How Cox models react to a study-specific confounder in a patient-level pooled dataset: random effects better cope with an imbalanced covariate across trials unless baseline hazards differ

Abstract Background Diabetes is a public health burden that disproportionately affects military veterans and racial minorities. Studies of racial disparities are inherently observational, and thus may require the use of methods such as Propensity Score Analysis (PSA). While traditional PSA accounts for patient-level factors, this may not be sufficient when patients are clustered at the geographic level and thus important confounders, whether observed or unobserved, vary by geographic location. Methods We employ a spatial propensity score matching method to account for “geographic confounding”, which occurs when the confounding factors, whether observed or unobserved, vary by geographic region. We augment the propensity score and outcome models with spatial random effects, which are assigned scaled Besag-York-Mollié priors to address spatial clustering and improve inferences by borrowing information across neighboring geographic regions. We apply this approach to a study exploring racial disparities in diabetes specialty care between non-Hispanic black and non-Hispanic white veterans. We construct multiple global estimates of the risk difference in diabetes care: a crude unadjusted estimate, an estimate based solely on patient-level matching, and an estimate that incorporates both patient and spatial information. Results In simulation we show that in the presence of an unmeasured geographic confounder, ignoring spatial heterogeneity results in increased relative bias and mean squared error, whereas incorporating spatial random effects improves inferences. In our study of racial disparities in diabetes specialty care, the crude unadjusted estimate suggests that specialty care is more prevalent among non-Hispanic blacks, while patient-level matching indicates that it is less prevalent. Hierarchical spatial matching supports the latter conclusion, with a further increase in the magnitude of the disparity. Conclusions These results highlight the importance of accounting for spatial heterogeneity in propensity score analysis, and suggest the need for clinical care and management strategies that are culturally sensitive and racially inclusive.

Download Full-text

Bayesian Estimation of Cox Models with Non-Nested Random Effects: An Application to the Ratification of Ilo Conventions by Developing Countries

SSRN Electronic Journal ◽

10.2139/ssrn.771949 ◽

2005 ◽

Cited By ~ 1

Author(s):

Guillaume Horny ◽

Bernhard Boockmann ◽

Dragana Djurdjevic ◽

Francois Laisney

Keyword(s):

Developing Countries ◽

Bayesian Estimation ◽

Random Effects ◽

Cox Models ◽

Nested Random Effects

Download Full-text

Drug-coated balloons for dysfunctional hemodialysis venous access: A patient-level meta-analysis of randomized controlled trials

10.1101/2020.12.10.20240069 ◽

2020 ◽

Author(s):

Fong Khi Yung ◽

Joseph J Zhao ◽

Eelin Tan ◽

Nicholas Syn ◽

Rehena Sultana ◽

...

Keyword(s):

Random Effects ◽

Meta Analysis ◽

Venous Access ◽

Primary Patency ◽

Two Stage ◽

Patient Level ◽

Randomized Controlled ◽

One Stage ◽

Shared Frailty ◽

Meta Analyses

ABSTRACTPurposeTo perform an individual patient data-level meta-analysis of randomized controlled trials comparing drug-coated balloon angioplasty (DCB) against conventional percutaneous transluminal angioplasty (PTA) in the treatment of dysfunctional hemodialysis venous access.MethodsA search was conducted from inception till 13th November 2020. Kaplan-Meier curves comparing DCB to PTA by target lesion primary patency (TLPP) and access circuit primary patency (ACPP) were graphically reconstructed to retrieve patient-level data. One-stage meta-analyses with Cox-models with random-effects gramma-frailties were conducted to determine hazard ratios (HRs). Dynamic restricted mean survival times (RMST) were conducted in view of violation of the proportional hazards assumption. Conventional two-stage meta-analyses and network meta-analyses under random-effects Frequentist models were conducted to determine overall and comparative outcomes of paclitaxel concentrations utilised. Where outliers were consistently detected through outlier and influence analyses, sensitivity analyses excluding those studies were conducted.ResultsAmong 10 RCTs (1,207 patients), HRs across all models favoured DCB (one-stage shared-frailty HR=0.62, 95%-CI: 0.53–0.73, P<0.001; two-stage random-effects HR=0.60, 95%-CI: 0.42–0.86, P=0.018, I2=65%) for TLPP. Evidence of time-varying effects (P=0.005) was found. TLPP RMST was +3.47 months (25.0%) longer in DCB-treated patients compared to PTA (P=0.001) at 3-years. TLPP at 6-months, 1-year and 2-years was 75.3% vs 58.0%, 51.1% vs 37.1% and 31.3% vs 26.0% for DCB and PTA respectively. P-Scores within the Frequentist network meta-analysis suggest that higher concentrations of paclitaxel were associated with better TLPP and ACPP. Among 6 RCTs (854 patients), the one-stage model favoured DCB (shared-frailty HR=0.72, 95%-CI: 0.60–0.87, P<0.001) for ACPP. Conversely, the two-stage random-effects model demonstrated no significant difference (HR=0.76, 95%-CI: 0.35–1.67, P=0.414, I2=81%). Sensitivity analysis excluding outliers significantly favoured DCB (HR=0.61, 95%-CI: 0.41–0.91, P=0.027, I2=62%).ConclusionOverall evidence suggests that DCB is favoured over PTA in TLPP and ACPP. The increased efficacy of higher concentrations of paclitaxel may warrant further investigation.

Download Full-text

Proportion of Individuals With Past-Year Suicidal Ideation Who Attempted Suicide Over the Past 10 Years in the United States, and the Influence of Age and Sex

Crisis ◽

10.1027/0227-5910/a000690 ◽

2020 ◽

pp. 1-5

Author(s):

Shannon Lange ◽

Courtney Bagge ◽

Charlotte Probst ◽

Jürgen Rehm

Keyword(s):

United States ◽

Suicidal Ideation ◽

Random Effects ◽

Attempted Suicide ◽

Suicide Attempts ◽

The United States ◽

Individual Level ◽

The Past ◽

Meta Analyses ◽

Age Category

Abstract. Background: In recent years, the rate of death by suicide has been increasing disproportionately among females and young adults in the United States. Presumably this trend has been mirrored by the proportion of individuals with suicidal ideation who attempted suicide. Aim: We aimed to investigate whether the proportion of individuals in the United States with suicidal ideation who attempted suicide differed by age and/or sex, and whether this proportion has increased over time. Method: Individual-level data from the National Survey on Drug Use and Health (NSDUH), 2008–2017, were used to estimate the year-, age category-, and sex-specific proportion of individuals with past-year suicidal ideation who attempted suicide. We then determined whether this proportion differed by age category, sex, and across years using random-effects meta-regression. Overall, age category- and sex-specific proportions across survey years were estimated using random-effects meta-analyses. Results: Although the proportion was found to be significantly higher among females and those aged 18–25 years, it had not significantly increased over the past 10 years. Limitations: Data were self-reported and restricted to past-year suicidal ideation and suicide attempts. Conclusion: The increase in the death by suicide rate in the United States over the past 10 years was not mirrored by the proportion of individuals with past-year suicidal ideation who attempted suicide during this period.

Download Full-text

Positional Estimation Within a Latent Space Model for Networks

Methodology ◽

10.1027/1614-2241.2.1.24 ◽

2006 ◽

Vol 2 (1) ◽

pp. 24-33 ◽

Cited By ~ 17

Author(s):

Susan Shortreed ◽

Mark S. Handcock ◽

Peter Hoff

Keyword(s):

Random Effects ◽

Network Data ◽

Random Effects Models ◽

Space Model ◽

Network Analyses ◽

Recent Advances ◽

Latent Space ◽

Latent Space Model ◽

Modeling Data

Recent advances in latent space and related random effects models hold much promise for representing network data. The inherent dependency between ties in a network makes modeling data of this type difficult. In this article we consider a recently developed latent space model that is particularly appropriate for the visualization of networks. We suggest a new estimator of the latent positions and perform two network analyses, comparing four alternative estimators. We demonstrate a method of checking the validity of the positional estimates. These estimators are implemented via a package in the freeware statistical language R. The package allows researchers to efficiently fit the latent space model to data and to visualize the results.

Download Full-text