Modelling recurrent events: a tutorial for analysis in epidemiology

Predicting urinary stone recurrence: a joint model analysis of repeated 24-hour urine collections from the MSTONE database

To address the limitations in existing urinary stone recurrence (USR) models, including failure to account for changes in 24-hour urine (24U) parameters over time and ignoring multiplicity of stone recurrences, we presented a novel statistical method to jointly model temporal trends in 24U parameters and multiple recurrent stone events. The MSTONE database spanning May 2001 to April 2015 was analyzed. A joint recurrent model was employed, combining a linear mixed-effects model for longitudinal 24U parameters and a recurrent event model with a dynamic first-order Autoregressive (AR(1)) structure. A mixture cure component was included to handle patient heterogeneity. Comparisons were made with existing methods, multivariable Cox regression and conditional Prentice-Williams-Peterson regression, both applied to established nomograms. Among 396 patients (median follow-up of 2.93 years; IQR, 1.53-4.36 years), 34.6% remained free of stone recurrence throughout the study period, 30.0% experienced a single recurrence, and 35.4% had multiple recurrences. The joint recurrent model with a mixture cure component identified significant associations between 24U parameters - including urine pH (adjusted HR = 1.991; 95% CI 1.490-2.660; p < 0.001), total volume (adjusted HR = 0.700; 95% CI 0.501-0.977; p = 0.036), potassium (adjusted HR = 0.983; 95% CI 0.974-0.991; p < 0.001), uric acid (adjusted HR = 1.528; 95% CI 1.105-2.113, p = 0.010), calcium (adjusted HR = 1.164; 95% CI 1.052-1.289; p = 0.003), and citrate (adjusted HR = 0.796; 95% CI 0.706-0.897; p < 0.001), and USR, achieving better predictive performance compared to existing methods. 24U parameters play an important role in prevention of USR, and therefore, patients with a history of stones are recommended to closely monitor for future recurrence by regularly conducting 24U tests.

A Simple Score (MOG-AR) to Identify Individuals at High Risk of Relapse After MOGAD Attack

BACKGROUND AND OBJECTIVES

To identify predictors for relapse in patients with myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) and to develop and validate a simple risk score for predicting relapse.

METHODS

In China National Registry of Neuro-Inflammatory Diseases (CNRID), we identified patients with MOGAD from March 2023 and followed up prospectively to September 2023. The primary endpoint was MOGAD relapse, confirmed by an independent panel. Patients were randomly divided into model development (75%) and internal validation (25%) cohorts. Prediction models were constructed and internally validated using Andersen-Gill models. Nomogram and relapse risk score were generated based on the final prediction models.

RESULTS

A total of 188 patients (comprising 612 treatment episodes) were included in cohorts. Female (HR: 0.687, 95% CI 0.524-0.899, p = 0.006), onset age 45 years or older (HR: 1.621, 95% CI 1.242-2.116, p < 0.001), immunosuppressive therapy (HR: 0.338, 95% CI 0.239-0.479, p < 0.001), oral corticosteroids >3 months (HR 0.449, 95% CI 0.326-0.620, p < 0.001), and onset phenotype (p < 0.001) were identified as factors associated with MOGAD relapse. A predictive score, termed MOG-AR (Immunosuppressive therapy, oral Corticosteroids, Onset Age, Sex, Attack phenotype), derived in prediction model, demonstrated strong predictive ability for MOGAD relapse. MOG-AR score of 13-16 indicates a higher risk of relapse (HR: 3.285, 95% CI 1.473-7.327, p = 0.004).

DISCUSSION

The risk of MOGAD relapse seems to be predictable. Further validation of MOG-AR score developed from this cohort to determine appropriate treatment and monitoring frequency is warranted.

TRIAL REGISTRATION INFORMATION

CNRID, NCT05154370, registered December 13, 2021, first enrolled December 15, 2021.

Changes in the association between educational achievement, attainment and subsequent mental health: A survival analysis of 21 Swedish graduation cohorts

BACKGROUND

Low academic achievement and low educational attainment in adolescence is associated with higher risks of internalizing disorders later in life. However, less is known regarding if these associations vary over time across cohorts. The aim of this study was to investigate temporal changes in the association between academic achievement or educational attainment and subsequent inpatient treatment for internalizing disorders among Swedish youths.

METHODS

Register data on all students graduating from compulsory school in Sweden between 1990 and 2010 (N = 2 252 703) were used. Students were followed for a maximum of 8 years using discrete time proportional hazard models. Internalized disorders were measured by specialized inpatient psychiatric care for depression or anxiety disorders. Academic achievement was measured by grades at the end of compulsory school, and educational attainment by completion of upper secondary school.

RESULTS

The positive association between inpatient treatment for internalizing disorders and both low compulsory school achievement and non-completion of upper secondary school became stronger in more recent cohorts. The results were completely driven by girls and native-born youth.

CONCLUSIONS

Low compulsory school achievements and failure to complete upper secondary school has become more important risk factors for inpatient treatment for internalizing disorders, particularly in native-born youth and girls. More research is needed to establish whether youth with internalizing disorders increasingly fail in school or whether low achievement has become more harmful for mental health.

Stereotactic radiosurgery (SRS) for patients with brainstem cerebral cavernous malformations (CCMs): an international, multicentric study

Brainstem cerebral cavernous malformations (CCM) are clinically more aggressive compared to superficial CCMs. Due to their location, resection can be challenging, making stereotactic radiosurgery (SRS) an attractive alternative for symptomatic patient. Brainstem CCM patients (n = 170) were treated with Gamma Knife SRS at 11 radiosurgical centers. Hemorrhagic risk reduction, risk factors of post-SRS hemorrhage, and clinical outcomes were retrospectively analyzed. Most patients had a single (165/170 patients) brainstem CCMs treated; the majority of CCMs (165/181) presented with bleeding. Single-session SRS decreased the risk of repeat hemorrhage in patients with hemorrhagic brainstem CCM (HR: 0.17, p < 0.001) using recurrent multivariate analysis. The annual hemorrhage rate decreased from 14.8 per 100 CCM-years before SRS to 2.3 after treatment. Using univariate Cox-analysis, the probability of a new hemorrhages after SRS was reduced for patient older than 35 years (HR = 0.21, p = 0.002) and increased with a margin dose > 13 Gy (HR = 2.57, p = 0.044). Adverse radiation effect (ARE) occurred in 9 patients (5.3%) and was symptomatic in four (2.4%). At a median follow-up of 3.4 years (Inter-quartile range: 5.4), 13 patients (8.0%) had a worsened clinical status, with the treated CCM being the cause in 5.6% (10) of the patients. Single-session SRS decreased the risk of repeat hemorrhage in patients with hemorrhagic brainstem CCM and conveyed this benefit with a low risk of advrse radiation effects (ARE) and worsening clinical status.

Association of Coprescribing of Gabapentinoid and Other Psychoactive Medications With Altered Mental Status and Falls in Adults Receiving Dialysis

RATIONALE & OBJECTIVE

Prescribing psychoactive medications for patients with kidney disease is common, but for patients receiving dialysis, some medications may be inappropriate. We evaluated the association of coprescribing gabapentinoids and other psychoactive potentially inappropriate medications (PPIMs) (e.g., sedatives, opioids) with altered mental status (AMS) and falls, and whether the associations are modified by frailty.

STUDY DESIGN

Observational cohort study.

SETTING

& Participants: Adults receiving dialysis represented in the United States Renal Data System who had an active gabapentinoid prescription and no other PPIM prescriptions in the prior 6 months.

EXPOSURE

PPIM coprescribing, or the presence of overlapping prescriptions of a gabapentinoid and ≥1 additional PPIM.

OUTCOMES

Acute care visits for AMS and injurious falls.

ANALYTICAL APPROACH

Prentice-Williams-Petersen Gap Time models estimated the association between PPIM coprescribing and each outcome, adjusting for demographics, comorbidities, and frailty (assessed by a validated frailty index (FI)). Each model tested for interaction between PPIM coprescribing and frailty.

RESULTS

Overall, PPIM coprescribing was associated with increased hazard of AMS (HR: 1.66 [95% CI 1.44, 1.92]) and falls (HR: 1.55 [95% CI 1.36, 1.77]). Frailty significantly modified the effect of PPIM coprescribing on the hazard of AMS (interaction p=0.01), but not falls. Among individuals with low frailty (FI=0.15), the hazard ratio for AMS with PPIM co-prescribing was 2.14 (95% CI: 1.69, 2.71); while for individuals with severe frailty (FI=0.34), the hazard ratio for AMS with PPIM coprescribing was 1.64 (95% CI: 1.42, 1.89). Individuals with PPIM coprescribing and severe frailty (FI =0.34) had the highest hazard of AMS [HR 4.04 (95% CI: 3.20, 5.10)] and falls [HR 2.77 (95% CI: 2.27, 3.38)] compared to non-frail individuals without PPIM coprescribing.

LIMITATIONS

Outcome ascertainment bias; residual confounding.

CONCLUSIONS

Compared to gabapentinoid prescriptions alone, PPIM coprescribing was associated with an increased risk of AMS and falls. Clinicians should consider these risks when coprescribing PPIMs to patients receiving dialysis.

Evaluating dynamic and predictive discrimination for recurrent event models: use of a time-dependent C-index

Interest in analyzing recurrent event data has increased over the past few decades. One essential aspect of a risk prediction model for recurrent event data is to accurately distinguish individuals with different risks of developing a recurrent event. Although the concordance index (C-index) effectively evaluates the overall discriminative ability of a regression model for recurrent event data, a local measure is also desirable to capture dynamic performance of the regression model over time. Therefore, in this study, we propose a time-dependent C-index measure for inferring the model's discriminative ability locally. We formulated the C-index as a function of time using a flexible parametric model and constructed a concordance-based likelihood for estimation and inference. We adapted a perturbation-resampling procedure for variance estimation. Extensive simulations were conducted to investigate the proposed time-dependent C-index's finite-sample performance and estimation procedure. We applied the time-dependent C-index to three regression models of a study of re-hospitalization in patients with colorectal cancer to evaluate the models' discriminative capability.

Cardiovascular Risks With SGLT2 Inhibitors in Clinical Practice Among Patients With Type 2 Diabetes

Importance

Cardiovascular disease (CVD) can be recurrent during type 2 diabetes (T2D) progression in this aging population. The effectiveness of sodium-glucose cotransporter 2 inhibitor (SGLT2i) therapy on total (ie, first and subsequent) CVD among patients with T2D in clinical practice remains uncertain.

Objective

To analyze the comparative association of SGLT2i vs dipeptidyl peptidase 4 inhibitor (DPP4i) therapy with total CVD among patients with T2D in clinical practice.

Design, Setting, and Participants

This retrospective cohort study used electronic medical records at the National Cheng Kung University Hospital, a leading medical center in Taiwan, from 2015 through 2021. Adult patients with T2D who initiated first use of the study drugs from 2016 through 2019, with up to 6 years of follow-up, were identified.

Main Outcomes and Measures

The primary outcomes included total composite CVD events and individual CVD subtypes (ie, atrial fibrillation, coronary heart disease, heart failure, stroke, myocardial infarction, and transient ischemic attack). A shared frailty model analysis was used to assess the association of treatment with repeat CVD events. Data from patients at high risk for CVD recurrence were further analyzed. Data were analyzed from September 1, 2022, to December 31, 2023.

Results

Overall, 8384 patients with T2D were identified (mean [SD] age, 63.7 [12.4] years; 4645 [55.4%] male). A total of 1632 propensity score-matched pairs of SGLT2i (mean [SD] age, 57.8 [12.0] years; 673 [41.2%] female and 959 [58.8%] male) and DPP4i (mean [SD] age, 58.2 [12.9] years; 655 [40.1%] female and 977 [59.9%] male) users were included. SGLT2i was associated with reduced total CVD risk vs DPP4i therapy (hazard ratio [HR], 0.82 [95% CI, 0.69-0.98]) but not the first CVD event (with the use of SGLT2i therapy were more prominent for patients at high risk of CVD (ie, HR, 0.70 [95% CI, 0.62-0.80] for individuals with estimated glomerular filtration rate lower than 60 mL/min/1.73 m2; HR, 0.70 [95% CI, 0.64-0.78]; for individuals having any diabetes-related complications; and HR, 0.72 [95% CI, 0.65-0.80] for individuals with a history of CVD) compared with the overall cohort. Among patients at high risk of CVD, greater reduced total CVD burden associated with SGLT2i therapy was observed for women vs men (eg, HR, 0.59 [95% CI, 0.49-0.72] in the subgroup with CVD history).

Conclusions and Relevance

In this cohort study of patients with T2D, the use of SGLT2is vs DPP4is was associated with reduced total cardiovascular burden, suggesting that long-term use of this therapy may optimize treatment benefit among patients with chronic CVD. The SGLT2i-associated benefit among patients with high risk of CVD encourages the prioritization of SGLT2i use for these vulnerable individuals.

Long-Term Exposure to Air Pollution and Risk of Acute Lower Respiratory Infections in the Danish Nurse Cohort

Rationale: Air pollution is a major risk factor for chronic cardiorespiratory diseases, affecting the immune and respiratory systems' functionality, but epidemiological evidence in respiratory infections remains sparse. Objectives: We aimed to assess the association of long-term exposure to ambient air pollution with the risk of developing new and recurrent acute lower respiratory infections (ALRIs), characterized by persistently severe symptoms necessitating hospital contact, and identify the potential susceptible populations by socioeconomic status, smoking, physical activity status, overweight, and comorbidity with chronic lung disease. Methods: We followed 23,912 female nurses from the Danish Nurse Cohort (age >44 yr) from baseline (1993 or 1999) until 2018 for incident and recurrent ALRIs defined by hospital contact (inpatient, outpatient, and emergency room) data from the National Patient Register. Residential annual mean concentrations of fine particulate matter, nitrogen dioxide (NO2), and black carbon were modeled using the Danish Eulerian Hemispheric Model/Urban Background Model/Air Geographic Information System. We used marginal Cox models with time-varying exposures to assess the association of 3-year running mean air pollution level with incident and recurrent ALRIs and examined effect modification by age, socioeconomic status, smoking, physical activity, body mass index, and comorbidity with asthma or chronic obstructive pulmonary disease (COPD). Results: During a 21.3-year mean follow-up, 4,746 ALRIs were observed, of which 2,553 were incident. We observed strong positive associations of all three pollutants with incident ALRIs, with hazard ratios and 95% confidence intervals of 1.19 (1.08-1.31) per 2.5 μg/m3 for fine particulate matter, 1.17 (1.11-1.24) per 8.0 μg/m3 for NO2, and 1.09 (1.05-1.12) per 0.3 μg/m3 for black carbon, and slightly stronger associations with recurrent ALRIs. Associations were strongest in patients with COPD and nurses with low physical activity. Conclusions: Long-term exposure to air pollution at low levels was associated with risks of new and recurrent ALRIs, with patients with COPD and physically inactive subjects most vulnerable.

Receipt of Opioid Agonist Treatment in provincial correctional facilities in British Columbia is associated with a reduced hazard of nonfatal overdose in the month following release

BACKGROUND

In many jurisdictions, policies restrict access to Opioid Agonist Treatment (OAT) in correctional facilities. Receipt of OAT during incarceration is associated with reduced risk of fatal overdose after release but little is known about the effect on nonfatal overdose. This study aimed to examine the association between OAT use during incarceration and nonfatal overdose in the 30 days following release.

METHODS AND FINDINGS

Using linked administrative healthcare and corrections data for a random sample of 20% of residents of British Columbia, Canada we examined releases from provincial correctional facilities between January 1, 2015 -December 1, 2018, among adults (aged 18 or older at the time of release) with Opioid Use Disorder. We fit Andersen-Gill models to examine the association between receipt of OAT in custody and the hazard of nonfatal following release. We conducted secondary analyses to examine the association among people continuing treatment initiated prior to their arrest and people who initiated a new episode of OAT in custody separately. We also conducted sex-based subgroup analyses. In this study there were 4,738 releases of 1,535 people with Opioid Use Disorder. In adjusted analysis, receipt of OAT in custody was associated with a reduced hazard of nonfatal overdose (aHR 0.55, 95% CI 0.41, 0.74). This was found for prescriptions continued from community (aHR 0.49, 95%CI 0.36, 0.67) and for episodes of OAT initiated in custody (aHR 0.58, 95%CI 0.41, 0.82). The effect was greater among women than men.

CONCLUSIONS

OAT receipt during incarceration is associated with a reduced hazard of nonfatal overdose after release. Policies to expand access to OAT in correctional facilities, including initiating treatment, may help reduce harms related to nonfatal overdose in the weeks following release. Differences in the effect seen among women and men indicate a need for gender-responsive policies and programming.

Application of optical coherence tomography in cardiovascular diseases: bibliometric and meta-analysis

Significance

Since the advent of Optical Coherence Tomography (OCT) two decades ago, there has been substantial advancement in our understanding of intravascular biology. Identifying culprit lesion pathology through OCT could precipitate a paradigm shift in the treatment of patients with Acute Coronary Syndrome. Given the technical prowess of OCT in the realm of cardiology, bibliometric analysis can reveal trends and research focal points in the application of OCT for cardiovascular diseases. Concurrently, meta-analyses provide a more comprehensive evidentiary base, supporting the clinical efficacy of OCT-guided Percutaneous Coronary Intervention (PCI).

Design

This study employs a dual approach of Bibliometric and Meta-analysis.

Methods

Relevant literature from 2003 to 2023 was extracted from the Web of Science Core Collection (WoSCC) and analyzed using VOSviewer, CiteSpace, and R for publication patterns, countries, institutions, authors, and research hotspots. The study compares OCT-guided and coronary angiography-guided PCI in treating adult coronary artery disease through randomized controlled trials (RCTs) and observational studies. The study has been reported in the line with PRISMA and AMSTAR Guidelines.

Results

Adhering to inclusion and exclusion criteria, 310 publications were incorporated, demonstrating a continual rise in annual output. Chinese researchers contributed the most studies, while American research wielded greater influence. Analysis of trends indicated that research on OCT and angiography-guided PCI has become a focal topic in recent cohort studies and RCTs. In 11 RCTs (n = 5,277), OCT-guided PCI was not significantly associated with a reduction in the risk of Major Adverse Cardiac Events (MACE) (Odds ratio 0.84, 95% CI 0.65-1.10), cardiac death (0.61, 0.36-1.02), all-cause death (0.7, 0.49-1.02), myocardial infarction (MI) (0.88, 0.69-1.13), target lesion revascularization (TLR) (0.94, 0.7-1.27), target vessel revascularization (TVR) (1.04, 0.76-1.43), or stent thrombosis (0.72, 0.38-1.38). However, in 7 observational studies (n = 4,514), OCT-guided PCI was associated with a reduced risk of MACE (0.66, 0.48-0.91) and TLR (0.39, 0.22-0.68).

Conclusion

Our comprehensive review of OCT in cardiovascular disease literature from 2004 to 2023, encompassing country and institutional origins, authors, and publishing journals, suggests that OCT-guided PCI does not demonstrate significant clinical benefits in RCTs. Nevertheless, pooled results from observational studies indicate a reduction in MACE and TLR.

Economic evaluation of the effect of needle and syringe programs on skin, soft tissue, and vascular infections in people who inject drugs: a microsimulation modelling approach

BACKGROUND

Needle and syringe programs (NSP) are effective harm-reduction strategies against HIV and hepatitis C. Although skin, soft tissue, and vascular infections (SSTVI) are the most common morbidities in people who inject drugs (PWID), the extent to which NSP are clinically and cost-effective in relation to SSTVI in PWID remains unclear. The objective of this study was to model the clinical- and cost-effectiveness of NSP with respect to treatment of SSTVI in PWID.

METHODS

We performed a model-based, economic evaluation comparing a scenario with NSP to a scenario without NSP. We developed a microsimulation model to generate two cohorts of 100,000 individuals corresponding to each NSP scenario and estimated quality-adjusted life-years (QALY) and cost (in 2022 Canadian dollars) over a 5-year time horizon (1.5% per annum for costs and outcomes). To assess the clinical effectiveness of NSP, we conducted survival analysis that accounted for the recurrent use of health care services for treating SSTVI and SSTVI mortality in the presence of competing risks.

RESULTS

The incremental cost-effectiveness ratio associated with NSP was $70,278 per QALY, with incremental cost and QALY gains corresponding to $1207 and 0.017 QALY, respectively. Under the scenario with NSP, there were 788 fewer SSTVI deaths per 100,000 PWID, corresponding to 24% lower relative hazard of mortality from SSTVI (hazard ratio [HR] = 0.76; 95% confidence interval [CI] = 0.72-0.80). Health service utilization over the 5-year period remained lower under the scenario with NSP (outpatient: 66,511 vs. 86,879; emergency department: 9920 vs. 12,922; inpatient: 4282 vs. 5596). Relatedly, having NSP was associated with a modest reduction in the relative hazard of recurrent outpatient visits (HR = 0.96; 95% CI = 0.95-0.97) for purulent SSTVI as well as outpatient (HR = 0.88; 95% CI = 0.87-0.88) and emergency department visits (HR = 0.98; 95% CI = 0.97-0.99) for non-purulent SSTVI.

CONCLUSIONS

Both the individuals and the healthcare system benefit from NSP through lower risk of SSTVI mortality and prevention of recurrent outpatient and emergency department visits to treat SSTVI. The microsimulation framework provides insights into clinical and economic implications of NSP, which can serve as valuable evidence that can aid decision-making in expansion of NSP services.

Stereotactic radiosurgery for haemorrhagic cerebral cavernous malformation: a multi-institutional, retrospective study

BACKGROUND

Cerebral cavernous malformations (CCMs) frequently manifest with haemorrhages. Stereotactic radiosurgery (SRS) has been employed for CCM not suitable for resection. Its effect on reducing haemorrhage risk is still controversial. The aim of this study was to expand on the safety and efficacy of SRS for haemorrhagic CCM.

METHODS

This retrospective multicentric study included CCM with at least one haemorrhage treated with single-session SRS. The annual haemorrhagic rate (AHR) was calculated before and after SRS. Recurrent event analysis and Cox regression were used to evaluate factors associated with haemorrhage. Adverse radiation effects (AREs) and occurrence of new neurological deficits were recorded.

RESULTS

The study included 381 patients (median age: 37.5 years (Q1-Q3: 25.8-51.9) with 414 CCMs. The AHR from diagnosis to SRS excluding the first haemorrhage was 11.08 per 100 CCM-years and was reduced to 2.7 per 100 CCM-years after treatment. In recurrent event analysis, SRS, HR 0.27 (95% CI 0.17 to 0.44), p<0.0001 was associated with a decreased risk of haemorrhage, and the presence of developmental venous anomaly (DVA) with an increased risk, HR 1.60 (95% CI 1.07 to 2.40), p=0.022. The cumulative risk of first haemorrhage after SRS was 9.4% (95% CI 6% to 12.6%) at 5 years and 15.6% (95% CI% 9 to 21.8%) at 10 years. Margin doses> 13 Gy, HR 2.27 (95% CI 1.20 to 4.32), p=0.012 and the presence of DVA, HR 2.08 (95% CI 1.00 to 4.31), p=0.049 were factors associated with higher probability of post-SRS haemorrhage. Post-SRS haemorrhage was symptomatic in 22 out of 381 (5.8%) patients, presenting with transient (15/381) or permanent (7/381) neurological deficit. ARE occurred in 11.1% (46/414) CCM and was responsible for transient neurological deficit in 3.9% (15/381) of the patients and permanent deficit in 1.1% (4/381) of the patients. Margin doses >13 Gy and CCM volume >0.7 cc were associated with increased risk of ARE.

CONCLUSION

Single-session SRS for haemorrhagic CCM is associated with a decrease in haemorrhage rate. Margin doses ≤13 Gy seem advisable.

Outcomes After Pediatric Pulmonary Valve Replacement in Patients With Tetralogy of Fallot

BACKGROUND

People with tetralogy of Fallot might require a transannular patch during primary repair, which leads to pulmonary valve regurgitation. Pulmonary valve replacement (PVR) is performed to prevent complications of chronic pulmonary regurgitation, but the optimal timing of PVR remains a matter of debate. In this study we aimed to assess the association of PVR performed in children younger than 18 years of age with the rate of hospitalizations, interventions, and mortality.

METHODS

This was a retrospective observational cohort of people with tetralogy of Fallot born in Québec between 1982 and 2015, using a combination of clinical and administrative data. Marginal means and rates models and survival curves were used to compare outcomes of patients with pediatric PVR (younger than 18 years) and those without. Outcomes of interest were rates of cardiac hospitalizations, all-cause hospitalizations, cardiac interventions, and mortality. Groups were balanced using models weighed on the inverse probability of receiving pediatric PVR.

RESULTS

Of the 316 eligible patients, 58 (18.4%) received a pediatric PVR. Compared with patients who did not receive pediatric PVR, they were at increased risk of cardiac hospitalizations, although the rates of cardiac hospitalization were low: 0.50 vs 0.09 hospitalizations per 20 years (hazard ratio [HR] = 4.71 [95% confidence interval (CI), 2.22-9.96]). Patients who received a pediatric PVR had a comparable risk of all-cause hospitalizations (HR = 0.95 [95% CI, 0.71-1.26]) and of cardiac interventions (HR = 1.13 [95% CI, 0.72-1.77]).

CONCLUSIONS

Patients who underwent pediatric PVR had higher rates of cardiac hospitalizations, but similar rates of all-cause hospitalizations, cardiac procedures, and mortality. In this observational cohort, pediatric PVR was not associated with an improved outcome.

Distributional imputation for the analysis of censored recurrent events

Longitudinal clinical trials for which recurrent events endpoints are of interest are commonly subject to missing event data. Primary analyses in such trials are often performed assuming events are missing at random, and sensitivity analyses are necessary to assess robustness of primary analysis conclusions to missing data assumptions. Control-based imputation is an attractive approach in superiority trials for imposing conservative assumptions on how data may be missing not at random. A popular approach to implementing control-based assumptions for recurrent events is multiple imputation (MI), but Rubin's variance estimator is often biased for the true sampling variability of the point estimator in the control-based setting. We propose distributional imputation (DI) with corresponding wild bootstrap variance estimation procedure for control-based sensitivity analyses of recurrent events. We apply control-based DI to a type I diabetes trial. In the application and simulation studies, DI produced more reasonable standard error estimates than MI with Rubin's combining rules in control-based sensitivity analyses of recurrent events.

Psychiatric disorders in childhood cancer survivors: A retrospective matched cohort study of inpatient hospitalisations and community-based mental health services utilisation in Western Australia

OBJECTIVE

We examined the impact of long-term mental health outcomes on healthcare services utilisation among childhood cancer survivors in Western Australia using linked hospitalisations and community-based mental healthcare records from 1987 to 2019.

METHOD

The study cohort included 2977 childhood cancer survivors diagnosed with cancer at age < 18 years in Western Australia from 1982 to 2014 and a matched non-cancer control group of 24,994 individuals. Adjusted hazard ratios of recurrent events were estimated using the Andersen-Gill model. The cumulative burden of events over time was assessed using the method of mean cumulative count. The annual percentage change in events was estimated using the negative binomial regression model.

RESULTS

The results showed higher community-based service contacts (rate/100 person-years: 30.2, 95% confidence interval = [29.7-30.7] vs 22.8, 95% confidence interval = [22.6-22.9]) and hospitalisations (rate/1000 person-years: 14.8, 95% confidence interval = [13.6-16.0] vs 12.7, 95% confidence interval = [12.3-13.1]) in childhood cancer survivors compared to the control group. Childhood cancer survivors had a significantly higher risk of any event (adjusted hazard ratio = 1.5, 95% confidence interval = [1.1-2.0]). The cumulative burden of events increased with time since diagnosis and across age groups. The annual percentage change for hospitalisations and service contacts significantly increased over time (p < 0.05). Substance abuse was the leading cause of hospitalisations, while mood/affective and anxiety disorders were common causes of service contacts. Risk factors associated with increased service events included cancer diagnosis at age < 5 years, leukaemia diagnosis, high socioeconomic deprivation, and an attained age of < 18 years.

CONCLUSIONS

The elevated utilisation of healthcare services observed among childhood cancer survivors emphasises the need for periodic assessment of psychiatric disorders, particularly in high-risk survivors, to facilitate early management and optimise healthcare resources.

Gut microbiome correlates of recurrent urinary tract infection: a longitudinal, multi-center study

Background

Urinary tract infections (UTI) affect approximately 250 million people annually worldwide. Patients often experience a cycle of antimicrobial treatment and recurrent UTI (rUTI) that is thought to be facilitated by a gut reservoir of uropathogenic Escherichia coli (UPEC).

Methods

125 patients with UTI caused by an antibiotic-resistant organism (ARO) were enrolled from July 2016 to May 2019 in a longitudinal, multi-center cohort study. Multivariate statistical models were used to assess the relationship between uropathogen colonization and recurrent UTI (rUTI), controlling for clinical characteristics. 644 stool samples and 895 UPEC isolates were interrogated for taxonomic composition, antimicrobial resistance genes, and phenotypic resistance. Cohort UTI gut microbiome profiles were compared against published healthy and UTI reference microbiomes, as well as assessed within-cohort for timepoint- and recurrence-specific differences.

Findings

Risk of rUTI was not independently associated with clinical characteristics. The UTI gut microbiome was distinct from healthy reference microbiomes in both taxonomic composition and antimicrobial resistance gene (ARG) burden, with 11 differentially abundant taxa at the genus level. rUTI and non-rUTI gut microbiomes in the cohort did not generally differ, but gut microbiomes from urinary tract colonized patients were elevated in E. coli abundance 7-14 days post-antimicrobial treatment. Corresponding UPEC gut isolates from urinary tract colonizing lineages showed elevated phenotypic resistance against 11 of 23 tested drugs compared to non-colonizing lineages.

Interpretation

The gut microbiome is implicated in UPEC urinary tract colonization during rUTI, serving as an ARG-enriched reservoir for UPEC. UPEC can asymptomatically colonize the gut and urinary tract, and post-antimicrobial blooms of gut E. coli among urinary tract colonized patients suggest that cross-habitat migration of UPEC is an important mechanism of rUTI. Thus, treatment duration and UPEC populations in both the urinary and gastrointestinal tract should be considered in treating rUTI and developing novel therapeutics.

Funding

This work was supported in part by awards from the U.S. Centers for Disease Control and Prevention Epicenter Prevention Program (grant U54CK000482; principal investigator, V.J.F.); to J.H.K. from the Longer Life Foundation (an RGA/Washington University partnership), the National Center for Advancing Translational Sciences (grants KL2TR002346 and UL1TR002345), and the National Institute of Allergy and Infectious Diseases (NIAID) (grant K23A1137321) of the National Institutes of Health (NIH); and to G.D. from NIAID (grant R01AI123394) and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (grant R01HD092414) of NIH. R.T.'s research was funded by the Deutsche Forschungsgemeinschaft (DFG; German Research Foundation; grant 402733540). REDCap is Supported by Clinical and Translational Science Award (CTSA) Grant UL1 TR002345 and Siteman Comprehensive Cancer Center and NCI Cancer Center Support Grant P30 CA091842. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies.

A Retrospective Cohort Analysis of Mental Health-Related Emergency Department Visits Among Veterans and Non-Veterans Residing in Ontario, Canada: Une analyse de cohorte rétrospective des visites au service d'urgence liées à la santé mentale parmi les vétérans et non-vétérans résidant en Ontario, Canada

OBJECTIVES

Emergency departments (EDs) are a vital part of healthcare systems, at times acting as a gateway to community-based mental health (MH) services. This may be particularly true for veterans of the Royal Canadian Mounted Police who were released prior to 2013 and the Canadian Armed Forces, as these individuals transition from federal to provincial healthcare coverage on release and may use EDs because of delays in obtaining a primary care provider. We aimed to estimate the hazard ratio (HR) of MH-related ED visits between veterans and non-veterans residing in Ontario, Canada: (1) overall; and by (2) sex; and (3) length of service.

METHODS

This retrospective cohort study used administrative healthcare data from 18,837 veterans and 75,348 age-, sex-, geography-, and income-matched non-veterans residing in Ontario, Canada between April 1, 2002, and March 31, 2020. Anderson-Gill regression models were used to estimate the HR of recurrent MH-related ED visits during the period of follow-up. Sex and length of service were used as stratification variables in the models.

RESULTS

Veterans had a higher adjusted HR (aHR) of MH-related ED visits than non-veterans (aHR, 1.97, 95% CI, 1.70 to 2.29). A stronger effect was observed among females (aHR, 3.29; 95% CI, 1.96 to 5.53) than males (aHR, 1.78; 95% CI, 1.57 to 2.01). Veterans who served for 5-9 years had a higher rate of use than non-veterans (aHR, 3.76; 95% CI, 2.34 to 6.02) while veterans who served for 30+ years had a lower rate compared to non-veterans (aHR, 0.60; 95% CI, 0.42 à 0.84).

CONCLUSIONS

Rates of MH-related ED visits are higher among veterans overall compared to members of the Ontario general population, but usage is influenced by sex and length of service. These findings indicate that certain subpopulations of veterans, including females and those with fewer years of service, may have greater acute mental healthcare needs and/or reduced access to primary mental healthcare.

Multiple Sclerosis, Rituximab, Hypogammaglobulinemia, and Risk of Infections

BACKGROUND AND OBJECTIVES

B-cell-depleting therapies increase the risk of infections and hypogammaglobulinemia. These relationships are poorly understood. The objectives of these analyses were to estimate how much of this rituximab-associated infection risk is mediated by hypogammaglobulinemia and to identify other modifiable risk factors in persons with multiple sclerosis (pwMS).

METHODS

We conducted a retrospective cohort study of rituximab-treated pwMS from January 1, 2008, to December 31, 2020, in Kaiser Permanente Southern California. Cumulative rituximab dose was defined as ≤2, >2 and ≤4, or >4 g. Serious infections were defined as infections requiring or prolonging hospitalizations, and recurrent outpatient infections as seeking care for ≥3 within 12 months. Exposures, outcomes, and covariates were collected from the electronic health record. Adjusted hazard ratios (aHRs) were estimated using Andersen-Gill hazards models, and generalized estimating equations were used to examine correlates of IgG values. Cross-sectional causal mediation analyses of rituximab and hypogammaglobulinemia were conducted.

RESULTS

We identified 2,482 pwMS who were treated with rituximab for a median of 2.4 years (interquartile range = 1.3-3.9). The average age at rituximab initiation was 43.0 years, 71.9% were female, 49.7% were White, non-Hispanic patients, and 29.6% had advanced disability (requiring walker or worse). Seven hundred patients (28.2%) developed recurrent outpatient infections, 155 (6.2%) developed serious infections, and only 248 (10.0%) had immunoglobulin G (IgG) < 700 mg/dL. Higher cumulative rituximab dose (>4 g) was correlated with lower IgG levels (Beta = -58.8, p < 0.0001, ref ≤2 g) and, in models mutually adjusted for hypogammaglobulinemia, both were independently associated with an increased risk of serious (>4 g, aHR = 1.56, 95% CI 1.09-2.24; IgG < 500, aHR = 2.98, 95% CI 1.56-5.72) and outpatient infections (>4 g, aHR = 1.73, 95% CI 1.44-2.06; IgG < 500 aHR = 2.06, 95% CI 1.52-2.80; ref = IgG ≥ 700). Hypogammaglobulinemia explained at most 17.9% (95% CI -47.2-119%) of serious infection risk associated with higher cumulative rituximab exposure but was not significant for outpatient infections. Other independent modifiable risk factors were advanced physical disability for serious (aHR = 5.51, 95% CI 3.71-8.18) and outpatient infections (aHR = 1.24, 95% CI 1.06-1.44) and COPD (aHR = 1.68, 95% CI 1.34-2.11) and obesity (aHR = 1.25, 95% CI 1.09-1.45) for outpatient infections.

DISCUSSION

Higher cumulative rituximab doses increase the risk of infections even in this population where 90% of patients maintained normal IgG levels. Clinicians should strive to use minimally effective doses of rituximab and other B-cell-depleting therapies and consider important comorbidities to minimize risks of infections.

Treatment patterns and persistence on disease modifying therapies for multiple sclerosis and its associated factors

BACKGROUND

Effective interventions for Multiple Sclerosis require timely treatment optimization which usually involves switching disease modifying therapies. The patterns of prescription and the reasons for changing treatment in people with MS, especially in low prevalence populations, are unknown.

OBJECTIVES

To describe the persistence, reasons of DMT switches and prescription patterns in a cohort of Colombian people with MS.

METHODS

We conducted a retrospective observational study including patients with confirmed MS with at least one visit at our centre. We estimated the overall incidence rate of medication changes and assessed the persistence on medication with Kaplan-Meier survival estimates for individual medications and according to efficacy and mode of administration. The factors associated with changing medications were assessed using adjusted Cox proportional-hazards models. The reasons for switching medication changes were described, and the prescription patterns were assessed using network analysis, with measures of centrality.

RESULTS

Seven hundred one patients with MS were included. Mean age was 44.3 years, and 67.9% were female. Mean disease duration was 11.3 years and 84.5% had relapsing MS at onset, with median EDSS of 1.0. Treatment was started in 659 (94%) of the patients after a mean of 3 years after MS symptom onset. Among them, 39.5% maintained their initial DMT, 29.9% experienced a single DMT change, while 18.7% went through two, and 11.9% had three or more DMT changes until the final follow-up. The total number of treatment modifications reached 720, resulting in an incidence rate of 1.09 (95% confidence interval: 1.01-1.17) per patient per year The median time to change after the first DMT was 3.75 years, and was not different according to the mode of administration or efficacy classification. The main reasons for changing DMT were MS activity (relapses, 56.7%; MRI activity, 18.6%), followed by non-serious adverse events (15.3%) and disability (11.1%). Younger age at MS onset, care under our centre and insurer status were the main determinants of treatment change. Network analysis showed that interferons and fingolimod were the most influential DMTs.

CONCLUSIONS

A majority of patients switch medications, mostly due to disease activity, and in association with age and insurer status.

Determinants and Clinical Outcomes of Patients With Tetralogy of Fallot Lost to Cardiology Follow-up

BACKGROUND

Various rates of loss to follow-up (LTFU) have been reported in patients with congenital heart disease, but return to follow-up is rarely considered in those analyses. Outcomes of LTFU patients are difficult to assess because the patients no longer attend cardiac care. We leveraged data from the TRIVIA cohort, which combines more than 30 years of clinical and administrative data, allowing us to study outcomes even after LTFU.

METHODS

This population-based cohort included 904 patients with tetralogy of Fallot (TOF) born from 1982 to 2015 in Québec, Canada. Risk factors for LTFU and outcomes were calculated by Cox models and marginal means/rates models. Outcomes of LTFU patients were compared with propensity score-matched non-LTFU patients.

RESULTS

The cumulative risk of experiencing 1 episode of LTFU was 50.3% at 30 years. However, return to follow-up was frequent and the proportion of patients actively followed was 85.9% at 10 years, 76.4% at 20 years, and 70.6% at 30 years. Factors associated with a reduced risk of LTFU were primary repair with conduit (hazard ratio [HR] 0.29, 95% confidence interval [CI] 0.15-0.58) and transannular patch (HR 0.60, 95% CI 0.46-0.79). LTFU patients had lower rates of cardiac hospitalisations (HR 0.49, 95% CI 0.42-0.56) and cardiac interventions (HR 0.32, 95% CI 0.25-0.42), but similar rates of cardiac mortality (HR 0.95, 95% CI 0.24-3.80).

CONCLUSIONS

There was a lower proportion of LTFU patients compared with previous studies. Factors associated with lower rates of LTFU were conduits and non-valve-sparing surgery. LTFU patients had lower rates of cardiac procedures and cardiac hospitalisations.

An Overview of Regression Models for Adverse Events Analysis

Over the last few years, several review articles described the adverse events analysis as sub-optimal in clinical trials. Indeed, the context surrounding adverse events analyses often imply an overwhelming number of events, a lack of power to find associations, but also a lack of specific training regarding those complex data. In randomized controlled trials or in observational studies, comparing the occurrence of adverse events according to a covariable of interest (e.g., treatment) is a recurrent question in the analysis of drug safety data, and adjusting other important factors is often relevant. This article is an overview of the existing regression models that may be considered to compare adverse events and to discuss model choice regarding the characteristics of the adverse events of interest. Many dimensions may be relevant to compare the adverse events between patients, (e.g., timing, recurrence, and severity). Recent efforts have been made to cover all of them. For chronic treatments, the occurrence of intercurrent events during the patient follow-up usually needs the modeling approach to be adapted (at least with regard to their interpretation). Moreover, analysis based on regression models should not be limited to the estimation of relative effects. Indeed, absolute risks stemming from the model should be presented systematically to help the interpretation, to validate the model, and to encourage comparison of studies.

The Association of Frailty and Neighborhood Disadvantage with Emergency Department Visits and Hospitalizations in Older Adults

BACKGROUND

Risk stratification and population management strategies are critical for providing effective and equitable care for the growing population of older adults in the USA. Both frailty and neighborhood disadvantage are constructs that independently identify populations with higher healthcare utilization and risk of adverse outcomes.

OBJECTIVE

To examine the joint association of these factors on acute healthcare utilization using two pragmatic measures based on structured data available in the electronic health record (EHR).

DESIGN

In this retrospective observational study, we used EHR data to identify patients aged ≥ 65 years at Atrium Health Wake Forest Baptist on January 1, 2019, who were attributed to affiliated Accountable Care Organizations. Frailty was categorized through an EHR-derived electronic Frailty Index (eFI), while neighborhood disadvantage was quantified through linkage to the area deprivation index (ADI). We used a recurrent time-to-event model within a Cox proportional hazards framework to examine the joint association of eFI and ADI categories with healthcare utilization comprising emergency visits, observation stays, and inpatient hospitalizations over one year of follow-up.

KEY RESULTS

We identified a cohort of 47,566 older adults (median age = 73, 60% female, 12% Black). There was an interaction between frailty and area disadvantage (P = 0.023). Each factor was associated with utilization across categories of the other. The magnitude of frailty's association was larger than living in a disadvantaged area. The highest-risk group comprised frail adults living in areas of high disadvantage (HR 3.23, 95% CI 2.99-3.49; P < 0.001). We observed additive effects between frailty and living in areas of mid- (RERI 0.29; 95% CI 0.13-0.45; P < 0.001) and high (RERI 0.62, 95% CI 0.41-0.83; P < 0.001) neighborhood disadvantage.

CONCLUSIONS

Considering both frailty and neighborhood disadvantage may assist healthcare organizations in effectively risk-stratifying vulnerable older adults and informing population management strategies. These constructs can be readily assessed at-scale using routinely collected structured EHR data.

Risk Factors and Inequities in Transportation Injury and Mortality in the Canadian Census Health and Environment Cohorts (CanCHECs)

BACKGROUND

Road traffic injury contributes substantially to morbidity and mortality. Canada stands out among developed countries in not conducting a national household travel survey, leading to a dearth of national transportation mode data and risk calculations that have appropriate denominators. Since traffic injuries are specific to the mode of travel used, these risk calculations should consider travel mode.

METHODS

Census data on mode of commute is one of the few sources of these data for persons aged 15 and over. This study leveraged a national data linkage cohort, the Canadian Census Health and Environment Cohorts, that connects census sociodemographic and commute mode data with records of deaths and hospitalizations, enabling assessment of road traffic injury associations by indicators of mode of travel (commuter mode). We examined longitudinal (1996-2019) bicyclist, pedestrian, and motor vehicle occupant injury and fatality risk in the Canadian Census Health and Environment Cohorts by commuter mode and sociodemographic characteristics using Cox proportional hazards models within the working adult population.

RESULTS

We estimated positive associations between commute mode and same mode injury and fatality, particularly for bicycle commuters (hazard ratios for bicycling injury was 9.1 and for bicycling fatality was 11). Low-income populations and Indigenous people had increased injury risk across all modes.

CONCLUSIONS

This study shows inequities in transportation injury risk in Canada and underscores the importance of adjusting for mode of travel when examining differences between population groups.

Association between sodium-glucose cotransporter-2 inhibitors and arrhythmic outcomes in patients with diabetes and pre-existing atrial fibrillation

AIMS

Prior studies suggest that sodium-glucose cotransporter-2 inhibitors (SGLT2is) may decrease the incidence of atrial fibrillation (AF). However, it is unknown whether SGLT2i can attenuate the disease course of AF among patients with pre-existing AF and Type II diabetes mellitus (DM). In this study, our objective was to examine the association between SGLT2i prescription and arrhythmic outcomes among patients with DM and pre-existing AF.

METHODS AND RESULTS

We conducted a population-based cohort study of adults with DM and AF between 2014 and 2019. Using a prevalent new-user design, individuals prescribed SGLT2i were matched 1:1 to those prescribed dipeptidyl peptidase-4 inhibitors (DPP4is) based on time-conditional propensity scores. The primary endpoint was a composite of AF-related healthcare utilization (i.e. hospitalization, emergency department visits, electrical cardioversion, or catheter ablation). Secondary outcome measures included all-cause mortality, heart failure (HF) hospitalization, and ischaemic stroke or transient ischaemic attack (TIA). Cox proportional hazard models were used to examine the association of SGLT2i with the study endpoint. Among 2242 patients with DM and AF followed for an average of 3.0 years, the primary endpoint occurred in 8.7% (n = 97) of patients in the SGLT2i group vs. 10.0% (n = 112) of patients in the DPP4i group [adjusted hazard ratio 0.73 (95% confidence interval 0.55-0.96; P = 0.03)]. Sodium-glucose cotransporter-2 inhibitors were associated with significant reductions in all-cause mortality and HF hospitalization, but there was no difference in the risk of ischaemic stroke/TIA.

CONCLUSION

Among patients with DM and pre-existing AF, SGLT2is are associated with decreased AF-related health resource utilization and improved arrhythmic outcomes compared with DPP4is.

Infant Whole-Cell Versus Acellular Pertussis Vaccination in 1997 to 1999 and Risk of Childhood Hospitalization for Food-Induced Anaphylaxis: Linked Administrative Databases Cohort Study

BACKGROUND

Evidence suggests that children who had received an initial priming dose of whole-cell pertussis (wP) vaccine, rather than acellular pertussis (aP) vaccine, had a lower risk of developing IgE-mediated food allergy, the most common cause of anaphylaxis-related hospital presentations of childhood.

OBJECTIVE

To assess the association between wP versus aP vaccination in infancy and subsequent hospital presentations for anaphylaxis.

METHODS

This study was preregistered under PMID 34874968. Perinatal records for a cohort of New South Wales-born children (1997-1999) receiving their first dose of pertussis-containing vaccine before age 4 months were probabilistically linked to hospital and immunization records. We used adjusted Cox models to estimate hazard ratios (aHRs) and 95% CIs for anaphylaxis-coded hospitalizations.

RESULTS

There were 218,093 New South Wales-born children who received a first dose of wP or aP before age 4 months. Among these children, 86 experienced at least one hospitalization for food-induced anaphylaxis at age 5-15 years (range of events per patient, one to three). The person-time of follow-up was 1,476,969 years, and 665,519 years for children vaccinated with wP as a first dose (wP-1 children) and aP as a first dose (aP-1 children), respectively. The incidence rates for first hospitalization for food anaphylaxis were 3.5 (95% CI, 2.6-4.6) and 5.1 (95% CI, 3.5-7.1) per 100,000 child-years among wP-1 children and aP-1 children, respectively (aHR for wP vs aP = 0.47; 95% CI, 0.26-0.83). For first admission for venom anaphylaxis, the incidence rate was 4.9 (95% CI, 3.9-6.2) per 100,000 child-years among wP-1 children and 5.1 (95% CI, 3.5-7.1) per 100,000 child-years among aP-1 children (aHR for wP vs aP = 0.92; 95% CI, 0.53-1.60), and for all-cause anaphylaxis, the incidence rate was 10.6 (95% CI, 9.0-12.4) per 100,000 child-years among wP-1 children and 12.8 (95% CI, 10.2-15.8) per 100,000 child-years among aP-1 children (aHR for wP vs aP = 0.92; 95% CI, 0.53-1.60).

CONCLUSION

Vaccination with wP in infancy was associated with a lower risk of hospitalizations for food-induced anaphylaxis (and therefore severe IgE-mediated food allergy) occurring in childhood.

Patterns of neuromyelitis optica spectrum disorder attacks in different age groups and sexes depending on the status of immunosuppressive therapy: A retrospective cohort study

BACKGROUND AND PURPOSE

The association between onset age and sex with relapse risk in neuromyelitis optica spectrum disorder (NMOSD) remains inconclusive. We aimed to describe the clinical features of patients with NMOSD in different age groups and sexes and to analyse relapse characteristics pre- and post-immunosuppressive therapy (IST).

METHODS

Patients with NMOSD were retrospectively reviewed from our clinical centre's database. Demographic and clinical data, attack presentation, and disease course pre- and post-IST were investigated. We also analysed the effect of onset age on the annualized relapse rate and relapse risk according to sex and IST status. Interactions on the additive scale between onset age and sex were analysed. A restricted cubic spline was used to analyse potential nonlinear correlations. Longitudinal changes in the Expanded Disability Status Scale score across NMOSD attacks were analysed using linear mixed-effect models.

RESULTS

In total, 533 patients experienced 1394 attacks pre-IST and 753 relapses post-IST. Older age at onset was correlated with more myelitis attacks but fewer optic neuritis attacks, with no sex-related differences in attack presentation. Pre-IST, relapse risk increased with age at onset in women, while a U-shaped correlation between onset age and relapse risk was found in men. Post-IST, an inverted U-shaped association between the predicted relapse risk and onset age was observed in women. Conversely, a negative correlation between the predicted relapse risk and onset age was found in men. Overall, a higher ratio of myelitis attacks was found post-IST.

CONCLUSIONS

Patients of different onset ages and sexes had different relapse patterns before and after IST.

ADHD medications use and risk of mortality and unintentional injuries: a population-based cohort study

We assessed the association between the use of medications for attention-deficit/hyperactivity disorder (ADHD) and the risk of all-cause mortality and unintentional injuries leading to emergency department (ED) or hospital admission in individuals aged ≤24 years with ADHD. We conducted a population-based retrospective cohort study between 2000 and 2021 using Quebec health administrative data. Individuals were followed from the first ADHD diagnosis or ADHD medication claim until turning 25, death, or study end. Exposure was defined as mutually exclusive episodes of ADHD medication use and/or coverage under the public provincial drug plan (PDP): 1) covered and not treated with ADHD medication; 2) covered and treated with ADHD medication; and 3) not covered under the PDP. The risk of all-cause mortality and unintentional injuries associated with exposure episodes was estimated using multivariable survival analyses. The cohort included n = 217 192 individuals aged 1-24 years with a male to female ratio of close to 2:1. Compared to non-medication use, episodes of ADHD medication use, overall, were associated with reduced all-cause mortality (adjusted hazard ratio, aHR 0.61, 95% CI 0.48-0.76) and unintentional injury leading to ED (0.75, 0.74-0.77) or hospitalisation (0.71, 0.68-0.75). Episodes of stimulants were associated with a lower risk of all-cause mortality and reduced risk of unintentional injuries, while episodes with non-stimulants and with both stimulants and non-stimulants concomitantly were associated with reduced risk of unintentional injuries, but not of all-cause mortality. Although residual confounding cannot be excluded, stimulants may have a protective effect in terms of risk of all-cause mortality and both stimulants and non-stimulants for ADHD may reduce the risk of unintentional injuries. The findings of the current study should inform clinical decision making on the choice of starting a pharmacological treatment for ADHD, when a balance needs to be struck between expected benefits and possible risks.

Associations between air pollution and the risk of first admission and multiple readmissions for cardiovascular diseases

OBJECTIVES

We aimed to investigate the associations between air pollutants and the risk of admission and multiple readmission events for cardiovascular disease (CVD).

METHODS

A total of 285 009 participants free of CVD at baseline from the UK Biobank were included in this analysis. Four major cardiovascular admission events were identified during the follow-up: chronic ischaemic heart disease (CIHD), cerebrovascular disease, atrial fibrillation and heart failure. We used Prentice, Williams and Peterson-Total Time model to examine the association between ambient air pollution and first admission, as well as multiple readmissions for these CVDs.

RESULTS

During a median follow-up of 12 years, 17 176 (6.03%) participants were hospitalised with CVDs, and 6203 (36.11%) patients with CVD had subsequent readmission events for CVDs. We observed significant associations between air pollution and both first admission and readmission for CVDs, with generally stronger associations on readmission for cardiovascular events. For example, the adjusted HRs for the first admission and subsequent readmission for cerebrovascular disease were 1.130 (95% CI 1.070 to 1.194) and 1.270 (95% CI 1.137 to 1.418) for each IQR increase of particulate matter with a diameter ≤2.5 µm. The corresponding HRs for CIHD were 1.060 (95% CI 1.008 to 1.114) and 1.120 (95% CI 1.070 to 1.171). Sex stratified analyses showed that the associations were generally more pronounced among females than males.

CONCLUSION

This study provides evidence that ambient air pollutants might play an important role in both first admission and readmission for cardiovascular events. In addition, patients with pre-existing CVDs may be more vulnerable to air pollution compared with healthy population.

Sample size estimation for recurrent event data using multifrailty and multilevel survival models

In epidemiology and clinical research, recurrent events refer to individuals who are likely to experience transient clinical events repeatedly over an observation period. Examples include hospitalizations in patients with heart failure, fractures in osteoporosis studies and the occurrence of new lesions in oncology. We provided an in-depth analysis of the sample size required for the analysis of recurrent time-to-event data using multifrailty or multilevel survival models. We covered the topic from the simple shared frailty model to models with hierarchical or joint frailties. We relied on a Wald-type test statistic to estimate the sample size assuming either a single or multiple endpoints. Simulations revealed that the sample size increased as heterogeneity increased. We also observed that it was more attractive to include more patients and reduce the duration of follow-up than to include fewer patients and increase the duration of follow-up to obtain the number of events required. Each model investigated can address the question of the number of subjects for recurrent events. However, depending on the research question, one model will be more suitable than another. We illustrated our methodology with the AFFIRM-AHF trial investigating the effect of intravenous ferric carboxymaltose in patients hospitalised for acute heart failure.

Association of a calcium channel blocker and diuretic prescribing cascade with adverse events: A population-based cohort study

BACKGROUND

Prescribing cascades occur when a drug adverse event is misinterpreted as a new medical condition and a second, potentially unnecessary drug, is prescribed to treat the adverse event. The population-level consequences of prescribing cascades remain unknown.

METHODS

This population-based cohort study used linked health administrative databases in Ontario, Canada. The study included community-dwelling adults, 66 years of age or older with hypertension and no history of heart failure (HF) or diuretic use in the prior year, newly dispensed a calcium channel blocker (CCB). Individuals subsequently dispensed a diuretic within 90 days of incident CCB dispensing were classified as the prescribing cascade group, and compared to those not dispensed a diuretic, classified as the non-prescribing cascade group. Those with and without a prescribing cascade were matched one-to-one on the propensity score and sex. The primary outcome was a serious adverse event (SAE), which was the composite of emergency room visits and hospitalizations in the 90-day follow-up period. We estimated hazard ratios (HRs) with 95% confidence intervals (CI) for SAE using an Andersen-Gill recurrent events regression model.

RESULTS

Among 39,347 older adults with hypertension and no history of HF who were newly dispensed a CCB, 1881 (4.8%) had a new diuretic dispensed within 90 days after CCB initiation. Compared to the non-prescribing cascade group, those in the prescribing cascade group had higher rates of SAEs (HR: 1.21, 95% CI: 1.02-1.43).

CONCLUSIONS

The CCB-diuretic prescribing cascade was associated with an increased rate of SAEs, suggesting harm beyond prescribing a second drug therapy. Our study raises awareness of the downstream impact of the CCB-diuretic prescribing cascade at a population level and provides an opportunity for clinicians who identify this prescribing cascade to review their patients' medications to determine if they can be optimized.

The Natural History of Hemorrhage in Brain Arteriovenous Malformations-Poisson Regression Analysis of 1066 Patients in a Single Institution

BACKGROUND AND OBJECTIVES

Natural history of hemorrhage in brain arteriovenous malformations (bAVM) is reported at 2%-4% per year. Published studies using survival analysis fail to account for recurrent hemorrhagic events. In this study, we present a large, single institution series to elucidate the natural history of bAVM using multivariable Poisson regression.

METHODS

This is a retrospective cohort study. All patients with bAVM seen at our institution from 1990 to 2021 were included. Hemorrhages after detection of bAVM during the untreated interval were recorded. Natural history of hemorrhage was calculated by dividing number of hemorrhages by untreated interval. The frequency of hemorrhages followed a Poisson distribution. Multivariable Poisson regression with an offset variable of untreated interval in patient-years was constructed. Model selection was through a stepwise Akaike information criterion method. Stratified hemorrhagic rate was presented using different combinations of significant factors.

RESULTS

A total of 1066 patients with nonhereditary hemorrhagic telangiectasia harboring a single bAVM were included. Ninety (8.44%) patients had 101 hemorrhages during an untreated interval of 3596.3344 patient-years, translating to an overall hemorrhagic rate of 2.81% per year. Significant factors increasing hemorrhage risk included ruptured presentation ( P < .001), increasing age ( P < .001), female sex ( P = .043), and deep location ( P = .040). Adult male patients with ruptured presentation and deep bAVMs sustained the highest annual risk at 10.81%, whereas no hemorrhages occurred in unruptured pediatric male patients or ruptured pediatric (younger than 18 years) male patients with superficial bAVMs.

CONCLUSION

Hemorrhage after bAVM detection occurs in 8.41% of all patients, and the rate averages 2.81% per year. However, this risk varies from 0.00% to 10.81% per year depending on various risk factor combinations. Efforts should be made to stratify bAVM hemorrhage rate by risk factors for more precise estimation of bleeding risk if left untreated.

Guideline implementation, drug sequencing, and quality of care in heart failure: design and rationale of TITRATE-HF

AIMS

Current heart failure (HF) guidelines recommend to prescribe four drug classes in patients with HF with reduced ejection fraction (HFrEF). A clear challenge exists to adequately implement guideline-directed medical therapy (GDMT) regarding the sequencing of drugs and timely reaching target dose. It is largely unknown how the paradigm shift from a serial and sequential approach for drug therapy to early parallel application of the four drug classes will be executed in daily clinical practice, as well as the reason clinicians may not adhere to new guidelines. We present the design and rationale for the real-world TITRATE-HF study, which aims to assess sequencing strategies for GDMT initiation, dose titration patterns (order and speed), intolerance for GDMT, barriers for implementation, and long-term outcomes in patients with de novo, chronic, and worsening HF.

METHODS AND RESULTS

A total of 4000 patients with HFrEF, HF with mildly reduced ejection fraction, and HF with improved ejection fraction will be enrolled in >40 Dutch centres with a follow-up of at least 3 years. Data collection will include demographics, physical examination and vital parameters, electrocardiogram, laboratory measurements, echocardiogram, medication, and quality of life. Detailed information on titration steps will be collected for the four GDMT drug classes. Information will include date, primary reason for change, and potential intolerances. The primary clinical endpoints are HF-related hospitalizations, HF-related urgent visits with a need for intravenous diuretics, all-cause mortality, and cardiovascular mortality.

CONCLUSIONS

TITRATE-HF is a real-world multicentre longitudinal registry that will provide unique information on contemporary GDMT implementation, sequencing strategies (order and speed), and prognosis in de novo, worsening, and chronic HF patients.

Cancer Diagnoses After Recent Weight Loss

Importance

Weight loss is common in primary care. Among individuals with recent weight loss, the rates of cancer during the subsequent 12 months are unclear compared with those without recent weight loss.

Objective

To determine the rates of subsequent cancer diagnoses over 12 months among health professionals with weight loss during the prior 2 years compared with those without recent weight loss.

Design, Setting, and Participants

Prospective cohort analysis of females aged 40 years or older from the Nurses' Health Study who were followed up from June 1978 until June 30, 2016, and males aged 40 years or older from the Health Professionals Follow-Up Study who were followed up from January 1988 until January 31, 2016.

Exposure

Recent weight change was calculated from the participant weights that were reported biennially. The intentionality of weight loss was categorized as high if both physical activity and diet quality increased, medium if only 1 increased, and low if neither increased.

Main Outcome and Measures

Rates of cancer diagnosis during the 12 months after weight loss.

Results

Among 157 474 participants (median age, 62 years [IQR, 54-70 years]; 111 912 were female [71.1%]; there were 2631 participants [1.7%] who self-identified as Asian, Native American, or Native Hawaiian; 2678 Black participants [1.7%]; and 149 903 White participants [95.2%]) and during 1.64 million person-years of follow-up, 15 809 incident cancer cases were identified (incident rate, 964 cases/100 000 person-years). During the 12 months after reported weight change, there were 1362 cancer cases/100 000 person-years among all participants with recent weight loss of greater than 10.0% of body weight compared with 869 cancer cases/100 000 person-years among those without recent weight loss (between-group difference, 493 cases/100 000 person-years [95% CI, 391-594 cases/100 000 person-years]; P < .001). Among participants categorized with low intentionality for weight loss, there were 2687 cancer cases/100 000 person-years for those with weight loss of greater than 10.0% of body weight compared with 1220 cancer cases/100 000 person-years for those without recent weight loss (between-group difference, 1467 cases/100 000 person-years [95% CI, 799-2135 cases/100 000 person-years]; P < .001). Cancer of the upper gastrointestinal tract (cancer of the esophagus, stomach, liver, biliary tract, or pancreas) was particularly common among participants with recent weight loss; there were 173 cancer cases/100 000 person-years for those with weight loss of greater than 10.0% of body weight compared with 36 cancer cases/100 000 person-years for those without recent weight loss (between-group difference, 137 cases/100 000 person-years [95% CI, 101-172 cases/100 000 person-years]; P < .001).

Conclusions and Relevance

Health professionals with weight loss within the prior 2 years had a significantly higher risk of cancer during the subsequent 12 months compared with those without recent weight loss. Cancer of the upper gastrointestinal tract was particularly common among participants with recent weight loss compared with those without recent weight loss.

The impact of falls on activities of daily living in older adults: A retrospective cohort analysis

BACKGROUND

Falls contribute to impairments in activities of daily living (ADLs), resulting in significant declines in the quality of life, safety, and functioning of older adults. Understanding the magnitude and duration of the effect of falls on ADLs, as well as identifying the characteristics of older adults more likely to have post-fall ADL impairment is critical to inform fall prevention and post-fall intervention. The purpose of this study is to 1) Quantify the association between falls and post-fall ADL impairment and 2) Model trajectories of ADL impairment pre- and post-fall to estimate the long-term impact of falls and identify characteristics of older adults most likely to have impairment.

METHOD

Study participants were from the Ginkgo Evaluation of Memory Study, a randomized controlled trial in older adults (age 75+) in the United States. Self-reported incident falls and ADL scores were ascertained every 6 months over a 7-year study period. We used Cox proportional hazards analyses (n = 2091) to quantify the association between falls and ADL impairment and latent class trajectory modeling (n = 748) to visualize trajectories of ADL impairment pre-and post-fall.

RESULTS

Falls reported in the previous 6 months were associated with impairment in ADLs (HR: 1.42; 95% CI 1.32, 1.52) in fully adjusted models. Based on trajectory modeling (n = 748), 19% (n = 139) of participants had increased, persistent ADL impairment after falling. Participants who were female, lived in a neighborhood with higher deprivation, or experienced polypharmacy were more likely to have ADL impairment post-fall.

CONCLUSIONS

Falls are associated with increased ADL impairment, and this impairment can persist over time. It is crucial that all older adults, and particularly those at higher risk of post-fall ADL impairment have access to comprehensive fall risk assessment and evidence-based fall prevention interventions, to help mitigate the negative impacts on ADL function.

Sensitivity Analysis for Observational Studies with Recurrent Events

We conduct an observational study of the effect of sickle cell trait Haemoglobin AS (HbAS) on the hazard rate of malaria fevers in children. Assuming no unmeasured confounding, there is strong evidence that HbAS reduces the rate of malarial fevers. Since this is an observational study, however, the no unmeasured confounding assumption is strong. A sensitivity analysis considers how robust a conclusion is to a potential unmeasured confounder. We propose a new sensitivity analysis method for recurrent event data and apply it to the malaria study. We find that for the causal conclusion that HbAS is protective against malarial fevers to be overturned, the hypothesized unmeasured confounder must be as influential as all but one of the measured confounders.

Outcomes of Integrated Home Dialysis Care: Results From the Canadian Organ Replacement Register

RATIONALE & OBJECTIVE

The integrated home dialysis model proposes the initiation of kidney replacement therapy (KRT) with peritoneal dialysis (PD) and a timely transition to home hemodialysis (HHD) after PD ends. We compared the outcomes of patients transitioning from PD to HHD with those initiating KRT with HHD.

STUDY DESIGN

Observational analysis of the Canadian Organ Replacement Register (CORR).

SETTINGS & PARTICIPANTS

All patients who initiated PD or HHD within the first 90 days of KRT between 2005 and 2018.

EXPOSURE

Patients transitioning from PD to HHD (PD+HHD group) versus patients initiating KRT with HHD (HHD group).

OUTCOME

(1) A composite of all-cause mortality and modality transfer (to in-center hemodialysis or PD for 90 days) and (2) all hospitalizations (considered as recurrent events).

ANALYTICAL APPROACH

A propensity score analysis for which PD+HHD patients were matched 1:1 to (1) incident HHD patients ("incident-match" analysis) or (2) HHD patients with a KRT vintage at least equivalent to the vintage of PD+HHD patients at the transition time ("vintage-matched" analysis). Cause-specific hazards models (composite outcome) and shared frailty models (hospitalization) were used to compare groups.

RESULTS

Among 63,327 individuals in the CORR, 163 PD+HHD patients (median of 1.9 years in PD) and 711 HHD patients were identified. In the incident-match analysis, compared to the HHD patients, the PD+HHD group had a similar risk of the composite outcome (HR, 0.88 [95% CI, 0.58-1.32]) and hospitalizations (HR, 1.04 [95% CI, 0.76-1.41]). In the vintage-match analysis, PD+HHD patients had a lower hazard for the composite outcome (HR, 0.61 [95% CI, 0.40-0.94]) but a similar hospitalization risk (HR, 0.85 [95% CI, 0.59-1.24]).

LIMITATIONS

Risk of survivor bias in the PD+HHD cohort and residual confounding.

CONCLUSIONS

Controlling for KRT vintage, the patients transitioning from PD to HHD had better clinical outcomes than the incident HHD patients. These data support the use of integrated home dialysis for patients initiating home-based KRT.

PLAIN-LANGUAGE SUMMARY

The integrated home dialysis model proposes the initiation of dialysis with peritoneal dialysis (PD) and subsequent transition to home hemodialysis (HHD) once PD is no longer feasible. It allows patients to benefit from initial lifestyle advantages of PD and to continue home-based treatments after its termination. However, some patients may prefer to initiate dialysis with HHD from the outset. In this study, we compared the long-term clinical outcomes of both approaches using a large Canadian dialysis register. We found that both options led to a similar risk of hospitalization. In contrast, the PD-to-HHD model led to improved survival when controlling for the duration of kidney failure.

Incidence of sexually transmitted infections and association with behavioural factors: Time-to-event analysis of a large pre-exposure prophylaxis (PrEP) cohort

OBJECTIVES

Our objective was to obtain long-term data on the incidence of sexually transmitted infections (STIs) and their association with behavioural factors after widespread pre-exposure prophylaxis (PrEP) implementation.

METHODS

This was a time-to-event analysis of a national PrEP cohort in Switzerland (SwissPrEPared study). Participants were people without HIV interested in taking PrEP with at least two STI screening visits. Primary outcomes were incidence rate of gonorrhoea, chlamydia, and syphilis. The association between behavioural factors and STI diagnosis was expressed using hazard ratios. We adjusted for testing frequency and calendar year.

RESULTS

This analysis included 3907 participants enrolled between April 2019 and April 2022, yielding 3815.7 person-years of follow-up for gonorrhoea (15 134 screenings), 3802.5 for chlamydia (15 141 screenings), and 3858.6 for syphilis (15 001 screenings). The median age was 39 years (interquartile range [IQR] 32-47), 93.8% (n = 3664) identified as men who have sex with men (MSM). The incidence was 22.8 (95% confidence interval [CI] 21.3-24.4) per 100 person-years for gonorrhoea, 26.3 (95% CI 24.7-28.0) for chlamydia, and 4.4 (95% CI 3.8-5.1) for syphilis. Yearly incidence rates decreased between 2019 (all bacterial STIs: 81.6; 95% CI 59.1-109.9) and 2022 (all bacterial STIs: 49.8; 95% CI 44.6-55.3). Participants reporting chemsex substance use were at higher risk of incident STIs, as were those reporting multiple sexual partners. Younger age was associated with a higher risk of gonorrhoea and chlamydia.

CONCLUSIONS

Incidence rates of bacterial STIs decreased over time. Young MSM, those with multiple partners, and those using chemsex substances were at increased risk of STIs.

Predicting recurrent interventions after radiocephalic arteriovenous fistula creation with machine learning and the PREDICT-AVF web app

OBJECTIVE

Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines discourage ongoing access salvage attempts after two interventions prior to successful use or more than three interventions per year overall. The goal was to develop a tool for prediction of radiocephalic arteriovenous fistula (AVF) intervention requirements to help guide shared decision-making about access appropriateness.

METHODS

Prospective cohort study of 914 adult patients in the United States and Canada undergoing radiocephalic AVF creation at one of the 39 centers participating in the PATENCY-1 or -2 trials. Clinical data, including demographics, comorbidities, access history, anatomic features, and post-operative ultrasound measurements at 4-6 and 12 weeks were used to predict recurrent interventions required at 1 year postoperatively. Cox proportional hazards, random survival forest, pooled logistic, and elastic net recurrent event survival prediction models were built using a combination of baseline characteristics and post-operative ultrasound measurements. A web application was created, which generates patient-specific predictions contextualized with the KDOQI guidelines.

RESULTS

Patients underwent an estimated 1.04 (95% CI 0.94-1.13) interventions in the first year. Mean (SD) age was 57 (13) years; 22% were female. Radiocephalic AVFs were created at the snuffbox (2%), wrist (74%), or proximal forearm (24%). Using baseline characteristics, the random survival forest model performed best, with an area under the receiver operating characteristic curve (AUROC) of 0.75 (95% CI 0.67-0.82) at 1 year. The addition of ultrasound information to baseline characteristics did not substantially improve performance; however, Cox models using either 4-6- or 12-week post-operative ultrasound information alone had the best discrimination performance, with AUROCs of 0.77 (0.70-0.85) and 0.76 (0.70-0.83) at 1 year. The interactive web application is deployed at https://predict-avf.com.

CONCLUSIONS

The PREDICT-AVF web application can guide patient counseling and guideline-concordant shared decision-making as part of a patient-centered end-stage kidney disease life plan.

Comparison of models for stroke-free survival prediction in patients with CADASIL

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy, which is caused by mutations of the NOTCH3 gene, has a large heterogeneous progression, presenting with declines of various clinical scores and occurrences of various clinical event. To help assess disease progression, this work focused on predicting the composite endpoint of stroke-free survival time by comparing the performance of Cox proportional hazards regression to that of machine learning models using one of four feature selection approaches applied to demographic, clinical and magnetic resonance imaging observational data collected from a study cohort of 482 patients. The quality of the modeling process and the predictive performance were evaluated in a nested cross-validation procedure using the time-dependent Brier Score and AUC at 5 years from baseline, the former measuring the overall performance including calibration and the latter highlighting the discrimination ability, with both metrics taking into account the presence of right-censoring. The best model for each metric was the componentwise gradient boosting model with a mean Brier score of 0.165 and the random survival forest model with a mean AUC of 0.773, both combined with the LASSO feature selection method.

Potential contribution of vaccination uptake to occupational differences in risk of SARS-CoV-2: analysis of the ONS COVID-19 Infection Survey

OBJECTIVES

To assess variation in vaccination uptake across occupational groups as a potential explanation for variation in risk of SARS-CoV-2 infection.

DESIGN

We analysed data from the UK Office of National Statistics COVID-19 Infection Survey linked to vaccination data from the National Immunisation Management System in England from 1 December 2020 to 11 May 2022. We analysed vaccination uptake and SARS-CoV-2 infection risk by occupational group and assessed whether adjustment for vaccination reduced the variation in risk between occupational groups.

RESULTS

Estimated rates of triple vaccination were high across all occupational groups (80% or above), but were lowest for food processing (80%), personal care (82%), hospitality (83%), manual occupations (84%) and retail (85%). High rates were observed for individuals working in health (95% for office based, 92% for those in patient-facing roles) and education (91%) and office-based workers not included in other categories (90%). The impact of adjusting for vaccination when estimating relative risks of infection was generally modest (ratio of hazard ratios across all occupational groups reduced from 1.37 to 1.32), but was consistent with the hypothesis that low vaccination rates contribute to elevated risk in some groups.

CONCLUSIONS

Variation in vaccination coverage might account for a modest proportion of occupational differences in infection risk. Vaccination rates were uniformly very high in this cohort, which may suggest that the participants are not representative of the general population. Accordingly, these results should be considered tentative pending the accumulation of additional evidence.

Factors associated with recurrent emergency department visits among people living with dementia: A retrospective cohort study

BACKGROUND

Research on factors associated with recurrent emergency department (ED) visits and their implications for improving dementia care is lacking. The objective of this study was to examine associations between the individual characteristics of older adults living with dementia and recurrent ED visits.

METHODS

We used health administrative databases to conduct a population-based retrospective cohort study among older adults with dementia in Ontario, Canada. We included community-dwelling adults 66 years and older who visited the ED between April 1, 2010, and March 31, 2019 and were discharged home. We recorded all ED visits within one year after the baseline visit. We used recurrent event Cox regression to examine associations between repeat ED visits and individual clinical, demographic, and health service use characteristics. We fit conditional inference trees to identify the most important factors and define subgroups of varying risk.

RESULTS

Our cohort included 175,863 older adults with dementia. ED use in the year prior to baseline had the strongest association with recurrent visits (3+ vs.0 adjusted hazard ratio (aHR): 1.92 (1.89, 1.94), 2vs.0 aHR: 1.45 (1.43, 1.47), 1vs.0 aHR: 1.23 (1.21, 1.24)). The conditional inference tree utilized history of ED visits and comorbidity count to define 12 subgroups with ED revisit rates ranging from 0.79 to 7.27 per year. Older adults in higher risk groups were more likely to live in rural and low-income areas and had higher use of anticonvulsants, antipsychotics, and benzodiazepines.

CONCLUSIONS

History of ED visits may be a useful measure to identify older adults with dementia who would benefit from additional interventions and supports. A substantial proportion of older adults with dementia have a pattern of recurrent visits and may benefit from dementia-friendly and geriatric-focused EDs. Collaborative medication review in the ED and closer follow-up and engagement with community supports could improve patient care and experience.

The impact of gender on the risk of cardiovascular events in older adults with advanced chronic kidney disease

Background

Patients with chronic kidney disease (CKD) are at a higher risk of major adverse cardiovascular events (MACE) compared with the general population, but gender differences in this risk, especially in older adults, are not fully known. We aim to identify gender differences in the risk of MACE in older European CKD patients, and explore factors that may explain these differences.

Methods

The European Quality study (EQUAL) is a prospective study on stage 4-5 CKD patients, ≥65 years old, not on dialysis, from Germany, Italy, the Netherlands, Poland, Sweden and the UK. Cox regression and cumulative incidence competing risk curves were used to identify gender differences in MACE risks. Mediation analysis was used to identify variables which may explain risk differences between men and women.

Results

A total of 417 men out of 1134 (37%) and 185 women out of 602 women (31%) experienced at least one MACE, over a follow-up period of 5 years. Women had an 18% lower risk of first MACE compared with men (hazard ratio 0.82; 95% confidence interval 0.69-0.97; P = .02), which was attenuated after adjusting for pre-existing cardiometabolic comorbidities and cardiovascular risk factors. There were no significant gender differences in the risk of recurrent MACE or fatal MACE. The risk difference in MACE by gender was larger in patients aged 65-75 years, compared with patients over 75 years.

Conclusions

In a cohort of older adults with advanced CKD, women had lower risks of MACE. These risk differences were partially explained by pre-existing cardiometabolic comorbidities and cardiovascular risk factors.

Post hoc analysis of the SONAR trial indicates that the endothelin receptor antagonist atrasentan is associated with less pain in patients with type 2 diabetes and chronic kidney disease

Pain is prevalent among patients with diabetes and chronic kidney disease (CKD). The management of chronic pain in these patients is limited by nephrotoxicity of commonly used drugs including non-steroidal anti-inflammatory drugs (NSAIDs) and opioids. Since previous studies implicated endothelin-1 in pain nociception, our post hoc analysis of the SONAR trial assessed the association between the endothelin receptor antagonist atrasentan and pain and prescription of analgesics. SONAR was a randomized, double-blind, placebo-controlled clinical trial that recruited participants with type 2 diabetes and CKD (estimated glomerular filtration rate 25-75 ml/min/1.73 m2; urinary albumin-to-creatinine ratio 300-5000 mg/g). Participants were randomized to receive atrasentan or placebo (1834 each arm). The main outcome was pain-related adverse events (AEs) reported by investigators. We applied Cox regression to assess the effect of atrasentan compared to placebo on the risk of the first reported pain-related AE and, secondly, first prescription of analgesics. We used the Anderson-Gill method to assess effects on all (first and subsequent) pain-related AEs. During 2.2-year median follow-up, 1183 pain-related AEs occurred. Rates for the first pain-related event were 138.2 and 170.2 per 1000 person-years in the atrasentan and placebo group respectively (hazard ratio 0.82 [95% confidence interval 0.72-0.93]). Atrasentan also reduced the rate of all (first and subsequent) pain-related AEs (rate ratio 0.80 [0.70-0.91]). These findings were similar after accounting for competing risk of death (sub-hazard ratio 0.81 [0.71-0.92]). Patients treated with atrasentan initiated fewer analgesics including NSAIDs and opioids compared to placebo during follow-up (hazard ratio = 0.72 [0.60-0.88]). Thus, atrasentan was associated with reduced pain-related events and pain-related use of analgesics in carefully selected patients with type 2 diabetes and CKD.

Secondary malignancies among mantle cell lymphoma patients

PURPOSE

With modern treatments, mantle cell lymphoma (MCL) patients more frequently experience long-lasting remission resulting in a growing population of long-term survivors. Follow-up care includes identification and management of treatment-related late-effects, such as secondary malignancies (SM). We conducted a population-based study to describe the burden of SM in MCL patients.

METHODS

All patients with a primary diagnosis of MCL, aged ≥ 18 years and diagnosed between 2000 and 2017 in Sweden were included along with up to 10 individually matched population comparators. Follow-up was from twelve months after diagnosis/matching until death, emigration, or December 2019, whichever occurred first. Rates of SM among patients and comparators were estimated using the Anderson-Gill method (accounting for repeated events) and presented as hazard ratios (HR) with 95% confidence intervals (CI) adjusted for age at diagnosis, calendar year, sex, and the number of previous events.

RESULTS

Overall, 1 452 patients and 13 992 comparators were followed for 6.6 years on average. Among patients, 230 (16%) developed at least one SM, and 264 SM were observed. Relative to comparators, patients had a higher rate of SM, HRadj= 1.6 (95%CI:1.4-1.8), and higher rates were observed across all primary treatment groups: the Nordic-MCL2 protocol, R-CHOP, R-bendamustine, ibrutinib, lenalidomide, and R-CHOP/Cytarabine. Compared to Nordic-MCL2, treatment with R-bendamustine was independently associated with an increased risk of SM, HRadj= 2.0 (95%CI:1.3-3.2). Risk groups among patients were those with a higher age at diagnosis (p < 0.001), males (p = 0.006), and having a family history of lymphoma (p = 0.009). Patients had preferably higher risk of melanoma, other neoplasms of the skin and other hematopoietic and lymphoid malignancies.

CONCLUSIONS

MCL survivors have an increased risk of SM, particularly if treated with R-bendamustine. The intensive treatments needed for long-term remissions are a concern, and transition to treatment protocols with sustained efficacy but with a lower risk of SM is needed.

Severe and common mental disorders and risk of emergency hospital admissions for ambulatory care sensitive conditions among the UK Biobank cohort

BACKGROUND

People with mental disorders have worse physical health compared with the general population, which could be attributable to receiving poorer quality healthcare.

AIMS

To examine the relationship between severe and common mental disorders and risk of emergency hospital admissions for ambulatory care sensitive conditions (ACSCs), and factors associated with increased risk.

METHOD

Baseline data for England (N = 445 814) were taken from UK Biobank, which recruited participants aged 37-73 years during 2006-2010, and linked to hospital admission records up to 31 December 2019. Participants were grouped into those with a history of either schizophrenia, bipolar disorder, depression or anxiety, or no mental disorder. Survival analysis was used to assess the risk of hospital admission for ACSCs among those with mental disorders compared with those without, adjusting for factors in different domains (sociodemographic, socioeconomic, health and biomarkers, health-related behaviours, social isolation and psychological).

RESULTS

People with schizophrenia had the highest (unadjusted) risk of hospital admission for ACSCs compared with those with no mental disorder (hazard ratio 4.40, 95% CI 4.04-4.80). People with bipolar disorder (hazard ratio 2.48, 95% CI 2.28-2.69) and depression or anxiety (hazard ratio 1.76, 95% CI 1.73-1.80) also had higher risk. Associations were more conservative when including all admissions, as opposed to first admissions only. The observed associations persisted after adjusting for a range of factors.

CONCLUSIONS

People with severe mental disorders have the highest risk of preventable hospital admissions. Ensuring people with mental disorders receive adequate ambulatory care is essential to reduce the large health inequalities they experience.

Delays in Hepatitis B Immunization Series Completion in People With Human Immunodeficiency Virus

Background

Studies have demonstrated low hepatitis B virus (HBV) vaccine series completion among persons with human immunodeficiency virus (HIV).

Methods

We conducted a retrospective record review of persons entering HIV care at 2 clinics in Houston, Texas, between 2010 and 2018. Kaplan-Meier curves summarized time to receipt of HBV vaccines for those eligible for vaccination. We estimated the proportions of patients who had received 1, 2, or 3 HBV vaccine doses at 12 and 24 months after entry to care. A Prentice Williams and Peterson total time model was used to evaluate associations between patient characteristics and time to vaccination.

Results

Of the 5357 patients who entered care, 2718 were eligible for HBV vaccination. After 2 years of follow-up, 51.2% of those eligible had received 1 HBV vaccine, 43.2% had received 2, and 28.4% received 3 vaccines. With adjustment for significant cofactors, patients whose CD4 cell count was ≥200/μL (adjusted hazard ratio [aHR], 1.43 [95% confidence interval (CI), 1.29-1.59]) and transgender patients (1.49 [1.08-2.04]) received any given vaccine dose sooner than those with CD4 cell counts <200/μL or cisgender patients, respectively. Compared with non-Hispanic whites, Hispanic patients were vaccinated sooner (aHR, 1.28 [95% CI, 1.07-1.53]). Those with an active substance use history had a significantly longer time to vaccination than those with no substance use history (aHR, 0.73 [95% CI, .62-.85]).

Conclusions

Strategies are needed to increase HBV vaccine completion rates in our study population, particularly among those with CD4 cell counts <200/μL or with a substance use disorder.

Recurrent Events in Cardiovascular Trials: JACC State-of-the-Art Review

Many randomized trials in cardiovascular disease have repeat nonfatal events (such as hospitalizations) occurring during patient follow-up; yet, it remains common practice to have time-to-first event as the primary outcome. We explore the value of analyses that include repeat events. Do they help us understand the effect of treatment and total disease burden? Do they enhance statistical power? Should they become a trial's primary analysis? It may also be difficult to choose which of the various statistical methods for analyzing repeat events to use, and we provide a nontechnical guide to what each method is doing. We compare several methods for repeat events: Lin Wei Yang Ying, negative binomial, joint frailty, win ratio, and area under the curve. We illustrate their performance in 5 large cardiovascular trials and compare them with time-to-first-event analyses. We review their use in recently published heart failure trials and make recommendations for their use in future trials.

Sociodemographic and Behavioral Predictors of Recurrent Bacterial Vaginosis Among Sexual Health Clinic Patients in New York City From 2014 to 2018

OBJECTIVE

Bacterial vaginosis (BV) is associated with adverse reproductive outcomes, and recurrence is common. We examined factors associated with BV recurrence using electronic medical record data for patients attending New York City Department of Health and Mental Hygiene sexual health clinics from 2014 to 2018.

METHODS

Clinician-diagnosed BV was defined using a clinical BV diagnosis code based on Amsel criteria. Recurrent BV was defined as any BV diagnosis occurring more than 30 days after the previous diagnosis. Adjusted hazard ratios (AHRs) for the relationship between potential risk factors and recurrent BV were estimated using conditional gap-time models.

RESULTS

The data set contained 14,858 patients with at least one BV diagnosis. Of these, 46.3% (n = 6882) had at least 1 follow-up visit to a sexual health clinic between January 2014 and December 2018. Of those with a follow-up visit, 53.9% (n = 3707) had ≥1 recurrent BV episode, with 33.7% (n = 2317) experiencing recurrence within 3 months. In the multivariable model, using a hormonal intrauterine device (IUD; AHR, 1.31; 95% confidence interval [CI], 1.14-1.49) or copper IUD (AHR, 1.17; 95% CI, 1.01-1.37), having a history of trichomonas (AHR, 1.23; 95% CI, 1.12-1.36), and being non-Hispanic Black (AHR, 1.11; 95% CI, 1.04-1.18) were associated with a higher risk of BV recurrence, whereas using non-IUD hormonal contraception was associated with reduced risk (AHR, 0.88; 95% CI, 0.80-0.98).

CONCLUSIONS

Risk of BV recurrence was increased among patients using an IUD, whereas it was reduced in patients using non-IUD hormonal contraception.

New horizons in the analysis of hospital readmissions of older adults

This short communication highlights analytical methods that can be usefully applied to the problem of hospital readmissions of older adults. The limitations of the models currently used in studies of hospital readmissions are described. In summary, analyses of hospital readmissions face two important methodological and statistical problems not accounted for by these currently used statistical models: the potential recurrence of readmissions, and death, a terminal event which absorbs the readmission process. Not addressing the issue raised by recurrent events and terminal event generates biased estimates. We discuss an approach for the analysis of hospital readmission risk and death in the same framework. Understanding the features of this kind of approaches is essential at a time when high-quality data on hospital readmission in older patients are becoming available to a large number of researchers. Models adapted for the analysis of recurrent and terminal events are presented, and their application to studies of hospital readmission are explained, with reference to two cohorts of several thousand older individuals.