Application of marginal structural models in pharmacoepidemiologic studies: a systematic review

CONTINUOUS KIDNEY REPLACEMENT THERAPY DOSAGE AND MORTALITY IN CRITICALLY ILL PATIENTS: A RETROSPECTIVE COHORT STUDY USING MARGINAL STRUCTURAL MODEL

ABSTRACT

Background : Continuous kidney replacement therapy (CKRT) is a crucial intervention for hemodynamically unstable patients with acute kidney injury (AKI). Despite the recommendations to offer a CKRT dose of 20 to 25 mL/kg/h, the optimal CKRT dose remains uncertain, especially whether low-dose CKRT is associated with poor outcomes. This study investigated the association between low CKRT dosage and 90-day mortality using a marginal structural model (MSM). Methods : Using the MIMIC-IV database, adult patients who received CKRT for more than 24 h were included. Data on time-fixed and time-dependent variables were collected. Patients were categorized based on CKRT dose thresholds of 13 and 20 mL/kg/h. Results : Among the 1,329 patients, the 90-day mortality rate was 49.6%. The median age of the patients was 62 years (IQR: 52-72). Changes in CKRT dosing during treatment were frequent. Patients with a reduced delivered CKRT dose (<20 and <13 mL/kg/h) generally exhibited low values during the initial days of CKRT, with an increase in the delivered CKRT dose. After adjusting only for baseline variables (traditional Cox regression model), patients receiving CKRT doses <13 mL/kg/h had significantly greater 90-day mortality (HR: 1.70, 95% CI 1.16-2.49) than those receiving CKRT doses ≥13 mL/kg/h. However, after adjusting for time-dependent variables, the CKRT dose was not significantly associated with mortality at either the 13 or 20 mL/kg/h threshold. Additionally, there were no significant associations between the delivered CKRT dose and 90-day mortality within the range of 5 to 40 mL/kg/h. Conclusion : This study highlights the impact of methodological approaches on the association between CKRT dose and mortality and that with personalized adjustments, there may not be a lower limit of the unsafe CKRT dose. However, lower CKRT doses were initially associated with higher mortality, and adjusting for time-dependent variables nullified this association.

Estimating the Effect of PrEP in Black Men Who Have Sex with Men: A Framework to Utilize Data from Multiple Non-Randomized Studies to Estimate Causal Effects

Black men who have sex with men (MSM) are disproportionately burdened by the HIV epidemic in the US. The effectiveness of pre-exposure prophylaxis (PrEP) in preventing HIV infection has been demonstrated through randomized placebo-controlled clinical trials in several populations. Importantly, no such trial has been conducted exclusively among Black MSM in the US, and it would be unethical and infeasible to do so now. To estimate the causal effects of PrEP access, initiation, and adherence on HIV risk, we utilized causal inference methods to combine data from two non-randomized studies that exclusively enrolled Black MSM. The estimated relative risks of HIV were: (i) 0.52 (95% confidence interval: 0.21, 1.22) for individuals with versus without PrEP access, (ii) 0.48 (0.12, 0.89) for individuals who initiated PrEP but were not adherent versus those who did not initiate, and (iii) 0.23 (0.02, 0.80) for individuals who were adherent to PrEP versus those who did not initiate. Beyond addressing the knowledge gap around the effect of PrEP in Black MSM in the US, which may have ramifications for public health, we have provided a framework to combine data from multiple non-randomized studies to estimate causal effects, which has broad utility.

Handling time-varying treatments in observational studies: A scoping review and recommendations

OBJECTIVE

Time-varying treatments are common in observational studies. However, when assessing treatment effects, the methodological framework has not been systematically established for handling time-varying treatments. This study aimed to examine the current methods for dealing with time-varying treatments in observational studies and developed practical recommendations.

METHODS

We searched PubMed from 2000 to 2021 for methodological articles about time-varying treatments, and qualitatively summarized the current methods for handling time-varying treatments. Subsequently, we developed practical recommendations through interactive internal group discussions and consensus by a panel of external experts.

RESULTS

Of the 36 eligible reports (22 methodological reviews, 10 original studies, 2 tutorials and 2 commentaries), most examined statistical methods for time-varying treatments, and only a few discussed the overarching methodological process. Generally, there were three methodological components to handle time-varying treatments. These included the specification of treatment which may be categorized as three scenarios (i.e., time-independent treatment, static treatment regime, or dynamic treatment regime); definition of treatment status which could involve three approaches (i.e., intention-to-treat, per-protocol, or as-treated approach); and selection of analytic methods. Based on the review results, a methodological workflow and a set of practical recommendations were proposed through two consensus meetings.

CONCLUSIONS

There is no consensus process for assessing treatment effects in observational studies with time-varying treatments. Previous efforts were dedicated to developing statistical methods. Our study proposed a stepwise workflow with practical recommendations to assist the practice.

Reporting, handling, and interpretation of time-varying drug treatments in observational studies using routinely collected healthcare data

BACKGROUND

Time-varying drug treatments are common in studies using routinely collected health data (RCD) for assessing treatment effects. This study aimed to examine how these studies reported, handled, and interpreted time-varying drug treatments.

METHODS

A systematic search was conducted on PubMed from 2018 to 2020. Eligible studies were those used RCD to explore drug treatment effects. We summarized the reporting characteristics and methods employed for handling time-varying treatments. Logistic regressions were performed to investigate the association between study characteristics and the reporting of time-varying treatments.

RESULTS

Two hundred and fifty-six studies were included, and 225 (87.9%) studies involved time-varying treatments. Of these, 24 (10.7%) reported the proportion of time-varying treatments and 105 (46.7%) reported methods used to handle time-varying treatments. Multivariable logistic regression showed that medical studies, prespecified protocol, and involvement of methodologists were associated with a higher likelihood of reporting the methods applied to handle time-varying treatments. Among the 105 studies that reported methods, as-treated analyses were the most commonly used analysis sets, which were employed in 73.9%, 75.3% and 88.2% of studies that reported approaches for treatment discontinuation, treatment switching and treatment add-on. Among the 225 studies involved time-varying treatments, 27 (12.0%) acknowledged the potential bias introduced by treatment change, of which 14 (51.9%) suggested that potential biases may impact acceptance or rejection of the null hypothesis.

CONCLUSIONS

Among observational studies using RCD, the underreporting about the presence and methods for handling time-varying treatments was largely common. The potential biases due to time-varying treatments have frequently been disregarded. Collaborative endeavors are strongly needed to enhance the prevailing practices.

Association of Longitudinal B-Type Natriuretic Peptide Monitoring With Kidney Failure in Patients With CKD: A Cohort Study

RATIONALE & OBJECTIVE

Both hypervolemia and hypovolemia are associated with chronic kidney disease (CKD) progression. Although longitudinal monitoring of B-type natriuretic peptide (BNP) may aid physicians' decision making about the optimization of volume status, its clinical benefit remains uncertain in CKD. This study assessed the association between BNP monitoring and the risk of incident kidney replacement therapy (KRT).

STUDY DESIGN

Retrospective cohort study.

SETTING & PARTICIPANTS

A total of 2,998 outpatients with stages 3-5 of nondialyzed CKD referred to the department of nephrology at an academic hospital.

EXPOSURE

BNP monitoring.

OUTCOME

KRT, acute kidney injury (AKI), and heart failure hospitalization.

ANALYTICAL APPROACH

Marginal structural models, which create a balanced pseudo population at each time point, were applied to account for potential time-dependent confounders. Inverse probability weighted pooled logistic regression models were employed to estimate hazard ratios.

RESULTS

At baseline, the median age and estimated glomerular filtration rate were 66 years and 38.1mL/min/1.73m2, respectively. During the follow-up period (median, 5.9 [IQR, 2.8-9.9] years), 449 patients required KRT, 765 had AKI, and 236 were hospitalized for heart failure. After adjustment for time-updated clinical characteristics and physician-specific practice styles, BNP monitoring was associated with lower risks of KRT (HR, 0.44 [95% CI, 0.21-0.92]), AKI (HR, 0.36 [95% CI, 0.18-0.72]), and heart failure hospitalization (HR, 0.37 [95% CI, 0.14-0.95]). The association between BNP monitoring and KRT was attenuated after additional adjustment for AKI or heart failure hospitalization as a time-varying covariate.

LIMITATIONS

Residual confounding by measured and unmeasured variables or indications for BNP measurements.

CONCLUSIONS

BNP monitoring was associated with a lower risk of KRT among patients with CKD that did not require dialysis. This association is potentially mediated through a reduced risk of AKI or heart failure hospitalization.

PLAIN-LANGUAGE SUMMARY

Both volume overload and volume depletion are deleterious to kidney function. B-type natriuretic peptide (BNP) is a biomarker that reflects volume status not only in heart failure but also in nondialysis chronic kidney disease (CKD). Although longitudinal BNP monitoring may aid physicians' decision making about the optimization of volume status, its clinical benefit remains uncertain in CKD. In this cohort study analyzing 2,998 patients with nondialyzed CKD, BNP monitoring was associated with a lower risk of kidney replacement therapy, acute kidney injury, and heart failure hospitalization over the follow-up period. The association with kidney replacement therapy may be mediated through a reduced risk of acute kidney injury or heart failure hospitalization. BNP monitoring may aid physicians in optimal fluid management, potentially conferring better kidney outcomes.

Real world data and data science in medical research: present and future

Real world data (RWD) are generating greater interest in recent times despite being not new. There are various purposes of the RWD analytics in medical research as follows: effectiveness and safety of medical treatment, epidemiology such as incidence and prevalence of disease, burden of disease, quality of life and activity of daily living, medical costs, etc. The RWD research in medicine is a mixture of digital transformation, statistics or data science, public health, and regulatory science. Most of the articles describing the RWD or real-world evidence (RWE) in medical research cover only a portion of these specializations, which might lead to an incomplete understanding of the RWD. This article summarizes the overview and challenges of the RWD analysis in medical fields from methodological perspectives. As the first step for the RWD analysis, data source of the RWD should be comprehended. The progress of the RWD is closely related to the digitization, especially of medical administrative data and medical records. Second, the selection of appropriate statistical and epidemiological methods is highly critical for an RWD analysis than those for randomized clinical trials. This is because it contains greater varieties of bias, which should be controlled by balancing the underlying risk between treatment groups. Last, the future of the RWD is discussed in terms of overcoming limited data by proxy confounders, using unstructured text data, linking of multiple databases, using the RWD or RWE for a regulatory purpose, and evaluating values and new aspects in medical research brought by the RWD.

Addressing Posttreatment Selection Bias in Comparative Effectiveness Research, Using Real-World Data and Simulation

To examine methodologies that address imbalanced treatment switching and censoring, 6 different analytical approaches were evaluated under a comparative effectiveness framework: intention-to-treat, as-treated, intention-to-treat with censor-weighting, as-treated with censor-weighting, time-varying exposure, and time-varying exposure with censor-weighting. Marginal structural models were employed to address time-varying exposure, confounding, and possibly informative censoring in an administrative data set of adult patients who were hospitalized with acute coronary syndrome and treated with either clopidogrel or ticagrelor. The effectiveness endpoint included first occurrence of death, myocardial infarction, or stroke. These methodologies were then applied across simulated data sets with varying frequencies of treatment switching and censoring to compare the effect estimate of each analysis. The findings suggest that implementing different analytical approaches has an impact on the point estimate and interpretation of analyses, especially when censoring is highly unbalanced.

Comparative effectiveness and safety of prasugrel versus ticagrelor following percutaneous coronary intervention: An observational study

BACKGROUND

Observational studies comparing ticagrelor and prasugrel in patients with acute coronary syndrome (ACS) undergoing percutaneous coronary intervention (PCI) have yielded contradictory results, but these studies often did not consider differential censoring (eg, for treatment switching or insurance disenrollment) or confounding by time-dependent factors.

OBJECTIVE

Our objective was to conduct a comparative effectiveness and safety analysis of ticagrelor and prasugrel in patients who underwent PCI after being hospitalized for an ACS.

METHODS

This study used the Optum's de-identified Clinformatics^® Data Mart Database and included patients aged 18 years or older with an index hospital admission between May 2012 and September 2015, a diagnosis of ACS managed with PCI, and treatment with either ticagrelor or prasugrel. The primary composite outcome was defined as the first occurrence of all-cause death, myocardial infarction (MI), or ischemic stroke. The secondary composite outcome included the first occurrence of gastrointestinal (GI) bleed, intracranial hemorrhage (ICH), or other major bleeds requiring hospitalization. Weighted Cox proportional hazard models and robust variance estimation were implemented to adjust for baseline comorbidities, time-varying exposure, time-dependent confounders, and differential censoring.

RESULTS

Included in the analysis were 2559 patients initiated on ticagrelor and 4456 patients initiated on prasugrel following PCI. Patients initiated on ticagrelor were 10% more likely to have eligibility disenrollment (Ticagrelor: 57%, Prasugrel: 47%, p < 0.01) and 7 percentage-points more likely to switch medication (Ticagrelor: 35%, Prasugrel: 28%, p < 0.01). After adjusting for multiple factors, including time-varying exposure and censoring imbalance, ticagrelor was associated with a higher risk of all-cause death, MI, or stroke when compared to prasugrel (Hazard ratio (HR): 1.33; 95% CI: 1.04-1.68). Similarly, ticagrelor was associated with a higher risk in bleeding events when compared with prasugrel (HR: 1.61; 95% CI: 1.19-2.17).

CONCLUSION

When compared with ticagrelor, prasugrel use following PCI for ACS was associated with a lower risk of death, MI, or stroke, as well as a reduced risk of major bleeding.

Prenatal exposure to non-steroidal anti-inflammatory drugs and risk of attention-deficit/hyperactivity disorder: A follow-up study in the Norwegian mother, father and child cohort

PURPOSE

To estimate the association between Attention-Deficit/Hyperactivity Disorder (ADHD) in children in preschool and primary school, and prenatal exposure to non-steroidal anti-inflammatory drugs (NSAIDs) by timing and duration.

METHODS

This study was based on the Norwegian Mother, Father and Child Cohort Study linked to the Medical Birth Registry of Norway, the Norwegian Patient Registry (NPR) and the Norwegian Prescription Database (NorPD). NSAID exposure was identified by maternal self-report in pregnancy. Child diagnosis of ADHD was obtained from NPR and NorPD. Symptoms of ADHD at age 5 years were measured using Conners' Parent Rating Scale-Revised, where higher scores correspond to more symptoms. To account for time-varying exposure and confounders, marginal structural models were fitted to estimate hazard ratios and mean difference in z-scores.

RESULTS

The analyses on ADHD diagnosis and ADHD symptoms included 56 340 and 34 961 children respectively. Children exposed to NSAIDs prenatally had no increased risk of ADHD diagnosis (first trimester: HR 1.12, 95% CI 0.86;1.45, second trimester: HR 0.98, 95% CI 0.69;1.38, third trimester: HR 0.68, 95% CI 0.31; 1.46) or ADHD symptoms (first trimester: standardized mean difference 0.03, 95% CI -0.03;0.09, second trimester: standardized mean difference 0.03, 95% CI -0.04;0.11, third trimester: standardized mean difference 0.11, 95% CI -0.03; 0.25). There was no duration-response relationship for either outcome.

CONCLUSION

Though non-differential misclassification of the exposure may have attenuated results, these findings are reassuring and suggest no substantially increased risk of ADHD diagnosis or symptoms in children prenatally exposed to NSAIDs, regardless of timing or duration.

Causal models adjusting for time-varying confounding-a systematic review of the literature

BACKGROUND

Obtaining unbiased causal estimates from longitudinal observational data can be difficult due to exposure-affected time-varying confounding. The past decade has seen considerable development in methods for analysing such complex longitudinal data. However, the extent to which those methods have been implemented is unclear. This study describes and characterizes the state of the field in methods adjusting for exposure-affected time-varying confounding, and examines their use in the literature.

METHODS

We systematically reviewed the literature from 2000 to 2016 for studies adjusting for time-dependent confounding, including use of specific methods like inverse probability of treatment weighting (IPTW). Articles were coded based on the methods used and, for applied articles, the topic areas covered.

RESULTS

We screened 4239 abstracts, and subsequently reviewed 1100 articles, leaving 542 relevant articles in the analyses. The number of published articles increased from two in 2000, to 112 in 2016. This increase was primarily in applied articles using IPTW, which increased from one study in 2000, to 90 in 2016. Of the 432 studies with applications to observed data, 60.9% were on at least one of: HIV (30.6%), cardiopulmonary health (13.2%), kidney disease (11.8%) or mental health (10.0%).

CONCLUSIONS

There has been marked growth in reports addressing exposure-affected time-varying confounding. This was driven by work in a small number of topic areas, with other areas showing relatively little uptake. In addition, despite developments in more advanced methods such doubly robust techniques and estimation via machine learning, implementation has been largely concentrated on the simpler, yet potentially less robust, IPTW.

Testing the Effectiveness and Cost-Effectiveness of a Combination HIV Prevention Intervention Among Young Cisgender Men Who Have Sex With Men and Transgender Women Who Sell or Exchange Sex in Thailand: Protocol for the Combination Prevention Effectiveness Study

Background

Pre-exposure prophylaxis (PrEP) is highly effective in the prevention of HIV acquisition, particularly for men who have sex with men (MSM). Questions remain on the benefits of PrEP and implementation strategies for those at occupational risk of HIV acquisition in sex work, as well as on methods to support adherence among young people who initiate PrEP.

Objective

The Combination Prevention Effectiveness study for young cisgender MSM and transgender women (TGW) aims to assess the effectiveness and cost-effectiveness of a combination intervention among HIV-uninfected young MSM and TGW engaged in sex work in Thailand.

Methods

This open-label, nonrandomized assessment compares the relative effectiveness of a combination prevention intervention with and without daily oral emtricitabine and tenofovir disoproxil fumarate (Truvada) PrEP with SMS-based adherence support. HIV-uninfected young MSM and TGW aged 18 to 26 years in Bangkok and Pattaya who self-report selling/exchanging sex at least once in the previous 12 months are recruited by convenience sampling and peer referral and are eligible regardless of their intent to initiate PrEP. At baseline, participants complete a standard assessment for PrEP eligibility and may initiate PrEP then or at any time during study participation. All participants complete a survey and HIV testing at baseline and every 3 months. Participants who initiate PrEP complete monthly pill pickups and may opt-in to SMS reminders. All participants are sent brief weekly SMS surveys to assess behavior with additional adherence questions for those who initiated PrEP. Adherence is defined as use of 4 or more pills within the last 7 days. The analytic plan uses a person-time approach to assess HIV incidence, comparing participant time on oral PrEP to participant time off oral PrEP for 12 to 24 months of follow-up, using a propensity score to control for confounders. Enrollment is based on the goal of observing 620 person-years (PY) on PrEP and 620 PY off PrEP.

Results

As of February 2019, 445 participants (417 MSM and 28 TGW) have contributed approximately 168 PY with 95% (73/77) retention at 12 months. 74.2% (330/445) of enrolled participants initiated PrEP at baseline, contributing to 134 PY of PrEP adherence, 1 PY nonadherence, and 33 PY PrEP nonuse/noninitiation. Some social harms, predominantly related to unintentional participant disclosure of PrEP use and peer stigmatization of PrEP and HIV, have been identified.

Conclusions

The majority of cisgender MSM and TGW who exchange sex and participate in this study are interested in PrEP, report taking sufficient PrEP, and stay on PrEP, though additional efforts are needed to address community misinformation and stigma. This novel multilevel, open-label study design and person-time approach will allow evaluation of the effectiveness and cost-effectiveness of combination prevention intervention in the contexts of both organized sex work and exchanged sex.

International Registered Report Identifier (IRRID)

RR1-10.2196/15354

Cardiovascular effect of discontinuing statins for primary prevention at the age of 75 years: a nationwide population-based cohort study in France

AIMS

The role of statin therapy in primary prevention of cardiovascular disease in persons older than 75 years remains a subject of debate with little evidence to support or exclude the benefit of this treatment. We assessed the effect of statin discontinuation on cardiovascular outcomes in previously adherent 75-year-olds treated for primary prevention.

METHODS AND RESULTS

A population-based cohort study using French national healthcare databases was performed, studying all subjects who turned 75 in 2012-14, with no history of cardiovascular disease and with a statin medication possession ratio ≥80% in each of the previous 2 years. Statin discontinuation was defined as three consecutive months without exposure. The outcome was hospital admission for cardiovascular event. The hazard ratio comparing statin discontinuation with continuation was estimated using a marginal structural model adjusting for both baseline and time-varying covariates (cardiovascular drug use, comorbidities, and frailty indicators). A total of 120 173 subjects were followed for an average of 2.4 years, of whom 17 204 (14.3%) discontinued statins and 5396 (4.5%) were admitted for a cardiovascular event. The adjusted hazard ratios for statin discontinuation were 1.33 [95% confidence interval (CI) 1.18-1.50] (any cardiovascular event), 1.46 (95% CI 1.21-1.75) (coronary event), 1.26 (95% CI 1.05-1.51) (cerebrovascular event), and 1.02 (95% CI 0.74-1.40) (other vascular event).

CONCLUSION

Statin discontinuation was associated with a 33% increased risk of admission for cardiovascular event in 75-year-old primary prevention patients. Future studies, including randomized studies, are needed to confirm these findings and support updating and clarification of guidelines on the use of statins for primary prevention in the elderly.

Multicenter Analysis of Dosing Protocols for Sotalol Initiation

Sotalol, a Vaughan-Williams Class III antiarrhythmic medication, is used to manage atrial arrhythmias. Due to its QT-prolonging effect and subsequent increased risk of torsade de pointes, many centers admit patients during the initial dosing period. Despite its widespread use, little information is available regarding dosing protocols during this period. In this multicenter investigation, dosing protocols in patients initiating sotalol therapy were examined to identify predictors of successful sotalol initiation. Over a 4-year period, patients admitted to 5 hospitals in the United States for inpatient telemetry monitoring during initiation for nonresearch purposes were enrolled. A 3-day course of 5 of 6 doses of sotalol was considered successful completion of the loading protocol. Of the 213 enrolled patients, over 90% were successfully discharged on sotalol. Significant bradycardia, ineffectiveness, and excessive QT prolongation were reasons for failed completion. Absence of a dose adjustment was a strong predictor of successful initiation (odds ratio: 6.6, 95% confidence interval: 1.3-32.7, P = .02). Hypertension, use of a calcium channel blocker, use of a separate β-blocker, and presence of a pacemaker were predictors of dose adjustments. Marginal structural models (ie, inverse probability weighting based on probability of a dose adjustment) verified that these factors also predicted successful initiation via preventing any dose adjustment and suggests that considering these factors may result in a higher likelihood of successful initiation in future investigations. In conclusion, we found that the majority of patients admitted for sotalol initiation are successfully discharged on the medication. The study findings suggest that factors predicting need for dose adjustment can be used to identify patients who could undergo outpatient initiation. Prospective studies are needed to verify this approach.

Nonrandomized studies using causal-modeling may give different answers than RCTs: a meta-epidemiological study

OBJECTIVES

To evaluate how estimated treatment effects agree between nonrandomized studies using causal modeling with marginal structural models (MSM-studies) and randomized trials (RCTs).

STUDY DESIGN

Meta-epidemiological study.

SETTING

MSM-studies providing effect estimates on any healthcare outcome of any treatment were eligible. We systematically sought RCTs on the same clinical question and compared the direction of treatment effects, effect sizes, and confidence intervals.

RESULTS

The main analysis included 19 MSM-studies (1,039,570 patients) and 141 RCTs (120,669 patients). MSM-studies indicated effect estimates in the opposite direction from RCTs for eight clinical questions (42%), and their 95% CI (confidence interval) did not include the RCT estimate in nine clinical questions (47%). The effect estimates deviated 1.58-fold between the study designs (median absolute deviation OR [odds ratio] 1.58; IQR [interquartile range] 1.37 to 2.16). Overall, we found no systematic disagreement regarding benefit or harm but confidence intervals were wide (summary ratio of odds ratios [sROR] 1.04; 95% CI 0.88 to 1.23). The subset of MSM-studies focusing on healthcare decision-making tended to overestimate experimental treatment benefits (sROR 1.44; 95% CI 0.99 to 2.09).

CONCLUSION

Nonrandomized studies using causal modeling with MSM may give different answers than RCTs. Caution is still required when nonrandomized "real world" evidence is used for healthcare decisions.

Missing Data in Marginal Structural Models: A Plasmode Simulation Study Comparing Multiple Imputation and Inverse Probability Weighting

BACKGROUND

The use of marginal structural models (MSMs) to adjust for time-varying confounding has increased in epidemiologic studies. However, in the setting of MSMs, recommendations for how best to handle missing data are contradictory. We present a plasmode simulation study to compare the validity and precision of MSMs estimates using complete case analysis (CC), multiple imputation (MI), and inverse probability weighting (IPW) in the presence of missing data on time-independent and time-varying confounders.

MATERIALS AND METHODS

Simulations were based on a cohort substudy using data from the Osteoarthritis Initiative which estimated the marginal causal effect of intra-articular injection use on yearly changes in knee pain. We simulated 81 scenarios with parameter values varied on missing mechanisms (MCAR, MAR, and MNAR), percentages of missing (10%, 20%, and 30%), type of confounders (time-independent, time-varying, either or both), and analytical approaches (CC, IPW, and MI). The performance of CC, IPW, and MI methods was compared using relative bias, mean squared error of the estimates of interest, and empirical power.

RESULTS

Across scenarios defined by missing data mechanism, extent of missing data, and confounder type, MI generally produced less biased estimates (range: 1.2%-6.7%) with better precision (range: 0.17-0.18) compared with IPW (relative bias: -5.3% to 8.0%; precision: 0.19-0.53). Empirical power was constant across the scenarios using MI.

CONCLUSIONS

Under simple yet realistically constructed scenarios, MI seems to confer an advantage over IPW in MSMs applications.

Clinical Events After Discontinuation of β‐Blockers in Patients Without Heart Failure Optimally Treated After Acute Myocardial Infarction

Background:

β-blockers have been among the first medications shown to improve outcomes after acute myocardial infarction (AMI). With the advent of reperfusion therapy and other secondary-prevention medications, their role has become uncertain, and large-scale experience after AMI in the contemporary era is lacking. In particular, the effect of stopping β-blockers in patients initially treated after AMI is unknown.

Methods and Results:

Using the French healthcare databases, 73 450 patients (<80 years of age), admitted for AMI in 2007 to 2012, without acute coronary syndrome (ACS) in the previous 2 years and no evidence of heart failure, having received optimal treatment with myocardial revascularization and all recommended medications in the 4 months after index admission, and not having discontinued β-blockers before 1 year, were followed for 3.8 years on average. β-Blocker discontinuation was defined as 4 consecutive months without exposure. If β-blocker treatment was resumed later on, follow-up was stopped. Both the risk of the composite outcome of death or admission for ACS and the risk of all-cause mortality were assessed in relation with β-blocker discontinuation during follow-up. Adjusted hazard ratios were estimated using marginal structural models accounting for time-varying confounders affected by previous exposure. A similar analysis was performed with statins. Of 204 592 patient-years, 12 002 (5.9%) corresponded to discontinued β-blocker treatment. For β-blocker discontinuation, the adjusted hazard ratio for death or ACS was 1.17 (95% confidence interval, 1.01–1.35); for all-cause death, the adjusted hazard ratio was 1.13 (95% confidence interval, 0.94–1.36). In contrast, for statin discontinuation, the adjusted hazard ratios for death or ACS and for all-cause death were 2.31 (95% confidence interval, 2.01–2.65) and 2.57 (95% confidence interval, 2.19–3.02), respectively.

Conclusions:

In routine care of patients without heart failure, revascularized and optimally treated after AMI, discontinuation of β-blockers beyond 1 year after AMI was associated with an increased risk of death or readmission for ACS, while statistical significance was not reached for the association with all-cause mortality. A contemporary randomized clinical trial is needed to precise the role of β-blockers in the long-term treatment after AMI.

Comparative cardiovascular safety of nonsteroidal anti-inflammatory drugs in patients with hypertension: a population-based cohort study

AIMS

Previous studies have suggested that nonsteroidal anti-inflammatory drugs (NSAIDs) may be associated with higher cardiovascular risks. However, few have been active comparison studies that directly assessed the potential differential cardiovascular risk between NSAID classes or across individual NSAIDs. We compared the risk of major cardiovascular events between cyclooxygenase 2 (COX-2)-selective and nonselective NSAIDs in patients with hypertension.

METHODS

We conducted a cohort study of patients with hypertension who initiated COX-2-selective or nonselective NSAIDs in a population-based Taiwanese database. The outcomes included hospitalization for the following major cardiovascular events: ischaemic stroke, acute myocardial infarction, congestive heart failure, transient ischaemic attack, unstable angina or coronary revascularization. We followed patients for up to 4 weeks, based on the as-treated principle. We used inverse probability weighting to control for baseline and time-varying covariates, and estimated the on-treatment hazard ratios (HRs) and 95% conservative confidence interval (CIs).

RESULTS

We identified 2749 eligible COX-2-selective NSAID users and 52 880 eligible nonselective NSAID users. The HR of major cardiovascular events comparing COX-2-selective with nonselective NSAIDs after adjusting for baseline and time-varying covariates was 1.07 (95% CI 0.65, 1.74). We did not observe a differential risk when comparing celecoxib to diclofenac (HR 1.17; 95% CI 0.61, 2.25), ibuprofen (HR 1.36; 95% CI 0.58, 3.18) or naproxen (HR 0.75; 95% CI 0.23, 2.44). There was an increased risk with COX-2-selective NSAIDs, however, when comparing COX-2-selective NSAIDs with mefenamic acid (HR 2.11; 95% CI 1.09, 4.09).

CONCLUSIONS

Our results provide important information about the comparative cardiovascular safety of NSAIDs in patients with hypertension.

Diagnostics for Confounding of Time-varying and Other Joint Exposures

The effects of joint exposures (or exposure regimes) include those of adhering to assigned treatment versus placebo in a randomized controlled trial, duration of exposure in a cohort study, interactions between exposures, and direct effects of exposure, among others. Unlike the setting of a single point exposure (e.g., propensity score matching), there are few tools to describe confounding for joint exposures or how well a method resolves it. Investigators need tools that describe confounding in ways that are conceptually grounded and intuitive for those who read, review, and use applied research to guide policy. We revisit the implications of exchangeability conditions that hold in sequentially randomized trials, and the bias structure that motivates the use of g-methods, such as marginal structural models. From these, we develop covariate balance diagnostics for joint exposures that can (1) describe time-varying confounding, (2) assess whether covariates are predicted by prior exposures given their past, the indication for g-methods, and (3) describe residual confounding after inverse probability weighting. For each diagnostic, we present time-specific metrics that encompass a wide class of joint exposures, including regimes of multivariate time-varying exposures in censored data, with multivariate point exposures as a special case. We outline how to estimate these directly or with regression and how to average them over person-time. Using a simulated example, we show how these metrics can be presented graphically. This conceptually grounded framework can potentially aid the transparent design, analysis, and reporting of studies that examine joint exposures. We provide easy-to-use tools to implement it.

Launching Effectiveness Research to Guide Practice in Neurosurgery: A National Institute Neurological Disorders and Stroke Workshop Report

This workshop addressed challenges of clinical research in neurosurgery. Randomized controlled clinical trials (RCTs) have high internal validity, but often insufficiently generalize to real-world practice. Observational studies are inclusive but often lack sufficient rigor. The workshop considered possible solutions, such as (1) statistical methods for demonstrating causality using observational data; (2) characteristics required of a registry supporting effectiveness research; (3) trial designs combining advantages of observational studies and RCTs; and (4) equipoise, an identified challenge for RCTs. In the future, advances in information technology potentially could lead to creation of a massive database where clinical data from all neurosurgeons are integrated and analyzed, ending the separation of clinical research and practice and leading to a new "science of practice."

Intensive care unit (ICU)-acquired bacteraemia and ICU mortality and discharge: addressing time-varying confounding using appropriate methodology

BACKGROUND

Studies often ignore time-varying confounding or may use inappropriate methodology to adjust for time-varying confounding.

AIM

To estimate the effect of intensive care unit (ICU)-acquired bacteraemia on ICU mortality and discharge using appropriate methodology.

METHODS

Marginal structural models with inverse probability weighting were used to estimate the ICU mortality and discharge associated with ICU-acquired bacteraemia among patients who stayed more than two days at the general ICU of a London teaching hospital and remained bacteraemia-free during those first two days. For comparison, the same associations were evaluated with (i) a conventional Cox model, adjusting only for baseline confounders and (ii) a Cox model adjusting for baseline and time-varying confounders.

FINDINGS

Using the marginal structural model with inverse probability weighting, bacteraemia was associated with an increase in ICU mortality (cause-specific hazard ratio (CSHR): 1.29; 95% confidence interval (CI): 1.02-1.63) and a decrease in discharge (CSHR: 0.52; 95% CI: 0.45-0.60). By 60 days, among patients still in the ICU after two days and without prior bacteraemia, 8.0% of ICU deaths could be prevented by preventing all ICU-acquired bacteraemia cases. The conventional Cox model adjusting for time-varying confounders gave substantially different results [for ICU mortality, CSHR: 1.08 (95% CI: 0.88-1.32); for discharge, CSHR: 0.68 (95% CI: 0.60-0.77)].

CONCLUSION

In this study, even after adjusting for the timing of acquiring bacteraemia and time-varying confounding using inverse probability weighting for marginal structural models, ICU-acquired bacteraemia was associated with a decreased daily ICU discharge risk and an increased risk of ICU mortality.

Estimating inverse probability weights using super learner when weight-model specification is unknown in a marginal structural Cox model context

Correct specification of the inverse probability weighting (IPW) model is necessary for consistent inference from a marginal structural Cox model (MSCM). In practical applications, researchers are typically unaware of the true specification of the weight model. Nonetheless, IPWs are commonly estimated using parametric models, such as the main-effects logistic regression model. In practice, assumptions underlying such models may not hold and data-adaptive statistical learning methods may provide an alternative. Many candidate statistical learning approaches are available in the literature. However, the optimal approach for a given dataset is impossible to predict. Super learner (SL) has been proposed as a tool for selecting an optimal learner from a set of candidates using cross-validation. In this study, we evaluate the usefulness of a SL in estimating IPW in four different MSCM simulation scenarios, in which we varied the specification of the true weight model specification (linear and/or additive). Our simulations show that, in the presence of weight model misspecification, with a rich and diverse set of candidate algorithms, SL can generally offer a better alternative to the commonly used statistical learning approaches in terms of MSE as well as the coverage probabilities of the estimated effect in an MSCM. The findings from the simulation studies guided the application of the MSCM in a multiple sclerosis cohort from British Columbia, Canada (1995-2008), to estimate the impact of beta-interferon treatment in delaying disability progression. Copyright © 2017 John Wiley & Sons, Ltd.

The impact of adherence to inhaled drugs on 5-year survival in COPD patients: a time dependent approach

Purpose

Whether inhaled medications improve long‐term survival in Chronic Obstructive Pulmonary Disease (COPD) is an open question. The purpose of this study is to assess the impact of adherence to inhaled drug use on 5‐year survival in COPD.

Methods

A population‐based cohort study in three Italian regions was conducted using healthcare linked datasets (hospitalization, mortality, drugs). Individuals (45+ years) discharged after COPD exacerbation in 2006–2009 were enrolled. Inhaled drug daily use during 5‐year follow‐up was determined through Proportion of Days Covered on the basis of Defined Daily Doses. Five levels of time‐dependent exposure were identified: (i) long‐acting β2 agonists and inhaled corticosteroids (LB/ICS) regular use; (ii) LB/ICS occasional use; (iii) LB regular use; (iv) LB occasional use; and (v) respiratory drugs other than LB. Cox regression models adjusted for baseline (socio‐demographic, comorbidities, drug use) and time‐dependent characteristics (COPD exacerbations, cardiovascular hospitalizations, cardiovascular therapy) were performed.

Results

A total of 12 124 individuals were studied, 46% women, mean age 73,8 years. Average follow‐up time 2,4 year. A total of 3415 subjects died (mortality rate = 11.9 per 100 person years). In comparison to LB/ICS regular use, higher risks of death for all remaining treatments were found, the highest risk for respiratory drugs other than LB category (HR = 1.63, 95%CI 1.43–1.87). Patients with regular LB use had higher survival than those with LB/ICS occasional use (HR = 0.89, 95%CI 0.79–0.99).

Conclusions

These findings support clinical guidelines and recommendations for the regular use of inhaled drugs to improve health status and prognosis among moderate–severe COPD patients. © 2016 The Authors. Pharmacoepidemiology and Drug Safety Published by John Wiley & Sons Ltd.

Continuous renal replacement therapy versus intermittent hemodialysis in intensive care patients: impact on mortality and renal recovery

PURPOSE

The best renal replacement therapy (RRT) modality remains controversial. We compared mortality and short- and long-term renal recovery between patients treated with continuous RRT and intermittent hemodialysis.

METHODS

Patients of the prospective observational multicenter cohort database OUTCOMEREA™ were included if they underwent at least one RRT session between 2004 and 2014. Differences in patients' baseline and daily characteristics between treatment groups were taken into account by using a marginal structural Cox model, allowing one to substantially reduce the bias resulting from confounding factors in observational longitudinal data analysis. The composite primary endpoint was 30-day mortality and dialysis dependency.

RESULTS

Among 1360 included patients with RRT, 544 (40.0 %) and 816 (60.0 %) were initially treated by continuous RRT and intermittent hemodialysis, respectively. At day 30, 39.6 % patients were dead. Among survivors, 23.8 % still required RRT. There was no difference between groups for the primary endpoint in global population (HR 1.00, 95 % CI 0.77-1.29; p = 0.97). In patients with higher weight gain at RRT initiation, mortality and dialysis dependency were significantly lower with continuous RRT (HR 0.54, 95 % CI 0.29-0.99; p = 0.05). Conversely, this technique appeared to be deleterious in patients without shock (HR 2.24, 95 % CI 1.24-4.04; p = 0.01). Six-month mortality and persistent renal dysfunction were not influenced by the RRT modality in patients with dialysis dependence at ICU discharge.

CONCLUSION

Continuous RRT did not appear to improve 30-day and 6-month patient outcomes. It seems beneficial for patients with fluid overload, but might be deleterious in the absence of hemodynamic failure.

The Choice of Analytical Strategies in Inverse-Probability-of-Treatment-Weighted Analysis: A Simulation Study

We sought to explore the impact of intention to treat and complex treatment use assumptions made during weight construction on the validity and precision of estimates derived from inverse-probability-of-treatment-weighted analysis. We simulated data assuming a nonexperimental design that attempted to quantify the effect of statin on lowering low-density lipoprotein cholesterol. We created 324 scenarios by varying parameter values (effect size, sample size, adherence level, probability of treatment initiation, associations between low-density lipoprotein cholesterol and treatment initiation and continuation). Four analytical approaches were used: 1) assuming intention to treat; 2) assuming complex mechanisms of treatment use; 3) assuming a simple mechanism of treatment use; and 4) assuming invariant confounders. With a continuous outcome, estimates assuming intention to treat were biased toward the null when there were nonnull treatment effect and nonadherence after treatment initiation. For each 1% decrease in the proportion of patients staying on treatment after initiation, the bias in estimated average treatment effect increased by 1%. Inverse-probability-of-treatment-weighted analyses that took into account the complex mechanisms of treatment use generated approximately unbiased estimates. Studies estimating the actual effect of a time-varying treatment need to consider the complex mechanisms of treatment use during weight construction.

Failure of empirical systemic antifungal therapy in mechanically ventilated critically ill patients

RATIONALE

Systemic antifungal treatments are empirically administered to the sickest critically ill patients, often without documented invasive fungal infection.

OBJECTIVES

To estimate the impact of systemic antifungal treatment on 30-day survival of patients suspected to have invasive candidiasis.

METHODS

All nonneutropenic, nontransplant recipients managed in five intensive care units intubated for at least 5 days, and free of invasive candidiasis, were included. To account for differences in patients' characteristics recorded daily before study end point, a causal model for longitudinal data was used to assess benefits from antifungal treatment. The composite primary end point was hospital mortality or occurrence of invasive candidiasis.

MEASUREMENTS AND MAIN RESULTS

Among 1,491 patients, 100 (6.7%) received antifungal treatment for a suspected infection. Patients treated with antifungals were more severely ill than untreated patients. Within the 30-day follow-up period, 363 (24.3%) patients died, and 22 (1.5%) exhibited documented invasive candidiasis. After adjustment on baseline and time-dependent confounders (underlying illness, severity, invasive procedures, Candida colonization), and using a marginal structural model for longitudinal data, treatment was not associated with a decreased risk of mortality or of occurrence of invasive candidiasis (hazard ratio, 1.05; 95% confidence interval, 0.56-1.96; P = 0.91).

CONCLUSIONS

This study failed to show outcome benefits for empirical systemic antifungal therapy in the sickest critically ill, nonneutropenic, nontransplanted patients. The post hoc power did not allow us to conclude to an absence of treatment effect especially for specific subgroups. Studies to refine indications for empirical treatment based on surrogate markers of invasive candidiasis are warranted.

Severe COPD Exacerbation Risk and Long-Acting Bronchodilator Treatments: Comparison of Three Observational Data Analysis Methods

OBJECTIVE

Results from three observational methods for assessing effectiveness of long-acting bronchodilator therapies for reducing severe exacerbations of chronic obstructive pulmonary disease (COPD) were compared: intent-to-treat (ITT), as protocol (AP), and an as-treated analysis that utilized a marginal structural model (MSM) incorporating time-varying covariates related to treatment adherence and moderate exacerbations.

STUDY DESIGN AND SETTING

Severe exacerbation risk was assessed over a 2-year period using claims data for patients aged ≥40 years who initiated long-acting muscarinic antagonist (LAMA), inhaled corticosteroid/long-acting beta-agonist (ICS/LABA), or triple therapy (LAMA + ICS/LABA).

RESULTS

A total of 5475 COPD patients met inclusion criteria. Six months post-initiation, 53.5 % of patients discontinued using any therapy. The ITT analysis found an increased severe exacerbation risk for triple therapy treatment (hazard ratio [HR] 1.24; 95 % confidence interval [CI] 1.00-1.53). No increased risk was found in the AP (HR 1.00; 95 % CI 0.73-1.36), or MSM analyses (HR 1.11; 95 % CI 0.68-1.81). The MSM highlighted important associations among post-index events.

CONCLUSION

Neglecting to adjust for treatment discontinuation may produce biased risk estimates. The MSM approach is a promising tool to compare chronic disease management by illuminating relationships between treatment decisions, adherence, patient choices, and outcomes.