Doubly robust adaptive LASSO for effect modifier discovery

Effect modification occurs when the effect of a treatment on an outcome differsaccording to the level of some pre-treatment variable (the effect modifier). Assessing an effect modifier is not a straight-forward task even for a subject matter expert. In this paper, we propose a two-stageprocedure to automatically selecteffect modifying variables in a Marginal Structural Model (MSM) with a single time point exposure based on the two nuisance quantities (the conditionaloutcome expectation and propensity score). We highlight the performance of our proposal in a simulation study. Finally, to illustrate tractability of our proposed methods, we apply them to analyze a set of pregnancy data. We estimate the conditional expected difference in the counterfactual birth weight if all women were exposed to inhaled corticosteroids during pregnancy versus the counterfactual birthweight if all women were not, using data from asthma medications during pregnancy.

Modeling treatment effect modification in multidrug-resistant tuberculosis in an individual patientdata meta-analysis

Effect modification occurs while the effect of the treatment is not homogeneous across the different strata of patient characteristics. When the effect of treatment may vary from individual to individual, precision medicine can be improved by identifying patient covariates to estimate the size and direction of the effect at the individual level. However, this task is statistically challenging and typically requires large amounts of data. Investigators may be interested in using the individual patient data from multiple studies to estimate these treatment effect models. Our data arise from a systematic review of observational studies contrasting different treatments for multidrug-resistant tuberculosis, where multiple antimicrobial agents are taken concurrently to cure the infection. We propose a marginal structural model for effect modification by different patient characteristics and co-medications in a meta-analysis of observational individual patient data. We develop, evaluate, and apply a targeted maximum likelihood estimator for the doubly robust estimation of the parameters of the proposed marginal structural model in this context. In particular, we allow for differential availability of treatments across studies, measured confounding within and across studies, and random effects by study.

Moderate alcohol consumption and depressive symptoms: A marginal structural model approach promoting causal inference

Importance: Prevention of depressive symptoms and disorders is a key public health priority but requires an improved understanding of modifiable risk and protective factors. A salient unanswered question in this context is whether the apparent protective effect of alcohol against depression may be causal.Objective: To compare the effects of consistent abstinence, occasional, moderate, and heavy alcohol consumption throughout early-to-middle adulthood on depressive symptoms at age 50.Design: This secondary analysis of the National Longitudinal Survey of Youth (NLSY79) cohort employed a marginal structural model approach in assessing the relationship between alcohol consumption in early-to-middle adulthood (29-37 through 41-49) and depressive symptoms at age 50. Alcohol consumption was based on measurements at 1994, 2002, and 2006, covariates at 1992, 1994, and age 40 (1998-2006), and outcome at age 50 (2008-2016). Setting: The NLSY79 is a nationally representative, population-based cohort study.Participants: 5,667 eligible participants at baseline provided valid data on alcohol consumption, depressive symptoms, and covariates of interest. Exposure: Alcohol consumption was categorised as either abstinence, occasional, moderate, or heavy drinking in 1994, 2002, and 2006.Main Outcome and Measure: Depressive symptoms at age 50 as measured by the Centre for Epidemiological Studies-Depression Scale short form (CES-D-SF).Results: Of the 5,667 eligible participants at baseline, 2,862 [50.50%] were female and the mean age was 30.81 [2.24], with 3,593 participants providing valid outcome data for analysis. Results of linear contrasts from marginal structural models were consistent with a J-shaped relationship, where both consistent occasional (b=-0.84, CI= -1.47, -.11) and consistent moderate (b=-1.08, CI=-1.88, -.20) drinkers had significantly reduced predicted CES-D-SF scores at age 50 compared to consistent abstainers. Consistent heavy drinkers were predicted to have increased depressive symptoms, but this was not statistically significant (b=0.34, CI=-0.62, 1.25). In sex-stratified analyses, results were similar for females and males.Conclusions and Relevance: In this secondary analysis of longitudinal data accounting for time-varying exposure and confounding, consistent low-to-moderate alcohol consumption in early-to-middle adulthood predicted lower depressive symptoms at age 50, compared with those abstaining from alcohol. This work offers preliminary evidence that such protective effects may be causal.

Effect of ACE2 Expression Inhibiting Drugs on COVID-19 Disease Severity, Outcome and Length of Admission in Ethiopian Patients: A Causal Inference Using Marginal Structural Model with Inverse Probability Weight

Background The role of drugs that inhibit ACE2 expression on COVID-19 disease severity, progression and outcome has been debatable with studies reporting contradictory findings. So far, there is no such study conducted in Africa. Having clarity on this issue is relevant as these drugs are the commonly prescribed medications for patients with co-morbid illnesses who are reported to be vulnerable to COVID-19 poor outcome. Therefore, the aim of this study was to assess the effect of acute or chronic ACEIs, ARBs and/or NSAIDs use on COVID-19 disease severity, outcome and length of admission among patients with COVID-19 admitted to the Millennium COVID-19 Care Center in Ethiopia. Methods A retrospective cohort study was conducted among 945 patients with COVID-19 who were on follow up from July 2nd to December 25th, 2020. Data was described using frequency tables and cross tabulations. To identify the effect of ACEIs, ARBs and/or NSAIDs use on COVID-19 disease severity, disease outcome and length of admission, Marginal Structural Model (MSM) with inverse probability weighting (IPW) approach was used. Results Among the 945 patients studied, 115 (12.2%) had a history of ACEIs, ARBs and/or NSAIDs use. At admission, the majority (39.6%) had mild disease and 272 (28.8%) had severe disease. Among the study participants, 900 (95.2%) were discharged improved and the rest 45 (4.8%) died. The median length of admission was 14.0 days (IQR, 13–16). Multinomial Logistic Regression, Log Binomial Regression and Negative Binomial Regression models were fitted to assess the effect of ACEIs, ARBs and/or NSAIDs use on disease severity, outcome and length of admission respectively. In all the three outcome models, ACEIs, ARBs and/or NSAIDs use didn’t show a statistically significant association with the outcomes. Conclusions Acute or chronic use of ACEIs, ARBs and/or NSAIDs showed no effect on COVID-19 disease severity, outcome and length of admission and therefore should not be withdrawn from patients who need these therapies.

Statins Are Associated with Improved 28-day Mortality in Patients Hospitalized with SARS-CoV-2 Infection

Background: Statins may be protective in viral infection and have been proposed as treatment in severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection. Objective: We evaluated the effect of statins on mortality in four groups hospitalized with (SARS-CoV-2) infection (continued statin, newly initiated statin, discontinued statin, never on statin). Design: In a single center cohort study of 1179 patients hospitalized with SARS-CoV-2 infection, the outcome of death, Intensive Care Unit (ICU) admission or hospital discharge was evaluated. Patients statin use, laboratory data, and co-morbidities were determined via chart review and electronic health records. Using marginal structural models to account for timing of statin initiation and competing risks, we compared the likelihood of severe outcomes in the four statin exposure groups. Setting: Academic medical center in the United States Participants: Patients hospitalized with SARS-CoV-2 infection Measurements: 28-day mortality, ICU admission, or discharge Results: Among 1179 patients, 360 were never on a statin, 311 were newly initiated on a statin, 466 were continued on a statin, and 42 had a statin discontinued. In this cohort, 154 (13.1%) patients died by 28-days. With marginal structural model analysis, statin use reduced the hazard of 28-day mortality (HR 0.566 [CI 0.372, 0.862], p = 0.008). Both new initiation of statins (HR 0.493 [CI 0.253, 0.963], p=0.038) and continuing statin therapy reduced the hazard of 28-day mortality (HR 0.270 [CI 0.114, 0.637], p=0.003). Sensitivity analysis found that statin use was associated with improved mortality for patients > 65 years, but not for patients 65 years or younger. Limitation: Observational design Conclusion: Statin therapy during hospitalization for SARS-CoV-2 infection, including new initiation and continuation of therapy, was associated with reduced short-term mortality.

The association between continuity of care and surgery in lumbar disc herniation patients

Continuity of care is a core dimension of high-quality care in the management of disease. The purpose of this study was to investigate the association between continuity of care and lumbar surgery in patients with moderate disc herniation. The Korean National Sample Cohort was used. The target population consisted of patients who have had disc herniation more than 6 months and didn’t get surgery and red flag signs within 6 months from onset. The population was enrolled from 2004 to 2013. The Bice-Boxerman Continuity of Care was used in measuring continuity of care. The marginal structural model with time dependent survival analysis was used. In total, 29,061 patients were enrolled in the cohort. High level of continuity of care was associated with a lower risk of lumbar surgery (HR, 0.27; 95% CI, 0.20–0.27). When the index was calculated only with outpatient visits to primary care with related specialty, the HR was 0.49 (95% CI: 0.43–0.57). In exploratory analysis, patients with lumbar stenosis and spondylolisthesis had higher risk of having a low level of continuity of care. These results indicate that continuity of care is associated with lower rates of lumbar surgery in patients with moderate disc herniation.

Consistent use of lipid lowering therapy in HIV infection is associated with low mortality

Background In people living with HIV (PLWH), statins may be disproportionately effective but remain underutilized. A large prospective trial in patients with low to moderate cardiovascular (ASCVD) risk will reveal whether they should be considered in all PLWH. But its effect size may not apply to real-world PLWH with higher ASCVD and mortality risk. Also, the clinical role of non-statin lipid-lowering therapy (LLT) and LLT adherence in this population is unknown. Methods Comparative multi-level marginal structural model for all-cause mortality examining four time-updated exposure levels to LLT, antihypertensives, and aspirin in a virtual cohort of older PLWH. Incident coronary, cerebrovascular, and overall ASCVD events, serious infections, and new cancer diagnoses served as explanatory outcomes. Results In 23,276 HIV-infected US-veterans who were followed for a median of 5.2 years after virologic suppression overall mortality was 33/1000 patient years: > 3 times higher than in the US population. Use of antihypertensives or aspirin was associated with increased mortality. Past LLT use (> 1 year ago) had no effect on mortality. LLT exposure in the past year was associated with a reduced hazard ratio (HR) of death: 0.59, 95% confidence interval (CI) 0.51–0.69, p < 0.0001 for statin containing LLT and 0.71 (CI: 0.54–0.93), p = 0.03 for statin-free LLT. For consistent LLT use (> 11/12 past months) the HR of death was 0.48 (CI: 0.35–0.66) for statin-only LLT, 0.34 (CI: 0.23–0.52) for combination LLT, and 0.27 (CI: 0.15–0.48) for statin-free LLT (p < 0.0001 for all). The ASCVD risk in these patients was reduced in similar fashion. Use of statin containing LLT was also associated with reduced infection and cancer risk. Multiple contrasting subgroup analyses yielded comparable results. Confounding is unlikely to be a major contributor to our findings. Conclusions In PLWH, ongoing LLT use may lead to substantially lower mortality, but consistent long-term adherence may be required to reduce ASCVD risk. Consistent non-statin LLT may be highly effective and should be studied prospectively.

Does Free-Market Reform Induce Protest? Selection, Post-Treatment Bias, and Depoliticization

Across myriad literatures, it is widely held that expanding economic grievances induce violence, protest, or other forms of backlash. In Latin America, where economic liberalization deepened the downturn of the ‘lost decade’ of the 1980s (and 1990s), reform has been tightly associated with protest and mobilization. At the same time, liberal economic reforms have proven to be remarkably durable, even where long-promised benefits are hard to discern. This article makes the case that economically liberal reforms, despite inducing or deepening severe and sustained economic downturns, have actually undermined political protest. Previous work confirming the conventional wisdom foundered on two main methodological problems. First, selection into economic reform was a consequence of the very economic pain and macroeconomic imbalances it also served to induce. Secondly, because of this, these key (macro)economic characteristics are both pre- and post-treatment. Utilizing a marginal structural model approach to assess the impact of economic liberalization on protest outcomes net of this selection process, and the prior history of treatment, the study finds that painful reform reduces political protest even as it heightens grievances. This depoliticizing dynamic helps to explain the surprising durability of liberal reforms in Latin America.

TNF-Inhibitoren verlangsamen die radiologische Progression bei Patienten mit ankylosierender Spondylitis

<b>Objectives:</b> Tumour necrosis factor inhibitors (TNFis) have been suggested to slow radiographic progression in patients with ankylosing spondylitis. However, limitations such as variations in disease activity, complex drug administration and short follow-up duration make it difficult to determine the effect of TNFis on radiographic progression. The aim of the study was to investigate whether long-term treatment with TNFis can reduce radiographic progression in patients with ankylosing spondylitis using 18-year longitudinal real-world data. <b>Methods:</b> This retrospective study was conducted between January 2001 and December 2018 at a single centre. Among the 1280 patients whose electronic medical records were reviewed, data of 595 patients exposed to TNFis at least once were included. Among them, time intervals of TNFi exposure or non-exposure were determined in 338 patients (‹on the TNFis› or ‹off the TNFis› intervals, respectively). The difference in the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) change rate between ‹on the TNFis› and ‹off the TNFis› intervals was investigated. <b>Results:</b> We obtained 2364 intervals of 338 patients (1281 ‹on the TNFis› and 1083 ‹off the TNFis› intervals). In the marginal structural model for inverse probability of treatment weighting, the change rate of mSASSS significantly decreased with the use of TNFis (β = –0.112, p = 0.004), and the adjusted mSASSS changes were 0.848 and 0.960 per year during ‹on the TNFis› and ‹off the TNFis› intervals, respectively. <b>Conclusion:</b> Compared with treatment without TNFis, treatment with TNFis slowed radiologic progression significantly.