NN), elastic net logistic regression (EN) and support vector machine (SVM). The predictors with the greatest contribution to predictive performance in the base models will be determined. For the restricted predictor model, clinical input will be used to select predictors based on the top predictors in the broad model, and inputted in each of the XG, NN, EN and SVM models. If any show superiority in predictive performance as defined by the area under the receiver-operator curve, this model will be selected as the final model; otherwise, the EN model will be selected. The training and testing samples will consist of data from 2007 to 2017 and from 2018 to 2020, respectively. We will additionally assess the final model performance in individuals receiving a depression diagnosis within 90 days before SSRI initiation.The aims are to (A) develop a model predicting suicidal behaviour risk after SSRI initiation among children and youths, using machine learning methods, and (B) develop a model with a restricted set of predictors, favouring transparency and scalability.

ETHICS AND DISSEMINATION

The research is approved by the Swedish Ethical Review Authority (2020-06540). We will disseminate findings by publishing in peer-reviewed open-access journals, and presenting at international conferences.

A multiple imputation-based sensitivity analysis approach for regression analysis with a missing not at random covariate

Missing covariate problems are common in biomedical and electrical medical record data studies while evaluating the relationship between a biomarker and certain clinical outcome, when biomarker data are not collected for all subjects. However, missingness mechanism is unverifiable based on observed data. If there is a suspicion of missing not at random (MNAR), researchers often perform sensitivity analysis to evaluate the impact of various missingness mechanisms. Under the selection modeling framework, we propose a sensitivity analysis approach with a standardized sensitivity parameter using a nonparametric multiple imputation strategy. The proposed approach requires fitting two working models to derive two predictive scores: one for predicting missing covariate values and the other for predicting missingness probabilities. For each missing covariate observation, the two predictive scores along with the pre-specified sensitivity parameter are used to define an imputing set. The proposed approach is expected to be robust against mis-specifications of the selection model and the sensitivity parameter since the selection model and the sensitivity parameter are not directly used to impute missing covariate values. A simulation study is conducted to study the performance of the proposed approach when MNAR is induced by Heckman's selection model. Simulation results show the proposed approach can produce plausible regression coefficient estimates. The proposed sensitivity analysis approach is also applied to evaluate the impact of MNAR on the relationship between post-operative outcomes and incomplete pre-operative Hemoglobin A1c level for patients who underwent carotid intervetion for advanced atherosclerotic disease.

Does the Presence of Missing Data Affect the Performance of the SORG Machine-learning Algorithm for Patients With Spinal Metastasis? Development of an Internet Application Algorithm

BACKGROUND

The Skeletal Oncology Research Group machine-learning algorithm (SORG-MLA) was developed to predict the survival of patients with spinal metastasis. The algorithm was successfully tested in five international institutions using 1101 patients from different continents. The incorporation of 18 prognostic factors strengthens its predictive ability but limits its clinical utility because some prognostic factors might not be clinically available when a clinician wishes to make a prediction.

QUESTIONS/PURPOSES

We performed this study to (1) evaluate the SORG-MLA's performance with data and (2) develop an internet-based application to impute the missing data.

METHODS

A total of 2768 patients were included in this study. The data of 617 patients who were treated surgically were intentionally erased, and the data of the other 2151 patients who were treated with radiotherapy and medical treatment were used to impute the artificially missing data. Compared with those who were treated nonsurgically, patients undergoing surgery were younger (median 59 years [IQR 51 to 67 years] versus median 62 years [IQR 53 to 71 years]) and had a higher proportion of patients with at least three spinal metastatic levels (77% [474 of 617] versus 72% [1547 of 2151]), more neurologic deficit (normal American Spinal Injury Association [E] 68% [301 of 443] versus 79% [1227 of 1561]), higher BMI (23 kg/m2 [IQR 20 to 25 kg/m2] versus 22 kg/m2 [IQR 20 to 25 kg/m2]), higher platelet count (240 × 103/µL [IQR 173 to 327 × 103/µL] versus 227 × 103/µL [IQR 165 to 302 × 103/µL], higher lymphocyte count (15 × 103/µL [IQR 9 to 21× 103/µL] versus 14 × 103/µL [IQR 8 to 21 × 103/µL]), lower serum creatinine level (0.7 mg/dL [IQR 0.6 to 0.9 mg/dL] versus 0.8 mg/dL [IQR 0.6 to 1.0 mg/dL]), less previous systemic therapy (19% [115 of 617] versus 24% [526 of 2151]), fewer Charlson comorbidities other than cancer (28% [170 of 617] versus 36% [770 of 2151]), and longer median survival. The two patient groups did not differ in other regards. These findings aligned with our institutional philosophy of selecting patients for surgical intervention based on their level of favorable prognostic factors such as BMI or lymphocyte counts and lower levels of unfavorable prognostic factors such as white blood cell counts or serum creatinine level, as well as the degree of spinal instability and severity of neurologic deficits. This approach aims to identify patients with better survival outcomes and prioritize their surgical intervention accordingly. Seven factors (serum albumin and alkaline phosphatase levels, international normalized ratio, lymphocyte and neutrophil counts, and the presence of visceral or brain metastases) were considered possible missing items based on five previous validation studies and clinical experience. Artificially missing data were imputed using the missForest imputation technique, which was previously applied and successfully tested to fit the SORG-MLA in validation studies. Discrimination, calibration, overall performance, and decision curve analysis were applied to evaluate the SORG-MLA's performance. The discrimination ability was measured with an area under the receiver operating characteristic curve. It ranges from 0.5 to 1.0, with 0.5 indicating the worst discrimination and 1.0 indicating perfect discrimination. An area under the curve of 0.7 is considered clinically acceptable discrimination. Calibration refers to the agreement between the predicted outcomes and actual outcomes. An ideal calibration model will yield predicted survival rates that are congruent with the observed survival rates. The Brier score measures the squared difference between the actual outcome and predicted probability, which captures calibration and discrimination ability simultaneously. A Brier score of 0 indicates perfect prediction, whereas a Brier score of 1 indicates the poorest prediction. A decision curve analysis was performed for the 6-week, 90-day, and 1-year prediction models to evaluate their net benefit across different threshold probabilities. Using the results from our analysis, we developed an internet-based application that facilitates real-time data imputation for clinical decision-making at the point of care. This tool allows healthcare professionals to efficiently and effectively address missing data, ensuring that patient care remains optimal at all times.

RESULTS

Generally, the SORG-MLA demonstrated good discriminatory ability, with areas under the curve greater than 0.7 in most cases, and good overall performance, with up to 25% improvement in Brier scores in the presence of one to three missing items. The only exceptions were albumin level and lymphocyte count, because the SORG-MLA's performance was reduced when these two items were missing, indicating that the SORG-MLA might be unreliable without these values. The model tended to underestimate the patient survival rate. As the number of missing items increased, the model's discriminatory ability was progressively impaired, and a marked underestimation of patient survival rates was observed. Specifically, when three items were missing, the number of actual survivors was up to 1.3 times greater than the number of expected survivors, while only 10% discrepancy was observed when only one item was missing. When either two or three items were omitted, the decision curves exhibited substantial overlap, indicating a lack of consistent disparities in performance. This finding suggests that the SORG-MLA consistently generates accurate predictions, regardless of the two or three items that are omitted. We developed an internet application (https://sorg-spine-mets-missing-data-imputation.azurewebsites.net/) that allows the use of SORG-MLA with up to three missing items.

CONCLUSION

The SORG-MLA generally performed well in the presence of one to three missing items, except for serum albumin level and lymphocyte count (which are essential for adequate predictions, even using our modified version of the SORG-MLA). We recommend that future studies should develop prediction models that allow for their use when there are missing data, or provide a means to impute those missing data, because some data are not available at the time a clinical decision must be made.

CLINICAL RELEVANCE

The results suggested the algorithm could be helpful when a radiologic evaluation owing to a lengthy waiting period cannot be performed in time, especially in situations when an early operation could be beneficial. It could help orthopaedic surgeons to decide whether to intervene palliatively or extensively, even when the surgical indication is clear.

How to apply variable selection machine learning algorithms with multiply imputed data: A missing discussion

Psychological researchers often use standard linear regression to identify relevant predictors of an outcome of interest, but challenges emerge with incomplete data and growing numbers of candidate predictors. Regularization methods like the LASSO can reduce the risk of overfitting, increase model interpretability, and improve prediction in future samples; however, handling missing data when using regularization-based variable selection methods is complicated. Using listwise deletion or an ad hoc imputation strategy to deal with missing data when using regularization methods can lead to loss of precision, substantial bias, and a reduction in predictive ability. In this tutorial, we describe three approaches for fitting a LASSO when using multiple imputation to handle missing data and illustrate how to implement these approaches in practice with an applied example. We discuss implications of each approach and describe additional research that would help solidify recommendations for best practices. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Downer cows: a reanalysis of an old data set

AIMS

To compare the performance of two predictive models for the survival of downer cows.

METHODS

The first model had been developed in 1987 using a dataset containing missing values, while the second, new model was developed on the same dataset but using modern data imputation and analytical methods. Missing data were imputed using multiple imputation by chained equations and a logistic regression model fitted to the imputed data, with survival or not as the outcome variable. The predictive ability of the model built on the imputed data was contrasted with the original prognostic model by testing them both on a second smaller but complete data set, collected contemporaneously with the development of the original model but from a different region of New Zealand. Sensitivity, specificity, accuracy, and cut point for the two models were calculated.

RESULTS

The original 1987 model had a slightly higher accuracy than that of the new one with a sensitivity of 0.85 (95% CI = 0.72-0.94) and a specificity of 0.82 (95% CI = 0.7-0.91), using a cut point for the probability of survival = 0.313.

CONCLUSIONS

The original prognostic formula published by Clark et al. in 1987 performed as well as a modern model built on an imputed data set.

CLINICAL RELEVANCE

The use of a prognostic test based on the Clark model should remain an important part of the clinical examination of downer cows by New Zealand veterinarians.Abbreviations: AUC: Area under the curve; AST: Aspartate transaminase activity; CK: Creatine phosphokinase activity; GAM: Generalised additive model; NSAID: Non-steroidal-anti-inflammatory drugs; PCV: Packed cell volume.

Mediation analysis of multiple mediators with incomplete omics data

There is an increasing interest in using multiple types of omics features (e.g., DNA sequences, RNA expressions, methylation, protein expressions, and metabolic profiles) to study how the relationships between phenotypes and genotypes may be mediated by other omics markers. Genotypes and phenotypes are typically available for all subjects in genetic studies, but typically, some omics data will be missing for some subjects, due to limitations such as cost and sample quality. In this article, we propose a powerful approach for mediation analysis that accommodates missing data among multiple mediators and allows for various interaction effects. We formulate the relationships among genetic variants, other omics measurements, and phenotypes through linear regression models. We derive the joint likelihood for models with two mediators, accounting for arbitrary patterns of missing values. Utilizing computationally efficient and stable algorithms, we conduct maximum likelihood estimation. Our methods produce unbiased and statistically efficient estimators. We demonstrate the usefulness of our methods through simulation studies and an application to the Metabolic Syndrome in Men study.

Contemporary screen time modalities and disruptive behavior disorders in children: a prospective cohort study

BACKGROUND

Cross-sectional studies have demonstrated associations between screen time and disruptive behavior disorders (conduct disorder and oppositional defiant disorder); however, prospective associations remain unknown. This study's objective was to determine the prospective associations of contemporary screen time modalities with conduct and oppositional defiant disorder in a national cohort of 9-11-year-old children.

METHODS

We analyzed data from the Adolescent Brain Cognitive Development (ABCD) Study (N = 11,875). Modified Poisson regression analyses were conducted to estimate the associations between baseline child-reported screen time (total and by modality) and parent-reported conduct or oppositional defiant disorder based on the Kiddie Schedule for Affective Disorders and Schizophrenia (KSADS-5) at 1-year follow-up, adjusting for potential confounders.

RESULTS

Participants reported an average of 4 hr of total screen time per day at baseline. Each hour of total screen time per day was prospectively associated with a 7% higher prevalence of conduct disorder (95% CI 1.03-1.11) and a 5% higher prevalence of oppositional defiant disorder (95% CI 1.03-1.08) at 1-year follow-up. Each hour of social media per day was associated with a 62% higher prevalence of conduct disorder (95% CI 1.39-1.87). Each hour of video chat (prevalence ratio [PR] 1.21, 95% CI 1.06-1.37), texting (PR 1.19, 95% CI 1.07-1.33), television/movies (PR 1.17, 95% CI 1.10-1.25), and video games (PR 1.14, 95% CI 1.07-1.21) per day was associated with a higher prevalence of the oppositional defiant disorder. When examining thresholds, exposure to >4 hr of total screen time per day was associated with a higher prevalence of conduct disorder (69%) and oppositional defiant disorder (46%).

CONCLUSIONS

Higher screen time was prospectively associated with a higher prevalence of new-onset disruptive behavior disorders. The strongest association was between social media and conduct disorder, indicating that future research and interventions may focus on social media platforms to prevent conduct disorder.

Effectiveness of 13-valent pneumococcal conjugate vaccine on radiological primary end-point pneumonia among cases of severe community acquired pneumonia in children: A prospective multi-site hospital-based test-negative study in Northern India

INTRODUCTION

Community acquired pneumonia (CAP) is a leading cause of under-five mortality in India and Streptococcus pneumoniae is the main bacterial pathogen for it. Pneumococcal Conjugate Vaccine 13 (PCV13) has been introduced in a phased manner, in the national immunization program of India since 2017/2018. The primary objective of this study was to evaluate the effectiveness of PCV13 on chest radiograph (CXR)-confirmed pneumonia, in children hospitalized with WHO-defined severe CAP.

METHODS

This prospective, multi-site test-negative study was conducted in a hospital-network situated in three districts of Northern India where PCV13 had been introduced. Children aged 2-23 months, hospitalized with severe CAP and with interpretable CXR were included after parental consent. Clinical data was extracted from hospital records. CXRs were interpreted by a panel of three independent blinded trained radiologists. Exposure to PCV13 was defined as ≥2 doses of PCV13 in children aged ≤ 12 months and ≥ 1 dose(s) in children > 12 months of age. Our outcome measures were CXR finding of primary endpoint pneumonia with or without other infiltrates (PEP±OI); vaccine effectiveness (VE) and hospital mortality.

RESULTS

From 1st June 2017-30th April 2021, among 2711 children included, 678 (25.0%) were exposed to PCV1. CXR positive for PEP±OI on CXR was found in 579 (21.4%), of which 103 (17.8%) were exposed to PCV. Adjusted odds ratio (AOR) for PEP±OI among the exposed group was 0.69 (95% CI, 0.54-0.89, p = 0.004). Adjusted VE was 31.0% (95% CI: 11.0-44.0) for PEP±OI. AOR for hospital mortality with PEP±OI was 2.65 (95% CI: 1.27-5.53, p = 0.01).

CONCLUSION

In severe CAP, children exposed to PCV13 had significantly reduced odds of having PEP±OI. Since PEP±OI had increased odds of hospital mortality due to CAP, countrywide coverage with PCV13 is an essential priority.

The Collective Influence of Social Determinants of Health on Individuals Who Underwent Lumbar Spine Revision Surgeries: A Retrospective Cohort Study

OBJECTIVE

To analyze the collective effect of social determinants of health (SDH) on lumbar spine revision surgery outcomes using a retrospective cohort study design.

METHODS

Data from the Quality Outcomes Database were used, including 7889 adults who received lumbar spine revision surgery and completed 3 and 12 months' follow-up. The SDH of interest included race/ethnicity, educational attainment, employment status, insurance payer, and sex. A stepwise regression model using each number of SDH conditions present (0 of 5, 1 of 5, 2 of 5, ≥3 of 5) was used to assess the collective influence of SDH. The odds of demonstrating a minimum clinically important difference was evaluated in back and leg, disability, quality of life, and patient satisfaction at 3-months and 12-months follow-up.

RESULTS

An additive effect for SDH was found across all outcome variables at 3 and 12 months. Individuals with ≥3 SDH were at the lowest odds of meeting the minimum clinically important difference of each outcome. At 12 months, individuals with ≥3 SDH had a 67%, 65%, 71%, 65%, and 46% decrease in the odds of a clinically meaningful outcome in back and leg pain, disability, quality of life, and patient satisfaction.

CONCLUSIONS

Health care teams should evaluate SDH in individuals who may be considered for lumbar spine revision surgery. Viewing social factors in aggregate may be useful as a screening tool for lumbar spine revision surgeries to identify at risk patients who may require pre-emptive care strategies and postoperative resources to mitigate these risks.

Assessing Alternative Imputation Strategies for Infrequently Missing Items on Multi-item Scales

Health-science researchers often measure psychological constructs using multi-item scales and encounter missing items on some participants. Multiple imputation (MI) has emerged as an alternative to ad-hoc methods (e.g., mean substitution) for handling incomplete data on multi-item scales, appealingly reflecting available information while accounting for uncertainty due to missing values in a unified inferential framework. However, MI can be implemented in a variety of ways. When the number of variables to impute gets large, some strategies yield unstable estimates of quantities of interest while others are not technically feasible to implement. These considerations raise pragmatic questions about the extent to which ad-hoc procedures would yield statistical properties that are competitive with theoretically motivated methods. Drawing on an HIV study where depression and anxiety symptoms are measured with multi-item scales, this empirical investigation contrasts ad-hoc methods for handling missing items with various MI implementations that differ as to whether imputation is at the item-level or scale-level and how auxiliary variables are incorporated. While the findings are consistent with previous reports favoring item-level imputation when feasible to implement, we found only subtle differences in statistical properties across procedures, suggesting that weaknesses of ad-hoc procedures may be muted when missing data percentages are modest.

Cervical spine and muscle adaptation after spaceflight and relationship to herniation risk: protocol from 'Cervical in Space' trial

Background

Astronauts have a higher risk of cervical intervertebral disc herniation. Several mechanisms have been attributed as causative factors for this increased risk. However, most of the previous studies have examined potential causal factors for lumbar intervertebral disc herniation only. Hence, we aim to conduct a study to identify the various changes in the cervical spine that lead to an increased risk of cervical disc herniation after spaceflight.

Methods

A cohort study with astronauts will be conducted. The data collection will involve four main components: a) Magnetic resonance imaging (MRI); b) cervical 3D kinematics; c) an Integrated Protocol consisting of maximal and submaximal voluntary contractions of the neck muscles, endurance testing of the neck muscles, neck muscle fatigue testing and questionnaires; and d) dual energy X-ray absorptiometry (DXA) examination. Measurements will be conducted at several time points before and after astronauts visit the International Space Station. The main outcomes of interest are adaptations in the cervical discs, muscles and bones.

Discussion

Astronauts are at higher risk of cervical disc herniation, but contributing factors remain unclear. The results of this study will inform future preventive measures for astronauts and will also contribute to the understanding of intervertebral disc herniation risk in the cervical spine for people on Earth. In addition, we anticipate deeper insight into the aetiology of neck pain with this research project.

Trial registration

German Clinical Trials Register, DRKS00026777. Registered on 08 October 2021.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-022-05684-0.

A new approach to estimation of the proportional hazards model based on interval-censored data with missing covariates

This paper discusses the fitting of the proportional hazards model to interval-censored failure time data with missing covariates. Many authors have discussed the problem when complete covariate information is available or the missing is completely at random. In contrast to this, we will focus on the situation where the missing is at random. For the problem, a sieve maximum likelihood estimation approach is proposed with the use of I-spline functions to approximate the unknown cumulative baseline hazard function in the model. For the implementation of the proposed method, we develop an EM algorithm based on a two-stage data augmentation. Furthermore, we show that the proposed estimators of regression parameters are consistent and asymptotically normal. The proposed approach is then applied to a set of the data concerning Alzheimer Disease that motivated this study.

The moderating role of gender in the association between quality of social relationships and sleep

To determine whether the association between perceived social support or strain in close relationships and sleep outcomes varies by gender. Participants were selected from the Biomarker projects of either the MIDUS II or MIDUS Refresher study if they were in a married-or married-like relationship and shared a bed with their partner (N = 989). A subsample also participated in a seven-day sleep study (n = 282). Perceived social support and strain from partner, family, and friends were examined by self-report questionnaires. We used the Pittsburgh Sleep Quality Index, sleep daily diary, and actigraphy to measure both subjective and objective sleep. Social support and strain were both associated with sleep outcomes. Specifically, higher social support was associated with fewer daily reports of light sleep and feeling more rested in the morning, while higher social strain was associated with higher clinical sleep disturbance. For women, but not men, social support was significantly associated with lower daily sleep disturbance while perceived social strain was significantly associated with higher daily sleep disturbance, lighter sleep, feeling less rested in the morning, lower sleep efficiency, and longer sleep onset latency. Mainly among women, social support and strain are associated with an important transdiagnostic health outcome-sleep-which may have implications for a wide range of health disparities. Interpersonal stressors may increase health risks differently for women compared to men and one mechanism that may link social relationships to long-term health outcomes is sleep.

Efficacy and safety of Mydriatic Microdrops for Retinopathy Of Prematurity Screening (MyMiROPS): study protocol for a non-inferiority crossover randomized controlled trial

Background

Retinopathy of prematurity (ROP) eye examination screening presupposes adequate mydriasis for an informative fundoscopy of preterm infants at risk, on a weekly basis. Systemic absorption of the instilled mydriatic regimens has been associated with various adverse events in this fragile population. This report aims to present the fully developed protocol of a full-scale trial for testing the hypothesis that the reduced mydriatic drop volume achieves adequate mydriasis while minimizing systemic adverse events.

Methods

A non-inferiority crossover randomized controlled trial will be performed to study the efficacy and safety of combined phenylephrine 1.67% and tropicamide 0.33% microdrops compared with standard drops in a total of 93 preterm infants requiring ROP screening. Primary outcome will be the pupil diameter at 45 (T45) min after instillation. Pupil diameter at T90 and T120 will constitute secondary efficacy endpoints. Mixed-effects linear regression models will be developed, and the 95% confidence interval approach will be used for assessing non-inferiority. Whole blood samples will be analyzed using hydrophilic liquid chromatography–tandem mass spectrometry method (HILIC–MS/MS), for gathering pharmacokinetic (PK) data on the instilled phenylephrine, at nine specific time points within 3 h from mydriasis. Pooled PK data will be used due to ethical restrictions on having a full PK profile per infant. Heart rate, oxygen saturation, blood pressure measurements, and 48-h adverse events will also be recorded.

Discussion

This protocol is designed for a study powered to assess non-inferiority of microdrops compared with standard dilating drops. If our hypothesis is confirmed, microdrops may become a useful tool in ROP screening.

Trial registration

ClinicalTrials.govNCT05043077. Registered on 2 September 2021

Dietary fat intake and risk of Parkinson disease: results from the Swedish National March Cohort

Background

Following progressive aging of the population worldwide, the prevalence of Parkinson disease is expected to increase in the next decades. Primary prevention of the disease is hampered by limited knowledge of preventable causes. Recent evidence regarding diet and Parkinson disease is inconsistent and suggests that dietary habits such as fat intake may have a role in the etiology.

Objective

To investigate the association between intake of total and specific types of fat with the incidence of Parkinson disease.

Methods

Participants from the Swedish National March Cohort were prospectively followed-up from 1997 to 2016. Dietary intake was assessed at baseline using a validated food frequency questionnaire. Food items intake was used to estimate fat intake, i.e. the exposure variable, using the Swedish Food Composition Database. Total, saturated, monounsaturated and polyunsaturated fat intake were categorized into quartiles. Parkinson disease incidence was ascertained through linkages to Swedish population-based registers. Cox proportional hazards regression models were used to estimate hazard ratios (HR) with 95% confidence intervals (CI) of the association between fat intake from total or specific types of fats and the incidence of Parkinson disease. The lowest intake category was used as reference. Isocaloric substitution models were also fitted to investigate substitution effects by replacing energy from fat intake with other macronutrients or specific types of fat.

Results

41,597 participants were followed up for an average of 17.6 years. Among them, 465 developed Parkinson disease. After adjusting for potential confounders, the highest quartile of saturated fat intake was associated with a 41% increased risk of Parkinson disease compared to the lowest quartile (HR Q4 vs. Q1: 1.41; 95% CI: 1.04–1.90; p for trend: 0.03). Total, monounsaturated or polyunsaturated fat intake were not significantly associated with Parkinson disease. The isocaloric substitution models did not show any effect.

Conclusions

We found that a higher consumption of large amounts of saturated fat might be associated with an increased risk of Parkinson disease. A diet low in saturated fat might be beneficial for disease prevention.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10654-022-00863-8.

Comparing single and multiple imputation strategies for harmonizing substance use data across HIV-related cohort studies

Background

Although standardized measures to assess substance use are available, most studies use variations of these measures making it challenging to harmonize data across studies. The aim of this study was to evaluate the performance of different strategies to impute missing substance use data that may result as part of data harmonization procedures.

Methods

We used self-reported substance use data collected between August 2014 and June 2019 from 528 participants with 2,389 study visits in a cohort study of substance use and HIV. We selected a low (heroin), medium (methamphetamine), and high (cannabis) prevalence drug and set 10–50% of each substance to missing. The data amputation mimicked missingness that results from harmonization of disparate measures. We conducted Monte Carlo simulations to evaluate the comparative performance of single and multiple imputation (MI) methods using the relative mean bias, root mean square error (RMSE), and coverage probability of the 95% confidence interval for each imputed estimate.

Results

Without imputation (i.e., listwise deletion), estimates of substance use were biased, especially for low prevalence outcomes such as heroin. For instance, even when 10% of data were missing, the complete case analysis underestimated the prevalence of heroin by 33%. MI, even with as few as five imputations produced the least biased estimates, however, for a high prevalence outcome such as cannabis with low to moderate missingness, performance of single imputation strategies improved. For instance, in the case of cannabis, with 10% missingness, single imputation with regression performed just as well as multiple imputation resulting in minimal bias (relative mean bias of 0.06% and 0.07% respectively) and comparable performance (RMSE = 0.0102 for both and coverage of 95.8% and 96.2% respectively).

Conclusion

Our results from imputation of missing substance use data resulting from data harmonization indicate that MI provided the best performance across a range of conditions. Additionally, single imputation for substance use data performed comparably under scenarios where the prevalence of the outcome was high and missingness was low. These findings provide a practical application for the evaluation of several imputation strategies and helps to address missing data problem when combining data from individual studies.

Impute the missing data using retrieved dropouts

Background

In the past few decades various methods have been proposed to handle missing data of clinical studies, so as to assess the robustness of primary results. Some of the methods are based on the assumption of missing at random (MAR) which assumes subjects who discontinue the treatment will maintain the treatment effect after discontinuation. The agency, however, has expressed concern over methods based on this overly optimistic assumption, because it hardly holds for subjects discontinuing the investigational drug. Although in recent years a good number of sensitivity analyses based on missing not at random (MNAR) assumptions have been proposed, some use very conservative assumption on which it might be hard for sponsors and regulators to reach common ground.

Methods

Here we propose a multiple imputation method targeting at “treatment policy” estimand based on the MNAR assumption. This method can be used as the primary analysis, in addition to serving as a sensitivity analysis. It imputes missing data using information from retrieved dropouts defined as subjects who remain in the study despite occurrence of intercurrent events. Then imputed data long with completers and retrieved dropouts are analyzed altogether and finally multiple results are summarized into a single estimate. According to definition in ICH E9 (R1), this proposed approach fully aligns with the treatment policy estimand but its assumption is much more realistic and reasonable.

Results

Our approach has well controlled type I error rate with no loss of power. As expected, the effect size estimates take into account any dilution effect contributed by retrieved dropouts, conforming to the MNAR assumption.

Conclusions

Although multiple imputation approaches are always used as sensitivity analyses, this multiple imputation approach can be used as primary analysis for trials with sufficient retrieved dropouts or trials designed to collect retrieved dropouts.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-022-01509-9.

Use of Linked Databases for Improved Confounding Control: Considerations for Potential Selection Bias

Pharmacoepidemiologic studies are increasingly conducted within linked databases, often to obtain richer confounder data. However, the potential for selection bias is frequently overlooked when linked data is available only for a subset of patients. We highlight the importance of accounting for potential selection bias by evaluating the association between antipsychotics and type 2 diabetes in youths within a claims database linked to a smaller laboratory database. We used inverse probability of treatment weights (IPTW) to control for confounding. In analyses restricted to the linked cohorts, we applied inverse probability of selection weights (IPSW) to create a population representative of the full cohort. We used pooled logistic regression weighted by IPTW only or IPTW and IPSW to estimate treatment effects. Metabolic conditions were more prevalent in linked cohorts compared with the full cohort. Within the full cohort, the confounding-adjusted hazard ratio was 2.26 (95% CI: 2.07, 2.49) comparing initiation of antipsychotics with initiation of control medications. Within the linked cohorts, a different magnitude of association was obtained without adjustment for selection, whereas applying IPSW resulted in point estimates similar to the full cohort's (e.g., an adjusted hazard ratio of 1.63 became 2.12). Linked database studies may generate biased estimates without proper adjustment for potential selection bias.

Methods for Modeling Autocorrelation and Handling Missing Data in Mediation Analysis in Single Case Experimental Designs (SCEDs)

Single-Case Experimental Designs (SCEDs) are increasingly recognized as a valuable alternative to group designs. Mediation analysis is useful in SCEDs contexts because it informs researchers about the underlying mechanism through which an intervention influences the outcome. However, methods for conducting mediation analysis in SCEDs have only recently been proposed. Furthermore, repeated measures of a target behavior present the challenges of autocorrelation and missing data. This paper aims to extend methods for estimating indirect effects in piecewise regression analysis in SCEDs by (1) evaluating three methods for modeling autocorrelation, namely, Newey-West (NW) estimation, feasible generalized least squares (FGLS) estimation, and explicit modeling of an autoregressive structure of order one (AR(1)) in the error terms and (2) evaluating multiple imputation in the presence of data that are missing completely at random. FGLS and AR(1) outperformed NW and OLS estimation in terms of efficiency, Type I error rates, and coverage, while OLS was superior to the methods in terms of power for larger samples. The performance of all methods is consistent across 0% and 20% missing data conditions. 50% missing data led to unsatisfactory power and biased estimates. In light of these findings, we provide recommendations for applied researchers.

Effects of chemotherapy on contralateral breast cancer risk in BRCA1 and BRCA2 mutation carriers: A nationwide cohort study

Aim

BRCA1/2 mutation carriers with primary breast cancer (PBC) are at high risk of contralateral breast cancer (CBC). In a nationwide cohort, we investigated the effects of chemotherapeutic agents given for PBC on CBC risk separately in BRCA1 and BRCA2 mutation carriers.

Patients and methods

BRCA1 or BRCA2 mutation carriers with an invasive PBC diagnosis from 1990 to 2017 were selected from a Dutch cohort. We estimated cumulative CBC incidence using competing risks analysis. Hazard ratios (HR) for the effect of neo-adjuvant or adjuvant chemotherapy and different chemotherapeutic agents on CBC risk were estimated using Cox regression.

Results

We included 1090 BRCA1 and 568 BRCA2 mutation carriers; median follow-up was 8.9 and 8.4 years, respectively. Ten-year cumulative CBC incidence for treatment with and without chemotherapy was 6.7% [95%CI: 5.1–8.6] and 16.7% [95%CI: 10.8–23.7] in BRCA1 and 4.8% [95%CI: 2.7–7.8] and 16.0% [95%CI: 9.3–24.4] in BRCA2 mutation carriers, respectively. Chemotherapy was associated with reduced CBC risk in BRCA1 (multivariable HR: 0.46, 95%CI: 0.29–0.74); a similar trend was observed in BRCA2 mutation carriers (HR: 0.63, 95%CI: 0.29–1.39). In BRCA1, risk reduction was most pronounced in the first 5 years (HR: 0.32, 95%CI: 0.17–0.61). Anthracyclines and the combination of anthracyclines with taxanes were associated with substantial CBC risk reduction in BRCA1 carriers (HR: 0.34, 95%CI: 0.17–0.68 and HR: 0.22, 95%CI: 0.08–0.62, respectively).

Conclusion

Risk-reducing effects of chemotherapy are substantial for at least 5 years and may be used in personalised CBC risk prediction in any case for BRCA1 mutation carriers.

•

Contralateral breast cancer (CBC) risk is high in BRCA1/2 mutation carriers.

•

Chemotherapy for primary breast cancer results in decreased CBC risk in BRCA1.

•

Anthracyclines with/without taxanes show the largest CBC risk reduction in BRCA1.

•

For BRCA2 similar trends are observed as in BRCA1 mutation carriers.

•

Chemotherapy must be considered in personalised CBC risk models.

Assessing performance of artificial neural networks and re-sampling techniques for healthcare datasets

Re-sampling methods to solve class imbalance problems have shown to improve classification accuracy by mitigating the bias introduced by differences in class size. However, it is possible that a model which uses a specific re-sampling technique prior to Artificial neural networks (ANN) training may not be suitable for aid in classifying varied datasets from the healthcare industry. Five healthcare-related datasets were used across three re-sampling conditions: under-sampling, over-sampling and combi-sampling. Within each condition, different algorithmic approaches were applied to the dataset and the results were statistically analysed for a significant difference in ANN performance. The combi-sampling condition showed that four out of the five datasets did not show significant consistency for the optimal re-sampling technique between the f1-score and Area Under the Receiver Operating Characteristic Curve performance evaluation methods. Contrarily, the over-sampling and under-sampling condition showed all five datasets put forward the same optimal algorithmic approach across performance evaluation methods. Furthermore, the optimal combi-sampling technique (under-, over-sampling and convergence point), were found to be consistent across evaluation measures in only two of the five datasets. This study exemplifies how discrete ANN performances on datasets from the same industry can occur in two ways: how the same re-sampling technique can generate varying ANN performance on different datasets, and how different re-sampling techniques can generate varying ANN performance on the same dataset.

Risk factors associated with severe disease in respiratory syncytial virus infected children under 5 years of age

OBJECTIVE

To evaluate risk factors for severe disease in children under 59 months of age hospitalized with respiratory syncytial virus (RSV) infection.

STUDY DESIGN

We prospectively enrolled 1,096 cases of laboratory confirmed RSV infection during three consecutive RSV seasons in 2015-2018. Potential risk factors for severe disease were retrieved through patient questionnaires and linkage to national health registries. Need for respiratory support (invasive ventilation, bi-level positive airway pressure, or continuous positive airway pressure), and length of stay exceeding 72 h were used as measures of disease severity. Associations were investigated using multivariable logistic regression analyses. Multiple imputation was used to avoid bias and inference induced by missing data.

RESULTS

Risk factors associated with a need for respiratory support included age younger than 3 months of age [aOR: 6.73 (95% CI 2.71-16.7)], having siblings [aOR: 1.65 (95% CI 1.05-2.59)] and comorbidity [aOR: 2.40 (95% CI 1.35-4.24)]. The length of hospital stay >72 h was significantly associated with being younger than 3 months of age [aOR: 3.52 (95% CI 1.65-7.54)], having siblings [aOR: 1.45 (95% CI 1.01-2.08)], and comorbidity [aOR: 2.18 (95% CI 1.31-3.61)]. Sub-group analysis of children younger than 6 months of age confirmed the association between both young age and having siblings and the need for respiratory support.

CONCLUSION

In a large cohort of children <59 months hospitalized with RSV infection, young age, comorbidity, and having siblings were associated with more severe disease.

On the Treatment of Missing Item Responses in Educational Large-Scale Assessment Data: An Illustrative Simulation Study and a Case Study Using PISA 2018 Mathematics Data

Missing item responses are prevalent in educational large-scale assessment studies such as the programme for international student assessment (PISA). The current operational practice scores missing item responses as wrong, but several psychometricians have advocated for a model-based treatment based on latent ignorability assumption. In this approach, item responses and response indicators are jointly modeled conditional on a latent ability and a latent response propensity variable. Alternatively, imputation-based approaches can be used. The latent ignorability assumption is weakened in the Mislevy-Wu model that characterizes a nonignorable missingness mechanism and allows the missingness of an item to depend on the item itself. The scoring of missing item responses as wrong and the latent ignorable model are submodels of the Mislevy-Wu model. In an illustrative simulation study, it is shown that the Mislevy-Wu model provides unbiased model parameters. Moreover, the simulation replicates the finding from various simulation studies from the literature that scoring missing item responses as wrong provides biased estimates if the latent ignorability assumption holds in the data-generating model. However, if missing item responses are generated such that they can only be generated from incorrect item responses, applying an item response model that relies on latent ignorability results in biased estimates. The Mislevy-Wu model guarantees unbiased parameter estimates if the more general Mislevy-Wu model holds in the data-generating model. In addition, this article uses the PISA 2018 mathematics dataset as a case study to investigate the consequences of different missing data treatments on country means and country standard deviations. Obtained country means and country standard deviations can substantially differ for the different scaling models. In contrast to previous statements in the literature, the scoring of missing item responses as incorrect provided a better model fit than a latent ignorable model for most countries. Furthermore, the dependence of the missingness of an item from the item itself after conditioning on the latent response propensity was much more pronounced for constructed-response items than for multiple-choice items. As a consequence, scaling models that presuppose latent ignorability should be refused from two perspectives. First, the Mislevy-Wu model is preferred over the latent ignorable model for reasons of model fit. Second, in the discussion section, we argue that model fit should only play a minor role in choosing psychometric models in large-scale assessment studies because validity aspects are most relevant. Missing data treatments that countries can simply manipulate (and, hence, their students) result in unfair country comparisons.

Missing values compensation in duplicates detection using hot deck method

Duplicate record is a common problem within data sets especially in huge volume databases. The accuracy of duplicate detection determines the efficiency of duplicate removal process. However, duplicate detection has become more challenging due to the presence of missing values within the records where during the clustering and matching process, missing values can cause records deemed similar to be inserted into the wrong group, hence, leading to undetected duplicates. In this paper, duplicate detection improvement was proposed despite the presence of missing values within a data set through Duplicate Detection within the Incomplete Data set (DDID) method. The missing values were hypothetically added to the key attributes of three data sets under study, using an arbitrary pattern to simulate both complete and incomplete data sets. The results were analyzed, then, the performance of duplicate detection was evaluated by using the Hot Deck method to compensate for the missing values in the key attributes. It was hypothesized that by using Hot Deck, duplicate detection performance would be improved. Furthermore, the DDID performance was compared to an early duplicate detection method namely DuDe, in terms of its accuracy and speed. The findings yielded that even though the data sets were incomplete, DDID was able to offer a better accuracy and faster duplicate detection as compared to DuDe. The results of this study offer insights into constraints of duplicate detection within incomplete data sets.

The Course of Weight/Shape Concerns and Disordered Eating Symptoms Among Adolescent and Young Adult Males

PURPOSE

Male weight concerns tend to focus on shape and muscularity as opposed to a desire for thinness and remain underdetected by conventional eating disorder assessments. We aimed to describe the longitudinal course of weight concerns and disordered eating behaviors among males across adolescence and young adulthood.

METHODS

We used prospective assessments of 4,489 U S. males, aged 11 to 18 years at baseline of analyses, in the Growing Up Today Study. We assigned mutually exclusive classifications of behaviors consistent with bulimia nervosa (BN), binge eating disorder (BED), purging disorder (PD); high levels of concern with thinness and/or muscularity; and use of muscle-enhancing products. We estimated the probability of maintenance, resolution, or transition to different weight concerns and/or disordered eating behaviors across consecutive survey waves.

RESULTS

Less than 1% of participants met full or partial criteria for BN, PD, or BED at baseline. One-quarter (25.4%, n = 1,137) of males reported high weight concerns during follow-up; nearly all these cases (93.7%, n = 1,065) had high muscularity concerns. The most common transition in concerns or behaviors involved the addition of muscularity concerns to a preoccupation with thinness. Eleven percent of participants used muscle-building products during follow-up. Multi-year product use (23.0% [standard deviation 1.0%] of males who used products) was more common than maintenance of bulimic behaviors (3.0% [.7%] of BN/PD, 10.5% [1.2%] of BED cases).

CONCLUSIONS

Integrating muscularity concerns and product use into health promotion and screening tools may improve prevention and early detection of harmful body image and weight control among adolescent and young adult males.

Prenatal exposure to a mixture of elements and neurobehavioral outcomes in mid-childhood: Results from Project Viva

BACKGROUND

Lead (Pb), manganese (Mn), selenium (Se) and methylmercury (MeHg) can be neurotoxic individually, despite Mn and Se also being essential elements. Little is known about the joint effects of essential and non-essential elements on neurobehavior, particularly for prenatal exposures.

OBJECTIVES

To evaluate associations of prenatal exposure to multiple elements with executive function and neurobehavior in children.

METHODS

Participants included 1009 mother-child pairs from the Project Viva pre-birth cohort. We estimated maternal erythrocyte Pb, Mn, Se, and Hg concentrations prenatally. In 6-11-year old children (median 7.6 years), parents and teachers rated children's executive function-related behaviors using the Behavior Rating Inventory of Executive Function (BRIEF) Global Executive Composite score and behavioral difficulties using the Strengths and Difficulties Questionnaire (SDQ) total difficulties score. We evaluated associations of element mixtures with neurobehavior using Bayesian kernel machine regression (BKMR), multivariable linear regression, and quantile g-computation.

RESULTS

Median erythrocyte Pb, Mn, Se, and Hg concentrations were 1.1 μg/dL, 33.1 μg/L, 204.5 ng/mL, and 3.1 ng/g, respectively. Findings from BKMR and quantile g-computation models both showed worse (higher) parent-rated BRIEF and SDQ z-scores with higher concentrations of the mixture, although estimates were imprecise. When remaining elements were set at their median within BKMR models, increases in Pb and Se from the 25th to 75th percentile of exposure distributions were associated with 0.08 (95% CI: 0.02, 0.19) and 0.07 (95% CI: 0.03, 0.16) standard deviation increases in parent-rated BRIEF scores, and 0.08 (95% CI: 0.02, 0.17) and 0.05 (95% CI: 0.03, 0.13) standard deviation increases in SDQ scores, respectively. There was no evidence of element interactions.

DISCUSSION

Although associations were small in magnitude, we found a trend of worsening neurobehavioral ratings with increasing prenatal exposure to an element mixture. However, we may be observing a limited range of dose-dependent impacts given the levels of exposure within our population.

Activities, Housing Situation and Other Factors Influencing Psychological Strain Experienced During the First COVID-19 Lockdown in Switzerland

Background: The Coronavirus disease 2019 (COVID-19) crisis and the corresponding first nationwide lockdown from mid-March to 10 May 2020 engendered considerable psychological strain among people in Switzerland. This study analyzes determinants of changes in subjective levels of psychological strain experienced during the lockdown. Methods: An online survey conducted as part of a larger mixed methods study examined the material and emotional aspects of individual reactions to the lockdown from a socio-ecological perspective. Participants (N = 5932) were asked about their personal and employment status, housing features, changes in various activities (e.g., physical activity, watching TV, social media use) and aspects of mental distress and well-being. Results: A substantial share of participants reported to feel depressed (33%) and anxious (43%) more often during the COVID-19 lockdown than before, whereas significantly (p < 0.001) less persons reported a decrease of these negative feelings (depressed 17%; anxious 14%). Women, single people, students and people who lost their jobs or were temporally unemployed due to the lockdown experienced a particularly strong increase of subjective psychological strain. Important residential factors reducing subjective psychological strain were the general comfort of the housing situation and having a private garden or multiple types of outdoor space. Considering leisure activities, the strongest positive psychological effect resulted from increased physical activities, followed by reading and cooking. However, 45% of the participants reported a decreased frequency of physical activity during the lockdown compared to before, whereas significantly less persons (26%) reported a corresponding increase (p < 0.001). Conclusion: Consistent with other studies, the results indicate a substantial reduction of subjective psychological well-being of the population during the first COVID-19 lockdown in Switzerland. The psychological burdens which the participants experienced differ depending on personal characteristics and situational factors. Negative psychological and economic consequences and gender inequalities should accordingly be carefully considered and actively prevented when designing COVID-19 measures. Supportive economic and social, cognitive and behavioral psychological interventions need to be designed and implemented to maintain the well-being of residents during lockdown.

Psychological factors and trajectories of post-surgical pain: A longitudinal prospective study

A significant proportion of patients do not experience relief from pain during the early postsurgical period after joint arthroplasty and are at risk for developing chronic pain. The objectives of this study were to identify biopsychosocial factors associated with acute postsurgical pain trajectories and with pain intensity and interference after 1, 3, and 12 months. Two hundred ten patients listed for joint arthroplasty filled a presurgical battery of questionnaires assessing presurgical pain intensity, catastrophizing, emotional distress, state anxiety and depression, self-efficacy, central sensitization, and executive functions. From the day after surgery, they were asked to fill a 7-day diary, including questions about postsurgical pain and postsurgical state catastrophizing. Finally, they provided data about pain intensity and interference after 1, 3, and 12 months. Predictors of acute pain trajectories were investigated using multilevel growth curve analysis. Results showed that central sensitization was a predictor of the intercept of pain trajectories and daily postsurgical catastrophizing was a significant covariate of pain intensity in the acute phase. Analyses of follow-up data showed that central sensitization was a predictor of pain intensity and pain interference at 3 and 12 months, that emotional distress was related with pain intensity and interference at 1 month, and with pain interference at 3 months, and that cognitive flexibility was associated with pain interference at 1 month. Assessment of these factors could enable to identify patients at risk for worse outcomes and to plan targeted treatments to be implemented during the patient's inward stay.

Revisiting performance metrics for prediction with rare outcomes

Machine learning algorithms are increasingly used in the clinical literature, claiming advantages over logistic regression. However, they are generally designed to maximize the area under the receiver operating characteristic curve. While area under the receiver operating characteristic curve and other measures of accuracy are commonly reported for evaluating binary prediction problems, these metrics can be misleading. We aim to give clinical and machine learning researchers a realistic medical example of the dangers of relying on a single measure of discriminatory performance to evaluate binary prediction questions. Prediction of medical complications after surgery is a frequent but challenging task because many post-surgery outcomes are rare. We predicted post-surgery mortality among patients in a clinical registry who received at least one aortic valve replacement. Estimation incorporated multiple evaluation metrics and algorithms typically regarded as performing well with rare outcomes, as well as an ensemble and a new extension of the lasso for multiple unordered treatments. Results demonstrated high accuracy for all algorithms with moderate measures of cross-validated area under the receiver operating characteristic curve. False positive rates were <1%, however, true positive rates were <7%, even when paired with a 100% positive predictive value, and graphical representations of calibration were poor. Similar results were seen in simulations, with the addition of high area under the receiver operating characteristic curve (>90%) accompanying low true positive rates. Clinical studies should not primarily report only area under the receiver operating characteristic curve or accuracy.

Development of a "meta-model" to address missing data, predict patient-specific cancer survival and provide a foundation for clinical decision support

Objective

Like most real-world data, electronic health record (EHR)–derived data from oncology patients typically exhibits wide interpatient variability in terms of available data elements. This interpatient variability leads to missing data and can present critical challenges in developing and implementing predictive models to underlie clinical decision support for patient-specific oncology care. Here, we sought to develop a novel ensemble approach to addressing missing data that we term the “meta-model” and apply the meta-model to patient-specific cancer prognosis.

Materials and Methods

Using real-world data, we developed a suite of individual random survival forest models to predict survival in patients with advanced lung cancer, colorectal cancer, and breast cancer. Individual models varied by the predictor data used. We combined models for each cancer type into a meta-model that predicted survival for each patient using a weighted mean of the individual models for which the patient had all requisite predictors.

Results

The meta-model significantly outperformed many of the individual models and performed similarly to the best performing individual models. Comparisons of the meta-model to a more traditional imputation-based method of addressing missing data supported the meta-model’s utility.

Conclusions

We developed a novel machine learning–based strategy to underlie clinical decision support and predict survival in cancer patients, despite missing data. The meta-model may more generally provide a tool for addressing missing data across a variety of clinical prediction problems. Moreover, the meta-model may address other challenges in clinical predictive modeling including model extensibility and integration of predictive algorithms trained across different institutions and datasets.

What happens when performance-based financing meets free healthcare? Evidence from an interrupted time-series analysis

In spite of the wide attention performance-based financing (PBF) has received over the past decade, no evidence is available on its impacts on quantity and mix of service provision nor on its interaction with parallel health financing interventions. Our study aimed to examine the PBF impact on quantity and mix of service provision in Burkina Faso, while accounting for the parallel introduction of a free healthcare policy. We used Health Management Information System data from 838 primary-level health facilities across 24 districts and relied on an interrupted time-series analysis with independent controls. We placed two interruptions, one to account for PBF and one to account for the free healthcare policy. In the period before the free healthcare policy, PBF produced significant but modest increases across a wide range of maternal and child services, but a significant decrease in child immunization coverage. In the period after the introduction of the free healthcare policy, PBF did not affect service provision in intervention compared with control facilities, possibly indicating a saturation effect. Our findings indicate that PBF can produce modest increases in service provision, without altering the overall service mix. Our findings, however, also indicate that the introduction of other health financing reforms can quickly crowd out the effects produced by PBF. Further qualitative research is required to understand what factors allow healthcare providers to increase the provision of some, but not all services and how they react to the joint implementation of PBF and free health care.