Bias and efficiency of multiple imputation compared with complete-case analysis for missing covariate values

Comparing methods for handling missing covariates in meta-regression

Meta-analysts often encounter missing covariate values when estimating meta-regression models. In practice, ad hoc approaches involving data deletion have been widely used. The current study investigates the performance of different methods for handling missing covariates in meta-regression, including complete-case analysis (CCA), shifting-case analysis (SCA), multiple imputation (MI), and full information maximum likelihood (FIML), assuming missing at random mechanism. According to the simulation results, we advocate the use of MI and FIML than CCA and SCA approaches in practice. In addition, we cautiously note the challenges and potential advantages of using MI in the meta-analysis context.

Clinical Phenotypes of Dual Kidney Transplant Recipients in the United States as Identified through Machine Learning Consensus Clustering

Background and Objectives: Our study aimed to cluster dual kidney transplant recipients using an unsupervised machine learning approach to characterize donors and recipients better and to compare the survival outcomes across these various clusters. Materials and Methods: We performed consensus cluster analysis based on recipient-, donor-, and transplant-related characteristics in 2821 dual kidney transplant recipients from 2010 to 2019 in the OPTN/UNOS database. We determined the important characteristics of each assigned cluster and compared the post-transplant outcomes between clusters. Results: Two clinically distinct clusters were identified by consensus cluster analysis. Cluster 1 patients was characterized by younger patients (mean recipient age 49 ± 13 years) who received dual kidney transplant from pediatric (mean donor age 3 ± 8 years) non-expanded criteria deceased donor (100% non-ECD). In contrast, Cluster 2 patients were characterized by older patients (mean recipient age 63 ± 9 years) who received dual kidney transplant from adult (mean donor age 59 ± 11 years) donor with high kidney donor profile index (KDPI) score (59% had KDPI ≥ 85). Cluster 1 had higher patient survival (98.0% vs. 94.6% at 1 year, and 92.1% vs. 76.3% at 5 years), and lower acute rejection (4.2% vs. 6.1% within 1 year), when compared to cluster 2. Death-censored graft survival was comparable between two groups (93.5% vs. 94.9% at 1 year, and 89.2% vs. 84.8% at 5 years). Conclusions: In summary, DKT in the United States remains uncommon. Two clusters, based on specific recipient and donor characteristics, were identified through an unsupervised machine learning approach. Despite varying differences in donor and recipient age between the two clusters, death-censored graft survival was excellent and comparable. Broader utilization of DKT from high KDPI kidneys and pediatric en bloc kidneys should be encouraged to better address the ongoing organ shortage.

Physically fit with a higher cancer risk? Influences of cervical cancer screening among a sample of physically active women ages 21-49 living in the United States

To achieve the lowest risk level for various cancers, individuals would engage in several healthy lifestyle behaviors and age-eligible cancer screenings as recommended. Nonetheless, research has largely omitted exploration of concurrent primary and secondary prevention behaviors. This study was designed to explore influences of cervical cancer screening among physically active women who reported participation in recreational sports. U.S. based women between the ages of 21-49, who had never been diagnosed with cancer, were eligible to complete a web-based survey. Logistic regression analyses were conducted using SAS 9.4. On average, women were 31 years of age (N = 394) and self-identified as Black (51.3 %). Although low overall (30.7 %), higher odds of cervical cancer screening were associated with age (OR = 1.06, 95 % CI = 1.03-1.10), employment (OR = 2.43, 95 % CI = 1.14-5.18), knowledge of cancer-related risk behaviors (OR = 4.04, 95 % CI = 1.33-12.28), routine doctor's visit (OR = 4.25, 95 % CI = 1.56-11.54), and team-based vs individual-based sport participation (OR = 1.95, 95 % CI = 1.13-3.34). Our study provides insight into the health profile of physically active women, ages 21-49, as it relates to risks for cervical cancer. Screening uptake among this diverse sample was much lower than the general population and national goals set by Healthy People 2030. Interventions should be tailored to increase knowledge of cancer-related risk behaviors, access to healthcare, and recommended cervical cancer screenings among even assumed-to-be healthy populations.

Clinical characteristics do not reliably identify non-adherence in patients with uncontrolled hypertension

PURPOSE

Chemical adherence testing is a reliable method to assess adherence to antihypertensive drugs. However, it is expensive and has limited availability in clinical practice. To reduce the number and costs of chemical adherence tests, we aimed to develop and validate a clinical screening tool to identify patients with a low probability of non-adherence in patients with uncontrolled hypertension.

MATERIALS AND METHODS

In 495 patients with uncontrolled hypertension referred to the University Medical Centre Utrecht (UMCU), the Netherlands, a penalised logistic regression model including seven pre-specified easy-to-measure clinical variables was derived to estimate the probability of non-adherence. Non-adherence was defined as not detecting at least one of the prescribed antihypertensive drugs in plasma or urine. Model performance and test characteristics were evaluated in 240 patients with uncontrolled hypertension referred to the Heartlands Hospital, United Kingdom.

RESULTS

Prevalence of non-adherence to antihypertensive drugs was 19% in the UMCU and 44% in the Heartlands Hospital population. After recalibration of the model's intercept, predicted probabilities agreed well with observed frequencies. The c-statistic of the model was 0.63 (95%CI 0.53-0.72). Predicted probability cut-off values of 15%-22.5% prevented testing in 5%-15% of the patients, carrying sensitivities between 97% (64-100) and 90% (80-95), and negative predictive values between 74% (10-99) and 70% (50-85).

CONCLUSION

The combination of seven clinical variables is not sufficient to reliably discriminate adherent from non-adherent individuals to safely reduce the number of chemical adherence tests. This emphasises the complex nature of non-adherence behaviour and thus the need for objective chemical adherence tests in patients with uncontrolled hypertension.

Characteristics of Kidney Recipients of High Kidney Donor Profile Index Kidneys as Identified by Machine Learning Consensus Clustering

Background: Our study aimed to characterize kidney transplant recipients who received high kidney donor profile index (KDPI) kidneys using unsupervised machine learning approach. Methods: We used the OPTN/UNOS database from 2010 to 2019 to perform consensus cluster analysis based on recipient-, donor-, and transplant-related characteristics in 8935 kidney transplant recipients from deceased donors with KDPI ≥ 85%. We identified each cluster’s key characteristics using the standardized mean difference of >0.3. We compared the posttransplant outcomes among the assigned clusters. Results: Consensus cluster analysis identified 6 clinically distinct clusters of kidney transplant recipients from donors with high KDPI. Cluster 1 was characterized by young, black, hypertensive, non-diabetic patients who were on dialysis for more than 3 years before receiving kidney transplant from black donors; cluster 2 by elderly, white, non-diabetic patients who had preemptive kidney transplant or were on dialysis less than 3 years before receiving kidney transplant from older white donors; cluster 3 by young, non-diabetic, retransplant patients; cluster 4 by young, non-obese, non-diabetic patients who received dual kidney transplant from pediatric, black, non-hypertensive non-ECD deceased donors; cluster 5 by low number of HLA mismatch; cluster 6 by diabetes mellitus. Cluster 4 had the best patient survival, whereas cluster 3 had the worst patient survival. Cluster 2 had the best death-censored graft survival, whereas cluster 4 and cluster 3 had the worst death-censored graft survival at 1 and 5 years, respectively. Cluster 2 and cluster 4 had the best overall graft survival at 1 and 5 years, respectively, whereas cluster 3 had the worst overall graft survival. Conclusions: Unsupervised machine learning approach kidney transplant recipients from donors with high KDPI based on their pattern of clinical characteristics into 6 clinically distinct clusters.

Methods for handling missing data in serially sampled sputum specimens for mycobacterial culture conversion calculation

BACKGROUND

The occurrence and timing of mycobacterial culture conversion is used as a proxy for tuberculosis treatment response. When researchers serially sample sputum during tuberculosis studies, contamination or missed visits leads to missing data points. Traditionally, this is managed by ignoring missing data or simple carry-forward techniques. Statistically advanced multiple imputation methods potentially decrease bias and retain sample size and statistical power.

METHODS

We analyzed data from 261 participants who provided weekly sputa for the first 12 weeks of tuberculosis treatment. We compared methods for handling missing data points in a longitudinal study with a time-to-event outcome. Our primary outcome was time to culture conversion, defined as two consecutive weeks with no Mycobacterium tuberculosis growth. Methods used to address missing data included: 1) available case analysis, 2) last observation carried forward, and 3) multiple imputation by fully conditional specification. For each method, we calculated the proportion culture converted and used survival analysis to estimate Kaplan-Meier curves, hazard ratios, and restricted mean survival times. We compared methods based on point estimates, confidence intervals, and conclusions to specific research questions.

RESULTS

The three missing data methods lead to differences in the number of participants achieving conversion; 78 (32.8%) participants converted with available case analysis, 154 (64.7%) converted with last observation carried forward, and 184 (77.1%) converted with multiple imputation. Multiple imputation resulted in smaller point estimates than simple approaches with narrower confidence intervals. The adjusted hazard ratio for smear negative participants was 3.4 (95% CI 2.3, 5.1) using multiple imputation compared to 5.2 (95% CI 3.1, 8.7) using last observation carried forward and 5.0 (95% CI 2.4, 10.6) using available case analysis.

CONCLUSION

We showed that accounting for missing sputum data through multiple imputation, a statistically valid approach under certain conditions, can lead to different conclusions than naïve methods. Careful consideration for how to handle missing data must be taken and be pre-specified prior to analysis. We used data from a TB study to demonstrate these concepts, however, the methods we described are broadly applicable to longitudinal missing data. We provide valuable statistical guidance and code for researchers to appropriately handle missing data in longitudinal studies.

Investigating change across time in prevalence or association: the challenges of cross-study comparative research and possible solutions

Cross-study research initiatives to understand change across time are an increasingly prominent component of social and health sciences, yet they present considerable practical, analytical and conceptual challenges. First, we discuss the key challenges to comparative research as a basis for detecting societal change, as well as possible solutions. We focus on studies which investigate changes across time in outcome occurrence or the magnitude and/or direction of associations. We discuss the use and importance of such research, study inclusion, sources of bias and mitigation, and interpretation. Second, we propose a structured framework (a checklist) that is intended to provide guidance for future authors and reviewers. Third, we outline a new open-access teaching resource that offers detailed instruction and reusable analytical syntax to guide newcomers on techniques for conducting comparative analysis and data visualisation (in both R and Stata formats).

Supplementary Information

The online version contains supplementary material available at 10.1007/s44155-022-00021-1.

Prior antibiotics and risk of subsequent Herpes zoster: A population-based case control study

Background

The effect of antibiotics on the human microbiome is now well established, but their indirect effect on the related immune response is less clear. The possible association of Herpes zoster, which involves a reactivation of a previous varicella zoster virus infection, with prior antibiotic exposure might indicate a potential link with the immune response.

Methods

A case-control study was carried out using a clinical database, the UK’s Clinical Practice Research Datalink. A total of 163,754 patients with varicella zoster virus infection and 331,559 age/sex matched controls were identified and their antibiotic exposure over the previous 10 years, and longer when data permitted, was identified. Conditional logistic regression was used to identify the association between antibiotic exposure and subsequent infection in terms of volume and timing.

Results

The study found an association of antibiotic prescription and subsequent risk of varicella zoster virus infection (adjusted odds ratio of 1.50; 95%CIs: 1.42–1.58). The strongest association was with a first antibiotic over 10 years ago (aOR: 1.92; 95%CIs: 1.88–1.96) which was particularly pronounced in the younger age group of 18 to 50 (aOR 2.77; 95%CIs: 1.95–3.92).

Conclusions

By finding an association between prior antibiotics and Herpes zoster this study has shown that antibiotics may be involved in the reactivation of the varicella zoster virus. That effect, moreover, may be relatively long term. This indirect effect of antibiotics on viruses, possibly mediated through their effect on the microbiome and immune system, merits further study.

Oncology Drug Effectiveness from Electronic Health Record Data Calibrated Against RCT Evidence: The PARSIFAL Trial Emulation

Background

The use of electronic health records (EHR) data to assess drug effectiveness in clinical oncology practice is of great interest to regulators, clinicians, and payers. However, the utility of EHR data in clinical effectiveness studies may be limited by missing data, unmeasured confounding, and imperfect outcome surveillance. This study sought to emulate and compare the results of a randomized controlled trial investigating the efficacy of palbociclib with fulvestrant vs letrozole in advanced breast cancer.

Methods

This was a cohort study using longitudinal EHR data derived from outpatient oncology practices in the United States. Eligibility criteria from the PARSIFAL trial were emulated as closely as possible. Patients were included if they had hormone-positive, human epidermal growth factor receptor - 2 (HER-2) negative metastatic breast cancer and had no record of prior treatment for metastatic disease. Patients initiating first-line treatment with palbociclib and fulvestrant following their first record of metastasis were compared to those initiating palbociclib and letrozole on the same day. Treatments were ascertained by oncology medication ordering records in the data source. The primary outcome was death as recorded in the oncologists' EHR systems.

Results

There were 1886 eligible women in the study cohort. Although the 3-year survival was meaningfully lower in clinical practice (59%) compared to the randomized trial (78%), the relative effect size was a hazard ratio (HR) of 1.07 (95% CI: 0.86-1.35), similar to the randomized trial (HR = 1.00; 95% CI: 0.68-1.48).

Conclusion

Despite common challenges encountered in EHR-based studies, it is possible to achieve similar conclusions to emulated randomized trials with the application of analytic approaches that address missing data, confounding, and selection bias. This is a promising finding in light of other emulations and ongoing efforts to improve data from clinical practice and causal analytics.

Machine Learning Approach to Predict In-Hospital Mortality in Patients Admitted for Peripheral Artery Disease in the United States

Background Peripheral artery disease (PAD) affects >10 million people in the United States. PAD is associated with poor outcomes, including premature death. Machine learning (ML) has been increasingly used on big data to predict clinical outcomes. This study aims to develop ML models to predict in-hospital mortality in patients hospitalized for PAD based on a national database. Methods and Results Inpatient hospitalization data were obtained from the 2016 to 2019 National Inpatient Sample. A total of 150 921 inpatients were identified with a primary diagnosis of PAD and PAD-related procedures using codes of the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) and International Classification of Diseases, Tenth Revision, Procedure Coding System (ICD-10-PCS). Four ML models, including logistic regression, random forest, light gradient boosting, and extreme gradient boosting models, were trained to predict the risk of in-hospital death based on a selection of variables, including patient characteristics, comorbidities, procedures, and hospital-related factors. In-hospital mortality occurred in 1.8% of patients. The performance of the 4 models was comparable, with the area under the receiver operating characteristic curve ranging from 0.83 to 0.85, sensitivity of 77% to 82%, and specificity of 72% to 75%. These results suggest adequate predictability for clinical decision-making. In all 4 models, the total number of diagnoses and procedures, age, endovascular revascularization procedure, congestive heart failure, diabetes, and diabetes with complications were critical predictors of in-hospital mortality. Conclusions This study demonstrates the feasibility of ML in predicting in-hospital mortality in patients with a primary PAD diagnosis. Findings highlight the potential of ML models in identifying high-risk patients for poor outcomes and guiding personalized intervention.

Multiple imputation for cause-specific Cox models: Assessing methods for estimation and prediction

In studies analyzing competing time-to-event outcomes, interest often lies in both estimating the effects of baseline covariates on the cause-specific hazards and predicting cumulative incidence functions. When missing values occur in these baseline covariates, they may be discarded as part of a complete-case analysis or multiply imputed. In the latter case, the imputations may be performed either compatibly with a substantive model pre-specified as a cause-specific Cox model [substantive model compatible fully conditional specification (SMC-FCS)], or approximately so [multivariate imputation by chained equations (MICE)]. In a large simulation study, we assessed the performance of these three different methods in terms of estimating cause-specific regression coefficients and predicting cumulative incidence functions. Concerning regression coefficients, results provide further support for use of SMC-FCS over MICE, particularly when covariate effects are large and the baseline hazards of the competing events are substantially different. Complete-case analysis also shows adequate performance in settings where missingness is not outcome dependent. With regard to cumulative incidence prediction, SMC-FCS and MICE are performed more similarly, as also evidenced in the illustrative analysis of competing outcomes following a hematopoietic stem cell transplantation. The findings are discussed alongside recommendations for practising statisticians.

Socioeconomic position, bronchiolitis and asthma in children: counterfactual disparity measures from a national birth cohort study

BACKGROUND

The debated link between severe respiratory syncytial virus (RSV) infection in early life and asthma has yet to be investigated within a social inequity lens. We estimated the magnitude of socioeconomic disparity in childhood asthma which would remain if no child were admitted to hospital for bronchiolitis, commonly due to RSV, during infancy.

METHODS

The cohort, constructed from national administrative health datasets, comprised 83853 children born in Scotland between 1 January 2007 and 31 June 2008. Scottish Index for Multiple Deprivation (SIMD) was used to capture socioeconomic position. Emergency admissions for bronchiolitis before age 1 year were identified from hospital records. Yearly indicators of asthma/wheeze from ages 2 to 9 years were created using dispensing data and hospital admission records.

RESULTS

Using latent class growth analysis, we identified four trajectories of asthma/wheeze: early-transient (2.2% of the cohort), early-persistent (2.0%), intermediate-onset (1.8%) and no asthma/wheeze (94.0%). The estimated marginal risks of chronic asthma (combining early-persistent and intermediate-onset groups) varied by SIMD, with risk differences for the medium and high deprivation groups, relative to the low deprivation group, of 7.0% (95% confidence interval: 3.7-10.3) and 13.0% (9.6-16.4), respectively. Using counterfactual disparity measures, we estimated that the elimination of bronchiolitis requiring hospital admission could reduce these risk differences by 21.2% (4.9-37.5) and 17.9% (10.4-25.4), respectively.

CONCLUSIONS

The majority of disparity in chronic asthma prevalence by deprivation level remains unexplained. Our paper offers a guide to using causal inference methods to study other plausible pathways to inequities in asthma using complex, linked administrative data.

Evaluating the Efficacy of a Social Media-Based Intervention (Warna-Warni Waktu) to Improve Body Image Among Young Indonesian Women: Parallel Randomized Controlled Trial (Preprint)

BACKGROUND

Body dissatisfaction is a global issue, particularly among adolescent girls and young women. Effective body image interventions exist but face barriers to scaling up, particularly in lower- and middle-income countries, such as Indonesia, where a need exists.

OBJECTIVE

We aimed to evaluate the acceptability and efficacy of Warna-Warni Waktu, a social media-based, fictional 6-episode video series with self-guided web-based activities for improving body image among young Indonesian adolescent girls and young women. We hypothesized that Warna-Warni Waktu would increase trait body satisfaction and mood and decrease internalization of appearance ideals and skin shade dissatisfaction relative to the waitlist control condition. We also anticipated improvements in state body satisfaction and mood immediately following each video.

METHODS

We conducted a web-based, 2-arm randomized controlled trial among 2000 adolescent girls and young women, aged 15 to 19 years, recruited via telephone by an Indonesian research agency. Block randomization (1:1 allocation) was performed. Participants and researchers were not concealed from the randomized arm. Participants completed self-report assessments of trait body satisfaction (primary outcome) and the internalization of appearance ideals, mood, and skin shade dissatisfaction at baseline (before randomization), time 2 (1 day after the intervention [T2]), and time 3 (1 month after the intervention [T3]). Participants also completed state body satisfaction and mood measures immediately before and after each video. Data were evaluated using linear mixed models with an intent-to-treat analysis. Intervention adherence was tracked. Acceptability data were collected.

RESULTS

There were 1847 participants. Relative to the control condition (n=923), the intervention group (n=924) showed reduced internalization of appearance ideals at T2 (F_1,1758=40.56, P<.001, partial η²=0.022) and T3 (F_1,1782=54.03, P<.001, partial η²=0.03) and reduced skin shade dissatisfaction at T2 (F_1,1744=8.05, P=.005, partial η²=0.005). Trait body satisfaction improvements occurred in the intervention group at T3 (F_{1, 1781}=9.02, P=.005, partial η²=0.005), which was completely mediated by the internalization change scores between baseline and T2 (indirect effect: β=.03, 95% CI 0.017-0.041; direct effect: β=.03, P=.13), consistent with the Tripartite Influence Model of body dissatisfaction. Trait mood showed no significant effects. Dependent sample t tests (2-tailed) found each video improved state body satisfaction and mood. Cumulative analyses found significant and progressive improvements in pre- and poststate body satisfaction and mood. Intervention adherence was good; participants watched an average of 5.2 (SD 1.66) videos. Acceptability scores were high for understandability, enjoyment, age appropriateness, usefulness, and likelihood to recommend.

CONCLUSIONS

Warna-Warni Waktu is an effective eHealth intervention to reduce body dissatisfaction among Indonesian adolescent girls and young women. Although the effects were small, Warna-Warni Waktu is a scalable, cost-effective alternative to more intense interventions. Initially, dissemination through paid social media advertising will reach thousands of young Indonesian women.

TRIAL REGISTRATION

ClinicalTrials.gov NCT05383807, https://clinicaltrials.gov/ct2/show/NCT05383807 ; ISRCTN Registry ISRCTN35483207, https://www.isrctn.com/ISRCTN35483207.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.2196/33596.

The neighbourhood environment and profiles of the metabolic syndrome

BACKGROUND

There is a dearth of studies on how neighbourhood environmental attributes relate to the metabolic syndrome (MetS) and profiles of MetS components. We examined the associations of interrelated aspects of the neighbourhood environment, including air pollution, with MetS status and profiles of MetS components.

METHODS

We used socio-demographic and MetS-related data from 3681 urban adults who participated in the 3rd wave of the Australian Diabetes, Obesity and Lifestyle Study. Neighbourhood environmental attributes included area socio-economic status (SES), population density, street intersection density, non-commercial land use mix, percentages of commercial land, parkland and blue space. Annual average concentrations of NO₂ and PM_2.5 were estimated using satellite-based land-use regression models. Latent class analysis (LCA) identified homogenous groups (latent classes) of participants based on MetS components data. Participants were then classified into five metabolic profiles according to their MetS-components latent class and MetS status. Generalised additive mixed models were used to estimate relationships of environmental attributes with MetS status and metabolic profiles.

RESULTS

LCA yielded three latent classes, one including only participants without MetS ("Lower probability of MetS components" profile). The other two classes/profiles, consisting of participants with and without MetS, were "Medium-to-high probability of high fasting blood glucose, waist circumference and blood pressure" and "Higher probability of MetS components". Area SES was the only significant predictor of MetS status: participants from high SES areas were less likely to have MetS. Area SES, percentage of commercial land and NO₂ were associated with the odds of membership to healthier metabolic profiles without MetS, while annual average concentration of PM_2.5 was associated with unhealthier metabolic profiles with MetS.

CONCLUSIONS

This study supports the utility of operationalising MetS as a combination of latent classes of MetS components and MetS status in studies of environmental correlates. Higher socio-economic advantage, good access to commercial services and low air pollution levels appear to independently contribute to different facets of metabolic health. Future research needs to consider conducting longitudinal studies using fine-grained environmental measures that more accurately characterise the neighbourhood environment in relation to behaviours or other mechanisms related to MetS and its components.

Factors Contributing to SARS-CoV-2 Vaccine Hesitancy of Hispanic Population in Rio Grande Valley

Hispanic communities have been disproportionately affected by economic disparities. These inequalities have put Hispanics at an increased risk for preventable health conditions. In addition, the CDC reports Hispanics to have 1.5× COVID-19 infection rates and low vaccination rates. This study aims to identify the driving factors for COVID-19 vaccine hesitancy of Hispanic survey participants in the Rio Grande Valley. Our analysis used machine learning methods to identify significant associations between medical, economic, and social factors impacting the uptake and willingness to receive the COVID-19 vaccine. A combination of three classification methods (i.e., logistic regression, decision trees, and support vector machines) was used to classify observations based on the value of the targeted responses received and extract a robust subset of factors. Our analysis revealed different medical, economic, and social associations that correlate to other target population groups (i.e., males and females). According to the analysis performed on males, the Matthews correlation coefficient (MCC) value was 0.972. An MCC score of 0.805 was achieved by analyzing females, while the analysis of males and females achieved 0.797. Specifically, several medical, economic factors, and sociodemographic characteristics are more prevalent in vaccine-hesitant groups, such as asthma, hypertension, mental health problems, financial strain due to COVID-19, gender, lack of health insurance plans, and limited test availability.

Longitudinal Patterns of Material Hardship among US Families

Material hardship has emerged as a direct measure of deprivation in the United States and an important complement to income poverty, providing different evidence about the ways in which deprivation may affect wellbeing. This study addresses gaps in our knowledge about deprivation as the first to examine patterns of material hardship over time. Using data from the Fragile Families and Child Well-Being Study, this study examined five material hardship types (food, housing, medical, utility, and bill-paying) experienced at five timepoints over 15 years. Employing latent class analysis and latent transition analysis, this study identified six longitudinal patterns of material hardship experience, characterized by trajectories of stability or movement and relative severity of material hardship experience over time. These findings improve our conceptual understanding of deprivation and move us towards understanding the impacts of material hardship on wellbeing and identifying policy approaches to prevent deprivation or mitigate negative consequences.

Implications of Recent Epidemiological Studies for Compensation of Veterans Exposed to Plutonium

ABSTRACT

The objective of this paper is to compare post-2007 epidemiological results for plutonium workers to risk predicted by the software program NIOSH-IREP (IREP for short), which is used to determine the lowest dose for a US veteran to obtain cancer compensation. IREP output and methodology were used to predict excess relative risk per Gy (ERR Gy -1 ) for lung cancer at the 99 th credibility percentile, which is used for compensation decisions. Also estimated were relative biological effectiveness factors (RBE) predicted for workers using IREP methodology. IREP predictions were compared to results for Mayak and Sellafield plutonium workers, separately and pooled. Indications that IREP might underpredict 99 th -percentile lung cancer plutonium risk came from (1) comparison of worker RBEs and (2) from comparison of Sellafield results separately. When Sellafield and Mayak data were pooled, ERR Gy -1 comparisons at the 99 th percentile roughly matched epidemiological data with regression dose range restricted to < 0.05 Gy, the most relevant region to veterans, but overpredicted for the full dose range. When four plausible distributions for lung cancer risk, including both new and old data, were combined using illustrative weighting factors, compensation cutoff dose for lung cancer matched current IREP values unless regression results below 0.05 were chosen for Sellafield, producing a two-fold reduction. A 1997 claim of a dose threshold in lung cancer dose response was not confirmed in later literature. The benefit of the doubt is given to claimants when the science is unclear. The challenge for NIOSH-IREP custodians is dealing with the Sellafield results, which might best match US claimants.

Use of tranexamic acid in major trauma: a sex-disaggregated analysis of the Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage (CRASH-2 and CRASH-3) trials and UK trauma registry (Trauma and Audit Research Network) data

BACKGROUND

Women are less likely than men to receive some emergency treatments. This study examines whether the effect of tranexamic acid (TXA) on mortality in trauma patients varies by sex and whether the receipt of TXA by trauma patients varies by sex.

METHODS

First, we conducted a sex-disaggregated analysis of data from the Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage (CRASH)-2 and CRASH-3 trials. We used interaction tests to determine whether the treatment effect varied by sex. Second, we examined data from the Trauma and Audit Research Network (TARN) to explore sex differences in the receipt of TXA. We used logistic regression models to estimate the odds ratio for receipt of TXA in females compared with males. Results are reported as n (%), risk ratios (RR), and odds ratios (OR) with 95% confidence intervals.

RESULTS

Overall, 20 211 polytrauma patients (CRASH-2) and 12 737 patients with traumatic brain injuries (CRASH-3) were included in our analysis. TXA reduced the risk of death in females (RR=0.69 [0.52-0.91]) and in males (RR=0.80 [0.71-0.90]) with no significant heterogeneity by sex (P=0.34). We examined TARN data for 216 364 patients aged ≥16 yr with an Injury Severity Score ≥9 with 98 879 (46%) females and 117 485 (54%) males. TXA was received by 7198 (7.3% [7.1-7.4%]) of the females and 19 697 (16.8% [16.6-17.0%]) of the males (OR=0.39 [0.38-0.40]). The sex difference in the receipt of TXA increased with increasing age.

CONCLUSIONS

Administration of TXA to patients with bleeding trauma reduces mortality to a similar extent in women and men, but women are substantially less likely to be treated with TXA.

Emotional Regulation and Its Influence on the Experience of Stuttering Across the Life Span

PURPOSE

This study evaluated the relationship between emotional regulation (ER) and adverse impact related to stuttering across the developmental spectrum, in preschool and school-age children, adolescents, and adults who stutter. An additional aim examined how these variables relate to the ways that individuals approach speaking (i.e., their agreement on whether their goal is to speak fluently).

METHOD

Participants were the parents of 60 preschoolers and younger school-age children (ages 3-9 years), 95 school-age children and adolescents who stutter (ages 7-18 years), and 180 adults who stutter (ages 18-81 years). All participants completed surveys with age-appropriate measures examining ER and the adverse impact of stuttering. Older children and adults who stutter also answered questions regarding their goals when speaking. Multiple regression and ordinal logistic regression were used to examine relationships among ER, adverse impact related to stuttering, and goal when speaking.

RESULTS

In preschool children, adverse impact was significantly predicted by a parent-reported measure of ER skills; in school-age children and adults, adverse impact was significantly predicted by measures of the ER strategies cognitive reappraisal (CR) and expressive suppression. Less frequent use of CR by adults was significantly associated with an increased likelihood of having "not stuttering" as a goal when speaking. Differences in the significance and magnitude of these relationships were found across the life span.

DISCUSSION

For both children and adults who stutter, ER is a significant factor related to the adverse impact of stuttering; the relationship between ER and adverse impact may change over development. Accounting for individual differences in ER can improve understanding of why a person copes with stuttering in the ways they do, and this has notable implications for individualizing intervention for both children and adults who stutter.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.20044469.

A Patient-specific Approach to Alignment and Proximal Junctional Kyphosis Risk Assessment in Adult Spinal Deformity Surgery: Development and Validation of a Predictive Tool

STUDY DESIGN

This was a single-institution, retrospective cohort study.

OBJECTIVE

We aimed to develop a predictive model for proximal junctional kyphosis (PJK) severity that considers multiple preoperative variables and modifiable surgical alignment.

SUMMARY OF BACKGROUND DATA

PJK is a common complication following adult deformity surgery. Current alignment targets account for age and pelvic incidence but not other risk factors.

MATERIALS AND METHODS

This is a single-institution, retrospective cohort study of adult deformity patients with a minimum 2-year follow-up undergoing instrumented fusion between 2009 and 2018. A proportional odds regression model was fit to estimate PJK probability and Hart-International Spine Study Group (ISSG) PJK severity score. Predictors included preoperative Charlson Comorbidity Index, vertebral Hounsfield Units near the upper instrumented vertebrae, pelvic incidence, T1-pelvic angle, and postoperative L1-L4 and L4-S1 lordosis. Predictor effects were assessed using adjusted odds ratios and a nomogram constructed for estimating PJK probability. Bootstrap resampling was used for internal validation.

RESULTS

Of 145 patients, 47 (32%) developed PJK. The median PJK severity score was 6 (interquartile range, 4-7.5). After adjusting for predictors, Charlson Comorbidity Index, Hounsfield Units, preoperative T1-pelvic angle, and postoperative L1-L4 and L4-S1 lordosis were significantly associated with PJK severity ( P <0.05). After adjusting for potential overfitting, the model showed acceptable discrimination [ C -statistic (area under the curve)=0.75] and accuracy (Brier score=0.10).

CONCLUSIONS

We developed a model to predict PJK probability, adjusted for preoperative alignment, comorbidity burden, vertebral bone density, and modifiable postoperative L1-L4 and L4-S1 lordosis. This approach may help surgeons assess the patient-specific risk of developing PJK and provide a framework for future predictive models assessing PJK risk after adult deformity surgery.

LEVEL OF EVIDENCE

Level III.

Propensity Score and Instrumental Variable Techniques in Observational Transplantation Studies: An Overview and Worked Example Relating to Pre-Transplant Cardiac Screening

Inferring causality from observational studies is difficult due to inherent differences in patient characteristics between treated and untreated groups. The randomised controlled trial is the gold standard study design as the random allocation of individuals to treatment and control arms should result in an equal distribution of known and unknown prognostic factors at baseline. However, it is not always ethically or practically possible to perform such a study in the field of transplantation. Propensity score and instrumental variable techniques have theoretical advantages over conventional multivariable regression methods and are increasingly being used within observational studies to reduce the risk of confounding bias. An understanding of these techniques is required to critically appraise the literature. We provide an overview of propensity score and instrumental variable techniques for transplant clinicians, describing their principles, assumptions, strengths, and weaknesses. We discuss the different patient populations included in analyses and how to interpret results. We illustrate these points using data from the Access to Transplant and Transplant Outcome Measures study examining the association between pre-transplant cardiac screening in kidney transplant recipients and post-transplant cardiac events.

Individual treatment effect estimation in the presence of unobserved confounding using proxies: a cohort study in stage III non-small cell lung cancer

Randomized Controlled Trials (RCT) are the gold standard for estimating treatment effects but some important situations in cancer care require treatment effect estimates from observational data. We developed “Proxy based individual treatment effect modeling in cancer” (PROTECT) to estimate treatment effects from observational data when there are unobserved confounders, but proxy measurements of these confounders exist. We identified an unobserved confounder in observational cancer research: overall fitness. Proxy measurements of overall fitness exist like performance score, but the fitness as observed by the treating physician is unavailable for research. PROTECT reconstructs the distribution of the unobserved confounder based on these proxy measurements to estimate the treatment effect. PROTECT was applied to an observational cohort of 504 stage III non-small cell lung cancer (NSCLC) patients, treated with concurrent chemoradiation or sequential chemoradiation. Whereas conventional confounding adjustment methods seemed to overestimate the treatment effect, PROTECT provided credible treatment effect estimates.

Using adverse childhood experiences to explore the usefulness of community health needs assessments to monitor complex determinants of health at the local level

BACKGROUND

Hospital led community health needs assessments (CHNAs) are intended to help medical care organizations assess determinants of health within the communities they serve. This study demonstrates the utility of data from non-profit hospital CHNAs to monitor complex health issues such as adverse childhood events (ACEs) at the local-level.

METHODS

CHNA data were collected from August to November 2019 and analyzed July 2021. A series of logistic regressions were used to analyze associations between ACEs, mental health conditions, and self-rated health from a convenience sample of 2831 adults from two regional hospitals that service five counties located in central Appalachia.

RESULTS

ACEs were associated with increased odds of experiencing all metal health conditions after adjusting for other exposures and demographics, including: bipolar disorder (AOR: 2.42, CL: 1.78, 3.30), chronic pain (AOR: 1.61, CL: 1.438, 1.87), depression (AOR: 2.05, CL: 1.76, 2.36), PTSD (AOR: 3.83, CL: 2.95, 4.98), and poor self-rated health (AOR: 1.88, CL: 1.65, 2.15).

CONCLUSION

Findings suggest hospital CHNAs are a useful way to assess local data and should include factors known to antecede disease including associated risks and outcomes. CHNAs may provide an opportunity to fill important gaps in community surveillance and inform local prevention and treatment strategies.

Handling Missing Values in Information Systems Research: A Review of Methods and Assumptions

Data have never been more essential to the success of decision making. However, data are often messy. A perennial data challenge is missing values, which frequently occur in real-world data, such as unreported data items in public firms’ financial statements and skipped product ratings from consumers. What is the influence of missing values and how should they be handled? Although we are in a big data era, missing values are not ignorable if data are missing for nonrandom reasons. In the case of product ratings, if only people who favor the product provide ratings while others put aside the product and do not respond, then even a simple mean estimation of the product rating would be significantly biased. Such bias challenges the validity of data analysis, and it cannot be eliminated simply by increasing the sample size of the data. To correct the bias arising from nonrandom missing values, it is necessary to examine and model what causes the missing values. We propose and demonstrate the superior performance of a Monte Carlo likelihood approach to correct the bias. Overall, we recommend well-designed data collection processes with documentation of the possible reasons for missing values, cautious adoption of missing value handling methods, and structured missing value reporting practices.

Illegal Drug Use, Depressive Symptoms, and General Health: Exploring Co-occurrence across 11 Years in a National Sample

The co-occurrence of illegal drug use, symptoms of depression, and a lower perception of general health among adolescents continues to be of substantive interest for researchers and the general public alike. Research on this topic, however, remains relatively stagnant, focusing on narrow developmental periods and each association independently, with limited consideration for the existence of a nexus between the three constructs as individuals age. Considering these limitations, the current study examines the longitudinal progression, from adolescence to early adulthood, of illegal drug use, symptoms of depression, and a lower perception of general health. The National Longitudinal Survey of Youth 1997 (NLSY97; N = 8,984), measures over an eleven-year data collection period, and between-and within-individual analytical strategies were used to evaluate the nexus between the constructs. The findings suggested that illegal drug use, depressive symptoms, and general health at previous time periods directly and indirectly predicted illegal drug use, depressive symptoms, and general health at subsequent time periods. Moreover, the within-individual change in illegal drug use was associated with the change in depressive symptoms, and the change in depressive symptoms was associated with the change in general health. Practitioners should consider this co-occurrence when treating symptoms related to illegal drug use, symptoms of depression, and physical health.

An Empirical Comparison of the Sales Forecasting Performance for Plastic Tray Manufacturing Using Missing Data

The problem of missing data is frequently met in time series analysis. If not appropriately addressed, it usually leads to failed modeling and distorted forecasting. To deal with high market uncertainty, companies need a reliable and sustainable forecasting mechanism. In this article, two propositions are presented: (1) a dedicated time series forecasting scheme, which is both accurate and sustainable, and (2) a practical observation of the data background to deal with the problem of missing data and to effectively formulate correction strategies after predictions. In the empirical study, actual tray sales data and a comparison of different models that combine missing data processing methods and forecasters are employed. The results show that a specific product needs to be represented by a dedicated model. For example, regardless of whether the last fiscal year was a growth or recession year, the results suggest that the missing data for products with a high market share should be handled by the zero-filling method, whereas the mean imputation method should be for the average market share products. Finally, the gap between forecast and actual demand is bridged by employing a validation set, and it is further used for formulating correction strategies regarding production volumes.

ARHGAP25 expression in colorectal cancer as a biomarker associated with favorable prognosis

Although progress has been made in the early diagnosis of colorectal cancer (CRC) and in the systemic therapy of patients with CRC, the prognosis for advanced CRC remains poor. Our previous study demonstrated that ARHGAP25 overexpression significantly inhibits CRC cell growth, invasion and migration. However, it was not possible to evaluate and analyze the overall survival (OS) rate of patients with CRC. Thus, the discovery of relevant factors and their expression on the basis of existing research is necessary to predict the OS rate of patients with advanced CRC. Therefore, the aim of the present study was to define the value of Rho GTPase-activating protein 25 (ARHGAP25) expression in predicting the OS rate in patients with CRC. The clinical data of 153 patients with CRC who underwent colorectal resection were retrospectively analyzed. In order to explore the expression of ARHGAP25, immunohistochemical analysis of the tumor tissues of these patients, was performed. Univariate Cox regression analysis was used to assess the prognostic value of ARHGAP25 expression for OS. Multivariate analysis was used to evaluate the effect of ARHGAP25 expression in the presence of other variables. Confounding factors and interaction were assessed by a stratified analysis using ARHGAP25 expression and other variables associated with survival. The univariate analysis revealed that, ARHGAP25 expression was associated with an improved OS in patients with CRC (P<0.05). The multivariate analysis revealed that ARHGAP25 expression was still correlated with an improved OS after adjusting for sex, age, invasion degree, lymph node metastasis, distant metastasis, TNM stage, tumor location, histological type, histological grade, tumor deposits, and postoperative treatment (P<0.05). The stratified analysis demonstrated that the predictive value of ARHGAP25 for the OS of patients with CRC was stronger in males, elderly patients (>70 years old), patients with T3 stage tumor, lymph node metastasis, TNM stage III, right hemicolon location and patients with a poorly differentiated tumor (P<0.05). Overall, our results demonstrated that ARHGAP25 may have an important potential value for improving the prognosis of patients with CRC.

Factors Affecting Young Adults' Decision Making to Undergo COVID-19 Vaccination: A Patient Preference Study

Young adults are a substantial driver of lagging vaccination against COVID-19 worldwide. We aimed to understand what vaccine or vaccination environment attributes may affect young adults' vaccine inclination. We contacted a convenience sample of 1415 students to recruit a minimum of 150 individuals for a web-based discrete choice experiment. The respondents were asked to choose one of two hypothetical vaccines, defined by six attributes-vaccine efficacy, risk of mild side effects, protection duration, administration route, recommender, and travel time to the vaccination site. Individual preferences were calculated with the Markov chain Monte Carlo hierarchical Bayes estimation. A total of 445 individuals (mean age 24.4 years, 272 (61.1%) women) completed the survey between 22 March and 3 May 2021. Vaccine protection duration (28.3 (95% CI, 27.0-29.6)) and vaccine efficacy in preventing COVID-19 (27.5 (95% CI, 26.3-28.8)) were the most important, followed by the risk of vaccine side effects (17.3 (95% CI, 16.2-18.4)). Individuals reluctant or unsure about vaccination (21.1%) prioritized the potential for mild side effects higher and vaccine efficacy lower than the vaccine-inclined individuals. New vaccination programs that target young adults should emphasize the protection duration, low risk of vaccine side effects, and high efficacy.

Sensitivity to missing not at random dropout in clinical trials: Use and interpretation of the trimmed means estimator

Outcome values in randomized controlled trials (RCTs) may be missing not at random (MNAR), if patients with extreme outcome values are more likely to drop out (eg, due to perceived ineffectiveness of treatment, or adverse effects). In such scenarios, estimates from complete case analysis (CCA) and multiple imputation (MI) will be biased. We investigate the use of the trimmed means (TM) estimator for the case of univariable missingness in one continuous outcome. The TM estimator operates by setting missing values to the most extreme value, and then “trimming” away equal fractions of both groups, estimating the treatment effect using the remaining data. The TM estimator relies on two assumptions, which we term the “strong MNAR” and “location shift” assumptions. We derive formulae for the TM estimator bias resulting from the violation of these assumptions for normally distributed outcomes. We propose an adjusted TM estimator, which relaxes the location shift assumption and detail how our bias formulae can be used to establish the direction of bias of CCA and TM estimates, to inform sensitivity analyses. The TM approach is illustrated in a sensitivity analysis of the CoBalT RCT of cognitive behavioral therapy (CBT) in 469 individuals with 46 months follow‐up. Results were consistent with a beneficial CBT treatment effect, with MI estimates closer to the null and TM estimates further from the null than the CCA estimate. We propose using the TM estimator as a sensitivity analysis for data where extreme outcome value dropout is plausible.

Availability of Multiple Organ Failure Score Components in Surgical Patients

Background: Scoring systems are often used describe the degree of multi-system organ failure (MOF), however, the data used to calculate these scores are often missing. Studies utilizing these scoring systems often underreport the frequency of missing data. No study has examined the availability of clinical data needed to calculate Sequential Organ Failure Assessment (SOFA), and other organ failure scores. The primary objective of this study is to observe how often emergency general surgery and trauma patients have missing data needed to calculate MOF scores. Patients and Methods: Patients admitted between June 2017 and September 2019 were evaluated. Data to calculate SOFA, quick SOFA (qSOFA), Marshall Multiple Organ Dysfunction Score (MODS), Denver Post-Injury Multiple Organ Failure, and systemic inflammatory response syndrome (SIRS) criteria, as well as demographic and general admission and discharge data, were collected. Results: Of the 238 patients included in this study, 66.4% were emergency general surgery and 33.6% were trauma patients. For all patients, the median intensive care unit (ICU) length of stay (LOS) was seven days (range, 4-12), the median hospital LOS was 14 days (range, 10-21), and 28 patients (11.8%) did not survive to hospital discharge. Sequential Organ Failure Assessment was calculable in 21.4%-18.1%, whereas MODS was calculable in 6.3%-5.0% on days three and five, respectively. The Denver score was calculable in 32.5%-28.8% of trauma patients on these days. Of the data points needed to calculate these scores, the partial pressure of oxygen (Pao₂)/fraction of inspired oxygen (FIo₂) ratio, central venous pressure (CVP), and bilirubin were the least available components. Conclusions: Data needed to fully calculate SOFA and other common MOF scores are often not readily available highlighting the degree of imputation required to calculate these scores. We recommend better reporting of the degree of missing data in the literature.

Predictive Factors for and Complications of Bronchiectasis in Common Variable Immunodeficiency Disorders

Bronchiectasis is a frequent complication of common variable immunodeficiency disorders (CVID). In a cohort of patients with CVID, we sought to identify predictors of bronchiectasis. Secondly, we sought to describe the impact of bronchiectasis on lung function, infection risk, and quality of life. We conducted an observational cohort study of 110 patients with CVID and an available pulmonary computed tomography scan. The prevalence of bronchiectasis was 53%, with most of these patients (54%) having mild disease. Patients with bronchiectasis had lower median serum immunoglobulin (Ig) concentrations, especially long-term IgM (0 vs 0.25 g/l; p < 0.01) and pre-treatment IgG (1.3 vs 3.7 g/l; p < 0.01). CVID patients with bronchiectasis had worse forced expiratory volume in one second (2.10 vs 2.99 l; p < 0.01) and an annual decline in forced expiratory volume in one second of 25 ml/year (vs 8 ml/year in patients without bronchiectasis; p = 0.01). Patients with bronchiectasis also reported more annual respiratory tract infections (1.77 vs 1.25 infections/year, p = 0.04) and a poorer quality of life (26 vs 14 points in the St George's Respiratory Questionnaire; p = 0.02). Low serum immunoglobulin M concentration identifies patients at risk for bronchiectasis in CVID and may play a role in pathogenesis. Bronchiectasis is relevant because it is associated with frequent respiratory tract infections, poorer lung function, a greater rate of lung function decline, and a lower quality of life.

Mortality associated with nonrestorative short sleep or nonrestorative long time-in-bed in middle-aged and older adults

Associations of sleep duration with human health could differ depending on whether sleep is restorative. Using data from 5804 participants of the Sleep Heart Health Study, we examined the longitudinal association of sleep restfulness combined with polysomnography-measured total sleep time (TST) or time in bed (TIB), representing different sleeping behaviors, with all-cause mortality. Among middle-aged adults, compared with restful intermediate TST quartile, the lowest TST quartile with feeling unrested was associated with higher mortality (hazard ratio [HR], 1.54; 95% confidence interval [CI] 1.01–2.33); the highest TST quartile with feeling rested was associated with lower mortality (HR, 0.55; 95% CI 0.32–0.97). Among older adults, the highest TIB quartile with feeling unrested was associated with higher mortality, compared with restful intermediate TIB quartile (HR, 1.57; 95% CI 1.23–2.01). Results suggest a role of restorative sleep in differentiating the effects of sleep duration on health outcomes in midlife and beyond.

Supervised machine learning in the mass spectrometry laboratory: A tutorial

As the demand for laboratory testing by mass spectrometry increases, so does the need for automated methods for data analysis. Clinical mass spectrometry (MS) data is particularly well-suited for machine learning (ML) methods, which deal nicely with structured and discrete data elements. The alignment of these two fields offers a promising synergy that can be used to optimize workflows, improve result quality, and enhance our understanding of high-dimensional datasets and their inherent relationship with disease. In recent years, there has been an increasing number of publications that examine the capabilities of ML-based software in the context of chromatography and MS. However, given the historically distant nature between the fields of clinical chemistry and computer science, there is an opportunity to improve technological literacy of ML-based software within the clinical laboratory scientist community. To this end, we present a basic overview of ML and a tutorial of an ML-based experiment using a previously published MS dataset. The purpose of this paper is to describe the fundamental principles of supervised ML, outline the steps that are classically involved in an ML-based experiment, and discuss the purpose of good ML practice in the context of a binary MS classification problem.

Overall and cause-specific hospitalisation and death after COVID-19 hospitalisation in England: A cohort study using linked primary care, secondary care, and death registration data in the OpenSAFELY platform

BACKGROUND

There is concern about medium to long-term adverse outcomes following acute Coronavirus Disease 2019 (COVID-19), but little relevant evidence exists. We aimed to investigate whether risks of hospital admission and death, overall and by specific cause, are raised following discharge from a COVID-19 hospitalisation.

METHODS AND FINDINGS

With the approval of NHS-England, we conducted a cohort study, using linked primary care and hospital data in OpenSAFELY to compare risks of hospital admission and death, overall and by specific cause, between people discharged from COVID-19 hospitalisation (February to December 2020) and surviving at least 1 week, and (i) demographically matched controls from the 2019 general population; and (ii) people discharged from influenza hospitalisation in 2017 to 2019. We used Cox regression adjusted for age, sex, ethnicity, obesity, smoking status, deprivation, and comorbidities considered potential risk factors for severe COVID-19 outcomes. We included 24,673 postdischarge COVID-19 patients, 123,362 general population controls, and 16,058 influenza controls, followed for ≤315 days. COVID-19 patients had median age of 66 years, 13,733 (56%) were male, and 19,061 (77%) were of white ethnicity. Overall risk of hospitalisation or death (30,968 events) was higher in the COVID-19 group than general population controls (fully adjusted hazard ratio [aHR] 2.22, 2.14 to 2.30, p < 0.001) but slightly lower than the influenza group (aHR 0.95, 0.91 to 0.98, p = 0.004). All-cause mortality (7,439 events) was highest in the COVID-19 group (aHR 4.82, 4.48 to 5.19 versus general population controls [p < 0.001] and 1.74, 1.61 to 1.88 versus influenza controls [p < 0.001]). Risks for cause-specific outcomes were higher in COVID-19 survivors than in general population controls and largely similar or lower in COVID-19 compared with influenza patients. However, COVID-19 patients were more likely than influenza patients to be readmitted or die due to their initial infection or other lower respiratory tract infection (aHR 1.37, 1.22 to 1.54, p < 0.001) and to experience mental health or cognitive-related admission or death (aHR 1.37, 1.02 to 1.84, p = 0.039); in particular, COVID-19 survivors with preexisting dementia had higher risk of dementia hospitalisation or death (age- and sex-adjusted HR 2.47, 1.37 to 4.44, p = 0.002). Limitations of our study were that reasons for hospitalisation or death may have been misclassified in some cases due to inconsistent use of codes, and we did not have data to distinguish COVID-19 variants.

CONCLUSIONS

In this study, we observed that people discharged from a COVID-19 hospital admission had markedly higher risks for rehospitalisation and death than the general population, suggesting a substantial extra burden on healthcare. Most risks were similar to those observed after influenza hospitalisations, but COVID-19 patients had higher risks of all-cause mortality, readmission or death due to the initial infection, and dementia death, highlighting the importance of postdischarge monitoring.

Factors associated with excess all-cause mortality in the first wave of the COVID-19 pandemic in the UK: A time series analysis using the Clinical Practice Research Datalink

BACKGROUND

Excess mortality captures the total effect of the Coronavirus Disease 2019 (COVID-19) pandemic on mortality and is not affected by misspecification of cause of death. We aimed to describe how health and demographic factors were associated with excess mortality during, compared to before, the pandemic.

METHODS AND FINDINGS

We analysed a time series dataset including 9,635,613 adults (≥40 years old) registered at United Kingdom general practices contributing to the Clinical Practice Research Datalink. We extracted weekly numbers of deaths and numbers at risk between March 2015 and July 2020, stratified by individual-level factors. Excess mortality during Wave 1 of the UK pandemic (5 March to 27 May 2020) compared to the prepandemic period was estimated using seasonally adjusted negative binomial regression models. Relative rates (RRs) of death for a range of factors were estimated before and during Wave 1 by including interaction terms. We found that all-cause mortality increased by 43% (95% CI 40% to 47%) during Wave 1 compared with prepandemic. Changes to the RR of death associated with most sociodemographic and clinical characteristics were small during Wave 1 compared with prepandemic. However, the mortality RR associated with dementia markedly increased (RR for dementia versus no dementia prepandemic: 3.5, 95% CI 3.4 to 3.5; RR during Wave 1: 5.1, 4.9 to 5.3); a similar pattern was seen for learning disabilities (RR prepandemic: 3.6, 3.4 to 3.5; during Wave 1: 4.8, 4.4 to 5.3), for black or South Asian ethnicity compared to white, and for London compared to other regions. Relative risks for morbidities were stable in multiple sensitivity analyses. However, a limitation of the study is that we cannot assume that the risks observed during Wave 1 would apply to other waves due to changes in population behaviour, virus transmission, and risk perception.

CONCLUSIONS

The first wave of the UK COVID-19 pandemic appeared to amplify baseline mortality risk to approximately the same relative degree for most population subgroups. However, disproportionate increases in mortality were seen for those with dementia, learning disabilities, non-white ethnicity, or living in London.

PACAP38- and VIP-induced cluster headache attacks are not associated with changes of plasma CGRP or markers of mast cell activation

BACKGROUND

Pituitary adenylate cyclase-activating polypeptide-38 (PACAP38) and vasoactive intestinal polypeptide can provoke cluster headache attacks in up to half of cluster headache patients in their active phase. At present, it is unknown whether provoked attacks are mediated via calcitonin gene-related peptide or mast cell activation.

METHODS

All enrolled patients with cluster headache were randomly allocated to receive a continuous infusion of either PACAP38 (10 pmol/kg/min) or vasoactive intestinal polypeptide (8 pmol/kg/min) over 20 min. We collected clinical data and measured plasma levels of calcitonin gene-related peptide and markers of mast cell activation (tryptase and histamine) at fixed time points: at baseline (T₀), at the end of the infusion (T₂₀), 10 min after the infusion (T₃₀), and 70 min after the infusion (T₉₀).

RESULTS

Blood was collected from episodic cluster headache patients in active phase (n = 14), episodic cluster headache patients in remission (n = 15), and chronic cluster headache patients (n = 15). At baseline, plasma levels of calcitonin gene-related peptide, tryptase, and histamine were not different among the three study groups. Plasma levels of calcitonin gene-related peptide (p = 0.7074), tryptase (p = 0.6673), or histamine (p = 0.4792) remained unchanged during provoked attacks compared to attack-free patients.

CONCLUSION

Cluster headache attacks provoked by either PACAP38 or vasoactive intestinal polypeptide were not accompanied by alterations of plasma calcitonin gene-related peptide, tryptase or histamine. The provoked attacks may not be mediated by calcitonin gene-related peptide or mast cell activation.Trial Registration: The study is registered at ClinicalTrials.gov (NCT03814226).

Optimal Methods for Reducing Proxy-Introduced Bias on Patient-Reported Outcome Measurements for Group-Level Analyses

BACKGROUND

Caregivers, or proxies, often complete patient-reported outcomes (PROs) on behalf of patients; yet, research has demonstrated proxies rate patient outcomes worse than patients rate their own outcomes. To improve interpretability of PROs in group-level analyses, our study aimed to identify optimal approaches for reducing proxy-introduced bias in the analysis of PROs.

METHODS

Data were simulated based on 200 patients with stroke and their proxies who both completed 9 PROMIS domains as part of a cross-sectional study. The sample size was varied as 50, 100, 200, and 500, and the proportion of patients with proxy-respondents was varied as 10%, 20%, and 50%. Six methods for handling proxy-completions were investigated: (1) complete case analysis; (2) proxy substitution; (3) Method 2 plus proxy adjustment; (4) Method 3 including inverse-probability of treatment weighting; (5) multiple imputation; (6) linear equating. These methods were evaluated by comparing average bias in PROMIS T-scores (estimated versus observed patient-only responses), as well as by comparing estimated regression coefficients to models using patient-only responses.

RESULTS

Overall mean T-score differences ranged from 0 to 1.75. The range of mean differences varied by the 6 methods with methods 1 and 5 providing estimates closest to the observed mean. In regression models, all but inverse-probability of treatment weighting resulted in low bias when proxy-completions were 10% to 20%. With 50% proxy-completions, method 5 resulted in less accurate estimations while methods 1 to 3 provided less proxy-introduced bias. Bias remained low across domain and varying sample sizes but increased with larger percentages of proxy-respondents.

CONCLUSIONS

Our study found modest proxy-introduced bias when estimating PRO scores or regression estimates across multiple domains of health. This bias remained low, even when sample size was 50 and there were large proportions of proxy-completions. While many of these methods can be chosen for including proxies in stroke PRO research with <20% proxy-respondents, proxy substitution with adjustment resulted in low bias with 50% proxy-respondents.

Antibiotics for lower respiratory tract infection in children presenting in primary care in England (ARTIC PC): a double-blind, randomised, placebo-controlled trial

BACKGROUND

Antibiotic resistance is a global public health threat. Antibiotics are very commonly prescribed for children presenting with uncomplicated lower respiratory tract infections (LRTIs), but there is little evidence from randomised controlled trials of the effectiveness of antibiotics, both overall or among key clinical subgroups. In ARTIC PC, we assessed whether amoxicillin reduces the duration of moderately bad symptoms in children presenting with uncomplicated (non-pneumonic) LRTI in primary care, overall and in key clinical subgroups.

METHODS

ARTIC PC was a double-blind, randomised, placebo-controlled trial done at 56 general practices in England. Eligible children were those aged 6 months to 12 years presenting in primary care with acute uncomplicated LRTI judged to be infective in origin, where pneumonia was not suspected clinically, with symptoms for less than 21 days. Patients were randomly assigned in a 1:1 ratio to receive amoxicillin 50 mg/kg per day or placebo oral suspension, in three divided doses orally for 7 days. Patients and investigators were masked to treatment assignment. The primary outcome was the duration of symptoms rated moderately bad or worse (measured using a validated diary) for up to 28 days or until symptoms resolved. The primary outcome and safety were assessed in the intention-to-treat population. The trial is registered with the ISRCTN Registry (ISRCTN79914298).

FINDINGS

Between Nov 9, 2016, and March 17, 2020, 432 children (not including six who withdrew permission for use of their data after randomisation) were randomly assigned to the antibiotics group (n=221) or the placebo group (n=211). Complete data for symptom duration were available for 317 (73%) patients; missing data were imputed for the primary analysis. Median durations of moderately bad or worse symptoms were similar between the groups (5 days [IQR 4-11] in the antibiotics group vs 6 days [4-15] in the placebo group; hazard ratio [HR] 1·13 [95% CI 0·90-1·42]). No differences were seen for the primary outcome between the treatment groups in the five prespecified clinical subgroups (patients with chest signs, fever, physician rating of unwell, sputum or chest rattle, and short of breath). Estimates from complete-case analysis and a per-protocol analysis were similar to the imputed data analysis.

INTERPRETATION

Amoxicillin for uncomplicated chest infections in children is unlikely to be clinically effective either overall or for key subgroups in whom antibiotics are commonly prescribed. Unless pneumonia is suspected, clinicians should provide safety-netting advice but not prescribe antibiotics for most children presenting with chest infections.

FUNDING

National Institute for Health Research.

Analyzing evidence-based falls prevention data with significant missing information using variable selection after multiple imputation

Falls are the leading cause of fatal and non-fatal injuries among older adults. Evidence-based fall prevention programs are delivered nationwide, largely supported by funding from the Administration for Community Living (ACL), to mitigate fall-related risk. This study utilizes data from 39 ACL grantees in 22 states from 2014 to 2017. The large amount of missing values for falls efficacy in this national database may lead to potentially biased statistical results and make it challenging to implement reliable variable selection. Multiple imputation is used to deal with missing values. To obtain a consistent result of variable selection in multiply-imputed datasets, multiple imputation-stepwise regression (MI-stepwise) and multiple imputation-least absolute shrinkage and selection operator (MI-LASSO) methods are used. To compare the performances of MI-stepwise and MI-LASSO, simulation studies were conducted. In particular, we extended prior work by considering several circumstances not covered in previous studies, including an extensive investigation of data with different signal-to-noise ratios and various missing data patterns across predictors, as well as a data structure that allowed the missingness mechanism to be missing not at random (MNAR). In addition, we evaluated the performance of MI-LASSO method with varying tuning parameters to address the overselection issue in cross-validation (CV)-based LASSO.

Clinical Predictors for Analgesic Response to Radiotherapy in Patients with Painful Bone Metastases

BACKGROUND

Radiotherapy (RT) reduces pain in about 60% of patients with painful bone metastases, leaving many patients without clinical benefit. This study assesses predictors for RT effectiveness in patients with painful bone metastases.

MATERIALS AND METHODS

We included adult patients receiving RT for painful bone metastases in a multicenter, multinational longitudinal observational study. Pain response within 8 weeks was defined as ≥2-point decrease on a 0-10 pain score scale, without increase in analgesics; or a decrease in analgesics of ≥25% without increase in pain score. Potential predictors were related to patient demographics, RT administration, pain characteristics, tumor characteristics, depression and inflammation (C-reactive protein [CRP]). Multivariate logistic regression analysis with multiple imputation of missing data were applied to identify predictors of RT response.

RESULTS

Of 513 eligible patients, 460 patients (90 %) were included in the regression model. 224 patients (44%, 95% confidence interval (CI) 39%-48%) responded to RT. Better Karnofsky performance status (Odds ratio (OR) 1.39, CI 1.15-1.68), breast cancer (OR 2.54, CI 1.12-5.73), prostate cancer (OR 2.83, CI 1.27-6.33) and soft tissue expansion (OR 2.00, CI 1.23-3.25) predicted RT response. Corticosteroids were a negative predictor (OR 0.57, CI 0.37-0.88). Single and multiple fraction RT had similar response. The discriminative ability of the model was moderate; C-statistic 0.69.

CONCLUSION

This study supports previous findings that better performance status and type of cancer diagnosis predicts analgesic RT response, and new data showing that soft tissue expansion predicts RT response and that corticosteroids is a negative predictor for RT response in patients with painful bone metastases.

Investigating the inequalities in route to diagnosis amongst patients with diffuse large B-cell or follicular lymphoma in England

INTRODUCTION

Diagnostic delay is associated with lower chances of cancer survival. Underlying comorbidities are known to affect the timely diagnosis of cancer. Diffuse large B-cell (DLBCL) and follicular lymphomas (FL) are primarily diagnosed amongst older patients, who are more likely to have comorbidities. Characteristics of clinical commissioning groups (CCG) are also known to impact diagnostic delay. We assess the association between comorbidities and diagnostic delay amongst patients with DLBCL or FL in England during 2005-2013.

METHODS

Multivariable generalised linear mixed-effect models were used to assess the main association. Empirical Bayes estimates of the random effects were used to explore between-cluster variation. The latent normal joint modelling multiple imputation approach was used to account for partially observed variables.

RESULTS

We included 30,078 and 15,551 patients diagnosed with DLBCL or FL, respectively. Amongst patients from the same CCG, having multimorbidity was strongly associated with the emergency route to diagnosis (DLBCL: odds ratio 1.56, CI 1.40-1.73; FL: odds ratio 1.80, CI 1.45-2.23). Amongst DLBCL patients, the diagnostic delay was possibly correlated with CCGs that had higher population densities.

CONCLUSIONS

Underlying comorbidity is associated with diagnostic delay amongst patients with DLBCL or FL. Results suggest a possible correlation between CCGs with higher population densities and diagnostic delay of aggressive lymphomas.

Association of Atopic Dermatitis and Mental Health Outcomes Across Childhood: A Longitudinal Cohort Study

IMPORTANCE

Research has highlighted associations between atopic dermatitis (AD) and mental health conditions in adults. However, literature on the development of mental health comorbidities in children is limited despite the large burden of pediatric AD worldwide.

OBJECTIVE

To examine the association between AD and internalizing behaviors and symptoms of depression at multiple points across childhood and adolescence and to explore potential mediating factors, including asthma/rhinitis, sleep, and inflammation.

DESIGN, SETTING, AND PARTICIPANTS

This longitudinal, population-based birth cohort study included children followed up from birth for a mean (SD) duration of 10.0 (2.9) years from the UK Avon Longitudinal Study of Parents and Children. Data were collected from September 6, 1990, to December 31, 2009. Data were analyzed from August 30, 2019, to July 30, 2020.

EXPOSURES

Annual period prevalence of AD assessed at 11 points from 6 months to 18 years of age, measured by standardized questions about flexural dermatitis.

MAIN OUTCOMES AND MEASURES

Symptoms of depression, measured using child-reported responses to the Short Moods and Feelings Questionnaire at 5 points from 10 to 18 years of age and internalizing behaviors, measured by maternal report of the Emotional Symptoms subscale of the Strength and Difficulties Questionnaire at 7 points from 4 to 16 years of age.

RESULTS

Among the 11 181 children included in the analysis (5721 male [51.2%]), the period prevalence of symptoms of depression ranged from 6.0% to 21.6%; for internalizing behaviors, from 10.4% to 16.0%. Although mild to moderate AD was not associated with symptoms of depression, it was associated with internalizing behaviors as early as 4 years of age (mean increased odds of 29%-84% across childhood in adjusted models). Severe AD was associated with symptoms of depression (adjusted odds ratio, 2.38; 95% CI, 1.21-4.72) and internalizing symptoms (adjusted odds ratio, 1.90; 95% CI, 1.14-3.16). Sleep quality mediated some of this association, but it was not explained by differences in sleep duration, asthma/rhinitis, or levels of inflammatory markers (interleukin 6 and C-reactive protein).

CONCLUSIONS AND RELEVANCE

Within this population-based birth cohort study in the UK, severe AD was associated with symptoms of depression and internalizing behaviors throughout childhood and adolescence. Risk of internalizing symptoms was increased even for children with mild AD beginning early in childhood, highlighting the importance of behavioral and mental health awareness in this population.