Understanding Causal Distributional and Subgroup Effects With the Instrumental Propensity Score.

To address issues with measured and unmeasured confounding in observational studies, we developed a unified approach to using an instrumental variable in more flexible ways to evaluate treatment effects. The approach is based on an instrumental propensity score conditional on baseline variables, which can then be incorporated in matching, regression, subclassification, or weighting along with various parametric, semiparametric, or nonparametric methods for the assessment of treatment effects. Therefore, the application of the instrumental propensity score allows different methods for outcome effect evaluations in addition to standard 2-stage least square models while controlling for unmeasured confounders. Several properties of the instrumental propensity score are discussed. The approach is then illustrated using subclassification along with a semiparametric density ratio model and empirical likelihood. This method allows us to evaluate distributional and subgroup treatment effects in addition to the overall average treatment effect. Simulation studies showed that the method works well. We applied our method to a study of the effects of attending a Catholic school versus a public school and found that attending a Catholic school had significant beneficial effects on subsequent wages among a subgroup of subjects.

The Fluorocycline TP-271 Is Efficacious in Models of Aerosolized Bacillus anthracis Infection in BALB/c Mice and Cynomolgus Macaques.

The fluorocycline TP-271 was evaluated in mouse and nonhuman primate (NHP) models of inhalational anthrax. BALB/c mice were exposed by nose-only aerosol to Bacillus anthracis Ames spores at a level of 18 to 88 lethal doses sufficient to kill 50% of exposed individuals (LD50). When 21 days of once-daily dosing was initiated at 24 h postchallenge (the postexposure prophylaxis [PEP] study), the rates of survival for the groups treated with TP-271 at 3, 6, 12, and 18 mg/kg of body weight were 90%, 95%, 95%, and 84%, respectively. When 21 days of dosing was initiated at 48 h postchallenge (the treatment [Tx] study), the rates of survival for the groups treated with TP-271 at 6, 12, and 18 mg/kg TP-271 were 100%, 91%, and 81%, respectively. No deaths of TP-271-treated mice occurred during the 39-day posttreatment observation period. In the NHP model, cynomolgus macaques received an average dose of 197 LD50 of B. anthracis Ames spore equivalents using a head-only inhalation exposure chamber, and once-daily treatment of 1 mg/kg TP-271 lasting for 14 or 21 days was initiated within 3 h of detection of protective antigen (PA) in the blood. No (0/8) animals in the vehicle control-treated group survived, whereas all 8 infected macaques treated for 21 days and 4 of 6 macaques in the 14-day treatment group survived to the end of the study (56 days postchallenge). All survivors developed toxin-neutralizing and anti-PA IgG antibodies, indicating an immunologic response. On the basis of the results obtained with the mouse and NHP models, TP-271 shows promise as a countermeasure for the treatment of inhalational anthrax.

The Fluorocycline TP-271 Is Efficacious in Models of Aerosolized Francisella tularensis SCHU S4 Infection in BALB/c Mice and Cynomolgus Macaques.

TP-271 is a novel, fully synthetic fluorocycline in development for complicated bacterial respiratory infections. TP-271 was active in vitro against a panel of 29 Francisella tularensis isolates, showing MICs against 50% and 90% of isolates of 0.25 and 0.5 µg/ml, respectively. In a mouse model of inhalational tularemia, animals were exposed by aerosol to 91 to 283 50% lethal doses (LD50)/mouse of F. tularensis SCHU S4. Following 21 days of once-daily intraperitoneal dosing with TP-271 at 3, 6, 12, and 18 mg/kg of body weight/day, initiating at 24 h postchallenge, survival was 80%, 100%, 100%, and 100%, respectively. When treatment was initiated at 72 h postchallenge, survival was 89%, 100%, 100%, and 100% in the 3-, 6-, 12-, and 18-mg/kg/day TP-271 groups, respectively. No mice treated with the vehicle control survived. Surviving mice treated with TP-271 showed little to no relapse during 14 days posttreatment. In a nonhuman primate model of inhalational tularemia, cynomolgus macaques received an average aerosol exposure of 1,144 CFU of F. tularensis SCHU S4. Once-daily intravenous infusion with 1 or 3 mg/kg TP-271, or vehicle control, for 21 days was initiated within 6 h of confirmed fever. All animals treated with TP-271 survived to the end of the study, with no relapse during 14 days after the last treatment, whereas no vehicle control-treated animals survived. The protection and low relapse afforded by TP-271 treatment in these studies support continued investigation of TP-271 for use in the event of aerosolized exposure to F. tularensis.

Improved outcomes using laparoscopy for emergency colectomy after mitigating bias by negative control exposure analysis.

BACKGROUND: Laparoscopy is superior to open surgery for elective colectomy, but its role in emergency colectomy remains unclear. Previous studies were small and limited by confounding because surgeons may have selected lower-risk patients for laparoscopy. We therefore studied the effect of attempting laparoscopy for emergency colectomies while adjusting for confounding using multiple techniques in a large, nationwide registry. METHODS: Using National Surgical Quality Improvement Program data, we identified emergency colectomy cases from 2014 to 2018. We first compared outcomes between patients who underwent laparoscopic versus open surgery, while adjusting for baseline variables using both propensity scores and regression. Next, we performed a negative control exposure analysis. By assuming that the group that converted to open did not benefit from the attempt at laparoscopy, we used the observed benefit to bound the effect of unmeasured confounding. RESULTS: Of 21,453 patients meeting criteria, 3,867 underwent laparoscopy, of which 1,375 converted to open. In both inverse probability of treatment weighting and regression analyses, attempting laparoscopy was associated with improved 30-day mortality, overall morbidity, anastomotic leak, surgical site infection, postoperative septic shock, and length of hospital stay compared with open surgery. These effects were consistent with the lower bounds computed from the converted group. CONCLUSION: Laparoscopic surgery for colorectal emergencies appears to improve outcomes compared with open surgery. The benefit is observed even after adjusting for both measured and unmeasured confounding using multiple statistical approaches, thus suggesting a benefit not attributable to patient selection.

Increased mortality with resuscitative endovascular balloon occlusion of the aorta only mitigated by strong unmeasured confounding: An expanded analysis using the National Trauma Data Bank.

BACKGROUND: Resuscitative endovascular balloon occlusion of the aorta (REBOA) is being increasingly adopted to manage noncompressible torso hemorrhage, but a recent analysis of the 2015 to 2016 Trauma Quality Improvement Project (TQIP) data set showed that placement of REBOA was associated with higher rates of death, lower extremity amputation, and acute kidney injury (AKI). We expand this analysis by including the 2017 data set, quantifying the potential role of residual confounding, and distinguishing between traumatic and ischemic lower extremity amputation. METHODS: This retrospective study used the 2015 to 2017 TQIP database and included patients older than 18 years, with signs of life on arrival, who had no aortic injury and were not transferred. Resuscitative endovascular balloon occlusions of the aorta placed after 2 hours were excluded. We adjusted for baseline variables using propensity scores with inverse probability of treatment weighting. A sensitivity analysis was then conducted to determine the strength of an unmeasured confounder (e.g., unmeasured shock severity/response to resuscitation) that could explain the effect on mortality. Finally, lower extremity injury patterns of patients undergoing REBOA were inspected to distinguish amputation indicated for traumatic injury from complications of REBOA placement. RESULTS: Of 1,392,482 patients meeting the inclusion criteria, 187 underwent REBOA. After inverse probability of treatment weighting, all covariates were balanced. The risk difference for mortality was 0.21 (0.14-0.29) and for AKI was 0.041 (-0.007 to 0.089). For the mortality effect to be explained by an unmeasured confounder, it would need to be stronger than any observed in terms of its relationship with mortality and with REBOA placement. Eleven REBOA patients underwent lower extremity amputation; however, they all suffered severe traumatic injury to the lower extremity. CONCLUSION: There is no evidence in the TQIP data set to suggest that REBOA causes amputation, and the evidence for its effect on AKI is considerably weaker than previously reported. The increased mortality effect of REBOA is confirmed and could only be nullified by a potent confounder. LEVEL OF EVIDENCE: Therapeutic/care management, level IV.

Optimal Allocation of Interviews to Baseline and Endline Surveys in Place-Based Randomized Trials and Quasi-Experiments.

BACKGROUND: Many place-based randomized trials and quasi-experiments use a pair of cross-section surveys, rather than panel surveys, to estimate the average treatment effect of an intervention. In these studies, a random sample of individuals in each geographic cluster is selected for a baseline (preintervention) survey, and an independent random sample is selected for an endline (postintervention) survey. OBJECTIVE: This design raises the question, given a fixed budget, how should a researcher allocate resources between the baseline and endline surveys to maximize the precision of the estimated average treatment effect? RESULTS: We formalize this allocation problem and show that although the optimal share of interviews allocated to the baseline survey is always less than one-half, it is an increasing function of the total number of interviews per cluster, the cluster-level correlation between the baseline measure and the endline outcome, and the intracluster correlation coefficient. An example using multicountry survey data from Africa illustrates how the optimal allocation formulas can be combined with data to inform decisions at the planning stage. Another example uses data from a digital political advertising experiment in Texas to explore how precision would have varied with alternative allocations.

A "placement of death" approach for studies of treatment effects on ICU length of stay.

Length of stay in the intensive care unit (ICU) is a common outcome measure in randomized trials of ICU interventions. Because many patients die in the ICU, it is difficult to disentangle treatment effects on length of stay from effects on mortality; conventional analyses depend on assumptions that are often unstated and hard to interpret or check. We adapt a proposal from Rosenbaum that addresses concerns about selection bias and makes its assumptions explicit. A composite outcome is constructed that equals ICU length of stay if the patient was discharged alive and indicates death otherwise. Given any preference ordering that compares death with possible lengths of stay, we can estimate the intervention's effects on the composite outcome distribution. Sensitivity analyses can show results for different preference orderings. We discuss methods for constructing approximate confidence intervals for treatment effects on quantiles of the outcome distribution or on proportions of patients with outcomes preferable to various cutoffs. Strengths and weaknesses of possible primary significance tests (including the Wilcoxon-Mann-Whitney rank sum test and a heteroskedasticity-robust variant due to Brunner and Munzel) are reviewed. An illustrative example reanalyzes a randomized trial of an ICU staffing intervention.

Expression of estrogen receptor-alpha and cFos in norepinephrine and epinephrine neurons of young and middle-aged rats during the steroid-induced luteinizing hormone surge.

Norepinephrine (NE) and epinephrine are important stimulators of GnRH release during the preovulatory surge in female rats. Previous studies have shown that the catecholaminergic neurons are sensitive to estradiol and that NE release in the hypothalamus is decreased in middle-aged rats at the time when the estrous cycles become irregular and later cease to exist. The aims of the present study were to determine whether the NE and epinephrine neurons continue to express estrogen receptor (ER)-alpha in middle-aged rats; temporal expression of ER-alpha and cFos changes with age during the steroid-induced surge; and tyrosine hydroxylase (TH), dopamine-beta-hydroxylase (DBH), and phenylethanol-N-methyltransferase mRNA content in catecholaminergic neurons of the brain stem changes during the surge with age. The results show that there was no difference in TH mRNA content; however, DBH mRNA levels in areas A1, A2, and C1 of the middle-aged animals did not rise during the surge as was observed in the young animals. Although the percentage of NE and epinephrine neurons that express ER-alpha was unchanged during the surge in both young and middle-aged animals, cFos expression was enhanced in areas A1 and A2 of the middle-aged animals but not in the young animals. Together the results suggest that NE and epinephrine neurons in the middle-aged rat continue to express appropriate basal levels of TH, DBH, and phenylethanol-N-methyltransferase mRNAs as well as ER-alpha and cFos; however, the enhancement of DBH expression, as seen in the young animals during the steroid-induced surge, was not detected in middle-aged animals. On the other hand, cFos expression in the middle-aged rat was higher in areas A1 and A2 during the surge. It is concluded that the reduced catecholamine release during the surge in middle-aged rats is caused, in part, by an altered sensitivity of the NE neurons to estradiol, which results in an aberrant cFos expression and probably not by major deficits in the expression of transmitter synthesizing enzymes or steroid receptors.

Distribution of vesicular glutamate transporter-2 messenger ribonucleic Acid and protein in the septum-hypothalamus of the rat.

The excitatory neurotransmitter glutamate is involved in the control of most, perhaps all, neuroendocrine systems, yet the sites of glutamatergic neurons and their processes are unknown. Here, we used in situ hybridization and immunohistochemistry for the neuron-specific vesicular glutamate transporter-2 (VGLUT2) to identify the neurons in female rats that synthesize the neurotransmitter glutamate as well as their projections throughout the septum-hypothalamus. The results show that glutamatergic neurons are present in the septum-diagonal band complex and throughout the hypothalamus. The preoptic area and ventromedial and dorsomedial nuclei are particularly rich in glutamatergic neurons, followed by the supraoptic, paraventricular, and arcuate nuclei, whereas the suprachiasmatic nucleus does not express detectable amounts of VGLUT2 mRNA. Immunoreactive neurites are seen in very high densities in all regions analyzed, particularly in the preoptic region, followed by the ventromedial, dorsomedial, and arcuate nuclei as well as the external layer of the median eminence, whereas the mammillary complex does not exhibit VGLUT2 immunoreactivity. Many VGLUT2 immunoreactive fibers also contained synaptophysin, suggesting that the transporter is indeed localized to presynaptic terminals. Together, the results identify glutamatergic cell bodies throughout the septum-hypothalamus in region-specific patterns and show that glutamatergic nerve terminals are present in very large numbers such that most neurons in these brain regions can receive glutamatergic input. We examined the GnRH system as an example of a typical neuroendocrine system and could show that the GnRH perikarya are closely apposed by many VGLUT2-immunoreactive boutons, some of which also contained synaptophysin. The presence of VGLUT2 mRNA-containing cells in specific nuclei of the hypothalamus indicates that many neuroendocrine neurons coexpress glutamate as neurotransmitter, in addition to neuropeptides. These systems include the oxytocin, vasopressin, or CRH neurons as well as many others in the periventricular and mediobasal hypothalamus. The presence of VGLUT2 mRNA in steroid-sensitive regions of the hypothalamus, such as the anteroventral periventricular, paraventricular, or ventromedial nuclei indicates that gonadal and adrenal steroid can directly alter the functions of these glutamatergic neurons.

Modulation of decompression sickness risk in pigs with caffeine during H(2) biochemical decompression.

In H(2) biochemical decompression, H(2)-metabolizing intestinal microbes remove gas stored in tissues of animals breathing hyperbaric H(2), thereby reducing decompression sickness (DCS) risk. We hypothesized that increasing intestinal perfusion in pigs would increase the activity of intestinal Methanobrevibacter smithii, lowering DCS incidence further. Pigs (Sus scrofa, 17-23 kg, n = 20) that ingested caffeine (5 mg/kg) increased O(2) consumption rate in 1 atm air by ~20% for at least 3 h. Pigs were given caffeine alone or caffeine plus injections of M. smithii. Animals were compressed to 24 atm (20.5-23.1 atm H(2), 0.3-0.5 atm O(2)) for 3 h, then decompressed and observed for signs of DCS. In previous studies, DCS incidence in animals without caffeine treatment was significantly (P < 0.05) lower with M. smithii injections (7/16) than in controls (9/10). However, contrary to our hypothesis, DCS incidence was marginally higher (P = 0.057) in animals that received caffeine and M. smithii (9/10) than in animals that received caffeine but no M. smithii (4/10). More information on gas kinetics is needed before extending H(2) biochemical decompression to humans.

Establishment of a Swiss Webster mouse model of pneumonic plague to meet essential data elements under the animal rule.

A recombinant vaccine (rF1V) is being developed for protection against pneumonic plague. This study was performed to address essential data elements to establish a well-characterized Swiss Webster mouse model for licensing the rF1V vaccine using the FDA's Animal Rule. These elements include the documentation of challenge material characteristics, aerosol exposure parameters, details of the onset and severity of clinical signs, pathophysiological response to disease, and relevance to human disease. Prior to animal exposures, an evaluation of the aerosol system was performed to determine and understand the variability of the aerosol exposure system. Standardized procedures for the preparation of Yersinia pestis challenge material also were developed. The 50% lethal dose (LD(50)) was estimated to be 1,966 CFU using Probit analysis. Following the LD(50) determination, pathology was evaluated by exposing mice to a target LD(99) (42,890 CFU). Mice were euthanized at 12, 24, 36, 48, 60, and 72 h postexposure. At each time point, samples were collected for clinical pathology, detection of bacteria in blood and tissues, and pathology evaluations. A general increase in incidence and severity of microscopic findings was observed in the lung, lymph nodes, spleen, and liver from 36 to 72 h postchallenge. Similarly, the incidence and severity of pneumonia increased throughout the study; however, some mice died in the absence of pneumonia, suggesting that disease progression does not require the development of pneumonia. Disease pathology in the Swiss Webster mouse is similar to that observed in humans, demonstrating the utility of this pneumonic plague model that can be used by researchers investigating plague countermeasures.

Protection in mice passively immunized with serum from cynomolgus macaques and humans vaccinated with recombinant plague vaccine (rF1V).

Passive transfer models were developed to evaluate the ability of antibodies generated in cynomolgus macaques and humans vaccinated with a recombinant plague vaccine (rF1V) to protect naïve Swiss Webster mice against pneumonic plague. Development of the passive transfer model is intended to support clinical and nonclinical development of the rF1V vaccine. To evaluate protection, unfractionated serum collected from rF1V vaccinated cynomolgus macaques and human volunteers with known antibody titers to rF1, rV and rF1V was transferred into naïve Swiss Webster mice via the intraperitoneal route. Results of these studies demonstrated that passive immunization protected mice from challenge or extended mean survival time and that the passive transfer assay can be used to evaluate the functional role of antibodies induced by rF1V vaccination in protection against aerosol exposure.

DNA microarray and proteomic analyses of the RpoS regulon in Geobacter sulfurreducens.

The regulon of the sigma factor RpoS was defined in Geobacter sulfurreducens by using a combination of DNA microarray expression profiles and proteomics. An rpoS mutant was examined under steady-state conditions with acetate as an electron donor and fumarate as an electron acceptor and with additional transcriptional profiling using Fe(III) as an electron acceptor. Expression analysis revealed that RpoS acts as both a positive and negative regulator. Many of the RpoS-dependent genes determined play roles in energy metabolism, including the tricarboxylic acid cycle, signal transduction, transport, protein synthesis and degradation, and amino acid metabolism and transport. As expected, RpoS activated genes involved in oxidative stress resistance and adaptation to nutrient limitation. Transcription of the cytochrome c oxidase operon, necessary for G. sulfurreducens growth using oxygen as an electron acceptor, and expression of at least 13 c-type cytochromes, including one previously shown to participate in Fe(III) reduction (MacA), were RpoS dependent. Analysis of a subset of the rpoS mutant proteome indicated that 15 major protein species showed reproducible differences in abundance relative to those of the wild-type strain. Protein identification using mass spectrometry indicated that the expression of seven of these proteins correlated with the microarray data. Collectively, these results indicate that RpoS exerts global effects on G. sulfurreducens physiology and that RpoS is vital to G. sulfurreducens survival under conditions typically encountered in its native subsurface environments.

The importance of porE and porF in the anabolic pyruvate oxidoreductase of Methanococcus maripaludis.

The operon of the anabolic pyruvate oxidoreductase (POR) of Methanococcus maripaludis encodes two genes ( porEF) whose functions are unknown. Because these genes possess sequence similarity to polyferredoxins, they may be electron carriers to the POR. To elucidate whether the methanococcal POR requires PorEF for activity, a deletion mutant, strain JJ150, lacking porEF was constructed. Compared to the wild-type strain JJ1, the mutant grew more slowly in minimal medium and minimal plus acetate medium, and pyruvate-dependent methanogenesis was inhibited. In contrast, the methyl-viologen-dependent pyruvate-oxidation activity of POR, carbon monoxide dehydrogenase, and hydrogenase activities of the mutant were similar to those of the wild-type. Upon genetic complementation of the mutant with porEF in the methanococcal shuttle vector pMEV2+ porEF, growth in minimal medium and pyruvate-dependent methanogenesis were restored to wild-type levels. Complementation with porE alone restored methanogenesis from pyruvate but not growth in minimal medium. Complementation with porF alone partially restored growth but not methanogenesis from pyruvate. Although the specific roles of porE and porF have not been determined, these results suggest that PorEF play important roles in the anabolic POR in vivo even though they are not required for the dye-dependent activity.

The anabolic pyruvate oxidoreductase from Methanococcus maripaludis.

In autotrophic methanogens, pyruvate oxidoreductase (POR) plays a key role in the assimilation of CO(2) and the biosynthesis of organic carbon. This enzyme has been purified to homogeneity, and the genes from Methanococcus maripaludis were sequenced. The purified POR contained five polypeptides with molecular masses of 47, 33, 25, 21.5 and 13 kDa. The N-terminal sequences of four of the polypeptides had high similarity to the subunits commonly associated with this enzyme from other archaea. However, the 21.5-kDa polypeptide had not been previously observed in PORs. Nucleotide sequencing of the gene cluster encoding the POR revealed six open reading frames ( porABCDEF). The genes porABCD corresponded to the subunits previously identified in PORs. On the basis of the N-terminal amino acid sequence, porE encoded the 21.5-kDa polypeptide and contained a high cysteinyl residue content and a motif indicative of a [Fe-S] cluster. porF also had a high sequence similarity to porE, a high cysteinyl residue content, and two [Fe-S] cluster motifs. Homologs to porE were also present in the genomic sequences of the autotrophic methanogens Methanocaldococcus jannaschii and Methanothermobacter thermautotrophicus. Based upon these results, it is proposed that PorE and PorF are components of a specialized system required to transfer low-potential electrons for pyruvate biosynthesis. Some biochemical properties of the purified methanococcal POR were also determined. This unstable enzyme was very sensitive to O(2 )and demonstrated high activity with pyruvate, oxaloacetate, and alpha-ketobutyrate. Methyl viologen, rubredoxin, FMN, and FAD were readily reduced. Activity was also observed with spinach and clostridial ferredoxins and cytochrome c. Coenzyme F(420) was not an electron acceptor for the purified enzyme.