On Variance of the Treatment Effect in the Treated When Estimated by Inverse Probability Weighting

Empirical Sandwich Variance Estimator for Iterated Conditional Expectation g-Computation

Iterated conditional expectation (ICE) g-computation is an estimation approach for addressing time-varying confounding for both longitudinal and time-to-event data. Unlike other g-computation implementations, ICE avoids the need to specify models for each time-varying covariate. For variance estimation, previous work has suggested the bootstrap. However, bootstrapping can be computationally intense. Here, we present ICE g-computation as a set of stacked estimating equations. Therefore, the variance for the ICE g-computation estimator can be consistently estimated using the empirical sandwich variance estimator. Performance of the variance estimator was evaluated empirically with a simulation study. The proposed approach is also demonstrated with an illustrative example on the effect of cigarette smoking on the prevalence of hypertension. In the simulation study, the empirical sandwich variance estimator appropriately estimated the variance. When comparing runtimes between the sandwich variance estimator and the bootstrap for the applied example, the sandwich estimator was substantially faster, even when bootstraps were run in parallel. The empirical sandwich variance estimator is a viable option for variance estimation with ICE g-computation.

Uncertain about uncertainty in matching-adjusted indirect comparisons? A simulation study to compare methods for variance estimation

In health technology assessment, matching-adjusted indirect comparison (MAIC) is the most common method for pairwise comparisons that control for imbalances in baseline characteristics across trials. One of the primary challenges in MAIC is the need to properly account for the additional uncertainty introduced by the matching process. Limited evidence and guidance are available on variance estimation in MAICs. Therefore, we conducted a comprehensive Monte Carlo simulation study to evaluate the performance of different statistical methods across 108 scenarios. Four general approaches for variance estimation were compared in both anchored and unanchored MAICs of binary and time-to-event outcomes: (1) conventional estimators (CE) using raw weights; (2) CE using weights rescaled to the effective sample size (ESS); (3) robust sandwich estimators; and (4) bootstrapping. Several variants of sandwich estimators and bootstrap methods were tested. Performance was quantified on the basis of empirical coverage probabilities for 95% confidence intervals and variability ratios. Variability was underestimated by CE + raw weights when population overlap was poor or moderate. Despite several theoretical limitations, CE + ESS weights accurately estimated uncertainty across most scenarios. Original implementations of sandwich estimators had a downward bias in MAICs with a small ESS, and finite sample adjustments led to marked improvements. Bootstrapping was unstable if population overlap was poor and the sample size was limited. All methods produced valid coverage probabilities and standard errors in cases of strong population overlap. Our findings indicate that the sample size, population overlap, and outcome type are important considerations for variance estimation in MAICs.

Neoadjuvant chemotherapy combined with immunotherapy versus neoadjuvant chemoradiotherapy in patients with locally advanced esophageal squamous cell carcinoma

BACKGROUND

To date, few studies have compared effectiveness and survival rates of neoadjuvant chemotherapy combined with immunotherapy (NACI) and conventional neoadjuvant chemoradiotherapy (NCRT) in patients with locally advanced esophageal squamous cell carcinoma (ESCC). The present study was conducted to compare therapeutic response and survival between NACI and NCRT.

METHODS

The study cohort comprised patients with locally advanced ESCC treated with either NACI or NCRT followed by surgery between June 2018 and March 2021. The 2 groups were compared for treatment response, 3-year overall survival (OS), and disease-free survival (DFS). Survival curves were created using the Kaplan-Meier method, differences were compared using the log-rank test, and potential imbalances were corrected for using the inverse probability of treatment weighting (IPTW) method.

RESULTS

Among 202 patients with locally advanced ESCC, 81 received NACI and 121 received conventional NCRT. After IPTW adjustment, the R0 resection rate (85.2% vs 92.3%; P = .227) and the pathologic complete response (pCR) rate (27.5% vs 36.4%; P = .239) were comparable between the 2 groups. Nevertheless, patients who received NACI exhibited both a better 3-year OS rate (91.7% vs 79.8%; P = .032) and a better 3-year DFS rate (87.4% vs 72.8%; P = .039) compared with NCRT recipients.

CONCLUSIONS

NACI has R0 resection and pCR rates comparable to those of NCRT and seems to be correlated with better prognosis than NCRT. NACI followed by surgery may be an effective treatment strategy for locally advanced ESCC.

Variation in outcomes and practice patterns among patients with localized pancreatic cancer: the impact of the pancreatic cancer multidisciplinary clinic

Introduction

Patients with localized pancreatic adenocarcinoma (PDAC) benefit from multi-modality therapy. Whether care patterns and oncologic outcomes vary if a patient was seen through a pancreatic multi-disciplinary clinic (PMDC) versus only individual specialty clinics is unclear.

Methods

Using institutional Pancreatic Cancer Registry, we identified patients with localized PDAC from 2019- 2022 who eventually underwent resection. It was our standard practice for borderline resectable (BRPC) patients to undergo ≤4 months of neoadjuvant chemotherapy, ± radiation, followed by exploration, while locally advanced (LAPC) patients were treated with 4-6 months of chemotherapy, followed by radiation and potential exploration. Descriptive and multivariable analyses (MVA) were performed to examine the association between clinic type (PMDC vs individual specialty clinics i.e. surgical oncology, medical oncology, or radiation oncology) and study outcomes.

Results

A total of 416 patients met inclusion criteria. Of these, 267 (64.2%) had PMDC visits. PMDC group received radiation therapy more commonly (53.9% versus 27.5%, p=0.001), as compared to individual specialty clinic group. Completion of neoadjuvant treatment (NAT) was far more frequent in patients seen through PMDC compared to patients seen through individual specialty clinics (69.3% vs 48.9%). On MVA, PMDC group was significantly associated with receipt of NAT per institutional standards (adjusted OR 2.23, 95% CI 1.46-7.07, p=0.006). Moreover, the average treatment effect of PMDC on progression-free survival (PFS) was 4.45 (95CI: 0.87-8.03) months. No significant association between overall survival (OS) and clinic type was observed.

Discussion

Provision of care through PMDC was associated with significantly higher odds of completing NAT per institutional standards as compared to individual specialty clinics, which possibly translated into improved PFS. The development of multidisciplinary clinics for management of pancreatic cancer should be incentivized, and any barriers to such development should be addressed.

On variance estimation of the inverse probability-of-treatment weighting estimator: A tutorial for different types of propensity score weights

Propensity score methods, such as inverse probability-of-treatment weighting (IPTW), have been increasingly used for covariate balancing in both observational studies and randomized trials, allowing the control of both systematic and chance imbalances. Approaches using IPTW are based on two steps: (i) estimation of the individual propensity scores (PS), and (ii) estimation of the treatment effect by applying PS weights. Thus, a variance estimator that accounts for both steps is crucial for correct inference. Using a variance estimator which ignores the first step leads to overestimated variance when the estimand is the average treatment effect (ATE), and to under or overestimated estimates when targeting the average treatment effect on the treated (ATT). In this article, we emphasize the importance of using an IPTW variance estimator that correctly considers the uncertainty in PS estimation. We present a comprehensive tutorial to obtain unbiased variance estimates, by proposing and applying a unifying formula for different types of PS weights (ATE, ATT, matching and overlap weights). This can be derived either via the linearization approach or M-estimation. Extensive R code is provided along with the corresponding large-sample theory. We perform simulation studies to illustrate the behavior of the estimators under different treatment and outcome prevalences and demonstrate appropriate behavior of the analytical variance estimator. We also use a reproducible analysis of observational lung cancer data as an illustrative example, estimating the effect of receiving a PET-CT scan on the receipt of surgery.

The performance of marginal structural models for estimating risk differences and relative risks using weighted univariate generalized linear models

We used Monte Carlo simulations to compare the performance of marginal structural models (MSMs) based on weighted univariate generalized linear models (GLMs) to estimate risk differences and relative risks for binary outcomes in observational studies. We considered four different sets of weights based on the propensity score: inverse probability of treatment weights with the average treatment effect as the target estimand, weights for estimating the average treatment effect in the treated, matching weights and overlap weights. We considered sample sizes ranging from 500 to 10,000 and allowed the prevalence of treatment to range from 0.1 to 0.9. We examined both the robust variance estimator when using generalized estimating equations with an independent working correlation matrix and a bootstrap variance estimator for estimating the standard error of the risk difference and the log-relative risk. The performance of these methods was compared with that of direct weighting. Both the direct weighting approach and MSMs based on weighted univariate GLMs resulted in the identical estimates of risk differences and relative risks. When sample sizes were small to moderate, the use of an MSM with a bootstrap variance estimator tended to result in the most accurate estimates of standard errors. When sample sizes were large, the direct weighting approach and an MSM with a bootstrap variance estimator tended to produce estimates of standard error with similar accuracy. When using a MSM to estimate risk differences and relative risks, in general it is preferable to use a bootstrap variance estimator than the robust variance estimator. We illustrate the application of the different methods for estimating risks differences and relative risks using an observational study on the effect on mortality of discharge prescribing of a beta-blocker in patients hospitalized with acute myocardial infarction.

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.

M-estimation for common epidemiological measures: introduction and applied examples

M-estimation is a statistical procedure that is particularly advantageous for some comon epidemiological analyses, including approaches to estimate an adjusted marginal risk contrast (i.e. inverse probability weighting and g-computation) and data fusion. In such settings, maximum likelihood variance estimates are not consistent. Thus, epidemiologists often resort to bootstrap to estimate the variance. In contrast, M-estimation allows for consistent variance estimates in these settings without requiring the computational complexity of the bootstrap. In this paper, we introduce M-estimation and provide four illustrative examples of implementation along with software code in multiple languages. M-estimation is a flexible and computationally efficient estimation procedure that is a powerful addition to the epidemiologist's toolbox.

Performance of modeling and balancing approach methods when using weights to estimate treatment effects in observational time-to-event settings

In observational studies weighting techniques are often used to overcome bias due to confounding. Modeling approaches, such as inverse propensity score weighting, are popular, but often rely on the correct specification of a parametric model wherein neither balance nor stability are targeted. More recently, balancing approach methods that directly target covariate imbalances have been proposed, and these allow the researcher to explicitly set the desired balance constraints. In this study, we evaluate the finite sample properties of different modeling and balancing approach methods, when estimating the marginal hazard ratio, through Monte Carlo simulations. The use of the different methods is also illustrated by analyzing data from the Swedish stroke register to estimate the effect of prescribing oral anticoagulants on time to recurrent stroke or death in stroke patients with atrial fibrillation. In simulated scenarios with good overlap and low or no model misspecification the balancing approach methods performed similarly to the modeling approach methods. In scenarios with bad overlap and model misspecification, the modeling approach method incorporating variable selection performed better than the other methods. The results indicate that it is valuable to use methods that target covariate balance when estimating marginal hazard ratios, but this does not in itself guarantee good performance in situations with, e.g., poor overlap, high censoring, or misspecified models/balance constraints.

Using Balancing Weights to Target the Treatment Effect on the Treated when Overlap is Poor

Inverse probability weights are commonly used in epidemiology to estimate causal effects in observational studies. Researchers often focus on either the average treatment effect or the average treatment effect on the treated with inverse probability weighting estimators. However, poor overlap in the baseline covariates between the treated and control groups can produce extreme weights that can result in biased treatment effect estimates. One alternative to inverse probability weights are overlap weights, which target the population with the most overlap on observed covariates. Although estimates based on overlap weights produce less bias in such contexts, the causal estimand can be difficult to interpret. An alternative to model-based inverse probability weights are balancing weights, which directly target imbalances during the estimation process, rather than model fit. Here, we explore whether balancing weights allow analysts to target the average treatment effect on the treated in cases where inverse probability weights lead to biased estimates due to poor overlap. We conduct three simulation studies and an empirical application. We find that balancing weights often allow the analyst to still target the average treatment effect on the treated even when overlap is poor. We show that although overlap weights remain a key tool, more familiar estimands can sometimes be targeted by using balancing weights.

Targeted Antibiotics for Lower Respiratory Tract Infection with Corynebacterium striatum

Purpose

To assess the impact of targeted antibiotic therapy on clinical outcomes of patients with lower respiratory tract (LRT) infection with Corynebacterium striatum (C. striatum).

Methods

A new propensity score-inverse probability of treatment weighting (IPTW) cohort study was conducted by using 10-year data. The study included LRT infection patients with respiratory secretions cultured positive for C. striatum simultaneously. The primary outcome was all-cause hospital mortality; the secondary outcomes included hospital stay, ICU stay and ventilation time. The safety outcomes were drug-related serum creatinine (Cr) increase and thrombocytopenia.

Results

A total of 339 patients were included in the cohort, and 84 (24.78%) initiated vancomycin or linezolid therapy. In the new IPTW cohort, targeted antibiotic therapy did not improve all-cause hospital mortality (P=0.632), and the OR (95% CI) was 0.879 (0.519-1.488). Moreover, targeted antibiotic therapy was not associated with hospital stay (P=0.415), ICU stay (P=0.945) or ventilation time (P=0.885). The side effects of drug-related higher serum Cr (P=0.044) and thrombocytopenic levels (P=0.038) cannot be ignored.

Conclusion

Clinical benefits by vancomycin or linezolid targeted against LRT infection with C. striatum were limited and with drug-related side effects. A prospectively designed study is needed to further confirm the results.

Pathologic responses and surgical outcomes after neoadjuvant immunochemotherapy versus neoadjuvant chemoradiotherapy in patients with locally advanced esophageal squamous cell carcinoma

BACKGROUND

Currently, the role of immunotherapy in neoadjuvant setting for patients with locally advanced esophageal squamous cell carcinoma (ESCC) is gradually attracting attention. Few studies compared the efficacy of neoadjuvant immunochemotherapy (NICT) and neoadjuvant chemoradiotherapy (NCRT). Our study aimed to compare treatment response and postoperative complications after NICT followed by surgery with that after conventional NCRT in patients with locally advanced ESCC.

METHODS

Of 468 patients with locally advanced ESCC, 154 received conventional NCRT, whereas 314 received NICT. Treatment response, postoperative complications and mortality between two groups were compared. Pathological response of primary tumor was evaluated using the Mandard tumor regression grade (TRG) scoring system. Pathological complete response (pCR) of metastatic lymph nodes (LNs) was defined as no viable tumor cell within all resected metastatic LNs. According to regression directionality, tumor regression pattern was summarized into four categories: type I, regression toward the lumen; type II, regression toward the invasive front; type III, concentric regression; and type IV, scattered regression. Inverse probability propensity score weighting was performed to minimize the influence of confounding factors.

RESULTS

After adjusting for baseline characteristics, the R0 resection rates (90.9% vs. 89.0%, P=0.302) and pCR (ypT0N0) rates (29.8% vs. 34.0%, P=0.167) were comparable between two groups. Patients receiving NCRT showed lower TRG score (P<0.001) and higher major pathological response (MPR) rate (64.7% vs. 53.6%, P=0.001) compared to those receiving NICT. However, NICT brought a higher pCR rate of metastatic LNs than conventional NCRT (53.9% vs. 37.1%, P<0.001). The rates of type I/II/III/IV regression patterns were 44.6%, 6.8%, 11.4% and 37.1% in the NICT group, 16.9%, 8.2%, 18.3% and 56.6% in the NCRT group, indicating a significant difference (P<0.001). Moreover, there were no significant differences in the incidence of total postoperative complications (35.8% vs. 39.9%, P=0.189) and 30-d mortality (0.0% vs. 1.1%, P=0.062).

CONCLUSION

For patients with locally advanced ESCC, NICT showed a R0 resection rate and pCR (ypT0N0) rate comparable to conventional NCRT, without increased incidence of postoperative complications and mortality. Notablely, NICT followed by surgery might bring a promising treatment response of metastatic LNs.