Variance estimation when using propensity-score matching with replacement with survival or time-to-event outcomes

Minimizing confounding in comparative observational studies with time-to-event outcomes: An extensive comparison of covariate balancing methods using Monte Carlo simulation

Observational studies are frequently used in clinical research to estimate the effects of treatments or exposures on outcomes. To reduce the effects of confounding when estimating treatment effects, covariate balancing methods are frequently implemented. This study evaluated, using extensive Monte Carlo simulation, several methods of covariate balancing, and two methods for propensity score estimation, for estimating the average treatment effect on the treated using a hazard ratio from a Cox proportional hazards model. With respect to minimizing bias and maximizing accuracy (as measured by the mean square error) of the treatment effect, the average treatment effect on the treated weighting, fine stratification, and optimal full matching with a conventional logistic regression model for the propensity score performed best across all simulated conditions. Other methods performed well in specific circumstances, such as pair matching when sample sizes were large (n = 5000) and the proportion treated was < 0.25. Statistical power was generally higher for weighting methods than matching methods, and Type I error rates were at or below the nominal level for balancing methods with unbiased treatment effect estimates. There was also a decreasing effective sample size with an increasing number of strata, therefore for stratification-based weighting methods, it may be important to consider fewer strata. Generally, we recommend methods that performed well in our simulations, although the identification of methods that performed well is necessarily limited by the specific features of our simulation. The methods are illustrated using a real-world example comparing beta blockers and angiotensin-converting enzyme inhibitors among hypertensive patients at risk for incident stroke.

Propensity score matching for estimating a marginal hazard ratio

Propensity score matching is commonly used to draw causal inference from observational survival data. However, its asymptotic properties have yet to be established, and variance estimation is still open to debate. We derive the statistical properties of the propensity score matching estimator of the marginal causal hazard ratio based on matching with replacement and a fixed number of matches. We also propose a double-resampling technique for variance estimation that takes into account the uncertainty due to propensity score estimation prior to matching.

Real-world evidence of survival benefit of remdesivir: study of 419 propensity score-matched patients hospitalized over the alpha and delta waves of COVID-19 in New Orleans, LA

Background

The direct acting antiviral remdesivir (RDV) has shown promising results in randomized clinical trials. This study is a unique report of real clinical practice RDV administration for COVID-19 from alpha through delta variant circulation in New Orleans, Louisiana (NOLA). Patients in NOLA have among US worst pre-COVID health outcomes, and the region was an early epicenter for severe COVID.

Methods

Data were directly extracted from electronic medical records through REACHnet. Of 9,106 adults with COVID, 1,928 were admitted to inpatient care within 7 days of diagnosis. The propensity score is based upon 22 selected covariates, related to both RDV assignment and outcome of interest. RDV and non-RDV patients were matched 1:1 with replacement, by location and calendar period of admission. Primary and secondary endpoints were, death from any cause and inpatient discharge, within 28 and 14 days after inpatient admission.

Results

Of 448 patients treated with RDV, 419 (94%) were successfully matched to a non-RDV patient. 145 (35%) patients received RDV for < 5 days, 235 (56%) for 5 days, and 39 (9%) for > 5 days. 96% of those on RDV received it within 2 days of admission. RDV was more frequently prescribed in patients with pneumonia (standardized difference: 0.75), respiratory failure, hypoxemia, or dependence on supplemental oxygen (0.69), and obesity (0.35) within 5 days prior to RDV initiation or corresponding day in non-RDV patients (index day). RDV patients were numerically more likely to be on steroids within 5 days prior to index day (86 vs. 82%) and within 7 days after inpatient admission (96 vs. 87%). RDV was significantly associated with lower risk of death within 14 days after admission (hazard ratio [HR]: 0.37, 95% CI: 0.19 to 0.69, p = 0.002) but not within 28 days (HR: 0.62, 95% CI: 0.36 to 1.07, p = 0.08). Discharge within 14 days of admission was significantly more likely for RDV patients (p < 0.001) and numerically more likely within 28 days after admission (p = 0.06).

Conclusion

Overall, our findings support recommendation of RDV administration for COVID-19 in a highly comorbid, highly impoverished population representative of both Black and White subjects in the US Gulf South.

Interpersonal violent injury and subsequent risk of suicide and deliberate self-harm: a register-based national cohort study from Norway, 2010-2018

Background

Interpersonal violence is a leading cause of morbidity, with potentially severe adverse consequences for the mental health of the injured persons. The extent to which violent injury is associated with subsequent suicidal behavior, however, remains unclear. This study aimed to examine how violent injury was associated with subsequent deliberate self-harm and death by suicide.

Methods

This retrospective cohort study used nationwide longitudinal registry data from Norway to identify all individuals presenting to emergency services in 2010-2018 with a violence-related injury, along with sex- and age-matched control individuals from the general population. The primary outcomes were any emergency visit for deliberate self-harm (DSH) and suicide death, observed through 31 December 2018. Rates of each outcome were compared between violence-injured patients and comparison individuals using stratified multivariable Cox regression models, controlling for sociodemographic characteristics as well as history of psychiatric treatment and DSH. Secondary analyses tested for moderation by sex, age, and prior psychiatric treatment.

Findings

Violence-injured patients (n = 28,276) had substantially higher rates of DSH (946.7 per 100,000 person-years) and suicide death (74.5 per 100,000) when compared to controls (n = 282,760; 90.0 and 15.2 per 100,000, respectively). The hazard ratios (HRs) remained significantly higher even after accounting for covariates (HRadj for DSH: 5.11; 95% CI: 4.62, 5.66; HRadj for suicide: 2.40; 95% CI: 1.78, 3.24). Sex differences in this association were negligible, but the association between violence injury and DSH increased with age. Violence-injured patients with prior psychiatric treatment had the highest risk of suicidal behavior.

Interpretation

Violence-injured patients experience significantly excess rates of suicidal behavior, a finding with potential to inform both clinical intervention and population-level suicide prevention strategies.

Funding

Fulbright Norway Scholarship.

Comparing quality indicator rates for home care clients receiving palliative and end-of-life care before and during the Covid-19 pandemic

BACKGROUND

The consensus among Canadians with regards to end-of-life preferences is that with adequate support the majority prefer to live and die at home.

PURPOSE

To compare quality indicator (QI) rates for home care clients receiving palliative and end-of-life care prior to and after the onset of the COVID-19 pandemic.

METHODS

A retrospective population-based cohort design was used. Sixteen QIs informed by existing literature and a preliminary set of QIs recently evaluated by a modified Delphi panel were compared. Data were obtained from the interRAI Palliative Care instrument for Ontario home care clients for two separate cohorts: the pre-COVID (January 14, 2019 to March 16, 2020) and COVID cohort (March 17, 2020 to May 18, 2021). A propensity score analysis was used to match (using nearest neighbour matching) on 21 covariates, resulting in a sample size of 2479 unique interRAI Palliative Care assessments in each cohort. Alternative propensity score methods were explored as part of a sensitivity analysis.

RESULTS

After matching the pre-COVID and COVID cohorts, five of the 16 QIs had statistically significant differences in the QI rates (change from pre-COVID to COVID): decrease in prevalence of severe or excruciating daily pain (p = 0.03, effect size=-0.08), decrease in prevalence of caregiver distress (p = 0.02, effect size=-0.06), decrease in prevalence of negative mood (p = 0.003, effect size=- 0.17), decrease in prevalence of a delirium-like syndrome (p = 0.001, effect size=-0.25) and decrease in prevalence of nausea or vomiting (p = 0.04, effect size=-0.06). While the alternative propensity score methods produced slightly different results, no clinically meaningful differences were seen between the cohorts when effect sizes were examined. All methods were in agreement regarding the highest QI rates, which included the prevalence of shortness of breath with activity, no advance directives, and fatigue.

CONCLUSION

This study is the first to examine differences in QI rates for home care clients receiving palliative and end-of-life care before and during COVID in Ontario. It appears that QI rates did not change over the course of the pandemic in this population. Future work should be directed to understanding the temporal variation in these QI rates, risk-adjusting the QI rates for further comparison among jurisdictions, provinces, and countries, and in creating benchmarks for determining acceptable rates of different QIs.

RKHS-based covariate balancing for survival causal effect estimation

Survival causal effect estimation based on right-censored data is of key interest in both survival analysis and causal inference. Propensity score weighting is one of the most popular methods in the literature. However, since it involves the inverse of propensity score estimates, its practical performance may be very unstable, especially when the covariate overlap is limited between treatment and control groups. To address this problem, a covariate balancing method is developed in this paper to estimate the counterfactual survival function. The proposed method is nonparametric and balances covariates in a reproducing kernel Hilbert space (RKHS) via weights that are counterparts of inverse propensity scores. The uniform rate of convergence for the proposed estimator is shown to be the same as that for the classical Kaplan-Meier estimator. The appealing practical performance of the proposed method is demonstrated by a simulation study as well as two real data applications to study the causal effect of smoking on survival time of stroke patients and that of endotoxin on survival time for female patients with lung cancer respectively.

Impact of the matching algorithm on the treatment effect estimate: A neutral comparison study

Propensity score matching is increasingly being used in the medical literature. Choice of matching algorithms, reporting quality, and estimands are oftentimes not discussed. We evaluated the impact of propensity score matching algorithms, based on a recent clinical dataset, with three commonly used outcomes. The resulting estimands for different strengths of treatment effects were compared in a neutral comparison study and based on a thoroughly designed simulation study. Different algorithms yielded different levels of balance after matching. Along with full matching and genetic matching with replacement, good balance was achieved with nearest neighbor matching with caliper but thereby more than one fifth of the treated units were discarded. Average marginal treatment effect estimates were least biased with genetic or nearest neighbor matching, both with replacement and full matching. Double adjustment yielded conditional treatment effects that were closer to the true values, throughout. The choice of the matching algorithm had an impact on covariate balance after matching as well as treatment effect estimates. In comparison, genetic matching with replacement yielded better covariate balance than all other matching algorithms. A literature review in the British Medical Journal including its subjournals revealed frequent use of propensity score matching; however, the use of different matching algorithms before treatment effect estimation was only reported in one out of 21 studies. Propensity score matching is a methodology for causal treatment effect estimation from observational data; however, the methodological difficulties and low reporting quality in applied medical research need to be addressed.

Disability specific equivalence scales: a case-control approach applied to the cost of acquired brain injuries

This study estimates the household costs resulting from acquired brain injuries in terms of a reduction in the standard of living. The application uses primary data collected in the Verona and Florence provinces of Italy integrating highly detailed health information with information about consumption, income, wealth, time-use and relational well-being describing the standard of living. In general, the estimates of disability costs in previous studies are obtained from survey data without a specific focus on individuals with disabilities but collect information on the general health status. In contrast, this study exploits highly detailed information on a sample of "cases" with a disability, whose intensity is measured by the highly precise European quality of life measure-5 domain-5 (EQ-5D) instrument, to be compared with a sample of "control" formed by households without disabled members. The disability scales have been estimated using a Structural Equation Modelling (SEM) based procedure. We then implement interpersonal comparisons on the health income dimension in a theoretically plausible way, testing the independence hypothesis of equivalence scales. Our study finds that on average disabled households need an additional amount of about €1800 per month to reach the same standard of livings as the non-disabled households corresponding to a scale of 1.78.

Effectiveness of Nirmatrelvir-Ritonavir Against the Development of Post-COVID-19 Conditions Among U.S. Veterans : A Target Trial Emulation

BACKGROUND

COVID-19 has been linked to the development of many post-COVID-19 conditions (PCCs) after acute infection. Limited information is available on the effectiveness of oral antivirals used to treat acute COVID-19 in preventing the development of PCCs.

OBJECTIVE

To measure the effectiveness of outpatient treatment of COVID-19 with nirmatrelvir-ritonavir in preventing PCCs.

DESIGN

Retrospective target trial emulation study comparing matched cohorts receiving nirmatrelvir-ritonavir versus no treatment.

SETTING

Veterans Health Administration (VHA).

PARTICIPANTS

Nonhospitalized veterans in VHA care who were at risk for severe COVID-19 and tested positive for SARS-CoV-2 during January through July 2022.

INTERVENTION

Nirmatrelvir-ritonavir treatment for acute COVID-19.

MEASUREMENTS

Cumulative incidence of 31 potential PCCs at 31 to 180 days after treatment or a matched index date, including cardiac, pulmonary, renal, thromboembolic, gastrointestinal, neurologic, mental health, musculoskeletal, endocrine, and general conditions and symptoms.

RESULTS

Eighty-six percent of the participants were male, with a median age of 66 years, and 17.5% were unvaccinated. Baseline characteristics were well balanced between participants treated with nirmatrelvir-ritonavir and matched untreated comparators. No differences were observed between participants treated with nirmatrelvir-ritonavir (n = 9593) and their matched untreated comparators in the incidence of most PCCs examined individually or grouped by organ system, except for lower combined risk for venous thromboembolism and pulmonary embolism (subhazard ratio, 0.65 [95% CI, 0.44 to 0.97]; cumulative incidence difference, -0.29 percentage points [CI, -0.52 to -0.05 percentage points]).

LIMITATIONS

Ascertainment of PCCs using International Classification of Diseases, 10th Revision, codes may be inaccurate. Evaluation of many outcomes could have resulted in spurious associations with combined thromboembolic events by chance.

CONCLUSION

Out of 31 potential PCCs, only combined thromboembolic events seemed to be reduced by nirmatrelvir-ritonavir.

PRIMARY FUNDING SOURCE

U.S. Department of Veterans Affairs.

How to perform prespecified subgroup analyses when using propensity score methods in the case of imbalanced subgroups

BACKGROUND

Looking for treatment-by-subset interaction on a right-censored outcome based on observational data using propensity-score (PS) modeling is of interest. However, there are still issues regarding its implementation, notably when the subsets are very imbalanced in terms of prognostic features and treatment prevalence.

METHODS

We conducted a simulation study to compare two main PS estimation strategies, performed either once on the whole sample ("across subset") or in each subset separately ("within subsets"). Several PS models and estimands are also investigated. We then illustrated those approaches on the motivating example, namely, evaluating the benefits of facial nerve resection in patients with parotid cancer in contact with the nerve, according to pretreatment facial palsy.

RESULTS

Our simulation study demonstrated that both strategies provide close results in terms of bias and variance of the estimated treatment effect, with a slight advantage for the "across subsets" strategy in very small samples, provided that interaction terms between the subset variable and other covariates influencing the choice of treatment are incorporated. PS matching without replacement resulted in biased estimates and should be avoided in the case of very imbalanced subsets.

CONCLUSIONS

When assessing heterogeneity in the treatment effect in small samples, the "across subsets" strategy of PS estimation is preferred. Then, either a PS matching with replacement or a weighting method must be used to estimate the average treatment effect in the treated or in the overlap population. In contrast, PS matching without replacement should be avoided in this setting.

Intraoperative hypotension in noncardiac surgery patients with chronic beta-blocker therapy: A matched cohort analysis

STUDY OBJECTIVE

To explore the incidence of intraoperative hypotension in patients with chronic beta-blocker therapy, expressed as time spent, area and time-weighted average under predefined mean arterial pressure thresholds.

DESIGN

Retrospective analysis of a prospective observational cohort registry.

SETTING

Patients ≥60 years undergoing intermediate- to high-risk noncardiac surgery with routine postoperative troponin measurements on the first three days after surgery.

PATIENTS

1468 matched sets of patients (1:1 ratio with replacement) with and without chronic beta-blocker treatment.

INTERVENTIONS

None.

MEASUREMENTS

The primary outcome was the exposure to intraoperative hypotension in beta-blocker users vs. non-users. Time spent, area and time-weighted average under predefined mean arterial pressure thresholds (55-75 mmHg) were calculated to express the duration and severity of exposure. Secondary outcomes included incidence of postoperative myocardial injury and thirty-day mortality, myocardial infarction (MI) and stroke. Furthermore, analyses for patient subgroup and beta-blocker subtype were conducted.

MAIN RESULTS

In patients with chronic beta-blocker therapy, no increased exposure to intraoperative hypotension was observed for all characteristics and thresholds calculated (all P > .05). Beta-blocker users had lower heart rate before, during and after surgery (70 vs. 74, 61 vs. 65 and 68 vs. 74 bpm, all P < .001, respectively). Postoperative myocardial injury (13.6% vs. 11.6%, P = .269) and thirty-day mortality (2.5% vs. 1.4%, P = .055), MI (1.4% vs. 1.5%, P = .944) and stroke (1.0% vs 0.7%, P =  .474) rates were comparable. The results were consistent in subtype and subgroup analyses.

CONCLUSIONS

In this matched cohort analysis, chronic beta-blocker therapy was not associated with increased exposure to intraoperative hypotension in patients undergoing intermediate- to high-risk noncardiac surgery. Furthermore, differences in patient subgroups and postoperative adverse cardiovascular events as a function of treatment regimen could not be demonstrated.

Economic evaluation of a multimorbidity patient centered care model implemented in the Chilean public health system

Multimorbidity and patient-centered care approaches are growing challenges for health systems and patients. The cost of multimorbidity patients and the transition to a new care strategy is still sightly explored. In Chile, more than 70% of the adult population suffer from multimorbidity, opening an opportunity to implement a Multimorbidity patient-centered care model. The objective of this study was to perform an economic evaluation of the model from the public health system perspective.The methodology used a cost-consequence evaluation comparing seven exposed with seven unexposed primary care centers, and their reference hospitals. It followed three steps. First, we performed a Time-Driven Activity-Based Costing with routinely collected data routinely collected. Second, we run a comparative analysis through a propensity score matching and an estimation of the attributable costs to health services utilization at primary, secondary and tertiary care and health outcomes. Third, we estimated implementation and transaction costs.Results showed savings in aggregate costs of the total population (-0.12 (0.03) p < 0.01) during the period under evaluation. Costs in primary care showed a significant increase, whereas tertiary care showed significant savings. Health outcomes were associated with higher survival in patients under the new care model (HR 0.70 (0.05) p < 0.01). Implementation and transaction costs increased as the number of pilot intervention centers increased, and they represented 0,07% of the total annual budget of the Servicio de Salud Metropolitano Sur Oriente. After three years of piloting, the implementation and transaction cost for the total period was USD 1,838,767 and 393,775, respectively.The study's findings confirm the purpose of the new model to place primary health care at the center of care for people with non-communicable chronic diseases. Thus, it is necessary to consider implementation and transaction costs to introduce a broad health system multimorbidity approach. The health system should assume some of them permanently to guarantee sustainability and facilitate scale-up.

Effect of extracurricular tutoring on adolescent students cognitive ability: A propensity score matching analysis

In recent years, there has been a substantial increase in the number of adolescent students attending extracurricular tutoring. However, extracurricular tutoring, being a distinct form of education, may have varying effects on cognitive capabilities compared to conventional education. Accordingly, the purpose of this paper is to conduct a thorough examination of the effects of extracurricular tutoring on cognitive abilities among adolescent students. The study utilized national data from the China Family Panel Study 2018 to explore the relationship between involvement in extracurricular tutoring among students and cognitive abilities. The study included 2567 valid respondents. A binary logistic regression model was built to determine the factors associated with tutoring involvement while controlling for 19 individual, family, and school factors. Furthermore, a propensity score-matched analysis was conducted in order to mitigate potential bias by using confounding variables identified in the previous step. The study results show that participation in extracurricular tutoring can significantly increase the level of cognitive ability of adolescent students, with adjustments made for age, gender, ethnicity, number of family members, net family income per capita, education and training expenditure in the past years, change of residence for enrollment, change of domicile address for enrollment, locality of the current school, class size, hold a position as a class cadre, average daily study time on weekdays, average study time per day during weekends. The findings imply that the government should provide tutorial subsidies to disadvantaged groups of adolescent students, allocate educational resources equitably, and invest more in education resources in less developed regions to foster fair and healthy development of education and improve the cognitive abilities of young students in the long-term.

Comparative Effectiveness of Rituximab- Versus Cyclophosphamide-Based Remission Induction Strategies in Antineutrophil Cytoplasmic Antibody-Associated Vasculitis for the Risk of Kidney Failure and Mortality

OBJECTIVE

To compare rituximab- versus cyclophosphamide-based remission induction strategies for the long-term risks of kidney failure and death in antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) in a real-world cohort.

METHODS

We performed a cohort study using the Mass General Brigham AAV Cohort, which includes proteinase 3-ANCA+ and myeloperoxidase (MPO)-ANCA+ AAV patients diagnosed from January 1, 2002 to December 31, 2019. We included cases in which the initial remission induction strategy was based either on rituximab or cyclophosphamide. The primary outcome was the composite outcome of kidney failure or death. We used multivariable Cox proportional hazards models and propensity score-matched analyses to assess the association of rituximab- versus cyclophosphamide-based treatment strategies with the composite outcome of kidney failure or death.

RESULTS

Of 595 included patients, 352 patients (~60%) received rituximab-based and 243 patients (~40%) received cyclophosphamide-based regimens. The mean age was 61 years, 58% of patients were female, 70% of patients were MPO-ANCA+, and 69% of patients had renal involvement (median estimated glomerular filtration rate 37.3 ml/minute/1.73 m2 ). There were 133 events at 5 years, and the incidence rates in rituximab- and cyclophosphamide-based regimens were 6.8 and 6.1 per 100 person-years, respectively. The risk of kidney failure or death was similar in both groups in multivariable-adjusted analyses (hazard ratio [HR] 1.03 [95% confidence interval (95% CI) 0.55-1.93]) and in propensity score-matched analyses (HR 1.05 [95% CI 0.55-1.99]) at 5 years. Our findings were similar when outcomes were assessed at 1 and 2 years as well as in subgroups stratified according to renal involvement and severity as well as major organ involvement.

CONCLUSION

Rituximab- and cyclophosphamide-based remission induction strategies for AAV are associated with similar risks of kidney failure and death.

A Large-Scale Multicenter Retrospective Study on Nephrotoxicity Associated With Empiric Broad-Spectrum Antibiotics in Critically Ill Patients

BACKGROUND

Evidence regarding acute kidney injury associated with concomitant administration of vancomycin and piperacillin-tazobactam is conflicting, particularly in patients in the ICU.

RESEARCH QUESTION

Does a difference exist in the association between commonly prescribed empiric antibiotics on ICU admission (vancomycin and piperacillin-tazobactam, vancomycin and cefepime, and vancomycin and meropenem) and acute kidney injury?

STUDY DESIGN AND METHODS

This was a retrospective cohort study using data from the eICU Research Institute, which contains records for ICU stays between 2010 and 2015 across 335 hospitals. Patients were enrolled if they received vancomycin and piperacillin-tazobactam, vancomycin and cefepime, or vancomycin and meropenem exclusively. Patients initially admitted to the ED were included. Patients with hospital stay duration of < 1 h, receiving dialysis, or with missing data were excluded. Acute kidney injury was defined as Kidney Disease: Improving Global Outcomes stage 2 or 3 based on serum creatinine component. Propensity score matching was used to match patients in the control (vancomycin and meropenem or vancomycin and cefepime) and treatment (vancomycin and piperacillin-tazobactam) groups, and ORs were calculated. Sensitivity analyses were performed to study the effect of longer courses of combination therapy and patients with renal insufficiency on admission.

RESULTS

Thirty-five thousand six hundred fifty-four patients met inclusion criteria (vancomycin and piperacillin-tazobactam, n = 27,459; vancomycin and cefepime, n = 6,371; vancomycin and meropenem, n = 1,824). Vancomycin and piperacillin-tazobactam was associated with a higher risk of acute kidney injury and initiation of dialysis when compared with that of both vancomycin and cefepime (Acute kidney injury: OR, 1.37 [95% CI, 1.25-1.49]; dialysis: OR, 1.28 [95% CI, 1.14-1.45]) and vancomycin and meropenem (Acute kidney injury: OR, 1.27 [95%, 1.06-1.52]; dialysis: OR, 1.56 [95% CI, 1.23-2.00]). The odds of acute kidney injury developing was especially pronounced in patients without renal insufficiency receiving a longer duration of vancomycin and piperacillin-tazobactam therapy compared with vancomycin and meropenem therapy.

INTERPRETATION

VPT is associated with a higher risk of acute kidney injury than both vancomycin and cefepime and vancomycin and meropenem in patients in the ICU, especially for patients with normal initial kidney function requiring longer durations of therapy. Clinicians should consider vancomycin and meropenem or vancomycin and cefepime to reduce the risk of nephrotoxicity for patients in the ICU.

HOSPITAL VOLUME, POSTOPERATIVE MORTALITY, AND COSTS AFTER GASTRECTOMY FOR GASTRIC CANCER IN COLOMBIA: IS THERE ANY ASSOCIATION?

BACKGROUND

There are no information in the literature associating the volume of gastrectomies with survival and costs for the health system in the treatment of patients with gastric cancer in Colombia.

AIMS

The aim of this study was to analyze how gastrectomy for gastric cancer is associated with hospital volume, 30-day and 180-day postoperative mortality, and healthcare costs in Bogotá, Colombia.

METHODS

A retrospective cohort study based on hospital data of all adult patients with gastric cancer who underwent gastrectomy between 2014 and 2016 using a paired propensity score. The surgical volume was identified as the average annual number of gastrectomies performed by the hospital.

RESULTS

A total of 743 patients were included in the study. Hospital mortality at 30 and 180 days postoperatively was 36 (4.85%) and 127 (17.09%) patients, respectively. The average health care cost was USD 3,200. A total of 26 or more surgeries were determined to be the high surgical volume cutoff. Patients operated on in hospitals with a high surgical volume had lower 6-month mortality (HR 0.44; 95%CI 0.27-0.71; p=0.001), and no differences were found in health costs (mean difference 398.38; 95%CI-418.93-1,215.69; p=0.339).

CONCLUSIONS

This study concluded that in Bogotá (Colombia), surgery in a high-volume hospital is associated with better 6-month survival and no additional costs to the health system.

Hybrid immunity against reinfection with SARS-CoV-2 following a previous SARS-CoV-2 infection and single dose of the BNT162b2 vaccine in children and adolescents: a target trial emulation

BACKGROUND

Although most children and adolescents have had a previous SARS-CoV-2 infection and many continue to receive COVID-19 vaccinations, studies of the effectiveness of hybrid immunity against reinfection with the omicron (B.1.1.529) variant are scarce. We aimed to examine the effectiveness of vaccination in convalescent children and adolescents against reinfection with the delta (B.1.617.2) variant and the BA.1 and BA.2 and BA.4 and BA.5 omicron subvariants.

METHODS

This retrospective cohort study was devised to emulate a target randomised control trial using a retrospective dataset of anonymised health records of children (5-11 years old) and adolescents (12-16 years old) who were members of the Maccabi Healthcare Services, Israel. The design emulated 91 randomised trials by devising a series of multiple nested trials, compiling the results into a single dataset, and fitting Cox proportional hazards models to estimate adjusted hazard ratios (HRs) with 95% CIs of each measured outcome. The primary aim was to assess the protection from reinfection with the delta variant and the BA.1 and BA.2 and BA.4 and BA.5 omicron subvariants associated with hybrid immunity as a result of a previous SARS-CoV-2 infection followed by vaccination with the BNT162b2 (Pfizer-BioNTech) vaccine.

FINDINGS

Data from between from March 1, 2020, to July 31, 2022, for 163 812 individuals (120 721 children [59 404 girls and 61 317 boys], median age 8·0 years [IQR 6·7 to 10·2]; and 43 091 adolescents [21 239 girls and 21 852 boys], median age 13·5 years [12·6 to 14·8]) were included in at least one trial. A single dose of the BNT162b2 vaccine in convalescent children and adolescents confers statistically significant protection against the delta variant (78% [95% CI 72 to 83] in adolescents and 64% [3 to 87] in children) and the omicron BA.1 and BA.2 subvariants (54% [50 to 57] in adolescents and 71% [67 to 73] in children) compared with children who had a previous infection but were unvaccinated. However, the vaccine was not found to confer statistically significant protection against the BA.4 and BA.5 omicron subvariants in adolescents (8% [-18 to 29]) and children (12% [-6 to 27]).

INTERPRETATION

Decision makers in BA.4 and BA.5 dominant regions should re-examine whether convalescent individuals aged 5-16 years should receive the BNT162b2 vaccine to prevent future reinfection, especially in light of reports that show that most children and adolescents have already been infected with SARS-CoV-2.

FUNDING

None.

Allogeneic transplantation in advanced cutaneous T-cell lymphomas (CUTALLO): a propensity score matched controlled prospective study

BACKGROUND

Advanced-stage cutaneous T-cell lymphomas (CTCLs) are rare, usually refractory, and fatal diseases. Case series have suggested that allogeneic haematopoietic stem cell transplantation (HSCT) might improve the prognosis of advanced-stage CTCLs. The objective of this study was to investigate the effect of allogeneic HSCT compared with non-HSCT therapy on the outcome of individuals with advanced-stage CTCLs.

METHODS

In this prospective, multicentre, matched controlled trial, conducted at 30 hospitals, participants with advanced CTCLs were allocated treatment: if they had an available compatible related donor they were assigned to allogeneic HSCT, or if not they were allocated to non-allogeneic HSCT therapy. Key inclusion criteria were participants aged 18-70 years, with advanced stage mycosis fungoides or Sézary syndrome, and at least one poor prognostic criteria. Participants were excluded if they were not in complete or partial remission of the disease. Propensity score 1:1 matching with replacement (ie, that each participant treated with HSCT was matched to the participant with the closest propensity score treated with non-HSCT therapy, even if they had already been matched) was used to handle confounding factors, with the balance of covariate distribution between HSCT and non-HSCT groups assessed using standardised mean differences. The primary endpoint was progression-free survival in the matched intention-to-treat population. This trial is registered with ClinicalTrials.gov (NCT02520908), and is currently active but not recruiting.

FINDINGS

From June 1, 2016, to March 3, 2022, total of 99 participants were enrolled at 17 centres in France. Participants with a sibling or matched unrelated donor were assigned to allogeneic HSCT (HSCT group, n=55 [56%]) and participants without a donor were assigned to non-allogeneic HSCT treatment (non-HSCT group, n=44 [44%]). The median follow-up among survivors was 12·6 months (IQR 11·0-35·2). In the HSCT group, 51 participants (93%) were 1:1 matched to participants from the non-HSCT group. In the intention-to-treat analysis, median progression-free survival was significantly longer in the HSCT group (9·0 months [95% CI 6·6-30·5]) than in the non-HSCT group (3·0 months [2·0-6·3]), with a hazard ratio of 0·38 (95% CI 0·21-0·69; p<0·0001). In the per-protocol population, 40 participants (78%) in the HSCT group had 101 serious events and 29 participants (67%) in the non-HSCT group had 70 serious adverse events. The most common serious adverse event other than graft-versus-host disease in both groups was infections, occurring in 30 participants (59%) in the HSCT group and in 19 participants (44%) in the non-HSCT group.

INTERPRETATION

Allogeneic HSCT was associated with significantly longer progression-free survival in participants with advanced-stage CTCLs. These results indicate that allogeneic HSCT treatment should be made available to individuals with high-risk, advanced-stage mycosis fungoides or Sézary syndrome who achieve pre-transplant disease remission.

FUNDING

French Ministry of Health, National Cancer Institute, Programme Hospitalier de Recherche Clinique en Cancérologie.

Effectiveness of COVID-19 Treatment With Nirmatrelvir-Ritonavir or Molnupiravir Among U.S. Veterans: Target Trial Emulation Studies With One-Month and Six-Month Outcomes

BACKGROUND

Information about the effectiveness of oral antivirals in preventing short- and long-term COVID-19-related outcomes in the setting of Omicron variant transmission and COVID-19 vaccination is limited.

OBJECTIVE

To measure the effectiveness of nirmatrelvir-ritonavir and molnupiravir for outpatient treatment of COVID-19.

DESIGN

Three retrospective target trial emulation studies comparing matched cohorts of nirmatrelvir-ritonavir versus no treatment, molnupiravir versus no treatment, and nirmatrelvir-ritonavir versus molnupiravir.

SETTING

Veterans Health Administration (VHA).

PARTICIPANTS

Nonhospitalized veterans in VHA care who were at risk for severe COVID-19 and tested positive for SARS-CoV-2 during January through July 2022.

INTERVENTION

Nirmatrelvir-ritonavir or molnupiravir pharmacotherapy.

MEASUREMENTS

Incidence of any hospitalization or all-cause mortality at 30 days and from 31 to 180 days.

RESULTS

Eighty-seven percent of participants were male; the median age was 66 years, and 18% were unvaccinated. Compared with matched untreated control participants, those treated with nirmatrelvir-ritonavir (n = 9607) had lower 30-day risk for hospitalization (22.07 vs. 30.32 per 1000 participants; risk difference [RD], -8.25 [95% CI, -12.27 to -4.23] per 1000 participants) and death (1.25 vs. 5.47 per 1000 participants; RD, -4.22 [CI, -5.45 to -3.00] per 1000 participants). Among persons alive at day 31, reductions were seen in 31- to 180-day incidence of death (hazard ratio, 0.66 [CI, 0.49 to 0.89]) but not hospitalization (subhazard ratio, 0.90 [CI, 0.79 to 1.02]). Molnupiravir-treated participants (n = 3504) had lower 30-day and 31- to 180-day risks for death (3.14 vs. 13.56 per 1000 participants at 30 days; RD, -10.42 [CI, -13.49 to -7.35] per 1000 participants; hazard ratio at 31 to 180 days, 0.67 [CI, 0.48 to 0.95]) but not hospitalization. A difference in 30-day or 31- to 180-day risk for hospitalization or death was not observed between matched nirmatrelvir- or molnupiravir-treated participants.

LIMITATION

The date of COVID-19 symptom onset for most veterans was unknown.

CONCLUSION

Nirmatrelvir-ritonavir was effective in reducing 30-day hospitalization and death. Molnupiravir was associated with a benefit for 30-day mortality but not hospitalization. Further reductions in mortality from 31 to 180 days were observed with both antivirals.

PRIMARY FUNDING SOURCE

U.S. Department of Veterans Affairs.

The mental health and wellbeing impact of a Community Wealth Building programme in England: a difference-in-differences study

BACKGROUND

Wide differences in health exist between places in the UK, underscored by economic inequalities. Preston, an economically disadvantaged city in England, implemented a new approach to economic development, known as the Community Wealth Building programme. Public and non-profit organisations modified their procurement policies to support the development of local supply chains, improve employment conditions, and increase socially productive use of wealth and assets. We aimed to investigate the effect of this programme on population mental health and wellbeing.

METHODS

Difference-in-differences techniques compared trends in mental health outcomes in Preston, relative to matched control areas before (2011-15) and after (2016-19) the introduction of the programme. Outcomes were antidepressant prescribing, prevalence of depression, and mental health related hospital attendance rates using data provided by National Health Service Digital, the Quality and Outcomes Framework, and the Office for National Statistics. Additional analysis compared local authority measures of life satisfaction, median wages, and employment with synthetic counterfactuals created using Bayesian Structural Time Series.

FINDINGS

The introduction of the Community Wealth Building programme was associated with reductions in the prescribing of antidepressants (1·3 average daily quantities per person [95% CI 0·72-1·78) and prevalence of depression (2·4 per 1000 population [0·42-4·46]), relative to the control areas. The local population also experienced a 9% improvement in life satisfaction (95% credible interval 0-19·6%) and 11% increase in median wages (1·8-18·9%), relative to expected trends. Associations with employment and mental health related hospital attendance outcomes did not reach statistical significance.

INTERPRETATION

During the period in which the Community Wealth Building programme was introduced, there were fewer mental health problems than would have been expected compared with other similar areas, as life satisfaction and economic measures improved. This approach potentially provides an effective model for economic regeneration potentially leading to substantial health benefits.

FUNDING

National Institute for Health Research.

Denosumab and incidence of type 2 diabetes among adults with osteoporosis: population based cohort study

OBJECTIVE

To estimate the effect of denosumab compared with oral bisphosphonates on reducing the risk of type 2 diabetes in adults with osteoporosis.

DESIGN

Population based study involving emulation of a randomized target trial using electronic health records.

SETTING

IQVIA Medical Research Data primary care database in the United Kingdom, 1995-2021.

PARTICIPANTS

Adults aged 45 years or older who used denosumab or an oral bisphosphonate for osteoporosis.

MAIN OUTCOME MEASURES

The primary outcome was incident type 2 diabetes, as defined by diagnostic codes. Cox proportional hazards models were used to estimate adjusted hazard ratios and 95% confidence intervals, comparing denosumab with oral bisphosphonates using an as treated approach.

RESULTS

4301 new users of denosumab were matched on propensity score to 21 038 users of an oral bisphosphonate and followed for a mean of 2.2 years. The incidence rate of type 2 diabetes in denosumab users was 5.7 (95% confidence interval 4.3 to 7.3) per 1000 person years and in oral bisphosphonate users was 8.3 (7.4 to 9.2) per 1000 person years. Initiation of denosumab was associated with a reduced risk of type 2 diabetes (hazard ratio 0.68, 95% confidence interval 0.52 to 0.89). Participants with prediabetes appeared to benefit more from denosumab compared with an oral bisphosphonate (hazard ratio 0.54, 0.35 to 0.82), as did those with a body mass index ≥30 (0.65, 0.40 to 1.06).

CONCLUSIONS

In this population based study, denosumab use was associated with a lower risk of incident type 2 diabetes compared with oral bisphosphonate use in adults with osteoporosis. This study provides evidence at a population level that denosumab may have added benefits for glucose metabolism compared with oral bisphosphonates.

Long-Term Survival of Multiple Versus Single Arterial Coronary Bypass Grafting in Elderly Patients

Multiple arterial grafting (MAG) utilizes more than one arterial graft with any additional grafts being saphenous vein grafts (SVG). It remains an infrequently used coronary surgical revascularization technique, especially in elderly patients. Our study aims to evaluate the age-related association with the relative outcomes of multiple versus single arterial grafting (SAG). The Australian and New Zealand national registry was used to identify adult patients undergoing primary isolated CABG with at least two grafts. Exclusion criteria included reoperations, concomitant or previous cardiac surgery, and the absence of arterial grafting. Propensity score matching was used to match patient groups. The primary outcome was all-cause late mortality and the secondary outcomes were 30-day mortality and 30-day hospital readmission. We selected 69,624 eligible patients with a mean (standard deviation) age of 65.0 (10.2) years old. Matching between MAG and SAG generated 16,882 pairs of patients < 70 years old and 10,921 pairs of patients ≥ 70 years old. At a median [interquartile range] follow-up duration of 5.9 [3.2-9.6] years, MAG was associated with significantly reduced mortality compared to SAG (hazard ratio [HR], 0.73; 95% confidence interval [CI], 0.68-0.78; p < 0.001) in the younger subgroup as well as the elderly subgroup (HR, 0.84; 95% CI, 0.79-0.88; p < 0.001). In conclusion, MAG offers a survival benefit over SAG, in both younger and elderly patients.

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.

Effectiveness of COVID-19 treatment with nirmatrelvir-ritonavir or molnupiravir among U.S. Veterans: target trial emulation studies with one-month and six-month outcomes

Background

Information about the effectiveness of oral antivirals in preventing short- and long-term COVID-19-related outcomes during the Omicron surge is limited. We sought to determine the effectiveness of nirmatrelvir-ritonavir and molnupiravir for the outpatient treatment of COVID-19.

Methods

We conducted three retrospective target trial emulation studies comparing matched patient cohorts who received nirmatrelvir-ritonavir versus no treatment, molnupiravir versus no treatment, and nirmatrelvir-ritonavir versus molnupiravir in the Veterans Health Administration (VHA). Participants were Veterans in VHA care at risk for severe COVID-19 who tested positive for SARS-CoV-2 in the outpatient setting during January and February 2022. Primary outcomes included all-cause 30-day hospitalization or death and 31-180-day incidence of acute or long-term care admission, death, or post-COVID-19 conditions. For 30-day outcomes, we calculated unadjusted risk rates, risk differences, and risk ratios. For 31-180-day outcomes, we used unadjusted time-to-event analyses.

Results

Participants were 90% male with median age 67 years and 26% unvaccinated. Compared to matched untreated controls, nirmatrelvir-ritonavir-treated participants (N=1,587) had a lower 30-day risk of hospitalization (27.10/1000 versus 41.06/1000, risk difference [RD] - 13.97, 95% CI -23.85 to -4.09) and death (3.15/1000 versus 14.86/1000, RD -11.71, 95% CI - 16.07 to -7.35). Among persons who were alive at day 31, further significant reductions in 31-180-day incidence of hospitalization (sub-hazard ratio 1.07, 95% CI 0.83 to 1.37) or death (hazard ratio 0.61, 95% CI 0.35 to 1.08) were not observed. Molnupiravir-treated participants aged ≥65 years (n=543) had a lower combined 30-day risk of hospitalization or death (55.25/1000 versus 82.35/1000, RD -27.10, 95% CI -50.63 to -3.58). A statistically significant difference in 30-day or 31-180-day risk of hospitalization or death was not observed between matched nirmatrelvir- or molnupiravir-treated participants. Incidence of most post-COVID conditions was similar across comparison groups.

Conclusions

Nirmatrelvir-ritonavir was highly effective in preventing 30-day hospitalization and death. Short-term benefit from molnupiravir was observed in older groups. Significant reductions in adverse outcomes from 31-180 days were not observed with either antiviral.

Effectiveness of mRNA COVID-19 Vaccine Boosters Against Infection, Hospitalization, and Death: A Target Trial Emulation in the Omicron (B.1.1.529) Variant Era

BACKGROUND

The effectiveness of a third mRNA COVID-19 vaccine dose (booster dose) against the Omicron (B.1.1.529) variant is uncertain, especially in older, high-risk populations.

OBJECTIVE

To determine mRNA booster vaccine effectiveness (VE) against SARS-CoV-2 infection, hospitalization, and death in the Omicron era by booster type, primary vaccine type, time since primary vaccination, age, and comorbidity burden.

DESIGN

Retrospective matched cohort study designed to emulate a target trial of booster vaccination versus no booster, conducted from 1 December 2021 to 31 March 2022.

SETTING

U.S. Department of Veterans Affairs health care system.

PARTICIPANTS

Persons who had received 2 mRNA COVID-19 vaccine doses at least 5 months earlier.

INTERVENTION

Booster monovalent mRNA vaccination (Pfizer-BioNTech's BNT162b2 or Moderna's mRNA-1273) versus no booster.

MEASUREMENTS

Booster VE.

RESULTS

Each group included 490 838 well-matched persons, who were predominantly male (88%), had a mean age of 63.0 years (SD, 14.0), and were followed for up to 121 days (mean, 79.8 days). Booster VE more than 10 days after a booster dose was 42.3% (95% CI, 40.6% to 43.9%) against SARS-CoV-2 infection, 53.3% (CI, 48.1% to 58.0%) against SARS-CoV-2-related hospitalization, and 79.1% (CI, 71.2% to 84.9%) against SARS-CoV-2-related death. Booster VE was similar for different booster types (BNT162b2 or mRNA-1273), age groups, and primary vaccination regimens but was significantly higher with longer time since primary vaccination and higher comorbidity burden.

LIMITATION

Predominantly male population.

CONCLUSION

Booster mRNA vaccination was highly effective in preventing death and moderately effective in preventing infection and hospitalization for up to 4 months after administration in the Omicron era. Increased uptake of booster vaccination, which is currently suboptimal, should be pursued to limit the morbidity and mortality of SARS-CoV-2 infection, especially in persons with high comorbidity burden.

PRIMARY FUNDING SOURCE

U.S. Department of Veterans Affairs.

Theory and Practice of Integrating Machine Learning and Conventional Statistics in Medical Data Analysis

The practice of medical decision making is changing rapidly with the development of innovative computing technologies. The growing interest of data analysis with improvements in big data computer processing methods raises the question of whether machine learning can be integrated with conventional statistics in health research. To help address this knowledge gap, this paper presents a review on the conceptual integration between conventional statistics and machine learning, focusing on the health research. The similarities and differences between the two are compared using mathematical concepts and algorithms. The comparison between conventional statistics and machine learning methods indicates that conventional statistics are the fundamental basis of machine learning, where the black box algorithms are derived from basic mathematics, but are advanced in terms of automated analysis, handling big data and providing interactive visualizations. While the nature of both these methods are different, they are conceptually similar. Based on our review, we conclude that conventional statistics and machine learning are best to be integrated to develop automated data analysis tools. We also strongly believe that machine learning could be explored by health researchers to enhance conventional statistics in decision making for added reliable validation measures.

Validation of Biomarker-Based ABCD Score in Atrial Fibrillation Patients with a Non-Gender CHA2DS2-VASc Score 0-1: A Korean Multi-Center Cohort

PURPOSE

Atrial fibrillation (AF) patients with low to intermediate risk, defined as non-gender CHA₂DS₂-VASc score of 0-1, are still at risk of stroke. This study verified the usefulness of ABCD score [age (≥60 years), B-type natriuretic peptide (BNP) or N-terminal pro-BNP (≥300 pg/mL), creatinine clearance (<50 mL/min/1.73 m²), and dimension of the left atrium (≥45 mm)] for stroke risk stratification in non-gender CHA₂DS₂-VASc score 0-1.

MATERIALS AND METHODS

This multi-center cohort study retrospectively analyzed AF patients with non-gender CHA₂DS₂-VASc score 0-1. The primary endpoint was the incidence of stroke with or without antithrombotic therapy (ATT). An ABCD score was validated.

RESULTS

Overall, 2694 patients [56.3±9.5 years; female, 726 (26.9%)] were followed-up for 4.0±2.8 years. The overall stroke rate was 0.84/100 person-years (P-Y), stratified as follows: 0.46/100 P-Y for an ABCD score of 0; 1.02/100 P-Y for an ABCD score ≥1. The ABCD score was superior to non-gender CHA₂DS₂-VASc score in the stroke risk stratification (C-index=0.618, p=0.015; net reclassification improvement=0.576, p=0.040; integrated differential improvement=0.033, p=0.066). ATT was prescribed in 2353 patients (86.5%), and the stroke rate was significantly lower in patients receiving non-vitamin K antagonist oral anticoagulant (NOAC) therapy and an ABCD score ≥1 than in those without ATT (0.44/100 P-Y vs. 1.55/100 P-Y; hazard ratio=0.26, 95% confidence interval 0.11-0.63, p=0.003).

CONCLUSION

The biomarker-based ABCD score demonstrated improved stroke risk stratification in AF patients with non-gender CHA₂DS₂-VASc score 0-1. Furthermore, NOAC with an ABCD score ≥1 was associated with significantly lower stroke rate in AF patients with non-gender CHA₂DS₂-VASc score 0-1.

Recommendations for the use of propensity score methods in multiple sclerosis research

BACKGROUND

With many disease-modifying therapies currently approved for the management of multiple sclerosis, there is a growing need to evaluate the comparative effectiveness and safety of those therapies from real-world data sources. Propensity score methods have recently gained popularity in multiple sclerosis research to generate real-world evidence. Recent evidence suggests, however, that the conduct and reporting of propensity score analyses are often suboptimal in multiple sclerosis studies.

OBJECTIVES

To provide practical guidance to clinicians and researchers on the use of propensity score methods within the context of multiple sclerosis research.

METHODS

We summarize recommendations on the use of propensity score matching and weighting based on the current methodological literature, and provide examples of good practice.

RESULTS

Step-by-step recommendations are presented, starting with covariate selection and propensity score estimation, followed by guidance on the assessment of covariate balance and implementation of propensity score matching and weighting. Finally, we focus on treatment effect estimation and sensitivity analyses.

CONCLUSION

This comprehensive set of recommendations highlights key elements that require careful attention when using propensity score methods.

A Review of Causal Inference for External Comparator Arm Studies

Randomized controlled trials (RCTs) are the gold standard design to establish the efficacy of new drugs and to support regulatory decision making. However, a marked increase in the submission of single-arm trials (SATs) has been observed in recent years, especially in the field of oncology due to the trend towards precision medicine contributing to the rise of new therapeutic interventions for rare diseases. SATs lack results for control patients, and information from external sources can be compiled to provide context for better interpretability of study results. External comparator arm (ECA) studies are defined as a clinical trial (most commonly a SAT) and an ECA of a comparable cohort of patients-commonly derived from real-world settings including registries, natural history studies, or medical records of routine care. This publication aims to provide a methodological overview, to sketch emergent best practice recommendations and to identify future methodological research topics. Specifically, existing scientific and regulatory guidance for ECA studies is reviewed and appropriate causal inference methods are discussed. Further topics include sample size considerations, use of estimands, handling of different data sources regarding differential baseline covariate definitions, differential endpoint measurements and timings. In addition, unique features of ECA studies are highlighted, specifically the opportunity to address bias caused by unmeasured ECA covariates, which are available in the SAT.

Levosimendan's effect on venoarterial extracorporeal membrane oxygenation weaning

BACKGROUND

Veno-arterial extracorporeal membrane oxygenation (V-A ECMO) provides temporary haemodynamic support in refractory cardiogenic shock. Recent retrospective studies on levosimendan on V-A ECMO weaning had conflicting results. This study aimed to determine the association between levosimendan on V-A ECMO weaning success in a tertiary centre in Hong Kong.

METHODS

This retrospective study was conducted in an intensive care unit in Hong Kong. All adult patients requiring V-A ECMO from January 2016 to September 2020 were included. Patients who were given levosimendan were compared to patients who were not, on rates of successful V-A ECMO weaning. The groups were also compared after propensity matching based on covariates closely associated with the use of levosimendan.

RESULTS

A total of 119 patients were included in the study, with 38 in the levosimendan group and 81 in the non-levosimendan group. Patients treated with levosimendan trended towards improved weaning success, but the difference was not statistically significant (63% vs 53%, p = 0.404). In the propensity-matched groups, there was no difference in weaning success (odds ratio 1.00, 95% CI 0.23-8.00). The levosimendan group was associated with lower vasopressor requirement, lower lactate levels, and more significant drop in lactate in the first 2 days of V-A ECMO. The levosimendan group had longer ECMO duration. There was no difference in other secondary outcomes including mortality, length of stay in ICU and hospital and duration of mechanical ventilation. There was no difference in the rate of ventricular arrhythmias.

CONCLUSION

Levosimendan did not improve V-A ECMO weaning success in our cohort with refractory cardiogenic shock.

Comparison of Moderna versus Pfizer-BioNTech COVID-19 vaccine outcomes: A target trial emulation study in the U.S. Veterans Affairs healthcare system

BACKGROUND

mRNA COVID-19 vaccines manufactured by Pfizer-BioNTech (BNT162b2) and Moderna (mRNA-1273) have been shown to be efficacious but have not been compared in head-to-head clinical trials.

METHODS

We designed this observational study to emulate a target trial of COVID-19 vaccination by BNT162b2 versus mRNA-1273 among persons who underwent vaccination in the national U.S. Veterans Affairs (VA) healthcare system from 11/12/2020 to 25/03/2021 using combined VA and Medicare electronic health records. We identified the best matching mRNA-1273 recipient(s) for each BNT162b2 recipient, using exact/coarsened-exact matching (calendar week, VA integrated service network, age buckets and Charlson comorbidity index buckets) followed by propensity score matching. Vaccine recipients were followed from the date of first vaccine dose until 25/08/2021 for the development of SARS-CoV-2 infection, SARS-CoV-2-related hospitalization or SARS-CoV-2-related death.

FINDINGS

Each group included 902,235 well-matched vaccine recipients, followed for a mean of 192 days, during which 16,890 SARS-CoV-2 infections, 3591 SARS-CoV-2-related hospitalizations and 381 SARS-CoV-2-related deaths were documented. Compared to BNT162b2, mRNA-1273 recipients had significantly lower risk of SARS-CoV-2 infection (adjusted hazard ratio [aHR] 0.736, 95% CI 0.696-0.779) and SARS-CoV-2-related hospitalization (aHR 0.633, 95% CI 0.562-0.713), which persisted across all age groups, comorbidity burden categories and black/white race. The differences between mRNA-1273 and BNT162b2 in risk of infection or hospitalization were progressively greater when the follow-up period was longer, i.e. extending to March 31, June 30 or August 25, 2021. These differences were more pronounced when we analyzed separately the outcomes that occurred during the follow-up period from July 1 to August 25, 2021 when the Delta variant became predominant in the U.S. (aHR for infection 0.584, 95% CI 0.533-0.639 and aHR for hospitalization 0.387, 95% 0.311-0.482). SARS-CoV-2-related deaths were less common in mRNA-1273 versus BNT162b2 recipients (168 versus 213) but this difference was not statistically significant (aHR 0.808, 95% CI 0.592-1.103).

INTERPRETATION

In conclusion, although absolute rates of infection, hospitalization and death in both vaccine groups were low regardless of the vaccine received, our data suggests that compared to BNT162b2, vaccination with mRNA-1273 resulted in significantly lower rates of SARS-CoV-2-infection and SARS-CoV-2-related hospitalization. These differences were greater with longer follow-up time since vaccination and even more pronounced in the Delta variant era.

FUNDING

U.S. Department of Veterans Affairs, grant numbers COVID19-8900-11 and C19 21-278.

The Protective Effects of Lipid-Lowering Agents on Cardiovascular Disease and Mortality in Maintenance Dialysis Patients: Propensity Score Analysis of a Population-Based Cohort Study

Lipid-lowering agents display limited benefits on cardiovascular diseases and mortality in patients undergoing dialysis. Therefore, they are not routinely recommended for dialysis patients. The aim of this study was to assess the effects of lipid-lowering agents on clinical outcomes in dialysis patients on the basis of real-world evidence. This research used Taiwan’s National Health Insurance Research Database to identify dialysis patients from January 2009 to December 2015; patients were then categorized into a case group treated with lipid-lowering agents (n = 3,933) and a control group without lipid-lowering agents (n = 24,267). Patients were matched by age, sex, and comorbidities in a 1:1 ratio. This study used the Cox regression model to estimate the hazard ratios (HRs) for mortality and major adverse cardiovascular events (MACEs) for events recorded until December 2017. During a mean follow-up period of approximately 3.1 years, 1726 [43.9%, incidence 0.123/person-year (PY)] deaths and 598 (15.2%, incidence 0.047/PY) MACEs occurred in the case group and 2031 (51.6%, incidence 0.153/PY) deaths and 649 (16.5% incidence 0.055/PY) MACEs occurred in the control group. In the multivariable analysis of the Cox regression model, lipid-lowering agent users showed a significantly lower risk of death [HR: 0.75; 95% confidence interval (CI): 0.70–0.80] and MACEs (HR: 0.88; 95% CI: 0.78–0.98) than lipid-lowering agent non-users. Moreover, the survival benefit of lipid-lowering agents was significant across most subgroups. Dialysis patients treated with lipid-lowering agents display a 25 and 12% reduction in their risk of mortality and MACEs, respectively. Therefore, lipid-lowering agents might be considered when treating dialysis patients with hyperlipidemia.

Complaints about excessive use of police force in women's neighborhoods and subsequent perinatal and cardiovascular health

There are substantial, unexplained racial disparities in women’s health. Some of the most pronounced involve elevated rates of preterm delivery (PTD) and cardiovascular disease (CVD) among Black women. We hypothesized that stress associated with excessive use of force by police may contribute to these disparities. In two prospective cohorts derived from electronic health records (pregnancy cohort, N = 67,976; CVD cohort, N = 6773), we linked formal complaints of excessive police force in patients’ neighborhoods with health outcomes. Exposed Black women were 1.19 times as likely to experience PTD [95% confidence interval (CI): 1.04 to 1.35] and 1.42 times as likely to develop CVD (95% CI: 1.12 to 1.79), even after adjustment for neighborhood disadvantage and homicide. The excess risks of PTD were also observed in maternal fixed-effects analyses comparing births to the same woman. These findings suggest police violence may be an unrecognized contributor to health inequity for Black women.

Applied usage and performance of statistical matching in bibliometrics: The comparison of milestone and regular papers with multiple measurements of disruptiveness as an empirical example

Controlling for confounding factors is one of the central aspects of quantitative research. Although methods such as linear regression models are common, their results can be misleading under certain conditions. We demonstrate how statistical matching can be utilized as an alternative that enables the inspection of post-matching balancing. This contribution serves as an empirical demonstration of matching in bibliometrics and discusses the advantages and potential pitfalls. We propose matching as an easy-to-use approach in bibliometrics to estimate effects and remove bias. To exemplify matching, we use data about papers published in Physical Review E and a selection classified as milestone papers. We analyze whether milestone papers score higher in terms of a proposed class of indicators for measuring disruptiveness than nonmilestone papers. We consider disruption indicators DI1, DI5, DI1n, DI5n, and DEP and test which of the disruption indicators performs best, based on the assumption that milestone papers should have higher disruption indicator values than nonmilestone papers. Four matching algorithms (propensity score matching (PSM), coarsened exact matching (CEM), entropy balancing (EB), and inverse probability weighting (IPTW)) are compared. We find that CEM and EB perform best regarding covariate balancing and DI5 and DEP performing well to evaluate disruptiveness of published papers.

Oversampling and replacement strategies in propensity score matching: a critical review focused on small sample size in clinical settings

Background

Propensity score matching is a statistical method that is often used to make inferences on the treatment effects in observational studies. In recent years, there has been widespread use of the technique in the cardiothoracic surgery literature to evaluate to potential benefits of new surgical therapies or procedures. However, the small sample size and the strong dependence of the treatment assignment on the baseline covariates that often characterize these studies make such an evaluation challenging from a statistical point of view. In such settings, the use of propensity score matching in combination with oversampling and replacement may provide a solution to these issues by increasing the initial sample size of the study and thus improving the statistical power that is needed to detect the effect of interest. In this study, we review the use of propensity score matching in combination with oversampling and replacement in small sample size settings.

Methods

We performed a series of Monte Carlo simulations to evaluate how the sample size, the proportion of treated, and the assignment mechanism affect the performances of the proposed approaches. We assessed the performances with overall balance, relative bias, root mean squared error and nominal coverage. Moreover, we illustrate the methods using a real case study from the cardiac surgery literature.

Results

Matching without replacement produced estimates with lower bias and better nominal coverage than matching with replacement when 1:1 matching was considered. In contrast to that, matching with replacement showed better balance, relative bias, and root mean squared error than matching without replacement for increasing levels of oversampling. The best nominal coverage was obtained by using the estimator that accounts for uncertainty in the matching procedure on sets of units obtained after matching with replacement.

Conclusions

The use of replacement provides the most reliable treatment effect estimates and that no more than 1 or 2 units from the control group should be matched to each treated observation. Moreover, the variance estimator that accounts for the uncertainty in the matching procedure should be used to estimate the treatment effect.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01454-z.

Endoscopic versus open surgery in patients with malignant sinonasal tumours and brain invasion. A case series study

OBJECTIVES

To determine the safety, effectiveness and perioperative costs of endonasal endoscopic approach in brain invasive malignant sinonsal tumours patients.

MATERIALS AND METHODS

This was a case series bidirectional study; that included 30 brain invasive malignant sinonsal tumours patients treated by endonasal endoscopic approach (2015-2017) and 53 by open surgery (2010-2015). Propensity score matching was used to compensate the prognostic factors; in a sample of 50 patients (25 per group). Primary response variables was local control and 3-years overall survival. Perioperative cost variables were analyzed.

RESULTS

A number of 50 patients were included after matching (25 in each therapeutic group). The age average was 55 years and male proportion was 62%. Squamous cell carcinoma and grade II lesions were the most represented in the sample. Endonasal endoscopic approach reduced surgical time in 1 h 20 min, transfusion needs in 5.5 fold and hospitalization in 19 days; in comparison with open technique. Oncologic control based on surgical free margins, local control, overall survival and progression free survival after three years was higher when the resection was performed endoscopically. Functional status was enhanced and complications diminished by using endoscopic approach. Saving was estimated in $7 355.18 per patient.

CONCLUSIONS

Endonasal endoscopic approach represents a safe, effective and economic procedure in selected patients with malignant sinonasal tumors and brain invasion.

Endoscopic versus open surgery in patients with malignant sinonasal tumors and brain invasion. A case series study

OBJECTIVES

To determine the safety, effectiveness and perioperative costs of endonasal endoscopic approach in brain invasive malignant sinonsal tumors patients.

MATERIALS AND METHODS

This was a case series bidirectional study; that included 30 brain invasive malignant sinonsal tumors patients treated by endonasal endoscopic approach (2015-2017) and 53 by open surgery (2010-2015). Propensity score matching was used to compensate the prognostic factors; in a sample of 50 patients (25 per group). Primary response variables was local control and 3-years overall survival. Perioperative cost variables were analyzed.

RESULTS

A number of 50 patients were included after matching (25 in each therapeutic group). The age average was 55 years and male proportion was 62%. Squamous cell carcinoma and grade II lesions were the most represented in the sample. Endonasal endoscopic approach reduced surgical time in 1 hour 20 minutes, transfusion needs in 5.5 fold and hospitalization in 19 days; in comparison with open technique. Oncologic control based on surgical free margins, local control, overall survival and progression free survival after three years was higher when the resection was performed endoscopically. Functional status was enhanced and complications diminished by using endoscopic approach. Saving was estimated in $7 355.18 per patient.

CONCLUSIONS

Endonasal endoscopic approach represents a safe, effective and economic procedure in selected patients with malignant sinonasal tumors and brain invasion.

Using Propensity Scores for Causal Inference: Pitfalls and Tips

Methods based on propensity score (PS) have become increasingly popular as a tool for causal inference. A better understanding of the relative advantages and disadvantages of the alternative analytic approaches can contribute to the optimal choice and use of a specific PS method over other methods. In this article, we provide an accessible overview of causal inference from observational data and two major PS-based methods (matching and inverse probability weighting), focusing on the underlying assumptions and decision-making processes. We then discuss common pitfalls and tips for applying the PS methods to empirical research and compare the conventional multivariable outcome regression and the two alternative PS-based methods (ie, matching and inverse probability weighting) and discuss their similarities and differences. Although we note subtle differences in causal identification assumptions, we highlight that the methods are distinct primarily in terms of the statistical modeling assumptions involved and the target population for which exposure effects are being estimated.

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.

Using propensity scores to estimate effects of treatment initiation decisions: State of the science

Confounding can cause substantial bias in nonexperimental studies that aim to estimate causal effects. Propensity score methods allow researchers to reduce bias from measured confounding by summarizing the distributions of many measured confounders in a single score based on the probability of receiving treatment. This score can then be used to mitigate imbalances in the distributions of these measured confounders between those who received the treatment of interest and those in the comparator population, resulting in less biased treatment effect estimates. This methodology was formalized by Rosenbaum and Rubin in 1983 and, since then, has been used increasingly often across a wide variety of scientific disciplines. In this review article, we provide an overview of propensity scores in the context of real-world evidence generation with a focus on their use in the setting of single treatment decisions, that is, choosing between two therapeutic options. We describe five aspects of propensity score analysis: alignment with the potential outcomes framework, implications for study design, estimation procedures, implementation options, and reporting. We add context to these concepts by highlighting how the types of comparator used, the implementation method, and balance assessment techniques have changed over time. Finally, we discuss evolving applications of propensity scores.

Propensity score methods for time-dependent cluster confounding

In observational studies, subjects are often nested within clusters. In medical studies, patients are often treated by doctors and therefore patients are regarded as nested or clustered within doctors. A concern that arises with clustered data is that cluster-level characteristics (e.g., characteristics of the doctor) are associated with both treatment selection and patient outcomes, resulting in cluster-level confounding. Measuring and modeling cluster attributes can be difficult and statistical methods exist to control for all unmeasured cluster characteristics. An assumption of these methods however is that characteristics of the cluster and the effects of those characteristics on the outcome (as well as probability of treatment assignment when using covariate balancing methods) are constant over time. In this paper, we consider methods that relax this assumption and allow for estimation of treatment effects in the presence of unmeasured time-dependent cluster confounding. The methods are based on matching with the propensity score and incorporate unmeasured time-specific cluster effects by performing matching within clusters or using fixed- or random-cluster effects in the propensity score model. The methods are illustrated using data to compare the effectiveness of two total hip devices with respect to survival of the device and a simulation study is performed that compares the proposed methods. One method that was found to perform well is matching within surgeon clusters partitioned by time. Considerations in implementing the proposed methods are discussed.

Variance estimation when using propensity-score matching with replacement with survival or time-to-event outcomes

Propensity‐score matching is a popular analytic method to estimate the effects of treatments when using observational data. Matching on the propensity score typically requires a pool of potential controls that is larger than the number of treated or exposed subjects. The most common approach to matching on the propensity score is matching without replacement, in which each control subject is matched to at most one treated subject. Failure to find a matched control for each treated subject can lead to “bias due to incomplete matching.” To avoid this bias, it is important to identify a matched control subject for each treated subject. An alternative to matching without replacement is matching with replacement, in which control subjects are allowed to be matched to multiple treated subjects. A limitation to the use of matching with replacement is that variance estimation must account for both the matched nature of the sample and for some control subjects being included in multiple matched sets. While a variance estimator has been proposed for when outcomes are continuous, no such estimator has been proposed for use with time‐to‐event outcomes, which are common in medical and epidemiological research. We propose a variance estimator for the hazard ratio when matching with replacement. We conducted a series of Monte Carlo simulations to examine the performance of this estimator. We illustrate the utility of matching with replacement to estimate the effect of smoking cessation counseling on survival in smokers discharged from hospital with a heart attack.