A tutorial on frailty models

Self-Rated Health in Middle Age and Risk of Hospitalizations and Death: Recurrent Event Analysis of the ARIC Study

BACKGROUND

Self-rated health is a simple measure that may identify individuals who are at a higher risk for hospitalization or death.

OBJECTIVE

To quantify the association between a single measure of self-rated health and future risk of recurrent hospitalizations or death.

PARTICIPANTS

Atherosclerosis Risk in Communities (ARIC) study, a community-based prospective cohort study of middle-aged men and women with follow-up beginning from 1987 to 1989.

MAIN MEASURES

We quantified the associations between initial self-rated health with risk of recurrent hospitalizations and of death using a recurrent events survival model that allowed for dependency between the rates of hospitalization and hazards of death, adjusted for demographic and clinical factors.

KEY RESULTS

Of the 14,937 ARIC cohort individuals with available self-rated health and covariate information, 34% of individuals reported "excellent" health, 47% "good," 16% "fair," and 3% "poor" at study baseline. After a median follow-up of 27.7 years, 1955 (39%), 3569 (51%), 1626 (67%), and 402 (83%) individuals with "excellent," "good," "fair," and "poor" health, respectively, had died. After adjusting for demographic factors and medical history, a less favorable self-rated health status was associated with increased rates of hospitalization and death. As compared to those reporting "excellent" health, adults with "good," "fair," and "poor" health had 1.22 (1.07 to 1.40), 2.01 (1.63 to 2.47), and 3.13 (2.39 to 4.09) times the rate of hospitalizations, respectively. The hazards of death also increased with worsening categories of self-rated health, with "good," "fair," and "poor" health individuals experiencing 1.30 (1.12 to 1.51), 2.15 (1.71 to 2.69), and 3.40 (2.54 to 4.56) times the hazard of death compared to "excellent," respectively.

CONCLUSIONS

Even after adjusting for demographic and clinical factors, having a less favorable response on a single measure of self-rated health taken in middle age is a potent marker of future hospitalizations and death.

Illicit Fentanyl Use and Hepatitis C Virus Seroconversion Among People Who Inject Drugs in Tijuana and San Diego: Results From a Binational Cohort Study

BACKGROUND

Illicitly manufactured fentanyl (IMF) increases overdose mortality, but its role in infectious disease transmission is unknown. We examined whether IMF use predicts hepatitis C virus (HCV) and human immunodeficiency virus (HIV) incidence among a cohort of people who inject drugs (PWID) in San Diego, California and Tijuana, Mexico.

METHODS

PWID were recruited during 2020-2022, undergoing semi-annual interviewer-administered surveys and HIV and HCV serological rapid tests through 2024. Cox regression was conducted to examine predictors of seroconversion considering self-reported IMF use as a 6-month lagged, time-dependent covariate.

RESULTS

Of 398 PWID at baseline, 67% resided in San Diego, 70% were male, median age was 43 years, 42% reported receptive needle sharing, and 25% reported using IMF. HCV incidence was 14.26 per 100 person-years (95% confidence interval [CI]: 11.49-17.02), and HIV incidence was 1.29 (95% CI: .49-2.10). IMF was associated with HCV seroconversion, with a univariable hazard ratio (HR) of 1.64 (95% CI: 1.09-2.40), and multivariable HR of 1.57 (95% CI: 1.03-2.40). The direction of the relationship with HIV was similar, albeit not significant (HR 2.39; 95% CI: .66-8.64).

CONCLUSIONS

We document a novel association between IMF and HCV seroconversion among PWID in Tijuana-San Diego. Few HIV seroconversions (n = 10) precluded our ability to assess if a similar relationship held for HIV. IMF's short half-life may destabilize PWID-increasing the need for repeat dosing and sharing smoking materials and syringes. New preventive care approaches may reduce HCV transmission in the fentanyl era.

The Rise of the Direct Anterior Approach: Trends, Learning Curves, and Patient Characteristics of 63,182 Primary Total Hip Arthroplasties in the Dutch Arthroplasty Register (LROI)

BACKGROUND

The use of the direct anterior approach (DAA) in total hip arthroplasty (THA) has steadily increased in the Netherlands since 2007. The aim of this study was to outline how the DAA has been implemented in the Netherlands. Moreover, we investigated the learning curve of the DAA at a hospital level, and explored patient characteristics of the DAA compared with other approaches and during the learning phase after implementing the DAA.

METHODS

In this population-based cohort study, we included all primary THAs between 2007 and 2020 (n = 342,473) from the Dutch Arthroplasty Register. For hospitals implementing the DAA (n > 20), patients were categorized in 4 experience groups using the date of surgery: 1 to 50, 51 to 100, 101 to 150, or > 150. Subsequently, data from different hospitals were pooled and survival rates were calculated using Kaplan-Meier survival analyses. Adjusted revision rates were calculated using mixed Cox proportional hazard models (frailty).

RESULTS

The use of the DAA gradually rose from 0.2% in 2007 to 41% of all primary THAs in 2020. A total of 64 (56%) hospitals implemented the DAA. However, not all hospitals continued using this approach. After implementation, the 5-year survival rate for the first 50 procedures was significantly lower (96% confidence interval [CI] 95.8 to 97.2) compared to >150 procedures (98% CI 97.7 to 98.1). Multivariable Cox hazard analyses demonstrated a higher risk of revision during the first 50 procedures compared with >150 procedures (hazard ratio 1.6, CI 1.3 to 2.0).

CONCLUSIONS

The use of DAA for primary THA significantly increased. For hospitals implementing DAA, a considerable learning curve with increased revision risk was seen.

Follow-up blood cultures do not reduce mortality in hospitalized patients with Gram-negative bloodstream infection: a retrospective population-wide cohort study

OBJECTIVES

The utility of follow-up blood cultures (FUBCs) in patients with Gram-negative bloodstream infection (GN-BSI) is controversial. Observational studies have suggested significant mortality benefit but may be limited by single-centre designs, immortal time bias, and residual confounding. We examined the impact of FUBCs on mortality in patients with GN-BSI in a retrospective population-wide cohort study in Ontario, Canada.

METHODS

Adult patients with GN-BSI hospitalized between April 2017 and December 2021 were included. Primary outcome was all-cause mortality within 30 days. FUBC was treated as a time-varying exposure. Secondary outcomes were 90-day mortality, length of stay, and number of days alive and out of hospital at 30 and 90 days.

RESULTS

Thirty-four thousand one hundred patients were included; 8807 (25.8%) patients received FUBC, of which 966 (11.0%) were positive. Median proportion of patients receiving FUBC was 18.8% (interquartile range, 10.0-29.7%; range, 0-66.1%) across 101 hospitals; this correlated with positivity and contamination rate. Eight hundred ninety (10.1%) patients in the FUBC group and 2263 (8.9%) patients in the no FUBC group died within 30 days. In the fully adjusted model, there was no association between FUBC and mortality (hazard ratio, 0.97; 95% CI, 0.90-1.04). Patients with FUBC had significantly longer length of stay (median, 11 vs. 7 days; adjusted risk ratio, 1.18; 95% CI, 1.16-1.21) and fewer number of days alive and out of hospital at 30 and 90 days.

DISCUSSION

FUBC collection in patients with GN-BSI varies widely across hospitals and may be associated with prolonged hospitalization without clear survival benefit. Residual confounding may be present given the observational design. Clear benefit should be demonstrated in a randomized trial before widespread adoption of routine FUBC.

Institution-level Patterns of Care for Early-stage Oropharynx Cancers in the United States

OBJECTIVES

The adoption of transoral robotic surgery and shifting epidemiology in oropharyngeal squamous cell cancer have stimulated debate over upfront and adjuvant treatment. Institutional variation in practice patterns can be obscured in patient-level analyses. We aimed to characterize institutional patterns of care as well as identify potential associations between patterns of care and survival.

METHODS

This was a retrospective cohort study of patients identified from 2004-2015 in the National Cancer Database. We analyzed 42,803 cases of oropharyngeal squamous cell cancer Stage cT1-2N0-2bM0 (AJCC 7th edition) treated with curative intent surgery and/or radiotherapy. We defined facility-4-year periods to account for changing institutional practice patterns. The 42,803 patients were treated within 2578 facility-4-year periods. We assessed institutional practice patterns, including the ratio of upfront surgery to definitive radiotherapy, case volumes, use of adjuvant therapies (radiotherapy or chemoradiotherapy), and margin positivity rates. Survival associations with institutional practice patterns were estimated with Cox regression.

RESULTS

The ratio of upfront surgery to definitive radiotherapy ranged from 80-to-1 to 1-to-23. The institution-level median rate of adjuvant radiotherapy was 69% (IQR 50%-100%), adjuvant chemoradiotherapy was 44% (IQR 0%-67%), and margin-positive resection was 33% (IQR 0%-50%). On patient-level MVA, worse overall survival was not significantly associated with institutional case volume, adjuvant radiotherapy, or adjuvant chemoradiotherapy utilization.

CONCLUSIONS

High rates of multimodal therapy and positive margins underscore the importance of multidisciplinary care and highlight variable patterns of care across institutions. Further work is warranted to explore indicators of high-quality care and to optimize adjuvant therapy in the HPV era.

Recurrence of new-onset post-operative AF after cardiac surgery: detected by implantable loop recorders: A systematic review and Meta-analysis

BACKGROUND

Atrial fibrillation (AF) is one of the most common complications after cardiac surgery. New-onset post-operative AF may signal an elevated risk of AF and associated outcomes in long-term follow-up. We aimed to estimate the rate of AF recurrence as detected by an implantable loop recorder (ILR) in patients experiencing post-operative AF within 30 days after cardiac surgery.

METHODS

We searched MEDLINE, Embase and Cochrane CENTRAL to April 2023 for studies of adults who did not have known AF, experienced new-onset AF within 30 days of cardiac surgery and received an ILR. We pooled individual participant data on timing of AF recurrence using a random-effects model with a frailty model applied to a Cox proportional hazard analysis.

RESULTS

From 8671 citations, 8 single-centre prospective cohort studies met eligibility criteria. Data were available from 185 participants in 7 studies, with a median follow-up of 1.7 (IQR: 1.3-2.8) years. All included studies were at a low risk of bias. Pooled AF recurrence rates following 30 post-operative days were 17.8% (95% CI 11.9%-23.2%) at 3 months, 24.4% (17.7%-30.6%) at 6 months, 30.1% (22.8%-36.7%) at 12 months and 35.3% (27.6%-42.2%) at 18 months.

CONCLUSIONS

In patients who experience new-onset post-operative AF after cardiac surgery, AF recurrence lasting at least 30 s occurs in approximately 1 in 3 in the first year after surgery. The optimal frequency and modality to use for monitoring for AF recurrence in this population remain uncertain.

Modelling the spatial risk pattern of dementia in Denmark using residential location data: A registry-based national cohort

Dementia is a major global public health concern that is increasingly leading to morbidity and mortality among older adults. While studies have focused on the risk factors and care provision, there is currently limited knowledge about the spatial risk pattern of the disease. In this study, we employ Bayesian spatial modelling with a stochastic partial differential equation (SPDE) approach to model the spatial risk using complete residential history data from the Danish population and health registers. The study cohort consisted of 1.6 million people aged 65 years and above from 2005 to 2018. The results of the spatial risk map indicate high-risk areas in Copenhagen, southern Jutland and Funen. Individual socioeconomic factors and population density reduce the intensity of high-risk patterns across Denmark. The findings of this study call for the critical examination of the contribution of place of residence in the susceptibility of the global ageing population to dementia.

ADHD medication adherence reduces risk of committing minor offenses in adolescents

BACKGROUND

This study aimed to investigate the association between adolescents' adherence to attention-deficit/hyperactivity disorder (ADHD) medication and their risk of committing minor offenses.

METHODS

Using two Dutch databases, Statistics Netherlands (CBS) and the Foundation for Pharmaceutical Statistics (SFK), we aimed to investigate the association between adherence to ADHD medication and registered minor offenses between 2005 and 2019 of 18,234 adolescents (12-18 years). We used Cox regression analyses to compare the rate of committing minor offenses of adolescents during periods of high ADHD medication adherence compared to periods of low adherence (i.e., periods with or without sufficient amounts of dispensed medication). We additionally tested associations with adherence to selective serotonin reuptake inhibitors (SSRIs) as control medication and analyzed potential reverse causation.

RESULTS

High ADHD medication adherence was associated with a reduced risk of committing a minor offense of between 33% and 38% compared to low adherence periods of ≥3 months (hazard ratio [HR] 0.67, confidence interval [CI] 0.64-0.71) or ≥6 months (HR 0.62, CI 0.59-0.65). The reduction in risk can likely be attributed to ADHD medication, given the absence of effects of SSRIs and no reverse causation. The reduction rate remained between 16% and 55% per sex, stimulant versus non-stimulant medication, different offense categories and further sensitivity analyses.

CONCLUSIONS

Among adolescents using ADHD medication, rates of criminality were lower during periods of high medication adherence, suggesting that adherence to ADHD medication may contribute to prevention of minor offenses in adolescents.

Impact of Prosthesis-Patient Mismatch After Surgical Aortic Valve Replacement: Systematic Review and Meta-Analysis of Reconstructed Time-to-Event Data of 122 989 Patients With 592 952 Patient-Years

BACKGROUND

It remains controversial whether prosthesis-patient mismatch (PPM) impacts long-term outcomes after surgical aortic valve replacement. We aimed to evaluate the association of PPM with mortality, rehospitalizations, and aortic valve reinterventions.

METHODS AND RESULTS

We performed a systematic review with meta-analysis of reconstructed time-to-event data of studies published by March 2023 (according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses). Sixty-five studies met our eligibility criteria and included 122 989 patients (any PPM: 68 332 patients, 55.6%). At 25 years of follow-up, the survival rates were 11.8% and 20.6% in patients with and without any PPM, respectively (hazard ratio [HR], 1.16 [95% CI, 1.13-1.18], P<0.001). At 20 years of follow-up, the survival rates were 19.5%, 12.1%, and 8.8% in patients with no, moderate, and severe PPM, respectively (moderate versus no PPM: HR, 1.09 [95% CI, 1.06-1.11], P<0.001; severe versus no PPM: HR, 1.29 [95% CI, 1.24-1.35], P<0.001). PPM was associated with higher risk of cardiac death, heart failure-related hospitalizations, and aortic valve reinterventions over time (P<0.001). Statistically significant associations between PPM and worse survival were observed regardless of valve type (bioprosthetic versus mechanical valves), contemporary PPM definitions unadjusted and adjusted for body mass index, and PPM quantification method (in vitro, in vivo, Doppler echocardiography). Our meta-regression analysis revealed that populations with more women tend to have higher HRs for all-cause death associated with PPM.

CONCLUSIONS

The results of the present study suggest that any degree of PPM is associated with poorer long-term outcomes following surgical aortic valve replacement and provide support for implementation of preventive strategies to avoid PPM after surgical aortic valve replacement.

Centre-level variation in the survival of patients receiving haemodialysis in India: findings from a nationwide private haemodialysis network

Background

There are no large studies examining survival in patients receiving haemodialysis in India or considering centre-level effects on survival. We measured survival variation between dialysis centres across India and evaluated the extent to which differences are explained by measured centre characteristics.

Methods

This is a multilevel analysis of patient survival in centres of the NephroPlus dialysis network consisting of 193 centres across India. Patients receiving haemodialysis at a centre for ≥90 days between April 2014 and June 2019 were included, with analyses restricted to centres with ≥10 such patients. The primary outcome was all-cause mortality, measured from 90 days after joining a centre. Proportional hazards models with shared frailty were used to model centre- and patient-level effects on survival.

Findings

Amongst 23,601 patients (median age 53 years; 29% female), the unadjusted centre-specific 180-day Kaplan-Meier survival estimates ranged between 55% (95% confidence interval [CI] 38-80%) and 100%, with a median of 88% (interquartile interval 83%-92%). After accounting for multilevel factors, estimated 180-day survival ranged between 83% (73-89%) and 97% (95-98%), with 90% 180-day survival in the average centre. The mortality rate in patients attending rural centres was 32% (Hazard Ratio 1.32; 95% CI 1.06-1.65) higher than those at urban centres in adjusted analyses. Multiple patient characteristics were associated with mortality.

Interpretation

This is the first national benchmark for survival amongst dialysis patients in India. Centre- and patient-level characteristics are associated with survival but there remains unexplained variation between centres. As India continues to widen dialysis access, ongoing quality improvement programs will be an important part of ensuring that patients experience the best possible outcomes at the point of care.

Funding

This project received no external funding.

The built-in selection bias of hazard ratios formalized using structural causal models

It is known that the hazard ratio lacks a useful causal interpretation. Even for data from a randomized controlled trial, the hazard ratio suffers from so-called built-in selection bias as, over time, the individuals at risk among the exposed and unexposed are no longer exchangeable. In this paper, we formalize how the expectation of the observed hazard ratio evolves and deviates from the causal effect of interest in the presence of heterogeneity of the hazard rate of unexposed individuals (frailty) and heterogeneity in effect (individual modification). For the case of effect heterogeneity, we define the causal hazard ratio. We show that the expected observed hazard ratio equals the ratio of expectations of the latent variables (frailty and modifier) conditionally on survival in the world with and without exposure, respectively. Examples with gamma, inverse Gaussian and compound Poisson distributed frailty and categorical (harming, beneficial or neutral) distributed effect modifiers are presented for illustration. This set of examples shows that an observed hazard ratio with a particular value can arise for all values of the causal hazard ratio. Therefore, the hazard ratio cannot be used as a measure of the causal effect without making untestable assumptions, stressing the importance of using more appropriate estimands, such as contrasts of the survival probabilities.

Low revision rate throughout the adoption of the direct superior approach in primary total hip arthroplasty: an analysis based on 1551 total hip arthroplasties from the Dutch Arthroplasty Register

BACKGROUND

Recently, surgeons introduced a minimally invasive modification on the classic posterolateral approach (PLA) in total hip arthroplasty (THA): the direct superior approach (DSA). We investigated the association between surgeon's experience and the risk of early revision of the DSA in primary THA, using data from the Dutch Arthroplasty Register (LROI).

METHODS

We retrieved all primary THAs performed using the DSA in 4 hospitals between 2016 and 2022 (n = 1551). Procedures were sorted in 5 groups using the date of operation and number of previous procedures per surgeon: 1-25; 26-50; 51-75; 76-100; >100. Subsequently, data from different surgeons were pooled together and the risk of revision was calculated via a multilevel time-to-event analysis.

RESULTS

The overall revision rate was 1.5% after a mean follow-up of 2 years. Patients from the 1-25 group had comparable risks of revision compared to patients in the >100 group (hazard ratio [HR] 1.0 [CI, 0.3-3.2]). The risk for patients in groups 26-50, 51-75 and 75-100 was also not statistically different from the >100 group (resp. HR 1.5 [CI, 0.5-5.0], 1.8 [CI, 0.5-6.4] and 0.5 [CI, 0.1-4.0]). Main reasons of revision were dislocation (0.5%) and infection (0.4%).

CONCLUSIONS

We did not identify an association between the surgeon's experience and the early risk of revision for the DSA in primary THA in the Netherlands. The DSA seems safe in the early adoption phases with a low risk of revision due to dislocation and revision for all other causes.

Trends in epidemiology and risk factors of opportunistic infections in kidney transplant recipients between 2004 and 2017

BACKGROUND

While opportunistic infections are a frequent and challenging problem in kidney transplant recipients, their long-term epidemiology remains hardly known.

METHODS

Opportunistic infections were recorded in 1144 recipients transplanted in our center between 2004 and 2015. Incidence rates and baseline risk factors were determined using joint frailty models.

RESULTS

After a median follow-up of 5.6 years, 544 opportunistic infections occurred in 373/1144 (33%) patients, dominated by viral infections (396/544, 72%), especially cytomegalovirus (CMV) syndromes and diseases (213/544, 39%). One-third of the infected patients experienced at least two opportunistic infections. The incidence of opportunistic infections was 10 times higher during the first year post-transplantation than after that (34.7 infections for 100 patient-years vs 3.64). Opportunistic infections associated with the age of the donor (P = .032), the age of the recipient (P = .049), the CMV serostatus (P < 10-6), a higher class II HLA mismatch (P = .032) and an induction treatment including rabbit anti-thymocyte globulins (P = .026). Repeated opportunistic infections associated with each other (P < 10-6) and with renal death (P < 10-6).

CONCLUSION

Opportunistic infections occur with a two-period incidence pattern and many susceptible patients suffer from repeated episodes. This knowledge may help tailor new prevention and follow-up strategies to reduce the burden of opportunistic infections and their impact on transplantation outcome.

Examining the role of firearm involvement in repeat intimate partner violence assaults

BACKGROUND

Intimate partner violence (IPV) remains a pervasive and complex issue with significant social and public health implications. The nexus of firearms and intimate partner violence (IPV) is an especially dangerous one. However, little is known about how firearm involvement can influence the risk of repeat IPV assaults.

METHODS

We use data from 346 male perpetrated IPV incidents reported to the Detroit Police Department between December 2016 and April 2017 to examine the role of firearm involvement in IPV recidivism during a 5 and half year follow up period. Employing a conditional gap-time frailty model that accommodates heterogeneity among individuals through a frailty term, we analyze time to multiple IPV assaults that occur over the follow up period. We identify various pathways through which firearms impact the likelihood of subsequent IPV incidents, including intimidation, threats, and use of firearms, while controlling for observable perpetrator characteristics to understand the explicit roles of firearms.

RESULTS

Firearm involvement at the index assault was not associated with IPV recidivism. However, involvement of firearms in past IPV assaults significantly increased the risk of subsequent physical IPV. The discrepancy is likely arising from a high degree of censoring among individuals who were armed with a firearm during the index assault.

CONCLUSION

Our research reveals a nuanced relationship between firearm involvement and IPV recidivism, shedding light on the multifaceted dynamics at play. By elucidating the intricate dynamics at the intersection of firearms and intimate partner violence, our study underscores the need for targeted policy interventions and preventative measures aimed at reducing IPV recidivism.

Mortality-related risk factors of inpatients with diabetes and COVID-19: A multicenter retrospective study in Belgium

BACKGROUND AND AIMS

We describe mortality-related risk factors of inpatients with diabetes and coronavirus disease 2019 (COVID-19) in Belgium.

METHODS

We conducted a multicenter retrospective study from March to May, 2020, in 8 Belgian centers. Data on admission of patients with diabetes and COVID-19 were collected. Survivors were compared to non-survivors to identify prognostic risk factors for in-hospital death using multivariate analysis in both the total population and in the subgroup of patients admitted in the intensive care unit (ICU).

RESULTS

The study included 375 patients. The mortality rate was 26.4% (99/375) in the total population and 40% (27/67) in the ICU. Multivariate analysis identified older age (HR 1.05 [CI 1.03-1.07], P<0.0001) and male gender (HR 2.01 [1.31-3.07], P=0.0013) as main independent risk factors for in-hospital death in the total population. Metformin (HR 0.51 [0.34-0.78], P=0.0018) and renin-angiotensin-aldosterone system blockers (HR 0.56 [0.36-0.86], P=0.0088) use before admission were independent protective factors. In the ICU, chronic kidney disease (CKD) was identified as an independent risk factor for death (HR 4.96 [2.14-11.5], P<0.001).

CONCLUSION

In-hospital mortality due to the first wave of COVID-19 pandemic in Belgium was high in patients with diabetes. We found that advanced age and male gender were independent risk factors for in-hospital death. We also showed that metformin use before admission was associated with a significant reduction of COVID-19-related in-hospital mortality. Finally, we showed that CKD is a COVID-19-related mortality risk factor in patients with diabetes admitted in the ICU.

Modeling the multi-state natural history of rare diseases with heterogeneous individual patient data: A simulation study

Multi-state survival models are used to represent the natural history of a disease, forming the basis of a health technology assessment comparing a novel treatment to current practice. Constructing such models for rare diseases is problematic, since evidence sources are typically much sparser and more heterogeneous. This simulation study investigated different one-stage and two-stage approaches to meta-analyzing individual patient data (IPD) in a multi-state survival setting when the number and size of studies being meta-analyzed are small. The objective was to assess methods of different complexity to see when they are accurate, when they are inaccurate and when they struggle to converge due to the sparsity of data. Biologically plausible multi-state IPD were simulated from study- and transition-specific hazard functions. One-stage frailty and two-stage stratified models were estimated, and compared to a base case model that did not account for study heterogeneity. Convergence and the bias/coverage of population-level transition probabilities to, and lengths of stay in, each state were used to assess model performance. A real-world application to Duchenne Muscular Dystrophy, a neuromuscular rare disease, was conducted, and a software demonstration is provided. Models not accounting for study heterogeneity were consistently out-performed by two-stage models. Frailty models struggled to converge, particularly in scenarios of low heterogeneity, and predictions from models that did converge were also subject to bias. Stratified models may be better suited to meta-analyzing disparate sources of IPD in rare disease natural history/economic modeling, as they converge more consistently and produce less biased predictions of lengths of stay.

Treatment in certified cancer centers is related to better survival in patients with colon and rectal cancer: evidence from a large German cohort study

BACKGROUND

Certified cancer centers aim to ensure high-quality care by establishing structural and procedural standards according to evidence-based guidelines. Despite the high clinical and health policy relevance, evidence from a nation-wide study for the effectiveness of care for colorectal cancer in certified centers vs. other hospitals in Germany is still missing.

METHODS

In a retrospective cohort study covering the years 2009-2017, we analyzed patient data using demographic information, diagnoses, and treatments from a nationwide statutory health insurance enriched with information on certification. We investigated whether patients with incident colon or rectal cancer did benefit from primary therapy in a certified cancer center. We used relative survival analysis taking into account mortality data of the German population and adjustment for patient and hospital characteristics via Cox regression with shared frailty for patients in hospitals with and without certification.

RESULTS

The cohorts for colon and rectal cancer consisted of 109,518 and 51,417 patients, respectively, treated in a total of 1052 hospitals. 37.2% of patients with colon and 42.9% of patients with rectal cancer were treated in a certified center. Patient age, sex, comorbidities, secondary malignoma, and distant metastases were similar across groups (certified/non-certified) for both colon and rectal cancer. Relative survival analysis showed significantly better survival of patients treated in a certified center, with 68.3% (non-certified hospitals 65.8%) 5-year survival for treatment of colon cancer in certified (p < 0.001) and 65.0% (58.8%) 5-year survival in case of rectal cancer (p < 0.001), respectively. Cox regression with adjustment for relevant covariates yielded a lower hazard of death for patients treated in certified centers for both colon (HR = 0.92, 95% CI = 0.89-0.95) and rectal cancer (HR = 0.92, 95% CI = 0.88-0.95). The results remained robust in a series of sensitivity analyses.

CONCLUSIONS

This large cohort study yields new important evidence that patients with colorectal cancer have a better chance of survival if treated in a certified cancer center. Certification thus provides one powerful means to improve the quality of care for colorectal cancer. To decrease the burden of disease, more patients should thus receive cancer care in a certified center.

A Bayesian joint model for multivariate longitudinal and time-to-event data with application to ALL maintenance studies

The most common type of cancer diagnosed among children is the Acute Lymphocytic Leukemia (ALL). A study was conducted by Tata Translational Cancer Research Center (TTCRC) Kolkata, in which 236 children (diagnosed as ALL patients) were treated for the first two years (approximately) with two standard drugs (6MP and MTx) and were then followed nearly for the next 3 years. The goal is to identify the longitudinal biomarkers that are associated with time-to-relapse, and also to assess the effectiveness of the drugs. We develop a Bayesian joint model in which a linear mixed model is used to jointly model three biomarkers (i.e. white blood cell count, neutrophil count, and platelet count) and a semi-parametric proportional hazards model is used to model the time-to-relapse. Our proposed joint model can assess the effects of different covariates on the progression of the biomarkers, and the effects of the biomarkers (and the covariates) on time-to-relapse. In addition, the proposed joint model can impute the missing longitudinal biomarkers efficiently. Our analysis shows that the white blood cell (WBC) count is not associated with time-to-relapse, but the neutrophil count and the platelet count are significantly associated with it. We also infer that a lower dose of 6MP and a higher dose of MTx jointly result in a lower relapse probability in the follow-up period. Interestingly, we find that relapse probability is the lowest for the patients classified into the "high-risk" group at presentation. The effectiveness of the proposed joint model is assessed through the extensive simulation studies.

Patient Readmission for Ischemic Stroke: Risk Factors and Impact on Mortality

Patient readmission for ischemic stroke significantly strains the healthcare and medical insurance systems. Current understanding of the risk factors associated with these readmissions, as well as their subsequent impact on mortality within China, remains insufficient. This is particularly evident in the context of comprehensive, contemporary population studies. This 4-year retrospective cohort study included 125 397 hospital admissions for ischemic stroke from 838 hospitals located in 22 regions (13 urban and 9 rural) of a major city in western China, between January 1, 2015 and December 31, 2018. The Chi-square tests were used in univariate analysis. Accounting for intra-subject correlations of patients' readmissions, accelerated failure time (AFT) shared frailty models were used to examine readmission events and pure AFT models for mortality. Risk factors for patient readmission after ischemic stroke include frequent admission history, male gender, employee's insurance, advanced age, residence in urban areas, index hospitalization in low-level hospitals, extended length of stay (LOS) during index hospitalization, specific comorbidities and subtypes of ischemic stroke. Furthermore, our findings indicated that an additional admission for ischemic stroke increased patient mortality by 16.4% (P < .001). Stroke readmission contributed to an increased risk of hospital mortality. Policymakers can establish more effective and targeted policies to reduce readmissions for stroke by controlling these risk factors.

Association between neuromuscular blocking agent use and outcomes among out-of-hospital cardiac arrest patients treated with extracorporeal cardiopulmonary resuscitation and target temperature management: A secondary analysis of the SAVE-J II study

Background

Neuromuscular blocking agents are used to control shivering in cardiac arrest patients treated with target temperature management. However, their effect on outcomes in patients treated with extracorporeal cardiopulmonary resuscitation is unclear.

Methods

This study was a secondary analysis of the SAVE-J II study, a retrospective multicenter study of 2175 out-of-hospital cardiac arrest patients treated with extracorporeal cardiopulmonary resuscitation in Japan. We classified patients into those who received neuromuscular blocking agents and those who did not and compared in-hospital mortality and incidence rates of favorable neurological outcome and in-hospital pneumonia between the groups using multivariable regression models and stabilized inverse probability weighting with propensity scores.

Results

Six hundred sixty patients from the SAVE-J II registry were analyzed. Neuromuscular blocking agents were used in 451 patients (68.3%). After adjusting for potential confounders, neuromuscular blocking agents use was not significantly associated with in-hospital mortality (aHR 0.88; 95% CI, 0.67-1.14), favorable neurological outcome (aOR 0.85; 95% CI, 0.60-1.11), or pneumonia (aOR 1.52; 95% CI, 0.85-2.71). The results for in-hospital mortality (aHR 0.89; 95% CI, 0.64-1.25), favorable neurological outcome (aOR 0.94; 95% CI, 0.59-1.48) and pneumonia (aOR 1.59; 95% CI, 0.74-3.41) were similar after weighting was performed.

Conclusions

Although data on the rationale for using neuromuscular blocking agents were unavailable, their use was not significantly associated with outcomes in out-of-hospital cardiac arrest patients treated with extracorporeal cardiopulmonary resuscitation and targeted temperature management. Neuromuscular blocking agents should be used based on individual clinical indications.

Bayesian inference for a principal stratum estimand on recurrent events truncated by death

Recurrent events are often important endpoints in randomized clinical trials. For example, the number of recurrent disease-related hospitalizations may be considered as a clinically meaningful endpoint in cardiovascular studies. In some settings, the recurrent event process may be terminated by an event such as death, which makes it more challenging to define and estimate a causal treatment effect on recurrent event endpoints. In this paper, we focus on the principal stratum estimand, where the treatment effect of interest on recurrent events is defined among subjects who would be alive regardless of the assigned treatment. For the estimation of the principal stratum effect in randomized clinical trials, we propose a Bayesian approach based on a joint model of the recurrent event and death processes with a frailty term accounting for within-subject correlation. We also present Bayesian posterior predictive check procedures for assessing the model fit. The proposed approaches are demonstrated in the randomized Phase III chronic heart failure trial PARAGON-HF (NCT01920711).

Critical Windows of Maternal Exposure to Biothermal Stress and Birth Weight for Gestational Age in Western Australia

BACKGROUND

There is limited and inconsistent evidence on the risk of ambient temperature on small for gestational age (SGA) and there are no known related studies for large for gestational age (LGA). In addition, previous studies used temperature rather than a biothermal metric.

OBJECTIVES

Our aim was to examine the associations and critical susceptible windows of maternal exposure to a biothermal metric [Universal Thermal Climate Index (UTCI)] and the hazards of SGA and LGA.

METHODS

We linked 385,337 singleton term births between 1 January 2000 and 31 December 2015 in Western Australia to daily spatiotemporal UTCI. Distributed lag nonlinear models with Cox regression and multiple models were used to investigate maternal exposure to UTCI from 12 weeks preconception to birth and the adjusted hazard ratios (HRs) of SGA and LGA.

RESULTS

Relative to the median exposure, weekly and monthly specific exposures showed potential critical windows of susceptibility for SGA and LGA at extreme exposures, especially during late gestational periods. Monthly exposure showed strong positive associations from the 6th to the 10th gestational months with the highest hazard of 13% for SGA (HR = 1.13 ; 95% CI: 1.10, 1.14) and 7% for LGA (HR = 1.07 ; 95% CI: 1.03, 1.11) at the 10th month for the 1st UTCI centile. Entire pregnancy exposures showed the strongest hazards of 11% for SGA (HR = 1.11 ; 95% CI: 1.04, 1.18) and 3% for LGA (HR = 1.03 ; 95% CI: 0.95, 1.11) at the 99th UTCI centile. By trimesters, the highest hazards were found during the second and first trimesters for SGA and LGA, respectively, at the 99th UTCI centile. Based on estimated interaction effects, male births, mothers who were non-Caucasian, smokers, ≥ 35 years of age, and rural residents were most vulnerable.

CONCLUSIONS

Both weekly and monthly specific extreme biothermal stress exposures showed potential critical susceptible windows of SGA and LGA during late gestational periods with disproportionate sociodemographic vulnerabilities. https://doi.org/10.1289/EHP12660.

Frailty and outcomes in heart failure patients from high-, middle-, and low-income countries

BACKGROUND AND AIMS

There is little information on the incremental prognostic importance of frailty beyond conventional prognostic variables in heart failure (HF) populations from different country income levels.

METHODS

A total of 3429 adults with HF (age 61 ± 14 years, 33% women) from 27 high-, middle- and low-income countries were prospectively studied. Baseline frailty was evaluated by the Fried index, incorporating handgrip strength, gait speed, physical activity, unintended weight loss, and self-reported exhaustion. Mean left ventricular ejection fraction was 39 ± 14% and 26% had New York Heart Association Class III/IV symptoms. Participants were followed for a median (25th to 75th percentile) of 3.1 (2.0-4.3) years. Cox proportional hazard models for death and HF hospitalization adjusted for country income level; age; sex; education; HF aetiology; left ventricular ejection fraction; diabetes; tobacco and alcohol use; New York Heart Association functional class; HF medication use; blood pressure; and haemoglobin, sodium, and creatinine concentrations were performed. The incremental discriminatory value of frailty over and above the MAGGIC risk score was evaluated by the area under the receiver-operating characteristic curve.

RESULTS

At baseline, 18% of participants were robust, 61% pre-frail, and 21% frail. During follow-up, 565 (16%) participants died and 471 (14%) were hospitalized for HF. Respective adjusted hazard ratios (95% confidence interval) for death among the pre-frail and frail were 1.59 (1.12-2.26) and 2.92 (1.99-4.27). Respective adjusted hazard ratios (95% confidence interval) for HF hospitalization were 1.32 (0.93-1.87) and 1.97 (1.33-2.91). Findings were consistent among different country income levels and by most subgroups. Adding frailty to the MAGGIC risk score improved the discrimination of future death and HF hospitalization.

CONCLUSIONS

Frailty confers substantial incremental prognostic information to prognostic variables for predicting death and HF hospitalization. The relationship between frailty and these outcomes is consistent across countries at all income levels.

On-Support and Postweaning Mortality in Postcardiotomy Extracorporeal Membrane Oxygenation

BACKGROUND

Postcardiotomy venoarterial extracorporeal membrane oxygenation (VA ECMO) is characterized by discrepancies between weaning and survival-to-discharge rates. This study analyzes the differences between postcardiotomy VA ECMO patients who survived, died on ECMO, or died after ECMO weaning. Causes of death and variables associated with mortality at different time points are investigated.

METHODS

The retrospective, multicenter, observational Postcardiotomy Extracorporeal Life Support Study (PELS) includes adults requiring postcardiotomy VA ECMO between 2000 and 2020. Variables associated with on-ECMO mortality and postweaning mortality were modeled using mixed Cox proportional hazards, including random effects for center and year.

RESULTS

In 2058 patients (men, 59%; median age, 65 years; interquartile range [IQR], 55-72 years), weaning rate was 62.7%, and survival to discharge was 39.6%. Patients who died (n = 1244) included 754 on-ECMO deaths (36.6%; median support time, 79 hours; IQR, 24-192 hours), and 476 postweaning deaths (23.1%; median support time, 146 hours; IQR, 96-235.5 hours). Multiorgan (n = 431 of 1158 [37.2%]) and persistent heart failure (n = 423 of 1158 [36.5%]) were the main causes of death, followed by bleeding (n = 56 of 754 [7.4%]) for on-ECMO mortality and sepsis (n = 61 of 401 [15.4%]) for postweaning mortality. On-ECMO death was associated with emergency surgery, preoperative cardiac arrest, cardiogenic shock, right ventricular failure, cardiopulmonary bypass time, and ECMO implantation timing. Diabetes, postoperative bleeding, cardiac arrest, bowel ischemia, acute kidney injury, and septic shock were associated with postweaning mortality.

CONCLUSIONS

A discrepancy exists between weaning and discharge rate in postcardiotomy ECMO. Deaths occurred during ECMO support in 36.6% of patients, mostly associated with unstable preoperative hemodynamics. Another 23.1% of patients died after weaning in association with severe complications. This underscores the importance of postweaning care for postcardiotomy VA ECMO patients.

Mortality in KPC-producing Klebsiella pneumoniae bloodstream infections: a changing landscape

OBJECTIVES

To assess the impact of carbapenem resistance on mortality in Klebsiella pneumoniae bloodstream infection (BSI) in the era of novel β-lactam/β-lactamase inhibitor combinations.

MATERIAL AND METHODS

Retrospective study of patients with K. pneumoniae BSI between January and August 2020 in 16 centres (CARBANEW study within the MULTI-SITA project).

RESULTS

Overall, 426 patients were included: 107/426 (25%) had carbapenem-resistant K. pneumoniae (CR-Kp) BSI and 319/426 (75%) had carbapenem-susceptible K. pneumoniae (CS-Kp) BSI. Crude cumulative 30 day mortality was 33.8% and 20.7% in patients with, respectively, CR-Kp BSI and CS-Kp BSI (P = 0.027). Carbapenemase production or carbapenemase-encoding genes were detected in 84/98 tested CR-Kp isolates (85.7%), mainly KPC (78/84; 92.9%). Ceftazidime/avibactam was the most frequently used appropriate therapy for CR-Kp BSI (80/107; 74.7%). In multivariable analyses, variables showing an unfavourable association with mortality after correction for multiple testing were age-adjusted Charlson comorbidity index (HR 1.20; 95% CI 1.10-1.31, P < 0.001) and Pitt score (HR 1.33; 95% CI 1.15-1.55, P < 0.001), but not carbapenem resistance (HR 1.28, 95% CI 0.74-2.22, P = 0.410). In a propensity score-matched analysis, there was no difference in mortality between patients appropriately treated with ceftazidime/avibactam for CR-Kp BSI and patients appropriately treated with other agents (mainly meropenem monotherapy or piperacillin/tazobactam monotherapy) for CS-Kp BSI (HR 1.07; 95% CI 0.50-2.29, P = 0.866).

CONCLUSIONS

Our results suggest that the increased mortality in CR-Kp BSI compared with CS-Kp BSI is not (or no longer) dependent on the type of therapy in areas where ceftazidime/avibactam-susceptible KPC-producing isolates are the most prevalent type of CR-Kp.

Surgical and Patient-Reported Outcomes of 694 Two-Stage Prepectoral versus Subpectoral Breast Reconstructions

BACKGROUND

Opinion regarding the optimal plane for prosthetic device placement in breast reconstruction patients has evolved. The purpose of this study was to assess the differences in complication rates and patient satisfaction between patients who underwent prepectoral and subpectoral implant-based breast reconstruction (IBR).

METHODS

The authors conducted a retrospective cohort study of patients who underwent two-stage IBR at their institution from 2018 to 2019. Surgical and patient-reported outcomes were compared between patients who received a prepectoral versus a subpectoral tissue expander.

RESULTS

A total of 694 reconstructions in 481 patients were identified (83% prepectoral, 17% subpectoral). The mean body mass index was higher in the prepectoral group (27 versus 25 kg/m 2 , P = 0.001), whereas postoperative radiotherapy was more common in the subpectoral group (26% versus 14%, P = 0.001). The overall complication rate was very similar, with 29.3% in the prepectoral and 28.9% in the subpectoral group ( P = 0.887). Rates of individual complications were also similar between the two groups. A multiple-frailty model showed that device location was not associated with overall complications, infection, major complications, or device explantation. Mean scores for Satisfaction with the Breast, Psychosocial Well-Being, and Sexual Well-Being were similar between the two groups. Median time to permanent implant exchange was significantly longer in the subpectoral group (200 versus 150 days, P < 0.001).

CONCLUSION

Prepectoral breast reconstruction results in similar surgical outcomes and patient satisfaction compared with subpectoral IBR.

CLINICAL QUESTION/LEVEL OF EVIDENCE

Therapeutic, III.

Initial Cancer Treatment in Certified Versus Non-Certified Hospitals

BACKGROUND

According to the National Cancer Plan in Germany, all cancer patients should receive high-quality care in accordance with evidence-based treatment guidelines. Certification programs were established for this purpose but have not yet been comprehensively evaluated.

METHODS

In the WiZen project, which was supported by the Innovation Fund (supported project number 01VSF17020), controlled cohort studies were performed to investigate whether initial treatment in hospitals with or without a certificate from the German Cancer Society was associated with a difference in overall survival (primary endpoint) in patients with cancer of the colon, rectum, lung, pancreas, breast, cervix, prostate, endometrium, and ovary, head and neck cancer, and neuro-oncological tumors. The studies were based on nationwide data from adult insurees of the AOK statutory health insurance carrier for the years 2009-2017.

RESULTS

The majority of patients with all entities except breast cancer received their initial treatment in non-certified hospitals. Initial treatment in a certified hospital was found to be beneficial in terms of overall survival for all cancer entities, even after extensive adjustment for patient- and hospital-related confounders. The hazard ratio (HR) ranged from 0.97 (95% CI: [0.94; 1.00]) for lung cancer to 0.77 [0.74; 0.81] for breast cancer, corresponding to an absolute risk reduction (ARR) for overall survival of 0.62 months for lung cancer to 4.61 months for cervical cancer.

CONCLUSION

The WiZen study shows for the entities studied that initial cancer treatment in a certified center is associated with lower mortality. Despite the recommendations of the National Cancer Plan, however, more than 40% of all cancer patients still receive their initial treatment in a non-certified hospital. The preferential provision of initial care in certified hospitals would be likely to improve overall survival. Although the study design does not permit any conclusion with regard to causality, the findings seem robust considering that a control group was used, confounders were taken into account, and the study population was of large size.

Impact of Obesity on Outcomes of Prepectoral vs Subpectoral Implant-Based Breast Reconstruction

BACKGROUND

The impact of obesity on outcomes of prepectoral vs subpectoral implant-based reconstruction (IBR) is not well-established.

OBJECTIVES

The goal of this study was to assess the surgical and patient-reported outcomes of prepectoral vs subpectoral IBR. The authors hypothesized that obese patients would have similar outcomes regardless of device plane.

METHODS

We conducted a retrospective review of obese patients who underwent 2-stage IBR from January 2017 to December 2019. The primary endpoint was the occurrence of any breast-related complication; the secondary endpoint was device explantation.

RESULTS

The authors identified a total of 284 reconstructions (184 prepectoral, 100 subpectoral) in 209 patients. Subpectoral reconstruction demonstrated higher rates of overall complications (50% vs 37%, P = .047) and device explantation (25% vs 12.5%, P = .008) than prepectoral reconstruction. In multivariable regression, subpectoral reconstruction was associated with higher risk of infection (hazard ratio [HR], 1.65; P = .022) and device explantation (HR, 1.97; P = .034). Subgroup analyses demonstrated significantly higher rates of complications and explantation in the subpectoral group in those with a body mass index (BMI) ≥ 35 and BMI ≥40. The authors found no significant differences in mean scores for satisfaction with the breast (41.57 ± 13.19 vs 45.50 ± 11.91, P = .469), psychosocial well-being (39.43 ± 11.23 vs 39.30 ± 12.49, P = .915), and sexual well-being (17.17 ± 7.83 vs 17.0 ± 9.03, P = .931) between subpectoral and prepectoral reconstruction.

CONCLUSIONS

Prepectoral reconstruction was associated with significantly decreased overall complications, infections, and device explantation in obese patients compared with subpectoral reconstruction. Prepectoral reconstruction provides superior outcomes to subpectoral reconstruction with comparable patient-reported outcomes.

LEVEL OF EVIDENCE: 4

Quantifying individual-level heterogeneity in infectiousness and susceptibility through household studies

The spread of SARS-CoV-2, like that of many other pathogens, is governed by heterogeneity. "Superspreading," or "over-dispersion," is an important factor in transmission, yet it is hard to quantify. Estimates from contact tracing data are prone to potential biases due to the increased likelihood of detecting large clusters of cases, and may reflect variation in contact behavior more than biological heterogeneity. In contrast, the average number of secondary infections per contact is routinely estimated from household surveys, and these studies can minimize biases by testing all members of a household. However, the models used to analyze household transmission data typically assume that infectiousness and susceptibility are the same for all individuals or vary only with predetermined traits such as age. Here we develop and apply a combined forward simulation and inference method to quantify the degree of inter-individual variation in both infectiousness and susceptibility from observations of the distribution of infections in household surveys. First, analyzing simulated data, we show our method can reliably ascertain the presence, type, and amount of these heterogeneities given data from a sufficiently large sample of households. We then analyze a collection of household studies of COVID-19 from diverse settings around the world, and find strong evidence for large heterogeneity in both the infectiousness and susceptibility of individuals. Our results also provide a framework to improve the design of studies to evaluate household interventions in the presence of realistic heterogeneity between individuals.

Characterization of direct and/or indirect genetic associations for multiple traits in longitudinal studies of disease progression

When quantitative longitudinal traits are risk factors for disease progression and subject to random biological variation, joint model analysis of time-to-event and longitudinal traits can effectively identify direct and/or indirect genetic association of single nucleotide polymorphisms (SNPs) with time-to-event. We present a joint model that integrates: (1) a multivariate linear mixed model describing trajectories of multiple longitudinal traits as a function of time, SNP effects, and subject-specific random effects and (2) a frailty Cox survival model that depends on SNPs, longitudinal trajectory effects, and subject-specific frailty accounting for dependence among multiple time-to-event traits. Motivated by complex genetic architecture of type 1 diabetes complications (T1DC) observed in the Diabetes Control and Complications Trial (DCCT), we implement a 2-stage approach to inference with bootstrap joint covariance estimation and develop a hypothesis testing procedure to classify direct and/or indirect SNP association with each time-to-event trait. By realistic simulation study, we show that joint modeling of 2 time-to-T1DC (retinopathy and nephropathy) and 2 longitudinal risk factors (HbA1c and systolic blood pressure) reduces estimation bias in genetic effects and improves classification accuracy of direct and/or indirect SNP associations, compared to methods that ignore within-subject risk factor variability and dependence among longitudinal and time-to-event traits. Through DCCT data analysis, we demonstrate feasibility for candidate SNP modeling and quantify effects of sample size and Winner's curse bias on classification for 2 SNPs identified as having indirect associations with time-to-T1DC traits. Joint analysis of multiple longitudinal and multiple time-to-event traits provides insight into complex traits architecture.

Developing prediction models to estimate the risk of two survival outcomes both occurring: A comparison of techniques

INTRODUCTION

This study considers the prediction of the time until two survival outcomes have both occurred. We compared a variety of analytical methods motivated by a typical clinical problem of multimorbidity prognosis.

METHODS

We considered five methods: product (multiply marginal risks), dual-outcome (directly model the time until both events occur), multistate models (msm), and a range of copula and frailty models. We assessed calibration and discrimination under a variety of simulated data scenarios, varying outcome prevalence, and the amount of residual correlation. The simulation focused on model misspecification and statistical power. Using data from the Clinical Practice Research Datalink, we compared model performance when predicting the risk of cardiovascular disease and type 2 diabetes both occurring.

RESULTS

Discrimination was similar for all methods. The product method was poorly calibrated in the presence of residual correlation. The msm and dual-outcome models were the most robust to model misspecification but suffered a drop in performance at small sample sizes due to overfitting, which the copula and frailty model were less susceptible to. The copula and frailty model's performance were highly dependent on the underlying data structure. In the clinical example, the product method was poorly calibrated when adjusting for 8 major cardiovascular risk factors.

DISCUSSION

We recommend the dual-outcome method for predicting the risk of two survival outcomes both occurring. It was the most robust to model misspecification, although was also the most prone to overfitting. The clinical example motivates the use of the methods considered in this study.

Considering Questions Before Methods in Dementia Research With Competing Events and Causal Goals

Studying causal exposure effects on dementia is challenging when death is a competing event. Researchers often interpret death as a potential source of bias, although bias cannot be defined or assessed if the causal question is not explicitly specified. Here we discuss 2 possible notions of a causal effect on dementia risk: the "controlled direct effect" and the "total effect." We provide definitions and discuss the "censoring" assumptions needed for identification in either case and their link to familiar statistical methods. We illustrate concepts in a hypothetical randomized trial on smoking cessation in late midlife, and emulate such a trial using observational data from the Rotterdam Study, the Netherlands, 1990-2015. We estimated a total effect of smoking cessation (compared with continued smoking) on 20-year dementia risk of 2.1 (95% confidence interval: -0.1, 4.2) percentage points and a controlled direct effect of smoking cessation on 20-year dementia risk had death been prevented of -2.7 (95% confidence interval: -6.1, 0.8) percentage points. Our study highlights how analyses corresponding to different causal questions can have different results, here with point estimates on opposite sides of the null. Having a clear causal question in view of the competing event and transparent and explicit assumptions are essential to interpreting results and potential bias.

How hazardous are hazard ratios? An empirical investigation of individual patient data from 27 large randomized clinical trials

The use of hazard ratios as the standard treatment effect estimators for randomized trials with time-to-event outcomes has been the subject of repeated criticisms in recent years, e.g., for its non-collapsibility or with respect to (causal) interpretation. Another important issue is the built-in selection bias, which arises when the treatment is effective and when there are unobserved or not included prognostic factors that influence time-to-event. In these cases, the hazard ratio has even been termed "hazardous" because it is estimated from groups that increasingly differ in their (unobserved or omitted) baseline characteristics, yielding biased treatment estimates. We therefore adapt the Landmarking approach to assess the effect of ignoring a gradually increasing proportion of early events on the estimated hazard ratio. We propose an extension called "Dynamic Landmarking". This approach is based on successive deletion of observations, refitting Cox models and balance checking of omitted but observed prognostic factors, to obtain a visualization that can indicate built-in selection bias. In a small proof-of-concept simulation, we show that our approach is valid under the given assumptions. We further use "Dynamic Landmarking" to assess the suspected selection bias in the individual patient data sets of 27 large randomized clinical trials (RCTs). Surprisingly, we find no empirical evidence of selection bias in these RCTs and thus conclude that the supposed bias of the hazard ratio is of little practical relevance in most cases. This is mainly due to treatment effects in RCTs being small and the patient populations being homogeneous, e.g., due to inclusion and exclusion criteria.

Patient and Management Variables Associated With Survival After Postcardiotomy Extracorporeal Membrane Oxygenation in Adults: The PELS-1 Multicenter Cohort Study

Background Extracorporeal membrane oxygenation (ECMO) has been increasingly used for postcardiotomy cardiogenic shock, but without a concomitant reduction in observed in-hospital mortality. Long-term outcomes are unknown. This study describes patients' characteristics, in-hospital outcome, and 10-year survival after postcardiotomy ECMO. Variables associated with in-hospital and postdischarge mortality are investigated and reported. Methods and Results The retrospective international multicenter observational PELS-1 (Postcardiotomy Extracorporeal Life Support) study includes data on adults requiring ECMO for postcardiotomy cardiogenic shock between 2000 and 2020 from 34 centers. Variables associated with mortality were estimated preoperatively, intraoperatively, during ECMO, and after the occurrence of any complications, and then analyzed at different time points during a patient's clinical course, through mixed Cox proportional hazards models containing fixed and random effects. Follow-up was established by institutional chart review or contacting patients. This analysis included 2058 patients (59% were men; median [interquartile range] age, 65.0 [55.0-72.0] years). In-hospital mortality was 60.5%. Independent variables associated with in-hospital mortality were age (hazard ratio [HR], 1.02 [95% CI, 1.01-1.02]) and preoperative cardiac arrest (HR, 1.41 [95% CI, 1.15-1.73]). In the subgroup of hospital survivors, the overall 1-, 2-, 5-, and 10-year survival rates were 89.5% (95% CI, 87.0%-92.0%), 85.4% (95% CI, 82.5%-88.3%), 76.4% (95% CI, 72.5%-80.5%), and 65.9% (95% CI, 60.3%-72.0%), respectively. Variables associated with postdischarge mortality included older age, atrial fibrillation, emergency surgery, type of surgery, postoperative acute kidney injury, and postoperative septic shock. Conclusions In adults, in-hospital mortality after postcardiotomy ECMO remains high; however, two-thirds of those who are discharged from hospital survive up to 10 years. Patient selection, intraoperative decisions, and ECMO management remain key variables associated with survival in this cohort. Registration URL: https://www.clinicaltrials.gov; Unique identifier: NCT03857217.

Time-varying effects are common in genetic control of gestational duration

Preterm birth is a major burden to neonatal health worldwide, determined in part by genetics. Recently, studies discovered several genes associated with this trait or its continuous equivalent-gestational duration. However, their effect timing, and thus clinical importance, is still unclear. Here, we use genotyping data of 31 000 births from the Norwegian Mother, Father and Child cohort (MoBa) to investigate different models of the genetic pregnancy 'clock'. We conduct genome-wide association studies using gestational duration or preterm birth, replicating known maternal associations and finding one new fetal variant. We illustrate how the interpretation of these results is complicated by the loss of power when dichotomizing. Using flexible survival models, we resolve this complexity and find that many of the known loci have time-varying effects, often stronger early in pregnancy. The overall polygenic control of birth timing appears to be shared in the term and preterm, but not very preterm, periods and exploratory results suggest involvement of the major histocompatibility complex genes in the latter. These findings show that the known gestational duration loci are clinically relevant and should help design further experimental studies.

Delineating COVID-19 subgroups using routine clinical data identifies distinct in-hospital outcomes

The COVID-19 pandemic has been a great challenge to healthcare systems worldwide. It highlighted the need for robust predictive models which can be readily deployed to uncover heterogeneities in disease course, aid decision-making and prioritise treatment. We adapted an unsupervised data-driven model-SuStaIn, to be utilised for short-term infectious disease like COVID-19, based on 11 commonly recorded clinical measures. We used 1344 patients from the National COVID-19 Chest Imaging Database (NCCID), hospitalised for RT-PCR confirmed COVID-19 disease, splitting them equally into a training and an independent validation cohort. We discovered three COVID-19 subtypes (General Haemodynamic, Renal and Immunological) and introduced disease severity stages, both of which were predictive of distinct risks of in-hospital mortality or escalation of treatment, when analysed using Cox Proportional Hazards models. A low-risk Normal-appearing subtype was also discovered. The model and our full pipeline are available online and can be adapted for future outbreaks of COVID-19 or other infectious disease.

Short- and Long-term Outcomes After Laparoscopic Emergency Resection of Left-Sided Obstructive Colon Cancer: A Nationwide Propensity Score-Matched Analysis

BACKGROUND

The role of laparoscopy for emergency resection of left-sided obstructive colon cancer remains unclear, especially regarding impact on survival.

OBJECTIVE

This study aimed to determine short- and long-term outcomes after laparoscopic versus open emergency resection of left-sided obstructive colon cancer.

DESIGN

This observational cohort study compared patients who underwent laparoscopic emergency resection to those who underwent open emergency resection between 2009 and 2016 by using 1:3 propensity-score matching. Matching variables included sex, age, BMI, ASA score, previous abdominal surgery, tumor location, cT4, cM1, multivisceral resection, small-bowel distention on CT, and subtotal colectomy.

SETTING

This was a nationwide, population-based study.

PATIENTS

Of 2002 eligible patients with left-sided obstructive colon cancer, 158 patients who underwent laparoscopic emergency resection were matched with 474 patients who underwent open emergency resection.

INTERVENTIONS

The intervention was laparoscopic versus open emergency resection.

MAIN OUTCOME MEASURES

The main outcome measures were 90-day mortality, 90-day complications, permanent stoma, disease recurrence, overall survival, and disease-free survival.

RESULTS

Intentional laparoscopy resulted in significantly fewer 90-day complications (26.6% vs 38.4%; conditional OR, 0.59; 95% CI, 0.39-0.87) and similar 90-day mortality. Laparoscopy resulted in better 3-year overall survival (81.0% vs 69.4%; HR, 0.54; 95% CI, 0.37-0.79) and disease-free survival (68.3% vs 52.3%; HR, 0.64; 95% CI, 0.47-0.87). Multivariable regression analyses of the unmatched 2002 patients confirmed an independent association of laparoscopy with fewer 90-day complications and better 3-year survival.

LIMITATIONS

Selection bias was the limitation that cannot be completely ruled out because of the retrospective nature of this study.

CONCLUSIONS

This population-based study with propensity score-matched analysis suggests that intentional laparoscopic emergency resection might improve outcomes in patients with left-sided obstructive colon cancer compared to open emergency resection. Management of those patients in the emergency setting requires proper selection for intentional laparoscopic resection if relevant expertise is available, thereby considering other alternatives to avoid open emergency resection (ie, decompressing stoma). See Video Abstract at http://links.lww.com/DCR/B972 .

RESULTADOS A CORTO Y LARGO PLAZO DESPUS DE LA RESECCIN LAPAROSCPICA DE EMERGENCIA EN CNCER DE COLON IZQUIERDO OBSTRUCTIVO UN ANLISIS EMPAREJADO POR PUNTAJE DE PROPENSIN A NIVEL NACIONAL

ANTECEDENTES:El papel de la laparoscopia en la resección de emergencia en cáncer de colon izquierdo obstructivo sigue sin estar claro, especialmente con respecto al impacto en la supervivencia.OBJETIVO:El objetivo de este estudio fue determinar los resultados a corto y largo plazo después de la resección de emergencia laparoscópica versus abierta en cáncer de colon izquierdo obstructivo.DISEÑO:Estudio observacional de cohortes comparó pacientes que se sometieron a resección de laparoscópica de emergencia versus resección abierta de emergencia entre 2009 y 2016, mediante el uso de emparejamineto por puntaje de propensión 1: 3. Las variables emparejadas incluyeron sexo, edad, IMC, puntaje ASA, cirugía abdominal previa, ubicación del tumor, cT4, cM1, resección multivisceral, distensión del intestino delgado en la TAC y colectomía subtotal.ENTORNO CLINICO:A nivel nacional, basado en la población.PACIENTES:De 2002 pacientes elegibles con cáncer de colon izquierdo obstructivo, 158 pacientes con resección laparoscópica s de emergencia e emparejaron con 474 pacientes con resección abierta de emergencia.INTERVENCIONES:Resección laparoscópica de emergencia versus abierta.PRINCIPALES MEDIDAS DE RESULTADO:Las medidas primarias fueron la mortalidad a 90 días, complicaciones a 90 días, estoma permanente, recurrencia de la enfermedad, supervivencia general y supervivencia libre de enfermedad.RESULTADOS:La laparoscopia intencional dió como resultado significativamente menos complicaciones a los 90 días (26,6 % vs 38,4 %, cOR 0,59, IC del 95 %: 0,39-0,87) y una mortalidad similar a los 90 días. La laparoscopia resultó en una mejor supervivencia general a los 3 años (81,0 % vs 69,4 %, HR 0,54, IC del 95 % 0,37-0,79) y supervivencia libre de enfermedad (68,3 % vs 52,3 %, HR 0,64, IC del 95 % 0,47-0,87). Los análisis de regresión multivariable de los 2002 pacientes no emparejados confirmaron una asociación independiente de la laparoscopia con menos complicaciones a los 90 días y una mejor supervivencia a los 3 años.LIMITACIONES:El sesgo de selección no se puede descartar por completo debido a la naturaleza retrospectiva de este estudio.CONCLUSIONES:Estudio poblacional con análisis emparejado por puntaje de propensión sugiere que la resección laparoscópica de emergencia intencional podría mejorar los resultados a corto y largo plazo en pacientes con cáncer de colon izquierdo obstructivo en comparación con resección abierta de emergencia, lo que justifica la confirmación en estudios futuros. El manejo de esos pacientes en el entorno de emergencia requiere una selección adecuada para la resección laparoscópica intencional si se dispone de experiencia relevante, considerando así otras alternativas para evitar la resección abierta de emergencia (es decir, ostomia descompresiva). Consulte Video Resumen en http://links.lww.com/DCR/B972 . (Traducción- Dr. Francisco M. Abarca-Rendon & Dr. Fidel Ruiz Healy).

A proposal to analyze the progression of non-dialytic chronic kidney disease by surrogate endpoints: introducing parametric survival models

Introduction

Chronic kidney disease (CDK) progression studies increasingly use surrogate endpoints based on the estimated glomerular filtration rate. The clinical characteristics of these endpoints bring new challenges in comparing groups of patients, as traditional Cox models may lead to biased estimates mainly because they do not assume a hazard function.

Objective

This study proposes the use of parametric survival analysis models with the three most commonly used endpoints in nephrology based on a case study. Estimated glomerular filtration rate (eGFR) decay > 5 mL/year, eGFR decline > 30%, and change in CKD stage were evaluated.

Method

The case study is a 5-year retrospective cohort study that enrolled 778 patients in the predialysis stage. Exponential, Weibull, Gompertz, lognormal, and logistic models were compared, and proportional hazard and accelerated failure time (AFT) models were evaluated.

Results

The endpoints had quite different hazard functions, demonstrating the importance of choosing appropriate models for each. AFT models were more suitable for the clinical interpretation of the effects of covariates on these endpoints.

Conclusion

Surrogate endpoints have different hazard distributions over time, which is already recognized by nephrologists. More flexible analysis techniques that capture these relevant clinical characteristics in decision-making should be encouraged and disseminated in nephrology research.

Intensivist physician-to-patient ratios and mortality in the intensive care unit

PURPOSE

A high daily census may hinder the ability of physicians to deliver quality care in the intensive care unit (ICU). We sought to determine the relationship between intensivist-to-patient ratios and mortality among ICU patients.

METHODS

We performed a retrospective cohort study of intensivist-to-patient ratios in 29 ICUs in 10 hospitals in the United States from 2018 to 2020. We used meta-data from progress notes in the electronic health record to determine an intensivist-specific caseload for each ICU day. We then fit a multivariable proportional hazards model with time-varying covariates to estimate the relationship between the daily intensivist-to-patient ratio and ICU mortality at 28 days.

RESULTS

The final analysis included 51,656 patients, 210,698 patient days, and 248 intensivist physicians. The average caseload per day was 11.8 (standard deviation: 5.7). There was no association between the intensivist-to-patient ratio and mortality (hazard ratio for each additional patient: 0.987, 95% confidence interval: 0.968-1.007, p = 0.2). This relationship persisted when we defined the ratio as caseload over the sample-wide average (hazard ratio: 0.907, 95% confidence interval: 0.763-1.077, p = 0.26) and cumulative days with a caseload over the sample-wide average (hazard ratio: 0.991, 95% confidence interval: 0.966-1.018, p = 0.52). The relationship was not modified by the presence of physicians-in-training, nurse practitioners, and physician assistants (p value for interaction term: 0.14).

CONCLUSIONS

Mortality for ICU patients appears resistant to high intensivist caseloads. These results may not generalize to ICUs organized differently than those in this sample, such as ICUs outside the United States.

Improving interventional causal predictions in regulatory risk assessment

In 2022, the US EPA published an important risk assessment concluding that "Compared to the current annual standard, meeting a revised annual standard with a lower level is estimated to reduce PM2.5-associated health risks in the 30 annually-controlled study areas by about 7-9% for a level of 11.0 µg/m3… and 30-37% for a level of 8.0 µg/m3." These are interventional causal predictions: they predict percentage reductions in mortality risks caused by different counterfactual reductions in fine particulate (PM2.5) levels. Valid causal predictions are possible if: (1) Study designs are used that can support valid causal inferences about the effects of interventions (e.g., quasi-experiments with appropriate control groups); (2) Appropriate causal models and methods are used to analyze the data; (3) Model assumptions are satisfied (at least approximately); and (4) Non-causal sources of exposure-response associations such as confounding, measurement error, and model misspecification are appropriately modeled and adjusted for. This paper examines two long-term mortality studies selected by the EPA to predict reductions in PM2.5-associated risk. Both papers use Cox proportional hazards (PH) models. For these models, none of these four conditions is satisfied, making it difficult to interpret or validate their causal predictions. Scientists, reviewers, regulators, and members of the public can benefit from more trustworthy and credible risk assessments and causal predictions by insisting that risk assessments supporting interventional causal conclusions be based on study designs, methods, and models that are appropriate for predicting effects caused by interventions.

Neighborhood Opportunity and Mortality Among Children and Adults in Their Households

BACKGROUND AND OBJECTIVES

Research has linked neighborhood opportunity to health outcomes in children and adults; however, few studies have examined neighborhood opportunity and mortality risk among children and their caregivers. The objective of this study was to assess associations of neighborhood opportunity and mortality risk in children and their caregivers over 11 years.

METHODS

Participants included 1 025 000 children drawn from the Mortality Disparities in American Communities study, a cohort developed by linking the 2008 American Community Survey to the National Death Index and followed for 11 years. Neighborhood opportunity was measured using the Child Opportunity Index, a measure designed to capture compounding inequities in access to opportunities for health.

RESULTS

Using hazard models, we observed inverse associations between Child Opportunity Index quintile and deaths among child and caregivers. Children in very low opportunity neighborhoods at baseline had 1.30 times the risk of dying over follow-up relative to those in very high opportunity neighborhoods (95% confidence interval [CI], 1.15-1.45), and this excess risk attenuated after adjustment for household characteristics (hazard ratio, 1.15; 95% CI, 0.98-1.34). Similarly, children in very low opportunity neighborhoods had 1.57 times the risk of experiencing the death of a caregiver relative to those in very high opportunity neighborhoods (95% CI, 1.50-1.64), which remained after adjustment (hazard ratio, 1.30; 95% CI, 1.23-1.38).

CONCLUSIONS

Our analyses advance understanding of the adverse consequences of inequitable neighborhood contexts for child well-being and underscore the potential importance of place-based policies for reducing disparities in child and caregiver mortality.

Individual frailty excess hazard models in cancer epidemiology

Unobserved individual heterogeneity is a common challenge in population cancer survival studies. This heterogeneity is usually associated with the combination of model misspecification and the failure to record truly relevant variables. We investigate the effects of unobserved individual heterogeneity in the context of excess hazard models, one of the main tools in cancer epidemiology. We propose an individual excess hazard frailty model to account for individual heterogeneity. This represents an extension of frailty modeling to the relative survival framework. In order to facilitate the inference on the parameters of the proposed model, we select frailty distributions which produce closed-form expressions of the marginal hazard and survival functions. The resulting model allows for an intuitive interpretation, in which the frailties induce a selection of the healthier individuals among survivors. We model the excess hazard using a flexible parametric model with a general hazard structure which facilitates the inclusion of time-dependent effects. We illustrate the performance of the proposed methodology through a simulation study. We present a real-data example using data from lung cancer patients diagnosed in England, and discuss the impact of not accounting for unobserved heterogeneity on the estimation of net survival. The methodology is implemented in the R package IFNS.

Utilizing shared frailty with the Cox proportional hazards regression: Post discharge survival analysis of CHF patients

Understanding patients' survival probability as well as the factors affecting it constitute a significant concern for researchers and practitioners, in particular for patients with severe chronic illnesses such as congestive heart failure (CHF). CHF is a clinical syndrome characterized by comorbidities and adverse medical events. Risk stratification to identify patients most likely to die shortly after hospital discharge can improve the quality of care by better allocating organizational resources and personalized interventions. Probability assessment improves clinical decision-making, contributes to personalized care, and saves costs. Although one of the most informative indices is the time to an adverse event for each patient, commonly analyzed using survival analysis methods, these are often challenging to implement due to the complexity of the medical data. Numerous studies have used the Cox proportional hazards (PH) regression method to generate the survival distribution pattern and factors affecting survival. This model, although advantageous for survival analysis, assumes the homogeneity of the hazard ratio across patients and independence of the observations in terms of survival time. These assumptions are often violated in real-world data, especially when the dataset is composed of readmission data for chronically ill patients, since these recurring observations are inherently dependent. This study ran the Cox PH regression on a feature set selected by machine learning algorithms from a rich hospital dataset. The event modeled here was patient mortality within 90 days post-hospital discharge. The sample was composed of medical records of patients hospitalized in the Israeli Sheba Medical Center more than once, with CHF as the primary diagnosis. We modeled the survival of CHF patients using the Cox PH regression with and without the shared frailty correction that addresses the shortcomings of the Cox Model. The results of the two models of the Cox PH regression - with and without the shared frailty correction were compared. The results demonstrate that the shared frailty correction, which was statistically significant in our analysis, improved the performance of the basic Cox PH model. While this is the main contribution, we also show that this model outperforms two commonly used measures (ADHERE and EFFECT) for predicting early mortality of CHF patients. Thus, the results illustrate how applying advanced analytics can outperform traditional methods. An additional contribution is the feature set selected using machine-learning methods that is different from those used in the extant literature.

Correcting for heterogeneity and non-comparability bias in multicenter clinical trials with a rescaled random-effect excess hazard model

In the presence of competing causes of event occurrence (e.g., death), the interest might not only be in the overall survival but also in the so-called net survival, that is, the hypothetical survival that would be observed if the disease under study were the only possible cause of death. Net survival estimation is commonly based on the excess hazard approach in which the hazard rate of individuals is assumed to be the sum of a disease-specific and expected hazard rate, supposed to be correctly approximated by the mortality rates obtained from general population life tables. However, this assumption might not be realistic if the study participants are not comparable with the general population. Also, the hierarchical structure of the data can induces a correlation between the outcomes of individuals coming from the same clusters (e.g., hospital, registry). We proposed an excess hazard model that corrects simultaneously for these two sources of bias, instead of dealing with them independently as before. We assessed the performance of this new model and compared it with three similar models, using extensive simulation study, as well as an application to breast cancer data from a multicenter clinical trial. The new model performed better than the others in terms of bias, root mean square error, and empirical coverage rate. The proposed approach might be useful to account simultaneously for the hierarchical structure of the data and the non-comparability bias in studies such as long-term multicenter clinical trials, when there is interest in the estimation of net survival.

Effectiveness of a community-based support programme to reduce social inequality in exclusive breastfeeding: study protocol for a cluster-randomised trial

BACKGROUND

Breastmilk is the ideal nutrition for infants, and breastfeeding protects infants and mothers from a range of adverse health outcomes. In Denmark, most mothers initiate breastfeeding but many cease within the first months resulting in just 14% reaching the World Health Organization recommendation of six months of exclusive breastfeeding. Furthermore, the low breastfeeding proportion at six months is characterised by a marked social inequality. A previous intervention tested in a hospital setting succeeded in increasing the proportion of mothers breastfeeding exclusively at six months. However, most breastfeeding support is provided within the Danish municipality-based health visiting programme. Therefore, the intervention was adapted to fit the health visiting programme and implemented in 21 Danish municipalities. This article reports the study protocol, which will be used to evaluate the adapted intervention.

METHODS

The intervention is tested in a cluster-randomised trial at the municipal level. A comprehensive evaluation approach is taken. The effectiveness of the intervention will be evaluated using survey and register data. Primary outcomes are the proportion of women who breastfeed exclusively at four months postpartum and duration of exclusive breastfeeding measured as a continuous outcome. A process evaluation will be completed to evaluate the implementation of the intervention; a realist evaluation will provide an understanding of the mechanisms of change characterising the intervention. Finally, a health economic evaluation will assess the cost-effectiveness and cost-utility of this complex intervention.

DISCUSSION

This study protocol reports on the design and evaluation of the Breastfeeding Trial - a cluster-randomised trial implemented within the Danish Municipal Health Visiting Programme from April 2022 to October 2023. The purpose of the programme is to streamline breastfeeding support provided across healthcare sectors. The evaluation approach is comprehensive using a multitude of data to analyse the effect of the intervention and inform future efforts to improve breastfeeding for all.

TRIAL REGISTRATION

Prospectively registered with Clinical Trials NCT05311631 https://clinicaltrials.gov/ct2/show/NCT05311631.

Validation of the INCREMENT-SOT-CPE score in a large cohort of liver transplant recipients with carbapenem-resistant Enterobacterales infection

BACKGROUND

Management of infections due to carbapenemase-resistant Enterobacterales (CRE) in solid organ transplant (SOT) recipients remains a difficult challenge. The INCREMENT-SOT-CPE score has been specifically developed from SOT recipients to stratify mortality risk, but an external validation is lacking.

METHODS

Multicenter retrospective cohort study of liver transplant (LT) recipients colonized with CRE infection who developed infection after transplant over 7-year period. Primary endpoint was all-cause 30-day mortality from infection onset. A comparison between INCREMENT-SOT-CPE and other selected scores was performed. A two-level mixed effects logistic regression model with random effects for the center was fitted. Performance characteristics at optimal cut-point were calculated. Multivariable Cox regression analysis of risk factors for all-cause 30-day mortality was carried out.

RESULTS

Overall, 250 CRE carriers developed infection after LT and were analyzed. The median age was 55 years (interquartile range [IQR]: 46-62) and 157 were males (62.8%). All-cause 30-day mortality was 35.6%. A sequential organ failure assessment (SOFA) score ≥ 11 showed a sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of 69.7%, 76.4%, 62.0%, 82.0%, and 74.0%, respectively. An INCREMENT-SOT-CPE ≥ 11 reported a sensitivity, specificity, PPV, NPV, and accuracy of 73.0%, 62.1%, 51.6%, 80.6% and 66.0%, respectively. At multivariable analysis acute renal failure, prolonged mechanical ventilation, INCREMENT-SOT-CPE score ≥ 11 and SOFA score ≥ 11 were independently associated with all-cause 30-day mortality, while a tigecycline-based targeted regimen was found to be protective.

CONCLUSIONS

Both INCREMENT-SOT-CPE ≥ 11 and SOFA ≥ 11 were identified as strong predictors of all-cause 30-day mortality in a large cohort of CRE carriers developing infection after LT.

A Bayesian joint model for multivariate longitudinal and time-to-event data with application to ALL maintenance studies

The most common type of cancer diagnosed among children is the acute lymphocytic leukemia (ALL). A study was conducted by Tata Translational Cancer Research Center (TTCRC) Kolkata, in which 236 children (diagnosed as ALL patients) were treated for the first two years (approximately) with two standard drugs (6MP and MTx) and were then followed nearly for the next three years. The goal is to identify the longitudinal biomarkers that are associated with time-to-relapse, and also to assess the effectiveness of the drugs. We develop a Bayesian joint model in which a linear mixed model is used to jointly model three biomarkers (i.e. white blood cell count, neutrophil count, and platelet count) and a semi-parametric proportional hazards model is used to model the time-to-relapse. Our proposed joint model can assess the effects of different covariates on the progression of the biomarkers, and the effects of the biomarkers (and the covariates) on time-to-relapse. In addition, the proposed joint model can impute the missing longitudinal biomarkers efficiently. Our analysis shows that the white blood cell (WBC) count is not associated with time-to-relapse, but the neutrophil count and the platelet count are significantly associated with it. We also infer that a lower dose of 6MP and a higher dose of MTx jointly result in a lower relapse probability in the follow-up period. Interestingly, we find that relapse probability is the lowest for the patients classified into the "high-risk" group at presentation. The effectiveness of the proposed joint model is assessed through the extensive simulation studies.

Time-varying effects are common in genetic control of gestational duration

Preterm birth is a major burden to neonatal health worldwide, determined in part by genetics. Recently, studies discovered several genes associated with this trait or its continuous equivalent - gestational duration. However, their effect timing, and thus clinical importance, is still unclear. Here, we use genotyping data of 31,000 births from the Norwegian Mother, Father and Child cohort (MoBa) to investigate different models of the genetic pregnancy "clock". We conduct genome-wide association studies using gestational duration or preterm birth, replicating known maternal associations and finding one new foetal variant. We illustrate how the interpretation of these results is complicated by the loss of power when dichotomizing. Using flexible survival models, we resolve this complexity and find that many of the known loci have time-varying effects, often stronger early in pregnancy. The overall polygenic control of birth timing appears to be shared in the term and preterm, but not very preterm periods, and exploratory results suggest involvement of the major histocompatibility complex genes in the latter. These findings show that the known gestational duration loci are clinically relevant, and should help design further experimental studies.

In-hospital and 6-month outcomes in patients with COVID-19 supported with extracorporeal membrane oxygenation (EuroECMO-COVID): a multicentre, prospective observational study

BACKGROUND

Extracorporeal membrane oxygenation (ECMO) has been widely used in patients with COVID-19, but uncertainty remains about the determinants of in-hospital mortality and data on post-discharge outcomes are scarce. The aims of this study were to investigate the variables associated with in-hospital outcomes in patients who received ECMO during the first wave of COVID-19 and to describe the status of patients 6 months after ECMO initiation.

METHODS

EuroECMO-COVID is a prospective, multicentre, observational study developed by the European Extracorporeal Life Support Organization. This study was based on data from patients aged 16 years or older who received ECMO support for refractory COVID-19 during the first wave of the pandemic-from March 1 to Sept 13, 2020-at 133 centres in 21 countries. In-hospital mortality and mortality 6 months after ECMO initiation were the primary outcomes. Mixed-Cox proportional hazards models were used to investigate associations between patient and management-related variables (eg, patient demographics, comorbidities, pre-ECMO status, and ECMO characteristics and complications) and in-hospital deaths. Survival status at 6 months was established through patient contact or institutional charts review. This study is registered with ClinicalTrials.gov, NCT04366921, and is ongoing.

FINDINGS

Between March 1 and Sept 13, 2020, 1215 patients (942 [78%] men and 267 [22%] women; median age 53 years [IQR 46-60]) were included in the study. Median ECMO duration was 15 days (IQR 8-27). 602 (50%) of 1215 patients died in hospital, and 852 (74%) patients had at least one complication. Multiorgan failure was the leading cause of death (192 [36%] of 528 patients who died with available data). In mixed-Cox analyses, age of 60 years or older, use of inotropes and vasopressors before ECMO initiation, chronic renal failure, and time from intubation to ECMO initiation of 4 days or more were associated with higher in-hospital mortality. 613 patients did not die in hospital, and 547 (95%) of 577 patients for whom data were available were alive at 6 months. 102 (24%) of 431 patients had returned to full-time work at 6 months, and 57 (13%) of 428 patients had returned to part-time work. At 6 months, respiratory rehabilitation was required in 88 (17%) of 522 patients with available data, and the most common residual symptoms included dyspnoea (185 [35%] of 523 patients) and cardiac (52 [10%] of 514 patients) or neurocognitive (66 [13%] of 512 patients) symptoms.

INTERPRETATION

Patient's age, timing of cannulation (<4 days vs ≥4 days from intubation), and use of inotropes and vasopressors are essential factors to consider when analysing the outcomes of patients receiving ECMO for COVID-19. Despite post-discharge survival being favourable, persisting long-term symptoms suggest that dedicated post-ECMO follow-up programmes are required.

FUNDING

None.

Increasing disparities in the proportions of active treatment and 5-year overall survival over time by age groups among older patients with gastric cancer in Korea

Purpose

As older patients with gastric cancer increase in Korea, no consensus indicative of anti-cancer treatment exists for the oldest old (age 85+). We investigated potential disparities in the proportion of surgery-including active treatment and the degree of survival improvement over time by age groups, and whether heterogeneity exists in the protective effect of time period on overall survival (OS) by age at diagnosis clusters.

Materials and methods

A nationwide cohort (N = 63,975) of older patients with gastric cancer (age at diagnosis 70+) in 2005-2012 were followed until the end of 2018. Patients were categorized into four time period groups by their year of diagnosis. Cancer treatment patterns and 5-year OS were analyzed accordingly, and a random coefficients Cox model with random intercepts and random slopes of time period by age at diagnosis clusters was employed.

Results

The mean age of patients was 76.4, and 60.4% were males. Most patients had 0-1 comorbidities (73.3%) and low-risk frailty scores (74.2%). Roughly two-thirds of patients received some form of anti-cancer treatment (62.4%), and while the number of comorbidities and the proportion of high-risk frailty scores trended toward an increase, the proportion of patients receiving anti-cancer treatment increased from 58% in 2005-2006 to 69.6% in 2011-2012. The proportion of surgery-including active treatment increased to over 70% in the 70-74 years old group, while stagnating at 10% in the 90+ years old group. Differences in the slope of 5-year OS improvement resulted in a widening survival gap between the old (age 70-84) and the oldest old. The protective effect of time period on OS hazard in the oldest old was not monotonically reduced with increasing "chronological" age but varied quite randomly, especially among female patients.

Conclusion

Our study showed no upper age limit in terms of benefiting from the advances in the detection and treatment of gastric cancer over time. Thus, "functional" age rather than "chronological" age should be the criterion for anti-cancer screening and treatment, and actual implementation of proven treatments in the oldest old patients to reduce their non-compliance with treatment in clinical practice is needed to improve gastric cancer survival for all.