Logistic Regression Risk Model Research Articles

Dose-response associations between physical activity and mental health among USA adolescents

Abstract Background Adolescent mental health is a critical public health concern. Physical activity has been shown to reduce risk of negative mental health outcomes. However, in adolescents little is known about the dose-response nature of this relationship. This study aims to fill this gap by examining the association between varying levels of self-reported physical activity and the odds of reporting depressive symptoms and poor mental health, among US high school students. Methods This is a cross-sectional study using data from the 2021 Youth Risk Behavior Survey. Physical activity exposure was assessed as students’ self-reported weekly frequency of ≥ 60min moderate-to-vigorous physical activity (MVPA), as well as frequency of participation in muscle-strengthening activities, sports teams, and physical education (PE) classes. For the dose-response analysis three categories were created for each item. Depressive symptoms and poor mental health were measured using established survey items. Multivariate logistic regression risk models were used. Results A total of 8,015 high-school students were included. Depressive symptoms were reported by 43% of students, and 28% reported poor mental health. Moderate doses (1-4 days) of muscle-strengthening activities reduced the odds of depressive symptoms by 23% and the odds of poor mental health by 38% compared to low doses (0 days). Higher reductions were observed for higher doses (5-7 days). Similar results were observed for sports teams (1 vs 0 teams). Moderate doses (1-4 days) of MVPA ≥60min showed a significant reduction in the odds for poor mental health (38%) but not depressive symptoms compared to low doses (0 days). No effect was observed for any dose of PE attendance. Conclusions The findings suggest a dose-response pattern in the association between physical activity and mental health in adolescents. Any increase in physical activity levels among young adults would be valuable in preventing poor mental health. Key messages • A dose-response pattern was observed between self-reported mental health and physical activity exposure in adolescents. • Moderate levels of physical activity significantly reduced the risk of self-reported poor mental health among adolescents.

Risk model for morbidity and mortality following liver surgery based on a national Japanese database.

We evaluated the morbidity and mortality associated with liver surgery in Japan and developed a risk model for liver resection using information from a national database. We retrospectively reviewed 73 861 Japanese patients who underwent hepatectomy between 2014 and 2019, using information from the National Clinical Database (NCD) registrations. The primary endpoints were 30 days and in-hospital mortality, and the secondary endpoints were postoperative complications. Logistic regression risk models for postoperative morbidity and mortality after hepatectomy were constructed based on preoperative clinical parameters and types of liver resection, and validated using a bootstrapping method. The 30-day and in-hospital mortality rates were 0.9% and 1.7%, respectively. Trisectionectomy, hepatectomy for gallbladder cancer, hepatectomy for perihilar cholangiocarcinoma, and poor activities of daily living were statistically significant risk factors with high odds ratios for both postoperative morbidity and mortality. Internal validations indicated that the c-indices for 30-day and in-hospital mortality were 0.824 and 0.839, respectively. We developed a risk model for liver resection by using a national surgical database that can predict morbidity and mortality based on preoperative factors.

Construction and validation of a nomogram to predict left ventricular hypertrophy in low-risk patients with hypertension.

Electrocardiography (ECG) is an accessible diagnostic tool for screening patients with hypertensive left ventricular hypertrophy (LVH). However, its diagnostic sensitivity is low, with a high probability of false-negatives. Thus, this study aimed to establish a clinically useful nomogram to supplement the assessment of LVH in patients with hypertension and without ECG-LVH based on Cornell product criteria (low-risk hypertensive population). A cross-sectional dataset was used for model construction and divided into development (n=2906) and verification (n=1447) datasets. A multivariable logistic regression risk model and nomogram were developed after screening for risk factors. Of the 4353 low-risk hypertensive patients, 673 (15.4%) had LVH diagnosed by echocardiography (Echo-LVH). Eleven risk factors were identified: hypertension awareness, duration of hypertension, age, sex, high waist-hip ratio, education level, tea consumption, hypochloremia, and other ECG-LVH diagnostic criteria (including Sokolow-Lyon, Sokolow-Lyon products, and Peguero-Lo Presti). For the development and validation datasets, the areas under the curve were 0.724 (sensitivity=0.606) and 0.700 (sensitivity=0.663), respectively. After including blood pressure, the areas under the curve were 0.735 (sensitivity=0.734) and 0.716 (sensitivity=0.718), respectively. This novel nomogram had a good predictive ability and may be used to assess the Echo-LVH risk in patients with hypertension and without ECG-LVH based on Cornell product criteria.

Predictive value of hemoglobin to serum creatinine ratio combined with serum uric acid for in-hospital mortality after emergency percutaneous coronary intervention in patients with acute myocardial infarction

To investigate the clinical value of hemoglobin to serum creatinine ratio (Hb/SCr) combined with blood uric acid (SUA) in predicting in-hospital mortality after emergency percutaneous coronary intervention (PCI) in patients with acute myocardial infarction (AMI). The clinical data of AMI patients who underwent emergency PCI in the First Affiliated Hospital of Kangda College of Nanjing Medical University from January 2017 to December 2021 were retrospectively analyzed. The general information, underlying medical history, blood routine, liver and kidney function, blood coagulation routine, SUA and other indicators were collected from patients. The primary composite endpoint was defined as in-hospital death, including in-hospital all-cause death during PCI and 15-day post-procedure hospitalization. Multivariate Logistic regression was used to analyze factors associated with in-hospital death after emergency PCI in patients with AMI. Multivariate Logistic regression was used to analyze the independent related factors and construct a risk prediction model. The Hosmer-Lemeshow method and receiver operator characteristic curve (ROC curve) were used to test the goodness of fit and predictive effect of the model and correlates, respectively. A total of 1 976 patients were enrolled, 92 died in hospital and 1 884 survived. SUA was higher in the death group than that in the survival group (μmol/L: 476.88±132.04 vs. 354.87±105.31, P < 0.01), and the Hb/SCr ratio was significantly lower than that in the survival group (13.84±5.48 vs.19.20±5.74, P < 0.01). Pearson analysis showed a linear negative correlation between SUA and Hb/SCr ratio (r = -0.502, P < 0.01). Logistic regression risk model analysis finally included age [odds ratio (OR) = 0.916], Hb/SCr ratio (OR = 0.182), white blood cell count (WBC, OR = 2.733), C-reactive protein (CRP, OR = 3.611), SUA (OR = 4.667), blood glucose (Glu, OR = 2.726), homocysteine (Hcy, OR = 2.688) 7 factors to construct a risk prediction model, which were independent correlation factors for in-hospital death in AMI patients after emergency PCI (all P < 0.05). Hosmer-Lemeshow test verified the fitting effect of the model, and the result showed P = 0.447. The area under the ROC curve (AUC) of the model for predicting in-hospital death in AMI patients after emergency PCI was 0.764 [95% confidence interval (95%CI) was 0.712-0.816, P = 0.001]. When the cut-off value was 0.565 8, the sensitivity was 70.7%, the specificity was 70.2%, and the Yoden index was 0.410. When Hb/SCr ratio+SUA, SUA, Hb/SCr ratio, Hb and SCr were used to predict in-hospital death in AMI patients after emergency PCI, the AUC of Hb/SCr ratio+SUA was the largest, which was 0.810. When the optimal cut-off value was -0.847, the sensitivity was 77.7%, the specificity was 74.5%, and the Youden index was 0.522. Age, SUA, Hb/SCr ratio, WBC, CRP, Glu, and Hcy are independent risk factors for in-hospital death after emergency PCI in AMI patients. The lower the Hb/SCr ratio and the higher the SUA at admission, the higher the risk of in-hospital death after emergency PCI in AMI patients. Hb/SCr ratio combined with SUA has a higher predictive value for in-hospital death after emergency PCI in AMI patients than single index, which is helpful for early identification of high-risk patients.

Acute seizure risk in patients with encephalitis: development and validation of clinical prediction models from two independent prospective multicentre cohorts

ObjectiveIn patients with encephalitis, the development of acute symptomatic seizures is highly variable, but when present is associated with a worse outcome. We aimed to determine the factors associated with...

Age Specific Models to Capture the Change in Risk Factor Contribution by Age to Short Term Primary Ischemic Stroke Risk.

Age is one of the most important risk factors when it comes to stroke risk prediction. However, including age as a risk factor in a stroke prediction model can give rise to a number of difficulties. Age often dominates the risk score, and also not all risk factors contribute proportionally to stroke risk by age. In this study we investigate a number of common stroke risk factors, using Framingham heart study data from the NHLBI Biologic Specimen and Data Repository Information Coordinating Center to determine if they appear to contribute proportionally by age to a stroke risk score. As we find evidence that there is some non-proportionality by age, we then create a set of logistic regression risk models that each predict the 5 year stroke risk for a different age group. The age group models are shown to be better calibrated when compared to a model for all ages that includes age as a risk factor. This suggests that to get better predictions for stroke risk it may be necessary to consider alternative methods for including age in stroke risk prediction models that account for the non-proportionality of the other risk factors as age changes.

Development and Validation of a Risk Prediction Model for Acute-on-Chronic Liver Failure in Chronic Hepatitis B Patients with Severe Acute Exacerbation: A Multi-Center Study

Background: Chronic hepatitis B patients with severe acute exacerbation is at progression stage of acute-on-chronic liver failure (ACLF). We aimed to develop and validate an accurate risk model to early identify the patients at high-risk of liver failure and predict patient’s survival. Methods: We selected the best variable combination using recursive feature elimination algorithm to develop and validate a multivariate logistic regression risk model as well as online application on cloud server from the training cohort with a total of 342 patients and two external cohorts with a sample size of 96 and 65 patients, respectively. Findings: An excellent prediction model called PATA model including four predictors, prothrombin time, age, total bilirubin and alanine aminotransferase could achieve AUROC of 0.959 (95% CI 0.941-0.977) in the development set and AUROC of 0.932 (95% CI 0.876-0.987) and 0.905 (95% CI 0.826-0.984) in two external validation cohorts, respectively. The C-index of the model was 0.720 (95% CI 0.675-0.765) in prognostic stratification of 90-days mortality for training set patients, while the end-stage liver disease score model had a C-index of 0.549 (95% CI 0.506-0.592). Interpretation: The highly-predictive risk model and easy-to-use online application can accurately predict the risk of ACLF and patient survival, which is expected to become a widely accepted prediction model and guide therapeutic options. Funding This work was supported by a grant for National Key RD National Natural Science Foundation of China; Major Science and Technology Projects in 13th Five-Year and Natural Science Foundation of Guangdong. Declaration of Interest: The authors declare that they have no competing interests. Ethical Approval: This study was approved by the Ethics Committee of the Third Affiliated Hospital of Sun Yat-sen University ([2015]2-206 No.1). This study was conducted according to the Declaration of Helsinki. All adult participants provided written informed consent.

Learning curve in laparoscopic liver surgery: A single-surgeon risk-adjusted CUSUM analysis for postoperative complications of the first 320 laparoscopic liver resections

Background: The Learning curve and the amount of procedures required to reach the 'mastery phase' in laparoscopic liver resections(LLR) are unknown. This study aimed to evaluate the short-term outcomes and investigate the learning curve(LC) during the first 320LLR performed by a single surgeon. Materials & Methods: Retrospective review of a prospective database of all patients who underwent LLR between October 2011 and November 2018. The IWATE Difficulty Score(IDS) was used to calculate a logistic regression risk model do adjust CUSUM for postoperative complications. Results: During 7years, 320/480 liver resections (66.7%) were performed by a pure laparoscopic approach. During these years, percentage of LLR increased from 45.0% to 69.2%(p=0.33). Malignancy was present in 275/320LLL (85.9%). Laparoscopic repeat hepatectomy was performed in 45/320(14.1%) cases. Median blood loss was 100ml(IQR:50-250ml) and median operative time was 130min (IQR:90-180min). Conversion rate was 2.5% and blood transfusion rate was 2.8%. Major complication rate (Clavien-Dindo=III) was 3.1% with 0.94% 90-day mortality. R0 resection rate was 93.6%. The single-surgeon CUSUM LC for postoperative complications showed the presence of three phases: an initiation phase of approximately 50 cases,an intermediate phase and a mastery phase beginning at approximately 140procedures. Important differences were found in complication rates for patients with different IDS (3.4% in IDS=0 VS 39.1% in IDS=10). A risk-adjusted CUSUM (RA-CUSUM) analysis showed an initial learning phase of 20 procedures,a plateau phase and a mastery phase after 60 cases with very low complication rates after 100 procedures. Conclusion: Even after a dedicated fellowship, with excellent short-term outcomes in early experience, a learning curve is present in laparoscopic liver surgery. The mastery phase is reached at approximately 60 cases. CUSUM and RA-CUSUM have an important role in monitoring surgical outcomes. However, adjustment for patient mix is critical. Difficulty scores can be used to adjust the CUSUM chart for different patient characteristics that significantly afffect the risk. CUSUM and RA-CUSUM charts for postoperative complications

Development and Validation of a Multivariable Lung Cancer Risk Prediction Model That Includes Low-Dose Computed Tomography Screening Results

Low-dose computed tomography lung cancer screening is most effective when applied to high-risk individuals. To develop and validate a risk prediction model that incorporates low-dose computed tomography screening results. A logistic regression risk model was developed in National Lung Screening Trial (NLST) Lung Screening Study (LSS) data and was validated in NLST American College of Radiology Imaging Network (ACRIN) data. The NLST was a randomized clinical trial that recruited participants between August 2002 and April 2004, with follow-up to December 31, 2009. This secondary analysis of data from the NLST took place between August 10, 2013, and November 1, 2018. Included were LSS (n = 14 576) and ACRIN (n = 7653) participants who had 3 screens, adequate follow-up, and complete predictor information. Incident lung cancers occurring 1 to 4 years after the third screen (202 LSS and 96 ACRIN). Predictors included scores from the validated PLCOm2012 risk model and Lung CT Screening Reporting & Data System (Lung-RADS) screening results. Overall, the mean (SD) age of 22 229 participants was 61.3 (5.0) years, 59.3% were male, and 90.9% were of non-Hispanic white race/ethnicity. During follow-up, 298 lung cancers were diagnosed in 22 229 individuals (1.3%). Eight result combinations were pooled into 4 groups based on similar associations. Adjusted for PLCOm2012 risks, compared with participants with 3 negative screens, participants with 1 positive screen and last negative had an odds ratio (OR) of 1.93 (95% CI, 1.34-2.76), and participants with 2 positive screens with last negative or 2 negative screens with last positive had an OR of 2.66 (95% CI, 1.60-4.43); when 2 or more screens were positive with last positive, the OR was 8.97 (95% CI, 5.76-13.97). In ACRIN validation data, the model that included PLCOm2012 scores and screening results (PLCO2012results) demonstrated significantly greater discrimination (area under the curve, 0.761; 95% CI, 0.716-0.799) than when screening results were excluded (PLCOm2012) (area under the curve, 0.687; 95% CI, 0.645-0.728) (P < .001). In ACRIN validation data, PLCO2012results demonstrated good calibration. Individuals who had initial negative scans but elevated PLCOm2012 six-year risks of at least 2.6% did not have risks decline below the 1.5% screening eligibility criterion when subsequent screens were negative. According to this analysis, some individuals with elevated risk scores who have negative initial screens remain at elevated risks, warranting annual screening. Positive screens seem to increase baseline risk scores and may identify high-risk individuals for continued screening and enrollment into clinical trials. ClinicalTrials.gov Identifier: NCT00047385.

Construction of a risk model through the fusion of experimental data and finite element modeling: Application to car crash-induced TBI

This article introduces a new approach for the construction of a risk model for the prediction of Traumatic Brain Injury (TBI) as a result of a car crash. The probability of TBI is assessed through the fusion of an experiment-based logistic regression risk model and a finite element (FE) simulation-based risk model. The proposed approach uses a multilevel framework which includes FE simulations of vehicle crashes with dummy and FE simulations of the human brain. The loading conditions derived from the crash simulations are transferred to the brain model thus allowing the calculation of injury metrics such as the Cumulative Strain Damage Measure (CSDM). The framework is used to propagate uncertainties and obtain probabilities of TBI based on the CSDM injury metric. The risk model from FE simulations is constructed from a support vector machine classifier, adaptive sampling, and Monte-Carlo simulations. An approach to compute the total probability of TBI, which combines the FE-based risk assessment as well as the risk prediction from the experiment-based logistic regression model is proposed. In contrast to previous published work, the proposed methodology includes the uncertainty of explicit parameters such as impact conditions (e.g., velocity, impact angle), and material properties of the brain model. This risk model can provide, for instance, the probability of TBI for a given assumed crash impact velocity.

Association Study with 77 SNPs Confirms the Robust Role for the rs10830963/G of MTNR1B Variant and Identifies Two Novel Associations in Gestational Diabetes Mellitus Development.

ContextGenetic variation in human maternal DNA contributes to the susceptibility for development of gestational diabetes mellitus (GDM).ObjectiveWe assessed 77 maternal single nucleotide gene polymorphisms (SNPs) for associations with GDM or plasma glucose levels at OGTT in pregnancy.Methods960 pregnant women (after dropouts 820: case/control: m99’WHO: 303/517, IADPSG: 287/533) were enrolled in two countries into this case-control study. After genomic DNA isolation the 820 samples were collected in a GDM biobank and assessed using KASP (LGC Genomics) genotyping assay. Logistic regression risk models were used to calculate ORs according to IADPSG/m’99WHO criteria based on standard OGTT values.ResultsThe most important risk alleles associated with GDM were rs10830963/G of MTNR1B (OR = 1.84/1.64 [IADPSG/m’99WHO], p = 0.0007/0.006), rs7754840/C (OR = 1.51/NS, p = 0.016) of CDKAL1 and rs1799884/T (OR = 1.4/1.56, p = 0.04/0.006) of GCK. The rs13266634/T (SLC30A8, OR = 0.74/0.71, p = 0.05/0.02) and rs7578326/G (LOC646736/IRS1, OR = 0.62/0.60, p = 0.001/0.006) variants were associated with lower risk to develop GDM. Carrying a minor allele of rs10830963 (MTNR1B); rs7903146 (TCF7L2); rs1799884 (GCK) SNPs were associated with increased plasma glucose levels at routine OGTT.ConclusionsWe confirmed the robust association of MTNR1B rs10830963/G variant with GDM binary and glycemic traits in this Caucasian case-control study. As novel associations we report the minor, G allele of the rs7578326 SNP in the LOC646736/IRS1 region as a significant and the rs13266634/T SNP (SLC30A8) as a suggestive protective variant against GDM development. Genetic susceptibility appears to be more preponderant in individuals who meet both the modified 99’WHO and the IADPSG GDM diagnostic criteria.

Managing pregnancy of unknown location based on initial serum progesterone and serial serum hCG levels: development and validation of a two-step triage protocol.

A uniform rationalized management protocol for pregnancies of unknown location (PUL) is lacking. We developed a two-step triage protocol to select PUL at high risk of ectopic pregnancy (EP), based on serum progesterone level at presentation (step 1) and the serum human chorionic gonadotropin (hCG) ratio, defined as the ratio of hCG at 48 h to hCG at presentation (step 2). This was a cohort study of 2753 PUL (301 EP), involving a secondary analysis of prospectively and consecutively collected PUL data from two London-based university teaching hospitals. Using a chronological split we used 1449 PUL for development and 1304 for validation. We aimed to assign PUL as low risk with high confidence (high negative predictive value (NPV)) while classifying most EP as high risk (high sensitivity). The first triage step assigned PUL as low risk using a threshold of serum progesterone at presentation. The remaining PUL were triaged using a novel logistic regression risk model based on hCG ratio and initial serum progesterone (second step), defining low risk as an estimated EP risk of < 5%. On validation, initial serum progesterone ≤ 2 nmol/L (step 1) classified 16.1% PUL as low risk. Second-step classification with the risk model selected an additional 46.0% of all PUL as low risk. Overall, the two-step protocol classified 62.1% of PUL as low risk, with an NPV of 98.6% and a sensitivity of 92.0%. When the risk model was used in isolation (i.e. without the first step), 60.5% of PUL were classified as low risk with 99.1% NPV and 94.9% sensitivity. PUL can be classified efficiently into being either high or low risk for complications using a two-step protocol involving initial progesterone and hCG levels and the hCG ratio. Copyright © 2016 ISUOG. Published by John Wiley & Sons Ltd.

MP39-02 DEVELOPMENT AND VALIDATION OF A RISK PREDICTION MODEL FOR PROSTATE CANCER DIAGNOSIS WITHIN A STATEWIDE QUALITY IMPROVEMENT COLLABORATIVE

You have accessJournal of UrologyProstate Cancer: Detection & Screening V1 Apr 2016MP39-02 DEVELOPMENT AND VALIDATION OF A RISK PREDICTION MODEL FOR PROSTATE CANCER DIAGNOSIS WITHIN A STATEWIDE QUALITY IMPROVEMENT COLLABORATIVE Gregory Auffenberg, Seliln Merdan, David Miller, Karandeep Singh, Benjamin Stockton, Brian Denton, and Khurshid Ghani Gregory AuffenbergGregory Auffenberg More articles by this author , Seliln MerdanSeliln Merdan More articles by this author , David MillerDavid Miller More articles by this author , Karandeep SinghKarandeep Singh More articles by this author , Benjamin StocktonBenjamin Stockton More articles by this author , Brian DentonBrian Denton More articles by this author , and Khurshid GhaniKhurshid Ghani More articles by this author View All Author Informationhttps://doi.org/10.1016/j.juro.2016.02.127AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail INTRODUCTION AND OBJECTIVES The Prostate Cancer Prevention Trial Risk Calculator (PCPTRC) is one of several publicly available tools to predict prostate cancer (CaP) risk prior to prostate needle biopsy (PNB). We explored whether more contemporary PNB registry data from a statewide quality improvement collaborative can be used to create a risk prediction model that better predicts cancer risk. METHODS The Michigan Urological Surgery Improvement Collaborative (MUSIC) is a consortium of 42 diverse community and academic urology practices. The collaborative maintains a prospective clinical registry with detailed and validated clinical data for all patients undergoing PNB. We identified all patients undergoing their first PNB, and excluded patients without complete pre-procedural data including: age, race, CaP family history, digital rectal examination findings, PSA, and practice site. A logistic regression risk model was fit using the above clinical factors (MUSIC prediction model) to predict the risk of biopsy-proven CaP. Predictive performance was verified using 10-fold cross validation. The accuracy and calibration of the MUSIC model was compared to the performance of the PCPTRC in the same cohort. RESULTS Results from 11,803 PNBs occurring in 38 practices between 2012 and 2015 were used to fit the MUSIC regression model. Practice location was a significant factor for outcome prediction (p<0.001) and was integrated into the regression model. Discrimination was improved when using the MUSIC model (area under the curve (AUC) for MUSIC model =0.68 vs. 0.66 for PCPTRC). Moreover, on the basis of Brier score, predictions from the MUSIC risk calculator agreed with observed outcomes more frequently (Brier score = 0.22 vs. 0.29). The PCPTRC underestimated the probability of prostate cancer within the MUSIC population (Figure 1). CONCLUSIONS Application of the publicly available version of the PCPTRC underestimates cancer risk in a contemporary cohort of patients undergoing prostate biopsy. We found that a predictive model specifically tailored to our population better estimates cancer risk. Further testing will be needed to determine the generalizability of these findings. © 2016FiguresReferencesRelatedDetails Volume 195Issue 4SApril 2016Page: e541-e542 Advertisement Copyright & Permissions© 2016MetricsAuthor Information Gregory Auffenberg More articles by this author Seliln Merdan More articles by this author David Miller More articles by this author Karandeep Singh More articles by this author Benjamin Stockton More articles by this author Brian Denton More articles by this author Khurshid Ghani More articles by this author Expand All Advertisement Advertisement PDF downloadLoading ...

Analysis of risk factors and the establishment of a risk model for peripherally inserted central catheter thrombosis

Abstract Objective To investigate the main risk factors of peripherally inserted central catheter (PICC) related upper extremity deep venous thrombosis and establish the risk predictive model of PICC-related upper extremity deep venous thrombosis. Methods Patients with PICC who were hospitalized between January 2014 and July 2015 were studied retrospectively; they were divided into a thrombosis group (n = 52), with patients who had a venous thrombosis complication after PICC, and a no-thrombosis group (n = 144), with patients without venous thrombosis. To compare between the two groups, significantly different variables were selected to perform multivariate logistic regression to establish the risk-predictive model. Results The PICC catheter history, catheter tip position, and diameter of blood vessel were the key factors for thrombosis. The logistic regression predictive model was as follows: Y = 3.338 + 2.040 × PICC catheter history +1.964× catheter tip position −1.572× diameter of vessel. The area under the receiver operating characteristic curve for the model was 0.872, 95%CI (0.817–0.927). The cut-off point was 0.801, the sensitivity of the model was 0.832, and the specificity was 0.745. Conclusions The PICC catheterization history, catheter tip position, the diameter of blood vessel were the key factors for thrombosis. The logistic regression risk model based on these factors is reliable for predicting PICC-related upper extremity deep venous thrombosis.

Strabismus at Age 2 Years in Children Born Before 28 Weeks' Gestation: Antecedents and Correlates.

Children born very preterm are at greater risk of ophthalmic morbidities, including strabismus, than children born at term. We evaluated perinatal factors associated with strabismus at age 2 years in a large population of infants delivered before 28 weeks' gestation. A total of 996 infants in the multicenter ELGAN (Extremely Low Gestational Age Newborn) study who had a retinal exam in infancy and a developmental assessment at 2 years corrected age are included. Their mothers were interviewed about the pregnancy, and both mother and newborn charts were reviewed. Certified examiners administered the Bayley Scales of Infant Development-II and performed an examination of ocular alignment. Time-oriented logistic regression risk models were created to evaluate the associations of characteristics and exposures with the development of strabismus. Overall, 14% (n = 141) of the children had strabismus at 2 years, and 80% of strabismic children had esotropia. Characteristics associated with strabismus were birth before 26 weeks' gestation, severe fetal growth restriction, and maternal history of aspirin ingestion. Associated postnatal factors included a SNAP-II (Score for Neonatal Acute Physiology) illness severity value ≥ 30, brain ventriculomegaly, type I retinopathy of prematurity, and ventilator-dependent severe bronchopulmonary dysplasia. Strabismus in very preterm populations is associated with a number of antenatal and postnatal antecedents as well as clinical and imaging correlates indicative of brain damage in these children. Routine ophthalmologic assessments in the early years can allow appropriate and timely interventions.

Predictors of clinical outcomes in acute decompensated heart failure: Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure outcome models

Patients hospitalized for acute decompensated heart failure (ADHF) are at high risk for early mortality and rehospitalization. Risk stratification of ADHF using clinically available data on admission is increasingly important to integrate with clinical pathways. Our goal was to create a simple method of screening patients upon admission to identify those with increased risk of future adverse events. Using ASCEND-HF, a pragmatic clinical trial conducted in 398 sites globally, we developed and validated logistic regression risk models for (a) 30-day mortality/HF rehospitalization, (b) 30-day mortality/all-cause rehospitalization, (c) 30-day all-cause mortality, and (d) 180-day all-cause mortality. Fifty-one candidate variables were evaluated based on prior publications and clinical review. Final models were selected based on stepwise selection with entry and a staying criterion of P < .01. The 30-day mortality model was externally validated, and coefficients were converted to an additive risk score. Among 7,141 patients, the median age was 67 years, 34% were female, and 80% had a left ventricular ejection fraction <40%. The models had between 5 and 12 risk factors with c-indices ranging from 0.68 to 0.75. A simplified score, including age, systolic blood pressure, sodium, blood urea nitrogen, and dyspnea at rest, discriminated 30-day mortality risk from 0.5% (score 0) to 53% (score 10). Commonly available clinical variables provide simple risk stratification for clinical outcomes among patients with ADHF, and these models may be considered for integration into routine clinical care.

Antecedents and correlates of visual field deficits in children born extremely preterm

AimWe sought to identify the antecedents and correlates of visual field deficits (VFDs) at age 2 years among infants born before the 28th week of gestation. MethodsThe visual fields of 1023 infants were assessed by confrontation at age 2 years. We compared the ante-and postnatal characteristics and exposures of the 65 infants with a VFD to their peers who did not have a VFD. We used time-oriented logistic regression risk models to assess the associations of potential antecedents and correlates with a VFD. ResultsIn the final regression model, VFD was associated with maternal consumption of aspirin during the current pregnancy, recurring/persistent acidemia during the first 3 postnatal days, cerebral ventriculomegaly seen on neonatal ultrasound, prethreshold retinopathy of prematurity (ROP), and supplemental oxygen and ventilator dependence at 36 weeks post-menstrual age. Birth before the 27th week was also associated with increased risk, but its significance was diminished by the addition of postnatal variables. ConclusionIn this sample of extremely preterm born infants, antenatal as well as early and late postnatal characteristics and exposures are associated with an increased risk of having a VFD. Our study adds to our knowledge about the complex etiology of visual deficits of prematurity, and supports a multifactorial cause of these deficits.

Antenatal and Early Postnatal Antecedents of Parent-Reported Attention Problems at 2 Years of Age

To assess antenatal and early postnatal antecedents of attention problems identified by the Child Behavior Checklist in extremely preterm children. In a cohort of 826 children born between 23 and 27 weeks' gestation, we collected demographic, birth, and postnatal information. We then identified behavior problems by using parent ratings from the Child Behavior Checklist at 2 years' adjusted age. We created time-oriented logistic regression risk models to identify significant risk factors for attention problems and Diagnostic and Statistical Manual of Mental Disorders-compatible attention deficit/hyperactivity problems (ADHP(DSM)). Children were at increased risk of both attention problems if they were born to a woman who had no formal education beyond high school and/or a woman who was exposed to secondhand smoke. Recovery of a single organism from the placenta was associated with increased risk of an attention problem, and fetal stem vessel thrombosis and recovery of Mycoplasma species were associated with increased risk of ADHP(DSM). Infants of multifetal gestations were at reduced risk of both attention problems. The only postnatal risk factor for an attention problem was recovery of bacteria from a tracheal aspirate. Among extremely preterm infants, several potentially modifiable antenatal and perinatal antecedents are associated with increased risk for attention problems and ADHP(DSM) at 2 years adjusted age.

Common Genetic Variants Associated with Breast Cancer in Korean Women and Differential Susceptibility According to Intrinsic Subtype

Recently identified genetic variants from genome-wide association studies (GWAS) on breast cancer have not been validated in Asian populations, except in China. In this study, we sought to confirm the association between known variants and breast cancer in Korean women and further evaluate the associations of individual single nucleotide polymorphisms (SNP) with different intrinsic subtypes of breast cancer. In total, 3,321 invasive breast cancer patients and 3,500 healthy controls were genotyped for five SNPs by using the TaqMan assay. The SNPs genotyped included rs2046210 (6q25.1), rs2981582 (FGFR2), rs889312 (MAP3K1), rs3803662 (TOX3/TNRC9), and rs4973768 (SLC4A7). Tumors were classified into four intrinsic subtypes based on estrogen receptor (ER), progesterone receptor, HER2, and Ki67 expression. All five SNPs were significantly associated with risk of breast cancer in dominant, recessive, and additive models. ORs per risk allele (95% CI) were 1.29 (1.16-1.43), 1.40 (1.18-1.68), 1.22 (1.06-1.41), 1.52 (1.30-1.77), and 1.20 (1.08-1.33) for rs2046210, rs2981582, rs889312, rs3803662, and rs4973768, respectively. A multigene logistic regression risk model was generated with the SNPs. In subtype analysis, all 5 SNPs were associated with the Luminal A subtype. Two SNPs (rs2046210 and rs3803662) were linked to the ER(-)HER2(+) subtype, and only rs2046210 SNP was associated with the triple-negative subtype. The five SNPs from GWAS were significantly associated with breast cancer risk in Korean women. Associations were heterogeneous according to the intrinsic subtype of breast cancer. Our result is an important contribution to the literature about genetic susceptibility for breast cancer in nonwhite populations.

Abstract P3-10-19: Breast Cancer Risk Prediction Model in Korean Women Using Five Polymorphisms Identified in Genome Wide Association Studies

Abstract Background: Recently indentified genetic variants from genome-wide association studies of breast cancer have not been validated in Asian population except for Chinese. We sought to confirm the association in ethnically distinct Korean women and to make genetic risk assessment model using multi-gene markers Materials and Methods: 3321 invasive breast cancer patients operated in Seoul National University Hospital and 3500 healthy control women from a population based cohort were genotyped for 5 single nucleotide polymorphisms (SNPs) using Taqman assay. The SNPs genotyped were rs2046210 (6q25.1), rs4973768 (3p), rs2981582 (FGFR2), rs3803662 (TNRC9), and rs889312 (MAP3K1). Results: The five SNPs were significantly associated with the risk of breast cancer in dominant, recessive, and co-dominant model (p-values from 0.012 to 3.41E-08). Minor allele frequencies were between 0.2 and 0.5. Odds ratios were between 1.14 and 1.51. Multi-gene logistic regression risk model was made with this 5 SNPs. Women who have all protective variants showed odds ratio of 0.43, while women who are homozygous for risk variants in all 5 SNPs showed odds ratio of 2.36 compared with women with the most common genotype. Conclusion: We found that 5 SNPs from GWAS in mostly women of European ancestry were also significantly associated with breast cancer risk in Korean women. The multi-gene risk assessment model might be useful to classify women into relevant risk groups. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P3-10-19.