Comparative Effectiveness of New Approaches to Improve Mortality Risk Models From Medicare Claims Data

IMPORTANCE

Risk adjustment models using claims-based data are central in evaluating health care performance. Although US Centers for Medicare & Medicaid Services (CMS) models apply well-vetted statistical approaches, recent changes in the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) coding system and advances in computational capabilities may provide an opportunity for enhancement.

OBJECTIVE

To examine whether changes using already available data would enhance risk models and yield greater discrimination in hospital-level performance measures.

DESIGN, SETTING, AND PARTICIPANTS

This comparative effectiveness study used ICD-9-CM codes from all Medicare fee-for-service beneficiary claims for hospitalizations for acute myocardial infarction (AMI), heart failure (HF), or pneumonia among patients 65 years and older from July 1, 2013, through September 30, 2015. Changes to current CMS mortality risk models were applied incrementally to patient-level models, and the best model was tested on hospital performance measures to model 30-day mortality. Analyses were conducted from April 19, 2018, to September 19, 2018.

MAIN OUTCOMES AND MEASURES

The main outcome was all-cause death within 30 days of hospitalization for AMI, HF, or pneumonia, examined using 3 changes to current CMS mortality risk models: (1) incorporating present on admission coding to better exclude potential complications of care, (2) separating index admission diagnoses from those of the 12-month history, and (3) using ungrouped ICD-9-CM codes.

RESULTS

There were 361 175 hospital admissions (mean [SD] age, 78.6 [8.4] years; 189 225 [52.4%] men) for AMI, 716 790 hospital admissions (mean [SD] age, 81.1 [8.4] years; 326 825 [45.6%] men) for HF, and 988 225 hospital admissions (mean [SD] age, 80.7 [8.6] years; 460 761 [46.6%] men) for pneumonia during the study; mean 30-day mortality rates were 13.8% for AMI, 12.1% for HF, and 16.1% for pneumonia. Each change to the models was associated with incremental gains in C statistics. The best model, incorporating all changes, was associated with significantly improved patient-level C statistics, from 0.720 to 0.826 for AMI, 0.685 to 0.776 for HF, and 0.715 to 0.804 for pneumonia. Compared with current CMS models, the best model produced wider predicted probabilities with better calibration and Brier scores. Hospital risk-standardized mortality rates had wider distributions, with more hospitals identified as good or bad performance outliers.

CONCLUSIONS AND RELEVANCE

Incorporating present on admission coding and using ungrouped index and historical ICD-9-CM codes were associated with improved patient-level and hospital-level risk models for mortality compared with the current CMS models for all 3 conditions.

Study of Rheology and Plug Assist Thermoforming of Linear and Branched PP Homopolymer and Impact Copolymer

Polypropylene (PP) is one of the fastest growing thermoplastic polymers in the world, second only to polyethylene. This is primarily due to its excellent balance of physical and chemical properties at a lower cost. PP however possesses low melt strength on account of its linear structure and hence is not easily amenable to processing techniques that involve free surface stretching deformations like thermoforming, blow molding and extrusion film casting. One way to enhance the melt strength of PP is to incorporate long chain branches in its molecular architecture. The present study focuses on the impact of rheology of linear and branched PP on their thermoforming characteristics. Two grades each of linear and long chain branched (LCB) PP homopolymer and impact copolymer (ICP) were used. It was observed that the LCB-PP homopolymer and LCB-ICP showed higher flow activation energy, reduced value of loss tangent and nearly equal frequency dependence of storage and loss moduli in shear rheology. Also, a strong strain hardening behavior was displayed in extensional rheology by the LCB grades. Plug assist thermoforming experiments were carried out to assess the effect of long chain branching on surface strain and thickness distribution for axisymmetric cups of two draw ratios. Biaxial surface strain maps of the formed cups were quantified using Grid Strain Analysis (GSA). Thermoformed cups made from LCB-PP homopolymer and LCB-impact copolymer showed lower surface strain and overall higher thickness as compared to cups made from their linear counterparts, which is in accordance with what might be expected from their rheology.

What do we actually see in intracellular SERS? Investigating nanosensor-induced variation

Plasmonic nanoparticles (NPs), predominantly gold (AuNPs), are easily internalised into cells and commonly employed as nanosensors for reporter-based and reporter-free intracellular SERS applications. While AuNPs are generally considered non-toxic to cells, many biological and toxicity studies report that exposure to NPs induces cell stress through the generation of reactive oxygen species (ROS) and the upregulated transcription of pro-inflammatory genes, which can result in severe genotoxicity and apoptosis. Despite this, the extent to which normal cellular metabolism is affected by AuNP internalisation remains a relative unknown along with the contribution of the uptake itself to the SERS spectra obtained from within so called 'healthy' cells, as indicated by traditional viability tests. This work aims to interrogate the perturbation created by treatment with AuNPs under different conditions and the corresponding effect on the SERS spectra obtained. We characterise the changes induced by varying AuNP concentrations and medium serum compositions using biochemical assays and correlate them to the corresponding intracellular reporter-free SERS spectra. The different serum conditions lead to different extents of nanoparticle internalisation. We observe that changes in SERS spectra are correlated to an increasing amount of internalisation, confirmed qualitatively and quantitatively by confocal imaging and ICP-MS analysis, respectively. We analyse spectra and characterise changes that can be attributed to nanoparticle induced changes. Thus, our study highlights a need for understanding condition-dependent NP-cell interactions and standardisation of nanoparticle treatments in order to establish the validity of intracellular SERS experiments for use in all arising applications.

Enhancing the prediction of acute kidney injury risk after percutaneous coronary intervention using machine learning techniques: A retrospective cohort study

BACKGROUND

The current acute kidney injury (AKI) risk prediction model for patients undergoing percutaneous coronary intervention (PCI) from the American College of Cardiology (ACC) National Cardiovascular Data Registry (NCDR) employed regression techniques. This study aimed to evaluate whether models using machine learning techniques could significantly improve AKI risk prediction after PCI.

METHODS AND FINDINGS

We used the same cohort and candidate variables used to develop the current NCDR CathPCI Registry AKI model, including 947,091 patients who underwent PCI procedures between June 1, 2009, and June 30, 2011. The mean age of these patients was 64.8 years, and 32.8% were women, with a total of 69,826 (7.4%) AKI events. We replicated the current AKI model as the baseline model and compared it with a series of new models. Temporal validation was performed using data from 970,869 patients undergoing PCIs between July 1, 2016, and March 31, 2017, with a mean age of 65.7 years; 31.9% were women, and 72,954 (7.5%) had AKI events. Each model was derived by implementing one of two strategies for preprocessing candidate variables (preselecting and transforming candidate variables or using all candidate variables in their original forms), one of three variable-selection methods (stepwise backward selection, lasso regularization, or permutation-based selection), and one of two methods to model the relationship between variables and outcome (logistic regression or gradient descent boosting). The cohort was divided into different training (70%) and test (30%) sets using 100 different random splits, and the performance of the models was evaluated internally in the test sets. The best model, according to the internal evaluation, was derived by using all available candidate variables in their original form, permutation-based variable selection, and gradient descent boosting. Compared with the baseline model that uses 11 variables, the best model used 13 variables and achieved a significantly better area under the receiver operating characteristic curve (AUC) of 0.752 (95% confidence interval [CI] 0.749-0.754) versus 0.711 (95% CI 0.708-0.714), a significantly better Brier score of 0.0617 (95% CI 0.0615-0.0618) versus 0.0636 (95% CI 0.0634-0.0638), and a better calibration slope of observed versus predicted rate of 1.008 (95% CI 0.988-1.028) versus 1.036 (95% CI 1.015-1.056). The best model also had a significantly wider predictive range (25.3% versus 21.6%, p < 0.001) and was more accurate in stratifying AKI risk for patients. Evaluated on a more contemporary CathPCI cohort (July 1, 2015-March 31, 2017), the best model consistently achieved significantly better performance than the baseline model in AUC (0.785 versus 0.753), Brier score (0.0610 versus 0.0627), calibration slope (1.003 versus 1.062), and predictive range (29.4% versus 26.2%). The current study does not address implementation for risk calculation at the point of care, and potential challenges include the availability and accessibility of the predictors.

CONCLUSIONS

Machine learning techniques and data-driven approaches resulted in improved prediction of AKI risk after PCI. The results support the potential of these techniques for improving risk prediction models and identification of patients who may benefit from risk-mitigation strategies.

D-Meson Azimuthal Anisotropy in Midcentral Pb-Pb Collisions at sqrt[s]_{NN}=5.02 TeV

The azimuthal anisotropy coefficient v_{2} of prompt D^{0}, D^{+}, D^{*+}, and D_{s}^{+} mesons was measured in midcentral (30%-50% centrality class) Pb-Pb collisions at a center-of-mass energy per nucleon pair sqrt[s_{NN}]=5.02  TeV, with the ALICE detector at the LHC. The D mesons were reconstructed via their hadronic decays at midrapidity, |y|<0.8, in the transverse momentum interval 1<p_{T}<24  GeV/c. The measured D-meson v_{2} has similar values as that of charged pions. The D_{s}^{+} v_{2}, measured for the first time, is found to be compatible with that of nonstrange D mesons. The measurements are compared with theoretical calculations of charm-quark transport in a hydrodynamically expanding medium and have the potential to constrain medium parameters.

Nitrosative stress indices in dogs with neurological form of canine distemper

This is a prospective, controlled, randomized clinical study to evidence the role of nitrosative stress in development of overt neurological sign of canine distemper (CD). The enrollment of cases wasa made on basis of overt sign of CD (n=139) and the findings were compared with apparently healthy dogs (n=15). The CD specific immunoglobulins resulted in 94 confirmed positive cases. The nitric oxide (NO) and neuronal nitric oxide synthase (nNOS) concentration in cerebrospinal fluid (CSF) (18.08 ± 2.76, 415.84 ± 46.24, respectively) and plasma (32.68 ± 4.31, 321.31 ± 102.30, respectively) were significantly (P<0.05) elevated as compared to healthy control group. The concentration of neuron specific enolase (NSE) in CSF and plasma were also significantly (P<0.05) higher in dogs suffering from CD. The significant differences in other biochemical parameters like total protein, albumin and globulin were found in the CSF of dogs compared with healthy control. The author concludes that NO plays a role in pathophysiology of neurological form of CD and nNOS and NSE estimation in CSF and plasma could help in making early diagnosis of clinical cases.

Anomalous Evolution of the Near-Side Jet Peak Shape in Pb-Pb Collisions at sqrt[s_{NN}]=2.76 TeV

The measurement of two-particle angular correlations is a powerful tool to study jet quenching in a p_{T} region inaccessible by direct jet identification. In these measurements pseudorapidity (Δη) and azimuthal (Δφ) differences are used to extract the shape of the near-side peak formed by particles associated with a higher p_{T} trigger particle (1<p_{T,trig}<8  GeV/c). A combined fit of the near-side peak and long-range correlations is applied to the data allowing the extraction of the centrality evolution of the peak shape in Pb-Pb collisions at sqrt[s_{NN}]=2.76  TeV. A significant broadening of the peak in the Δη direction at low p_{T} is found from peripheral to central collisions, which vanishes above 4  GeV/c, while in the Δφ direction the peak is almost independent of centrality. For the 10% most central collisions and 1<p_{T,assoc}<2  GeV/c, 1<p_{T,trig}<3  GeV/c a novel feature is observed: a depletion develops around the center of the peak. The results are compared to pp collisions at the same center of mass energy and ampt model simulations. The comparison to the investigated models suggests that the broadening and the development of the depletion is connected to the strength of radial and longitudinal flow.

Traumatic stress-induced persistent changes in DNA methylation regulate neuropeptide Y expression in rat jejunum

BACKGROUND

Stress-induced chronic neuropsychiatric conditions such as anxiety are often co-morbid with gastrointestinal malfunctions. While we find enduring anxiety-like symptoms following minimal traumatic brain injury (MTBI) in rats, gastrointestinal consequences of MTBI remain elusive.

METHODS

In this study, we examined the effects of MTBI on a major gut peptide, neuropeptide Y (NPY) and gut motility. DNA methylation was studied as a possible epigenetic mechanism operative in the regulation of NPY expression in the gut.

KEY RESULTS

Minimal traumatic brain injury reduced the gut motility 48 hours and 30 days after trauma. The expression of DNA methyltransferase isoforms (DNMT1, DNMT3a, and DNMT3b) was altered in the jejunum 48 hours and 30 days after MTBI. However, the mRNA levels of growth arrest and DNA damage 45 (GADD45) isoforms, GADD45a, and GADD45b, which are believed to be involved in active DNA demethylation, initially decreased at 48 hours but subsequently increased after 30 days of trauma. Similarly, DNA hypomethylation at the NPY promoter region in the jejunum was correlated with the increase in NPY mRNA and protein levels 30 days post-trauma. On the other hand, DNA hypomethylation at 48 hours was associated with a decline in NPY expression. Treatment with 5-azacytidine (5-AzaC), a DNMT inhibitor, retarded DNA methylation and restored the NPY mRNA levels in the jejunum of MTBI-induced rats.

CONCLUSIONS & INFERENCES

These results suggest that DNA demethylation could be operative as an epigenetic mechanism in the long-term regulation of NPY gene expression to alter the gut motility during traumatic stress.

Energy dependence of forward-rapidity [Formula: see text] and [Formula: see text] production in pp collisions at the LHC

We present results on transverse momentum ([Formula: see text]) and rapidity ([Formula: see text]) differential production cross sections, mean transverse momentum and mean transverse momentum square of inclusive [Formula: see text] and [Formula: see text] at forward rapidity ([Formula: see text]) as well as [Formula: see text]-to-[Formula: see text] cross section ratios. These quantities are measured in pp collisions at center of mass energies [Formula: see text] and 13 TeV with the ALICE detector. Both charmonium states are reconstructed in the dimuon decay channel, using the muon spectrometer. A comprehensive comparison to inclusive charmonium cross sections measured at [Formula: see text], 7 and 8 TeV is performed. A comparison to non-relativistic quantum chromodynamics and fixed-order next-to-leading logarithm calculations, which describe prompt and non-prompt charmonium production respectively, is also presented. A good description of the data is obtained over the full [Formula: see text] range, provided that both contributions are summed. In particular, it is found that for [Formula: see text] GeV/c the non-prompt contribution reaches up to 50% of the total charmonium yield.

Production of Σ(1385)± and Ξ(1530)0 in p-Pb collisions at sNN=5.02 TeV

The transverse momentum distributions of the strange and double-strange hyperon resonances (Σ ( 1385 ) ± , Ξ ( 1530 ) 0 ) produced in p-Pb collisions at s NN = 5.02 TeV were measured in the rapidity range - 0.5 < y CMS < 0 for event classes corresponding to different charged-particle multiplicity densities, ⟨ dN ch /dη lab ⟩ . The mean transverse momentum values are presented as a function of ⟨ dN ch /dη lab ⟩ , as well as a function of the particle masses and compared with previous results on hyperon production. The integrated yield ratios of excited to ground-state hyperons are constant as a function of ⟨ dN ch /dη lab ⟩ . The equivalent ratios to pions exhibit an increase with ⟨ dN ch /dη lab ⟩ , depending on their strangeness content.

Macronutrients in Breastmilk of Mothers of Preterm Infants

OBJECTIVE

To evaluate the variability in macronutrient and energy content of breastmilk of Indian women delivering at ≤34 weeks of gestation.

METHODS

In this cross-sectional study, samples of breastmilk expressed manually for feeding of preterm neonates were collected from 106 mothers at 3±1 (n=26), 7±2 (n=34), 14±2 (n=24), 21±3 (n=12) and 28±3 (n=10) days after birth. Protein, fat and carbohydrate content were estimated and total energy content was calculated.

RESULTS

Protein content in the human milk declined from 4.1±2.1 g/dL on the 3rd postpartum day to 2.2±0.6 g/dL by the 28th day postpartum. Lactose (from 2.2±0.7 g/dL to 3.0±0.9 g/dL), fat (1.9±1.8 g/dL to 3.4±2.1 g/dL) and energy (42.3±18.8 Kcal/dL to 51.9±21.5 Kcal/dL) contents increased from day 3 to day 28.

CONCLUSION

Preterm human milk has high temporal and inter-individual variation in the macronutrient composition and without fortification is unlikely to meet the nutritional requirement of preterm neonates.

Production of [Formula: see text] and [Formula: see text] mesons up to high transverse momentum in pp collisions at 2.76 TeV

The invariant differential cross sections for inclusive [Formula: see text] and [Formula: see text] mesons at midrapidity were measured in pp collisions at [Formula: see text] TeV for transverse momenta [Formula: see text] GeV/c and [Formula: see text] GeV/c, respectively, using the ALICE detector. This large range in [Formula: see text] was achieved by combining various analysis techniques and different triggers involving the electromagnetic calorimeter (EMCal). In particular, a new single-cluster, shower-shape based method was developed for the identification of high-[Formula: see text] neutral pions, which exploits that the showers originating from their decay photons overlap in the EMCal. Above 4 GeV/[Formula: see text], the measured cross sections are found to exhibit a similar power-law behavior with an exponent of about 6.3. Next-to-leading-order perturbative QCD calculations differ from the measured cross sections by about 30% for the [Formula: see text], and between 30-50% for the [Formula: see text] meson, while generator-level simulations with PYTHIA 8.2 describe the data to better than 10-30%, except at [Formula: see text] GeV/[Formula: see text]. The new data can therefore be used to further improve the theoretical description of [Formula: see text] and [Formula: see text] meson production.

Flow Dominance and Factorization of Transverse Momentum Correlations in Pb-Pb Collisions at the LHC

We present the first measurement of the two-particle transverse momentum differential correlation function, P_{2}≡⟨Δp_{T}Δp_{T}⟩/⟨p_{T}⟩^{2}, in Pb-Pb collisions at sqrt[s_{NN}]=2.76  TeV. Results for P_{2} are reported as a function of the relative pseudorapidity (Δη) and azimuthal angle (Δφ) between two particles for different collision centralities. The Δϕ dependence is found to be largely independent of Δη for |Δη|≥0.9. In the 5% most central Pb-Pb collisions, the two-particle transverse momentum correlation function exhibits a clear double-hump structure around Δφ=π (i.e., on the away side), which is not observed in number correlations in the same centrality range, and thus provides an indication of the dominance of triangular flow in this collision centrality. Fourier decompositions of P_{2}, studied as a function of the collision centrality, show that correlations at |Δη|≥0.9 can be well reproduced by a flow ansatz based on the notion that measured transverse momentum correlations are strictly determined by the collective motion of the system.

Measurement of azimuthal correlations of D mesons with charged particles in pp collisions at [Formula: see text] TeV and p-Pb collisions at [Formula: see text] TeV

The azimuthal correlations of D mesons with charged particles were measured with the ALICE apparatus in pp collisions at [Formula: see text] and p-Pb collisions at [Formula: see text] at the Large Hadron Collider. [Formula: see text], [Formula: see text], and [Formula: see text] mesons and their charge conjugates with transverse momentum [Formula: see text] and rapidity in the nucleon-nucleon centre-of-mass system [Formula: see text] (pp collisions) and [Formula: see text] (p-Pb collisions) were correlated to charged particles with [Formula: see text]. The yield of charged particles in the correlation peak induced by the jet containing the D meson and the peak width are compatible within uncertainties in the two collision systems. The data are described within uncertainties by Monte-Carlo simulations based on PYTHIA, POWHEG, and EPOS 3 event generators.