The effect of a constraint on the maximum number of controls matched to each treated subject on the performance of full matching on the propensity score when estimating risk differences

Association between early echocardiography screening of low systemic blood flow and intraventricular hemorrhage in preterm infants: a multicenter cohort study

OBJECTIVE

To determine whether early echocardiography screening of low systemic blood flow reduces intraventricular hemorrhage in preterm infants.

STUDY DESIGN

Prospective multicenter study in preterm infants below 33 weeks of gestational age at nine neonatal units. Five units performed early echocardiography screening for low systemic blood flow and guided clinical management (exposure group) and 4 units did not (control group). Our main outcome was ≥grade II intraventricular hemorrhage or death within the first 7 days of life. The main analysis used the inverse probability of treatment weighting.

RESULTS

Three hundred and thirty-two preterm infants (131 in the exposure group and 201 in the control group) were included. Exposure to early echocardiography screening was associated with a significant reduction in ≥grade II intraventricular hemorrhage or early death [odds ratio 0.285 (95% CI: 0.133-0.611); p = 0.001].

CONCLUSIONS

Early echocardiography screening for low systemic blood flow may reduce the incidence of intraventricular hemorrhage in preterm infants.

The iterative bisection procedure: a useful tool for determining parameter values in data-generating processes in Monte Carlo simulations

BACKGROUND

Data-generating processes are key to the design of Monte Carlo simulations. It is important for investigators to be able to simulate data with specific characteristics.

METHODS

We described an iterative bisection procedure that can be used to determine the numeric values of parameters of a data-generating process to produce simulated samples with specified characteristics. We illustrated the application of the procedure in four different scenarios: (i) simulating binary outcome data from a logistic model such that the prevalence of the outcome is equal to a specified value; (ii) simulating binary outcome data from a logistic model based on treatment status and baseline covariates so that the simulated outcomes have a specified treatment relative risk; (iii) simulating binary outcome data from a logistic model so that the model c-statistic has a specified value; (iv) simulating time-to-event outcome data from a Cox proportional hazards model so that treatment induces a specified marginal or population-average hazard ratio.

RESULTS

In each of the four scenarios the bisection procedure converged rapidly and identified parameter values that resulted in the simulated data having the desired characteristics.

CONCLUSION

An iterative bisection procedure can be used to identify numeric values for parameters in data-generating processes to generate data with specified characteristics.

Pre-processing data to reduce biases: full matching incorporating an instrumental variable in population-based studies

BACKGROUND

Epidemiologists are often concerned with unobserved biases that produce confounding in population-based studies. We introduce a new design approach-'full matching incorporating an instrumental variable (IV)' or 'Full-IV Matching'-and illustrate its utility in reducing observed and unobserved biases to increase inference accuracy. Our motivating example is tailored to a central question in humanitarian emergencies-the difference in sexual violence risk by displacement setting.

METHODS

We conducted a series of 1000 Monte Carlo simulations generated from a population-based survey after the 2010 Haitian earthquake and included earthquake damage severity as an IV and the unmeasured variable of 'social capital'. We compared standardized mean differences (SMDs) for covariates after different designs to understand potential biases. Mean risk differences (RDs) were used to assess each design's accuracy in estimating the oracle of the simulated data set.

RESULTS

Naive analysis and pair matching equivalently performed. Full matching reduced imbalances between exposed and comparison groups across covariates, except for the unobserved covariate of 'social capital'. Pair and full matching overstated differences in sexual violence risk when displaced to a camp vs a community [pair: RD = 0.13, 95% simulation interval (SI) 0.09-0.16; full: RD = 0.11, 95% SI 0.08-0.14). Full-IV Matching reduced imbalances across observed covariates and importantly 'social capital'. The estimated risk difference (RD = 0.07, 95% SI 0.03-0.11) was closest to the oracle (RD = 0.06, 95% SI 0.4-0.8).

CONCLUSION

Full-IV Matching is a novel approach that is promising for increasing inference accuracy when unmeasured sources of bias likely exist.

Relative effectiveness of the adjuvanted vs non-adjuvanted seasonal influenza vaccines against severe laboratory-confirmed influenza among hospitalized Italian older adults

OBJECTIVES

In this study, we aimed to investigate the relative vaccine effectiveness (rVE) of the MF59-adjuvanted trivalent (aTIV) and non-adjuvanted quadrivalent (QIVe) egg-based standard-dose vaccines against severe laboratory-confirmed influenza.

METHODS

This test-negative case-control study was conducted in a hospital setting during four recent Italian influenza seasons (from 2018/19 to 2021/22). The clinical outcome was severe acute respiratory infection (SARI) with laboratory confirmation diagnosed among subjects aged ≥65 years. rVE of aTIV versus QIVe was estimated through propensity score matching followed by logistic regression.

RESULTS

The influenza virus circulated to a significant extent only during the 2018/19 and 2019/20 seasons. The final population included 512 vaccinated older adults, of which 83 were cases and 429 were test-negative controls. aTIV and QIVe users differed substantially from the point of view of several baseline characteristics. The propensity score adjusted rVE of aTIV vs QIVe was 59.2% (95% CI: 14.6%, 80.5%), 54.7% (95% CI: -28.7%, 84.0%) and 56.9% (95% CI: -7.8%, 82.8%) against any influenza, A(H1N1)pdm09 and A(H3N2), respectively.

CONCLUSION

aTIV was more effective than QIVe in preventing laboratory-confirmed SARI. The benefits of aTIV may be obscured by confounding indication.

Body sway and movement strategies for control of postural stability in people with spinocerebellar ataxia type 3: A cross-sectional study

BACKGROUND

Postural instability with an excessive body sway is a disabling manifestation in spinocerebellar ataxia type 3. Whether the larger body sway reflects distinct movement strategies for postural control remains uncertain. This study compared the control of postural stability of people with spinocerebellar ataxia type 3 with healthy subjects using body sway and movement strategy analyses derived from bi- and three-dimensional posturography.

METHODS

Twenty-three patients (7 men, 16 women, 47 ± 11 years) and 102 healthy participants (34 men, 68 women; 44 ± 22 years) underwent posturography while standing with eyes open/closed tasks. Postural stability was assessed using elliptical area and average velocity of body sway. Spatial patterns (single-, double-, or multi-centered) were derived from the number of high-density regions in the three-dimensional statokinesigram.

FINDINGS

Repeated measures two-way analysis-of-variance showed a vision-by-group interaction effect for area (F_1,122 = 28.831, P < 0.001, η² = 0.037) and velocity (F_1,123 = 59.367, P < 0.001, η² = 0.073); sway area and velocity were higher in spinocerebellar ataxia type 3 and increased under eyes-closed condition, with a higher increase in the spinocerebellar ataxia type 3. A main effect for group (F_1,123 = 11.702, P < 0.001, η² = 0.061) but not vision (F_1,123 = 2.257, P = 0.136, η² = 0.005) was found for the number of high-density regions. Spatial patterns were different between groups under trials with eyes closed (χ²_2,125 = 7.46, P = 0.023) but not open (χ²_2,125 = 2.026, P = 0.363), with a shift from single- to double- or multi-centered spatial patterns.

INTERPRETATION

Compared to healthy subjects, a larger body displacement and velocity in spinocerebellar ataxia type 3, mainly under visual constraints, are not related to the predominance of either ankle or hip movement strategies.

Propensity Score Analysis with Partially Observed Baseline Covariates: A Practical Comparison of Methods for Handling Missing Data

(1) Background: Propensity score methods gained popularity in non-interventional clinical studies. As it may often occur in observational datasets, some values in baseline covariates are missing for some patients. The present study aims to compare the performances of popular statistical methods to deal with missing data in propensity score analysis. (2) Methods: Methods that account for missing data during the estimation process and methods based on the imputation of missing values, such as multiple imputations, were considered. The methods were applied on the dataset of an ongoing prospective registry for the treatment of unprotected left main coronary artery disease. The performances were assessed in terms of the overall balance of baseline covariates. (3) Results: Methods that explicitly deal with missing data were superior to classical complete case analysis. The best balance was observed when propensity scores were estimated with a method that accounts for missing data using a stochastic approximation of the expectation-maximization algorithm. (4) Conclusions: If missing at random mechanism is plausible, methods that use missing data to estimate propensity score or impute them should be preferred. Sensitivity analyses are encouraged to evaluate the implications methods used to handle missing data and estimate propensity score.

Matching Methods for Confounder Adjustment: An Addition to the Epidemiologist's Toolbox

Propensity score weighting and outcome regression are popular ways to adjust for observed confounders in epidemiological research. Here, we provide an introduction to matching methods, which serve the same purpose but can offer advantages in robustness and performance. A key difference between matching and weighting methods is that matching methods do not directly rely on the propensity score and so are less sensitive to its misspecification or to the presence of extreme values. Matching methods offer many options for customization, which allow a researcher to incorporate substantive knowledge and carefully manage bias/variance trade-offs in estimating the effects of nonrandomized exposures. We review these options and their implications, providing guidance for their use, and comparison with weighting methods. Because of their potential advantages over other methods, matching methods should have their place in an epidemiologist's methodological toolbox.

The effect of a constraint on the maximum number of controls matched to each treated subject on the performance of full matching on the propensity score when estimating risk differences

Many observational studies estimate causal effects using methods based on matching on the propensity score. Full matching on the propensity score is an effective and flexible method for utilizing all available data and for creating well-balanced treatment and control groups. An important component of the full matching algorithm is the decision about whether to impose a restriction on the maximum ratio of controls matched to each treated subject. Despite the possible effect of this restriction on subsequent inferences, this issue has not been examined. We used a series of Monte Carlo simulations to evaluate the effect of imposing a restriction on the maximum ratio of controls matched to each treated subject when estimating risk differences. We considered full matching both with and without a caliper restriction. When using full matching with a caliper restriction, the imposition of a subsequent constraint on the maximum ratio of the number of controls matched to each treated subject had no effect on the quality of inferences. However, when using full matching without a caliper restriction, the imposition of a constraint on the maximum ratio of the number of controls matched to each treated subject tended to result in an increase in bias in the estimated risk difference. However, this increase in bias tended to be accompanied by a corresponding decrease in the sampling variability of the estimated risk difference. We illustrate the consequences of these restrictions using observational data to estimate the effect of medication prescribing on survival following hospitalization for a heart attack.