Risk prediction models are widely used to guide real-world decision-making in areas such as healthcare and economics, and they also play a key role in estimating nuisance parameters in semiparametric inference. The super learner is a machine learning framework that combines a library of prediction algorithms into a meta-learner using cross-validated loss. In the context of right-censored data, careful consideration must be given to both the choice of the loss function and the estimation of expected loss. Moreover, estimators such as the inverse probability of censoring weighting requires accurate modeling and an estimator of the censoring distribution. We propose a novel approach to super learning for survival analysis that jointly evaluates the candidate learners for both the event-time distribution and the censoring distribution. Our method imposes no restrictions on the algorithms included in the library, accommodates competing risks, and does not rely on a single prespecified estimator of the censoring distribution. We establish a finite-sample bound on the average price we pay for using cross-validation, and show that this price vanishes asymptotically, up to poly-logarithmic terms, provided that the size of the library does not grow faster than at a polynomial rate in the sample size. We demonstrate the practical utility of our method using prostate cancer data and compare it to existing super learner algorithms for survival analysis using synthesized data.

Disadvantaged groups are often defined by characteristics such as income or ethnicity. Reducing health disparities by directly manipulating such exposures may be infeasible. Instead, interventions can target mediators between these exposures and health outcomes. Indirect effects estimated using mediation analysis, interventional effects or the interventional disparity measure can quantify the expected impact of such disparity-reducing interventions. They capture the impact of changing the mediator distribution evaluated among the total population. This means keeping individuals in their exposure group but hypothetically assigning them the mediator distribution of another group. However, when indirect effects are intended to inform about disparity-reducing interventions implemented among disadvantaged groups, estimating effects in the total population does not quantify the effect among those targeted. Instead, we propose evaluating the interventional disparity indirect effect directly among the disadvantaged individuals. We introduce the estimand and illustrate it using a register-based study examining a potential intervention improving medication initiation in low-income heart failure patients. We compare the expected change in 1-year mortality in a hypothetical world where low-income patients were as likely to initiate medication as high-income patients. We included 1700 patients and assessed intervention effects in low-income patients and the total population, respectively. Under the intervention, the 1-year mortality declined from 10.3% to 9.3% (95% CI 8.6% to 10.1%) among low-income patients but 6.6% to 6.2% (95% CI 6.0% to 6.5%) in the total population. In disparity research, evaluating intervention effects in the total population, rather than among disadvantaged groups, may impact the effect size. Therefore, when guiding future disparity-targeted interventions, measuring effects within disadvantaged groups is important.

OBJECTIVE:To assess adherence to the International Ovarian Tumor Analysis (IOTA) terminology in Danish routine clinical practice and to evaluate how non-adherence affects the diagnostic performance and calibration of the Assessment of Different NEoplasias in the adneXa (ADNEX) model and the Two-Step Strategy with modified benign descriptors (BD) with Cancer Antigen 125 (CA125). METHODS:This prospective, multicenter cohort study included patients ≥18 years with adnexal masses across 14 gynecology departments and general gynecology practices. Reference standard was histopathology for surgically managed patients and clinical follow-up for conservatively managed patients. Ultrasound descriptions recorded at recruitment using IOTA terminology by examining clinicians with varying experience and IOTA certification were reviewed by three blinded IOTA-certified experts to identify deviations from IOTA definitions. Based on expert reassessment of stored representative images, obvious non-adherent terminology was corrected. Agreement in modified BD applicability was assessed using Cohen's kappa (κ). Performance and calibration were compared using predictive values (10% threshold), area under the curve (AUC), prediction error, and calibration plots. RESULTS:Of 1065 enrolled patients, 948 constituted the complete-case cohort. Non-adherence to the IOTA terminology was identified in 198 (20.9%) clinician-recorded ultrasound descriptions. Agreement on the modified BD category was 90.0%, κ = 0.79. NPVs increased for both models (ADNEX: 95.7% to 96.8%; Two-Step: 95.4% to 96.9%), as did PPVs (ADNEX: 48.6% to 52.0%; Two-Step: 50.4% to 52.8%) and AUCs (ADNEX: 90.8% to 93.4%; Two-Step: 90.3% to 93.6%). Prediction error decreased, while overall calibration remained unchanged. CONCLUSION:Non-adherence to IOTA terminology is a potential barrier to successful implementation and highlights the need for strategies that promote consistent use of standardized IOTA terminology. Expert reassessment was based on stored still ultrasound images and may not fully capture dynamic features of real-time examination.

BACKGROUND:Observational and genetic evidence show associations of high remnant cholesterol levels with atherosclerotic cardiovascular disease (ASCVD). New drugs have been developed that substantially lower remnant cholesterol; however, the corresponding absolute risk reduction of ASCVD remains unclear. Remnant cholesterol can be measured directly or calculated, but few studies have analyzed the effects of directly measured remnant cholesterol. OBJECTIVE:To estimate the 10-year absolute risk reductions of ASCVD according to proportional reduction of individual very low-density lipoprotein (VLDL) cholesterol levels among individuals with levels above 1 mmol/L (39 mg/dL). METHODS:We used VLDL cholesterol measured by nuclear magnetic resonance spectroscopy to quantify directly measured remnant cholesterol. We estimated the reduction in the average 10-year ASCVD risk associated with an intervention targeting the 2021 individuals in the Copenhagen General Population Study with VLDL cholesterol levels above 1 mmol/L (39 mg/dL), assuming a proportional reduction in their individual VLDL cholesterol levels. RESULTS:We found that a 50% or 80% proportional reduction in VLDL cholesterol was associated with a 10-year absolute risk reduction of ASCVD of 3.0% (95% CI: 2.6%-3.4%) and 4.5% (3.9%-5.1%), respectively. CONCLUSION:This suggests a clinically meaningful benefit from lowering of VLDL cholesterol in primary prevention.

AIMS:Chronic kidney disease (CKD) remains a leading complication of diabetes, and understanding population-level trends is crucial for prevention and healthcare planning. We examined national trends in CKD prevalence, incidence, and mortality among people with diabetes. METHODS:In this nationwide register-based cohort study, we identified people with diabetes in Denmark between 2016 and 2024 using linked national registers. CKD was defined by ≥ 2 measurements of estimated glomerular filtration rate < 60 mL/min/1.73 m2 and urinary albumin-creatinine ratio ≥ 30 mg/g, or a hospital diagnosis or procedure code. Annual CKD prevalence was calculated on 1 January each year. CKD incidence and mortality rates were calculated annually and reported per 1000 person-years (PY), separately for type 1 diabetes (T1D) and type 2 diabetes (T2D) and stratified by sex and age. RESULTS:Between 2016 and 2024, CKD prevalence increased from 20.5% to 26.4% in T1D and from 24.8% to 37.3% in T2D. CKD incidence declined from 21.5 to 14.9 per 1000 PY in T1D and from 55.3 to 43.7 per 1000 PY in T2D, with the largest reductions in older people (aged ≥ 70 years with T2D [93.6 to 74.3 per 1000 PY]). Albuminuria prevalence increased, whereas incidence declined in both diabetes types. Mortality declined in people with and without CKD but remained substantially higher among those with CKD. CONCLUSION:From 2016 to 2024, CKD incidence declined while prevalence increased among people with diabetes in Denmark. Mortality remained high among people with CKD, highlighting the ongoing clinical burden and need for improved prevention and early detection.

Causal mediation analysis with random interventions has become an area of significant interest for understanding time-varying effects with longitudinal and survival outcomes. To tackle causal and statistical challenges due to the complex longitudinal data structure with time-varying confounders, competing risks, and informative censoring, there exists a general desire to combine machine learning techniques and semiparametric theory. In this article, we focus on targeted maximum likelihood estimation (TMLE) of longitudinal natural direct and indirect effects defined with random interventions. The proposed estimators are multiply robust, locally efficient, and directly estimate and update the conditional densities that factorize data likelihoods. We utilize the highly adaptive lasso (HAL) and projection representations to derive new estimators (HAL-EIC) of the efficient influence curves (EICs) of longitudinal mediation problems and propose a fast one-step TMLE algorithm using HAL-EIC while preserving the asymptotic properties. The proposed method can be generalized for other longitudinal causal parameters that are smooth functions of data likelihoods, and thereby provides a novel and flexible statistical toolbox.

The causal roadmap is a formal framework for causal and statistical inference that supports clear specification of the causal question, interpretable and transparent statement of required causal assumptions, robust inference, and optimal precision. The roadmap is thus particularly well-suited to evaluating longitudinal causal effects using large scale registries; however, application of the roadmap to registry data also introduces particular challenges. In this paper we provide a detailed case study of the longitudinal causal roadmap applied to the Danish National Registry to evaluate the comparative effectiveness of second-line diabetes drugs on dementia risk. Specifically, we evaluate the difference in counterfactual five-year cumulative risk of dementia if a target population of adults with type 2 diabetes had initiated and remained on GLP-1 receptor agonists (a second-line diabetes drug) compared to a range of active comparator protocols. Time-dependent confounding is accounted for through use of the iterated conditional expectation representation of the longitudinal g-formula as a statistical estimand. Statistical estimation uses longitudinal targeted maximum likelihood, incorporating machine learning. We provide practical guidance on the implementation of the roadmap using registry data, and highlight how rare exposures and outcomes over long-term follow up can raise challenges for flexible and robust estimators, even in the context of the large sample sizes provided by the registry. We demonstrate how simulations can be used to help address these challenges by supporting careful estimator pre-specification. We find a protective effect of GLP-1RAs compared to some but not all other second-line treatments.

Bacteremia is a well-known complication to surgery and may result in infective endocarditis (IE). Transurethral resection of the prostate (TUR-P) may give rise to bacteremia, but the associated risk of IE is not well described. We aimed to examine risk of infective endocarditis following TUR-P. We examined risk of IE following TUR-P between 2010 and 2020 in comparison with an age-matched (match-ratio 1:1) cohort from the background population. Patients were considered exposed to TUR-P related IE 6 months after TUR-P. Comparisons were estimated using cumulative incidences and multivariable time-dependent Cox regression models. A total of 25,781 males underwent TUR-P (11.4

We consider estimation of conditional hazard functions and densities over the class of multivariate càdlàg functions with uniformly bounded sectional variation norm when data are either fully observed or subject to right-censoring. We demonstrate that the empirical risk minimizer is either not well-defined or not consistent for estimation of conditional hazard functions and densities. Under a smoothness assumption about the data-generating distribution, a highly-adaptive lasso estimator based on a particular data-adaptive sieve achieves the same convergence rate as has been shown to hold for the empirical risk minimizer in settings where the latter is well-defined. We use this result to study a highly-adaptive lasso estimator of a conditional hazard function based on right-censored data. We also propose a new conditional density estimator and derive its convergence rate. Finally, we show that the result is of interest also for settings where the empirical risk minimizer is well-defined, because the highly-adaptive lasso depends on a much smaller number of basis function than the empirical risk minimizer.

INTRODUCTION:Obstetrics research has predominantly focused on the management and identification of factors associated with labor dystocia. Despite these efforts, clinicians currently lack the necessary tools to effectively predict a woman's risk of experiencing labor dystocia. Therefore, the objective of this study was to create a predictive model for labor dystocia. MATERIAL AND METHODS:The study population included nulliparous women with a single baby in the cephalic presentation in spontaneous labor at term. With a cohort-based registry design utilizing data from the Copenhagen Pregnancy Cohort and the Danish Medical Birth Registry, we included women who had given birth from 2014 to 2020 at Copenhagen University Hospital-Rigshospitalet, Denmark. Logistic regression analysis, augmented by a super learner algorithm, was employed to construct the prediction model with candidate predictors pre-selected based on clinical reasoning and existing evidence. These predictors included maternal age, pre-pregnancy body mass index, height, gestational age, physical activity, self-reported medical condition, WHO-5 score, and fertility treatment. Model performance was evaluated using the area under the receiver operating characteristics curve (AUC) for discriminative capacity and Brier score for model calibration. RESULTS:A total of 12,445 women involving 5,525 events of labor dystocia (44%) were included. All candidate predictors were retained in the final model, which demonstrated discriminative ability with an AUC of 62.3% (95% CI:60.7-64.0) and Brier score of 0.24. CONCLUSIONS:Our model represents an initial advancement in the prediction of labor dystocia utilizing readily available information obtainable upon admission in active labor. As a next step further model development and external testing across other populations is warranted. With time a well-performing model may be a step towards facilitating risk stratification and the development of a user-friendly online tool for clinicians.

The aim of this study was to investigate the potential for pulp revascularization in relation to patient age at the time of injury following luxation injury of mature anterior permanent teeth. A total of 93 teeth from 70 patients were included. The patients were divided into subgroups based on their age at the time of the injury. Statistics: the Aalen–Johansen method was used to estimate the risks of pulp canal obliteration (PCO) and pulp necrosis (PN). The absolute 2 year risks of PCO and PN were obtained with cause-specific Cox regression and reported separately for each cohort, standardised to age at injury and degree of repositioning. For the group younger than 15 years of age, the risk of PN after 12 months was 62.3

BACKGROUND:Glucagon-like peptide-1 receptor agonists (GLP-1 RA) are increasingly being prescribed in drug-naive patients. We aimed to contrast add-on therapy, adherence, and changes in biomarkers, 1 year after treatment initiation with GLP-1 RA or metformin. METHODS:Using Danish nationwide registers, we included incident GLP-1 RA or metformin users from 2018 to 2021 with glycated hemoglobin (HbA1c) ≥ 42 mmol/mol. GLP-1 RA initiators were matched to metformin initiators in a ratio of 1:1 to assess outcomes in prediabetes and diabetes. Main outcomes analyzed were 1-year risk of add-on glucose-lowering medication and 1-year risk of nonadherence. One-year risks were estimated with multiple logistic regression and standardized. Multiple linear regression was used to estimate the average differences in biomarker changes. RESULTS:In total, 1778 individuals initiating GLP-1 RA and metformin were included. After standardizing for various factors, GLP-1 RA compared with metformin was associated with reduced 1-year risk of add-on glucose-lowering treatment in patients with prediabetes (1-year risk ratio [RR]: 0.27, 95% confidence interval [CI]: 0.10-0.44) and diabetes (RR: 0.67, 95% CI: 0.37-0.98). GLP-1 RA was associated with higher 1-year risk of nonadherence among patients with prediabetes (RR: 1.60, 95% CI: 1.45-1.75), but no difference in patients with diabetes (RR: 0.88, 95% CI: 0.70-1.06). Compared to metformin, GLP-1 RA was associated with greater HbA1c reduction (prediabetes: -2.59 mmol/mol 95% CI: -3.10 to -2.09, diabetes: -3.79 mmol/mol, 95% CI: -5.28 to -2.30). CONCLUSIONS:GLP-1 RA was associated with a reduced risk of additional glucose-lowering medication, achieving better glycated hemoglobin control overall. However, among patients with prediabetes, metformin was associated with better adherence.

IntroductionThis study investigated the efficacy of a digital health solution utilizing smartphone images of colorimetric test-strips for home-based salivary uric acid (sUA) measurement to predict pre-eclampsia (PE), pregnancy-induced hypertension (PIH), and intrauterine growth restriction (IUGR).Methods495 pregnant women were included prospectively at Zealand University Hospital, Denmark. They performed weekly self-tests from mid-pregnancy until delivery and referred these for analysis by a smartphone-app. Baseline characteristics were obtained at recruitment and pregnancy outcomes from the journals. The mean compliance rate of self-testing was assessed. For the statistical analyses, standard color analyses deduced the images into the red-green-blue (RGB) color model value, to observe the individual, longitudinal pattern throughout the pregnancy for each outcome. Extended color analyses were applied, deducing the images into 72 individual color variables that reflected the four dominant color models. The individual discriminatory ability was assessed by calculating the area under the curve for the outcome of PE, and the outcome of hypertensive pregnancy disorders solely or combined with IUGR at 25 weeks of gestation and for the weekly color change between 20 and 25 weeks of gestation.ResultsThirty-four women (6.9%) developed PE, 17 (3.4%) PIH, and 10 (2.0%) IUGR. The overall mean compliance rate was 67%, increasing to 77% after updating the smartphone-app halfway through the study. The longitudinal pattern of the RGB value showed a wide within-person variability, and discrimination was not achieved. However, it was noted that all women with IUGR repeatedly had RGB values below 110, contrasting women with non-IUGR. Significant discriminatory ability was achieved for 8.2% of the analyses of individual color variables, of which 27.4% summarized the Hue color variable. However, the analyses lacked consistency regarding outcome group and gestational age.ConclusionThis study is the first proof-of-concept that digital self-tests utilizing colorimetric sUA measurement for the prediction of PE, PIH, and IUGR is acceptable to pregnant women. The discriminatory ability was not found be sufficient to have clinical value. However, being the first study that compares individual color variables of the four dominant color models, this study adds important methodological insights into the expanding field of smartphone-assisted colorimetric test-strips.

Abstract Background and Aims The effect of vaccinations against SARS-CoV-2 on risk of infection and subsequent adverse outcomes in patients with kidney dysfunction beyond end-stage kidney disease remains uncertain. Based on nationwide data from multiple health care registers, the study aims to evaluate vaccination effect on rates of SARS-CoV-2 infection and subsequent risk of associated mortality in patients with kidney disease beyond end-stage kidney disease. Method A population-based retrospective cohort study on all Danish residents ≥18 years with an eGFR ≤90 ml/min/1.73 m2 identified between February 1st 2021 and June 27th 2022 corresponding to predominance of Alpha-, Delta-, and Omicron-variant. Outcomes compared based on vaccination state (unvaccinated, vaccinated, and booster vaccinated) across predefined eGFR categories (eGFR 60-90 mL/min/1.73 m2; eGFR 45-59 mL/min/1.73 m2, eGFR 30-44 mL/min/1.73 m2; eGFR <30 mL/min/1.73 m2). Risk of infection reported as age- and gender-standardized rate ratios of positive SARS-CoV-2 PCR and subsequent risk of death reported as age- and gender standardized 30-day mortality risk. Standardized risks computed based on hazards obtained in multiple cause-specific Cox regression models. Results A total of 982,047 adults with eGFR ≤90 ml/min/1.73 m2 identified. Gender distribution was 43.9% male, median age was 71 [IQR 61–78] years, median eGFR 75 [IQR 63-84], and 13.1% diabetes. Population rates of vaccination were 3.5%, 80.1%, and 95.4% for Alpha-, Delta-, and Omicron-variant, with 66.2% booster vaccinated at Omicron-variant baseline (Pfizer BionTech 88.3%, Moderna 9.4%, other vaccinations 2.3%). In total 359,428 SARS-CoV-2 infections were identified (Alpha: n = 7,861, Delta: n = 27,868, Omicron: n = 323,699), corresponding to a crude rate of positive SARS-CoV-2 PCR test of 629.0 per 100,000 person-weeks overall. Vaccination and booster vaccination were associated with progressively lower rate of SARS-CoV-2 infection. Age- and gender-standardized rate ratios are shown in Fig. 1. Overall, mean 30-day risks of death were 7.1‰ [95% CI 6.7-7.6], 3.8‰ [95% CI 3.5-4.0], and 1.4‰ [95% CI 1.3-1.4] in unvaccinated, vaccinated, and booster vaccinated persons, respectively. Vaccination and booster vaccination were associated with benefit on 30-day mortality across all strata of eGFR despite progressive increase in risk of death observed with declining kidney function. Age- and gender standardized 30-day mortality rates following a positive PCR for SARS-CoV-2 across granular eGFR are shown in Fig. 2. Conclusion Based on comprehensive nationwide data covering Alpha-, Delta-, and Omicron-variant predominance in Denmark, vaccination and booster vaccination against COVID-19 were associated with lower rates of SARS-CoV-2 infection and benefit on subsequent risk of 30-day mortality across all strata of eGFR.

Objective To train and test a super learner strategy for risk prediction of kidney failure and mortality in people with incident moderate to severe chronic kidney disease (stage G3b to G4). Design Multinational, longitudinal, population based, cohort study. Settings Linked population health data from Canada (training and temporal testing), and Denmark and Scotland (geographical testing). Participants People with newly recorded chronic kidney disease at stage G3b-G4, estimated glomerular filtration rate (eGFR) 15-44 mL/min/1.73 m 2 . Modelling The super learner algorithm selected the best performing regression models or machine learning algorithms (learners) based on their ability to predict kidney failure and mortality with minimised cross-validated prediction error (Brier score, the lower the better). Prespecified learners included age, sex, eGFR, albuminuria, with or without diabetes, and cardiovascular disease. The index of prediction accuracy, a measure of calibration and discrimination calculated from the Brier score (the higher the better) was used to compare KDpredict with the benchmark, kidney failure risk equation, which does not account for the competing risk of death, and to evaluate the performance of KDpredict mortality models. Results 67 942 Canadians, 17 528 Danish, and 7740 Scottish residents with chronic kidney disease at stage G3b to G4 were included (median age 77-80 years; median eGFR 39 mL/min/1.73 m 2 ). Median follow-up times were five to six years in all cohorts. Rates were 0.8-1.1 per 100 person years for kidney failure and 10-12 per 100 person years for death. KDpredict was more accurate than kidney failure risk equation in prediction of kidney failure risk: five year index of prediction accuracy 27.8% (95% confidence interval 25.2% to 30.6%) versus 18.1% (15.7% to 20.4%) in Denmark and 30.5% (27.8% to 33.5%) versus 14.2% (12.0% to 16.5%) in Scotland. Predictions from kidney failure risk equation and KDpredict differed substantially, potentially leading to diverging treatment decisions. An 80-year-old man with an eGFR of 30 mL/min/1.73 m 2 and an albumin-to-creatinine ratio of 100 mg/g (11 mg/mmol) would receive a five year kidney failure risk prediction of 10% from kidney failure risk equation (above the current nephrology referral threshold of 5%). The same man would receive five year risk predictions of 2% for kidney failure and 57% for mortality from KDpredict. Individual risk predictions from KDpredict with four or six variables were accurate for both outcomes. The KDpredict models retrained using older data provided accurate predictions when tested in temporally distinct, more recent data. Conclusions KDpredict could be incorporated into electronic medical records or accessed online to accurately predict the risks of kidney failure and death in people with moderate to severe CKD. The KDpredict learning strategy is designed to be adapted to local needs and regularly revised over time to account for changes in the underlying health system and care processes.

This study used demographic data in a novel prediction model to identify areas with high risk of out-of-hospital cardiac arrest (OHCA) in order to target prehospital preparedness. We combined data from the nationwide Danish Cardiac Arrest Registry with geographical- and demographic data on a hectare level. Hectares were classified in a hierarchy according to characteristics and pooled to square kilometers (km2). Historical OHCA incidence of each hectare group was supplemented with a predicted annual risk of at least 1 OHCA to ensure future applicability. We recorded 19,090 valid OHCAs during 2016 to 2019. The mean annual OHCA rate was highest in residential areas with no point of public interest and 100 to 1000 residents per hectare (9.7/year/km2) followed by pedestrian streets with multiple shops (5.8/year/km2), areas with no point of public interest and 50 to 100 residents (5.5/year/km2), and malls with a mean annual incidence per km2 of 4.6. Other high incidence areas were public transport stations, schools and areas without a point of public interest and 10 to 50 residents. These areas combined constitute 1496 km2 annually corresponding to 3.4% of the total area of Denmark and account for 65% of the OHCA incidence. Our prediction model confirms these areas to be of high risk and outperforms simple previous incidence in identifying future risk-sites. Two thirds of out-of-hospital cardiac arrests were identified in only 3.4% of the area of Denmark. This area was easily identified as having multiple residents or having airports, malls, pedestrian shopping streets or schools. This result has important implications for targeted intervention such as automatic defibrillators available to the public. Further, demographic information should be considered when implementing such interventions.

Background: Heart failure (HF) is associated with increased risk of death and a hospitalization, but for patients initiating guideline directed medical therapy, it is unknown how high these risks are compared to the general population - and how this may vary depending on age and comorbidity. Methods: In this retrospective cohort study, we identified patients diagnosed with HF in the period 2011-2017, surviving the initial 120 days after diagnosis. Patients who were on angiotensin converting enzyme inhibitor (ACEi)/ angiotensin receptor blocker (ARB) and beta-blocker were included and matched to 5 non-HF individuals from the background population each based on age and sex. We assessed the 5-year risk of all-cause death, HF and non-HF hospitalization according to sex and age and baseline comorbidity. Results: We included 35,367 patients with HF and 176,835 matched non-HF individuals. Patients with HF had a five-year excess risk (absolute risk difference) of death of 13% (31% [for HF] - 18% [for non-HF]), of HF hospitalization of 17% and of non-HF hospitalization of 24%. Excess risk of death increased with increasing age, whereas the relative risk decreased- for women in their twenties, the excess risk was 7%, risk ratio 7.2, while the excess risk was 18%, risk ratio 1.5 for women in their eighties. Having HF as a 60-year old man was associated with a five-year risk of death similar to a 75-year old man without HF. Further, HF was associated with an excess risk of non-HF hospitalization, ranging from 8% for patients >85 years to 30% for patients <30 years. Conclusion: Regardless of age, sex and comorbidity, HF was associated with excess risk of mortality and non-HF hospitalizations, but the relative risk ratio diminishes sharply with advancing age, which may influence allocation of resources for medical care across populations.