High-throughput and high-dimensional single-cell analysis of antigen-specific CD8+ T cells.

Although critical to T cell function, antigen specificity is often omitted in high-throughput multiomics-based T cell profiling due to technical challenges. We describe a high-dimensional, tetramer-associated T cell antigen receptor (TCR) sequencing (TetTCR-SeqHD) method to simultaneously profile cognate antigen specificities, TCR sequences, targeted gene expression and surface-protein expression from tens of thousands of single cells. Using human polyclonal CD8+ T cells with known antigen specificity and TCR sequences, we demonstrate over 98% precision for detecting the correct antigen specificity. We also evaluate gene expression and phenotypic differences among antigen-specific CD8+ T cells and characterize phenotype signatures of influenza- and Epstein-Barr virus-specific CD8+ T cells that are unique to their pathogen targets. Moreover, with the high-throughput capacity of profiling hundreds of antigens simultaneously, we apply TetTCR-SeqHD to identify antigens that preferentially enrich cognate CD8+ T cells in patients with type 1 diabetes compared to healthy controls and discover a TCR that cross-reacts with diabetes-related and microbiome antigens. TetTCR-SeqHD is a powerful approach for profiling T cell responses in humans and mice.

Epigenetic Regulation of DNA Repair Pathway Choice by MacroH2A1 Splice Variants Ensures Genome Stability.

The inactive X chromosome (Xi) is inherently susceptible to genomic aberrations. Replication stress (RS) has been proposed as an underlying cause, but the mechanisms that protect from Xi instability remain unknown. Here, we show that macroH2A1.2, an RS-protective histone variant enriched on the Xi, is required for Xi integrity and female survival. Mechanistically, macroH2A1.2 counteracts its structurally distinct and equally Xi-enriched alternative splice variant, macroH2A1.1. Comparative proteomics identified a role for macroH2A1.1 in alternative end joining (alt-EJ), which accounts for Xi anaphase defects in the absence of macroH2A1.2. Genomic instability was rescued by simultaneous depletion of macroH2A1.1 or alt-EJ factors, and mice deficient for both macroH2A1 variants harbor no overt female defects. Notably, macroH2A1 splice variant imbalance affected alt-EJ capacity also in tumor cells. Together, these findings identify macroH2A1 splicing as a modulator of genome maintenance that ensures Xi integrity and may, more broadly, predict DNA repair outcome in malignant cells.

TISSUE: uncertainty-calibrated prediction of single-cell spatial transcriptomics improves downstream analyses.

Whole-transcriptome spatial profiling of genes at single-cell resolution remains a challenge. To address this limitation, spatial gene expression prediction methods have been developed to infer the spatial expression of unmeasured transcripts, but the quality of these predictions can vary greatly. Here we present Transcript Imputation with Spatial Single-cell Uncertainty Estimation (TISSUE) as a general framework for estimating uncertainty for spatial gene expression predictions and providing uncertainty-aware methods for downstream inference. Leveraging conformal inference, TISSUE provides well-calibrated prediction intervals for predicted expression values across 11 benchmark datasets. Moreover, it consistently reduces the false discovery rate for differential gene expression analysis, improves clustering and visualization of predicted spatial transcriptomics and improves the performance of supervised learning models trained on predicted gene expression profiles. Applying TISSUE to a MERFISH spatial transcriptomics dataset of the adult mouse subventricular zone, we identified subtypes within the neural stem cell lineage and developed subtype-specific regional classifiers.

Principled and interpretable alignability testing and integration of single-cell data.

Single-cell data integration can provide a comprehensive molecular view of cells, and many algorithms have been developed to remove unwanted technical or biological variations and integrate heterogeneous single-cell datasets. Despite their wide usage, existing methods suffer from several fundamental limitations. In particular, we lack a rigorous statistical test for whether two high-dimensional single-cell datasets are alignable (and therefore should even be aligned). Moreover, popular methods can substantially distort the data during alignment, making the aligned data and downstream analysis difficult to interpret. To overcome these limitations, we present a spectral manifold alignment and inference (SMAI) framework, which enables principled and interpretable alignability testing and structure-preserving integration of single-cell data with the same type of features. SMAI provides a statistical test to robustly assess the alignability between datasets to avoid misleading inference and is justified by high-dimensional statistical theory. On a diverse range of real and simulated benchmark datasets, it outperforms commonly used alignment methods. Moreover, we show that SMAI improves various downstream analyses such as identification of differentially expressed genes and imputation of single-cell spatial transcriptomics, providing further biological insights. SMAI's interpretability also enables quantification and a deeper understanding of the sources of technical confounders in single-cell data.

SPRITE: improving spatial gene expression imputation with gene and cell networks.

MOTIVATION: Spatially resolved single-cell transcriptomics have provided unprecedented insights into gene expression in situ, particularly in the context of cell interactions or organization of tissues. However, current technologies for profiling spatial gene expression at single-cell resolution are generally limited to the measurement of a small number of genes. To address this limitation, several algorithms have been developed to impute or predict the expression of additional genes that were not present in the measured gene panel. Current algorithms do not leverage the rich spatial and gene relational information in spatial transcriptomics. To improve spatial gene expression predictions, we introduce Spatial Propagation and Reinforcement of Imputed Transcript Expression (SPRITE) as a meta-algorithm that processes predictions obtained from existing methods by propagating information across gene correlation networks and spatial neighborhood graphs. RESULTS: SPRITE improves spatial gene expression predictions across multiple spatial transcriptomics datasets. Furthermore, SPRITE predicted spatial gene expression leads to improved clustering, visualization, and classification of cells. SPRITE can be used in spatial transcriptomics data analysis to improve inferences based on predicted gene expression. AVAILABILITY AND IMPLEMENTATION: The SPRITE software package is available at https://github.com/sunericd/SPRITE. Code for generating experiments and analyses in the manuscript is available at https://github.com/sunericd/sprite-figures-and-analyses.

Association of Patient Race and Hospital with Utilization of Regional Anesthesia for Treatment of Postoperative Pain in Total Knee Arthroplasty: A Retrospective Analysis Using Medicare Claims.

BACKGROUND: Regional anesthesia for total knee arthroplasty has been deemed high priority by national and international societies, and its use can serve as a measure of healthcare equity. The association between utilization of regional anesthesia for postoperative pain and (1) race and (2) hospital in patients undergoing total knee arthroplasty was estimated. The hypothesis was that Black patients would be less likely than White patients to receive regional anesthesia, and that variability in regional anesthesia would more likely be attributable to the hospital where surgery occurred than race. METHODS: This study used Medicare fee-for-service claims for patients aged 65 yr or older who underwent primary total knee arthroplasty between January 1, 2011, and December 31, 2016. The primary outcome was administration of regional anesthesia for postoperative pain, defined as any peripheral (femoral, lumbar plexus, or other) or neuraxial (spinal or epidural) block. The primary exposure was self-reported race (Black, White, or Other). Clinical significance was defined as a relative difference of 10% in regional anesthesia administration. RESULTS: Data from 733,406 cases across 2,507 hospitals were analyzed: 90.7% of patients were identified as White, 4.7% as Black, and 4.6% as Other. Median hospital-level prevalence of use of regional anesthesia was 51% (interquartile range, 18 to 79%). Black patients did not have a statistically different probability of receiving a regional anesthetic compared to White patients (adjusted estimates: Black, 53.3% [95% CI, 52.5 to 54.1%]; White, 52.7% [95% CI, 52.4 to 54.1%]; P = 0.132). Findings were robust to alternate specifications of the exposure and outcome. Analysis of variance revealed that 42.0% of the variation in block administration was attributable to hospital, compared to less than 0.01% to race, after adjusting for other patient-level confounders. CONCLUSIONS: Race was not associated with administration of regional anesthesia in Medicare patients undergoing primary total knee arthroplasty. Variation in the use of regional anesthesia was primarily associated with the hospital where surgery occurred.

A Cost and Waste-Savings Comparison Between Single-Use and Reusable Pulse Oximetry Sensors Across US Operating Rooms.

BACKGROUND: Operating room (OR) expenditures and waste generation are a priority, with several professional societies recommending the use of reprocessed or reusable equipment where feasible. The aim of this analysis was to compare single-use pulse oximetry sensor stickers ("single-use stickers") versus reusable pulse oximetry sensor clips ("reusable clips") in terms of annual cost savings and waste generation across all ORs nationally. METHODS: This study did not involve patient data or research on human subjects. As such, it did not meet the requirements for institutional review board approval. An economic model was used to compare the relative costs and waste generation from using single-use stickers versus reusable clips. This model took into account: (1) the relative prices of single-use stickers and reusable clips, (2) the number of surgeries and ORs nationwide, (3) the workload burden of cleaning the reusable clips, and (4) the costs of capital for single-use stickers and reusable clips. In addition, we also estimated differences in waste production based on the raw weight plus unit packaging of single-use stickers and reusable clips that would be disposed of over the course of the year, without any recycling interventions. Estimated savings were rounded to the nearest $0.1 million. RESULTS: The national net annual savings of transitioning from single-use stickers to reusable clips in all ORs ranged from $510.5 million (conservative state) to $519.3 million (favorable state). Variability in savings estimates is driven by scenario planning for replacement rate of reusable clips, workload burden of cleaning (ranging from an additional expense of $618k versus a cost savings of $309k), and cost of capital-interest gained on investment of capital that is freed up by the monetary savings of a transition to reusable clips contributes between $541k (low-interest rates of 2.85%) and $1.3 million (high-interest rates of 7.08%). The annual waste that could be diverted from landfill by transitioning to reusable clips was found to be between 587 tons (conservative state) up to 589 tons (favorable state). If institutions need to purchase new vendor monitors or cables to make the transition, that may increase the 1-time capital disbursement. CONCLUSIONS: Using reusable clips versus single-use stickers across all ORs nationally would result in appreciable annual cost savings and waste generation reduction impact. As both single-use stickers and reusable clips are equally accurate and reliable, this cost and waste savings could be instituted without a compromise in clinical care.

Association Between Naloxone Coprescription Mandates and Postoperative Outcomes.

INTRODUCTION: The opioid epidemic is a public health issue in the United States. The objective of this study was to evaluate the association between naloxone coprescription mandates and postoperative outcomes. BACKGROUND: Data on naloxone coprescription mandates show mixed evidence for fatal overdoses in the broader population. How these mandates have impacted surgical patients has not been fully explored. METHODS: Healthcare claims data were used to identify all patients undergoing 1 of 50 common procedures between January 1, 2004, and June 30, 2019, and categorized as high risk for opioid overdose. The primary outcomes were an emergency department visit or hospital admission within 30 postoperative days. To reduce confounding, the association between this outcome and the implementation of naloxone coprescription mandates was estimated using a difference-in-differences approach. RESULTS: The study included 429,878 surgical patients with an average age of 54.8 years (SD=15.9 years) and with 257,728 females (60.0%). There was no significant association between naloxone prescribing mandates and the primary outcomes. After adjustment for potential confounders, the incidence of hospital admission was 3.26% after implementation of a naloxone coprescription mandate compared with 3.33% before (difference change: -0.08%, 95% CI: -0.44% to 0.29%, P =0.68). The incidence of an emergency department visit was 7.06% after implementation of a naloxone coprescription mandate compared with 7.73% before (difference: -0.67%, 95% CI: -1.39% to 0.05%, P =0.07). These results were robust to a variety of sensitivity and subgroup analyses. CONCLUSIONS: Naloxone coprescription mandates were not associated with a statistically or clinically significant change in emergency department visits or hospital admissions within 30 postoperative days.

Association Between State Limits on Opioid Prescribing and the Incidence of Persistent Postoperative Opioid Use Among Surgical Patients.

OBJECTIVE: To examine whether laws limiting opioid prescribing have been associated with reductions in the incidence of persistent postoperative opioid use. SUMMARY OF BACKGROUND DATA: In an effort to address the opioid epidemic, 26 states (as of 2018) have passed laws limiting opioid prescribing for acute pain. However, it is unknown whether these laws have achieved their reduced the risk of persistent postoperative opioid use. METHODS: We identified 957,639 privately insured patients undergoing one of 10 procedures between January 1, 2004 and September 30, 2018. We then estimated the association between persistent postoperative opioid use, defined as having filled ≥10 prescriptions or ≥120 days supply of opioids during postoperative days 91-365, and whether opioid prescribing limits were in effect on the day of surgery. States were classified as having: no limits, a limit of ≤7 days supply, or a limit of >7 days supply. The regression models adjusted for observable confounders such as patient comorbidities and also utilized a difference-in-differences approach, which relied on variation in state laws over time, to further minimize confounding. RESULTS: The adjusted incidence of persistent postoperative opioid use was 3.5% (95%CI 3.3%-3.7%) for patients facing a limit of ≤7 days supply, compared with 3.3% (95%CI 3.3%-3.3%) for patients facing no prescribing limits ( P = 0.13 for difference compared to no prescribing limits) and 3.4%, (95%CI 3.2%-3.6%) for patients facing a limit of >7 days supply ( P = 0.43 for difference compared to no prescribing limits). CONCLUSIONS: Laws limiting opioid prescriptions were not associated with subsequent reductions in persistent postoperative opioid use.

Association between "Balance Billing" Legislation and Anesthesia Payments in California: A Retrospective Analysis.

BACKGROUND: Insured patients who receive out-of-network care may receive a "balance bill" for the difference between the practitioner's charge and their insurer's contracted rate. In 2017, California banned balance billing for anesthesia care. This study examined the association between California's law and subsequent payments for anesthesia care. The authors hypothesized that, after the law's implementation, there would be no change in in-network payment amounts, and that out-of-network payment amounts and the portion of claims occurring out-of-network would decline. METHODS: The study used average, quarterly, California county-level payment data (2013 to 2020) derived from a claims database of commercially insured patients. Using a difference-in-differences approach, the change was estimated in payment amounts for intraoperative or intrapartum anesthesia care, along with the portion of claims occurring out-of-network, after the law's implementation. The comparison group was office visit payments, expected to be unaffected by the law. The authors prespecified that they would refer to differences of 10% or greater as policy significant. RESULTS: The sample consisted of 43,728 procedure code-county-quarter-network combinations aggregated from 4,599,936 claims. The law's implementation was associated with a significant 13.6% decline in payments for out-of-network anesthesia care (95% CI, -16.5 to -10.6%; P < 0.001), translating to an average $108 decrease across all procedures (95% CI, -$149 to -$64). There was a statistically significant 3.0% increase in payments for in-network anesthesia care (95% CI, 0.9 to 5.1%; P = 0.007), translating to an average $87 increase (95% CI, $64 to $110), which may be notable in some circumstances but did not meet the study threshold for identifying a change as policy significant. There was a nonstatistically significant increase in the portion of claims occurring out-of-network (10.0%, 95% CI, -4.1 to 24.2%; P = 0.155). CONCLUSIONS: California's balance billing law was associated with significant declines in out-of-network anesthesia payments in the first 3 yr after implementation. There were mixed statistical and policy significant results for in-network payments and the proportion of out-of-network claims.

Meeting Report: First Cardiovascular Outcomes Research in Perioperative Medicine Conference.

The first Cardiovascular Outcomes Research in Perioperative Medicine (COR-PM) conference took place on May 13, 2022, in Palm Springs, CA, and online. Here, we: (1) summarize the background, objective, and aims of the COR-PM meeting; (2) describe the conduct of the meeting; and (3) outline future directions for scientific meetings aimed at fostering high-quality clinical research in the broader perioperative medicine community.

Optimal control of aging in complex networks.

Many complex systems experience damage accumulation, which leads to aging, manifest as an increasing probability of system collapse with time. This naturally raises the question of how to maximize health and longevity in an aging system at minimal cost of maintenance and intervention. Here, we pose this question in the context of a simple interdependent network model of aging in complex systems and show that it exhibits cascading failures. We then use both optimal control theory and reinforcement learning alongside a combination of analysis and simulation to determine optimal maintenance protocols. These protocols may motivate the rational design of strategies for promoting longevity in aging complex systems with potential applications in therapeutic schedules and engineered system maintenance.

Full-Time Cardiac Intensive Care Unit Staffing by Heart Failure Specialists and its Association with Mortality Rates.

BACKGROUND: Cardiac intensive care units (CICUs) serve medically complex patients with multiorgan dysfunction. Whether a CICU that is staffed full time by heart failure (HF) specialists is associated with decreased mortality is unclear. METHODS AND RESULTS: A retrospective review of consecutive CICU admissions from January 1, 2012, to December 31, 2016, was performed. In January 2014, the CICU changed from an open unit staffed by any cardiologist to a closed unit managed by HF specialists. Patients' baseline characteristics were determined, and a multivariate regression analysis was performed to ascertain mortality rates in the CICU. Baseline severity of illness was higher in the closed/HF specialist CICU model (P< 0.001). Death occurred in 101 of 1185 patients admitted to the CICU (8.5%) in the open-unit model and in 139 of 2163 patients (6.4%) admitted to the closed/HF specialist model (absolute risk reduction 2.1%, 95% confidence interval [CI] 0.1-4.0%; P = 0.01). The transition from an open to a closed/HF specialist model was associated with a lower overall CICU mortality rate (odds ratio [OR] 0.63; 95% CI 0.43-0.93). Prespecified interaction with a mechanical circulatory support device and unit model showed that treatment with such a device was associated with lower mortality rates in the closed/HF specialist model of a CICU (OR 0.6; 95% CI 0.18-0.78; P for interaction <0.01). CONCLUSION: Transition to a closed unit model staffed by a dedicated HF specialist is associated with lower CICU mortality rates.

Surgeon Variation in Perioperative Opioid Prescribing and Medium- or Long-term Opioid Utilization after Total Knee Arthroplasty: A Cross-sectional Analysis.

BACKGROUND: Whether a particular surgeon's opioid prescribing behavior is associated with prolonged postoperative opioid use is unknown. This study tested the hypothesis that the patients of surgeons with a higher propensity to prescribe opioids are more likely to utilize opioids long-term postoperatively. METHODS: The study identified 612,378 Medicare fee-for-service patients undergoing total knee arthroplasty between January 1, 2011, and December 31, 2016. "High-intensity" surgeons were defined as those whose patients were, on average, in the upper quartile of opioid utilization in the immediate perioperative period (preoperative day 7 to postoperative day 7). The study then estimated whether patients of high-intensity surgeons had higher opioid utilization in the midterm (postoperative days 8 to 90) and long-term (postoperative days 91 to 365), utilizing an instrumental variable approach to minimize confounding from unobservable factors. RESULTS: In the final sample of 604,093 patients, the average age was 74 yr (SD 5), and there were 413,121 (68.4%) females. A total of 180,926 patients (30%) were treated by high-intensity surgeons. On average, patients receiving treatment from a high-intensity surgeon received 36.1 (SD 35.0) oral morphine equivalent (morphine milligram equivalents) per day during the immediate perioperative period compared to 17.3 morphine milligram equivalents (SD 23.1) per day for all other patients (+18.9 morphine milligram equivalents per day difference; 95% CI, 18.7 to 19.0; P < 0.001). After adjusting for confounders, receiving treatment from a high-intensity surgeon was associated with higher opioid utilization in the midterm opioid postoperative period (+2.4 morphine milligram equivalents per day difference; 95% CI, 1.7 to 3.2; P < 0.001 [11.4 morphine milligram equivalents per day vs. 9.0]) and lower opioid utilization in the long-term postoperative period (-1.0 morphine milligram equivalents per day difference; 95% CI, -1.4 to -0.6; P < 0.001 [2.8 morphine milligram equivalents per day vs. 3.8]). While statistically significant, these differences are clinically small. CONCLUSIONS: Among Medicare fee-for-service patients undergoing total knee arthroplasty, surgeon-level variation in opioid utilization in the immediate perioperative period was associated with statistically significant but clinically insignificant differences in opioid utilization in the medium- and long-term postoperative periods.

Association Between Changes in Postoperative Opioid Utilization and Long-Term Health Care Spending Among Surgical Patients With Chronic Opioid Utilization.

BACKGROUND: There is growing interest in identifying and developing interventions aimed at reducing the risk of increased, long-term opioid use among surgical patients. While understanding how these interventions impact health care spending has important policy implications and may facilitate the widespread adoption of these interventions, the extent to which they may impact health care spending among surgical patients who utilize opioids chronically is unknown. METHODS: This study was a retrospective analysis of administrative health care claims data for privately insured patients. We identified 53,847 patients undergoing 1 of 10 procedures between January 1, 2004, and September 30, 2018 (total knee arthroplasty, total hip arthroplasty, laparoscopic cholecystectomy, open cholecystectomy, laparoscopic appendectomy, open appendectomy, cesarean delivery, functional endoscopic sinus surgery, transurethral resection of the prostate, or simple mastectomy) who had chronic opioid utilization (≥10 prescriptions or ≥120-day supply in the year before surgery). Patients were classified into 3 groups based on differences in opioid utilization, measured in average daily oral morphine milligram equivalents (MMEs), between the first postoperative year and the year before surgery: "stable" (<20% change), "increasing" (≥20% increase), or "decreasing" (≥20% decrease). We then examined the association between these 3 groups and health care spending during the first postoperative year, using a multivariable regression to adjust for observable confounders, such as patient demographics, medical comorbidities, and preoperative health care utilization. RESULTS: The average age of the sample was 62.0 (standard deviation [SD] 13.1) years, and there were 35,715 (66.3%) women. Based on the change in average daily MME between the first postoperative year and the year before surgery, 16,961 (31.5%) patients were classified as "stable," 15,463 (28.7%) were classified as "increasing," and 21,423 (39.8%) patients were classified as "decreasing." After adjusting for potential confounders, "increasing" patients had higher health care spending ($37,437) than "stable" patients ($31,061), a difference that was statistically significant ($6377; 95% confidence interval [CI], $5669-$7084; P < .001), while "decreasing" patients had lower health care spending ($29,990), a difference (-$1070) that was also statistically significant (95% CI, -$1679 to -$462; P = .001). These results were generally consistent across an array of subgroup and sensitivity analyses. CONCLUSIONS: Among patients with chronic opioid utilization before surgery, subsequent increases in opioid utilization during the first postoperative year were associated with increased health care spending during that timeframe, while subsequent decreases in opioid utilization were associated with decreased health care spending.

Association of Medicaid Expansion With Neuraxial Labor Analgesia Use in the United States: A Retrospective Cross-Sectional Analysis.

BACKGROUND: The Affordable Care Act has been associated with increased Medicaid coverage for childbirth among low-income US women. We hypothesized that Medicaid expansion was associated with increased use of labor neuraxial analgesia. METHODS: We performed a cross-sectional analysis of US women with singleton live births who underwent vaginal delivery or intrapartum cesarean delivery between 2009 and 2017. Data were sourced from births in 26 US states that used the 2003 Revised US Birth Certificate. Difference-in-difference linear probability models were used to compare changes in the prevalence of neuraxial labor analgesia in 15 expansion and 11 nonexpansion states before and after Medicaid expansion. Models were adjusted for potential maternal and obstetric confounders with standard errors clustered at the state level. RESULTS: The study sample included 5,703,371 births from 15 expansion states and 5,582,689 births from 11 nonexpansion states. In the preexpansion period, the overall rate of neuraxial analgesia in expansion and nonexpansion states was 73.2% vs 76.3%. Compared with the preexpansion period, the rate of neuraxial analgesia increased in the postexpansion period by 1.7% in expansion states (95% CI, 1.6-1.8) and 0.9% (95% CI, 0.9-1.0) in nonexpansion states. The adjusted difference-in-difference estimate comparing expansion and nonexpansion states was 0.47% points (95% CI, -0.63 to 1.57; P = .39). CONCLUSIONS: Medicaid expansion was not associated with an increase in the rate of neuraxial labor analgesia in expansion states compared to the change in nonexpansion states over the same time period. Increasing Medicaid eligibility alone may be insufficient to increase the rate of neuraxial labor analgesia.

How Can Anesthesiologists Influence Policymaking? Reflections from a Year at the Council of Economic Advisers.

From September 2019 to August 2020, the author served as a senior economist on the Council of Economic Advisers, a government agency charged with providing economic analysis and advice to the President of the United States and senior government officials. Working with the Council yielded many useful lessons on how anesthesiologists can influence healthcare policy. First, because the President has wide latitude over many areas of health policy that directly impact patient care and anesthesiologists' working environment, anesthesiologists should focus their efforts on influencing policymakers within the executive branch of government in addition to influencing lawmakers. Second, policymakers are busy and typically do not have a technical background, so anesthesiologists must learn how to communicate with them succinctly and at an appropriate level. Finally, because policymakers often need analysis quickly, anesthesiologists must meet these needs even if the underlying analysis is rougher and less precise that what would normally be needed for peer review.

Preoperative Opioid Utilization Patterns and Postoperative Opioid Utilization: A Retrospective Cohort Study.

BACKGROUND: Among chronic opioid users, the association between decreasing or increasing preoperative opioid utilization and postoperative outcomes is unknown. The authors hypothesized that decreasing utilization would be associated with improved outcomes and increasing utilization with worsened outcomes. METHODS: Using commercial insurance claims, the authors identified 57,019 chronic opioid users (10 or more prescriptions or 120 or more days supplied during the preoperative year), age 18 to 89 yr, undergoing one of 10 surgeries between 2004 and 2018. Patients with a 20% or greater decrease or increase in opioid utilization between preoperative days 7 to 90 and 91 to 365 were compared to patients with less than 20% change (stable utilization). The primary outcome was opioid utilization during postoperative days 91 to 365. Secondary outcomes included alternative measures of postoperative opioid utilization (filling a minimum number of prescriptions during this period), postoperative adverse events, and healthcare utilization. RESULTS: The average age was 63 ± 13 yr, with 38,045 (66.7%) female patients. Preoperative opioid utilization was decreasing for 12,347 (21.7%) patients, increasing for 21,330 (37.4%) patients, and stable for 23,342 (40.9%) patients. Patients with decreasing utilization were slightly less likely to fill an opioid prescription during postoperative days 91 to 365 compared to stable patients (89.2% vs. 96.4%; odds ratio, 0.323; 95% CI, 0.296 to 0.352; P < 0.001), though the average daily doses were similar among patients who continued to utilize opioids during this timeframe (46.7 vs. 46.5 morphine milligram equivalents; difference, 0.2; 95% CI, -0.8 to 1.2; P = 0.684). Of patients with increasing utilization, 93.6% filled opioid prescriptions during this period (odds ratio, 0.57; 95% CI, 0.52 to 0.62; P < 0.001), with slightly lower average daily doses (44.3 morphine milligram equivalents; difference, -2.2; 95% CI, -3.1 to -1.3; P < 0.001). Except for alternative measures of persistent postoperative opioid utilization, there were no clinically significant differences for the secondary outcomes. CONCLUSIONS: Changes in preoperative opioid utilization were not associated with clinically significant differences for several postoperative outcomes including postoperative opioid utilization.

The link between health and economic preferences: Evidence from 22 OECD countries.

We study the link between health status and economic preferences using survey data from 22 Organisation for Economic Co-operation and Development (OECD) countries. We hypothesize that there is a relationship between poor health and the preferences that people hold, and therefore their choices and decisions. We find that individuals with a limiting health condition are more risk averse and less patient, and that this is true for physical and mental health conditions. The magnitudes of the health gap are approximately 60% and 70% of the gender gap in risk and time preferences, respectively. Importantly, the health gaps are large for males, females, young, old, school dropouts, degree holders, employed, nonemployed, rich, and poor. They also hold for countries with different levels of gross domestic product (GDP), inequality, social expenditure, and disease burden.

ImageNet-trained deep neural networks exhibit illusion-like response to the Scintillating grid.

Deep neural network (DNN) models for computer vision are capable of human-level object recognition. Consequently, similarities between DNN and human vision are of interest. Here, we characterize DNN representations of Scintillating grid visual illusion images in which white disks are perceived to be partially black. Specifically, we use VGG-19 and ResNet-101 DNN models that were trained for image classification and consider the representational dissimilarity (\(L^1\) distance in the penultimate layer) between pairs of images: one with white Scintillating grid disks and the other with disks of decreasing luminance levels. Results showed a nonmonotonic relation, such that decreasing disk luminance led to an increase and subsequently a decrease in representational dissimilarity. That is, the Scintillating grid image with white disks was closer, in terms of the representation, to images with black disks than images with gray disks. In control nonillusion images, such nonmonotonicity was rare. These results suggest that nonmonotonicity in a deep computational representation is a potential test for illusion-like response geometry in DNN models.