The median hazard ratio: a useful measure of variance and general contextual effects in multilevel survival analysis

Graft Survival of En Bloc Deceased Donor Kidneys Transplants Compared With Single Kidney Transplants

BACKGROUND

The US Kidney Allocation System allocates en bloc deceased donor kidney grafts from donors <18 kg in sequence A along with single kidney transplants (SKTs) from kidney donor profile index (KDPI) top 20% donors. Although en bloc grafts outperform SKT grafts holding donor weight constant, it is unclear if en bloc grafts from the smallest pediatric donors perform the same as top 20% KDPI SKTs.

METHODS

Using the Scientific Registry of Transplant Recipients, we compared the donor characteristics and graft survival of en bloc grafts from the smallest donors (<8 kg) and from larger donors (≥8 kg) with SKTs by KDPI sequence for transplants performed in 2021.

RESULTS

Larger donor en blocs had similar 1-y survival to sequence A SKTs estimated by the Kaplan-Meier method (96% versus 96%, P = 0.9), but the smallest donor en blocs had significantly shorter 1-y survival than those SKTs (80% versus 96%, P < 0.01). Using transplants from 2010 to 2012, the smallest donor en blocs had similar 10-y survival to sequence A SKTs (69% versus 64%, P  = 0.3).

CONCLUSIONS

These findings suggest that future updates of the Kidney Allocation System should include a score specific to pediatric donors to account for these differences in en bloc graft survival.

Predictors of COVID-19 infection and hospitalization in group homes for individuals with intellectual and/or developmental disabilities

BACKGROUND

More than seven million people with intellectual and/or developmental disabilities (ID/DD) live in the US and may face an elevated risk for COVID-19.

OBJECTIVE

To identify correlates of COVID-19 and related hospitalizations among people with ID/DD in group homes in Massachusetts.

METHODS

We collected data during March 1, 2020-June 30, 2020 (wave 1) and July 1, 2020-March 31, 2021 (wave 2) from the Massachusetts Department of Public Health and six organizations administering 206 group homes for 1035 residents with ID/DD. The main outcomes were COVID-19 infections and related hospitalizations. We fit multilevel Cox proportional hazards models to estimate associations with observed predictors and assess contextual home- and organizational-level effects.

RESULTS

Compared with Massachusetts residents, group home residents had a higher age-adjusted rate of COVID-19 in wave 1 (incidence rate ratio [IRR], 12.06; 95 % confidence interval [CI], 10.51-13.84) and wave 2 (IRR, 2.47; 95 % CI, 2.12-2.88) and a higher age-adjusted rate of COVID-19 hospitalizations in wave 1 (IRR, 17.64; 95 % CI, 12.59-24.70) and wave 2 (IRR, 4.95; 95 % CI, 3.23-7.60). COVID-19 infections and hospitalizations were more likely among residents aged 65+ and in group homes with 6+ resident beds and recent infection among staff and residents.

CONCLUSIONS

Aggressive efforts to decrease resident density, staff-to-resident ratios, and staff infections through efforts such as vaccination, in addition to ongoing access to personal protective equipment and COVID-19 testing, may reduce COVID-19 and related hospitalizations in people with ID/DD living in group homes.

Frequency of antimicrobial-resistant bloodstream infections in 111 hospitals in Thailand, 2022

OBJECTIVES

To evaluate the frequency of antimicrobial-resistant bloodstream infections (AMR BSI) in Thailand.

METHODS

We analyzed data from 2022, generated by 111 public hospitals in health regions 1 to 12, using the AutoMated tool for Antimicrobial resistance Surveillance System (AMASS), and submitted to the Ministry of Public Health, Thailand. Multilevel Poisson regression models were used.

RESULTS

The most common cause of community-origin AMR BSI was third-generation cephalosporin-resistant Escherichia coli (3GCREC, 65.6%; 5101/7773 patients) and of hospital-origin AMR BSI was carbapenem-resistant Acinetobacter baumannii (CRAB, 51.2%, 4968/9747 patients). The percentage of patients tested for BSI was negatively associated with the frequency of community-origin 3GCREC BSI and hospital-origin CRAB BSI (per 100,000 tested patients). Hospitals in health regions 4 (lower central region) had the highest frequency of community-origin 3GCREC BSI (adjusted incidence rate ratio, 2.06; 95% confidence interval: 1.52-2.97). Health regions were not associated with the frequency of hospital-origin CRAB BSI, and between-hospital variation was high, even adjusting for hospital level and size.

CONCLUSION

The high between-hospital variation of hospital-origin CRAB BSI suggests the importance of hospital-specific factors. Our approach and findings highlight health regions and hospitals where actions against AMR infection, including antimicrobial stewardship and infection control, should be prioritized.

Variability in short-term mortality following repair of ruptured abdominal aortic aneurysms across centers and physicians

BACKGROUND

Variation in the care management of repairs for ruptured infrarenal abdominal aortic aneurysms between centers and physicians, such as procedural volumes, may explain differences in mortality outcomes. First, we quantified the center and physician variability associated with 30- and 90-day mortality risk after ruptured open surgical repair (rOSR) and ruptured endovascular aneurysm repair (rEVAR). Second, we explored wheter part of this variability was attributable to procedural volume at the center and physician levels.

METHODS

Two cohorts including rOSR and rEVAR procedures between 2013 and 2019 were analyzed from the Vascular Quality Initiative database. Thirty- and 90-day all-cause mortality rates were derived from linked Medicare claims data. The median odds ratio (MOR) (median mortality risk from low- to high-risk cluster) and intraclass correlation coefficient (ICC) (variability attributable to each cluster) for 30- and 90-day mortality risks associated with center and physician variability were derived using patient-level adjusted multilevel logistic regression models. Procedural volume was calculated at the center and physician levels and stratified by quartiles. The models were sequentially adjusted for volumes, and the difference in ICCs (without vs with accounting for volume) was calculated to describe the center and physician variability in mortality risk attributable to volumes.

RESULTS

We included 450 rOSRs (mean age, 74.5 ± 7.6 years; 23.5% female) and 752 rEVARs (76.4 ± 8.4 years; 26.1% female). After rOSRs, the 30- and 90-day mortality rates were 32.9% and 38.7%, respectively. No variability across centers and physicians was noted (30- and 90-day MORs ≈1 and ICCs ≈0%). Neither center nor physician volume was associated with 30-day (P = .477 and P = .796) or 90-day mortality (P = .098 and P = .559). After rEVAR, the 30- and 90-day mortality rates were 21.3% and 25.5%, respectively. Significant center variability (30-day MOR, 1.82 [95% confidence interval (CI), 1.33-2.22]; ICC, 11% [95% CI, 2%-36%]; and 90-day MOR, 1.76 [95% CI, 1.37-2.09]; ICC, 10% [95% CI, 3%-30%]), but negligeable variability across physicians (30- and 90-day MORs ≈1 and ICCs ≈0%) were noted. Neither center nor physician volume were associated with 30-day (P = .076 and P = .336) or 90-day mortality risk (P = .066 and P = .584). The center variability attributable to procedural volumes was negligeable (difference in ICCs, 1% for 30-day mortality; 0% for 90-day mortality).

CONCLUSIONS

Variability in practice from center to center was associated with short-term mortality outcomes in rEVAR, but not for rOSR. Physician variability was not associated with short-term mortality for rOSR or rEVAR. Annualized center and physician volumes did not significantly explain these associations. Further work is needed to identify center-level factors affecting the quality of care and outcomes for ruptured abdominal aortic aneurysms.

Survival Analysis Without Sharing of Individual Patient Data by Using a Gaussian Copula

Cox regression and Kaplan-Meier estimations are often needed in clinical research and this requires access to individual patient data (IPD). However, IPD cannot always be shared because of privacy or proprietary restrictions, which complicates the making of such estimations. We propose a method that generates pseudodata replacing the IPD by only sharing non-disclosive aggregates such as IPD marginal moments and a correlation matrix. Such aggregates are collected by a central computer and input as parameters to a Gaussian copula (GC) that generates the pseudodata. Survival inferences are computed on the pseudodata as if it were the IPD. Using practical examples we demonstrate the utility of the method, via the amount of IPD inferential content recoverable by the GC. We compare GC to a summary-based meta-analysis and an IPD bootstrap distributed across several centers. Other pseudodata approaches are also considered. In the empirical results, GC approximates the utility of the IPD bootstrap although it might yield more conservative inferences and it might have limitations in subgroup analyses. Overall, GC avoids many legal problems related to IPD privacy or property while enabling approximation of common IPD survival analyses otherwise difficult to conduct. Sharing more IPD aggregates than is currently practiced could facilitate "second purpose"-research and relax concerns regarding IPD access.

A Radiomic Approach for Evaluating Intra-Subgroup Heterogeneity in SHH and Group 4 Pediatric Medulloblastoma: A Preliminary Multi-Institutional Study

Medulloblastoma (MB) is the most frequent malignant brain tumor in children with extensive heterogeneity that results in varied clinical outcomes. Recently, MB was categorized into four molecular subgroups, WNT, SHH, Group 3, and Group 4. While SHH and Group 4 are known for their intermediate prognosis, studies have reported wide disparities in patient outcomes within these subgroups. This study aims to create a radiomic prognostic signature, medulloblastoma radiomics risk (mRRisk), to identify the risk levels within the SHH and Group 4 subgroups, individually, for reliable risk stratification. Our hypothesis is that this signature can comprehensively capture tumor characteristics that enable the accurate identification of the risk level. In total, 70 MB studies (48 Group 4, and 22 SHH) were retrospectively curated from three institutions. For each subgroup, 232 hand-crafted features that capture the entropy, surface changes, and contour characteristics of the tumor were extracted. Features were concatenated and fed into regression models for risk stratification. Contrasted with Chang stratification that did not yield any significant differences within subgroups, significant differences were observed between two risk groups in Group 4 (p = 0.04, Concordance Index (CI) = 0.82) on the cystic core and non-enhancing tumor, and SHH (p = 0.03, CI = 0.74) on the enhancing tumor. Our results indicate that radiomics may serve as a prognostic tool for refining MB risk stratification, towards improved patient care.

Variation in the Care of Children with Inflammatory Bowel Disease Within and Across Canadian Provinces: A Multi-Province Population-Based Cohort Study

Purpose

The incidence of childhood-onset inflammatory bowel disease (IBD) is rising. We described variation in health services utilization and need for surgery among children with IBD between six and 60 months following IBD diagnosis across Canadian pediatric centers and evaluated the associations between care provided at diagnosis at each center and the variation in these outcomes.

Patients and Methods

Using population-based deterministically-linked health administrative data from four Canadian provinces (Alberta, Manitoba, Nova Scotia, Ontario) we identified children diagnosed with IBD <16 years of age using validated algorithms. Children were assigned to a pediatric center of care using a hierarchical approach based on where they received their initial care. Outcomes included IBD-related hospitalizations, emergency department (ED) visits, and IBD-related abdominal surgery occurring between 6 and sixty months after diagnosis. Mixed-effects meta-analysis was used to pool results and examine the association between center-level care provision and outcomes.

Results

We identified 3784 incident cases of pediatric IBD, of whom 2937 (77.6%) were treated at pediatric centers. Almost a third (31.4%) of children had ≥1 IBD-related hospitalization and there were 0.66 hospitalizations per person during follow-up. More than half (55.8%) of children had ≥1 ED visit and there were 1.64 ED visits per person. Between-center heterogeneity was high for both outcomes; centers where more children visited the ED at diagnosis had more IBD-related hospitalizations and more ED visits during follow-up. Between-center heterogeneity was high for intestinal resection in Crohn's disease but not colectomy in ulcerative colitis.

Conclusion

There is variation in health services utilization among children with IBD and risk of undergoing intestinal resection in those with Crohn's disease, but not colectomy among children with ulcerative colitis, across Canadian pediatric tertiary-care centers. Improvements in clinical care pathways are needed to ensure all children have equitable and timely access to high quality care.

COVID-19-Related Outcomes Among Group Home Residents with Serious Mental Illness in Massachusetts in the First Year of the Pandemic

This study examined COVID-19 infection and hospitalizations among people with serious mental illness who resided in residential care group homes in Massachusetts during the first year of the COVID-19 pandemic. The authors analyzed data on 2261 group home residents and COVID-19 data from the Massachusetts Department of Public Health. Outcomes included positive COVID-19 tests and COVID-19 hospitalizations March 1, 2020-June 30, 2020 (wave 1) and July 1, 2020-March 31, 2021 (wave 2). Associations between hazard of outcomes and resident and group home characteristics were estimated using multi-level Cox frailty models including home- and city-level frailties. Between March 2020 and March 2021, 182 (8%) residents tested positive for COVID-19, and 51 (2%) had a COVID-19 hospitalization. Compared with the Massachusetts population, group home residents had age-adjusted rate ratios of 3.0 (4.86 vs. 1.60 per 100) for COVID infection and 13.5 (1.99 vs. 0.15 per 100) for COVID hospitalizations during wave 1; during wave 2, the rate ratios were 0.5 (4.55 vs. 8.48 per 100) and 1.7 (0.69 vs. 0.40 per 100). In Cox models, residents in homes with more beds, higher staff-to-resident ratios, recent infections among staff and other residents, and in cities with high community transmission risk had greater hazard of COVID-19 infection. Policies and interventions that target group home-specific risks are needed to mitigate adverse communicable disease outcomes in this population.Clinical Trial Registration Number This study provides baseline (i.e., pre-randomization) data from a clinical trial study NCT04726371.

Biological Age Mediates the Effects of Perceived Neighborhood Problems on Heart Failure Risk Among Black Persons

OBJECTIVE

We assessed whether biological age, measured by the epigenetic clock GrimAge, mediates the association of objective and subjective neighborhood disadvantage with incident HF among Black persons.

METHODS

Participants were 1448 self-reported Black adults (mean age (standard deviation, SD) = 64.3 (5.5)) dually enrolled in two community-based cohorts in Jackson, Mississippi, the ARIC and JHS cohorts, who were free of HF as of January 1, 2000. Incident HF events leading to hospitalization through December 31, 2017, were classified using ICD-9 discharge codes of HF. Multilevel age- and sex-adjusted Cox causal mediation models were used to examine whether biological age (at the person and neighborhood level) mediated the effects of objective (the National Area Deprivation Index, ADI) and subjective (perceived neighborhood problems) neighborhood disadvantage on incident HF.

RESULTS

A total of 334 incident hospitalized HF events occurred over a median follow-up of 18.0 years. The total effect of the ADI and perceived neighborhood problems (SD units) on HF was hazard ration (HR) = 1.26 and 95% confidence interval (CI) 0.98-1.56 and HR = 1.26 and 95% CI 1.10-1.41, respectively. GrimAge mediated a majority of the effect of perceived neighborhood problems on HF (person-level indirect effect HR = 1.07; 95% CI 1.02-1.12 and neighborhood-level indirect effect HR = 1.18; 95% CI 1.03-1.34), with the combined indirect effect explaining 94.8% of the relationship. The combined indirect effect of ADI on incident HF was comparable but not statistically significant.

CONCLUSIONS

Subjective neighborhood disadvantage may confer an increased risk of HF among Black populations.

Distinct moderating pathways for psychosocial risk and resilience in the association of neighborhood disadvantage with incident heart failure among Black persons

Objective

To assess whether psychosocial factors moderate the associations between neighborhood disadvantage and incident heart failure (HF).

Methods

Among 1448 Non-Hispanic (NH) Black persons dually enrolled in two community-based cohorts in Jackson, Mississippi who were free of HF as of January 1, 2000, 336 HF events classified by reviewer panel accrued through December 31, 2017. Multilevel, multivariable Cox regression models were used to examine whether optimism and negative affect moderated the associations of two measures of neighborhood characteristics (the national Area Deprivation Index (ADI) and perceived neighborhood problems) on incident hospitalized HF.

Results

Optimism moderated the association of the ADI with incident HF. Compared to participants reporting the lowest tertile of optimism, those in the highest tertile of optimism had a 29% lower rate of HF associated with increasing ADI in fully adjusted models. We found no evidence for a moderating effect of negative affect.

Conclusions

This study supports optimism as a source of resilience to the detrimental effects of neighborhood disadvantage on HF risk. Population-level strategies to promote sociocultural antecedents to optimism may serve as a viable method of reducing the disproportionate burden of HF among NH Black persons.

Construction of multilevel statistical models in health research: Foundations and generalities

This topic review aims to present a global vision of multilevel analysis models’ applicability to health research, explaining its theoretical, methodological, and statistical foundations. We describe the basic steps to build these models and examples of their application according to the data hierarchical structure. It ir worth noticing that before using these models, researchers must have a rationale for needing them, and a statistical evaluation accounting for the variance percentage explained by the observations grouping effect. The requirements to conduct this type of analysis depends on special conditions such as the type of variables, the number of units per level, or the type of hierarchical structure. We conclude that multilevel analysis models are a useful tool to integrate information, considering the complexity of the relationships and interactions involved in most health conditions, including the loss of independence between observation units.

Comparing High- and Low-Model for End-Stage Liver Disease Living-Donor Liver Transplantation to Determine Clinical Efficacy: A Systematic Review and Meta-Analysis (CHALICE Study)

INTRODUCTION

Various studies have demonstrated that low-Model for End-Stage Liver Disease (MELD) living-donor liver transplant (LDLT) recipients have better outcomes with improved patient survival than deceased-donor liver transplantation (DDLT) recipients. LDLT recipients gain the most from being transplanted at MELD <25-30; however, some existing data have outlined that LDLT may provide equivalent outcomes in high-MELD and low-MELD patients, although the term "high" MELD is arbitrarily defined in the literature and various cut-off scores are outlined between 20 and 30, although most commonly, the dividing threshold is 25. The aim of this meta-analysis was to compare LDLT in high-MELD with that in low-MELD recipients to determine patient survival and graft survival, as well as perioperative and postoperative complications.

METHODS

Following PROSPERO registration CRD-42021261501, a systematic database search was conducted for the published literature between 1990 and 2021 and yielded a total of 10 studies with 2183 LT recipients; 490 were HM-LDLT recipients and 1693 were LM-LDLT recipients.

RESULTS

Both groups had comparable mortality at 1, 3 and 5 years post-transplant (5-year HR 1.19; 95% CI 0.79-1.79; p-value 0.40) and graft survival (HR 1.08; 95% CI 0.72, 1.63; p-value 0.71). No differences were observed in the rates of major morbidity, hepatic artery thrombosis, biliary complications, intra-abdominal bleeding, wound infection and rejection; however, the HM-LDLT group had higher risk for pulmonary infection, abdominal fluid collection and prolonged ICU stay.

CONCLUSIONS

The high-MELD LDLT group had similar patient and graft survival and morbidities to the low-MELD LDLT group, despite being at higher risk for pulmonary infection, abdominal fluid collection and prolonged ICU stay. The data, primarily sourced from high-volume Asian centers, underscore the feasibility of living donations for liver allografts in high-MELD patients. Given the rising demand for liver allografts, it is sensible to incorporate these insights into U.S. transplant practices.

The Mutation Portraits of Oncogenes and Tumor Supressor Genes in Predicting the Overall Survival in Pancreatic Cancer: A Bayesian Network Meta-Analysis

INTRODUCTION

In pancreatic cancer, the carcinogenesis can not be separated from genetics mutations. The portraits of genes alterations majorily including oncogenes (KRAS, HER2, PD-L1) and tumor supressor genes (P53, CDKN2A, SMAD4). Besides being notorious a screening marker, the genetic mutations were related to the prognosis of pancreatic cancer. The aim of this study is to determine the genetic mutations portrait in predicting the overall survival in pancreatic cancer.

METHODS

The network meta analysis (NMA) was registered in PROSPERO (CRD42023397976) and conducted in accordance with the PRISMA-P (Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols) in addition of NMA extension guidance. Comprehensive searches were done including all studies which reported the overall survival of pancreatic cancer subjects with KRAS, HER2, PD-L1, P53, CDKN2A, SMAD4. Data were collected and analysis will be done based on Bayesian method, Markov Chain Monte Carlo algorithm, using BUGSnet package in R studio. Transivity was controlled by methods and consistency of the NMA will be fitted by deviance information criterion. Data analysis in NMA were presented in Sucra plot, league table, and forest plot.

RESULTS

Twenty-four studies were included in this NMA with 4613 total subjects. The NMA was conducted in random-effects, consistent, and convergence model. Relative to control, the genetic mutation of SMAD4 (HR 1.84; 95%CI 1.39-2.46), HER2 (HR 1.76; 95%CI 1.14-2.71), and KRAS (HR 1.7; 95%CI 1.19-2.48) were significant to have worse survival. The mutations of PD-L1, P53, and CDKN2A also showed poor survival, but not statistically significant compared to control.

CONCLUSION

In pancreatic cancer, the mutation of SMAD4 predicted the worst overall survival, compared to control, also mutation of HER2, KRAS, PD-L1, P53, and CDKN2A.

Anticoagulation improves survival in patients with cirrhosis and portal vein thrombosis: The IMPORTAL competing-risk meta-analysis

BACKGROUND & AIMS

Previous meta-analyses demonstrated the safety and efficacy of anticoagulation in the recanalization of portal vein thrombosis in patients with cirrhosis. Whether this benefit translates into improved survival is unknown. We conducted an individual patient data (IPD) meta-analysis to assess the effect of anticoagulation on all-cause mortality in patients with cirrhosis and portal vein thrombosis.

METHODS

In this IPD meta-analysis, we selected studies comparing anticoagulation vs. no treatment in patients with cirrhosis and portal vein thrombosis from PubMed, Embase, and Cochrane databases (until June 2020) (PROSPERO no.: CRD42020140026). IPD were subsequently requested from authors. The primary outcome - the effect of anticoagulation on all-cause mortality - was assessed by a one-step meta-analysis based on a competing-risk model with liver transplantation as the competing event. The model was adjusted for clinically relevant confounders. A multilevel mixed-effects logistic regression model was used to determine the effect of anticoagulation on recanalization.

RESULTS

Individual data on 500 patients from five studies were included; 205 (41%) received anticoagulation and 295 did not. Anticoagulation reduced all-cause mortality (adjusted subdistribution hazard ratio 0.59; 95% CI 0.49-0.70), independently of thrombosis severity and recanalization. The effect of anticoagulation on all-cause mortality was consistent with a reduction in liver-related mortality. The recanalization rate was higher in the anticoagulation arm (adjusted odds ratio 3.45; 95% CI 2.22-5.36). The non-portal-hypertension-related bleeding rate was significantly greater in the anticoagulation group.

CONCLUSIONS

Anticoagulation reduces all-cause mortality in patients with cirrhosis and portal vein thrombosis independently of recanalization, but at the expense of increasing non-portal hypertension-related bleeding.

PROSPERO REGISTRATION NUMBER

CRD42020140026.

IMPACT AND IMPLICATIONS

Anticoagulation is effective in promoting recanalization of portal vein thrombosis in patients with cirrhosis, but whether this benefit translates into improved survival is controversial. Our individual patient data meta-analysis based on a competing-risk model with liver transplantation as the competing event shows that anticoagulation reduces all-cause mortality in patients with cirrhosis and portal vein thrombosis independently of recanalization. According to our findings, portal vein thrombosis may identify a group of patients with cirrhosis that benefit from long-term anticoagulation.

The role of geography and distance on physician follow-up after a first hospitalization with a diagnosis of a schizophrenia spectrum disorder: A retrospective population-based cohort study in Ontario, Canada

BACKGROUND

Timely follow-up after hospitalization for a schizophrenia spectrum disorder (SSD) is an important quality indicator. We examined the proportion of individuals who received physician follow-up within 7 and 30 days post-discharge by health region and estimated the effect of distance between a person's residence and discharging hospital on follow-up.

METHODS

We created a retrospective population-based cohort of incident hospitalizations with a discharge diagnosis of a SSD between 01/01/2012 and 30/03/2019. The proportion of follow-up with a psychiatrist and family physician within 7 and 30 days were calculated for each region. The effect of distance between a person's residence and discharging hospital on follow-up was estimated using adjusted multilevel logistic regression models.

RESULTS

We identified 6,382 incident hospitalizations for a SSD. Only 14.2% and 49.2% of people received follow-up care with a psychiatrist within 7 and 30 days of discharge, respectively, and these proportions varied between regions. Although distance from hospital was not associated with follow-up within 7 days of discharge, increasing distance was associated with lower odds of follow-up with a psychiatrist within 30 days.

CONCLUSION

Post-discharge follow-up is poor across the province. Geospatial factors may impact post-discharge care and should be considered in further evaluation of quality of care.

An observational study of ear-tagged calf mortality (1 to 100 days) on Irish dairy farms and associations between biosecurity practices and calf mortality on farms participating in a Johne's disease control program

Postnatal mortality among replacement stock has a detrimental effect on the social, economic, and environmental sustainability of dairy production. Calf mortality rates vary between countries and show differences in temporal trends; most, however, are characterized by high levels of between-farm variability. Explaining this variation can be difficult because herd-level information on management practices relevant to calf health is often not available. The Irish Johne's Control Programme (IJCP) contains a substantial on-farm monitoring program called the Veterinary Risk Assessment and Management Plan (VRAMP). Although this risk assessment is largely focused on factors relevant to the transmission of paratuberculosis, many of its principles are good practice biocontainment policies that are also advocated for the protection of calf health. The objectives of this study were (1) to quantify mortality in ear-tagged Irish dairy calves between 2016 and 2020 using both survival and risk approaches, (2) to determine risk factors for 100-d cumulative mortality hazard in ear-tagged Irish dairy calves between 2016 and 2020, (3) to determine whether 100-d cumulative mortality hazard was higher in ear-tagged calves within herds registered in the IJCP versus those that were not registered in the IJCP and whether there were differences between these cohorts over time, and (4) within IJCP herds, to determine whether VRAMP score or changes in VRAMP score were associated with 100-d cumulative mortality hazard. Excluding perinatal mortality, the overall 100-d cumulative mortality hazard was 4.1%. Calf mortality was consistently underestimated using risk approaches that did not account for calf censoring. Cox proportional hazards models showed that cumulative mortality hazard was greater in male calves; particularly, calves born to Jersey breed dams and those with a beef breed sire. Mortality hazard increased with increasing herd size, was highest in calves born in herds that contract-reared heifers, and lowest in those born in mixed dairy-beef enterprises. Mortality hazard decreased over time with the mortality hazard in 2020 being 0.83 times that of 2016. Mortality hazard was higher in IJCP-registered herds than nonregistered herds (hazard ratio 1.06, 95% CI 1.01-1.12), likely reflecting differences in herds that enrolled in the national program. However, we detected a significant interaction between IJCP status (enrolled vs. not enrolled) and year (hazard ratio 0.96, 95% CI 0.92-1.00), indicating that the decrease in mortality hazard between 2016 and 2020 was greater in IJCP herds versus non-IJCP herds. Finally, increasing VRAMP scores (indicating higher risk for paratuberculosis transmission) were positively associated with increased calf mortality hazard. Postnatal calf mortality rates in Irish dairy herds declined between 2016 and 2020. Our study suggests that implementation of recommended biocontainment practices to control paratuberculosis in IJCP herds was associated with a reduction in calf mortality hazard.

Neighbourhood-level social capital, marginalisation, and the incidence of schizophrenia and schizoaffective disorder in Toronto, Canada: a retrospective population-based cohort study

BACKGROUND

Studies have shown mixed results regarding social capital and the risk of developing a psychotic disorder, and this has yet to be studied in North America. We sought to examine the relationship between neighbourhood-level marginalisation, social capital, and the incidence of schizophrenia and schizoaffective disorder in Toronto, Canada.

METHODS

We used a retrospective population-based cohort to identify incident cases of schizophrenia and schizoaffective disorder over a 10 year period and accounted for neighbourhood-level marginalisation and a proxy indicator of neighbourhood social capital. Mixed Poisson regression models were used to estimate adjusted incidence rate ratios (aIRRs).

RESULTS

In the cohort (n = 649 020) we identified 4841 incident cases of schizophrenia and schizoaffective disorder. A 27% variation in incidence was observed between neighbourhoods. All marginalisation dimensions, other than ethnic concentration, were associated with incidence. Compared to areas with low social capital, areas with intermediate social capital in the second [aIRR = 1.17, 95% confidence interval (CI) 1.03-1.33] and third (aIRR = 1.23, 95% CI 1.08-1.40) quintiles had elevated incidence rates after accounting for marginalisation. There was a higher risk associated with the intermediate levels of social capital (aIRR = 1.18, 95% CI 1.00-1.39) when analysed in only the females in the cohort, but the CI includes the possibility of a null effect.

CONCLUSIONS

The risk of developing schizophrenia and schizoaffective disorder in Toronto varies by neighbourhood and is associated with socioenvironmental exposures. Social capital was not linearly associated with risk, and risk differs by sex and social capital quintile. Future research should examine these relationships with different forms of social capital and examine how known individual-level risk factors impact these findings.

A Multi-Level Analysis of Individual and Neighborhood Factors Associated with Patient Portal Use among Adult Emergency Department Patients with Multimorbidity

Background: Patient portals tethered to electronic health records (EHR) have become vital to patient engagement and better disease management, specifically among adults with multimorbidity. We determined individual and neighborhood factors associated with patient portal use (MyChart) among adult patients with multimorbidity seen in an Emergency Department (ED). Methods: This study adopted a cross-sectional study design and used a linked database of EHR from a single ED site to patients’ neighborhood characteristics (i.e., zip code level) from the American Community Survey. The study population included all adults (age > 18 years), with at least one visit to an ED and multimorbidity between 1 January 2019 to 31 December 2020 (N = 40,544). Patient and neighborhood characteristics were compared among patients with and without MyChart use. Random-intercept multi-level logistic regressions were used to analyze the associations of patient and neighborhood factors with MyChart use. Results: Only 19% (N = 7757) of adults with multimorbidity used the patient portal. In the fully adjusted multi-level model, at the patient level, having a primary care physician (AOR = 5.55, 95% CI 5.07−6.07, p < 0.001) and health insurance coverage (AOR = 2.41, 95% CI 2.23−2.61, p < 0.001) were associated with MyChart use. At the neighborhood level, 4.73% of the variation in MyChart use was due to differences in neighborhood factors. However, significant heterogeneity existed in patient portal use when neighborhood characteristics were included in the model. Conclusions: Among ED patients with multimorbidity, one in five adults used patient portals. Patient-level factors, such as having primary care physicians and insurance, may promote patient portal use.

Variation in Benzodiazepine and Antipsychotic Prescribing Among Hospice Agencies

BACKGROUND

Benzodiazepines and antipsychotics are routinely prescribed for symptom management in hospice. There is minimal evidence to guide prescribing in this population, and little is known about how prescribing varies across hospice agencies.

OBJECTIVE

Examine patient- and hospice agency-level characteristics associated with incident prescribing of benzodiazepines and antipsychotics in hospice.

DESIGN

Retrospective cohort study of a 20% sample of Medicare beneficiaries newly enrolled in hospice.

PARTICIPANTS

Medicare hospice beneficiaries ≥ 65 years old between 2014 and 2016, restricted to those without benzodiazepine (N = 169,688) or antipsychotic (N = 190,441) prescription fills in the 6 months before hospice enrollment.

MAIN MEASURES

The primary outcome was incident (i.e., new) prescribing of a benzodiazepine or antipsychotic. A series of multilevel Cox regression models with random intercepts for hospice agency were fit to examine the association of incident benzodiazepine and antipsychotic prescribing with patient and hospice agency characteristics.

KEY RESULTS

A total of 91,728 (54.1%) and 58,175 (30.5%) hospice beneficiaries were newly prescribed an incident benzodiazepine or antipsychotic. The prescribing rate of the hospice agency was the strongest predictor of incident prescribing: Compared to patients in bottom-quartile benzodiazepine-prescribing agencies, those in top-quartile agencies were 10.7 times more likely to be prescribed an incident benzodiazepine (adjusted hazard ratio [AHR] 10.7, 95% CI 10.1-11.3). For incident antipsychotic prescribing, patients in top-quartile agencies were 51.7 times more likely to receive an antipsychotic (AHR 51.7, 95% CI 44.3-60.4) compared to those in the bottom quartile. Results remained consistent accounting for comfort kit prescribing.

CONCLUSIONS

The pattern of benzodiazepine or antipsychotic prescribing of a hospice agency strongly predicts whether a hospice enrollee is prescribed these medications, exceeding every other patient-level factor. While the appropriate level of prescribing in hospice is unclear, this variation may reflect a strong local prescribing culture across individual hospice agencies.

Neighborhood social context and suicide mortality: A multilevel register-based 5-year follow-up study of 2.7 million individuals

Previous studies have linked neighborhood social characteristics to suicide mortality. However, the effects of the operational definition of neighborhoods and the general importance of neighborhood context on suicide mortality have received little attention, with most studies using various administrative areas as neighborhood delineations. In this study, neighborhoods were delineated by micro-areas generated with an automated redistricting algorithm and divided by physical barriers, such as large roads. The geographic data were linked to register data on the Danish adult population in the age range of 20-59 years in December 2013 (N = 2,672,799 individuals nested into 7943 neighborhoods). This cohort was followed for five years to evaluate the association between suicide mortality and neighborhood socioeconomic deprivation, social fragmentation, and population density. We used the median hazard ratio (MHR) to quantify the general contextual effect (GCE) of neighborhoods on suicide mortality and hazard ratios to quantify the specific contextual effects (SCEs) using multilevel survival models stratified by age group. The results showed a larger GCE and larger SCEs of neighborhoods on suicide mortality for individuals aged 20-39 years compared with those aged 40-59 years. After controlling for individual characteristics, higher suicide mortality was observed for individuals living in the least densely populated neighborhoods and the most socially fragmented neighborhoods for both age groups. We found cross-level interactions between neighborhood population density and gender and ethnicity for those aged 40-59 years, as well as between neighborhood social fragmentation and ethnicity for those aged 20-39 years. The results indicate that beyond individual characteristics, the neighborhood social context may affect the risk of suicide, especially for people aged 20-39 years.

Duration of inter-pregnancy interval and its predictors among pregnant women in urban South Ethiopia: Cox gamma shared frailty modeling

Background

Short inter-pregnancy interval is a public health concern because it results in adverse perinatal outcomes such as postpartum hemorrhage, anemia, premature birth, low birth weight, and perinatal deaths. Although it is critical to understand the factors that contribute to short inter-pregnancy interval to reduce the risk of these negative outcomes, adequate evidence about the factors in the urban context is lacking. Therefore, we aimed to assess the duration of the inter-pregnancy interval and its predictors among pregnant women in urban South Ethiopia.

Methods

A community-based retrospective follow-up study was conducted among 2171 pregnant women in five geographically diverse urban settings in South Ethiopia. For the analysis, a Cox gamma shared frailty (random-effect) model was used. Adjusted hazard ratio (AHR) with a 95% CI was used to assess significant predictors. The median hazard ratio (MHR) used to report clustering effect.

Results

The median duration of the inter-pregnancy interval was 22 months, 95% CI (21, 23), with an inter-quartile range of 14 months. Maternal age ≥30 years [AHR = 0.75, 95% CI: 0.58, 0.97], having no formal education [AHR = 0.60, 95% CI: 0.46, 0.78], contraceptive non-use [AHR = 2.27, 95% CI: 1.94, 2.66], breastfeeding for <24 months [AHR = 4.92, 95% CI: 3.95, 6.12], death of recent child [AHR = 2.90, 95% CI: 1.41, 5.97], plan pregnancy within 24 months [AHR = 1.72, 95% CI: 1.26, 2.35], lack of discussion with husband [AHR = 1.33, 95% CI: 1.10, 1.60] and lack of husband encouragement about pregnancy spacing [AHR = 1.25, 95% CI: 1.05, 1.48] were predictors of short inter-pregnancy interval. Adjusting for predictors, the median increase in the hazard of short inter-pregnancy interval in a cluster with higher short inter-pregnancy interval is 30% [MHR = 1.30, 95% CI: 1.11, 1.43] than lower cluster.

Conclusions

In the study settings, the duration of the inter-pregnancy interval was shorter than the World Health Organization recommendation. There is a need to improve contraceptive use and breastfeeding duration to maximize the inter-pregnancy interval. Men’s involvement in reproductive health services and advocacy for women’s reproductive decision-making autonomy are fundamental. The contextual disparities in the inter-pregnancy interval suggests further study and interventions.

Radiomic Deformation and Textural Heterogeneity (R-DepTH) Descriptor to Characterize Tumor Field Effect: Application to Survival Prediction in Glioblastoma

The concept of tumor field effect implies that cancer is a systemic disease with its impact way beyond the visible tumor confines. For instance, in Glioblastoma (GBM), an aggressive brain tumor, the increase in intracranial pressure due to tumor burden often leads to brain herniation and poor outcomes. Our work is based on the rationale that highly aggressive tumors tend to grow uncontrollably, leading to pronounced biomechanical tissue deformations in the normal parenchyma, which when combined with local morphological differences in the tumor confines on MRI scans, will comprehensively capture tumor field effect. Specifically, we present an integrated MRI-based descriptor, radiomic-Deformation and Textural Heterogeneity (r-DepTH). This descriptor comprises measurements of the subtle perturbations in tissue deformations throughout the surrounding normal parenchyma due to mass effect. This involves non-rigidly aligning the patients' MRI scans to a healthy atlas via diffeomorphic registration. The resulting inverse mapping is used to obtain the deformation field magnitudes in the normal parenchyma. These measurements are then combined with a 3D texture descriptor, Co-occurrence of Local Anisotropic Gradient Orientations (COLLAGE), which captures the morphological heterogeneity and infiltration within the tumor confines, on MRI scans. In this work, we extensively evaluated r-DepTH for survival risk-stratification on a total of 207 GBM cases from 3 different cohorts (Cohort 1 ( n1 = 53 ), Cohort 2 ( n2 = 75 ), and Cohort 3 ( n3 = 79 )), where each of these three cohorts was used as a training set for our model separately, and the other two cohorts were used for testing, independently, for each training experiment. When employing Cohort 1 for training, r-DepTH yielded Concordance indices (C-indices) of 0.7 and 0.65, hazard ratios (HR) and Confidence Intervals (CI) of 10 (6 - 19) and 5 (3 - 8) on Cohorts 2 and 3, respectively. Similarly, training on Cohort 2 yielded C-indices of 0.6 and 0.7, HR and CI of 1 (0.7 - 2) and 3 (2 - 5) on Cohorts 1 and 3, respectively. Finally, training on Cohort 3 yielded C-indices of 0.75 and 0.63, HR and CI of 24 (10 - 57) and 12 (6 - 21) on Cohorts 1 and 2, respectively. Our results show that r-DepTH descriptor may serve as a comprehensive and a robust MRI-based prognostic marker of disease aggressiveness and survival in solid tumors.

The Association of Modifiable Postresuscitation Management and Annual Case Volume With Survival After Extracorporeal Cardiopulmonary Resuscitation

It is not know if hospital-level extracorporeal cardiopulmonary resuscitation (ECPR) case volume, or postcannulation clinical management associate with survival outcomes.

OBJECTIVES

To describe variation in postresuscitation management practices, and annual hospital-level case volume, for patients who receive ECPR and to determine associations between these management practices and hospital survival.

DESIGN

Observational cohort study using case-mix adjusted survival analysis.

SETTING AND PARTICIPANTS

Adult patients greater than or equal to 18 years old who received ECPR from the Extracorporeal Life Support Organization Registry from 2008 to 2019.

MAIN OUTCOMES AND MEASURES

Generalized estimating equation logistic regression was used to determine factors associated with hospital survival, accounting for clustering by center. Factors analyzed included specific clinical management interventions after starting extracorporeal membrane oxygenation (ECMO) including coronary angiography, mechanical unloading of the left ventricle on ECMO (with additional placement of a peripheral ventricular assist device, intra-aortic balloon pump, or surgical vent), placement of an arterial perfusion catheter distal to the arterial return cannula (to mitigate leg ischemia); potentially modifiable on-ECMO hemodynamics (arterial pulsatility, mean arterial pressure, ECMO flow); plus hospital-level annual case volume for adult ECPR.

RESULTS

Case-mix adjusted patient-level management practices varied widely across individual hospitals. We analyzed 7,488 adults (29% survival); median age 55 (interquartile range, 44-64), 68% of whom were male. Adjusted hospital survival on ECMO was associated with mechanical unloading of the left ventricle (odds ratio [OR], 1.3; 95% CI, 1.08-1.55; p = 0.005), performance of coronary angiography (OR, 1.34; 95% CI, 1.11- 1.61; p = 0.002), and placement of an arterial perfusion catheter distal to the return cannula (OR, 1.39; 95% CI, 1.05-1.84; p = 0.022). Survival varied by 44% across hospitals after case-mix adjustment and was higher at centers that perform more than 12 ECPR cases/yr (OR, 1.23; 95% CI, 1.04-1.45; p = 0.015) versus medium- and low-volume centers.

CONCLUSIONS AND RELEVANCE

Modifiable ECMO management strategies and annual case volume vary across hospitals, appear to be associated with survival and should be the focus of future research to test if these hypothesis-generating associations are causal in nature.

Risk factors and clustering of mortality among older adults in the India Human Development Survey

With wide socioeconomic mortality differential among older adults in India, a constant question of death clustering across high-risk families and communities arises. The present study uses a follow-up survey from India to investigate the socioeconomic, demographic and health predictors of old-age mortality clustering. Data of 16,964 older adults nested within 12,981 households from 2352 communities were used from India Human Development Survey (IHDS) round-I (2005) who were further tracked down in round-II (2012). Bivariate association between the determinants of old-age mortality was investigated using the log-rank test. The multivariate analysis involved estimating the random-intercept Weibull proportional hazard model with three levels-individual (level 1), family (level 2) and community (level 3). We analyzed the sensitivity of multivariate results to unobservable variable and selection biases using the e-value method. The empirical analysis confirms that the risk of mortality is significantly heterogeneous between the families. The health status of older adults and the family's socioeconomic status in the early years emerged as prominent predictors of a longer lifespan. With a strong association between household income and mortality hazard risk, the present study urges early life interventions as those started in late-life might have negligible impact on keeping the older adults alive and healthy.

Drivers and outcomes of variation in surgical versus transcatheter aortic valve replacement in Ontario, Canada: a population-based study

Objectives

To understand the patient and hospital level drivers of the variation in surgical versus trascatheter aortic valve replacement (SAVR vs TAVR) for patients with aortic stenosis (AS) and to explore whether this variation translates into differences in clinical outcomes.

Background

Adoption of TAVR has grown exponentially worldwide. Notwithstanding, a wide variation in TAVR rates has been seen within and between countries and in some jurisdictions AS is still primarily being managed by SAVR.

Methods

We conducted a population-based retrospective cohort study in Ontario, Canada, including individuals who received TAVR or SAVR between 2016 and 2020. We developed iterative hierarchical logistic regression models for the likelihood of receiving TAVR instead of SAVR examining sequentially patient characteristics, hospital factors and year of procedure, calculating the median ORs and variance partition coefficients for each. Using Cox proportional hazards models, we examined the relationship between TAVR/SAVR ratio on all-cause mortality and readmissions.

Results

Annual procedures rates per million population increased from 171 to 201, mainly driven by the expansion of TAVR. TAVR/SAVR ratios differed substantially between hospitals, from 0.21 to 3.27. Neither patient nor hospital factors explained the between-hospital variation in AS treatment. The TAVR/SAVR ratio was significantly associated with clinical outcomes with high ratio hospitals having lower mortality and rehospitalisations.

Conclusions

Despite the expansion of TAVR, dramatic variation exists that is not explained by patient or hospital factors. This variation was associated with differences in clinical outcomes, suggesting that further work is needed in understanding and addressing inequity of access.

Healthy lifestyle behaviours and all-cause and cardiovascular mortality among 0.9 million Chinese adults

Background

Healthy lifestyle behaviours are effective means to reduce the burden of diseases. This study was aimed to fill the knowledge gaps on the distribution, associated factors, and potential health benefits on mortality of four healthy lifestyle behaviours in China.

Methods

During 2015–2019, participants aged 35–75 years from 31 provinces were recruited by the China PEACE Million Persons Project. Four healthy lifestyle behaviours were investigated in our study, including non-smoking, none or moderate alcohol use, sufficient leisure time physical activity (LTPA), and healthy diet.

Results

Among 903,499 participants, 74.1% were non-smokers, 96.0% had none or moderate alcohol use, 23.6% had sufficient LTPA, 11.1% had healthy diet, and only 2.8% had all the four healthy lifestyle behaviours. The adherence varied across seven regions; the highest median of county-level adherence to all the four healthy lifestyle behaviours was in North China (3.3%) while the lowest in the Southwest (0.8%) (p < 0.05). Participants who were female, elder, non-farmers, urban residents, with higher income or education, hypertensive or diabetic, or with a cardiovascular disease (CVD) history were more likely to adhere to all the four healthy lifestyle behaviours (p < 0.001). County-level per capital Gross Domestic Product (GDP) was positively associated with sufficient LTPA (p < 0.05 for both rural and urban areas) and healthy diet (p < 0.01 for urban areas), while negatively associated with none or moderate alcohol use (p < 0.01 for rural areas). Average annual temperature was negatively associated with none or moderate alcohol use (p < 0.05 for rural areas) and healthy diet (p < 0.001 for rural areas). Those adhering to all the four healthy lifestyle behaviours had lower risks of all-cause mortality (HR 0.64 [95% CI: 0.52, 0.79]) and cardiovascular mortality (HR 0.53 [0.37, 0.76]) after a median follow-up of 2.4 years.

Conclusions

Adherence to healthy lifestyle behaviours in China was far from ideal. Targeted health promotion strategies were urgently needed.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12966-021-01234-4.

Hospital-, anaesthetist-, and patient-level variation in peripheral nerve block utilisation for hip fracture surgery: a population-based cross-sectional study

BACKGROUND

Unwarranted variation in anaesthesia practice is associated with adverse outcomes. Despite high-certainty evidence of benefit, a minority of hip fracture surgery patients receive a peripheral nerve block. Our objective was to estimate variation in peripheral nerve block use at the hospital, anaesthetist, and patient levels, while identifying predictors of peripheral nerve block use in hip fracture patients.

METHODS

After protocol registration (https://osf.io/48bvp/), we conducted a population-based cross-sectional study using linked administrative data in Ontario, Canada. We included adults >65 yr of age having emergency hip fracture surgery from April 1, 2012 to March 31, 2018. Logistic mixed models were used to estimate the variation in peripheral nerve block use attributable to hospital-, anaesthetist-, and patient-level factors with use of peripheral nerve block, quantified using the variance partition coefficient and median odds ratio. Predictors of peripheral nerve block use were estimated and temporally validated.

RESULTS

Of 50 950 patients, 9144 (18.5%) received a peripheral nerve block within 1 day of surgery. Patient-level factors accounted for 14% of variation, whereas 42% and 44% were attributable to the hospital and anaesthetist providing care, respectively. The median odds ratio for receiving a peripheral nerve block was 5.73 at the hospital level and 5.97 at the anaesthetist level. No patient factors had large associations with receipt of a peripheral nerve block (odds ratios significant at the 5% level ranged from 0.86 to 1.35).

CONCLUSIONS

Patient factors explain the minimal variation in peripheral nerve block use for hip fracture surgery. Interventions to increase uptake of peripheral nerve blocks for hip fracture patients will likely need to focus on structures and processes at the hospital and anaesthetist levels.

Hospital Variation in Cancer Treatments and Survival OutComes of Advanced Melanoma Patients: Nationwide Quality Assurance in The Netherlands

BACKGROUND

To assure a high quality of care for patients treated in Dutch melanoma centers, hospital variation in treatment patterns and outcomes is evaluated in the Dutch Melanoma Treatment Registry. The aim of this study was to assess center variation in treatments and 2-year survival probabilities of patients diagnosed between 2013 and 2017 in the Netherlands.

METHODS

We selected patients diagnosed between 2013 and 2017 with unresectable IIIC or stage IV melanoma, registered in the Dutch Melanoma Treatment Registry. Centers' performance on 2-year survival was evaluated using Empirical Bayes estimates calculated in a random effects model. Treatment patterns of the centers with the lowest and highest estimates for 2-year survival were compared.

RESULTS

For patients diagnosed between 2014 and 2015, significant center variation in 2-year survival probabilities was observed even after correcting for case-mix and treatment with new systemic therapies. The different use of new systemic therapies partially explained the observed variation. From 2016 onwards, no significant difference in 2-year survival was observed between centers.

CONCLUSION

Our data suggest that between 2014 and 2015, after correcting for patient case-mix, significant variation in 2-year survival probabilities between Dutch melanoma centers existed. The use of new systemic therapies could partially explain this variation. In 2013 and between 2016 and 2017, no significant variation between centers existed.

Utilization and effect of neuromuscular blockade in a randomized trial of high-frequency oscillation

PURPOSE

We evaluated characteristics associated with neuromuscular blockade (NMB) use, center-level variation, and whether NMB mediated excess mortality among patients assigned to high-frequency oscillatory ventilation (HFOV) in the OSCILLATE trial.

MATERIALS AND METHODS

NMB exposure was defined as receipt after randomization; the primary outcome was hospital mortality. Descriptive analyses compared NMB-exposed vs unexposed patients. Multivariable analyses included patients not on baseline NMB. Cox regression evaluated associations of patient- and center-level variables with NMB use. A log-normal frailty model evaluated center effects. Mediation analysis examined the effect of NMB in HFOV-assigned patients.

RESULTS

376/548 patients (39 centers) received post-randomization NMB, of whom 165 received baseline NMB. Patients receiving post-randomization NMB (vs. not) had worse lung mechanics and gas exchange, received more sedation and vasopressors (p < 0.05), and had higher hospital mortality (44% vs. 34%, p = 0.03). Mean airway pressure ≥ 24 cmH₂O, randomization to HFOV, and intensive care unit size ≥ 31 beds were associated with post-randomization NMB. After adjustment, center had a negligible effect on post-randomization NMB (median hazard ratio 1.01, p = 0.047). NMB use did not mediate excess mortality among HFOV-allocated patients (p = 0.80).

CONCLUSIONS

In OSCILLATE, receipt of post-randomization NMB was associated with worse outcomes, but NMB use did not mediate HFOV-associated higher mortality.

Changes in regional variation in mortality over five decades - The contribution of age and socioeconomic population composition

Existing evidence suggests that within-country area variation in mortality has increased in several high-income countries. Little is known about the role of changes in the population composition of areas in these trends. In this study, we look at mortality variation across Finnish municipalities over five decades. We examine trends by sex, age categories and two broad cause of death groups and assess the role of individual-level compositional factors. Analyses rely on individual-level register data on the total Finnish population aged 30 years and over. We estimated two-level Weibull survival-models with individuals nested in areas for 10 periods between 1972 and 2018 to assess municipal-level variation in mortality. Median hazard ratio (MHR) was used as our summary measure and analyses were adjusted for age and socioeconomic characteristics. The results show a clear overall growth in area variation in mortality with MHR increasing from 1.14 (95% CI 1.12-1.15) to 1.28 (CI 1.26-1.30) among men and 1.17 (CI 1.15-1.18) to 1.30 (CI 1.27-1.32) among women. This growth, however, was fully attenuated by adjustment for age. Area differentials were largest and increased most among men at ages 30-49, and particularly for external causes. This increase was largely due to increasing differentiation in the socioeconomic composition of municipalities. In conclusion, our study shows increases in mortality differentials across municipalities that are mostly attributable to increasing differentiation between municipalities in terms of individual compositional factors.

Geographic disparities in heart transplantation persist under the new allocation policy

BACKGROUND

This study evaluated the impact of the 2018 heart allocation policy change on geographic disparities in United States orthotopic heart transplantation (OHT).

METHODS

The United Network for Organ Sharing registry was queried to measure geographic disparity in OHT rates between pre-policy and post-policy change eras. We performed multilevel Poisson regression to measure region-level OHT rates. We derived an allocation priority-adjusted median incidence rate ratio (MIRR) for each policy era, a measure of median change in OHT rates between regions.

RESULTS

5958.78 waitlist person-years were analyzed, comprising 6596 OHT procedures (3890 pre-policy and 2706 post-policy). Median region-level OHT rate was .94 transplants/person-years before and 1.51 transplants/person-years after the policy change (P < .001). The unadjusted OHT MIRR across regions was 1.29 (95% CI 1.00-1.50) pre-policy change and 1.17 (95% CI 1.00-1.43) post-policy change, suggesting that the region-related variance in OHT rates decreased under the new allocation. After adjustment for allocation priority risk factors, the MIRR pre-policy change was 1.13 (95% CI 1.01-1.32) and post-policy change was 1.15 (95% CI 1.00-1.35).

CONCLUSIONS

Geography accounts for ∼10% of the disparity among United States OHT rates. Despite broader heart sharing, the updated allocation policy did not substantially alter the existing geographic disparities among OHT recipients.

Association Between Patient Survival and Clinician Variability in Treatment Rates for Aortic Valve Stenosis

Background Patients with symptomatic severe aortic stenosis (ssAS) have a high mortality risk and compromised quality of life. Surgical/transcatheter aortic valve replacement (AVR) is a Class I recommendation, but it is unclear if this recommendation is uniformly applied. We determined the impact of managing cardiologists on the likelihood of ssAS treatment. Methods and Results Using natural language processing of Optum electronic health records, we identified 26 438 patients with newly diagnosed ssAS (2011-2016). Multilevel, multivariable Fine-Gray competing risk models clustered by cardiologists were used to determine the impact of cardiologists on the likelihood of 1-year AVR treatment. Within 1 year of diagnosis, 35.6% of patients with ssAS received an AVR; however, rates varied widely among managing cardiologists (0%, lowest quartile; 100%, highest quartile [median, 29.6%; 25th-75th percentiles, 13.3%-47.0%]). The odds of receiving AVR varied >2-fold depending on the cardiologist (median odds ratio for AVR, 2.25; 95% CI, 2.14-2.36). Compared with patients with ssAS of cardiologists with the highest treatment rates, those treated by cardiologists with the lowest AVR rates experienced significantly higher 1-year mortality (lowest quartile, adjusted hazard ratio, 1.22, 95% CI, 1.13-1.33). Conclusions Overall AVR rates for ssAS were low, highlighting a potential challenge for ssAS management in the United States. Cardiologist AVR use varied substantially; patients treated by cardiologists with lower AVR rates had higher mortality rates than those treated by cardiologists with higher AVR rates.

Association of Annual Intensive Care Unit Sepsis Caseload With Hospital Mortality From Sepsis in the United Kingdom, 2010-2016

IMPORTANCE

Sepsis is associated with a high burden of inpatient mortality. Treatment in intensive care units (ICUs) that have more experience treating patients with sepsis may be associated with lower mortality.

OBJECTIVE

To assess the association between the volume of patients with sepsis receiving care in an ICU and hospital mortality from sepsis in the UK.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study used data from adult patients with sepsis from 231 UK ICUs between 2010 and 2016. Demographic and clinical data were extracted from the Intensive Care National Audit & Research Centre (ICNARC) Case Mix Programme database. Data were analyzed from January 1, 2010, to December 31, 2016.

EXPOSURES

Annual sepsis case volume in an ICU in the year of a patient's admission.

MAIN OUTCOMES AND MEASURES

Hospital mortality after ICU admission for sepsis assessed using a mixed-effects logistic model in a 3-level hierarchical structure based on the number of individual patients nested in years nested within ICUs.

RESULTS

Among 273 001 patients included in the analysis, the median age was 66 years (interquartile range, 53-76 years), 148 149 (54.3%) were male, and 248 275 (91.0%) were White. The mean ICNARC-2018 illness severity score was 21.0 (95% CI, 20.9-21.0). Septic shock accounted for 19.3% of patient admissions, and 54.3% of patients required mechanical ventilation. The median annual sepsis volume per ICU was 242 cases (interquartile range, 177-334 cases). The study identified a significant association between the volume of sepsis cases in the ICU and mortality from sepsis; in the logistic regression model, hospital mortality was significantly lower among patients admitted to ICUs in the highest quartile of sepsis volume compared with the lowest quartile (odds ratio [OR], 0.89; 95% CI, 0.82-0.96; P = .002). With volume modeled as a restricted cubic spline, treatment in a larger ICU was associated with lower hospital mortality. A lower annual volume threshold of 215 patients above which hospital mortality decreased significantly was found; 38.8% of patients were treated in ICUs below this threshold volume. There was no significant interaction between ICU volume and severity of illness as described by the ICNARC-2018 score (β [SE], -0.00014 [0.00024]; P = .57).

CONCLUSIONS AND RELEVANCE

The findings suggest that patients with sepsis in the UK have higher odds of survival if they are treated in an ICU with a larger sepsis case volume. The benefit of a high sepsis case volume was not associated with the severity of the sepsis episode.

Risk factors associated with bleeding after prophylactic endoscopic variceal ligation in cirrhosis

BACKGROUND

Prophylactic endoscopic variceal band ligation (EVL) is frequently performed in patients with liver cirrhosis. The aim of our study was to identify factors associated with early upper gastrointestinal bleeding (UGIB) in cirrhosis patients after prophylactic EVL.

METHODS

787 nonemergency EVLs performed in 444 patients in two German University medical centers were analyzed retrospectively.

RESULTS

Within 30 days after EVL, 38 UGIBs were observed (4.8 % of all procedures). Bilirubin levels (hazard ratio [HR] 1.5, 95 % confidence interval [CI] 1.2-2.0 for a 2-fold increase) and presence of varices grade III/IV according to Paquet (HR 2.6, 95 %CI 1.3-5.0 compared with absence or smaller sized varices) were independently associated with UGIB following EVL. International normalized ratio (INR) was associated with bleeding events in the univariate analysis but did not reach statistical significance after adjustment for bilirubin and presence of varices grade III/IV (HR 1.2, 95 %CI 0.9-1.6 for an increase by 0.25). There was no statistically significant association between platelet count or fibrinogen levels and UGIB. Substitution of coagulation products did not affect incidence of bleeding after EVL, which also applied to patients with "coagulopathy" (INR > 1.5 and/or platelet count < 50 × 10⁹/L). No association between proton pump inhibitor therapy and post-EVL UGIB was observed.

CONCLUSIONS

EVL is a safe procedure and immediate bleeding complications are rare. Serum bilirubin levels and size of varices, rather than coagulation indices, are associated with UGIB after EVL. Our data do not support the preventive substitution of blood or coagulation products.

A first step towards a global nomogram to predict disease progression for men on active surveillance

BACKGROUND

Signs of disease progression (28%) and conversion to active treatment without evidence of disease progression (13%) are the main reasons for discontinuation of active surveillance (AS) in men with localised prostate cancer (PCa). We aimed to develop a nomogram to predict disease progression in these patients.

METHODS

As a first step in the development of a nomogram, using data from Movembers' GAP3 Consortium (n=14,380), we assessed heterogeneity between centres in terms of risk of disease progression. We started with assessment of baseline hazards for disease progression based on grouping of centres according to follow-up protocols [high: yearly; intermediate: ~2 yearly; and low: at year 1, 4 & 7 (i.e., PRIAS)]. We conducted cause-specific random effect Cox proportional hazards regression to estimate risk of disease progression by centre in each group.

RESULTS

Disease progression rates varied substantially between centres [median hazard ratio (MHR): 2.5]. After adjustment for various clinical factors (age, year of diagnosis, Gleason grade group, number of positive cores and PSA), substantial heterogeneity in disease progression remained between centres.

CONCLUSIONS

When combining worldwide data on AS, we noted unexplained differences of disease progression rate even after adjustment for various clinical factors. This suggests that when developing a global nomogram, local adjustments for differences in risk of disease progression and competing outcomes such as conversion to active treatment need to be considered.

Variation in prophylactic tranexamic acid administration among anesthesiologists and surgeons in orthopedic surgery: a retrospective cohort study

PURPOSE

Tranexamic acid (TXA) reduces red blood cell transfusion in various orthopedic surgeries, yet the degree of practice variation in its use among anesthesiologists and surgeons has not been described. To target future knowledge transfer and implementation strategies, and to better understand determinants of variability in prophylactic TXA use, our primary objective was to evaluate the influence of surgical team members on the variability of prophylactic TXA administration.

METHODS

This was a retrospective cohort study of all adult patients undergoing primary total hip arthroplasty (THA), hip fracture surgery, and spine fusion ± vertebrectomy at two Canadian hospitals between January 2014 and December 2016. We used Canadian Classification of Health Interventions procedure codes within the Discharge Abstract Database which we linked to the Ottawa Data Warehouse. We described the percentage of patients that received TXA by individual surgery, the specifics of TXA dosing, and estimated the effect of anesthesiologists and surgeons on prophylactic TXA using multivariable mixed-effects logistic regression analyses.

RESULTS

In the 3,900 patients studied, TXA was most commonly used in primary THA (85%; n = 1,344/1,582), with lower use in hip fracture (23%; n = 342/1,506) and spine fusion surgery (23%; n = 186/812). The median [interquartile range] total TXA dose was 1,000 [1,000-1,000] mg, given as a bolus in 92% of cases. Anesthesiologists and surgeons added significant variability to the odds of receiving TXA in hip fracture surgery and spine fusion, but not primary THA. Most of the variability in TXA use was attributed to patient and other factors.

CONCLUSION

We confirmed the routine use of TXA in primary THA, while observing lower utilization with more variability in hip fracture and spine fusion surgery. Further study is warranted to understand variations in use and the barriers to TXA implementation in a broader population of orthopedic surgical patients at high risk for transfusion.

Center-level Variation in HLA-incompatible Living Donor Kidney Transplantation Outcomes

BACKGROUND

Desensitization protocols for HLA-incompatible living donor kidney transplantation (ILDKT) vary across centers. The impact of these, as well as other practice variations, on ILDKT outcomes remains unknown.

METHODS

We sought to quantify center-level variation in mortality and graft loss following ILDKT using a 25-center cohort of 1358 ILDKT recipients with linkage to Scientific Registry of Transplant Recipients for accurate outcome ascertainment. We used multilevel Cox regression with shared frailty to determine the variation in post-ILDKT outcomes attributable to between-center differences and to identify any center-level characteristics associated with improved post-ILDKT outcomes.

RESULTS

After adjusting for patient-level characteristics, only 6 centers (24%) had lower mortality and 1 (4%) had higher mortality than average. Similarly, only 5 centers (20%) had higher graft loss and 2 had lower graft loss than average. Only 4.7% of the differences in mortality (P < 0.01) and 4.4% of the differences in graft loss (P < 0.01) were attributable to between-center variation. These translated to a median hazard ratio of 1.36 for mortality and 1.34 of graft loss for similar candidates at different centers. Post-ILDKT outcomes were not associated with the following center-level characteristics: ILDKT volume and transplanting a higher proportion of highly sensitized, prior transplant, preemptive, or minority candidates.

CONCLUSIONS

Unlike most aspects of transplantation in which center-level variation and volume impact outcomes, we did not find substantial evidence for this in ILDKT. Our findings support the continued practice of ILDKT across these diverse centers.

Cross-classified Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) to evaluate hospital performance: the case of hospital differences in patient survival after acute myocardial infarction

OBJECTIVE

To describe a novel strategy, Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) to evaluate hospital performance, by analysing differences in 30-day mortality after a first-ever acute myocardial infarction (AMI) in Sweden.

DESIGN

Cross-classified study.

SETTING

68 Swedish hospitals.

PARTICIPANTS

43 247 patients admitted between 2007 and 2009, with a first-ever AMI.

PRIMARY AND SECONDARY OUTCOME MEASURES

We evaluate hospital performance by analysing differences in 30-day mortality after a first-ever AMI using a cross-classified multilevel analysis. We classified the patients into 10 categories according to a risk score (RS) for 30-day mortality and created 680 strata defined by combining hospital and RS categories.

RESULTS

In the cross-classified multilevel analysis the overall RS adjusted hospital 30-day mortality in Sweden was 4.78% and the between-hospital variation was very small (variance partition coefficient (VPC)=0.70%, area under the curve (AUC)=0.54). The benchmark value was therefore achieved by all hospitals. However, as expected, there were large differences between the RS categories (VPC=34.13%, AUC=0.77) CONCLUSIONS: MAIHDA is a useful tool to evaluate hospital performance. The benefit of this novel approach to adjusting for patient RS is that it allowed one to estimate separate VPCs and AUC statistics to simultaneously evaluate the influence of RS categories and hospital differences on mortality. At the time of our analysis, all hospitals in Sweden were performing homogeneously well. That is, the benchmark target for 30-day mortality was fully achieved and there were not relevant hospital differences. Therefore, possible quality interventions should be universal and oriented to maintain the high hospital quality of care.

Variation in Arterial and Central Venous Catheter Use in Pediatric Intensive Care Units

OBJECTIVE

Describe patient and hospital characteristics associated with Arterial Catheter (AC) or Central Venous Catheter (CVC) use among pediatric intensive care units (ICUs).

DESIGN

Hierarchical mixed effects analyses were used to identify patient and hospital characteristics associated with AC or CVC placement. The ICU adjusted median odds ratios (ICU-AMOR) for the admission ICU, marginal R², and conditional intraclass correlation coefficient were reported.

SETTING

166 PICUs in the Virtual PICU Systems (VPS, LLC) Database.

PATIENTS

682,791 patients with unscheduled admissions to the PICU.

INTERVENTION

None.

MEASURES AND MAIN RESULTS

ACs were placed in (median, [interquartile range]) 8.2% [4.9%-11.3%] of admissions, and CVCs were placed in 14.9% [10.4%-19.3%] of admissions across cohort ICUs. Measured patient characteristics explained about 25% of the variability in AC and CVC placement. Higher Pediatric Index of Mortality 2 (PIM2) illness severity scores were associated with increased odds of placement (Odds Ratio (95th% Confidence Interval)) AC: 1.88 (1.87-1.89) and CVC: 1.82 (1.81-1.83) per 1 unit increase in PIM2 score. Primary diagnoses of cardiovascular, gastrointestinal, hematology/oncology, infectious, renal/genitourinary, rheumatology, and transplant were associated with increased odds of AC or CVC placement compared to a primary respiratory diagnosis. Presence of in-house attendings 24/7 was associated with increased odds of AC placement 1.32 (1.11-1.57). Admission ICU explained 4.9% and 3.5% of the variability in AC or CVC placement, respectively. The ICU-AMOR showed a patient would have a median increase in odds of 55% and 43% for AC or CVC placement, respectively, if the same patient moved from an ICU with lower odds of placement to an ICU with higher odds of placement.

CONCLUSIONS

Variation in AC or CVC use exists among PICUs. The admission ICU was more strongly associated with AC than with CVC placement. Further study is needed to understand unexplained variation in AC and CVC use.

Evidence-practice gaps in P2Y12 inhibitor use after hospitalisation for acute myocardial infarction: findings from a new population-level data linkage in Australia

Background

P2Y₁₂ inhibitor therapy is recommended for 12 months in patients hospitalised for acute myocardial infarction (AMI) unless the bleeding risk is high.

Aims

To describe real‐world use of P2Y₁₂ inhibitor therapy following AMI hospitalisation.

Methods

We used population‐level linked hospital data to identify all patients discharged from a public hospital with a primary diagnosis of AMI between July 2011 and June 2013 in New South Wales and Victoria, Australia. We used dispensing claims to examine dispensing of a P2Y₁₂ inhibitor (clopidogrel, prasugrel or ticagrelor) within 30 days of discharge and multilevel models to identify predictors of post‐discharge dispensing and persistence of therapy to 1 year.

Results

We identified 31 848 patients hospitalised for AMI, of whom 56.8% were dispensed a P2Y₁₂ inhibitor within 30 days of discharge. The proportion of patients with post‐discharge dispensing varied between hospitals (interquartile range: 25.0–56.5%), and significant between‐hospital variation remained after adjusting for patient characteristics. Patient factors associated with the lowest likelihood of post‐discharge dispensing were: having undergone coronary artery bypass grafting (odds ratio (OR): 0.17; 95% confidence intervals (CI): 0.15–0.20); having oral anticoagulants dispensed 180 days before or 30 days after discharge (OR: 0.39, 95% CI: 0.35–0.44); major bleeding (OR: 0.68, 95% CI: 0.61–0.76); or being aged ≥85 years (OR: 0.68, 95% CI: 0.62–0.75). A total of 26.8% of patients who were dispensed a P2Y₁₂ inhibitor post‐discharge discontinued therapy within 1 year.

Conclusion

Post‐hospitalisation use of P2Y₁₂ inhibitor therapy in AMI patients is low and varies substantially by hospital of discharge. Our findings suggest strategies addressing both health system (hospital and physician) and patient factors are needed to close this evidence‐practice gap.

Number of lung resections performed and long-term mortality rates of patients after lung cancer surgery: evidence from an Italian investigation

OBJECTIVES

Although it has been postulated that patients might benefit from the centralization of high-volume specialized centres, conflicting results have been reported on the relationship between the number of lung resections performed and the long-term, all-cause mortality rates among patients who underwent surgery for lung cancer. A population-based observational study was performed to contribute to the ongoing debate.

METHODS

The 2613 patients, all residents of the Lombardy region (Italy), who underwent lung resection for lung cancer from 2012 to 2014 were entered into the cohort and were followed until 2018. The hospitals were classified according to the annual number of pulmonary resections performed. Three categories of lung resection cases were identified: low (≤30), intermediate (31-95) and high (>95). The outcome of interest was all-cause death. A frailty model was used to estimate the death risk associated with the categories of numbers of lung resections performed, taking into account the multilevel structure of the data. A set of sensitivity analyses was performed to account for sources of systematic uncertainty.

RESULTS

The 1-year and 5-year survival rates of cohort members were 90% and 63%. Patients operated on in high-volume centres were on average younger and more often women. Compared to patients operated on in a low-volume centre, the mortality risk exhibited a significant, progressive reduction as the numbers of lung resections performed increased to intermediate (-13%; 95% confidence interval +10% to -31%) and high (-26%; 0% to -45%). Sensitivity analyses revealed that the association was consistent.

CONCLUSIONS

Further evidence that the volume of lung resection cases performed strongly affects the long-term survival of lung cancer patients has been supplied.

Understanding the large heterogeneity in hospital readmissions and mortality for acute myocardial infarction

This study aims to investigate the variation in two acute myocardial infarction (AMI) outcomes across public hospitals in Portugal. In-hospital mortality and 30-day unplanned readmissions were studied using two distinct AMI cohorts of adults discharged from all acute care public hospital centers in Portugal from 2012-2015. Hierarchical generalized linear models were used to assess the association between patient and hospital characteristics and hospital variability in the two outcomes. Our findings indicate that hospitals are not performing homogeneously-the risk of adverse events tends to be consistently larger in some hospitals and consistently lower in other hospitals. While patient characteristics accounted for a larger share of the explained between-hospital variance, hospital characteristics explain an additional 8% and 10% of hospital heterogeneity in the mortality and the readmission cohorts respectively. Admissions to hospitals with low AMI caseloads or located in Alentejo/Algarve and Lisbon had a higher risk of mortality. Discharges from larger-sized hospitals were associated with increased risk of readmissions. Future health policies should incorporate these findings in order to incentivize more consistent health care outcomes across hospitals. Further investigation addressing geographical disparities, hospital caseload and practices is needed to direct actions of improvement to specific hospitals.

Analysis of mortality after hip fracture on patient, hospital, and regional level in Germany

Knowledge about risk factors of mortality after hip fracture might encourage prevention and further improvements in care. This study identified patient risk factors as well as hospital and regional characteristics associated with a decreased risk. Variation of mortality was largest on patient level and modest on hospital and regional level.

INTRODUCTION

Among numerous studies analyzing mortality as worst consequence after hip fracture, the majority focused on patient level and fewer on hospital and regional level. Comprehensive knowledge about contributing factors on all levels might help to reveal relevant inequalities, which would encourage prevention and further improvements in care. This study aimed at investigating variation of mortality after hip fracture on patient, hospital, and regional level in Germany.

METHODS

We performed a retrospective cohort study on hip fracture patients aged 65 and older using statutory health insurance claims data from Jan 2009 through Dec. 2012 and additional information from the Federal Statistical Office Germany. Regions were classified based on two-digit postal code. We applied a multilevel Cox proportional hazard model with random intercepts on hospital and regional level to investigate the risk factors for mortality within 6 and 12 months after hip fracture.

RESULTS

The dataset contained information on 123,119 hip fracture patients in 1014 hospitals in 95 German regions. Within 6/12 months, 20.9%/27.6% of the patients died. On patient level, male sex, increasing age, increased pre-fracture care level, and increasing comorbidity were associated with an increased hazard of mortality. Hospitals with increasing hip fracture volume or with orthogeriatric co-management and regions with increased population density were associated with a decreased hazard. Variation was largest on patient level and rather modest on hospital and regional level.

CONCLUSIONS

The identification of patient-related risk factors enables prognosticating mortality after hip fracture. After adjusting for those, variation seemed to be attributable rather to hospitals than to regions.

Multilevel analysis of hemodialysis-associated infection among end-stage renal disease patients: results of a retrospective cohort study utilizing the insurance claim data of Fukuoka Prefecture, Japan

The presence of comorbid conditions along with heterogeneity in terms of healthcare practices and service delivery could have a significant impact on the patient's outcomes. With a strong interest in social epidemiology to examine the impact of health services and variations on health outcomes, the current study was conducted to analyse the incidence of hemodialysis-associated infection (HAI) as well as its associated factors, and to quantify the extent to which the contextual effects of the care facility and regional variations influence the risk of HAI.A total of 6111 patients with end-stage renal disease who received hemodialysis treatment between 1 October 2015 and 31 March 2016 were identified from the insurance claim database as a population-based, close-cohort retrospective study. Patients were followed for one year from April 1, 2016 to March 31, 2017. A total of 200 HAI cases were observed during the follow-up and 12 patients died within 90 days of the onset of HAI. Increased risks for HAI were associated with moderate (HR 1.73, 95% confidence interval [CI] 1.00-2.98) and severe (HR 1.87, 95% CI 1.11-3.14) comorbid conditions as well as malignancy (HR 1.36, 95% CI 1.00-1.85). Increased risk was also seen among patients who received hemodialysis treatment from clinics (HR 2.49, 95% CI 1.1-5.33). However, these statistics were no longer significant when variations at the level of care facilities were statistically controlled. In univariate analyses, no statistically significant association was observed between 90-day mortality and baseline patients, and the characteristics of the care facility.The results of the multivariate, multilevel analyses indicated that HAI variations were only significant at the care facility level (σ 2.07, 95% CI 1.3-3.2) and were largely explained by the heterogeneity between care facilities. The results of this study highlight the need to look beyond the influence of patient-level characteristics when developing policies that aim at improving the quality of hemodialysis healthcare and service delivery in Japan.

Individual participant data meta-analysis of intervention studies with time-to-event outcomes: A review of the methodology and an applied example

Many randomized trials evaluate an intervention effect on time-to-event outcomes. Individual participant data (IPD) from such trials can be obtained and combined in a so-called IPD meta-analysis (IPD-MA), to summarize the overall intervention effect. We performed a narrative literature review to provide an overview of methods for conducting an IPD-MA of randomized intervention studies with a time-to-event outcome. We focused on identifying good methodological practice for modeling frailty of trial participants across trials, modeling heterogeneity of intervention effects, choosing appropriate association measures, dealing with (trial differences in) censoring and follow-up times, and addressing time-varying intervention effects and effect modification (interactions).We discuss how to achieve this using parametric and semi-parametric methods, and describe how to implement these in a one-stage or two-stage IPD-MA framework. We recommend exploring heterogeneity of the effect(s) through interaction and non-linear effects. Random effects should be applied to account for residual heterogeneity of the intervention effect. We provide further recommendations, many of which specific to IPD-MA of time-to-event data from randomized trials examining an intervention effect.We illustrate several key methods in a real IPD-MA, where IPD of 1225 participants from 5 randomized clinical trials were combined to compare the effects of Carbamazepine and Valproate on the incidence of epileptic seizures.

Long-Term Survival After Heart Transplantation at Centers Stratified by Short-Term Performance

BACKGROUND

Center differences in short-term survival after heart transplant (HT) are known. We sought to compare long-term graft survival (freedom from death or retransplantation) at currently active United States HT centers stratified by performance for short-term survival.

METHODS

We used the Organ Procurement and Transplant Network database to identify subjects ≥18 years old who received primary HT during 2000 to 2014 at US centers active during 2013 and 2014. Follow-up was available until March 2016. Center case-mix was assessed by computing expected 90-day mortality and short-term performance by 90-day standardized mortality ratio (SMR; observed/expected mortality). Centers were stratified by case-mix as transplanting low-, intermediate-, and high-risk patients and by short-term performance as SMR quintiles. Center-level differences in long-term graft survival were assessed by risk-adjusted, mixed-effects Weibull survival models with center as a random effect.

RESULTS

We analyzed 25 467 HT recipients at 96 centers. Those receiving HT at centers with superior (lower) 90-day SMR had longer graft survival (P for trend <0.001). Survival difference among SMR groups remained significant in 90-day conditional survivors (P for trend <0.001). There was significant center-level variation in risk-adjusted graft survival censored at 5 years (P<0.001) and with all follow-up (P<0.001). Adjusting for 90-day SMR was associated with 62% reduction in center variation in 5-year graft survival and 56% reduction in center variation in overall graft survival.

CONCLUSIONS

HT recipients at centers with superior short-term outcomes have longer graft survival on long-term follow-up. Allocating resources to improve patient care processes and transplant expertise at high-SMR centers may improve short-term and overall survival after HT.

Causal variance decompositions for institutional comparisons in healthcare

There is increasing interest in comparing institutions delivering healthcare in terms of disease-specific quality indicators (QIs) that capture processes or outcomes showing variations in the care provided. Such comparisons can be framed in terms of causal models, where adjusting for patient case-mix is analogous to controlling for confounding, and exposure is being treated in a given hospital, for instance. Our goal here is to help identify good QIs rather than comparing hospitals in terms of an already chosen QI, and so we focus on the presence and magnitude of overall variation in care between the hospitals rather than the pairwise differences between any two hospitals. We consider how the observed variation in care received at patient level can be decomposed into that causally explained by the hospital performance adjusting for the case-mix, the case-mix itself, and residual variation. For this purpose, we derive a three-way variance decomposition, with particular attention to its causal interpretation in terms of potential outcome variables. We propose model-based estimators for the decomposition, accommodating different link functions and either fixed or random effect models. We evaluate their performance in a simulation study and demonstrate their use in a real data application.

Geographic Disparity in Deceased Donor Liver Transplant Rates Following Share 35

BACKGROUND

The Organ Procurement and Transplantation Network implemented Share 35 on June 18, 2013, to broaden deceased donor liver sharing within regional boundaries. We investigated whether increased sharing under Share 35 impacted geographic disparity in deceased donor liver transplantation (DDLT) across donation service areas (DSAs).

METHODS

Using Scientific Registry of Transplant Recipients June 2009 to June 2017, we identified 86 083 adult liver transplant candidates and retrospectively estimated Model for End-Stage Liver Disease (MELD)-adjusted DDLT rates using nested multilevel Poisson regression with random intercepts for DSA and transplant program. From the variance in DDLT rates across 49 DSAs and 102 programs, we derived the DSA-level median incidence rate ratio (MIRR) of DDLT rates. MIRR is a robust metric of heterogeneity across each hierarchical level; larger MIRR indicates greater disparity.

RESULTS

MIRR was 2.18 pre-Share 35 and 2.16 post-Share 35. Thus, 2 candidates with the same MELD in 2 different DSAs were expected to have a 2.2-fold difference in DDLT rate driven by geography alone. After accounting for program-level heterogeneity, MIRR was attenuated to 2.10 pre-Share 35 and 1.96 post-Share 35. For candidates with MELD 15-34, MIRR decreased from 2.51 pre- to 2.27 post-Share 35, and for candidates with MELD 35-40, MIRR increased from 1.46 pre- to 1.51 post-Share 35, independent of program-level heterogeneity in DDLT. DSA-level heterogeneity in DDLT rates was greater than program-level heterogeneity pre- and post-Share 35.

CONCLUSIONS

Geographic disparity substantially impacted DDLT rates before and after Share 35, independent of program-level heterogeneity and particularly for candidates with MELD 35-40. Despite broader sharing, geography remains a major determinant of access to DDLT.