Nonrandom Treatment Assignment Research Articles

Tightening Blocks in Complementary Analyses of Observational Studies: Optimization Algorithm and Examples

An observational block design has I blocks matched for covariates and J individuals per block, but treatments were not randomly assigned to individuals within blocks, as would have been done in an experiment. Tightening an observational block design means selecting J ′ < J individuals from each block, and possibly I ′ ≤ I blocks, to construct a new observational block design that, in some way, addresses unmeasured biases from nonrandom treatment assignment. Tightening must preserve covariate balance while altering the design to achieve some additional objective. An optimization algorithm is introduced that achieves this while maintaining the block structure by finely balancing covariates across blocks and through optimal subset matching. An example is considered in detail, both to motivate and illustrate the tightening of an observational block design. Two tightened designs are built from a study of light daily alcohol consumption and its possible effects on HDL cholesterol. One tightened design adjusts for an outcome tentatively presuming it was unaffected by the treatment. The second tightened design uses a differential effect to remove bias from an unobserved general disposition that promotes several treatments. An R package tightenBlock implements the method, contains the data, and in that package the help-file for the function tighten reproduces the example.

Impact of Quitline Services on Tobacco Cessation: An Application of Modern Epidemiological Methods.

This study investigated the effectiveness of quitline service intensity (high vs. low) on past 30-day tobacco abstinence at 7-months follow-up, using observational data from the Oklahoma Tobacco Helpline (OTH) between April 2020 and December 2021. To assess the impact of loss to follow-up and non-random treatment assignment, we fit the parameters of a marginal structural model to estimate inverse probability weights for censoring (IPCW) and treatment (IPTW) and combined (IPCTW). The Risk Ratio (RR) was estimated using modified Poisson regression with robust variance estimator. Of the 4,695 individuals included in the study, 64% received high-intensity cessation services, and 53% were lost to follow-up. Using the conventional complete case analysis (responders only), high-intensity cessation services were associated with abstinence (RR=1.18; 95 CI: 1.04, 1.34). The effect estimate was attenuated after accounting for censoring (RR=1.14; 95% CI: 1.00, 1.30). After adjusting for both baseline confounding and selection bias via IPTCW, high-intensity cessation services were associated with 1.23 times (95% CI: 1.08, 1.41) the probability of abstinence compared to low-intensity services. Despite relatively high loss to follow-up, accounting for selection bias and confounding did not notably impact quit rates or the relationship between intensity of quitline services and tobacco cessation among OTH participants.

Developing an individualized treatment rule for Veterans with major depressive disorder using electronic health records.

Efforts to develop an individualized treatment rule (ITR) to optimize major depressive disorder (MDD) treatment with antidepressant medication (ADM), psychotherapy, or combined ADM-psychotherapy have been hampered by small samples, small predictor sets, and suboptimal analysis methods. Analyses of large administrative databases designed to approximate experiments followed iteratively by pragmatic trials hold promise for resolving these problems. The current report presents a proof-of-concept study using electronic health records (EHR) of n = 43,470 outpatients beginning MDD treatment in Veterans Health Administration Primary Care Mental Health Integration (PC-MHI) clinics, which offer access not only to ADMs but also psychotherapy and combined ADM-psychotherapy. EHR and geospatial databases were used to generate an extensive baseline predictor set (5,865 variables). The outcome was a composite measure of at least one serious negative event (suicide attempt, psychiatric emergency department visit, psychiatric hospitalization, suicide death) over the next 12 months. Best-practices methods were used to adjust for nonrandom treatment assignment and to estimate a preliminary ITR in a 70% training sample and to evaluate the ITR in the 30% test sample. Statistically significant aggregate variation was found in overall probability of the outcome related to baseline predictors (AU-ROC = 0.68, S.E. = 0.01), with test sample outcome prevalence of 32.6% among the 5% of patients having highest predicted risk compared to 7.1% in the remainder of the test sample. The ITR found that psychotherapy-only was the optimal treatment for 56.0% of patients (roughly 20% lower risk of the outcome than if receiving one of the other treatments) and that treatment type was unrelated to outcome risk among other patients. Change in aggregate treatment costs of implementing this ITR would be negligible, as 16.1% fewer patients would be prescribed ADMs and 2.9% more would receive psychotherapy. A pragmatic trial would be needed to confirm the accuracy of the ITR.

Comparative Effectiveness of Mepolizumab, Benralizumab, and Dupilumab among Patients with Difficult-to-Control Asthma: A Multicenter Retrospective Propensity-matched Analysis.

Rationale: The comparative effectiveness of biologic agents used as add-on therapy in the management of difficult-to-control asthma is unclear. Objective: To compare the effectiveness of dupilumab, mepolizumab, and benralizumab among patients with difficult-to-control asthma. Methods: Retrospective multicenter cohort study of adult patients with difficult-to-control asthma starting treatment with dupilumab, mepolizumab, or benralizumab as documented in a multicenter electronic health record and claims-based database between October 19, 2018, and September 30, 2022. Propensity-score matching was used to minimize bias from nonrandomized treatment assignment; a prespecified α-level was set at 0.017 to account for three primary comparisons. The exposure of interest was the new initiation of dupilumab, benralizumab, or mepolizumab treatment. The primary outcome was the rate of asthma exacerbations in the 1 year after initiation of biologic therapy modeled using a negative binomial approach. Results: Among 893,668 patients with asthma who were prescribed an inhaled corticosteroid and were ⩾12 years old (65% female; mean age, 49 yr), 3,943 started dupilumab, 1,902 started benralizumab, and 2,012 started mepolizumab, all without an alternative indication for biologic therapy. After matching, there were 1,805 patients in each group for comparisons between dupilumab and benralizumab, 1,865 for comparisons between dupilumab and mepolizumab, and 1,721 for comparisons between mepolizumab and benralizumab. For all pairwise comparisons, covariates were well balanced after matching (all standardized mean differences <0.1). Patients who initiated dupilumab had a significantly lower rate of asthma exacerbations (1.07 per year) compared with benralizumab (1.47 per year), with a rate ratio (RR) of 0.73 (95% confidence interval, 0.63-0.85), and also had a significantly lower rate of asthma exacerbations compared with mepolizumab (1.04 per year vs. 1.45 per year), with an RR of 0.72 (0.62-0.84). There was no statistically significant difference in the rate of asthma exacerbations between mepolizumab (1.40 per year) and benralizumab (1.41 per year), with an RR of 1.00 (0.85-1.17). Conclusions: In patients with difficult-to-control asthma who had newly initiated biologic therapy, dupilumab was associated with a decreased rate of asthma exacerbations in the 1 year after initiation compared with mepolizumab or benralizumab.

EFFECT OF RISE ON REPEAT ELDER ABUSE AND SELF-NEGLECT INVESTIGATIONS BY ADULT PROTECTIVE SERVICES

Abstract Adult Protective Services (APS) is the primary agency responsible for investigating Elder Abuse and Self-Neglect (EASN) allegations in the US. The harms of EASN are well established; however, APS lacks a conceptually derived, evidenced-based intervention phase. RISE is a community-based EASN intervention designed to complement and augment APS that provides enhanced services and a longer intervention phase. The objective of this study was to test whether exposure to the APS/RISE collaboration reduced the case outcome of recurrence (repeat investigations) compared to usual APS only services. This study was based on a retrospective observational design (n = 1947) in two randomly selected counties of Maine where RISE was available to provide enhanced services to persons referred from APS. Analysis used an extended regression endogenous treatment Probit model using APS administrative data to predict case recurrence. Between 2019 and 2021, 154 cases participated in RISE and 1793 received usual APS only services. 49% of cases referred to RISE had a complex history with 2 or more prior substantiated allegations versus 6% for those receiving usual APS only services. After accounting for the non-random treatment assignment, RISE significantly lowered the likelihood of recurrence compared to persons receiving usual care provided by APS (probability of recurrence reduced by 0.55 for the Average Treatment Effect on the Treated and 0.26 for the Average Treatment Effect). A reduction in recurrence carries important implications for APS costs, resources, and workflow. It may also serve as a proxy indicating a reduction in re-victimization and harm for EASN victims.

Comparative Effectiveness of Adalimumab vs Tofacitinib in Patients With Rheumatoid Arthritis in Australia

There is a need for observational studies to supplement evidence from clinical trials, and the target trial emulation (TTE) framework can help avoid biases that can be introduced when treatments are compared crudely using observational data by applying design principles for randomized clinical trials. Adalimumab (ADA) and tofacitinib (TOF) were shown to be equivalent in patients with rheumatoid arthritis (RA) in a randomized clinical trial, but to our knowledge, these drugs have not been compared head-to-head using routinely collected clinical data and the TTE framework. To emulate a randomized clinical trial comparing ADA vs TOF in patients with RA who were new users of a biologic or targeted synthetic disease-modifying antirheumatic drug (b/tsDMARD). This comparative effectiveness study emulating a randomized clinical trial of ADA vs TOF included Australian adults aged 18 years or older with RA in the Optimising Patient Outcomes in Australian Rheumatology (OPAL) data set. Patients were included if they initiated ADA or TOF between October 1, 2015, and April 1, 2021; were new b/tsDMARD users; and had at least 1 component of the disease activity score in 28 joints using C-reactive protein (DAS28-CRP) recorded at baseline or during follow-up. Treatment with either ADA (40 mg every 14 days) or TOF (10 mg daily). The main outcome was the estimated average treatment effect, defined as the difference in mean DAS28-CRP among patients receiving TOF compared with those receiving ADA at 3 and 9 months after initiating treatment. Missing DAS28-CRP data were multiply imputed. Stable balancing weights were used to account for nonrandomized treatment assignment. A total of 842 patients were identified, including 569 treated with ADA (387 [68.0%] female; median age, 56 years [IQR, 47-66 years]) and 273 treated with TOF (201 [73.6%] female; median age, 59 years [IQR, 51-68 years]). After applying stable balancing weights, mean DAS28-CRP in the ADA group was 5.3 (95% CI, 5.2-5.4) at baseline, 2.6 (95% CI, 2.5-2.7) at 3 months, and 2.3 (95% CI, 2.2-2.4) at 9 months; in the TOF group, it was 5.3 (95% CI, 5.2-5.4) at baseline, 2.4 (95% CI, 2.2-2.5) at 3 months, and 2.3 (95% CI, 2.1-2.4) at 9 months. The estimated average treatment effect was -0.2 (95% CI, -0.4 to -0.03; P = .02) at 3 months and -0.03 (95% CI, -0.2 to 0.1; P = .60) at 9 months. In this study, there was a modest but statistically significant reduction in DAS28-CRP at 3 months for patients receiving TOF compared with those receiving ADA and no difference between treatment groups at 9 months. Three months of treatment with either drug led to clinically relevant average reductions in mean DAS28-CRP, consistent with remission.

Comparative Effectiveness of First-Line Baricitinib in Patients With Rheumatoid Arthritis in the Australian OPAL Data Set.

To analyze comparative treatment persistence for first-line baricitinib (BARI) versus first-line tumor necrosis factor inhibitor (TNFi) in patients with rheumatoid arthritis (RA) and for first-line BARI initiated as monotherapy versus first-line BARI initiated with at least one conventional synthetic disease-modifying antirheumatic drug (csDMARD). Patients with RA who initiated BARI or TNFi as first-line biologic or targeted synthetic DMARD from October 1, 2015, to September 30, 2021, were identified in the OPAL data set. Drug survival times to 6, 12, and 24 months were analyzed using restricted mean survival time (RMST). Multiple imputation and inverse probability of treatment weighting were used to address missing data and nonrandom treatment assignment. A total of 545 patients initiated first-line BARI, including 118 as monotherapy and 427 as csDMARD combination therapy. Three thousand five hundred patients initiated first-line TNFi. There was no difference in drug survival to 6 or 12 months for BARI compared with TNFi; differences in RMST were 0.02 months (95% CI: -0.08 to 0.013; P = 0.65) and 0.31 months (95% CI: -0.02 to 0.63; P = 0.06), respectively. Patients in the BARI group had 1.00 month (95% CI: 0.14 to 1.86; P = 0.02) longer drug survival to 24 months. There was no difference in drug survival for BARI monotherapy compared with combination therapy, with differences in RMST to 6, 12, and 24 months of -0.19 months (95% CI: -0.50 to 0.12; P = 0.12), -0.35 months (95% CI: -1.17 to 0.42; P = 0.41), and -0.56 months (95% CI: -2.66 to 1.54; P = 0.60), respectively. In this comparative analysis, treatment persistence up to 24 months was significantly longer for first-line BARI compared with TNFi, but the effect size of 1.00 month is not clinically meaningful. There was no difference in persistence for BARI monotherapy versus combination therapy.

Antibiotic Prophylaxis of Surgical Site Infections in Cutaneous Surgery: A Prospective Observational Study.

The use of perioperative antibiotic prophylaxis in cutaneous surgery is controversial due to unclear efficacy and, thus, potentially unnecessary side-effects. This prospective observational study analysed the efficacy of oral perioperative antibiotic prophylaxis in preventing surgical site infections. Adult patients undergoing cutaneous surgery between August 2020 and May 2021 at Ludwig-Maximilian University Hospital Munich, Germany, without prior signs of infection were eligible. Propensity score weighting was used for covariate adjustment to account for non-randomized treatment assignment. Of 758 included patients, 23 received perioperative antibiotic prophylaxis (3.0%). In this group, a surgical site infection occurred in 1 of 45 lesions (2.2%) compared with 76 of 1,189 lesions (6.5%) in the group without perioperative antibiotic prophylaxis (735 patients, 97.0%). With covariate adjustment, the odds ratio for the occurrence of a surgical site infection in patients receiving perioperative antibiotic prophylaxis was 0.114 (95% confidence interval 0.073-0.182; p <0.001) on a per lesion level. The number of lesions needed to treat to prevent 1 surgical site infection was 17.6 (95% confidence interval 16.8-19.2). This prospective observational study shows a reduction in the incidence of surgical site infection in cutaneous surgery performed with perioperative antibiotic prophylaxis. The large size difference between the 2 study groups limits the study.

Imputation-based Q-learning for optimizing dynamic treatment regimes with right-censored survival outcome.

Q-learning has been one of the most commonly used methods for optimizing dynamic treatment regimes (DTRs) in multistage decision-making. Right-censored survival outcome poses a significant challenge to Q-Learning due to its reliance on parametric models for counterfactual estimation which are subject to misspecification and sensitive to missing covariates. In this paper, we propose an imputation-based Q-learning (IQ-learning) where flexible nonparametric or semiparametric models are employed to estimate optimal treatment rules for each stage and then weighted hot-deck multiple imputation (MI) and direct-draw MI are used to predict optimal potential survival times. Missing data are handled using inverse probability weighting and MI, and the nonrandom treatment assignment among the observed is accounted for using a propensity-score approach. We investigate the performance of IQ-learning via extensive simulations and show that it is more robust to model misspecification than existing Q-Learning methods, imputes only plausible potential survival times contrary to parametric models and provides more flexibility in terms of baseline hazard shape. Using IQ-learning, we developed an optimal DTR for leukemia treatment based on a randomized trial with observational follow-up that motivated thisstudy.

Can conditioned taste aversion be deployed at a landscape level to mitigate the impact of invasive cane toads on northern quolls?

Context Animals can be taught new behaviours to mitigate threatening processes. However, it is yet to be confirmed if such teaching can be deployed in the field. Here we test this possibility using the invasion of cane toads because they are highly toxic novel prey items to many predators across northern Australia. Aims Research has shown that training predators to avoid toads, using conditioned taste aversion (CTA), significantly improves the survival rates of individuals. We sought to determine if deployment of CTA baits, in situ, could reduce cane toad impacts on a mammalian predator, the northern quoll (Dasyurus hallucatus). Methods The work was conducted in the Kimberley, Western Australia, where we treated three quoll populations with training baits from November 2015–November 2017, and kept four populations as controls. We used camera traps and Bayesian hierarchal modelling to estimate the population size of quolls on up to four occasions before, and two occasions after, cane toad arrival. Key results We observed a 65% reduction in quoll population size at control sites and a 94% reduction at treatment sites: a significant effect of aversion training, but in the direction opposite to that expected. Conclusions Two complexities – decay of aversion, and individual variation – together, may explain our result. Our trials indicate that most animals are no longer averse within 120 days post-training. Earlier studies indicating that aversion training lasts longer may have inadvertently observed innate (i.e. genetically based) aversion to cane toads. Another possibility is that our dose rate of thiabendazole within the CTA baits was, in fact, too low; or the result may be an artefact of the non-random assignment of treatment and control groups. Ultimately, there is no way of determining exactly why our outcome occurred without further laboratory/captive trails. Regardless, our study demonstrates a failure of CTA training to mitigate the impact of cane toads on northern quolls in this system. Implications Our result calls into question the value of this approach for conserving quoll populations, at least in its current form. More generally, our results point to the often-unexpected complexities encountered as ideas progress from captive trials to field deployment.

Radiofrequency-assisted transection of the pancreas vs stapler in distal pancreatectomy: a propensity score matched cohort analysis.

To demonstrate the efficacy of radiofrequency for pancreatic stump closure in reducing the incidence of postoperative pancreatic fistula (POPF) in distal pancreatectomy (DP) compared with mechanical transection methods. Despite all the different techniques of pancreatic stump closure proposed for DP, best practice for avoiding POPF remains an unresolved issue, with an incidence of up to 30% regardless of center volume or surgical expertise. DP was performed in a cohort of patients by applying radiofrequency to stump closure (RF Group) and compared with mechanical closure (Control Group). A propensity score (PS) matched cohort study was carried out to minimize bias from nonrandomized treatment assignment. Cohorts were matched by PS accounting for factors significantly associated with either undergoing RF transection or mechanical closure through logistic regression analysis. The primary end-point was the incidence of clinically relevant POPF (CR-POPF). Of 89 patients included in the whole cohort, 13 case patients from the RF-Group were 1:1 matched to 13 control patients. In both the first independent analysis of unmatched data and subsequent adjustment to the overall propensity score-matched cohort, a higher rate of CR-POPF in the Control Group compared with the RF-Group was detected (25.4% vs 5.3%, p = 0.049 and 53.8% vs 0%; p = 0.016 respectively). The RF Group showed better outcomes in terms of readmission rate (46.2% vs 0%, p = 0.031). No significant differences were observed in terms of mortality, major complications (30.8% vs 0%, p = 0.063) or length of hospital stay (5.7 vs 5.2 days, p = 0.89). Findings suggest that the RF-assisted technique is more efficacious in reducing CR-POPF than mechanical pancreatic stump closure.

Long-term age-stratified survival following endovascular and open abdominal aortic aneurysm repair

ObjectiveThe long-term survival differences between endovascular repair (EVAR) and open repair for abdominal aortic aneurysms (AAAs) and specifically the impact of age on these differences remain a topic of debate. Therefore, we compared the long-term mortality between EVARand open abdominal aneurysm repair for patients of different ages. MethodsThis was a retrospective cohort study of prospectively collected data from patients undergoing elective EVAR or open repair for infrarenal AAAs within the Vascular Quality Initiative multinational clinical registry (2003-2021). The primary outcome was long-term all-cause mortality comparing EVAR and open repair for patients aged less than 65 years, between 65 and 79 years, and those aged 80 and older. In addition, we investigated the interaction between repair modality and 10-year hazard of mortality for sex, aneurysm diameter, and several preoperative comorbid conditions within each age category. To account for the nonrandom assignment of treatment, we used propensity scores and inverse probability weighted Cox proportional hazard analysis. ResultsWe identified 48,074 patients undergoing elective infrarenal abdominal aneurysm repair (89% EVAR) within the study period, including 7940 patients aged less than 65, 29,555 aged between 65 and 79, and 10,579 aged 80 years or more. EVAR was associated with a higher propensity score-adjusted long-term hazard of mortality compared with open repair in the cohort aged less than 65 years (hazard ratio [HR], 1.39; 95% confidence interval [CI], 1.04-1.86; P = .026). The mortality was similar in the age cohort between 65 and 79 (HR, 0.94; 95% CI, 0.79-1.10; P = .43), whereas EVAR was associated with a lower hazard of mortality in the cohort aged 80 years or more (HR, 0.63; 95% CI, 0.46-0.86; P = .004). In patients aged less than 65 years, the hazard of mortality was higher with EVAR compared with open repair in those with female sex (HR, 4.40; 95% CI, 1.75-11.0), an aneurysm diameter of more than 65 mm (HR, 2.19; 95% CI, 1.11-4.34), and an absence of coronary artery disease (HR, 1.26; 95% CI, 0.83-1.91), congestive heart failure (HR, 1.41; 95% CI, 1.03-1.92), and renal dysfunction (HR, 1.46; 95% CI, 1.04-2.05). In the patient cohort aged 80 and older, a lower hazard of mortality for EVAR versus open repair was observed for male patients or those with small aneurysms or certain comorbidities. ConclusionsIn a selected group of young patients with a substantial life expectancy, the long-term mortality is higher with EVAR compared with open repair for infrarenal AAAs. Long-term mortality with EVAR is similar in the middle cohort and lower in the elderly cohort compared with open repair.

All-Cause Mortality in Ischemic Heart Failure Patients with Functional Mitral Regurgitation Undergoing Percutaneous Coronary Intervention

This study aimed to evaluate the association between percutaneous coronary intervention (PCI) treatment and all-cause mortality in patients with ischemic heart failure with left ventricular systolic dysfunction and functional mitral regurgitation (FMR). We included 1,483 patients of which 39.5% (n=586) had moderate-to-severe FMR. Multivariable Cox proportional hazard model was used to assess the association between PCI treatment and all-cause mortality. Furthermore, propensity score matching was used to account for nonrandom treatment assignment. In those with none-to-mild FMR, after a median follow-up of 3.1years, the cumulative rate of all-cause mortality between the PCI and non-PCI groups was comparable (10.1% vs 14.2%), with an adjusted hazard ratio (HR) of 0.731 (95% confidence interval [CI] 0.438 to 1.221, p=0.232). In those with moderate-to-severe FMR, after a median follow-up of 2.9years, the cumulative rate of all-cause mortality was lower in the PCI group (20.4% vs 31.6%), with an adjusted HR of 0.660 (95% CI 0.469 to 0.929, p=0.017). The result was confirmed with propensity matching (HR 0.596 and 95% CI 0.363 to 0.977, p=0.038). The mortality benefit associated with PCI treatment in patients with moderate-to-severe FMR was consistent regardless of the age, gender, reason for admission, presence of diabetes mellitus, left ventricular ejection fraction value, left main and 3-vessels disease. In conclusion, in patients with ischemic heart failure with left ventricular systolic dysfunction and moderate-to-severe FMR, PCI treatment was associated with improvement in all-cause mortality. Randomized clinical trials are needed to confirm the current results.

Early Education and Adult Health: Age 37 Impacts and Economic Benefits of the Child-Parent Center Preschool Program.

This paper evaluates the long-term impacts of the Chicago Child-Parent Center (CPC) program, a comprehensive early childhood program launched in the 1960s, on the physical and mental health outcomes. This study follows a cohort of 1539 participants born in 1979-1980 and surveyed most recently at age 35-37 by employing a matched study design created by including all students who were enrolled in kindergarten classrooms in CPC school sites as well as entire kindergarten classrooms in a matched set of similar high-poverty schools. Using propensity score weighting that addresses potential issues with differential attrition and nonrandom treatment assignment, results reveal that CPC preschool participation is associated with significantly lower rates of adverse health outcomes such as smoking and diabetes. Further, evaluating the economic impacts of the preschool component of the program, the study finds a benefit-cost ratio in the range of 1.35 to 3.66 (net benefit: $3,896) indicating that the health benefits of the program by themselves offset the costs of the program even without considering additional benefits arising from increased educational attainment and reduced involvement in crime reported in earlier cost-benefit analyses. The findings are robust to corrections for multiple hypothesis testing, sensitivity analysis using a range of discount rates, and Monte Carlo analysis to account for uncertainty in outcomes.

Steroids and mortality in non‐critically ill COVID‐19 patients: a propensity score‐weighted study in a Chilean cohort

ObjectivesThe aim of this study was to evaluate the impact on 30-day mortality of early use of corticosteroids in COVID-19 patients with supplementary oxygen requirements and without invasive mechanical ventilation at the initiation of therapy. MethodsAll patients hospitalized with COVID-19 between April 15 and July 15, 2020, and requiring supplementary oxygen, were prospectively included in a database. Patients who died or required intubation within the first 48 hours were excluded. Patients who received corticosteroids within the first 5 days of hospitalization and at least 24 hours prior to intubation were allocated to the ‘early corticosteroids’ group. To compare both populations and adjust for non-random treatment assignment bias, a weight-adjusted propensity score model was used. ResultsIn total, 571 patients met the inclusion criteria, 520 had sufficient information for the analysis. Of these, 233 received early corticosteroids and 287 did not. Analysis showed a reduction of 8.5% (p = 0.038) in 30-day mortality in the early corticosteroid group. The reduction in mortality was not significant when patients with corticosteroid initiation between day 5 and day 8 of hospitalization were included. ConclusionEarly corticosteroid use reduced mortality in patients with pneumonia due to COVID-19, who required supplementary oxygen but not initial invasive mechanical ventilation.

Modelling hospital outcome: problems with endogeneity

BackgroundMortality modelling in the critical care paradigm traditionally uses logistic regression, despite the availability of estimators commonly used in alternate disciplines. Little attention has been paid to covariate endogeneity and the status of non-randomized treatment assignment. Using a large registry database, various binary outcome modelling strategies and methods to account for covariate endogeneity were explored.MethodsPatient mortality data was sourced from the Australian & New Zealand Intensive Society Adult Patient Database for 2016. Hospital mortality was modelled using logistic, probit and linear probability (LPM) models with intensive care (ICU) providers as fixed (FE) and random (RE) effects. Model comparison entailed indices of discrimination and calibration, information criteria (AIC and BIC) and binned residual analysis. Suspect covariate and ventilation treatment assignment endogeneity was identified by correlation between predictor variable and hospital mortality error terms, using the Stata™ “eprobit” estimator. Marginal effects were used to demonstrate effect estimate differences between probit and “eprobit” models.ResultsThe cohort comprised 92,693 patients from 124 intensive care units (ICU) in calendar year 2016. Patients mean age was 61.8 (SD 17.5) years, 41.6% were female and APACHE III severity of illness score 54.5(25.6); 43.7% were ventilated. Of the models considered in predicting hospital mortality, logistic regression (with or without ICU FE) and RE logistic regression dominated, more so the latter using information criteria indices. The LPM suffered from many predictions outside the unit [0,1] interval and both poor discrimination and calibration. Error terms of hospital length of stay, an independent risk of death score and ventilation status were correlated with the mortality error term. Marked differences in the ventilation mortality marginal effect was demonstrated between the probit and the "eprobit" models which were scenario dependent. Endogeneity was not demonstrated for the APACHE III score.ConclusionsLogistic regression accounting for provider effects was the preferred estimator for hospital mortality modelling. Endogeneity of covariates and treatment variables may be identified using appropriate modelling, but failure to do so yields problematic effect estimates.

Propensity score–matched comparison between totally laparoscopic right hemicolectomy with transcolonic natural orifice specimen extraction and conventional laparoscopic surgery with mini-laparotomy in the treatment of ascending colon cancer (with video)

Now that the debate about the safety and effectiveness of laparoscopic versus open surgery is over, attention has turned to innovations that can verify whether minimizing the impact of laparoscopy on the abdominal wall can further reduce pain, improve patient comfort, lead to superior cosmesis, and reduce morbidity. The aim of this study was to further explore the application value of totally laparoscopic right hemicolectomy with transcolonic natural orifice specimen extraction (NOSE) and to evaluate the short-term efficacy of transcolonic NOSE surgery for resecting specimens of ascending colon cancer. From January 2016 to May 2017, a retrospective study was conducted in Guangxi. Propensity score matching was used to minimize the bias from nonrandomized treatment assignment. Patients were followed up through May2020. Forty-nine patients underwent totally laparoscopic right hemicolectomy with transcolonic NOSE and 116 patients laparoscopic right hemicolectomy with mini-laparotomy (ML) procedures at our institution. After propensity score matching, each group included 45 patients, and all covariate imbalances were alleviated. The transcolonic NOSE group and the ML group did not differ significantly in terms of baseline clinical characteristics. The transcolonic NOSE group was associated with a shorter time to first flatus (NOSE vs ML: 1.8 ± .5 vs 3.2 ± .8, P= .032), a shorter length of hospital stay (11.3 ± 2.5 days vs 13.0 ± 3.1 days, P= .034), a shorter time to first liquid intake (2.6 ± .8 vs 3.8 ± .9, P= .068), less pain (1.8 ± .8 vs 4.2 ± .7, P= .013), less analgesia requirement (6 [13.3%] vs 21 [46.7%], P= .001), and lower C-reactive protein levels on postoperative day 1 (3.6 ± 1.7 vs 8.2 ± 2.2, P= .001) and postoperative day 3 (NOSE 2.4 ± 1.4 vs M: 4.6 ± 1.7 [P= .013]) than the ML group. The median follow-up was 28.4 months (interquartile range, 18.0-36.0). The 3-year overall survival rates were similar between the transcolonic NOSE group and the ML group. In total, laparoscopic right hemicolectomy with transcolonic specimen extraction appears to be safe for selected patients with ascending colon cancer as a minimally invasive surgery.

Therapeutic plasma exchange for coronavirus disease-2019 triggered cytokine release syndrome; a retrospective propensity matched control study.

Cytokine release syndrome (CRS) plays a pivotal role in the pathophysiology and progression of Coronavirus disease-2019 (COVID-19). Therapeutic plasma exchange (TPE) by removing the pathogenic cytokines is hypothesized to dampen CRS. To evaluate the outcomes of the patients with COVID-19 having CRS being treated with TPE compared to controls on the standard of care. Retrospective propensity score-matched analysis in a single centre from 1st April to 31st July 2020. We retrospectively analyzed data of 280 hospitalized patients developing CRS initially. PSM was used to minimize bias from non-randomized treatment assignment. Using PSM 1:1, 90 patients were selected and assigned to 2 equal groups. Forced matching was done for disease severity, routine standard care and advanced supportive care. Many other Co-variates were matched. Primary outcome was 28 days overall survival. Secondary outcomes were duration of hospitalization, CRS resolution time and timing of viral clearance on Polymerase chain reaction testing. After PS-matching, the selected cohort had a median age of 60 years (range 32-73 in TPE, 37-75 in controls), p = 0.325 and all were males. Median symptoms duration was 7 days (range 3-22 days' TPE and 3-20 days controls), p = 0.266. Disease severity in both groups was 6 (6.6%) moderate, 40 (44.4%) severe and 44 (49%) critical. Overall, 28-day survival was significantly superior in the TPE group (91.1%), 95% CI 78.33-97.76; as compared to PS-matched controls (61.5%), 95% CI 51.29-78.76 (log rank 0.002), p<0.001. Median duration of hospitalization was significantly reduced in the TPE treated group (10 days vs 15 days) (p< 0.01). CRS resolution time was also significantly reduced in the TPE group (6 days vs. 12 days) (p< 0.001). In 71 patients who underwent TPE, the mortality was 0 (n = 43) if TPE was done within the first 12 days of illness while it was 17.9% (deaths 5, n = 28 who received it after 12th day (p = 0.0045). An earlier use of TPE was associated with improved overall survival, early CRS resolution and time to discharge compared to SOC for COVID-19 triggered CRS in this selected cohort of PS-matched male patients from one major hospital in Pakistan.

Towards zero postoperative pancreatic fistula after distal pancreatectomy: a propensity score matching analysis of pancreatic stump closure with radiofrequency vs stapler

Purpose: To demonstrate the efficacy of radiofrequency for pancreatic stump closure in reducing the incidence of postoperative pancreatic fistula (POPF) in distal pancreatectomy (DP) compared with mechanical transection methods. Despite all the different techniques of pancreatic stump closure proposed for DP, best practice for avoiding POPF remains an unresolved issue, with an incidence of up to 30% regardless of center volume or surgical expertise. Method: DP was performed in a cohort of patients by applying radiofrequency to stump closure (RF Group) and compared with mechanical closure (Control Group). A propensity score (PS) matched cohort study was carried out to minimize bias from nonrandomized treatment assignment. Cohorts were matched by PS accounting for factors significantly associated with either undergoing RF transection or mechanical closure through logistic regression analysis. The primary end-point was the incidence of clinically relevant POPF (CR-POPF). Results: Of 89 patients included in the whole cohort, 13 case patients from the RF- Group were 1:1 matched to 13 control patients. In both the first independent analysis of unmatched data and subsequent adjustment to the overall propensity score-matched cohort, a higher rate of CR-POPF in the Control Group compared with the RF-Group was detected (25.4% vs. 5.3%, p=0.049 and 53.8% vs. 0%; p = 0.016 respectively). The RF Group showed better outcomes in terms of readmission rate (46.2% vs 0%, p=0.031). No significant differences were observed in terms of mortality, major complications (30.8% vs 0%, p=0.063) or length of hospital stay (5.7 vs 5.2 days, p=0.89). Conclusion: Findings suggest that the RF-assisted technique is more efficacious in reducing CR-POPF than mechanical pancreatic stump closure. I confirm the evaluation and review of the work of the ethics committee in accordance with Law 14/2007 on Biomedical Research, Principles of the Declaration of Helsinki, and other applicable ethical principles.

Evaluation of Adherence to Guideline for Heart Failure with Reduced Ejection Fraction in Heart Failure with Preserved Ejection Fraction and with or without Atrial Fibrillation.

BackgroundThis study evaluated the relationship between guideline adherence for heart failure (HF) with reduced ejection fraction (HFrEF) at discharge and relevant clinical outcomes in patients with acute HF with preserved ejection fraction (HFpEF) with or without atrial fibrillation (AF).MethodsWe analyzed Korean Acute Heart Failure Registry data for 707 patients with HFpEF with documented AF and 687 without AF. Guideline adherence was defined as good or poor according to the prescription of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, β-blockers, and mineralocorticoid receptor antagonists. Anticoagulation adherence was also incorporated for the AF group.ResultsAmong patients with normal sinus rhythm, those with poor guideline adherence had a reduced prevalence of comorbidities and favorable clinical characteristics when compared with those with good guideline adherence. Using inverse probability of treatment weighting (IPTW) to address the bias of nonrandom treatment assignment, good adherence was associated with a poor 60-day composite endpoint in the multivariable Cox model (weighted hazard ratio [wHR], 1.74; 95% confidence interval [CI], 1.01–3.00; P = 0.045). For patients with AF, baseline clinical characteristics were similar according to the degree of adherence. The IPTW-adjusted analysis indicated that good adherence was significantly associated with the 60-day composite endpoint (wHR, 0.47; 95% CI, 0.27–0.79; P = 0.005). In the analysis excluding warfarin, good adherence was associated with 60-day re-hospitalization (wHR, 0.60; 95% CI, 0.37–0.98; P = 0.040), 1-year re-hospitalization (wHR, 0.67; 95% CI, 0.48–0.93; P = 0.018), and the composite endpoint (wHR, 0.77; 95% CI, 0.59–0.99; P = 0.041).ConclusionOur findings indicate that good adherence to guidelines for HFrEF is associated with a better 60-day composite endpoint in patients with HFpEF with AF.