Propensity score specification for optimal estimation of average treatment effect with binary response.

Propensity score methods are commonly used in statistical analyses of observational data to reduce the impact of confounding bias in estimations of average treatment effect. While the propensity score is defined as the conditional probability of a subject being in the treatment group given that subject's covariates, the most precise estimation of average treatment effect results from specifying the propensity score as a function of true confounders and predictors only. This property has been demonstrated via simulation in multiple prior research articles. However, we have seen no theoretical explanation as to why this should be so. This paper provides that theoretical proof. Furthermore, this paper presents a method for performing the necessary variable selection by means of elastic net regression, and then estimating the propensity scores so as to obtain optimal estimates of average treatment effect. The proposed method is compared against two other recently introduced methods, outcome-adaptive lasso and covariate balancing propensity score. Extensive simulation analyses are employed to determine the circumstances under which each method appears most effective. We applied the proposed methods to examine the effect of pre-cardiac surgery coagulation indicator on mortality based on a linked dataset from a retrospective review of 1390 patient medical records at Jewish Hospital (Louisville, KY) with the Society of Thoracic Surgeons database.

Keratin 18 Is a Diagnostic and Prognostic Factor for Acute Alcoholic Hepatitis.

BACKGROUND & AIMS: Acute alcoholic hepatitis (AAH) is a major cause of liver-related morbidity and mortality; there are no good blood biomarkers for diagnosis or determining magnitude of cell death. Keratin 18 (KRT18, also called K18), found in epithelial cells, is released from hepatocytes upon death. We investigated whether level of K18 is a better marker of hepatocyte death than standard biomarkers and might be used to identify patients with AAH at risk for death within 90 days. METHODS: We analyzed data from 173 participants in a large trial performed at 4 medical centers. Participants with AAH were classified as severe (n = 57, model for end-stage liver disease [MELD] scores above 20) or moderate (n = 27, MELD scores from 12 to 19); 38 participants had alcohol use disorder with mild (n = 28) or no liver injury (n = 10); 34 participants had nonalcoholic steatohepatitis; and 17 participants were healthy (controls). We quantified serum levels of K18 using ELISAs and APOPTOSENSE kits. RESULTS: Serum level of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and the ratio of AST:ALT did not correlate with MELD scores. Patients with alcohol use disorder had higher serum levels of ALT than patients with severe AAH. Levels of K18M65 and K18M30 had statistically significant increases as liver disease worsened, as did the degree of necrosis (ratio of K18 M65:M30). The ratio of K18M65:ALT was increased in serum from patients with AAH compared with controls. Serum levels of K18 identified patients who died within 90 days with greater accuracy than commonly used static biomarkers. CONCLUSIONS: There is a stronger association between serum level of keratin 18 and amount of hepatocyte death and liver disease severity than for other biomarkers (AST, ALT, and the AST:ALT ratio). The ratio of K18M65:M30 might be used as marker of mechanism of hepatocyte death, and the ratio of K18M65:ALT might be used to distinguish patients with AAH from patients with nonalcoholic steatohepatitis. Serum levels of K18 might be used to identify patients with severe AAH at risk for death. ClinicalTrials.gov identifier # NCT01922895 and NCT01809132.

Continuous infusion of low-dose unfractionated heparin after aneurysmal subarachnoid hemorrhage: a preliminary study of cognitive outcomes.

OBJECTIVECognitive dysfunction occurs in up to 70% of aneurysmal subarachnoid hemorrhage (aSAH) survivors. Low-dose intravenous heparin (LDIVH) infusion using the Maryland protocol was recently shown to reduce clinical vasospasm and vasospasm-related infarction. In this study, the Montreal Cognitive Assessment (MoCA) was used to evaluate cognitive changes in aSAH patients treated with the Maryland LDIVH protocol compared with controls.METHODSA retrospective analysis of all patients treated for aSAH between July 2009 and April 2014 was conducted. Beginning in 2012, aSAH patients were treated with LDIVH in the postprocedural period. The MoCA was administered to all aSAH survivors prospectively during routine follow-up visits, at least 3 months after aSAH, by trained staff blinded to treatment status. Mean MoCA scores were compared between groups, and regression analyses were performed for relevant factors.RESULTSNo significant differences in baseline characteristics were observed between groups. The mean MoCA score for the LDIVH group (n = 25) was 26.4 compared with 22.7 in controls (n = 22) (p = 0.013). Serious cognitive impairment (MoCA ≤ 20) was observed in 32% of controls compared with 0% in the LDIVH group (p = 0.008). Linear regression analysis demonstrated that only LDIVH was associated with a positive influence on MoCA scores (ß = 3.68, p =0.019), whereas anterior communicating artery aneurysms and fevers were negatively associated with MoCA scores. Multivariable linear regression analysis resulted in all 3 factors maintaining significance. There were no treatment complications.CONCLUSIONSThis preliminary study suggests that the Maryland LDIVH protocol may improve cognitive outcomes in aSAH patients. A randomized controlled trial is needed to determine the safety and potential benefit of unfractionated heparin in aSAH patients.