Bayesian estimation of potential outcomes for mediation analysis of racial disparity for infant mortality

Background

There is a need for novel methods to determine preventable causes of racial health disparities. This need has been met with the development of improved methods for mediation modeling. Current mediational analysis methods call for an evaluation of statistical interaction or effect modification between the investigated cause and mediator. For racial disparity, this approach facilitates the estimation of racially specific risks for infant mortality. However, current methods for evaluating multiple interacting mediators are inadequate. The first objective of the study was to compare Bayesian estimation of potential outcomes to other approaches to mediation analysis that included interaction. The second objective was to evaluate three potentially interacting mediators of racial disparity for infant mortality by modeling the large dataset from the National Natality Database using Bayesian estimation of potential outcomes.

Methods

A random sample of observations from the 2003 National Natality Database was used to compare the currently promoted methods for mediation modeling. Racial disparity was modeled as a separate function for each of three potential mediators, (i) maternal smoking, (ii) low birth weight and (iii) teenage maternity. As a second objective, direct Bayesian estimation of potential outcomes modeled infant mortality as function of the interactions among the three mediators and race using the full National Natality Database for the years 2016 to 2018.

Results

The counterfactual model was inaccurate in estimating the proportion of racial disparity that was attributable to either maternal smoking or teenage maternity. The counterfactual approach did not accurately estimate the probabilities defined by counterfactual definitions. The error was a result of modeling the excess relative risk instead of the risk probabilities. Bayesian approaches did estimate the probabilities of the counterfactual definitions. Results showed that 73% of the racial disparity for infant mortality was attributed to infants born with low birth weight.

Conclusions

Bayesian estimation of potential outcomes could evaluate whether proposed public health programs would affect races differently and decisions could include consideration of the causal effect the program may have on racial disparity. The large contribution of low birth weight to racial disparity for infant mortality should be further investigated to identify preventable factors for low birth weight.

Examining the Relationships Among Treatment, Pain, and Physical Function in Patients With Osteoarthritis: A Mediation-Modeling Approach

OBJECTIVES

To better understand the relationships among treatment, pain, and physical function (PF).

METHODS

Data were collected from 2 published randomized clinical trials of osteoarthritis patients who received tanezumab or a placebo. PF was measured by the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) PF domain. Pain (WOMAC pain domain) was a mediator of the effect of treatment on PF. A set of mediation models were investigated. Variables were treatment (tanezumab vs placebo), WOMAC pain domain, and WOMAC PF domain. Cross-sectional mediation models were assessed separately at different weeks. Longitudinal mediation models used data from all weeks simultaneously. Results could identify a steady-state period.

RESULTS

The cross-sectional and longitudinal mediation models showed a stable indirect effect of treatment through the pain on PF across time, indicating that a pseudo-steady-state model was appropriate. Therefore, the longitudinal steady-state mediation models were used with all available data assuming relationships among variables in the model being the same at all time points; results showed that the indirect effect of the treatment on PF was 77.8% in study 1 (NCT02697773) and 74.1% in study 2 (NCT02709486), both P <0.0001, whereas the direct effect was 22.2% for study 1 ( P = 0.0003) and 25.9% for study 2 ( P = 0.0019).

DISCUSSION

At least 75% of the treatment effect of tanezumab on physical functioning can be explained by the improvements in pain. However, tanezumab had an additional effect on physical functioning (~25%) that, was independent of improvements in pain. Such independent effects are of considerable interest and require further research to determine their mechanisms.

The mediating role of Courageous coping in the relations between spirituality and social support with resilience among adolescents with hemophilia

In addition to physical disorders, people with hemophilia face complex psychological problems. Hence, protective factors such as coping strategies, social support, and spirituality to improve psychological health should be determined. This study examined the role of courageous coping as a mediator in the interrelationships between spirituality as well as social support and resilience among adolescents with hemophilia. In this descriptive-analytical study, the participants were 372 adolescents with hemophilia aged 11-21 years. Connor-Davidson resilience scale, Jalowiec Coping scale, Perceived Social Support Scale, and spirituality scale were the instruments used in this study. The analysis results of Pearson correlation revealed a significant positive relationship between spirituality as well as social support and resilience. Also, it was determined that courageous coping is a thorough mediator between spirituality and resilience. The mediating role of courageous coping did not reach meaningful levels between social support and resilience, meaning that zero was included in bootstrap interval, CI 95% (β =.026, CI [-.002, .089]). Social support, adopting a spiritual strategy, and using coping strategies played a crucial role among the adolescents to promote their level of resilience to cope with hemophilia.

Predicting outcomes of phase III oncology trials with Bayesian mediation modeling of tumor response

Pivotal cancer trials often fail to yield evidence in support of new therapies thought to offer promising alternatives to standards-of-care. Conducting randomized controlled trials in oncology tends to be considerably more expensive than studies of other diseases with comparable sample size. Moreover, phase III trial design often takes place with a paucity of survival data for experimental therapies. Experts have explained the failures on the basis of design flaws which produce studies with unrealistic expectations. This article presents a framework for predicting outcomes of phase III oncology trials using Bayesian mediation models. Predictions, which arise from interim analyses, derive from multivariate modeling of the relationships among treatment, tumor response, and their conjoint effects on survival. Acting as a safeguard against inaccurate pre-trial design assumptions, the methodology may better facilitate rapid closure of negative studies. Additionally the models can be used to inform re-estimations of sample size for under-powered trials that demonstrate survival benefit via tumor response mediation. The methods are applied to predict the outcomes of two colorectal cancer studies. Simulation is used to evaluate and compare models in the absence versus presence of reliable surrogate markers of survival.

Understanding differences in prescription drug misuse between two Texas border communities

Objectives: The misuse of prescription drugs in the U.S. is an alarming public health crisis. Prior research at the U.S.-Mexico border has found high rates of prescription drug misuse, but with rates varying significantly across border communities. We aimed to examine a model of permissive climate measures and stress exposures as potential mediators of community differences in prescription drug misuse at the U.S.-Mexico border.Design: We analyzed data from the U.S.-Mexico Study of Alcohol and Related Conditions (UMSARC). Household, in-person interviews were conducted with Mexican-origin residents of the Texas border cities Laredo (n = 751) and Brownsville/McAllen (n = 814). Interviews assessed past-year misuse of any and pain-reliever prescription drugs. Drug availability, neighborhood safety, exposure to violence/crime, and social support were examined as potential mediators. Analyses were stratified by gender and employed regressions and mediation analysis with Mplus.Results: The past-year prevalence of any prescription drug misuse in Laredo was 26.3% among women and 24.4% among men, and in Brownsville/McAllen was 12.4% among men, and 6.7% among women. Mediation analysis revealed site effects via some of the hypothesized risk factors for men, but not for women. Specifically, for men, site effects on any and pain reliever prescription drug misuse were partially mediated via high drug availability and low family support.Conclusions: Past-year prescription drug misuse was over 3 times the 2015 national prevalence among both men and women in Laredo and calls for immediate attention. Findings regarding the model suggest drug availability and social support may be relevant to understanding community differences in prescription drug misuse among men living at the border, and that additional factors should be investigated to understand misuse among women living at the border.

Itch as Major Mediator of Effect of Tofacitinib on Health-Related Quality of Life in Psoriatic Arthritis: A Mediation Analysis

Patients with psoriatic arthritis (PsA) experience impaired health-related quality of life (HRQoL). Tofacitinib is an oral Janus kinase inhibitor for the treatment of PsA, which has been associated with improvements in dermatologic endpoints in patients with PsA. To assess the extent to which tofacitinib affects patient HRQoL via improvements in dermatologic symptoms, including itch, data were pooled from patients with PsA who received tofacitinib in phase III studies (NCT01866668 and NCT01882439). Mediation modeling assessed the indirect effects (via Itch Severity Item [ISI] and Physician’s Global Assessment of Psoriasis [PGA-PsO]) and direct effects (via all other factors) of tofacitinib treatment on dermatology-specific HRQoL (measured by Dermatology Life Quality Index [DLQI]). In the initial model, the treatment effect on DLQI was largely mediated by itch (ISI; p < 0.0001) and PGA-PsO (p < 0.01). The model was re-specified to assess the indirect effects only of itch and PGA-PsO on DLQI. Here, 17.7% of the treatment effect on DLQI was attributable to PGA-PsO (p = 0.0006), and 82.3% to itch (p < 0.0001). Tofacitinib-dependent improvements in DLQI were primarily mediated by itch relief, in addition to improvements in PGA-PsO.

Does the Presence of Anxiety and Depression Symptoms Mediate the Association Between Family Functions and Self-Efficacy in Pregnant Women in the Third Trimester?: A Community-Based Cross-Sectional Survey

Objectives: Pregnant women in the third trimester can be more vulnerable to adverse mental health outcomes, but there is limited research on the association between family function and self-efficacy and potential mediation by symptoms of anxiety and depression. Methods: The cross-sectional study enrolled 813 pregnant women in the third trimester from 14 communities of Hengyang city, Hunan province of China. All of the participants completed a battery of self-report measures of family function (Family Adaptation Partnership Growth and Resolve Index, APGAR-family), self-efficacy (General Self-Efficacy Scale, GSES), anxiety (measured by Generalized Anxiety Disorder scale with seven items, GAD-7), and depression symptoms (Patient Health Questionnaire with nine items, PHQ-9). The correlation analysis was performed using Spearman's correlation coefficient. Baron and Kenny's method and multiple mediation models with bootstrapping were used to determine whether the symptoms of anxiety and depression mediated the association between family functions and self-efficacy. Results: There were 22.6% pregnant women in the low level of self-efficacy, with 60.9% in the moderate level and 16.5% of respondents in the high level. Self-efficacy had significant correlations with both anxiety symptoms (r = -0.19, p < 0.05), depression symptoms (r = -0.22, p < 0.05), and family function (r = 0.31, p < 0.05). Anxiety symptoms were significantly associated with self-efficacy (β = -0.016, p < 0.05). Depression symptoms were significantly associated with self-efficacy (β = -0.024, p < 0.05). Anxiety and depression symptoms partly mediated the association between family function and self-efficacy, accounting for 11.4 and 16.4% of total effect, respectively. It was indicated that pregnant women with a high degree of family function are less likely to have emotional symptoms and predicted to have higher levels of self-efficacy. Conclusions: Anxiety and depression show mediating effects in the association between family function and self-efficacy. Improved family function can have a positive impact on pregnant women in the third trimester.