Multivariate Multilevel Analysis Research Articles

Timely completion of vaccination and its determinants among children in northwest, Ethiopia: a multilevel analysis

BackgroundTimely vaccination is key to prevent unnecessary childhood mortality from vaccine-preventable diseases. Despite the substantial efforts to improve vaccination completeness, the effort towards timeliness of vaccination is limited with non-attendance and delays to vaccination appointments remaining a big challenge in developing countries. There is also a limited evidence on timeliness of vaccination. Therefore, this study aimed to determine the magnitude and associated factors for timely completion of vaccination among children in Gondar city administration, north-west, Ethiopia.MethodThis study employed a community-based cross sectional study design. A sample size of 821 children aged 12 to 23 months were considered. Two stages random sampling technique was used to select study subjects. To account the effect of clustering, bivariable and multivariable multilevel logistic regression analysis were applied. The measures of association estimates were expressed as adjusted odds ratio (AOR) with their 95% confidence intervals (CIs).ResultsOf the 774 children included for analysis, 498 (64.3%) were fully vaccinated while 247 (31.9%) were fully vaccinated on-time. Caregivers who had secondary education and above (AOR = 2.391; 95% CI: 1.317–4.343), from richest households (AOR = 2.381; 95% CI: 1.502–3.773), children whose mother attended four or more ante natal care visits (AOR = 2.844; 95% CI: 1.310–6.174) and whose mother had two or more post natal care visits (AOR = 2.054; 95%CI:1.377–3.063) were positively associated with on-time full vaccination. In contrary, caregivers aged above 35 years (AOR = 0.469; 95 % CI: 0.253-0.869], being vaccinated at health post (AOR = 0.144; 95%CI: 0.048–0.428) and travelling more than 30 min to the vaccination site (AOR = 0.158; 95%CI: 0.033–0.739) were negatively associated with on-time full vaccination. The random effects indicated that 26% of the variability in on-time full vaccination was attributable to differences between communities.ConclusionIn this study, untimely vaccination was found to be high. Different individual and contextual factors were found to be associated with on-time full vaccination. Therefore, tailored strategies have to be designed and implemented to address people and the communities where they live. Moreover, timeliness of vaccination should be considered as important indicator of the immunization program performance in Ethiopia.

Socioeconomic determinants of excess weight and obesity among Indigenous women: findings from the First National Survey of Indigenous People's Health and Nutrition in Brazil.

This article assesses the nutritional status of Indigenous women from 14 to 49 years of age in Brazil. Sample size was calculated for each region considering a prevalence of 50 % for all disease outcomes, a relative error of 5 % and a CI of 95 %. In the initial data analysis, the prevalence of excess weight and obesity was calculated according to independent variables. Multivariate multilevel hierarchical analyses were conducted based on a theoretical model of two ranked blocks. The 2010 Indigenous population in Brazil was 896 000, with approximately 300 Indigenous ethnic groups, making Brazil one of the most ethnically diverse countries in the Americas and the world. Of the total target sample of 6722 women evaluated by the National Survey, thirty did not participate, 939 were not eligible for analyses due to pregnancy or unknown pregnancy status, and thirty-nine were excluded due to missing anthropometric data. The evaluation of nutritional status was completed for 5714 non-pregnant women (99·3 % of eligible participants for this outcome). High prevalence rates were encountered for both excess weight (46·2 %) and obesity (15·8 %) among the sampled women. In the multivariate analyses, higher socioeconomic indicators, market-integrated living conditions and less reliance on local food production, as well as increased age and parity were associated with excess weight and obesity. Results point to distinct patterns of associations between socioeconomic indicators and the occurrence of excess weight and obesity among Indigenous women, which have potentially significant implications from a public policy perspective for Indigenous peoples in Brazil.

Magnitude and associated factors of unintended pregnancy in Ethiopia: a multilevel analysis using 2016 EDHS data

BackgroundUnintended pregnancy has become a significant public health and reproductive health problem that has had a substantial and appreciable adverse impact on mother, child, and the general public. Despite the paramount negative effects of unintended pregnancy, many pregnancies are unintended in Ethiopia. Therefore, this study aimed to determine the prevalence and associated factors of unintended pregnancy in Ethiopia.MethodsThis study was based on the nationally representative 2016 Ethiopian Demographic and Health Survey data. We used a total weighted sample of 7590 reproductive-aged women who gave birth in the 5 years preceding the survey. A multi-level logistic regression analysis was used to account for the hierarchal nature of the DHS data. In the multivariable multilevel analysis, those variables with p-value < 0.05 were considered to be significantly associated with unintended pregnancy.ResultThe prevalence of unintended pregnancy was 26.6% [95%CI: 25.6, 27.6]. In the multivariable multilevel logistic regression analysis; individual level variables such as being in the age group 20 to 34 [adjusted odds ratio (AOR) = 0.57; 95%CI: 0.41, 0.79] and 35 to 49 [AOR = 0.68; 95%CI: 0.47, 0.97], being follower of Muslim religion [AOR = 0.73; 95%CI: 0.60, 0.88], being married [AOR = 0.46; 95%CI: 0.37–0.58], household size of four to six [AOR = 1.38; 95%CI: 1.10, 1.69] and seven and above [AOR = 1.54; 95%CI: 1.20, 1.99], and being multiparous [AOR = 1.36; 95%CI: 1.10, 1.69] and grand multiparous [AOR = 1.92; 95%CI: 1.47, 2.52] were significantly associated with unintended pregnancy. Among community level variables; being living in large central [AOR = 2.56; 95%CI: 2.06, 3.17] and metropolitan regions [AOR = 1.91; 95%CI: 1.44, 2.53] were significantly associated with unintended pregnancy.ConclusionIn this study the prevalence of unintended pregnancy was high. Maternal age, religion, marital status, household size, parity, and region were the most important factors associated with unintended pregnancy. Special attention should, therefore, be given to younger, single, multiparous and grand multiparous women, and not follower of Muslim religion as well as mothers from large central and metropolitan regions in terms of increasing accessibility and affordability of maternal health services, which could minimize unintended pregnancy.

Insight into bladder cancer care: study protocol of a large nationwide prospective cohort study (BlaZIB)

BackgroundDespite the embedding of bladder cancer management in European guidelines, large variation in clinical practice exists for applied diagnostics and treatments. This variation may affect patients’ outcomes including complications, disease recurrence, progression, survival, and health-related quality of life (HRQL). Lack of detailed clinical data and HRQL data hampers a comprehensive evaluation of bladder cancer care. Through prospective data registration, this study aims to provide insight in bladder cancer care in the Netherlands and to identify barriers and modulators of optimal bladder cancer care.MethodsThis study is a nationwide prospective cohort study including all patients who were newly diagnosed with high-risk non-muscle invasive bladder cancer (HR-NMIBC; Tis and/or T1, N0, M0/x) or non-metastatic muscle invasive bladder cancer (MIBC; ≥T2, N0/x-3, M0/x) in the Netherlands between November 1st 2017 and October 31st 2019. Extensive data on patient- and tumor characteristics, diagnostics, treatment and follow-up up to 2 years after diagnosis will be collected prospectively from electronic health records in the participating hospitals by data managers of the Netherlands Cancer Registry (NCR). Additionally, patients will be requested to participate in a HRQL survey shortly after diagnosis and subsequently at 6, 12 and 24 months. The HRQL survey includes six standardized questionnaires, e.g. SCQ Comorbidity score, EQ-5D-5 L, EORTC-QLQ-C30, EORTC-QLQ-BLM30, EORTC-QLQ-NMIBC24 and BCI. Variation in care and deviation from the European guidelines will be assessed through descriptive analyses and multivariable multilevel analyses. Survival analyses will be used to assess the association between variation in care and relevant outcomes such as survival.DiscussionThe results of this observational study will guide modifications of clinical practice and/or adaptation of guidelines and may set the agenda for new specific research questions in the management of bladder cancer.Trial registrationRetrospectively registered in the Netherlands Trial Register. Trial identification number: NL8106. Registered on October 22nd 2019.

A multivariate multilevel analysis of the risk factors associated with anthropometric indices in Iranian mid-adolescents

BackgroundThe present study was conducted to jointly assess some specific factors related to body fat measures using a multivariate multilevel analysis in a representative sample of Iranian mid-adolescents.MethodsThis study was conducted among 2538 students (1286 boys) aged 14–20 years old, who were randomly selected among 16 public high schools by multi-stage random sampling procedure from all education districts of Shiraz, Iran. Data on demographic characteristics, family history of obesity, physical activity, socio-economic (SES) variables and screen time were collected. Height, weight, triceps (TST), abdominal (AST), and subscapular (SST) skinfold thickness were measured and their body mass index (BMI) was calculated. A multivariate multilevel approach was used to analyze the factors associated with obesity measures of the TST, AST, SST at the child and district levels.ResultsIn this study, the prevalence of overweight and obesity was estimated to be 10.2 and 5.1%, respectively. Overall, the major portion of the total variance in TST (97.1%), AST (97.7%), and SST (97.5%) was found at the child level. The results of multivariate multilevel method revealed that being girls, having a family history of obesity, and SES were significantly associated with increasing of three body fat measures (all the p-values were less than 0.05). There were significant positive associations between moderate to vigorous physical activities with AST and SST (for AST: β =2.54, SE = 1.40, p = 0.05; for SST: β =2.24, SE = 1.20, p = 0.05). Compared to children in 14–16 age group, children in age group 16–18 years had less TST (β = − 0.67, SE = 0.34, p = 0.04). Furthermore, other age groups and screen time did not play an important role in three outcome variables.ConclusionsThe results showed some factors that contribute to three body fat measures. Therefore, it is necessary to develop effective interventions to prevent the effects of individual and environmental undesirable factors on childhood obesity in both family and community levels.

MP38: The impact of physician handoffs on the outcomes of emergency department patients: a medical administrative database retrospective cohort study

Introduction: A physician handoff is the process through which physicians transfer the primary responsibility of a care unit. The emergency department (ED) is a fast-paced and crowded environment where the risk of information loss between shifts is significant. Yet, the impact of handoffs between emergency physicians on patient outcomes remains understudied. We performed a retrospective cohort study in the ED to determine if handed-off patients, when compared to non-handed-off patients, were at higher risk of negative outcomes. Methods: We included every adult patient first assessed by an emergency physician and subsequently admitted to hospital in one of the five sites of the CHU de Québec-Université Laval during fiscal year 2016-17. Data were extracted from the local hospital discharge database and the ED information system. Primary outcome was mortality. Secondary outcomes were incidence of ICU admission and surgery and hospital length of stay. We conducted multilevel multivariate regression analyses, accounting for patient and hospital clusters and adjusting for demographics, CTAS score, comorbidities, admitting department delay before evaluation by an emergency physician and by another specialty, emergency department crowding, initial ED orientation and handoff timing. We conducted sensitivity analyses excluding patients that had an ED length of stay &gt; 24 hours or events that happened after 72 hours of hospitalization. Results: 21,136 ED visits and 17,150 unique individuals were included in the study. Median[Q1-Q3] age, Charlson index score, door-to-emergency-physician time and ED length of stay were 71[55-83] years old, 3[1-4], 48 [24,90] minutes, 20.8[9.9,32.7] hours, respectively. In multilevel multivariate analysis (OR handoff/no handoff [CI95%] or GMR[SE]), handoff status was not associated with mortality 0.89[0.77,1.02], surgery 0.95[0.85,1.07] or hospital length of stay (-0.02[0.03]). Non-handed-off patients had an increased risk of ICU admission (0.75[0.64,0.87]). ED occupancy rate was an independent predictor of mortality and ICU admission rate irrespectively of handoff status. Sensitivity and sub-group based analyses yielded no further information. Conclusion: Emergency physicians’ handoffs do not seem to increase the risk of severe in-hospital adverse events. ED occupancy rate is an independent predictor of mortality. Further studies are needed to explore the impact of ED handoffs on adverse events of low and moderate severity.

Impact of Serum Estradiol Levels Prior to Progesterone Administration in Artificially Prepared Frozen Embryo Transfer Cycles.

Background: The need for endocrine monitoring in artificial cycles for frozen embryo transfer (FET) remains unclear and, more specifically, the value of the late-proliferative phase serum estradiol (E2) levels is with conflicting evidence in current literature.Objective: To investigate whether artificial FET cycles require endocrine monitoring for the serum E2 level prior to initiation of exogenous progesterone administration after an endometrial thickness of 6.5 mm has been reached.Design: One thousand two hundred and twenty-two (n = 1,222) artificial FETs performed in a tertiary center between 2010 and 2015 were subdivided into 3 groups according to the following late-proliferative serum E2 level percentiles: ≤p10 (E2 ≤144 pg/ml; n = 124), p11–p90 (E2 from 145 to 438 pg/ml; n = 977) and >p90 (E2 >439 pg/ml; n = 121).A mixed-effects multilevel multivariable regression analysis was performed to assess the potential effect of the late-proliferative E2 level on the live birth rate (LBR).Results: The level of late-proliferative circulating E2 showed no significant difference in terms of LBR after FET. Specifically, the multivariable regression model demonstrated a LBR of 19.5% for the p11–p90 reference group, compared to 24.4% for the ≤p10 (p = 0.251) and 19.5% for the >p90 group (p = 0.989).Conclusion: In this large retrospective dataset, no association was observed between late-proliferative phase serum E2 levels and LBR following FET in artificially prepared cycles. Although, caution is warranted due to the retrospective nature of the analysis and the potential for unmeasured confounding, we argue that monitoring of the late-proliferative serum E2 levels and using them to guide clinical decision-making (e.g., medication step-up, cycle prolongation or cancelation) may be of questionable value.

Mapping non-response in a prevention program for cardiometabolic diseases in primary care: How to improve participation?

Non-response in prevention programs for cardiometabolic diseases (CMD) in primary care is often overlooked. The aim for this study was to define factors that influence the primary response to a selective CMD prevention program and to determine response-enhancing strategies that influence the willingness to participate. We conducted a non-response analysis within a randomized controlled trial evaluating a selective CMD prevention program, the study was conducted from 2013 to 2018 in Netherlands. A random sample of 5616 patients from 15 general practices were invited to complete a risk score (RS) as initial step of the program. Non-responders received an additional questionnaire. The response on the risk score was 51% (n=2872). From the 3558 non-response questionnaires sent, 786 (22%) were returned. In a multivariable multilevel regression analysis smoking was independently associated with non-response. Of all reported reasons for non-response 'forgot/no time' accounted for 45%. In total, 73% of the non-responders indicated to reconsider participation when approached differently. A personal approach by the patients' own GP, using advertisements and informative campaigns are potentially the best methods to enhance the response. Although a relatively high proportion did not respond to the invitation for the risk score, the majority of them indicated to be willing to participate if a different invitation strategy would be used. With more time and energy, response rates for CMD prevention programs could possibly increase substantially. A next logical step in this process is to test potential response enhancing strategies in research setting.

Abstract P326: Multilevel Analysis Of Concentrated Neighborhood Disadvantage And Cardiometabolic Factors In People Living With Hiv

Aging persons living with HIV (PLWH) are a vulnerable population at higher risk for chronic illnesses such as metabolic alterations, diabetes, cardiovascular disease, and cancer. Community and specifically, neighborhood, factors may play a key role in the development of comorbidities. Availability of healthy food stores, physical activity resources, walking/physical activity environment, and neighborhood socioeconomic status have all been individually associated with a higher odds of having ideal cardiovascular health scores. We hypothesized that exposure concentrated disadvantage at the community level impact cardio metabolic outcomes in PLWH. We examined this hypothesis in a cross-sectional analysis of the N ew O rleans A lcohol use in H IV [NOAH] Study of adult PLWH under care (n=365). Hypertension (HTN) was defined as a mean systolic ≥130-139 or diastolic blood pressure ≥80-89 mmHg. Metabolic syndrome (MS) was defined as having 3 of 5 conditions: HTN, low high-density lipoproteins (≤40 (men)/50 (women)), high triglycerides (≥ 150mg/dl), high glucose (≥100 mg/dl), waist circumference &gt;88(women)/102(men). Multivariable logistic regression and multilevel binomial regression analyses were performed adjusting for age, race, smoking, and viral load. The majority of participants were African American (83.6%) with a mean age 48.2 ±10.4, 68.8% male and 31.2% female. MS was prevalent in 36.1%, HTN 66.0%, diabetes 14%, dyslipidemia 28.2%, and 27.4% were obese. There was substantial clustering of cardiometabolic outcomes by neighborhood, with intra-class correlations (ICC) near or greater than 20% for BMI, diabetes, hypertension, and metabolic syndrome. Neighborhood concentrated disadvantage explained a significant proportion of the variance in diabetes (ICC drop from 18% to 8%) and the likelihood of having diabetes increases by 33% for each unit increase in neighborhood concentrated disadvantage (odds ratio = 1.33, 95% CI=1.02, 1.74). Our results suggest a significant role of environmental factors on cardio metabolic comorbidities among PLWH. There is a critical need for a deeper understanding of the roles that social determinants at the community and interpersonal levels play in chronic comorbidities in PLWH. Neighborhoods and social spaces are important contexts through which social determinants act to shape health and health behavior. (Supported by: NIH P60AA009803)

Faculty Perceptions of Gatekeeping and Student Suitability in the Context of Traditional and Online Social Work Programs

Gatekeeping and the process of assessing student suitability for social work education is increasingly being impacted by the rapid growth of online programs. It remains unclear if suitability concerns within social work programs are operationalized differently between traditional and online/hybrid programs. This quantitative study used previously established student suitability constructs of social consciousness, ethical suitability, practice suitability, and personal suitability to examine attitudes of social work educators and administrators regarding gatekeeping in their respective online/hybrid and traditional social work programs. The study utilized a multivariate multilevel regression analysis to investigate the effect of participant background characteristics regarding perceived emphasis on gatekeeping within participant’s own programs. Findings showed a large discrepancy between what was important to participants and what was perceived to be emphasized in their respective programs regarding suitability and gatekeeping. This study suggests that social work faculty and administrators need to pay particular attention to uneven gatekeeping practices related to technological and pedagogical shifts in the evolving social work education landscape.

Factors affecting current khat chewing among male adults 15\u201359\u2009years in Ethiopia, 2016: a multi-level analysis from Ethiopian Demographic Health Survey

BackgroundRegular khat chewing causes gingivitis, tooth loss, gastric disorders, cardiac complications, male impotence, sleeplessness, and several mental health problems. Based on the Ethiopian Demographic and Health Survey (EDHS) 2016, 12% of women and 27% of men have reported having ever chewed khat. Even though khat addiction is a major public health problem, studies that consider both individual level and community level factors are limited. Therefore, this study aimed to determine the prevalence and factors affecting current khat chewing among male adults in Ethiopia.MethodsData from EDHS, a community-based cross-sectional study conducted from January 18, 2016, to June 27, 2016, was used. A multistage stratified cluster sampling technique was used to select participants. Both descriptive and analytical statistics were done. Bi-variable and multivariable multilevel logistic regression analyses were performed to determine factors affecting current khat chewing. Adjusted Odds ratio (AOR) with 95% Confidence Interval (CI) for variables with P-value < 0.05 was used as a measure of association.ResultA total of 12,595 male adults were included. The prevalence of current khat chewing was 23.61% (95% CI: 22.87, 24.36). Age 20–24 years; (AOR = 2.68, 95% CI: 2.02, 3.56), being uneducated (AOR = 1.62, 95% CI: 1.10, 2.39), professional/technical/managerial job (AOR = 3.59, 95% CI: 2.18, 5.91), Muslim religion (AOR = 18.30, 95% CI: 13.54, 24.74), poorest wealth index (AOR = 0.67, 95% CI: 0.51, 0.89), being divorced (AOR = 0.38, 95% CI: 0.21, 0.69), history of alcohol drinking in the last 30 days (AOR = 2.15, 95% CI: 1.69, 2.73), and history of cigarette smoking in the last 30 days (AOR = 14.92, 95% CI: 10.88, 20.47), and Amhara region (AOR = 0.07, 95% CI: 0.04, 0.14) were significantly associated with khat chewing.ConclusionKhat chewing remains high in Ethiopia with certain regional variations. The uneducated, older age, Alcohol and cigarette users, Muslims, and professional workers were at higher risk of khat chewing whereas the poorest wealth index and being divorced reduced its risk. Policymakers should consider a multi-faceted policy approach that accounts for regional variation and the identified risk factors to alleviate the problem.

Study of migration and later life health in Nepal

PurposeThis study investigates long-term consequences of individual migration experience on later life health, specifically self-rated health and functional difficulty. Design/methodology/approachThe study uses multiple community-, household-, and individual-level data sets from the Chitwan Valley Family Study (CVFS) in Nepal. The CVFS selected a systematic probability sample of 151 neighborhoods in Western Chitwan and collected information on all households and individuals residing in the selected sample neighborhoods. This study uses data from multiple surveys featuring detailed migration histories of 1,373 older adults, and information on their health outcomes, households, and communities. FindingsResults of the multi-level multivariate analysis show a negative association between number of years of migration experience and self-rated health, and a positive association between migration and functional difficulty. These findings suggest a negative relationship between migration experience and later life health. Research limitations/implicationsAlthough we collected health outcome measures after the measurement of explanatory and control measures—a unique strength of this study—we were unable to control for baseline health outcomes. Also, due to the lack of time-varying measures of household socioeconomic status in the survey, this investigation was unable to control for measures associated with the economic prosperity hypothesis. Future research is necessary to develop panel data with appropriately timed measures. Practical implicationsThe findings provide important insights that may help shape individual's and their family's migration decisions. Originality/valueThis research provides important insight to individuals lured by potential short-term economic prospects in destination places, as well as to scholars and policy makers from migrant-sending settings that are grappling with skyrocketing medical expenses, rapid population aging, and old age security services.

Community Determinants of Physical Growth and Cognitive Development among Indian Children in Early Childhood: A Multivariate Multilevel Analysis.

Inadequate child physical growth and cognitive development share common individual-level risk factors. Less understood is how outcomes co-cluster at the community level and to what extent certain community-level characteristics influence the clustering. This study aims to quantify the extent to which child growth and development co-occur across communities, and to identify community-level characteristics associated with the clustering of the two development dimensions. We used longitudinal data from 1824 children (aged 5 years) across 98 communities in Andhra Pradesh, India in round 2 (2006) of the Young Lives study, who were followed up 3 years later in round 3 (2009). A multivariate, multilevel statistical model was estimated wherein the responses were nested within individuals, and communities. We used z-scores of height-for-age, weight-for-age, Peabody Picture Vocabulary Test, and a mathematics test in 2009 as outcome variables. At the community level, we included compositional variables representing community characteristics while controlling for child socio-demographic characteristics at the individual level. At the community level, children’s physical growth and cognitive development were strongly correlated (coefficient: 0.55–0.76) and, even after controlling for individual-level covariables, a more pronounced correlation was shown at the community level than individual level correlation. Greater local healthcare resources were associated with better physical growth. More local programs run by government and NGOs/charities were associated with higher child language skills. Local social problems were inversely associated with math scores. Our study showed that physical growth and cognitive development tended to be clustered and co-occurred within communities as well as individual children.

Trends of Hospital Palliative Care Utilization and Its Associated Factors Among Patients With Systemic Lupus Erythematosus in the United States From 2005 to 2014.

To investigate trends and associated factors of utilization of hospital palliative care among patients with systemic lupus erythematosus (SLE) and analyze its impact on length of hospital stay, hospital charges, and in-hospital mortality. Using the 2005-2014 National Inpatient Sample in the United States, the compound annual growth rate was used to investigate the temporal trend of utilization of hospital palliative care. Multivariate multilevel logistic regression analyses were performed to analyze the association with patient-related factors, hospital factors, length of stay, in-hospital mortality, and hospital charges. The overall proportion of utilization of hospital palliative care for the patient with SLE was 0.6% over 10 years. It increased approximately 12-fold from 0.1% (2005) to 1.17% (2014). Hospital palliative care services were offered more frequently to older patients, patients with high severity illnesses, and in urban teaching hospitals or large size hospitals. Patients younger than 40 years, the lowest household income group, or Medicare beneficiaries less likely received palliative care during hospitalization. Hospital palliative care services were associated with increased length of stay (β = 1.407, P < .0001) and in-hospital mortality (odds ratio, 48.18; 95% confidence interval, 41.59-55.82), and reduced hospital charge (β = -0.075, P = .009). Hospital palliative care service for patients with SLE gradually increased during the past decade in US hospitals. However, this showed disparities in access and was associated with longer hospital length of stay and higher in-hospital mortality. Nevertheless, hospital palliative care services yielded a cost-saving effect.

Ecological-transaction model approach of adolescents’ parental maltreatment and peer-bullying: the moderating role of bullying at the classroom

Introduction. The present study examined the ecological-transactional model delineated by Cicchetti and Lynch (1993) and the nature of the association between adolescents’ parental maltreatment and victimization and bullying at school.Method. Multivariate multilevel regression analyses were conducted on a sample of 2.852 adolescents between the ages of 12 and 17 from 25 secondary schools. Data was nested across 133 classrooms. Classrooms level variables and individual variables in relationship to parental maltreatment, behavioural disorders, sex, and bullying and victimization were included in the analysis.Results. Results indicated that adolescents who reported suffering violence at home, showed more vulnerability to becoming victims of bullying at school, with the relationship being moderated by the level of bullying in the classroom. It was also examined the role of behavioural disorders and a relation between these factors was found amongst the bullies but not the victims.Discussion and Conclusion. This study uses the systemic perspective of ecological-transactional model on child maltreatment to show the importance between the school and home microsystems in the perpetuation of victimization. The results imply that what children experience at home might cause emotional and behavioural differences in varying classroom climates at school. Therefore, understanding the interactions between systems’ transaction of socialisation mechanisms might contribute for effective anti-bullying programs.

Abstract 426: Factors Contributing to Out-Of-Hospital Cardiac Arrest Mortality Trends in the City of Detroit (2014-2018)

Introduction: Survival with good neurologic outcome following out-of-hospital cardiac arrest (OHCA) requires intensive post-ROSC care. In response to a recent decline in OHCA survival rates in Detroit, we conducted an analysis to identify factors contributing to this regression. Methods: CARES data from Detroit EMS transports were analyzed, including adult non-traumatic OHCA patients treated from 2014-2018 at four hospitals. Patient demographics, conditions of the arrest, and post-ROSC interventions including left heart catheterization, (LHC) and targeted temperature management, (TTM), were correlated to survival to hospital discharge and discharge with CPC 1 or 2. We calculated crude and adjusted survival rates across sites. Survival rates were adjusted for Utstein criteria using multilevel multi-variable regression analyses. Results: A total of 5,175 CARES patients were included. There was no difference in Utstein criteria between years, other than a small increase in the rate of witnessed arrest. We noted a steady decrease in LHC, and steady increase in TTM rate over the 5 years. There was a system-wide decrease in survival to hospital discharge during 2018, though the percentage of patients discharged with CPC 1 or 2 was similar. We found a strong, though insignificant, trend for LHC rate and survival: 0.72 (p = 0.1057), R 2 : 0.88, Adjusted R 2 : 0.76; and for survival with CPC 1 or 2: 0.59 (p = 0.5172), R 2 : 0.72, Adjusted R 2 : 0.44. In 2018, the rate of patients with LHC and good CPC nearly doubled. Conclusion: During the study period, we found a steady increase in TTM rate and concomitant decrease in LHC rate. The rate of LHC correlated best with survival to hospital discharge. In 2018, an increase in the proportion of patients with CPC 1 or 2 receiving LHC occurred, despite a decline in the overall rate of LHC.

Abstract 430: Variation of Post Arrest Survival to Discharge and Cerebral Performance Category 1/2 in State of Michigan

Introduction: Resuscitation from out-of-hospital cardiac arrest (OHCA) requires success in the entire chain of survival. Our prior work demonstrated substantial variation in provision of left heart catheterization (LHC) and TTM within Michigan (MI) hospitals but did not evaluate impact on survival. The objective of this study is to characterize the variation in post arrest survival to discharge and CPC-1/2 in MI Hospitals. Methods: We used the MI Cardiac Arrest to Enhance Survival (CARES) registry, to analyze all adult OHCA patients with ROSC from 2014 - 2017 that survived to hospital admission. Hospitals (N=40) were included if they managed &gt; 30 cases/4-years for all rhythms and &gt; 20 cases/4-years for shockable patients. We excluded transferred patients; those resuscitated with bystander CPR only and patients &lt; 18 years old. We assessed provision of LHC and TTM by hospital, survival to hospital discharge and survival with good neurologic outcome (CPC 1 or 2). We report crude and adjusted survival rates by hospital with median and range. To account for variations between hospitals, the adjusted survival rates were estimated using multilevel multi-variable regression analyses, adjusting for OHCA variables predictive of survival. Results: There were 5486 patients included CARES patients, 4,715 ultimately included for analyses of survival to discharge and survival with good neurological outcome. Of included hospitals (N=40), the median of crude rate was 31.4% [Range: 12.5%, 51.9%] for survival at discharge and 25.1% [Range: 5.2%, 42.2%] for survival rate with CPC-1/2. In the multivariable analyses, adjusting for patient characteristics, the median of adjusted rate was 27.6% [Range: 18.1%, 42.0%] for survival and 21.3% [Range: 9.7%, 32.8%] for survival to discharge with CPC-1/2. Conclusion: We observed substantial disparities in inter-hospital rates of survival at discharge, with a 4-fold range of survival and 8-fold range of survival with CPC-1/2. This survival variation was ameliorated but still persisted in the adjusted modeling analysis. Variation in post arrest survival by hospital is not explained by patient and arrest characteristics, and identifies a need for hospital quality improvement activities to improve post arrest patient survival.

Associations of Perioperative Variables With the 30-Day Risk of Stroke or Death in Carotid Endarterectomy for Symptomatic Carotid Stenosis.

Background and Purpose- This analysis was performed to assess the association between perioperative and clinical variables and the 30-day risk of stroke or death after carotid endarterectomy for symptomatic carotid stenosis. Methods- Individual patient-level data from the 5 largest randomized controlled carotid trials were pooled in the Carotid Stenosis Trialists' Collaboration database. A total of 4181 patients who received carotid endarterectomy for symptomatic stenosis per protocol were included. Determinants of outcome included carotid endarterectomy technique, type of anesthesia, intraoperative neurophysiological monitoring, shunting, antiplatelet medication, and clinical variables. Stroke or death within 30 days after carotid endarterectomy was the primary outcome. Adjusted risk ratios (aRRs) were estimated in multilevel multivariable analyses using a Poisson regression model. Results- Mean age was 69.5±9.2 years (70.7% men). The 30-day stroke or death rate was 4.3%. In the multivariable regression analysis, local anesthesia was associated with a lower primary outcome rate (versus general anesthesia; aRR, 0.70 [95% CI, 0.50-0.99]). Shunting (aRR, 1.43 [95% CI, 1.05-1.95]), a contralateral high-grade carotid stenosis or occlusion (aRR, 1.58 [95% CI, 1.02-2.47]), and a more severe neurological deficit (mRS, 3-5 versus 0-2: aRR, 2.51 [95% CI, 1.30-4.83]) were associated with higher primary outcome rates. None of the other characteristics were significantly associated with the perioperative stroke or death risk. Conclusions- The current results indicate lower perioperative stroke or death rates in patients operated upon under local anesthesia, whereas a more severe neurological deficit and a contralateral high-grade carotid stenosis or occlusion were identified as potential risk factors. Despite a possible selection bias and patients not having been randomized, these findings might be useful to guide surgeons and anesthetists when treating patients with symptomatic carotid disease.

Variation in the Prescription of Androgen Deprivation Therapy in Intermediate- and High-risk Prostate Cancer Patients Treated with Radiotherapy in the Netherlands, and Adherence to European Association of Urology Guidelines: A Population-based Study

BackgroundAccording to (inter-)national guidelines, (neo-)adjuvant and concurrent androgen deprivation therapy (ADT) in combination with external beam radiotherapy (EBRT) is optional for intermediate-risk prostate cancer (PCa) patients and is the recommended standard treatment for high-risk PCa patients. ObjectiveThe aim of this study is to provide insight into the prescription of ADT in intermediate- and high-risk PCa patients treated with EBRT in the Netherlands, and to evaluate adherence to European Association of Urology guidelines and factors affecting prescription. Design, setting, and participantsAll intermediate- and high-risk PCa patients between October 2015 and April 2016 were identified through the population-based Netherlands Cancer Registry. Variation in the prescription of ADT in patients with EBRT was evaluated. Multivariable multilevel logistic regression analyses were performed to determine the probability of ADT and to examine the role of patient-, tumour-, and hospital-related factors. Results and limitationsOverall, 29% of patients with intermediate-risk PCa received ADT varying from 3% to 73% between institutions. From the multivariable regression analysis, higher Gleason grade, magnetic resonance imaging, and computed tomography (CT)-positron-emission tomography/CT prior to radiotherapy appeared to be associated with increased prescription of ADT. Among high-risk patients, 83% received ADT, varying from 57% to 100% between departments. A higher prostate-specific antigen level, more advanced tumour stage, and a higher Gleason grade were associated with increased prescription. ConclusionsLess than one-third of intermediate-risk PCa patients treated with EBRT receive ADT. The variation in the prescription of ADT between different institutions is substantial. This suggests that the prescription is largely dependent on different institutional policies. The guideline adherence in high-risk PCa is fairly good, as the vast majority of patients received ADT as recommended. However, given the clear recommendations in the guidelines, adherence could be improved. Patient summaryIn this review, we looked at the variation of hormonal treatment in intermediate- and high-risk prostate cancer patients. We found substantial variation between institutions.

Automated Scoring of Students’ Use of Text Evidence in Writing

AbstractDespite the importance of analytic text‐based writing, relatively little is known about how to teach to this important skill. A persistent barrier to conducting research that would provide insight on best practices for teaching this form of writing is a lack of outcome measures that assess students’ analytic text‐based writing development and that are feasible to implement at scale. Automated essay‐scoring (AES) technologies offer one potential approach to increasing the feasibility of research in this area, provided that the scores yield information about substantive dimensions of writing aligned to new standards and are sensitive to variation in literacy instruction. The authors describe an approach to using AES technologies to provide information about students’ skills at marshaling text evidence in the upper elementary grades. Specifically, the authors examined 1,529 responses to a response‐to‐text assessment (RTA) from 65 fifth‐ and sixth‐grade language arts classrooms, from which the authors also collected data on instruction via logs, text‐based writing assignments, and surveys. Through correlational, univariate, and multilevel multivariate analyses, the authors found validity evidence supporting automated scoring of the RTA: The authors found close correspondence of human and AES scores, alignment of AES scores with components of instruction that the authors expected would predict variation in students’ writing quality, and association between AES scores and other expected measures of student achievement. These findings provide encouraging evidence that AES technologies as applied to the RTA can generate valid inferences about students’ ability to marshal text evidence in writing and, thus, could be a useful tool for advancing large‐scale writing research.