Exchangeable Working Correlation Research Articles

P-429 Measuring serum progesterone levels on the day of a fresh blastocyst transfer is debatable as there is no evidence for correlation with ongoing pregnancy results

Abstract Study question Does serum progesterone (P4) level on the day of a fresh blastocyst transfer relate to ongoing pregnancy rate (OPR)? Summary answer OPR after a fresh single blastocyst transfer seems not to be associated with serum P4 level on the day of transfer. What is known already Several studies have suggested an optimal serum P4 threshold for embryo transfers in frozen embryo transfer cycles. However, for fresh embryo transfers, data are scarce and conflicting. Study design, size, duration A retrospective single-center cohort study was performed on 598 complete data cycles, retained out of 631 consecutive fresh single blastocyst transfer cycles. Data were collected from patients undergoing a fresh IVF/ICSI cycle with autologous oocytes at a tertiary fertility center between June 2022 and August 2023. Participants/materials, setting, methods A total of 482 unique patients underwent one or more single fresh blastocyst transfers using micronized vaginal progesterone (MVP) 800mg BID for luteal phase support starting day after oocyte retrieval. Serum P4 levels were determined on day of transfer. The primary endpoint was OPR, positive hCG-test and early pregnancy loss (EPL) were secondary endpoints. Unadjusted and adjusted generalized estimating equations (GEE) assessed P4 level association with the endpoints, using an exchangeable working correlation matrix. Main results and the role of chance No association was identified between P4 levels on day of transfer and OPR after accounting for maternal age, blastocyst quality, P4 and estradiol level at last ultrasound, BMI and endometrial thickness. Data were grouped according to 10/50/90 and 25/50/75 percentiles, resulting in groups < = or > 27ng/ml and < = or > 40ng/ml, with omission of 90th and 75th percentile groups due to right-skewed distribution. With serum P4 levels higher than 40ng/ml, the probability of a positive hCG-test significantly increased (OR 1.79, CI 1.11-2.91; p = 0.018). No significant difference was observed in patients with serum P4 < = or > 27ng/ml (OR 1.23, CI 0.66-2.30; p = 0.506). Regarding OPR, no significant differences were observed when comparing patients with serum P4 < = or > 27/ng/ml or < = or > 40ng/ml (OR 0.88, CI 0.42-18.85; p = 0.736 and OR 1.41, CI 0.83-2.37; p = 0.203, respectively). Moreover, EPL did not significantly differ between serum P4 level groups (OR 1.93, CI 0.61-6.13; p = 0.265 and OR 1.2, CI 0.51-2.84; p = 0.674). Age (p < 0.001) and embryo quality (p < 0.001) are significantly related with OPR. Additionally, late follicular progesterone levels and OPR are significantly associated (OR 2.39, CI 1.16-4.95 in patients with P4 < = or > 27/ng/ml and OR 2.21, CI 1.07-4.55 in patients with < =40ng/ml and >40ng/ml). Limitations, reasons for caution Limitations include the retrospective design and the lack of live birth outcome data. Progesterone levels exceeding 60 ng/ml were not further analyzed to their precise value but grouped as 60 plus which did not allow to identify potential detrimental effects of very high P4 levels on treatment outcome. Wider implications of the findings Our data suggest that measuring P4 on the day of a single blastocyst transfer with MVP luteal phase support may not be indicative of OPR or the risk of EPL. This raises questions regarding the necessity for routine P4 measurements as well as the development of rescue strategies. Trial registration number ONZ-2023-0628

Modeling Longitudinal Outcomes in a Small Matched-Pair Sample Motivated by Cardiovascular Data: A Simulation Study

Introduction: In cardiovascular research, matched pairs with longitudinal outcomes are used to assess time and treatment effects on relevant cardiovascular parameters of interest. Generalized estimating equations (GEE) with independent working correlation are commonly employed for unbiased estimation, but their consistency depends on large sample properties. This study investigates the validity of independent GEE for small sample matched pair data under various working correlations, comparing the results with quasi-least squares (QLS) estimates through simulation. Methods: We simulate a hospital cohort with longitudinal outcomes for two exposure groups and individual random effects. After obtaining the propensity score-matched sample, we performed different simulation scenarios across various mortality rates and the random effect. We compare results from GEE and QLS methods under independence, exchangeable, and AR1 working correlation structures. Results: The independence structure often yields a wider range of relative biases and higher standard errors when there is considerable drop-out. Conversely, the exchangeable structure appears as the true correlation structure, providing more accurate and reliable estimates. No significant discrepancies are observed between the results generated by GEE and QLS methods in this study, and the performance of both approaches is significantly influenced by mortality rates and the standard deviation of the random intercept. Conclusion: Our findings suggest that GEE with an independent working correlation structure is misspecified due to incorrect correlation assumptions and is less efficient. Therefore, correctly specifying the working correlation structure is important for small sample data.

Terapi Back Massage untuk Meningkatkan Kualitas Tidur dan Menurunkan Kecemasan Pasien Di ICU

Massage can relax muscles, improve blood circulation, reduce pain and anxiety, and improve sleep quality by increasing comfort. However, there is little research on whether back massage improves sleep quality in intensive care unit (ICU) patients. This study investigated the effect of back massage on improving vital signs, sleep quality, anxiety, and depression in ICU patients. This study used a quasi-experimental design with a convenience sampling approach to sample ICU patients. The experimental group received back massages for three consecutive days (n = 20), whereas the control group received usual care (n = 20). The Verran and Snyder-Halpern scales and the Hospital Anxiety and Depression Scale were used, and sleep time was recorded. The effect of the intervention was examined using a generalized estimation equation model with robust standard errors and a time-adjusted exchanMassage can relax muscles, improve blood circulation, reduce pain and anxiety, and improve sleep quality by increasing comfort. However, there is little research on whether back massage improves sleep quality in intensive care unit (ICU) patients. This study investigated the effect of back massage on improving vital signs, sleep quality, anxiety, and depression in ICU patients. This study used a quasi-experimental design with a convenience sampling approach to sample ICU patients. The experimental group received back massages for three consecutive days (n = 20), whereas the control group received usual care (n = 20). The Verran and Snyder-Halpern scales and the Hospital Anxiety and Depression Scale were used, and sleep time was recorded. The effect of the intervention was examined using a generalized estimation equation model with robust standard errors and a time-adjusted exchangeable working correlation matrix. The results showed that the average observed sleep time (measured by nurses) was 39 hours. Back massage improves breathing in patients, improves sleep quality, and causes significant changes in anxiety. These findings indicate that 10 minutes of back massage can improve sleep quality, sleep duration, breathing, and anxiety in ICU patients. The application of back massage shows a positive increase in the sleep quality of ICU patients. Massage training and intervention theory should be further applied when developing critical care nursing courses. geable working correlation matrix. The results showed that the average observed sleep time (measured by nurses) was 39 hours. Back massage improves breathing in patients, improves sleep quality, and causes significant changes in anxiety. These findings indicate that 10 minutes of back massage can improve sleep quality, sleep duration, breathing, and anxiety in ICU patients. The application of back massage shows a positive increase in the sleep quality of ICU patients. Massage training and intervention theory should be further applied when developing critical care nursing courses.

An overview of optimal designs under a given budget in cluster randomized trials with a binary outcome.

Cluster randomized trial design may raise financial concerns because the cost to recruit an additional cluster is much higher than to enroll an additional subject in subject-level randomized trials. Therefore, it is desirable to develop an optimal design. For local optimal designs, optimization means the minimum variance of the estimated treatment effect under the total budget. The local optimal design derived from the variance needs the input of an association parameter in terms of a "working" correlation structure in the generalized estimating equation models. When the range of instead of an exact value is available, the parameter space is defined as the range of and the design space is defined as enrollment feasibility, for example, the number of clusters or cluster size. For any value within the range, the optimal design and relative efficiency for each design in the design space is obtained. Then, for each design in the design space, the minimum relative efficiency within the parameter space is calculated. MaxiMin design is the optimal design that maximizes the minimum relative efficiency among all designs in the design space. Our contributions are threefold. First, for three common measures (risk difference, risk ratio, and odds ratio), we summarize all available local optimal designs and MaxiMin designs utilizing generalized estimating equation models when the group allocation proportion is predetermined for two-level and three-level parallel cluster randomized trials. We then propose the local optimal designs and MaxiMin designs using the same models when the group allocation proportion is undecided. Second, for partially nested designs, we develop the optimal designs for three common measures under the setting of equal number of subjects per cluster and exchangeable working correlation structure in the intervention group. Third, we create three new Statistical Analysis System (SAS) macros and update two existing SAS macros for all the optimal designs. We provide two examples to illustrate our methods.

A modified generalized estimating equation approach for simultaneous spatial durbin panel model: Case study of economic growth in ASEAN countries

This article briefly explains the simultaneous spatial durbin panel (SSDP) model. The study of the SSDP model is substantial because it can explain the interaction between geographic units, is more informative, diverse, efficient, exhaustive, and accurate in reaching conclusions that influence the policy determination. This article’s intention is to derive a parameter estimation method from the SSDP model using a modified generalized estimating equation approach, which is then used to model economic growth in ASEAN nations. This article compares the SSDP model with rook contiguity, 2-nearest neighbors, and a customized spatial weighted matrix in relation to an independent, first-order autoregressive, exchangeable working correlation structure. To model economic growth in ASEAN countries, a customized weighted matrix with first-order autoregressive and exchangeable working correlations is chosen based on the CIC value. The parameter analysis outcomes indicate: 1) it is a significant spatial dependence among ASEAN countries; 2) it is a significant simultaneous interaction among the gross domestic product (GDP) and foreign direct investment (FDI); 3) GDP has a greater influence on FDI than FDI does on GDP; 4) The economic growth is directly affected by the labor force total; and 5) trade openness has a direct effect on FDI.

Abstract A105: Participants with lower socioeconomic status or lacking access to health care benefited more from a decision support navigation trial to increase colorectal cancer screening: Study of Chinese and Korean American primary care patients

Abstract Background: Although colorectal cancer (CRC) is the second most common cancer among Asian Americans, cancer screening rates among Chinese- and Korean Americans are lower than in other racial/ethnic groups in the United States. To address this disparity, we designed, implemented, and evaluated the effectiveness of a culturally adapted decision support navigation trial. Methods: Twenty-four Chinese and Korean primary care physicians (PCPs) from Maryland and northern Virginia were recruited to the STOP CRC (Screening TO Prevent CRC) program, a cluster randomized controlled trial. The PCPs’ clinics were randomized into intervention or control group. Then, 400 Chinese and Korean Americans were recruited through the PCPs’ clinics. Participants randomized to a decision support navigation intervention group (n=200) received an informational booklet, instructions to access colonoscopy screening, and a free fecal immunochemical test (FIT) kit at baseline. A bilingual patient navigator also guided them through an exercise to facilitate screening test choices. Six months after enrollment, participants randomized to the control group (n=200) received the informational booklet, colonoscopy instructions, and a FIT kit. Participants were offered a choice to select between the FIT or colonoscopy. Screening (completion of FIT or colonoscopy) was measured at 6 and 12 months after enrollment by medical record review and self-report. Generalized estimating equations (GEE) models with an exchangeable working correlation matrix to account for clustering effects at the PCP level were employed to examine intervention effects on CRC screening outcomes. Results: Intervention and control participants were comparable in their socio-demographic background and access to healthcare at baseline. At 12 months, screening adherence in the intervention group was (93%) substantially higher than in the control group (63%) (Prevalence Ratio: 1.51, 95% confidence interval: 1.37-1.66). Interestingly, the effect of the intervention on CRC screening was significantly higher among those with less than or equal to high school education (PR=1.95; 95% CI: 1.89-2.02) compared to those with higher than high school education (PR=1.37; 95% CI: 1.37; 95% CI: 1.22-1.54). Moreover, the impact of the intervention on CRC screening was higher among participants with lower income (<$100,000: PR=1.60; 95% CI: 1.56-1.65) than the participants with higher income (≥$100,000: PR=1.18; 96% CI: 0.85-1.64). The same impact was higher among those without health insurance (PR=1.70; 95% CI: 1.28-2.25) and those with Medicare/Medicaid (PR=1.80, 95% CI: 1.29-2.52) compared to those who have private health insurance (PR=1.35, 95% CI: 1.19-1.53). Conclusions: The STOP CRC program, a culturally and linguistically adapted decision support navigation intervention, is effective in increasing CRC screening among Chinese and Korean American primary care patients who are not up-to-date with the screening. This impact is more pronounced among participants with lower socioeconomic status or lacking access to health care. Citation Format: Sunmin Lee, Soomin Ryu, Carissa Kwan, Ronald E. Myers, Yuxi Shi, Luohua Jiang, Brittany N. Morey. Participants with lower socioeconomic status or lacking access to health care benefited more from a decision support navigation trial to increase colorectal cancer screening: Study of Chinese and Korean American primary care patients [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr A105.

The association of gabapentin use and change of cognitive and functional status in older adults with normal cognition

AbstractBackgroundAlthough gabapentin has been increasingly prescribed to older adults, the relationship between gabapentin and cognition/functional status is not well studied. Therefore, we aimed to examine the association of gabapentin use and cognitive/functional change in older adults.MethodsA retrospective cohort study was conducted using National Alzheimer’s Coordinating Center Uniform Data Set (2005‐March 2021). Gabapentin new users with normal cognition at the visit of gabapentin initiation (i.e., index visit) were included. New‐users were matched on year of first enrollment and time of gabapentin initiation since enrollment to randomly select nonusers with replacement. The decline of cognitive/functional status, including Clinical Dementia Rating global score (CDRGLOB), Clinical Dementia Rating sum of boxes (CDRSUM), Mini‐Mental State Examination (MMSE), sum of Functional Activities Questionnaire (FAQ), and mean of FAQ, was assessed from index to index+1 and index+2 visits. The criteria of cognitive/functional status decline were any increase of CDRGLOB, 1 point increase of CDRSUM, 1 point decrease of MMSE, 3 point increase of sum of FAQ, and 0.3 point increase of mean of FAQ. To mitigate confounding and selection bias, stabilized inverse probability of treatment weights and stabilized inverse probability of censoring weights were used. Generalized estimating equations with an exchangeable working correlation structure were used to account for within‐participant correlations and obtain robust standard errors.ResultsWe included 496 new‐users (mean age [SD]: 78.0 [7.1]; male 35.9%) and 4,464 nonusers (77.4 [7.1]; 35.1%) with normal cognition at index. Among the participants who had information of seizure history (57.1% of new‐users and 58.0% of nonusers), 2.2% and 1.6% of new‐users and nonusers, respectively, reported having seizure history, indicating that gabapentin is widely prescribed off‐label in older adults. Gabapentin use was significantly associated with cognitive/functional status decline with outcomes, including CDRGLOB at index+1 visit (odds ratio [95% confidence interval]: 1.57 [1.08, 2.31]); CDRSUM at index+1 visit (1.89 [1.16, 3.06]) and at index+2 visit (1.66 [1.05, 2.63]); and mean of FAQ at index+2 visit (2.23 [1.22, 4.07]) (Figure 1).ConclusionsGabapentin use was significantly associated with decline in cognitive and functional status among older adults with initially normal cognition. Further studies are needed to examine the association.

Cluster randomised trials with a binary outcome and a small number of clusters: comparison of individual and cluster level analysis method

BackgroundCluster randomised trials (CRTs) are often designed with a small number of clusters, but it is not clear which analysis methods are optimal when the outcome is binary. This simulation study aimed to determine (i) whether cluster-level analysis (CL), generalised linear mixed models (GLMM), and generalised estimating equations with sandwich variance (GEE) approaches maintain acceptable type-one error including the impact of non-normality of cluster effects and low prevalence, and if so (ii) which methods have the greatest power. We simulated CRTs with 8–30 clusters, altering the cluster-size, outcome prevalence, intracluster correlation coefficient, and cluster effect distribution. We analysed each dataset with weighted and unweighted CL; GLMM with adaptive quadrature and restricted pseudolikelihood; GEE with Kauermann-and-Carroll and Fay-and-Graubard sandwich variance using independent and exchangeable working correlation matrices. P-values were from a t-distribution with degrees of freedom (DoF) as clusters minus cluster-level parameters; GLMM pseudolikelihood also used Satterthwaite and Kenward-Roger DoF.ResultsUnweighted CL, GLMM pseudolikelihood, and Fay-and-Graubard GEE with independent or exchangeable working correlation matrix controlled type-one error in > 97% scenarios with clusters minus parameters DoF. Cluster-effect distribution and prevalence of outcome did not usually affect analysis method performance. GEE had the least power. With 20–30 clusters, GLMM had greater power than CL with varying cluster-size but similar power otherwise; with fewer clusters, GLMM had lower power with common cluster-size, similar power with medium variation, and greater power with large variation in cluster-size.ConclusionWe recommend that CRTs with ≤ 30 clusters and a binary outcome use an unweighted CL or restricted pseudolikelihood GLMM both with DoF clusters minus cluster-level parameters.

Comparing Generalized Estimating Equation and Linear Mixed Effects Model for Estimating Marginal Association with Bivariate Continuous Outcomes

ABSTRACT Purpose Both linear regression with generalized estimating equations (GEE) and linear mixed-effects models (LMEM) can be used to estimate the marginal association of an exposure with clustered continuous outcomes. This study compares their performance for bivariate continuous outcomes which are common in eye studies. Methods Parametric and non-parametric simulations were used to compare the GEE models including independent, exchangeable, and unstructured working correlation structures and LMEM including random intercept only and random intercept and slope models in R and SAS. Data generation referenced the data distributions from a real-world study for estimating ocular structure-visual function relationships in patients with retinitis pigmentosa. Results From both parametric and non-parametric simulations, comparing the random intercept LMEM and GEE exchangeable model, bias was similar; coverage probability of the 95% confidence interval (CI) from the random intercept LMEM was often closer to 95%, especially when the sample size was small; the power for testing the association of the exposure was higher from the GEE exchangeable model, but its type-I error rate might be inflated especially when the sample size was small. The type-I error rate from the random intercept LMEM was closer to 0.05, but it might be under 0.05 and coverage probability might be over 95%. The GEE independent model performed worst and the LMEM with both random intercept and slope might not converge. Conclusion To estimate marginal exposure-outcome association with bivariate continuous outcomes, the random intercept LMEM may be preferred. It has the best coverage probability of 95% CI and is the only model with correct type-I error rates in this study. However, it may have low power and overly wide CI in studies with small sample size or low inter-eye correlation.

Estimands in cluster-randomized trials: choosing analyses that answer the right question.

Cluster-randomized trials (CRTs) involve randomizing groups of individuals (e.g. hospitals, schools or villages) to different interventions. Various approaches exist for analysing CRTs but there has been little discussion around the treatment effects (estimands) targeted by each. We describe the different estimands that can be addressed through CRTs and demonstrate how choices between different analytic approaches can impact the interpretation of results by fundamentally changing the question being asked, or, equivalently, the target estimand. CRTs can address either the participant-average treatment effect (the average treatment effect across participants) or the cluster-average treatment effect (the average treatment effect across clusters). These two estimands can differ when participant outcomes or the treatment effect depends on the cluster size (referred to as 'informative cluster size'), which can occur for reasons such as differences in staffing levels or types of participants between small and large clusters. Furthermore, common estimators, such as mixed-effects models or generalized estimating equations with an exchangeable working correlation structure, can produce biased estimates for both the participant-average and cluster-average treatment effects when cluster size is informative. We describe alternative estimators (independence estimating equations and cluster-level analyses) that are unbiased for CRTs even when informative cluster size is present. We conclude that careful specification of the estimand at the outset can ensure that the study question being addressed is clear and relevant, and, in turn, that the selected estimator provides an unbiased estimate of the desired quantity.

Abstract CT164: A culturally adapted decision support navigation trial to increase colorectal cancer screening among Chinese and Korean American primary care patients

Abstract Background: Colorectal cancer (CRC) is the second most common cancer among Asian Americans. However, cancer screening rates among Chinese- and Korean Americans are lower than other racial/ethnic groups in the United States. To address this disparity, we tested the effectiveness of a culturally adapted decision support navigation trial. Methods: The STOP CRC (Screening TO Prevent CRC) program, a cluster randomized controlled trial, recruited 24 Chinese and Korean primary care physicians (PCP) from Maryland and northern Virginia. Then, 400 Chinese and Korean Americans were recruited through the PCPs’ clinics. Participants randomized to a decision support navigation intervention group (n=200) received an informational booklet, instructions to access colonoscopy screening, and a free fecal immunochemical test (FIT) kit at baseline. A bilingual patient navigator also guided them through an exercise to facilitate screening test choices. Six months after enrollment, participants randomized to the control group (n=200) received the informational booklet, colonoscopy instructions, and a FIT kit. Participants selected a method of CRC screening between the FIT or colonoscopy. Screening (completion of FIT or colonoscopy) was measured at 6 and 12 months after enrollment by medical record review and self-report. Generalized estimating equations (GEE) models with an exchangeable working correlation matrix to account for clustering effects at the PCP level were employed to examine intervention effects on CRC screening outcomes. Results: Intervention and control participants were comparable on socio-demographic background and access to healthcare at baseline. The mean age of participants was 58 years, and 53% of participants were female. The majority of participants (85%) were married, 34% had less than or equal to high school education, 32% had annual household income less than $40,000, and 21% did not have health insurance. At 6 months, screening adherence in the intervention group was (91%) predominantly greater than in the control group (8%) (Prevalence Ratio: 11.31, 95% confidence interval: 6.95-18.43). At 12 months, although the gap between the two groups was reduced, the screening adherence in the intervention group was (93%) still substantially higher than in the control group (63%) (Prevalence Ratio: 1.51, 95% confidence interval: 1.37-1.66). Conclusions: The STOP CRC program, a culturally and linguistically adapted decision support navigation intervention, is effective in increasing CRC screening among Chinese and Korean American primary care patients who are not up-to-date with the screening. This program can potentially be adapted and implemented for other Asian Americans who are not up-to-date with the CRC screening in the U.S. Citation Format: Sunmin Lee, Soomin Ryu, Carissa Kwan, Ronald E. Myers, Yuxi Shi, Luohua Jiang, Brittany N. Morey. A culturally adapted decision support navigation trial to increase colorectal cancer screening among Chinese and Korean American primary care patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr CT164.

Smoking-Cessation Assistance Among Older Adults by Ethnicity/Language Preference

Although smoking prevalence is lower among Hispanic adults than among non-Hispanic White adults, smoking remains a leading cause of preventable death among older Hispanics. This study examines the differences in tobacco assessment and smoking-cessation assistance among older patients seen in community health centers by ethnicity and language preference. Electronic health record data were extracted from the Accelerating Data Value Across a National Community Health Center Network of community health centers from patients aged 55‒80 years with ≥1 primary care visit between January 1, 2017 and December 31, 2018. Binary outcomes included tobacco use assessment and, among those with ≥1 status indicating current smoking, having a smoking-cessation medication ordered. The independent variable combined ethnicity and language preference, categorized as non-Hispanic White (reference), Spanish-preferring Hispanic, and English-preferring Hispanic. Multivariable generalized estimating equation logistic regressions, clustering by primary care clinic using an exchangeable working correlation structure, modeled the odds of tobacco use assessment and cessation medication orders by ethnicity/preferred language, adjusting for patient covariates, health system, and clinic location. Analyses were conducted in 2021. The study included 116,328 patients. Spanish-preferring Hispanic patients had significantly lower odds of having tobacco use assessed than non-Hispanic White patients (AOR=0.89, 95% CI=0.82, 0.95). Both Spanish- and English-preferring Hispanic patients had lower odds of having a smoking-cessation medication ordered (AOR=0.53, 95% CI=0.47, 0.60; AOR=0.77, 95% CI=0.67, 0.89, respectively) than non-Hispanic White patients. Significant disparities were found in tobacco assessment and cessation assistance by ethnicity and language preference among older adults seen in safety-net clinics. Future research is needed to understand the etiology of these smoking-related disparities.

Illness Representation and Associated Factors among Hypertensive Patients in Central Ethiopia.

Illness representation is an implicit belief system about an illness constructed by an individual to give meaning to their illness. An individual's belief about his/her illness, treatment, and own control are known to influence an individual's ability to cope with the illness and sustain the health-related quality of life. However, how Ethiopians perceive hypertension has not been studied well. This study aimed to assess illness representation and associated factors among hypertensive patients in Central Ethiopia. A facility-based cross-sectional study was conducted in four public hospitals in Central Ethiopia. A total of 989 patients participated in the study. The revised Illness Representation Questionnaire was used to collect relevant data. Data were analyzed using the Generalized Estimating Equation with an ordinal logistic regression model and exchangeable working correlation matrix. A P-Value of less than 0.05 was indicated statistical significance. Overall, 64.3% (95% CI: 61.3, 67.4) of the respondents reported low to moderate Illness representation about their hypertension. Respondents who were housewife [AOR: 1.48, 95% CI= 1.05, 2.08], in older age category 50-64 years [AOR: 1.92, 95% CI= 1.19, 3.09] and ≥ 65 years [AOR: 2.38, 95% CI= 1.43, 3.96], and had no family support [AOR: 1.98, 95% CI= 1.44, 2.73] showed a significant association with Illness Representation. This study revealed that about two-thirds of hypertensive patients in Central Ethiopia perceived hypertension as a low to moderately threatening illness. Such low illness representation undermines initiation of treatment and effective control of blood pressure. Health care providers need to strengthen strategies that increase their patient's illness representation.

Two weights make a wrong: Cluster randomized trials with variable cluster sizes and heterogeneous treatment effects

In cluster randomized trials (CRTs), the hierarchical nesting of participants (level 1) within clusters (level 2) leads to two conceptual populations: clusters and participants. When cluster sizes vary and the goal is to generalize to a hypothetical population of clusters, the unit average treatment effect (UATE), which averages equally at the cluster level rather than equally at the participant level, is a common estimand of interest. From an analytic perspective, when a generalized estimating equations (GEE) framework is used to obtain averaged treatment effect estimates for CRTs with variable cluster sizes, it is natural to specify an inverse cluster size weighted analysis so that each cluster contributes equally and to adopt an exchangeable working correlation matrix to account for within-cluster correlation. However, such an approach essentially uses two distinct weights in the analysis (i.e. both cluster size weights and covariance weights) and, in this article, we caution that it will lead to biased and/or inefficient treatment effect estimates for the UATE estimand. That is, two weights “make a wrong” or lead to poor estimation characteristics. These findings are based on theoretical derivations, corroborated via a simulation study, and illustrated using data from a CRT of a colorectal cancer screening program. We show that, an analysis with both an independence working correlation matrix and weighting by inverse cluster size is the only approach that always provides valid results for estimation of the UATE in CRTs with variable cluster sizes.

Predictors of prolonged hospitalization after vaginal birth in Ghana: A comparative study.

Early discharge after child delivery although indispensable, but maybe precluded by several factors. The effect of these factors on prolonged length of stay (LOS) after vaginal delivery has been sparsely investigated in Ghana. This limits understanding of potential leading indicators to inform intervention efforts and optimize health care delivery. This study examined factors associated with prolonged LOS after vaginal birth in two time-separated cohorts in Ghana. We analyzed data from Ghana's demographic and health surveys in 2007 and 2017. Our comparative analysis is based on subsamples in 2007 cohort (n = 2,486) and 2017 cohort (n = 8,065). A generalized estimating equation (GEE) with logistic regression was used to examine predictors of prolonged LOS after vaginal delivery. The cluster effect was accounted for using the exchangeable working correlation. The odds ratios (OR) and 95% confidence interval were reported. We found that 62.4% (1551) of the participants in 2007 had prolonged LOS after vaginal delivery, whereas the prevalence of LOS in the 2017 cohorts was 44.9% (3617). This constitutes a 17.5% decrease over the past decade investigated. Advanced maternal age (AOR = 1.24, 95% Cl 1.01-1.54), place of delivery (AOR = 1.18, 95% Cl 1.02-1.37), child's size below average (AOR = 1.14; 95% Cl 1.03-1.25), and problems suffered during/after delivery (AOR = 1.60; 95% Cl 1.43-1.80) were significantly associated with prolonged (≥ 24 hours) length of hospitalization after vaginal delivery in 2017. However, among variables that were available in 2007, only those who sought delivery assistance from non-health professionals (AOR = 1.89, 95% CI: 1.00-3.61) were significantly associated with prolonged LOS in the 2007 cohort. Our study provides suggestive evidence of a reduction in prolonged LOS between the two-time points. Despite the reduction observed, more intervention targeting the identified predictors of LOS is urgently needed to further reduce post-vaginal delivery hospital stay. Also, given that LOS is an important indicator of medical services use, an accurate understanding of its prevalence and associated predictors are useful in assessing the efficiency of hospital management practices and the quality of care of patients in Ghana.

Power and sample size calculations for cluster randomized trials with binary outcomes when intracluster correlation coefficients vary by treatment arm.

Generalized estimating equations are commonly used to fit logistic regression models to clustered binary data from cluster randomized trials. A commonly used correlation structure assumes that the intracluster correlation coefficient does not vary by treatment arm or other covariates, but the consequences of this assumption are understudied. We aim to evaluate the effect of allowing variation of the intracluster correlation coefficient by treatment or other covariates on the efficiency of analysis and show how to account for such variation in sample size calculations. We develop formulae for the asymptotic variance of the estimated difference in outcome between treatment arms obtained when the true exchangeable correlation structure depends on the treatment arm and the working correlation structure used in the generalized estimating equations analysis is: (i) correctly specified, (ii) independent, or (iii) exchangeable with no dependence on treatment arm. These formulae require a known distribution of cluster sizes; we also develop simplifications for the case when cluster sizes do not vary and approximations that can be used when the first two moments of the cluster size distribution are known. We then extend the results to settings with adjustment for a second binary cluster-level covariate. We provide formulae to calculate the required sample size for cluster randomized trials using these variances. We show that the asymptotic variance of the estimated difference in outcome between treatment arms using these three working correlation structures is the same if all clusters have the same size, and this asymptotic variance is approximately the same when intracluster correlation coefficient values are small. We illustrate these results using data from a recent cluster randomized trial for infectious disease prevention in which the clusters are groups of households and modest in size (mean 9.6 individuals), with intracluster correlation coefficient values of 0.078 in the control arm and 0.057 in an intervention arm. In this application, we found a negligible difference between the variances calculated using structures (i) and (iii) and only a small increase (typically ) for the independent correlation structure (ii), and hence minimal effect on power or sample size requirements. The impact may be larger in other applications if there is greater variation in the ICC between treatment arms or with an additional covariate. The common approach of fitting generalized estimating equations with an exchangeable working correlation structure with a common intracluster correlation coefficient across arms likely does not substantially reduce the power or efficiency of the analysis in the setting of a large number of small or modest-sized clusters, even if the intracluster correlation coefficient varies by treatment arm. Our formulae, however, allow formal evaluation of this and may identify situations in which variation in intracluster correlation coefficient by treatment arm or another binary covariate may have a more substantial impact on power and hence sample size requirements.

Racial‐ethnic differences in baseline and longitudinal change in neuropsychological test scores in the NACC Uniform Data Set 3.0

AbstractBackgroundRacial/ethnic differences in cross‐sectional assessment of cognitive test performance are well known. However, longitudinal studies of differences in cognitive decline have been mixed. The purpose of this study was to examine racial/ethnic differences in baseline and longitudinal change on the Uniform Data Set (UDS) version 3 of the NIA Alzheimer’s Disease Research Centers program.MethodLongitudinal data from 2,806 participants (2336 non‐Hispanic White, 350 non‐Hispanic Black, 120 Hispanics tested in Spanish) were acquired from the National Alzheimer’s Coordinating Center (NACC), and included baseline and at least two follow‐up visits. We used marginal linear regression models to examine racial/ethnic differences in standardized test scores, by including racial group indicators, time (in years) since initial visit and their interaction, controlling for baseline age, sex, education and changes in global CDR scores during the follow‐up. Additional models examining longitudinal change also controlled for baseline score. The outcome variables included the Montreal Cognitive Assessment (MoCA, Number Span (Forward/Backward), Craft Story 21 Recall (Immediate/Delayed), Multilingual Naming Test (MINT), Category Naming (animals and vegetables), Trail Making A and B, and Benson Figure (Copy/Recall). Estimates were obtained from generalized estimating equations with an exchangeable working correlation, and robust standard errors were used to construct confidence intervals and compute p‐values.ResultBlack and Hispanic participants had lower baseline scores on all tests (difference in standard deviation units: ‐0.029 to ‐0.858 for blacks, and ‐0.014 to ‐1.214 for Hispanics), but showed attenuated decline on all tests (difference in standard deviation units per year: 0.002 to 0.102 for blacks, 0.022 to 0.296 for Hispanics) compared to non‐Hispanic Whites. When baseline test scores were controlled, the differences in longitudinal changes became mostly statistically non‐significant.ConclusionDespite large racial/ethnic differences in baseline test scores, there were no racial/ethnic differences in longitudinal change over time once baseline differences were controlled. Further, results suggest that the utility of baseline test scores in UDS version 3 to predict longitudinal change may not be compromised in racially diverse populations.

Efficient Clinical Implementation of an MRI-Guided Adaptive Radiation Therapy Program

<h3>Purpose/Objective(s)</h3> Magnetic Resonance Image-guided Adaptive Radiation Therapy (MRgART) is ideal for intra-abdominal tumors as it accounts for daily anatomical changes in target volumes and nearby organs at risk (OAR) caused by soft-tissue deformation, respiration, organ filling/emptying and patient weight changes. The dosimetric gains associated with MRgART may be undermined by longer treatment times, resulting in unaccounted for intra-fractional OAR changes and patient motion. In this study, we aim to evaluate each step in our MRgART workflow for optimal resource utilization and potential treatment time savings. <h3>Materials/Methods</h3> This IRB approved retrospective review included pancreas cancer patients treated at our institution between 2018 and 2020 with MR guidance. We examined MRgART prescription details, and steps in the adaptive workflow, including MR simulation, image fusion, contouring, re-plan/QA and tumor tracking for resource allocation and clinical workflow efficiency. A generalized estimating equation (GEE) model with an exchangeable working correlation structure and Wilcoxon signed rank test were used for statistical time analysis. <h3>Results</h3> The study included 17 eligible pancreas cancer patients treated consecutively on our 0.35 T MR-Linac with MRgART. We examined 115 treated fractions (77 adapted). Two of the patients received conventional fractionation, and 15 were treated with stereotactic body radiation therapy (SBRT). SBRT prescription dose was 40 Gy to PTV in 5 fractions with 6MV FFF X-rays. Non-adaptive fractions averaged 38.3 ± 3.2 min (mean ± s.d.) compared to significantly longer adaptive treatments lasting 57.8 ± 4.3 min (an increase of 19.6 min or 51.3%, <b>P</b> < 0.01). Within each adapted treatment, the greatest time was spent on OAR contouring (13.6 min), followed by patient setup (10.6 min), imaging (9.1 min), re-plan (7.3 min), and QA (2.2 min). Beam-on time was 26% of overall treatment duration. Major contributors to increased treatment time with MRgART were contouring, physics review, replan and QA. Other sub-categories, such as, patient setup, imaging and beam-on times remained unchanged when compared to non-adaptive treatments. On average, treatment times decreased by 13.6% (61.4 min to 53 min, <b>P</b> = 0.06) as patients progressed through their treatment course reflecting improved patient confidence, compliance and comfort with their treatments and breathing instructions. Gains in treatment time also resulted from limited OAR contouring around the PTV and audio/visual coaching of patients leading to longer duty cycle. <h3>Conclusion</h3> Efficient resource utilization is important for safe and effective implementation of MRgART. Online OAR contouring, imaging and patient setup are potential targets for further time improvements and could be achieved <b>via</b> offline staff training, emphasis on robust AI based imaging, contouring and planning strategies, and in-room monitor device for improved active patient participation, comfort and compliance.

101 Uncommon alleles in FLG2 and TCHHL1 are associated with remission of atopic dermatitis

Atopic dermatitis (AD) is a relapsing inflammatory skin disease; filaggrin (FLG) variation has been consistently associated with its pathogenesis. Filaggrin-2 (FLG2) and trichohyalin-like-1 (TCHHL1) are members of the same protein family (S100 fused type proteins), are similar in structure to FLG, and may be involved in AD pathogenesis. We sought to evaluate the association between variation in FLG2, TCHHL1 and AD remission. We sequenced FLG2 and TCHHL1 in a longitudinal cohort of individuals with AD using targeted capture based massively parallel sequencing. The association between individual alleles and AD remission was evaluated with generalized estimating equations (GEE) for binary outcomes, assuming an exchangeable working correlation structure with empirical standard errors. The association between groups of alleles and AD remission was evaluated using a genetic algorithm, implemented via the rbga.bin function in R, to group alleles and GEE models to evaluate group associations with AD remission. We identified 3 loss-of-function (LoF) mutations in FLG2 (Ser2377Ter, Ter2392Ser, Arg2207Ter) and 2 LoF mutations in TCHHL1 (Gln656Ter, Gln294Ter), none of which were associated with AD remission. Common (MAF>5%), non-LoF alleles in FLG2 were similarly unassociated with AD. One common allele in TCHHL1, 152057002A>G, was associated with increased AD remission (OR 1.48; 95%CI 1.06-2.06; p=0.022). Within the self-described white population, a group of 34 uncommon alleles in FLG2 were associated with increased AD remission (OR 7.64e17; 95%CI 4.41e17-1.32e18; adjusted p<1.0e-16). 12 uncommon alleles in TCHHL1 were associated with increased AD remission (OR 23.46; 95%CI 7.07-77.89; adjusted p=0.064). Within the self-described African American population, 13 uncommon FLG2 alleles were associated with increased AD remission (OR 21.01; 95%CI 11.90-37.09; adjusted p<1.0e-16). No TCHHL1 uncommon allele groups were associated with AD remission. Our study supports the role of uncommon alleles in FLG2 and TCHHL1 in AD pathogenesis.

CAN THE ENDOMETRIUM BE TOO THICK? AN ANALYSIS OF SINGLE EUPLOID EMBRYO TRANSFER (SEET) CYCLE PREGNANCY OUTCOMES

Estrogen is used to prepare the endometrium prior to a frozen embryo transfer. An endometrial thickness (EMT) ≥7mm on the day of embryo transfer (ET) is associated with increased likelihood of pregnancy. However, few studies have investigated whether there is an upper limit of endometrial thickness at ET that affects implantation potential.1 The objective of this study is to determine whether there is a correlation between increasing EMT and ART cycle outcomes in SEET cycles. The study included patients who underwent an autologous or donor egg SEET cycle from September 2016 to March 2021 with an EMT ≥7mm on cycle day 10-15 after preparation with oral Estradiol. Demographic and cycle characteristics including oocyte age, BMI, AMH, peak E2, P4 prior to ET, day of embryo biopsy, embryo quality and days of estrogen treatment were collected. Primary outcome was ongoing clinical pregnancy rate (OPR). Secondary outcomes included implantation rate (IR), biochemical pregnancy rate, and early pregnancy loss rate. Data was analyzed using student’s t-test, chi-square, and Spearman’s correlation. General estimating equations (GEE) model with exchangeable working correlation structure was used to account for multiple cycles from the same patients. A total of 4,911 SEET cycles were identified and included in the study. Included patients had a mean oocyte age of 35.63, BMI of 24.2, and an average of 18.2 days of oral estradiol prior to ET. Implantation, biochemical pregnancy, early pregnancy loss, and ongoing clinical pregnancy rates were calculated for each mm of endometrial thickness starting at 7mm. Increasing EMT was correlated with IR (r = 0.50, p = <0.0001) and OPR (r = 0.84, p = <0.0001). The highest IR was observed for women with an EMT of 14mm (79%), above which the IR decreased to 71% at 15mm, and 60% at ≥16mm. The highest OPR occurred at an EMT of 15mm (67%), above which it decreased to 47%. However this decrease was not statistically significant after adjusting for oocyte age, BMI, peak E2, P4 prior to transfer, endometrial type, AMH, uterine factor, embryo quality, prior transfer outcomes and embryo biopsy day for both IR (β= 0.27, p = 0.61), and OPR (β= 0.26, p = 0.64). Our study demonstrates a positive correlation between increasing EMT ≥ 7mm and OPR in SEET cycles up to 15mm. Future studies evaluating endometrial vascularity and endometrial metabolic profiles and their association with endometrial thickness may further expand our understanding of endometrial physiology and its impact on clinical outcomes.