Andersen-Gill Model Research Articles

Predictors of relapse risk and treatment response in AQP4-IgG positive and seronegative NMOSD: A multicentre study

BackgroundNeuromyelitis optica spectrum disorder (NMOSD) can be categorised into aquaporin-4 antibody (AQP4-IgG) NMOSD or seronegative NMOSD. While our knowledge of AQP4-IgG NMOSD has evolved significantly in the past decade, seronegative...

Variable selection and estimation for recurrent event model with covariates subject to measurement error

This article focuses on variable selection in the Andersen–Gill model for recurrent event analysis, particularly when covariates are subject to measurement errors. We propose a comprehensive three-stage procedure that incorporates simulation–extrapolation with various penalty functions. This approach allows for the simultaneous selection of significant covariates, estimation of regression parameters, and adjustment for measurement errors. Through extensive simulation studies, we demonstrate that our method outperforms approaches that fail to account for measurement errors or the need for variable selection. Specifically, our procedure excels in removing unimportant error-prone covariates and accurately estimating the coefficients of important variables. The results also reveal that the magnitude of measurement error has a substantial negative impact on variable selection outcomes. Additionally, we apply our method to a real-world dataset, further illustrating its practical effectiveness and robustness.

22 Association of a germline single nucleotide polymorphism (SNP) in the interleukin-7 (IL7) gene with immune-related adverse events (irAEs)

Abstract Background Adverse events (AEs) can limit treatment immune checkpoint inhibitors (ICI) efficacy and worsen patient outcomes. Our group recently identified a germline IL7 SNP as a potential biomarker for prediction of irAEs (Groha, Nat Med 2022). In the current study, we sought to replicate the association between this IL7 SNP (rs16906115) and AEs in two clinical trials of patients with cancer treated with ICI regimens. Methods In the CheckMate-025 (CM025) trial (NCT01668784), involving patients with metastatic renal cell carcinoma (mRCC) randomized to either nivolumab (NIVO) or everolimus (EVE), whole-exome sequencing (WES) data from tumor and peripheral blood samples were analyzed. The Cancer Genome Analysis pipeline was utilized to identify somatic alterations, and the STITCH pipeline was employed to determine SNP carrier status. In the BinTA-0037 (BTA-0037) trial (NCT03631706), focusing on patients with metastatic non-small cell lung cancer (mNSCLC), the carrier status of a surrogate SNP rs16906062 (R2=0.66) was determined from tumor WES in the pembrolizumab (PEMBRO) arm. Within each treatment arm, time to incident AEs were compared between carriers (SNP+) and non-carriers (SNP-) via multivariable Cox regression, controlling for age, sex, race, ECOG and sample purity. A SNP´treatment interaction term was also included in the entire CM025 cohort. Censoring for AEs occurred at death or last follow-up. A recurrent event analysis for AEs was conducted using the Andersen-Gill model, controlling for the same variables. Additionally, overall survival (OS) and progression-free survival (PFS) were also assessed. Results In total, 534 pts were included (NIVO: n=189, PEMBRO: n=152, EVE: n=193), among which 82 (15.4%) were SNP+. There were no differences in clinical and pathological characteristics between SNP+ and SNP-, except for sex (SNP+ 16.1% vs. SNP- 30.1% in females, P=0.046). Similarly, no differences in somatic alterations, including single nucleotide and copy number variants were seen between SNP+ and SNP- in CM025. SNP carrier status had no effect on OS nor PFS in all treatment arms (all P≥0.22). The rate of grade 2+ AEs was significantly higher in SNP+ vs. SNP- in the NIVO arm (Figure A, HR=2.91[1.48-5.72]), but not in the EVE (Figure B, control, non-ICI) arm (HR=0.63[0.3-1.29], SNP´treatment Pinteraction=0.002). Similarly, the rate of all grade AEs in the PEMBRO arm of BTA-0037 was higher in SNP+ vs. SNP- (Figure C, HR=2.30[1.60-4.60]). The rate of recurrent grade 2+ AEs was also significantly higher in SNP+ vs. SNP- in the NIVO arm (HR=3.43[1.83-6.43]), whereas a trend for fewer recurrent grade 2+ AEs was seen in SNP+ vs. SNP- in the EVE arm (HR=0.46[0.17-1.25], SNP´treatment Pinteraction=0.0005). Figure: Kaplan-Meier curves showing the cumulative rate of adverse events in the NIVO (A) and EVE (B) arms of CM025, as well as the PEMBRO arm of BTA0037 (C), in SNP- and SNP+. Conclusions The IL7 SNP (rs16906115) is associated with significantly higher rates of grade 2+ AEs, including recurrent events, in pts with mRCC or mNSCLC treated with single agent PD-1 inhibitors but not with non-ICI regimens, with no effect on survival and efficacy outcomes. These results confirm the SNP’s predictive potential as a biomarker for irAEs to guide therapeutic decisions in pts treated with ICIs.

Epidemiology and risk factors of community-acquired pneumonia in patients with different causes of immunosuppression.

Immunosuppression constitutes a significant risk for community-acquired pneumonia (CAP). Nevertheless, specific causes of immunosuppression and their relevance for incidence, etiology and prognosis of CAP are insufficiently investigated.We conducted a population-based cohort study within a statutory health insurance in Germany from 2015 to 2018. CAP was retrieved by ICD-10-GM codes. Episodes of immunosuppression were identified by coded conditions (hematologic neoplasms, stem cell or organ transplantation, neutropenia, HIV, primary immunosuppressive syndromes) or treatments (immunosuppressants, antineoplastic drugs, systemic steroids). Endpoints were defined as occurrence of CAP (primary), hospitalization, 30-day mortality and CAP associated with rare pathogens. Our analysis utilized the Andersen-Gill model adjusted for sex, age, level of long-term care, vaccination status, community type and comorbidities.942,008 individuals with 54,781 CAPs were included (hospitalization 55%, 30-day mortality 14.5%). 6% of individuals showed at least one episode of immunosuppression during the study period with systemic steroids (39.8%) and hematologic neoplasms (26.7%) being most common. Immunosuppression was recorded in 7.7% of CAPs. Besides classical risk factors such as age and level of long-term care, immunosuppressed patients were most prone to CAP (HR 2.4[2.3-2.5]) and consecutive death (HR 1.9[1.8-2.1]). Organ and stem cell transplantation (HR 3.2[2.6-4.0] and 2.8[2.1-3.7], respectively), HIV (HR 3.2[1.9-5.4]) and systemic steroids (> 20mg prednisone daily dose equivalent (HR 2.7[2.4-3.1])) showed the highest risk for contracting CAP. CAP by rare pathogens was strongly associated with immunosuppression (HR 17.1[12.0-24.5]), especially HIV (HR 34.1[7.6-153]) and systemic steroids (HR 8.2[4.6-14.8]).Our study elucidates the relevance of particular immunosuppressive conditions including systemic steroids for occurrence and prognosis of CAP.

Association of a germline single nucleotide polymorphism (SNP) in the interleukin-7 (IL7) gene with immune-related adverse events (irAEs) in the CheckMate 025 trial.

12140 Background: Immune checkpoint inhibitors (ICIs) have dramatically improved outcomes of patients (pts) with metastatic renal cell carcinoma (mRCC). Unfortunately, adverse events (AEs) can limit treatment efficacy and worsen patient outcomes. Recently, a germline IL7 SNP (rs16906115) was identified as a potential biomarker for prediction of irAEs. We aimed to characterize the association between a germline IL7 SNP (rs16906115) and AEs in a prospective clinical trial of patients with mRCC treated with nivolumab (NIVO) or everolimus (EVE). Methods: Whole-exome sequencing (WES) data of tumor and peripheral blood samples from CheckMate 025 (NCT01668784) were used to infer somatic alterations using the Cancer Genome Analysis pipeline, as well as SNP carrier status using the STITCH pipeline. Within each treatment arm, time to incident adverse events (AEs) were compared between carriers (SNP+) and non-carriers (SNP-) via multivariable Cox regression, controlling for age, sex and sample purity, followed by a SNP´treatment interaction term in the whole cohort. Overall survival (OS) and progression-free survival (PFS) were also assessed. Finally, a recurrent event analysis for AEs was conducted using the Andersen-Gill model, controlling for the same variables. Results: In total, 382 pts were included (NIVO: n=189, EVE: n=193), among which 56 (14.7%) were SNP+. There were no differences in clinical and pathological characteristics between SNP+ and SNP-, except for sex (SNP+ 16.1% vs. SNP- 30.1% females, P=0.046). Similarly, no differences in somatic alterations, including single nucleotide and copy number variants were seen between SNP+ and SNP-. SNP carrier status had no effect on OS nor PFS in both treatment arms (all P≥0.47). Regarding AEs, 63 pts (33.3%) in the NIVO arm experienced at least 1 grade 2+ AE, compared to 78 pts (40.4%) in the EVE arm. The most common types of grade 2+ AEs were cutaneous (21.0%), hepatobiliary (16.8%) and endocrine (15.8%) in the NIVO arm, and respiratory (30.1%), cutaneous (22.3%) and gastrointestinal (19.4%) in the EVE arm. The rate of grade 2+ AEs was significantly higher in SNP+ vs. SNP- in the NIVO arm (HR=2.91[1.48-5.72]), but not in the EVE arm (HR=0.63[0.3-1.29], SNP´treatment Pinteraction=0.002). The rate of recurrent grade 2+ AEs was also significantly higher in SNP+ vs. SNP- in the NIVO arm (HR=3.43[1.83-6.43]), whereas a trend for fewer recurrent grade 2+ AEs was seen in SNP+ vs. SNP- in the EVE arm (HR=0.46[0.17-1.25], SNP´treatment Pinteraction=0.0005). Conclusions: The IL7 SNP (rs16906115) is associated with significantly higher rates of grade 2+ AEs, including recurrent events, in pts with mRCC treated with NIVO but not with EVE, with no effect on survival outcomes. These results affirm the SNP’s predictive potential as a biomarker for irAEs to guide therapeutic decisions in pts treated with ICIs.

Abstract 2750: Leveraging machine-learning approaches to dissect drivers of clinical metastatic dynamics in lung adenocarcinoma

Abstract Background: Non-small cell lung cancer (NSCLC) survival is closely linked to metastatic progression. 5-year survival rates drop from >65% when localized to ~9% when distant metastases occur. Despite this impact on mortality, drivers of overall metastasis and site-specific organotropism are not well understood in NSCLC. To investigate these questions, we focused on Lung adenocarcinoma (LUAD), the most common subtype of NSCLC. Although previous efforts have been made to examine LUAD organotropism, we sought to expand upon longitudinal modeling approaches and explore associations in an independent large clinical and genomic annotated cohort. Methods: Using the artificial intelligence model from (Kehl et al. 2021), we annotated a clinico-genomic cohort of 2777 patients with lung adenocarcinoma, resulting in 59,177 annotated imaging reports. This resulted in time to metastatic site annotations for each patient, with median of 2 years follow up. Metastatic sites included brain, bone, adrenal, liver, lymph node, and mesentery. All patients were evaluated at the Dana-Farber Cancer Institute, and each patient had at least one tumor biopsy sequenced with targeted panel sequencing of 400+ cancer related genes. To infer genomic and clinical correlates of site-specific metastasis, we performed cause-specific and Fine-Gray hazard modeling in this cohort. We used the recurrent event Andersen-Gill model to infer predictors of longitudinal metastasis rates. Results: Site-specific analyses yielded genes whose mutant status were significantly associated with a change in risk and/or rate of certain metastatic sites. We found that NOTCH1 is significantly associated with an increased rate and risk of brain metastases (HR=2.46, P<0.001). SETD2 is associated with decreased incidence of brain (HR=0.42, P<0.001). SMARCA4 is significantly associated with increased incidence and rate of adrenal metastasis (HR=1.64, P<0.01), while NF1 is significantly associated with decreased rate of adrenal metastasis (HR=0.72, P<0.01). In a multivariable model, Female patients and younger patients had a lower rate of new metastatic sites. As previously seen, TP53 and SMARCA4 mutations associated with higher rates of new metastatic sites (HR=1.17, P<0.001; HR=1.36, P<0.001 respectively), and SETD2 is associated with a lower rate of new metastatic sites (HR=0.71, P<0.001). Conclusions: By using time-to-event analysis with longitudinal metastatic annotations, we identify potential drivers of overall metastasis and site-specific organotropism in LUAD patients. Identification of new associations and concordance of our results with previous studies also supports the utility of artificial-intelligence-annotated datasets, allowing for larger cohorts without the need for manual annotation. Citation Format: Tyler Aprati, Michael Manos, Giuseppe Tarantino, Marc Glettig, Maryclare Griffin, Alexander Gusev, Kenneth Kehl, David Liu. Leveraging machine-learning approaches to dissect drivers of clinical metastatic dynamics in lung adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2750.

Joint analysis of vaccination effectiveness and antiviral drug effectiveness for COVID-19: a causal inference approach

ObjectivesThis study aims to estimate the causal effects of oral antivirals and vaccinations in the prevention of all-cause mortality and progression to severe COVID-19 in an integrative setting with both antivirals and vaccinations considered as interventions. MethodsWe identified hospitalized adult patients (i.e. aged 18 or above) in Hong Kong with confirmed SARS-CoV-2 infection between March 16, 2022, and December 31, 2022. An inverse probability-weighted (IPW) Andersen-Gill model with time-dependent predictors was used to address immortal time bias and produce causal estimates for the protection effects of oral antivirals and vaccinations against severe COVID-19. ResultsGiven prescription is made within 5 days of confirmed infection, nirmatrelvir-ritonavir is more effective in providing protection against all-cause mortality and development into severe COVID-19 than molnupiravir. There was no significant difference between CoronaVac and Comirnaty in the effectiveness of reducing all-cause mortality and progression to severe COVID-19. ConclusionsThe use of oral antivirals and vaccinations causes lower risks of all-cause mortality and progression to severe COVID-19 for hospitalized SARS-CoV-2 patients.

Investigating Nonspecific Effects of the Live-Attenuated Japanese Encephalitis Vaccine on Lower Respiratory Tract Infections in Children Aged 25-35 Months: Retrospective Cohort Study.

Live attenuated vaccines may be used to prevent nontargeted diseases such as lower respiratory tract infections (LRTIs) due to their nonspecific effects (NSEs). We aimed to analyze the NSEs of the Japanese encephalitis vaccine on pediatric LRTIs in children aged 25 months to 35 months. A retrospective cohort study was conducted by using a population-based electronic health record database in Zhejiang, China. Enrolled participants were children born from January 1, 2017, to December 31, 2017, and who were inoculated with the live-attenuated Japanese encephalitis vaccine (JE-L) or inactivated Japanese encephalitis vaccine (JE-I) as the most recent vaccine at 24 months of age. The study was carried out between January 1, 2019, and December 31, 2019. All inpatient and outpatient hospital visits for LRTIs among children aged 25 months to 35 months were recorded. The Andersen-Gill model was used to assess the NSEs of JE-L against LRTIs in children and compared with those of JE-I as the most recent vaccine. A total of 810 children born in 2017 were enrolled, of whom 585 received JE-L (JE-L cohort) and 225 received JE-I (JE-I cohort) as their last vaccine. The JE-L cohort showed a reduced risk of LRTIs (adjusted hazard ratio [aHR] 0.537, 95% CI 0.416-0.693), including pneumonia (aHR 0.501, 95% CI 0.393-0.638) and acute bronchitis (aHR 0.525, 95% CI 0.396-0.698) at 25 months to 35 months of age. The NSEs provided by JE-L were especially pronounced in female children (aHR 0.305, 95% CI 0.198-0.469) and children without chronic diseases (aHR 0.553, 95% CI 0.420-0.729), without siblings (aHR 0.361, 95% CI 0.255-0.511), with more than 30 inpatient and outpatient hospital visits prior to 24 months of age (aHR 0.163, 95% CI 0.091-0.290), or with 5 to 10 inpatient and outpatient hospital visits due to infectious diseases prior to 24 months old (aHR 0.058, 95% CI 0.017-0.202). Compared with JE-I, receiving JE-L as the most recent vaccine was associated with lower risk of inpatient and outpatient hospital visits for LRTIs among children aged 25 months to 35 months. The nature of NSEs induced by JE-L should be considered for policymakers and physicians when recommending JE vaccines to those at high risk of infection from the Japanese encephalitis virus.

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

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. 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. 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. 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.

Estimation of trajectory of protective efficacy in infectious disease prevention trials using recurrent event times.

In studies of infectious disease prevention, the level of protective efficacy of medicinal products such as vaccines and prophylactic drugs tends to vary over time. Many products require administration of multiple doses at scheduled times, as opposed to one-off or continual intervention. Accurate information on the trajectory of the level of protective efficacy over time facilitates informed clinical recommendations and implementation strategies, for example, with respect to the timing of administration of the doses. Based on concepts from pharmacokinetic and pharmacodynamic modeling, we propose a non-linear function for modeling the trajectory after each dose. The cumulative effect of multiple doses of the products is captured by an additive series of the function. The model has the advantages of parsimony and interpretability, while remaining flexible in capturing features of the trajectories. We incorporate this series into the Andersen-Gill model for analysis of recurrent event time data and compare it with alternative parametric and non-parametric functions. We use data on clinical malaria disease episodes from a trial of four doses of an anti-malarial drug combination for chemoprevention to illustrate, and evaluate the performance of the methods using simulation. The proposed method out-performed the alternatives in the analysis of real data in terms of Akaike and Bayesian Information Criterion. It also accurately captured the features of the protective efficacy trajectory such as the area under curve in simulations. The proposed method has strong potential to enhance the evaluation of disease prevention measures and improve their implementation strategies.

Regional differences in the effects of healthy aging on depressive symptoms: a Korean longitudinal study of aging (2006-2020).

Depression is a widely prevalent, often recurrent condition. To analyze the regional differences in depressive symptoms over time, we investigated urban-rural differences in change in depression over time in South Korea and the association between healthy aging and depressive symptoms among middle-aged and older adults. Data collected in the Korean Longitudinal Study of Aging, from 2006 to 2020, of adult participants aged ≥45 years without depressive symptoms were analyzed. Healthy aging was defined under five principal components: absence of chronic disease, good physical function, normal cognitive function, active social engagement, and good psychological adaptation. Depressive symptoms were measured using the short version of the Center for Epidemiologic Studies Depression Scale. Using the Andersen-Gill model for recurrent time-to-event, we examined the effect of healthy aging on depressive symptoms, with a subgroup analysis based on the residential area. Of the 7,708 participants, 78.2% lived in urban areas and 39.4% achieved healthy aging. In 2008, rural residents had a higher incidence of depressive symptoms (rural 11.8%; urban 8.9%); however, after 2016, the depressive symptoms of urban residents gradually increased (rural 6.4%; urban 12.1%). Unhealthy aging (adjusted hazard ratio = 3.04, 95% confidence interval: 2.72-3.39) and urban residence (adjusted hazard ratio = 1.15, 95% confidence interval: 1.06-1.24) were risk factors for depressive symptoms. The subgroup analysis revealed that individuals who did not achieve healthy aging had an increased risk of depressive symptoms, regardless of their residential area (hazard ratio [95% confidence interval]: urban, 3.13 [2.75-3.55]; rural 2.59 [2.05-3.28]). As urbanization accelerates, urban residents have a higher risk of depressive symptoms than rural residents. Healthy aging is an essential factor in reducing depressive symptoms. To achieve healthy aging, appropriate interventions and policies that target the middle-aged adults and gradually extend to older adults are needed, considering individual and regional factors.

Longer is Better for Endoscopic Follow-up of Upper Tract Urothelial Carcinoma After Ureteroscopic Treatment: An Evaluation Spanning 10 Years of Data

BackgroundTumour recurrences are frequent among patients with upper tract urothelial carcinoma (UTUC) treated with ureteroscopy (URS). Therefore, guidelines recommend a strict follow-up regimen, but there is little evidence on how to do this. ObjectiveTo analyse outcomes during our follow-up regimen and the impact on treatment in terms of ipsilateral UTUC recurrence, treatment conversion, and tumour upgrading, and to evaluate potential prognostic factors, including second-look URS outcomes. A secondary objective was to evaluate survival outcomes. Design, setting, and participantsThe single-centre cohort included all adult patients with nonmetastatic UTUC treated with URS from January 2010 to December 2020. Follow-up involved endoscopy at 3-mo intervals in the first year, then at 6-mo intervals up to year 3, and yearly thereafter. Outcome measurements and statistical analysisDescriptive analyses were performed for the follow-up outcomes. The Andersen-Gill model for recurrent event analysis was used to analyse tumour recurrences, and multivariable Cox regression to analyse for predictors for treatment conversion in low-grade tumours. Results and limitationsWe analysed 71 patients with median follow-up of 49.5 mo. The overall 2-yr recurrence-free survival (RFS) rate was 22%. In low-grade disease, the 1-yr RFS rate was 50% and the 2-yr RFS rate was 29%. Treatment was converted to radical nephroureterectomy for 23 patients, at a median time to conversion of 9.9 mo. Upgrading was seen in 13 patients, at a median time to upgrading of 21.9 mo. No factors were prognostic for either tumour recurrence or treatment conversion. The 5-yr OS, CSS, and MFS rates were 82%, 86%, and 84%, respectively. ConclusionsOur data show that it is rational to extend endoscopic follow-up for UTUC treated with URS, as clinically relevant events (treatment conversion and tumour upgrading) occur beyond the current 6-mo guideline recommendation. Second-look URS outcomes were not prognostic for tumour recurrence or treatment conversion during follow-up. Patient summaryOur study results show that for patients with cancer of the upper urinary tract treated with kidney-sparing surgery through a small telescope called a ureteroscope (URS), most of the clinically relevant events (treatment conversion and tumour upgrading) occur outside the current recommended follow-up of 6 months. Therefore, URS follow-up should be extended for these patients.

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

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. 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. 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). 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.

Commuting and sick leave: a retrospective longitudinal study among a Belgian military population

ObjectivesIn a military context, people often have to deal with long commuting distance. The aim of the current study is to investigate to what extent commuting distances predict sickness absence...

Sex differences in acute health service contact after release from prison in Australia: a data linkage study

ObjectivesWomen released from prison typically experience worse health outcomes than their male counterparts. We examined sex differences in the patterns, characteristics, and predictors of acute health service contact (AHSC) (i.e. ambulance and/or emergency department use) after release from prison. Study designData linkage study. MethodsBaseline survey data from 1307 adults (21% women) within six weeks of expected release from prisons in Queensland, Australia (2008–2010) were linked prospectively with state-wide ambulance and emergency department, correctional, mental health, and death records. Crude and adjusted incidence rates and incidence rate ratios of AHSC were calculated overall and by sex. An Andersen–Gill model was fit to examine whether sex predicted AHSC. The interaction effect between sex and each model covariate was tested. ResultsThe crude incidence rates of AHSC after release from prison were 1.4 (95% confidence interval [CI]: 1.3–1.5) and 1·1 (95%CI: 1.1–1.2) per person-year for women and men, respectively. The relationship between perceived physical health–related functioning at the baseline and AHSC was modified by sex (P = 0·039). The relationship between perceived health-related functioning and AHSC also differed among women. Compared to women who perceived their physical health as fair or good at the baseline, women who perceived their physical health as poor were at greater risk of AHSC (hazard ratio = 2.4, 95%CI: 1.4–3·9, P = 0.001) after release from prison. ConclusionsAmong people released from prison, women's and men's AHSC differs depending on how they perceive their own physical health. The specific needs of women and men must be considered in transitional support policy and planning to improve their health outcomes.

Hospitalisations and Cost of Inpatient Care for Physical Diseases in Survivors of Childhood Cancer in Western Australia: a Longitudinal Matched Cohort Study.

The long-term effects of childhood cancer are unclear in the Australian context. We examined hospitalisation trends for physical diseases and estimated the associated inpatient care costs in all 5-year childhood cancer survivors (CCS) diagnosed in Western Australia (WA) from 1982-2014. Hospitalisation records for 2,938 CCS and 24,792 comparisons were extracted from 1987- 2019 (median follow-up=12 years, min=1, max=32). The adjusted hazard ratio (aHR) of hospitalisation with 95% confidence intervals (CI) was estimated using the Andersen-Gill model for recurrent events. The cumulative burden of hospitalisations over time was assessed using the mean cumulative count method. The adjusted mean cost of hospitalisation was estimated using the generalised linear models. We identified a higher risk of hospitalisation for all-cause (aHR=2.0, 95%CI= 1.8-2.2) physical disease in CCS than comparisons, with the highest risk for subsequent malignant neoplasms (aHR=15.0, 95%CI 11.3-19.8) and blood diseases (aHR=6.9, 95%CI 2.6-18.2). Characteristics associated with higher hospitalisation rates included female gender, diagnosis with bone tumours, cancer diagnosis age between 5-9 years, multiple childhood cancer diagnoses, multiple comorbidities, higher deprivation, increased remoteness, and Indigenous status. The difference in the mean total hospitalisation costs for any disease was significantly higher in survivors than comparisons (publicly funded $11,483 United States Dollar, p<0.05). The CCS population face a significantly higher risk of physical morbidity and higher cost of hospital-based care than the comparisons. Our study highlights the need for long-term follow-up healthcare services to prevent disease progression and mitigate the burden of physical morbidity on patients and hospital services.

PO-05-186 LEVERAGING A NATURAL EXPERIMENT TO DETERMINE THE EFFECT OF AIR POLLUTION ON VENTRICULAR ARRHYTHMIAS AND PHYSICAL ACTIVITY: THE 2020 U.S. WILDFIRES

Particulate matter (PM) air pollution is associated with atherosclerotic cardiovascular disease. While the biological processes that link PM exposure to atherosclerotic disease may also trigger arrhythmias, there are limited data on the association of PM with ventricular arrhythmias. The high levels of PM pollution as a consequence of the 2020 U.S. wildfires provide a unique opportunity to explore the effects of air pollution on cardiac arrhythmias, physical activity, and heart failure markers using device remote monitoring data. To evaluate the association of PM exposure to ventricular arrhythmias, physical activity, and markers of heart failure. We performed a retrospective cohort study using the CERTITUDE database of patients with Biotronik CIEDs. We included patients with at least one remote data transmission on or after 1/1/2020. The primary predictors were Air Quality Index (AQI), PM < 10 μm in diameter (PM10), and PM < 2.5 μm in diameter (PM2.5). We evaluated the predictors as continuous and binary variables for values ≧ the 90th percentile for the 2020 calendar year. We determined day-level air pollutant levels using US EPA data cross-linked to patients’ Zip codes as a proxy for patient location, and we examined the outcomes of interest listed in the Table. We evaluated the association of air pollution to the outcomes of interest with (1) a case-crossover analysis using a conditional logistic regression model followed by (2) a time-varying exposure analysis using an Andersen-Gill model and day level air-pollutant levels as a time-varying exposure. The study cohort included 28349 patients (32% female, age 72.8±11.9), of whom 1822 (6.4%) had loop recorders, 17448 (61.6%) had pacemakers, and 9079 (32%) had defibrillators. In the case-crossover analysis, AQI and PM2.5 demonstrated non-significant trends toward higher ATP and shocks (Table). AQI and PM2.5 had significant but non-meaningful associations with change in heart rate, and PM2.5 with physical activity and thoracic impedance. The time-varying exposure analysis demonstrated significant but non-meaningful associations (ORs 0.99-1). In contrast to prior literature on ambient air pollution, these natural experiment analyses did not demonstrate acute or sub-acute association between AQI, PM2.5, or PM10 with ventricular arrhythmias, physical activity, or heart failure markers in CIED patients. Mediators of these findings merit further investigation to clarify causal and non-causal mechanisms.

Use of the FreeStyle Libre system and diabetes treatment progression in T2DM: Results from a retrospective cohort study using a Canadian private payer claims database.

Up to one-third of Canadians are estimated to be living with prediabetes or diabetes. A retrospective study using Canadian private drug claims data was conducted to investigate whether flash glucose monitoring using the FreeStyle Libre system (FSL) among people with type 2 diabetes mellitus (T2DM) in Canada can be associated with changes in treatment intensification when compared with blood glucose monitoring (BGM) alone. Using a Canadian national private drug claims database comprising approximately 50% coverage of insured individuals in Canada, cohorts of people with T2DM using FSL or BGM were identified algorithmically based on treatment history and followed over a 24-month study period, tracking their progression in diabetes treatment therapy. The Andersen-Gill model for recurrent time-to-event data was used to evaluate whether the rate of treatment progression differs between the FSL and BGM treatment cohorts. The survival function was used to calculate comparative treatment progression probabilities between the cohorts. In total, 373 871 people with T2DM met the inclusion criteria. Across treatment (FSL) and control (BGM) groups, people using FSL had a higher probability of treatment progression compared with BGM alone, with a relative risk ranging between 1.86 and 2.81 (p < .001). A higher probability of treatment progression was independent of the diabetes treatment at the enrolment date (index date) or the patient status, and independent of whether patients were treatment naïve or on established diabetes therapy. Assessment of the ending treatment relative to the starting therapy indicated that dynamic treatment changes were most evident for patients in the FSL cohort and that the FSL cohort had a much greater portion of patients who ended with insulin treatment (when they started with non-insulin treatment) compared with the BGM cohort. People with T2DM using FSL had a greater probability for treatment progression compared with BGM alone, irrespective of the starting therapy, which may suggest that FSL can be used to support escalation of diabetes therapy to improve therapeutic inertia in T2DM.

1888. Implementation of COVID-19 Surveillance Testing Procedures in Dental School

Abstract Background Dental practitioners and students of dentistry are potentially at increased risk of COVID-19 infection due to frequent usage of aerosol-generating procedures. To mitigate risk to patients and providers, the University of Utah School of Dentistry began regular surveillance PCR testing of its patient-facing faculty, staff, and students in May 2020. Methods Surveillance testing occurred every other week for non-vaccinated individuals and continued through February 2022. After May 2021, fully vaccinated individuals were tested monthly and encouraged to seek additional testing if symptoms or an exposure occurred. We assessed risk of positive test among faculty, student, and staff groups through a Cox proportional hazards regression, accounting for multiple events and time-dependent variables with the Andersen-Gill model. To account for inconsistent testing after vaccination, time was examined as number of tests rather than calendar time. Results In total, 410 participants were followed during the observation period, with an average of 22 (SD 10.0, RNG 1-50) tests per person. A total of 9,452 tests were performed. There were 158 positive tests, with 60 (38%) occurring in January 2022 alone. When analyzed by themselves, staff and student groups were significantly more likely to test positive (HR 1.98, 95% CI 1.15-3.42; HR 2.16, 95% CI 1.29-3.63 respectively) compared to faculty. However, once additional covariates were accounted for, the relationship was no longer significant (Staff: HR 2.15, 95% CI 0.92-5.05; Students: HR 2.38, 95% CI 0.88-6.40). Risk of COVID-19 within Dental School Hazard Ratios for testing positive for COVID-19 among different groups within the dental school. Vaccination is accounted for as time since last vaccine, with separate categories for one dose, and 2 or more doses combined. Time examined as test number. Conclusion More than a third of all positive tests during the 22-month study occurred during one month of the Omicron wave. This sudden increase in positive tests was not observed in previous surges, and demonstrates the intensity of the Omicron wave. Additionally, we did not find a significant difference between patient-facing groups who had different work exposures. While this may be due to effective preventative measures, within the dental setting we do not see evidence that work role and resulting exposures increase risk. Disclosures All Authors: No reported disclosures.

Association of Opioid and Stimulant Use Disorder Diagnoses With Fatal and Nonfatal Overdose Among People With a History of Incarceration

Studies have suggested a rise in opioid- and stimulant-involved overdoses in recent years in North America. This risk may be acute for individuals who have had contact with the criminal justice system, who are particularly vulnerable to overdose risk. To examine the association of opioid and/or stimulant use disorder diagnoses with overdose (fatal and nonfatal) among people with histories of incarceration. In this cohort study, population-based health and corrections data were retrieved from the British Columbia Provincial Overdose Cohort, which contains a 20% random sample of residents of British Columbia. The analysis included all people in the 20% random sample who had a history of incarceration between January 1, 2010, and December 31, 2014. Outcomes were derived from 5-years of follow-up data (January 1, 2015, to December 31, 2019). Statistical analysis took place from January 2022 to June 2022. Substance use disorder diagnosis type (ie, opioid use disorder, stimulant use disorder, both, or neither), sociodemographic, health, and incarceration characteristics. Hazard ratios (HRs) are reported from an Andersen-Gill model for recurrent nonfatal overdose events and from a Fine and Gray competing risk model for fatal overdose events. The study identified 6816 people (5980 male [87.7%]; 2820 aged <30 years [41.4%]) with histories of incarceration. Of these, 293 (4.3%) had opioid use disorder only, 395 (6.8%) had stimulant use disorder only, and 281 (4.1%) had both diagnoses. During follow-up, 1655 people experienced 4026 overdoses including 3781 (93.9%) nonfatal overdoses, and 245 (6.1%) fatal overdoses. In adjusted analyses, the hazard of both fatal (HR, 2.39; 95% CI, 1.48-3.86) and nonfatal (HR, 2.45; 95% CI, 1.94-3.11) overdose was highest in the group with both opioid and stimulant use disorder diagnoses. This cohort study of people with a history of incarceration found an elevated hazard of fatal and nonfatal overdose among people with both opioid and stimulant use disorder diagnoses. This study suggests an urgent need to address the service needs of individuals who have had contact with the criminal justice system and who co-use opioids and stimulants.