Search

OBJECTIVE. Several short-term readmission and mortality prediction models have been developed using clinical risk factors or biomarkers among patients undergoing coronary artery bypass graft (CABG) surgery. The use of biomarkers for long-term prediction of readmission and mortality is less well understood. Given the established association of cardiac biomarkers with short-term adverse outcomes, we hypothesized that 5-year prediction of readmission or mortality may be significantly improved using cardiac biomarkers. MATERIALS AND METHODS. Plasma biomarkers from 1,149 patients discharged alive after isolated CABG surgery from eight medical centers were measured in a cohort from the Northern New England Cardiovascular Disease Study Group (NNE) between 2004 and 2007. We assessed the added predictive value of a biomarker panel with a clinical model against the clinical model alone and compared the model discrimination using the area under the receiver operating characteristic (AUROC) curves. RESULTS. In our cohort, 461 (40%) patients were readmitted or died within 5 years. Long-term outcomes were predicted by applying the STS ASCERT clinical model with an AUROC of 0.69. The biomarker panel with the clinical model resulted in a significantly improved AUROC of 0.74 (p-value

Background Mendelian randomization (MR) applies instrumental variable (IV) methods to observational data using a genetic variant as an IV. Several Monte-Carlo studies investigate the performance of MR methods with binary outcomes, but few consider them in conjunction with binary risk factors. Objective To develop a novel MR estimator for scenarios with a binary risk factor and outcome; and compare to existing MR estimators via simulations and real data analysis. Methods A bivariate Bernoulli distribution is adapted to the IV setting. Empirical bias and asymptotic coverage probabilities are estimated via simulations. The proposed method is compared to the Wald method, two-stage predictor substitution (2SPS), two-stage residual inclusion (2SRI), and the generalized method of moments (GMM). An analysis is performed using existing data from the CLEAR study to estimate the potential causal effect of smoking on rheumatoid arthritis risk in African Americans. Results Bias was low for the proposed method and comparable to 2SPS. The Wald method was often biased towards the null. Coverage was adequate for the proposed method, 2SPS, and 2SRI. Coverage for the Wald and GMM methods was poor in several scenarios. The causal effect of ever smoking on rheumatoid arthritis risk was not statistically significant using a variety of genetic instruments. Conclusions Simulations suggest the proposed MR method is sound with binary risk factors and outcomes, and comparable to 2SPS and 2SRI in terms of bias. The proposed method also provides more natural framework for hypothesis testing compared to 2SPS or 2SRI, which require ad-hoc variance adjustments.

The USA leads the world in healthcare spending but trails dozens of countries in life expectancy. Government spending may reduce overall mortality by redistributing resources from the rich to the poor. We linked mortality data from 2006 to 2015 to municipal and state government spending in 149 of the largest American cities. We modeled the association of mortality with city and state government spending per capita in 2005 using weighted linear regression. A 10% increase in state government expenditures was associated (P = 0.008) with a 1.4% (95%CI: 0.4–2.4%) reduction in mortality in American cities. Total city government expenditures were not associated with mortality (P > 0.10). However, among Whites, increases in city government spending were associated with a reduction in mortality of 4.8% (2.1–7.5%), but among Blacks and Asians, increased city government spending was associated with respective mortality increases of 1.7% (0.6–2.9%) and 5.1% (2.1–6.2%). State government spending is associated with reduced mortality in American cities. City government spending appears to benefit White longevity and hurt non-White longevity. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11524-021-00516-3.

OBJECTIVE This study was undertaken to evaluate the influence that subject-specific factors have on intracranial interictal epileptiform discharge (IED) rates in persons with refractory epilepsy. METHODS One hundred fifty subjects with intracranial electrodes performed multiple sessions of a free recall memory task; this standardized task controlled for subject attention levels. We utilized a dominance analysis to rank the importance of subject-specific factors based on their relative influence on IED rates. Linear mixed-effects models were employed to comprehensively examine factors with highly ranked importance. RESULTS Antiseizure medication (ASM) status, time of testing, and seizure onset zone (SOZ) location were the highest-ranking factors in terms of their impact on IED rates. The average IED rate of electrodes in SOZs was 34% higher than the average IED rate of electrodes outside of SOZs (non-SOZ; p

Study design We combined elements of cohort and crossover-cohort design. Objective The objective of this study was to compare long-term outcomes for Spinal Manipulative Therapy (SMT) and Opioid Analgesic Therapy (OAT) regarding escalation of care for patients with chronic low back pain (cLBP). Summary of background data Current evidence-based guidelines for clinical management of cLBP include both OAT and SMT. For long-term care of older adults, the efficiency and value of continuing either OAT or SMT are uncertain. Methods We examined Medicare claims data spanning a five-year period. We included older Medicare beneficiaries with an episode of cLBP beginning in 2013. All patients were continuously enrolled under Medicare Parts A, B, and D. We analyzed the cumulative frequency of encounters indicative of an escalation of care for cLBP, including hospitalizations, emergency department visits, advanced diagnostic imaging, specialist visits, lumbosacral surgery, interventional pain medicine techniques, and encounters for potential complications of cLBP. Results SMT was associated with lower rates of escalation of care as compared to OAT. The adjusted rate of escalated care encounters was approximately 2.5 times higher for initial choice of OAT vs. initial choice of SMT (with weighted propensity scoring: rate ratio 2.67, 95% CI 2.64-2.69, p Conclusions Among older Medicare beneficiaries who initiated long-term care for cLBP with opioid analgesic therapy, the adjusted rate of escalated care encounters was significantly higher as compared to those who initiated care with spinal manipulative therapy.Level of Evidence: 3.

Among people with multiple sclerosis (MS), yoga has potential to improve fatigue and other symptoms that undermine quality of life. The aim of this study was to assess the feasibility, acceptability, and effectiveness of LoveYourBrain Yoga, a six-week yoga with psychoeducation program, on fatigue and other health-related outcomes among people with MS in a rural, community-based setting in the United States.This non-randomized 2x2 crossover pilot trial compared LoveYourBrain Yoga to a control among 15 people with MS. People were eligible if they were adults with MS (EDSS score ≤6), English-speaking, and ambulatory. Paired t-tests and Wilcoxon signed rank sum analyses assessed mean differences in PROMIS-FatigueSignificant improvements in fatigue (MD -4.34, SD 5.26, p = 0.012), positive affect and wellbeing (MD 2.76, SD 3.99, p = 0.028), and anxiety (MD -4.42, SD 5.36, p = 0.012) were found after LoveYourBrain Yoga compared to the control. Participants reported high satisfaction (M 9.5, SD 1.4) and a majority (92.3%) reported 'Definitely, yes' to recommending it to a friend or family.LoveYourBrain Yoga is feasible and acceptable when implemented in a rural, community-based setting for people with MS. It may improve a range of MS symptoms and offer a means for acquiring new skills for stress reduction, anxiety management, and overall wellbeing.

Study design Retrospective observational study. Objective Opioid Analgesic Therapy (OAT) and Spinal Manipulative Therapy (SMT) are evidence-based strategies for treatment of chronic low back pain (cLBP), but the long-term safety of these therapies is uncertain. The objective of this study was to compare OAT versus SMT with regard to risk of adverse drug events (ADEs) among older adults with cLBP. Summary of background data We examined Medicare claims data spanning a 5-year period on fee-for-service beneficiaries aged 65 to 84 years, continuously enrolled under Medicare Parts A, B, and D for a 60-month study period, and with an episode of cLBP in 2013. We excluded patients with a diagnosis of cancer or use of hospice care. Methods All included patients received long-term management of cLBP with SMT or OAT. We assembled cohorts of patients who received SMT or OAT only, and cohorts of patients who crossed over from OAT to SMT or from SMT to OAT. We used Poisson regression to estimate the adjusted incidence rate ratio for outpatient ADE among patients who initially chose OAT as compared with SMT. Results With controlling for patient characteristics, health status, and propensity score, the adjusted rate of ADE was more than 42 times higher for initial choice of OAT versus initial choice of SMT (rate ratio 42.85, 95% CI 34.16-53.76, P Conclusion Among older Medicare beneficiaries who received long-term care for cLBP the adjusted rate of ADE for patients who initially chose OAT was substantially higher than those who initially chose SMT.Level of Evidence: 2.

Background Incomplete resection of neoplastic polyps is considered an important reason for the development of colorectal cancer. However, there are no data on the natural history of polyps that were incompletely removed. Objective To examine the risk for metachronous neoplasia during surveillance colonoscopy after documented incomplete polyp resection. Design Observational cohort study of patients who participated in the CARE (Complete Adenoma REsection) study (2009 to 2012). Setting 2 academic medical centers. Patients Patients who had resection of a 5- to 20-mm neoplastic polyp, had a documented complete or incomplete resection, and had a surveillance examination. Measurements Segment metachronous neoplasia, defined as the proportion of colon segments with at least 1 neoplastic polyp at first surveillance examination, was measured. Segment metachronous neoplasia was compared between segments with a prior incomplete polyp resection (incomplete segments) and those with a prior complete resection (complete segments), accounting for clustering of segments within patients. Results Of 233 participants in the original study, 166 (71%) had at least 1 surveillance examination. Median time to surveillance was shorter after incomplete versus complete resection (median, 17 vs. 45 months). The risk for any metachronous neoplasia was greater in segments with incomplete versus complete resection (52% vs. 23%; risk difference [RD], 28% [95% CI, 9% to 47%]; P = 0.004). Incomplete segments also had a greater number of neoplastic polyps (mean, 0.8 vs. 0.3; RD, 0.50 [CI, 0.1 to 0.9]; P = 0.008) and greater risk for advanced neoplasia (18% vs. 3%; RD, 15% [CI, 1% to 29%]; P = 0.034). Incomplete resection was the strongest independent factor associated with metachronous neoplasia (odds ratio, 3.0 [CI, 1.12 to 8.17]). Limitation Potential patient selection bias due to incomplete follow-up. Conclusion This natural history study found a statistically significantly greater risk for future neoplasia and advanced neoplasia in colon segments after incomplete resection compared with segments with complete resection. Primary funding source None.

OBJECTIVES: The prevalence of obesity with sarcopenia is increasing in adults aged ≥65 years. This geriatric syndrome places individuals at risk for synergistic complications that leads to long-term functional decline. We ascertained the relationship between sarcopenic obesity and incident long-term impaired global cognitive function in a representative US population. DESIGN: A longitudinal, secondary data set analysis using the National Health and Aging Trends Survey. SETTING: Community-based older adults in the United States. PARTICIPANTS: Participants without baseline impaired cognitive function aged ≥65 years with grip strength and body mass index measures. METHODS: Sarcopenia was defined using the Foundation for the National Institutes of Health Sarcopenia Project grip strength cut points (men

Background Estimation that employs instrumental variables (IV) can reduce or eliminate bias due to confounding. In observational studies, instruments result from natural experiments such as the effect of clinician preference or geographic distance on treatment selection. In randomized studies the randomization indicator is typically a valid instrument, especially if the study is blinded, e.g. no placebo effect. Estimation via instruments is a highly developed field for linear models but the use of instruments in time-to-event analysis is far from established. Various IV-based estimators of the hazard ratio (HR) from Cox’s regression models have been proposed. Methods We extend IV based estimation of Cox’s model beyond proportionality of hazards, and address estimation of a log-linear time dependent hazard ratio and a piecewise constant HR. We estimate the marginal time-dependent hazard ratio unlike other approaches that estimate the hazard ratio conditional on the omitted covariates. We use estimating equations motivated by Martingale representations that resemble the partial likelihood score statistic. We conducted simulations that include the use of copulas to generate potential times-to-event that have a given marginal structural time dependent hazard ratio but are dependent on omitted covariates. We compare our approach to the partial likelihood estimator, and two other IV based approaches. We apply it to estimation of the time dependent hazard ratio for two vascular interventions. Results The method performs well in simulations of a stepwise time-dependent hazard ratio, but illustrates some bias that increases as the hazard ratio moves away from unity (the value that typically underlies the null hypothesis). It compares well to other approaches when the hazard ratio is stepwise constant. It also performs well for estimation of a log-linear hazard ratio where no other instrumental variable approaches exist. Conclusion The estimating equations we propose for estimating a time-dependent hazard ratio using an IV perform well in simulations. We encourage the use of our procedure for time-dependent hazard ratio estimation when unmeasured confounding is a concern and a suitable instrumental variable exists.

Purpose Obesity associated fat infiltration of organ systems is accompanied by organ dysfunction and poor cancer outcomes. Obese women demonstrate variable degrees of fat infiltration of axillary lymph nodes (LNs), and they are at increased risk for node-positive breast cancer. However, the relationship between enlarged axillary nodes and axillary metastases has not been investigated. The purpose of this study is to evaluate the association between axillary metastases and fat-enlarged axillary nodes visualized on mammograms and breast MRI in obese women with a diagnosis of invasive breast cancer. Methods This retrospective case–control study included 431 patients with histologically confirmed invasive breast cancer. The primary analysis of this study included 306 patients with pre-treatment and pre-operative breast MRI and body mass index (BMI) > 30 (201 node-positive cases and 105 randomly selected node-negative controls) diagnosed with invasive breast cancer between April 1, 2011, and March 1, 2020. The largest visible LN was measured in the axilla contralateral to the known breast cancer on breast MRI. Multivariate logistic regression models were used to assess the association between node-positive status and LN size adjusting for age, BMI, tumor size, tumor grade, tumor subtype, and lymphovascular invasion. Results A strong likelihood of node-positive breast cancer was observed among obese women with fat-expanded lymph nodes (adjusted OR for the 4th vs. 1st quartile for contralateral LN size on MRI: 9.70; 95% CI 4.26, 23.50; p Conclusion Fat expansion of axillary lymph nodes was associated with a high likelihood of axillary metastases in obese women with invasive breast cancer independent of BMI and tumor characteristics.

Objective Spinal manipulation (SM) is recommended for first-line treatment of patients with low back pain. Inadequate access to SM may result in inequitable spine care for older US adults, but the supply of clinicians who provide SM under Medicare is uncertain. The purpose of this study was to measure temporal trends and geographic variations in the supply of clinicians who provide SM to Medicare beneficiaries. Methods Medicare is a US government–administered health insurance program that provides coverage primarily for older adults and people with disabilities. We used a serial cross-sectional design to examine Medicare administrative data from 2007 to 2015 for SM services identified by procedure code. We identified unique providers by National Provider Identifier and distinguished between chiropractors and other specialties by Physician Specialty Code. We calculated supply as the number of providers per 100 000 beneficiaries, stratified by geographic location and year. Results Of all clinicians who provide SM to Medicare beneficiaries, 97% to 98% are doctors of chiropractic. The geographic supply of doctors of chiropractic providing SM services in 2015 ranged from 20/100 000 in the District of Columbia to 260/100 000 in North Dakota. The supply of other specialists performing the same services ranged from fewer than 1/100 000 in 11 states to 8/100 000 in Colorado. Nationally, the number of Medicare-active chiropractors declined from 47 102 in 2007 to 45 543 in 2015. The count of other clinicians providing SM rose from 700 in 2007 to 1441 in 2015. Conclusion Chiropractors constitute the vast majority of clinicians who bill for SM services to Medicare beneficiaries. The supply of Medicare-active SM providers varies widely by state. The overall supply of SM providers under Medicare is declining, while the supply of nonchiropractors who provide SM is growing.

Key Points Question Can machine learning deployed in electronic health records be used to improve readmission risk estimation for patients following acute myocardial infarction? Findings In this cohort study examining externally validated machine learning risk models for 30-day readmission of 10 187 patients following hospitalization for acute myocardial infarction, good discrimination performance was noted at the development site, but the best discrimination did not result in the best calibration. External validation yielded significant declines in discrimination and calibration. Meaning The findings of this study highlight that robust calibration assessments are a necessary complement to discrimination when machine learning models are used to predict post–acute myocardial infarction readmission; challenges with data availability across sites, even in the presence of a common data model, limit external validation performance., Importance In the US, more than 600 000 adults will experience an acute myocardial infarction (AMI) each year, and up to 20% of the patients will be rehospitalized within 30 days. This study highlights the need for consideration of calibration in these risk models. Objective To compare multiple machine learning risk prediction models using an electronic health record (EHR)–derived data set standardized to a common data model. Design, Setting, and Participants This was a retrospective cohort study that developed risk prediction models for 30-day readmission among all inpatients discharged from Vanderbilt University Medical Center between January 1, 2007, and December 31, 2016, with a primary diagnosis of AMI who were not transferred from another facility. The model was externally validated at Dartmouth-Hitchcock Medical Center from April 2, 2011, to December 31, 2016. Data analysis occurred between January 4, 2019, and November 15, 2020. Exposures Acute myocardial infarction that required hospital admission. Main Outcomes and Measures The main outcome was thirty-day hospital readmission. A total of 141 candidate variables were considered from administrative codes, medication orders, and laboratory tests. Multiple risk prediction models were developed using parametric models (elastic net, least absolute shrinkage and selection operator, and ridge regression) and nonparametric models (random forest and gradient boosting). The models were assessed using holdout data with area under the receiver operating characteristic curve (AUROC), percentage of calibration, and calibration curve belts. Results The final Vanderbilt University Medical Center cohort included 6163 unique patients, among whom the mean (SD) age was 67 (13) years, 4137 were male (67.1%), 1019 (16.5%) were Black or other race, and 933 (15.1%) were rehospitalized within 30 days. The final Dartmouth-Hitchcock Medical Center cohort included 4024 unique patients, with mean (SD) age of 68 (12) years; 2584 (64.2%) were male, 412 (10.2%) were rehospitalized within 30 days, and most of the cohort were non-Hispanic and White. The final test set AUROC performance was between 0.686 to 0.695 for the parametric models and 0.686 to 0.704 for the nonparametric models. In the validation cohort, AUROC performance was between 0.558 to 0.655 for parametric models and 0.606 to 0.608 for nonparametric models. Conclusions and Relevance In this study, 5 machine learning models were developed and externally validated to predict 30-day readmission AMI hospitalization. These models can be deployed within an EHR using routinely collected data., This cohort study compares multiple externally validated machine learning models using electronic health record data to predict 30-day readmission among patients hospitalized for acute myocardial infarction.

Hazard ratios (HR) associated with the well-known proportional hazard Cox regression models are routinely used for measuring the impact of one factor of interest on a time-to-event outcome. However, if the underlying real model does not fit with the theoretical requirements, the interpretation of those HRs is not clear. We propose a new index, gHR, which generalizes the HR beyond the underlying survival model. We consider the case in which the study factor is a binary variable and we are interested in both the unadjusted and adjusted effect of this factor on a time-to-event variable, potentially, observed in a right-censored scenario. We propose non-parametric estimations for unadjusted gHR and semi-parametric regression-induced techniques for the adjusted case. The behavior of those estimators is studied in both large and finite sample situations. Monte Carlo simulations reveal that both estimators provide good approximations of their respective inferential targets. Data from the Health and Lifestyle Study are used for studying the relationship of the tobacco use and the age of death and illustrate the practical application of the proposed technique. gHR is a promising index which can help facilitate better understanding of the association of one study factor on a time-dependent outcome.

Background Older adults with obesity residing in rural areas have reduced access to weight management programs. We determined the feasibility, acceptability and preliminary outcomes of an integrated technology-based health promotion intervention in rural-living, older adults using remote monitoring and synchronous video-based technology. Methods A 6-month, non-randomized, non-blinded, single-arm study was conducted from October 2018 to May 2020 at a community-based aging center of adults aged ≥65 years with a body mass index (BMI) ≥30 kg/m2. Weekly dietitian visits focusing on behavior therapy and caloric restriction and twice-weekly physical therapist-led group strength, flexibility and balance training classes were delivered using video-conferencing to participants in their homes. Participants used a Fitbit Alta HR for remote monitoring with data feedback provided by the interventionists. An aerobic activity prescription was provided and monitored. Results Mean age was 72.9±3.9 years (82% female). Baseline anthropometric measures of weight, BMI, and waist circumference were 97.8±16.3 kg, 36.5±5.2 kg/m2, and 115.5±13.0 cm, respectively. A total of 142 participants were screened (n=27 ineligible), and 53 consented. There were nine dropouts (17%). Overall satisfaction with the trial (4.7+ 0.6, scale: 1 (low) to 5 (high)) and with Fitbit (4.2+ 0.9) were high. Fitbit was worn an average of 81.7±19.3% of intervention days. In completers, mean weight loss was 4.6±3.5 kg or 4.7±3.5% (pp=0.005) but no differences were observed in gait speed or grip strength. Subjective measures of late-life function improved (3.4±4.7 points, p Conclusions A technology-based obesity intervention is feasible and acceptable to older adults with obesity and may lead to weight loss and improved physical function. Clinical trial registration Registered on Clinicaltrials.gov #NCT03104205. Registered on April 7, 2017. First participant enrolled on October 1st, 2018.

PURPOSE: The study objective was to understand if features derived from continuous pulse oximetry data can provide advanced warning of pulseless electrical activity arrest in the general care inpatient setting. METHODS: Retrospective analysis of SpO2 and pulse rate data derived from continuous pulse oximetry was performed for pulseless electrical activity (n=38) and control (n=42) patient cohorts. Measures of central tendency and variation over time intervals ranging from 1 minute to 1 hour were used for inter- and intra-group comparisons. Logistic regression was applied to understand ability of features to predict pulseless electrical activity in future time intervals. RESULTS: Overall, the pulseless electrical activity arrest group tended to have lower mean SpO2 and higher mean pulse rate values than the control group. SpO2 and pulse rate variability was higher in the pulseless electrical activity arrest cohort. Changes in variability were observed beginning several hours prior to the rescue event. Up to 20 minutes before rescue events, pulse rate features were significantly different from feature values for the preceding 30-minute interval (>10% difference in mean, >46% difference in range). Similar results were found for SpO2 features 10 minutes before the event (>4% difference in mean, >60% difference in range). CONCLUSIONS: There is a significant difference in SpO2 and pulse rate features derived from continuous pulse oximetry between pulseless electrical activity and control groups. Integration of automated feature calculation and clinician notification into clinical monitoring and information systems may increase patient safety by supporting early detection of such events.

Failure to rescue events, or events involving preventable deaths from complications, are a significant contributor to inpatient mortality. While many interventions have been designed and implemented over several decades, this patient safety issue remains at the forefront of concern for most hospitals. In the first part of this study, the development and implementation of one type of highly studied and widely adopted rescue intervention, algorithm-based patient assessment tools, is examined. The analysis summarizes how a lack of systems-oriented approaches in the design and implementation of these tools has resulted in suboptimal understanding of patient risk of mortality and complications and the early recognition of patient deterioration. The gaps identified impact several critical aspects of excellent patient care, including information-sharing across care settings, support for the development of shared mental models within care teams, and access to timely and accurate patient information. This chapter describes the use of several system-oriented design and implementation activities to establish design objectives, model clinical processes and workflows, and create an extensible information system model to maximize the benefits of patient state and risk assessment tools in the inpatient setting. A prototype based on the product of the design activities is discussed along with system-level considerations for implementation. This study also demonstrates the effectiveness and impact of applying systems design principles and practices to real-world clinical applications.

Psychiatric consultation services, particularly for emergencies, are limited in many parts of the United States. Telepsychiatry services are helping to bridge the gap, and are gaining acceptance and popularity. There is paucity of publications regarding comparison of turnaround time for consultations between video conferencing and traditional face-to-face psychiatric consultations in general hospital nonpsychiatric emergency departments (EDs). Our study aimed to address turnaround time and patient satisfaction.Data regarding the turnaround time for emergency psychiatric consultations using telepsychiatry in general hospital EDs was collected retrospectively and compared with the time for face-to-face traditional consultations. A patient satisfaction survey was also conducted after the telepsychiatry consultation. Statistical analysis of the data was done after the study was completed.The telepsychiatry group included 206 participants and the control group had 186 participants. There was an 84% reduction in the turnaround time for telepsychiatry consults (95% confidence interval, 81% to 86%). A patient satisfaction survey showed 97% satisfaction with telepsychiatry services. Gender and age did not modify the effect of telepsychiatry on time to consult (P.10).The reduction in the turnaround time and improved patient satisfaction indicate that telepsychiatry services can improve the quality of care for patients in need of emergency services.

Estimation that employs instrumental variables (IV) can reduce or eliminate bias due to confounding. In observational studies instruments result from natural experiments such as the effect of clinician preference or geographic distance on treatment selection. In randomized studies the randomization indicator is an instrument, especially if the study is blinded, e.g. no placebo effect. Estimation via instruments is a highly developed field for linear models but the use of instruments in time-to-event analysis is far from established. Various IV-based estimators of the hazard ratio (HR) from Cox's regression models have been proposed. We extend IV based estimation of Cox's models beyond proportionality of hazards, and address estimation of a log-linear time dependent hazard ratio and a piecewise constant HR. We estimate the marginal time-dependent hazard ratio unlike other approaches that estimate the hazard ratio conditional on the omitted covariates. Due to the non-collapsibility of the Cox's models these two estimands are not identical. We report the results of simulations that includes the use of copulas to generate potential times-to-event that have a given marginal structural time dependent hazard ratio but are dependent on omitted covariates. We demonstrate the method to estimate the time dependent hazard ratio for two vascular interventions.

During hospital quality improvement activities, statistical approaches are critical to help assess hospital performance for benchmarking. Current statistical approaches are used primarily for research and reimbursement purposes. In this multiinstitutional study, these established statistical methods were evaluated for quality improvement applications. Leveraging a dataset of 42,199 patients who underwent coronary artery bypass grafting surgery from 2014 to 2016 across 90 hospitals, six statistical approaches were applied. The non-shrinkage methods were: (1) indirect standardization without hospital effect; (2) indirect standardization with hospital fixed effect; (3) direct standardization with hospital fixed effect. The shrinkage methods were: (4) indirect standardization with hospital random effect; (5) direct standardization with hospital random effect; (6) Bayesian method. Hospital performance related to operative mortality and major morbidity or mortality was compared across methods based on variation in adjusted rates, rankings, and performance outliers. Method performance was evaluated across procedure volume terciles: small (< 96 cases/year), medium (96-171), and large (> 171). Shrinkage methods reduced inter-hospital variation (min-max) for mortality (observed: 0%-10%; adjusted: 1.5%-2.4%) and major morbidity or mortality (observed: 2.6%-35%; adjusted: 6.9%-17.5%). Shrinkage methods shrunk hospital rates toward the group mean. Direct standardization with hospital random effect, compared to fixed effect, resulted in 16.7%-38.9% of hospitals changing quintile mortality ranking. Indirect standardization with hospital random effect resulted in no performance outliers among small and medium hospitals for mortality, while logistic and fixed effect methods identified one small and three medium outlier hospitals. The choice of statistical method greatly impacts hospital ranking and performance outlier’ status. These findings should be considered when benchmarking hospital performance for hospital quality improvement activities.

OBJECTIVE: Using screening mammography, this study investigated the association between obesity and axillary lymph node (LN) size and morphology. METHODS: We conducted a retrospective review of 188 females who underwent screening mammography at an academic medical centre. Length and width of the LN and hilum were measured in the largest, mammographically visible axillary node. The hilo-cortical ratio (HCR) was calculated as the hilar width divided by the cortical width. Measurements were performed by a board certified breast radiologist and a resident radiology physician. Inter-rater agreement was assessed with Pearson correlation coefficient. We performed multivariable regression analysis for associations of LN measurements with body mass index (BMI), breast density and age. RESULTS: There was a strong association between BMI and LN dimensions, hilum dimensions and HCR (p < 0.001 for all metrics). There was no significant change in cortex width with increasing BMI (p = 0.15). Increases in LN length and width were found with increasing BMI [0.6 mm increase in length per unit BMI, 95% CI (0.4–0.8), p < 0.001 and0.3 mm increase in width per unit BMI, 95% CI(0.2–0.4), p < 0.001, respectively]. Inter-rater reliability for lymph node and hilum measurements was 0.57–0.72. CONCLUSION: We found a highly significant association between increasing BMI and axillary LN dimensions independent of age and breast density with strong interobserver agreement. The increase in LN size was driven by expansion of the LN hilum secondary to fat infiltration. ADVANCES IN KNOWLEDGE: This preliminary work determined a relationship between fat infiltrated axillary lymph nodes and obesity.

Instrumental variable (IV) methods are widely used for estimating average treatment effects in the presence of unmeasured confounders. However, the capability of existing IV procedures, and most notably the two-stage residual inclusion (2SRI) procedure recommended for use in nonlinear contexts, to account for unmeasured confounders in the Cox proportional hazard model is unclear. We show that instrumenting an endogenous treatment induces an unmeasured covariate, referred to as an individual frailty in survival analysis parlance, which if not accounted for leads to bias. We propose a new procedure that augments 2SRI with an individual frailty and prove that it is consistent under certain conditions. The finite sample-size behavior is studied across a broad set of conditions via Monte Carlo simulations. Finally, the proposed methodology is used to estimate the average effect of carotid endarterectomy versus carotid artery stenting on the mortality of patients suffering from carotid artery disease. Results suggest that the 2SRI-frailty estimator generally reduces the bias of both point and interval estimators compared to traditional 2SRI., Comment: 27 pages, 8 figures, 4 tables