Multivariable Conditional Regression Models Research Articles

Disparities in Access to Outpatient Surgery Related to Removal of Procedures from Medicare's Inpatient Only List.

We assessed the shift from inpatient to outpatient surgical care related to changes to the Inpatient Only List in 2020 and 2021 compared to 2019. The extent to which procedures shift from the inpatient to outpatient setting following removal from Medicare's Inpatient Only List is unknown. Many health systems also encouraged a shift from inpatient to outpatient surgery during the COVID-19 pandemic. Assessing the relative change in outpatient surgical utilization for procedures removed from the Inpatient Only List during COVID-19 would provide empirical data on whether reimbursement policy changes or inpatient capacity needs during the pandemic were more likely to shift care from the inpatient to outpatient setting. We used administrative data from the PINC AI Healthcare Database across 723 hospitals to determine the within-facility relative change in outpatient vs inpatient procedural volume in 2020 and 2021 compared to 2019 using a multivariable conditional fixed-effects Poisson regression model. We also assessed whether outpatient surgical utilization varied by race and ethnicity. Using a multivariable linear probability model, we assessed the absolute change in risk-adjusted 30-day complication, readmission, and mortality rates for inpatient and outpatient surgical procedures. In 2020 and 2021 compared to 2019 respectively, there was a 5.3% (95% CI, 1.4% to 9.5%) and 41.3% (95% CI 33.1% to 50.0%) relative increase in outpatient elective procedural volume. Outpatient procedural volume increased most significantly for hip replacement which was removed from the Inpatient Only List in 2020 (increase in outpatient surgical utilization of 589.3% (95% CI, 524.9% to 660.3%)). The shift to outpatient hip replacement procedures was concentrated among White patients; in 2021, hip replacement procedural volume increased by 271.1% (95% CI, 241.2% and 303.7%) for White patients and 29.5% (95% CI, 24.4% and 34.9%) for Black patients compared to 2019 levels. There were no consistent or large changes in 30-day complication, readmission, or mortality risk in 2020 and 2021 compared to 2019. There was a modest increase in elective outpatient surgeries and a pronounced increase in outpatient orthopedic surgeries which were removed from the Inpatient Only List during the COVID-19 pandemic. Utilization of outpatient surgical procedures was concentrated among White patients.

Periodontitis, Dental Procedures, and Young-Onset Cryptogenic Stroke

Periodontitis is associated with an increased risk of ischemic stroke, and the risk may be particularly high among young people with unexplained stroke etiology. Thus, we investigated in a case-control study whether periodontitis or recent invasive dental treatments are associated with young-onset cryptogenic ischemic stroke (CIS). We enrolled participants from a multicenter case-control SECRETO study including adults aged 18 to 49 y presenting with an imaging-positive first-ever CIS and stroke-free age- and sex-matched controls. Thorough clinical and radiographic oral examination was performed. Furthermore, we measured serum lipopolysaccharide (LPS) and lipotechoic acid (LTA) levels. Multivariate conditional regression models were adjusted for stroke risk factors, regular dentist visits, and patent foramen ovale (PFO) status. We enrolled 146 case-control pairs (median age 41.9 y; 58.2% males). Periodontitis was diagnosed in 27.5% of CIS patients and 20.1% of controls (P < 0.001). In the fully adjusted models, CIS was associated with high periodontal inflammation burden (odds ratio [OR], 95% confidence interval) with an OR of 10.48 (3.18–34.5) and severe periodontitis with an OR of 7.48 (1.24–44.9). Stroke severity increased with the severity of periodontitis, having an OR of 6.43 (1.87–23.0) in stage III to IV, grade C. Invasive dental treatments performed within 3 mo prestroke were associated with CIS, with an OR of 2.54 (1.01–6.39). Association between CIS and invasive dental treatments was especially strong among those with PFO showing an OR of 6.26 (1.72–40.2). LPS/LTA did not differ between CIS patients and controls but displayed an increasing trend with periodontitis severity. Periodontitis and recent invasive dental procedures were associated with CIS after controlling for multiple confounders. However, the role of bacteremia as a mediator of this risk was not confirmed.

Associations of Dietary Factors with Cutaneous Melanoma: A Case-Control Study in Greece with Literature Review

Introduction: It has been postulated that nutrition may influence the risk for cutaneous melanoma (CM); therefore, we aimed to assess the associations of food groups and individual nutrient intakes with CM in a Greek population. Methods: In this case-control study, 151 patients with histologically confirmed CM, newly diagnosed and treated in the Oncology Department of the “Laikon” University Hospital (Athens, Greece), and 151 age- and sex-matched healthy individuals residing in the Athens metropolitan area, recruited among participants for routine health examinations, were included. All participants completed a questionnaire comprising anthropometric measurements, sociodemographic, lifestyle, and health-related variables. A validated, semiquantitative food frequency questionnaire was used to assess average consumption of 136 food items during the 12 months preceding the onset of disease. Multivariate conditional regression models were used to derive odds ratios (ORs) with 95% confidence intervals (95% CI) regarding the association of nine food groups and seven macronutrients with CM. Results: Statistically significant positive associations with CM were found with higher energy intake (OR: 1.67, 95% CI: 1.22–2.30) and intake of saturated fatty acids (OR: 2.28, 95% CI: 1.00–5.28), after adjusting for sun sensitivity, major depression history, and alcohol intake. Inverse associations with higher intake of milk and dairy products (OR: 0.65, 95% CI: 0.48–0.88), fruits (OR: 0.68, 95% CI: 0.51–0.90), added lipids (OR: 0.65, 95% CI: 0.47–0.91), and sugars and syrups (OR: 0.70, 95% CI: 0.53–0.93) were also observed. Conclusions: Beyond intrinsic risk factors, our results support associations of CM with multiple food groups and nutrients; if confirmed by prospective studies, these findings can add further knowledge about this fatal cancer.

Carotid intima-media thickness in adults with and without psoriasis - a nested case-control study from baseline data of ELSA-Brasil cohort.

There is a lack of consensus about the association between psoriasis (PSO) and carotid intima-media thickness (cIMT) in literature, since previous studies considered dermatologic clinic patients or general population. This study aimed to compare cIMT levels according to PSO in a sample of 10,530 civil servants form the ELSA-Brasil cohort study and analyze its association with the disease. The PSO cases and disease duration were identified by medical diagnosis self-reported at study enrollment. A paired group was identified by propensity score matching among all the participants without PSO. Mean cIMT values were considered for continuous analysis while cIMT above 75th percentile was considered for categorical analysis. Multivariate conditional regression models were used to analyze association between cIMT and PSO diagnosis, by comparing PSO cases against paired controls and overall sample without disease. A total of n = 162 PSO cases were identified (1.54%) and no difference in cIMT values was observed between participants with PSO and overall sample or control group. PSO was not associated with linear increment of cIMT (vs. overall sample: β = 0.003, p = 0.690; vs. matched controls: β = 0.004, p = 0.633) neither with increased chance of having cIMT above 75th percentile (vs. overall sample: OR = 1.06, p = 0.777; vs. matched controls: OR = 1.19, p = 0.432; conditional regression: OR = 1.31, p = 0.254). There was no relationship between disease duration and cIMT (β = 0.000, p = 0.627). Although no significant relationship between mild cases of psoriasis and cIMT was observed among a wide cohort of civil servants, longitudinal investigation about cIMT progression and severity of disease are still needed.

Antibiotic therapy is associated with an increased incidence of cancer

PurposeThere is a growing body of evidence suggesting the decisive involvement of the human microbiome in cancer development. The consumption of antibiotics may fundamentally change the microbiome and thereby create a precancerous environment promoting cancer development and growth. However, clinical data on the association between the consumption of antibiotics and cancer incidence have remained inconclusive. In this study, we quantified the association between the intake of different antibiotics and various cancer entities among outpatients from Germany.MethodsThis retrospective case–control study based on the IQVIA Disease Analyzer database included 111,828 cancer patients and 111,828 non-cancer controls who were matched to cancer cases using propensity scores. Patients were categorized as non-users, low-consumption (up to 50th percentile), and high-consumption (above 50th percentile) users of antibiotics overall and for each antibiotic class. Multivariable logistic conditional regression models were used to study the association between antibiotic intake within 5 years prior to the index date (first cancer diagnosis for cases or randomly selected date for controls) and cancer incidence.ResultsThe probability of cancer was significantly higher among patients with a history of antibiotic intake than in matched controls. Patients using penicillin or cephalosporins displayed a higher incidence of cancer, while the intake of tetracyclines and macrolides actually reduced the risk of cancer development slightly. A complex picture was observed in our cancer site-stratified analyses. Most notably, the consumption of penicillin was significantly and positively associated with cancer development in the respiratory organs only (low consumption OR: 1.33, 95% CI 1.20–1.47; high consumption OR 1.42, 95% CI 1.22–1.64) and cephalosporin consumption was significantly associated with respiratory organ cancer (low consumption OR: 1.32, 95% CI 1.17–1.48, high consumption OR: 1.47, 95% CI 1.29–1.66), breast cancer (high consumption OR: 1.40, 95% CI 1.25–1.56), and lymphoid and hematopoietic tissue cancer (high consumption OR: 1.50, 95% CI 1.35–1.66).ConclusionOur data strongly support the hypothesis that the intake of antibiotics is positively associated with the risk of cancer development.

Comparing the Occurrence of Healthcare-Associated Infections in Patients with and without COVID-19 Hospitalized during the Pandemic: A 16-Month Retrospective Cohort Study in a Hospital Intensive Care Unit

The COVID-19 pandemic has increased the healthcare-associated infection (HAI) risk in intensive care unit (ICU) patients. However, a comparison between patients with and without COVID-19 in terms of HAI incidence has been rarely explored. In this study, we characterized the occurrence of HAI among patients with and without COVID-19 admitted to the ICU of the Umberto I hospital of Rome during the first 16 months of the pandemic and also identified risk factors for HAI acquisition. Patients were divided into four groups according to their ICU admission date. A multivariable conditional risk set regression model for multiple events was constructed for each admission period. Adjusted hazard ratios and 95% confidence intervals were calculated. Overall, 352 COVID-19 and 130 non-COVID-19 patients were included, and a total of 361 HAIs were recorded. We found small differences between patients with and without COVID-19 in the occurrence and type of HAI, but the infections in the two cohorts mostly involved different microorganisms. The results indicate that patient management was likely an important factor influencing the HAI occurrence during the pandemic. Effective prevention and control strategies to reduce HAI rates should be implemented.

Dietary Total Antioxidant Capacity and Risk of Polycystic Ovary Syndrome: A Case-Control Study.

Background:Among multiple factors that affect the etiology of polycystic ovary syndrome (PCOS), diet has an important contribution. Chronic oxidative stress has also been implicated in the development of PCOS. The present study is an attempt to evaluate dietary total antioxidant capacity (TAC) and its relationship with odds of PCOS in Iran.Materials and Methods:The study was carried out as a case-control study in hospital outpatient clinics, Tehran, Iran. Totally, 310 female participants with a history of PCOS and 602 age-matched controls took part in this study between June 2015 and December 2018. A reproducible and valid 168-item semi-quantitative food frequency inventory was utilized to determine the entire antioxidants of the usual diet in order to calculate dietary TAC. The relationship of dietary TAC with odds of PCOS were assessed adjusting for potential confounders through an estimation of two multivariable conditional regression models. The first tertile was presented as a reference category.Results:In a fully adjusted model, the highest tertile of dietary TAC was associated with a reduced odds of PCOS [odds ratio (OR): 0.81, 95% confidence interval (95% CI): 0.59, 0.96, P for trend: 0.038). In addition, PCOS odds decreased in the highest tertile of α-tocopherol intake (OR: 0.73, 95% CI: 0.56, 0.88, P for trend: 0.023). The adjusted ORs in the highest tertile of vitamin C, β-carotene and magnesium were 0.79 (95% CI: 0.83-0.97), 0.81 (95% CI: 0.67-0.98) and 0.91 (95% CI: 0.55-0.9 8) respectively, with a significant trend.Conclusion:Our results provide evidence that there was a relationship between high TAC diets and lower odds of PCOS.

Diet and Genetic Risk Factors of Colorectal Cancer in Palestine: A Case-Control Study

To add evidence to the limited data available on colorectal cancer (CRC) from Palestine, we examine the risk factors associated with CRC using a matched hospital-based case-control study. A structured questionnaire was used to collect data from 105 cases and 105 controls. A multivariable conditional regression model was used to adjust for the association between study factors and CRC risk. In the model, compared with controls, cases from villages were significantly less likely to have CRC (Adjusted Odds Ratio, AOR = 0.194); taking aspirin lowered the likelihood of CRC by 24%; and having a multiple birth sibling by 33%. Also, the likelihood of CRC was lowered significantly by consuming five servings of fruits/vegetables per week or more (5–6 servings: AOR = 0.21, 7–8 servings per week: AOR = 0.04). However, cases had a significantly higher likelihood of CRC if they consumed 2–4 servings of grilled red meat per week (AOR = 4.25); smoked (AOR = 4.38); had a sedentary lifestyle (AOR = 2.53); reported parental consanguinity (AOR = 3.88); or had a family history of cancer (AOR = 6.39). Our results confirmed the association between CRC and red meat intake and smoking, and proved that parental consanguinity and family history of cancer are also risk factors for CRC.

Association Between Serum Cadmium and Arsenic Levels with Bladder Cancer: A Case-control Study

Background: Bladder cancer is the second most common cancer of the genitourinary system and the eighth cause of cancer death. In addition to known risk factors such as smoking and urinary stones, trace elements are also effective in causing bladder cancer and other cancers. Objectives: This study was conducted to determine the association between bladder cancer and the carcinogens cadmium and arsenic. Methods: This case-control study was conducted on 40 patients with bladder cancer admitted to Shahid Bahonar Hospital and a control group of 40 healthy individuals in Kerman, Iran, in 2018. The serum levels of arsenic and cadmium were measured by atomic absorptiometry. The paired t-test and chi-square tests were employed to assess the difference between cases and controls groups. An unadjusted and 2 multivariable conditional regression models were separately adjusted on sex, family cancer history, residence, occupation, and smoking and were used to estimate odds ratios (OR) and 95% confidence intervals (95% CI) to assess the association between arsenic and cadmium levels and bladder cancer. The statistical software SPSS version 26 and R software version 3.6.3 were used to perform the statistical analyses. Results: The mean cadmium level was 2.99 ± 1.45 and 2.59 ± 0.46 in the case and control groups, respectively, with no statistically significant difference between the groups (P = 0.100). The mean arsenic level was 2.12 ± 1.04 and 1.43 ± 0.73 in the case and control groups, respectively, demonstrating a statistically significant difference between the groups (P = 0.001). Unadjusted and adjusted conditional logistic regression models indicated significant association between arsenic levels and bladder cancer (unadjusted: odds ratio (OR) (95% CI): 0.66 (0.46 - 0.94), P-value = 0.022; adjusted: OR (95% CI): 0.64 (0.44 - 0.92), P-value = 0.018). Conclusions: Overall, cadmium and arsenic levels are higher in patients with bladder cancer, with a statistically significant difference for arsenic. However, these elements are not interrelated and are not related to other factors.

Risk factors for candidemia: a prospective matched case-control study

BackgroundCandidemia is an opportunistic infection associated with high morbidity and mortality in patients hospitalized both inside and outside intensive care units (ICUs). Identification of patients at risk is crucial to ensure prompt antifungal therapy. We sought to assess risk factors for candidemia and death, both outside and inside ICUs.MethodsThis prospective multicenter matched case-control study involved six teaching hospitals in Switzerland and France. Cases were defined by positive blood cultures for Candida sp. Controls were matched to cases using the following criteria: age, hospitalization ward, hospitalization duration, and, when applicable, type of surgery. One to three controls were enrolled by case. Risk factors were analyzed by univariate and multivariate conditional regression models, as a basis for a new scoring system to predict candidemia.ResultsOne hundred ninety-two candidemic patients and 411 matched controls were included. Forty-four percent of included patients were hospitalized in ICUs, and 56% were hospitalized outside ICUs. Independent risk factors for candidemia in the ICU population included total parenteral nutrition, acute kidney injury, heart disease, prior septic shock, and exposure to aminoglycoside antibiotics. Independent risk factors for candidemia in the non-ICU population included central venous catheter, total parenteral nutrition, and exposure to glycopeptides and nitroimidazoles. The accuracy of the scores based on these risk factors is better in the ICU than in the non-ICU population. Independent risk factors for death in candidemic patients included septic shock, acute kidney injury, and the number of antibiotics to which patients were exposed before candidemia.DiscussionWhile this study shows a role for known and novel risk factors for candidemia, it specifically highlights important differences in their distribution according to the hospital setting (ICU versus non-ICU).ConclusionThis study provides novel risk scores for candidemia accounting for the hospital setting and recent progress in patients’ management strategies and fungal epidemiology.

Previous Live Births and Induced Abortions May Precede Later Development of Graves’ Hyperthyroidism

Objective: To investigate the association between reproductive history and later development of various nosological subtypes of overt hyperthyroidism. Study Design: From the Danish population, we included incident hyperthyroid women, and for each case we recruited 4 euthyroid age-sex-region-matched controls from the same sub-population. Hyperthyroid cases/controls were: Graves’ disease (GD, n = 232/928), multinodular toxic goitre (MNTG, n = 91/364), solitary toxic adenoma (STA, n = 21/84). Patients diagnosed with hyperthyroidism within 1 year after delivery including post-partum GD were excluded. In multivariate conditional regression models (reference: no reproductive events), we analysed the association between development of GD/MNTG/STA and reproductive factors such as age at menarche/menopause, reproductive span, number of pregnancies/childbirths/abortions, investigations for infertility, and years on oral contraceptives. We adjusted for possible confounders such as alcohol intake, smoking, co-morbidity, and education. Age was studied as a potential effect measure modifier. Results: GD patients diagnosed before the age of 40 years had given births more often than control subjects (OR [95% CI] for 1/2/3+ births [ref.: nulliparous] were 1.57 [0.80–3.11]/2.06 [1.001–4.22]/3.07 [1.50–6.26]), and they had induced abortions performed more often (OR for 1/2+ induced abortions [ref.: no: events] were 0.99 [0.54–1.84]/2.24 [1.12–4.45]). No associations were observed between any reproductive factor and the development of MNTG or STA. Conclusions: Childbirths and induced abortions may be followed by development of Graves’ hyperthyroidism after the post-partum period. This was not the case for the non-autoimmune subtypes of hyperthyroidism.

Patterns of antiepileptic drug use in patients with potential refractory epilepsy in Texas Medicaid

ObjectivesAntiepileptic drug (AED) monotherapy is usually effective in 60% of the patients with epilepsy while the remaining patients have refractory epilepsy. This study compared treatment patterns (adherence, persistence, addition, and switching) associated with refractory and nonrefractory epilepsy. MethodsTexas Medicaid claims from 09/01/07–12/31/13 were analyzed, and patients eligible for the study 1) were between 18 and 62 years of age, 2) had a prescription claim for an AED during the identification period (03/01/08–12/31/11) with no prior baseline AED use (6-month), and 3) had evidence of epilepsy diagnosis within 6 months of AED use. Based on AED use in the identification period, patients were categorized into “refractory” (≥3AEDs) and “nonrefractory” (<3AEDs) cohorts. The index date was the date of the first AED claim. Patients in both cohorts were matched 1:1 using propensity scoring and compared for adherence (proportion of days covered (PDC) ≥80% vs. <80%), persistence, addition (yes/no), and switching (yes/no) using multivariate conditional regression models. Conditional logistic regression and Cox proportional hazard models were used to address the study objectives. ResultsOf the 10,599 eligible patients, 2798 (26.5%) patients in the refractory cohort were matched to patients in the nonrefractory cohort. Patients in the refractory cohort had significantly higher (p < 0.005) mean (±Standard deviation (SD)) adherence (88.6% (±19.1%) vs. 77.0% ± (25.8%)) and persistence (328.0 (±87.3) days vs. 294.9 ± (113.4) days) as compared with patients in the nonrefractory cohort. Compared with patients with nonrefractory epilepsy, patients with refractory epilepsy were 3.6 times (odds ratio (OR) = 3.553; 95% confidence interval (CI) = 3.060–4.125; p < 0.0001) more likely to adhere to AEDs and had a 34.7% (hazard ratio (HR) = 0.653; 95% CI = 0.608–0.702; p < 0.0001) lower hazard rate of discontinuation of AEDs. Also, patients with refractory epilepsy were 3.7 times (OR = 3.723; 95% CI = 2.902–4.776; p < 0.0001) more likely to add an alternative AED and 3.6 times (OR = 3.591; 95% CI = 3.010–4.284; p < 0.0001) more likely to switch to an alternative AED. ConclusionPatients with refractory epilepsy were significantly more likely to adhere and persist to AED regimen and were significantly more likely to add and switch to an alternative AED than patients with nonrefractory epilepsy.

Risk of stroke associated with use of nonsteroidal anti-inflammatory drugs during acute respiratory infection episode.

Previous studies suggested that acute respiratory infection (ARI) could trigger stroke and that use of nonsteroidal anti-inflammatory drugs (NSAIDs) was associated with increased risk of stroke. In many countries, NSAIDs have been widely used among patients with ARI or common cold for pain and fever relief. However, studies evaluating whether NSAIDs use during ARI episodes may further increase the risk of stroke were very limited. During 2007 to 2011, 29518 patients with an incident hospitalization of stroke were identified. The date of admission was defined as the index date. Using case-crossover design, we compared the following exposure status between the case period (1- to 7-d period before the index date) and matched control period (366- to 372-d period before the index date): NSAIDs use during ARI episodes, ARI episodes without NSAIDs use, NSAIDs use only, or no exposure. Multivariable conditional regression models were used to estimate odds ratios adjusting potential confounders. The results suggested that NSAIDs use during ARI episodes was associated with a 2.3-fold increased risk of stroke (ischemic: adjusted odds ratio, aOR 2.27, 95% confidence interval, 95% CI, 2.00-2.58; hemorrhagic: aOR 2.28, 95% CI, 1.71-3.02). We also determined that parenteral NSAIDs were associated with much higher risk of stroke in patients with ARI. Nonsteroidal anti-inflammatory drugs use during ARI episodes, especially parenteral NSAIDs use, was associated with a further increased risk of stroke.

Determinants of Unintentional Leaks During CPAP Treatment in OSA

Unintentional leakage from the mouth or around the mask may lead to cessation of CPAP treatment; however, the causes of unintentional leaks are poorly understood. The objectives of this study were (1) to identify determining factors of unintentional leakage and (2) to determine the effect of the type of mask (nasal/oronasal) used on unintentional leakage. Seventy-four polysomnograms from patients with OSA syndrome treated with auto-CPAP were analyzed (23 women; 56 ± 13 years; BMI, 32.9kg/m2 (range, 29.0-38.0kg/m2). Polysomnographic recordings were obtained under auto-CPAP, and mandibular behavior was measured with a magnetic sensor. After sleep and respiratory scoring, polysomnographic signals were computed as mean values over nonoverlapping 10-s intervals. The presence/absence of unintentional leakage was dichotomized for each 10-s interval (yes/no). Univariate and multivariate conditional regression models estimated the risk of unintentional leaks during an interval "T" based on the explanatory variables from the previous interval "T-1." A sensitivity analysis for the type of mask was then conducted. The univariate analysis showed that mandibular lowering (mouth opening), a high level of CPAP, body position (other than supine), and rapid eye movement (REM) sleep increased the risk of unintentional leaks and microarousal decreased it. In the multivariate analysis, the same variables remained independently associated with an increased risk of unintentional leakage. The sensitivity analysis showed that oronasal masks reduced the risk of unintentional leaks in cases of mouth opening and REM sleep. Mouth opening, CPAP level, sleep position, and REM sleep independently contribute to unintentional leakage. These results provide a strong rationale for the definition of phenotypes and the individual management of leaks during CPAP treatment.

Tyrosine kinase-targeting drugs-associated heart failure

Background:The impact of cancer therapies on cardiac disease in the general adult cancer survivor population is largely unknown. Our objective was to evaluate which tyrosine kinase-targeting drugs are associated with greater risk for new-onset heart failure (HF).Methods:A nested case–control analysis was conducted within a cohort of 27 992 patients of Clalit Health Services, newly treated with a tyrosine kinase-targeting, and/or chemotherapeutic drug, for a malignant disease, between 1 January 2005 and 31 December 2012. Each new case of HF was matched to up to 30 controls from the cohort on calendar year of cohort entry, age, gender, and duration of follow-up. Main outcome measure was odds ratio (OR) with 95% confidence interval (CI) of new-onset HF.Results:There were 936 incident cases of HF during 71 742 person-years of follow-up. Trastuzumab (OR 1.90, 95% CI 1.46–2.49), cetuximab (OR 1.72, 1.10–2.69), panitumumab (OR 3.01, 1.02–8.85), and sunitinib (OR 3.39, 1.78–6.47) were associated with increased HF risk. Comorbidity independently associated with higher risk in a multivariable conditional regression model was diabetes mellitus, hypertension, chronic renal failure, ischaemic heart disease, valvular heart disease, arrhythmia, and smoking.Conclusions:Trastuzumab, cetuximab, panitumumab, and sunitinib are associated with increased risk for new-onset HF.

Corticosteroid Risk Function of Severe Infection in Primary Immune Thrombocytopenia Adults. A Nationwide Nested Case-Control Study

Corticosteroid (CS)-related infection risk in immune thrombocytopenia (ITP) is unknown. The aim of this study was to assess the adjusted CS risk function of severe infection in persistent or chronic primary ITP adults. We designed a nested case-control study in the FAITH cohort. This cohort is built through the French national health insurance database named SNIIRAM and includes all treated incident persistent or chronic primary ITP adults in France (ENCePP n°4574). Patients who entered the FAITH cohort between 2009 and 2012 were eligible (n = 1805). Cases were patients with infection as primary diagnosis code during hospitalization. Index date was the date of first hospitalization for infection. A 2:1 matching was performed on age and entry date in the cohort. Various CS exposure time-windows were defined: current user, exposure during the 1/3/6 months preceding index date and from the entry date. CS doses were converted in prednisone equivalent (PEQ). The cumulative CS doses were averaged in each time-window to obtain daily PEQ dosages. Each CS exposure definition was assessed using multivariate conditional regression models. During the study period, 161 cases (9 opportunistic) occurred. The model with the best goodness of fit was CS exposure during the month before the index date (OR: 2.48, 95% CI: 1.61–3.83). The dose-effect relation showed that the risk existed from averaged daily doses ≥5 mg PEQ (vs. <5 mg: 2.09, 95% CI: 1.17–3.71). The risk of infection was mainly supported by current or recent exposure to CS, even with low doses.

Joint estimation of sparse multivariate regression and conditional graphical models

Multivariate regression model is a natural generalization of the classical univariate regression model for fitting multiple responses. In this paper, we propose a high-dimensional multivariate conditional regression model for constructing sparse estimates of the multivariate regression coefficient matrix that accounts for the dependency structure among the multiple responses. The proposed method decomposes the multivariate regression problem into a series of penalized conditional log-likelihood of each response conditional on the covariates and other responses. It allows simultaneous estimation of the sparse regression coefficient matrix and the sparse inverse covariance matrix. The asymptotic selection consistency and normality are established for the diverging dimension of the covariates and number of responses. The effectiveness of the proposed method is demonstrated in a variety of simulated examples as well as an application to the Glioblastoma multiforme cancer data.

Acute pancreatitis in patients with type 2 diabetes mellitus treated with dipeptidyl peptidase-4 inhibitors: a population-based nested case-control study.

Concern about an increasing risk of acute pancreatitis associated with incretin-based drugs, including dipeptidyl peptidase-4 (DPP-4) inhibitors and glucagon-like peptide-1 analogs, has emerged recently. This nested case-control study examined the association between the use of DPP-4 inhibitors and acute pancreatitis using Taiwan's National Health Insurance Research Database. From a study cohort of patients with type 2 diabetes mellitus, we identified 1,957 acute pancreatitis cases (patients who had been admitted with a diagnosis of acute pancreatitis) and 7,828 age-, sex-, and cohort entry year-matched controls between 2000 and 2011. Multivariate conditional regression models were used to estimate the association between the use of DPP-4 inhibitors and acute pancreatitis. Sensitivity analyses were conducted by varying the definitions of timing of exposure to DPP-4 inhibitors. The risks of acute pancreatitis among current and past users of DPP-4 inhibitors were comparable with those of non-users (current users: adjusted odds ratio (aOR) 1.04; 95% CI [0.89-1.21]; past users: aOR 1.61 [0.93-2.77]). Similar results were found in sensitivity analyses with various definitions of "current users" of DPP-4 inhibitors. Nevertheless, the adjusted risk of acute pancreatitis was found to be increased significantly in patients with gallstone disease (aOR 5.89 [4.71-7.35]), alcohol-related disease (aOR 5.36 [4.05-7.08]), hypertriglyceridemia (aOR 1.80 [1.26-2.56]), pancreatic disease (aOR 17.29 [10.60-28.19]), and a higher Diabetes Complications Severity Index (DCSI) score (DCSI 3-4: aOR 1.49 [1.21-1.84]; DCSI≥5: aOR 1.32 [1.01-1.73]). This population-based study extends previous evidence by exploring the potential association between DPP-4 inhibitor use and the risk of acute pancreatitis in an ethnic Chinese type 2 diabetic cohort. We found that underlying diseases and severity of diabetes but not DPP-4 inhibitor use were associated with acute pancreatitis.

Statins and the risk of liver injury: a population-based case-control study.

This case-control study investigated the association between statin use and liver injury using Taiwan's National Health Insurance Research Database. Our study subjects included 4165 cases (patients who had been admitted with a primary diagnosis of liver injury between 2002 and 2009) and 16 660 age-matched, sex-matched and index date-matched controls. Multivariable conditional regression models were used to estimate the association between statin use and liver injury. Users of statins were not associated with risk of liver injury (adjusted odds ratio [aOR] 1.04; 95% confidence interval [0.90-1.19]) when compared with nonusers. Nevertheless, a higher dose of statin (≥1 defined daily dose; aOR 1.55 [1.14-2.11]) and use of rosuvastatin before event of liver injury (aOR 1.38 [1.03-1.85]) were significantly associated with liver injury. This population-based study extends previous evidence by exploring the potential association between statins use and risk of liver injury. Overall, we found that statin was not associated with risk of liver injury. Nevertheless, special concern should be paid to those who used statin ≥1 defined daily dose and rosuvastatin.

Study on environmental and psychological risk factors for female infertility

To investigate the environmental and psychological risk factors for female infertility and to provide a scientific basis for the prevention and control of female infertility. In a hospital-based case-control study, a self-designed questionnaire was used to survey the cases and controls (1:1) with nation and age (± 2 years) as matching variables. Univariate and multivariate conditional logistic regression models were employed to analyze the datasets. The univariate analysis showed that female infertility was related to the following factors: eating fried foods, alcohol consumption, smoking, staying up late, perm, housing decoration, contact with heavy metals, exposure to radiation, contact with pesticides, working in hot environment, mental stress, uneasiness, helplessness, and despair. The multivariate analysis showed that staying up late (OR = 2.937), housing decoration (OR = 2.963), exposure to radiation (OR = 2.506), contact with pesticides (OR = 2.908), and mental stress (OR = 4.101) were the main risk factors for female infertility. Furthermore, there was an interaction between staying up late and mental stress. Female infertility is caused by multiple factors including staying up late, housing decoration, exposure to radiation, contact with pesticides, and mental stress, and there is an interaction between staying up late and mental stress.