Epidemiological characteristics of first-time SARS-CoV-2 Omicron infection among hospital staff in Chengdu, China.

BACKGROUND: After China ended its 'dynamic zero-COVID policy' on 7 December 2022, a large-scale outbreak of SARS-CoV-2 Omicron infections emerged across the country. We conducted a hospital-wide prospective study to document the epidemiological characteristics of the outbreak among healthcare workers in a hospital of Chengdu, where no previous staff SARS-CoV-2 infections were detected. METHODS: All hospital staff members were invited to complete an online questionnaire on COVID-19 in January 2023, and SARS-CoV-2 infection cases were followed up by telephone in June 2023 to collect data on long COVID. Univariable and multivariable logistic regression analyses were performed to evaluate factors associated with SARS-CoV-2 infection. RESULTS: A total of 2,899 hospital staff (93.5%) completed the online questionnaire, and 86.4% were infected with SARS-CoV-2 Omicron. The clinical manifestations of these patients were characterized by a high incidence of systemic symptoms. Cough (83.4%), fatigue (79.8%) and fever (74.3%) were the most frequently reported symptoms. Multivariable logistic analysis revealed that females [adjusted odds ratio (aOR): 1.42, 95% confidence interval (CI): 1.07-1.88] and clinical practitioners (aOR: 10.32, 95% CI: 6.57-16.20) were associated with an increased risk of SARS-CoV-2 infection, whereas advanced age ≥ 60 years (aOR: 0.30, 95% CI: 0.19-0.49) and a three-dose COVID-19 vaccination with the most recent dose administered within 3 months before 7 December 2022 (aOR: 0.44, 95% CI: 0.23-0.87 for within 1 month; aOR: 0.46, 95% CI: 0.22-0.97 for within 1-3 months) were associated with reduced risk. Among the cases, 4.27% experienced long COVID of fatigue, brain fog or both, with the majority reporting minor symptoms. CONCLUSION: Our findings provide a snapshot of the epidemiological situation of SARS-CoV-2 infection among healthcare workers in Chengdu after China's deregulation of COVID-19 control. Data in the study can aid in the development and implementation of effective measures to protect healthcare workers and maintain the integrity of healthcare systems during challenging times such as a rapid and widespread Omicron outbreak.

Development and validation of a nomogram for predicting cancer-specific survival in small-bowel adenocarcinoma patients using the SEER database.

BACKGROUND: Small bowel adenocarcinoma (SBA) is a rare gastrointestinal malignancy forwhich survival is hampered by late diagnosis, complex responses to treatment, and poor prognosis. Accurate prognostic tools are crucial for optimizing treatment strategies and improving patient outcomes. This study aimed to develop and validate a nomogram based on the Surveillance, Epidemiology, and End Results (SEER) database to predict cancer-specific survival (CSS) in patients with SBA and compare it to traditional American Joint Committee on Cancer (AJCC) staging. METHODS: We analyzed data from 2,064 patients diagnosed with SBA between 2010 and 2020 from the SEER database. Patients were randomly assigned to training and validation cohorts (7:3 ratio). Kaplan‒Meier survival analysis, Cox multivariate regression, and nomograms were constructed for analysis of 3-year and 5-year CSS. The performance of the nomograms was evaluated using Harrell's concordance index (C-index), the area under the receiver operating characteristic (ROC) curve, calibration curves, decision curve analysis (DCA), net reclassification improvement (NRI), and integrated discrimination improvement (IDI). RESULTS: Multivariate Cox regression identified sex, age at diagnosis, marital status, tumor site, pathological grade, T stage, N stage, M stage, surgery, retrieval of regional lymph nodes (RORLN), and chemotherapy as independent covariates associated with CSS. In both the training and validation cohorts, the developed nomograms demonstrated superior performance to that of the AJCC staging system, with C-indices of 0.764 and 0.759, respectively. The area under the curve (AUC) values obtained by ROC analysis for 3-year and 5-year CSS prediction significantly surpassed those of the AJCC model. The nomograms were validated using calibration and decision curves, confirming their clinical utility and superior predictive accuracy. The NRI and IDI indicated the enhanced predictive capability of the nomogram model. CONCLUSION: The SEER-based nomogram offers a significantly superior ability to predict CSS in SBA patients, supporting its potential application in clinical decision-making and personalized approaches to managing SBA to improve survival outcomes.

Combining KRAS gene status with preoperative D­dimer levels as a predictive marker of venous thromboembolism risk in patients with resectable colorectal cancer: A prospective cohort study.

Colorectal cancer (CRC), one of the most prevalent types of cancer, is accompanied by a notably high incidence of thrombotic complications. The present study aimed to elucidate the association between KRAS mutations and hypercoagulability in operable CRC. The prognostic value of preoperative D-dimer levels was also investigated, thus providing novel insights into the development of therapeutic strategies to enhance patient survival and diminish morbidity. Therefore, a prospective analysis of 333 CRC cases post-surgery at Yan'an Hospital Affiliated to Kunming Medical University, between May 2019 and October 2022 was performed. Data on demographics, tumor characteristics and D-dimer levels were compiled from the electronic health records. Venous thromboembolism (VTE) was diagnosed by doppler or computed tomography angiography, with D-dimer thresholds set at 550 and 1,650 µg/l. KRAS mutations at codons 12 and 13 were assessed in a subset of 56 cases. Subsequently, the factors affecting the hypercoagulable state in these patients were prospectively analyzed, focusing on the pivotal role of KRAS. The results showed that KRAS mutations were associated with elevated preoperative D-dimer levels, with 1,076 µg/l compared with 485 µg/l in the wild-type cohort, indicative of a hypercoagulable state. Increased D-dimer levels were also associated with vascular invasion, distant metastases and a heightened risk of postoperative VTE. Furthermore, multivariate analyses identified KRAS mutations, distant metastases and vascular invasion as independent predictors of elevated D-dimer levels, with relative risk values of 2.912, 1.884 and 1.525, respectively. Conversely, sex, age, tumor location, differentiation grade, Ki67 index and tumor stage could not significantly affect D-dimer levels, thus indicating a complex interplay between tumor genetics and coagulation dysfunction in CRC. The current study suggested that the KRAS mutation status, distant metastasis and vascular invasion could be considered as independent risk factors of blood hypercoagulability in patients with CRC, potentially serving as prognostic factors for VTE risk.

Developing Sampling Weights for Statistical Analysis of Parent-Child Pair Data From the National Health Interview Survey.

The National Health Interview Survey (NHIS), conducted by the National Center for Health Statistics since 1957, is the principal source of information on the health of the U.S. civilian noninstitutionalized population. NHIS selects one adult (Sample Adult) and, when applicable, one child (Sample Child) randomly within a family (through 2018) or a household (2019 and forward). Sampling weights for the separate analysis of data from Sample Adults and Sample Children are provided annually by the National Center for Health Statistics. A growing interest in analysis of parent-child pair data using NHIS has been observed, which necessitated the development of appropriate analytic weights. Objective This report explains how dyad weights were created such that data users can analyze NHIS data from both Sample Children and their mothers or fathers, respectively. Methods Using data from the 2019 NHIS, adult-child pair-level sampling weights were developed by combining each pair's conditional selection probability with their household-level sampling weight. The calculated pair weights were then adjusted for pair-level nonresponse, and large sampling weights were trimmed at the 99th percentile of the derived sampling weights. Examples of analyzing parent-child pair data by means of domain estimation methods (that is, statistical analysis for subpopulations or subgroups) are included in this report. Conclusions The National Center for Health Statistics has created dyad or pair weights that can be used for studies using parent-child pairs in NHIS. This method could potentially be adapted to other surveys with similar sampling design and statistical needs.

Untangling the effects of multiple exposures with a common reference group in an epidemiologic study: A practical revisit.

When assessing multiple exposures in epidemiologic studies, epidemiologists often use multivariable regression models with main effects only to control for confounding. This method can mask the true effects of individual exposures, potentially leading to wrong conclusions. We revisited a simple, practical, and often overlooked approach to untangle effects of the exposures of interest, in which the combinations of all levels of the exposures of interest are recoded into a single, multicategory variable. One category, usually the absence of all exposures of interest, is selected as the common reference group (CRG). All other categories representing individual and joint exposures are then compared to the CRG using indicator variables in a regression model or in a 2×2 contingency table analysis. Using real data examples, we showed that using the CRG analysis results in estimates of individual and joint effects that are mutually comparable and free of each other's confounding effects, yielding a clear, accurate, intuitive, and simple summarization of epidemiologic study findings involving multiple exposures of interest.

[Remediation Effect and Mechanism of Biochar in Combination with Nitrogen Fertilizer on Cd-contaminated Paddy Soil].

Cadmium (Cd) heavy metal pollution has posed serious threats to soil health and the safe production utilization of agricultural products. A pot experiment was conducted to study the effects of biochar (BC) and nitrogen fertilizer with three levels, namely 2.6 g·pot-1 (N1), 3.5 g·pot-1 (N2), 4.4 g·pot-1 (N3) biochar combined with nitrogen fertilizer (BCN1, BCN2, and BCN3), on soil Cd fractions, Cd enrichment, the transport of rice, and soil enzyme activity, as well as the changes in microbial community composition and complex interactions between microorganisms through high-throughput sequencing. The results showed that biochar combined with nitrogen fertilizer led to the transformation of Cd from the exchangeable state to the residue state, and the proportion of the exchangeable state was significantly reduced by 6.2%-14.7%; by contrast, the proportion of the residue state increased by 18.6%-26.4% relative to that in CK. In addition, singular treatments of nitrogen fertilizer enhanced the accumulation capacities of Cd in roots, which increased by 22%-33.5% compared with that in CK. By contrast, the BC and BCN treatments reduced Cd accumulation in roots and the transfer capacity from stems to rice husks and husk to rice. Furthermore, the BCN treatments promoted soil enzyme activities (urease, acid phosphatase, invertase, and catalase). MiSeq sequencing showed that BCN treatments increased the abundance of the main species of soil bacterial microbes (such as Acidobacteriales, Solibacterales, Pedosphaerales, and Nitrospirales). Moreover, co-occurrence network analysis showed that the complexity of the soil bacterial network was enhanced under the N, BC, and BCN treatments. Overall, biochar combined with nitrogen fertilizer reduced soil Cd availability, inhibited the capacity of Cd accumulation and the transport of rice, and improved the soil eco-environmental quality. Thus, using BCN could be a feasible practice for the remediation of Cd-polluted agricultural soil.

National Center for Health Statistics' 2019 Research and Development Survey, RANDS 3.

Objective This report on the third round of the Research and Development Survey (RANDS 3) provides a general description of RANDS 3 and presents percentage estimates of selected demographic and health-related variables from the overall sample and by one set of experimental groups embedded in the survey. Statistical tests comparing estimates for the two randomized groups were conducted to evaluate the randomization. Methods NORC at the University of Chicago conducted RANDS 3 for the National Center of Health Statistics in 2019 using its AmeriSpeak Panel in web-only mode. To assess question-response patterns, probe questions and four sets of experiments were embedded in RANDS 3, with panelists randomized into two groups for each set of experiments. Participants in each group received questions with differences in wording, question-andresponse formats, or question order. Results Of the 4,255 people sampled, 2,646 completed RANDS 3 for a completion rate of 62.2% and a weighted cumulative response rate of 18.1%. Iterative raking was performed using demographic and selected health condition variables to calibrate the RANDS 3 sample to 2019 National Health Interview Survey (NHIS) estimates. As a result, the overall demographic distribution and percentages of asthma, diabetes, hypertension, and high cholesterol for the calibrated RANDS 3 sample aligned with the percentages estimated from the 2019 NHIS. The distributions of demographic and healthrelated variables were comparable between the two randomized groups examined except for ever-diagnosed hypertension. Conclusion As part of a research series using probability-based survey panels, RANDS 3 included health-related questions with a focus on disability and opioids. Because RANDS is an ongoing research platform, a variety of persistent and emergent research questions relating to survey methodology will continue to be examined in current and future rounds of RANDS.

Multiple Imputation of Missing Complex Survey Data using SAS®: A Brief Overview and An Example Based on the Research and Development Survey (RANDS).

Multiple imputation (MI) is a widely used analytic approach to address missing data problems. SAS® (SAS Institute Inc, Cary, N.C.) has established MI procedures including PROC MI and PROC MIANALYZE. We illustrate the use of these procedures for conducting MI analysis of complex survey data by an example from RANDS. Section 1 contains the introduction. Section 2 includes some necessary methodological background. Section 3 shows a MI example with an arbitrary missing data pattern. Section 4 concludes the paper with a discussion.

A multiple imputation-based sensitivity analysis approach for regression analysis with a missing not at random covariate.

Missing covariate problems are common in biomedical and electrical medical record data studies while evaluating the relationship between a biomarker and certain clinical outcome, when biomarker data are not collected for all subjects. However, missingness mechanism is unverifiable based on observed data. If there is a suspicion of missing not at random (MNAR), researchers often perform sensitivity analysis to evaluate the impact of various missingness mechanisms. Under the selection modeling framework, we propose a sensitivity analysis approach with a standardized sensitivity parameter using a nonparametric multiple imputation strategy. The proposed approach requires fitting two working models to derive two predictive scores: one for predicting missing covariate values and the other for predicting missingness probabilities. For each missing covariate observation, the two predictive scores along with the pre-specified sensitivity parameter are used to define an imputing set. The proposed approach is expected to be robust against mis-specifications of the selection model and the sensitivity parameter since the selection model and the sensitivity parameter are not directly used to impute missing covariate values. A simulation study is conducted to study the performance of the proposed approach when MNAR is induced by Heckman's selection model. Simulation results show the proposed approach can produce plausible regression coefficient estimates. The proposed sensitivity analysis approach is also applied to evaluate the impact of MNAR on the relationship between post-operative outcomes and incomplete pre-operative Hemoglobin A1c level for patients who underwent carotid intervetion for advanced atherosclerotic disease.

Calibration Weighting Methods for the National Center for Health Statistics Research and Development Survey.

Objectives The Research and Development Survey (RANDS) is a series of web-based, commercial panel surveys that have been conducted by the National Center for Health Statistics (NCHS) since 2015. RANDS was designed for methodological research purposes,including supplementing NCHS' evaluation of surveys and questionnaires to detect measurement error, and exploring methods to integrate data from commercial survey panels with high-quality data collections to improve survey estimation. The latter goal of improving survey estimation is in response to limitations of web surveys, including coverage and nonresponse bias. To address the potential bias in estimates from RANDS,NCHS has investigated various calibration weighting methods to adjust the RANDS panel weights using one of NCHS' national household surveys, the National Health Interview Survey. This report describes calibration weighting methods and the approaches used to calibrate weights in web-based panel surveys at NCHS.

Reduced Access to Preventive Care Due to the COVID-19 Pandemic, by Chronic Disease Status and Race and Hispanic Origin, United States, 2020-2021.

OBJECTIVES: The COVID-19 pandemic has disproportionately affected racial and ethnic minority populations in the United States. The National Center for Health Statistics adapted the Research and Development Survey (RANDS), a commercial panel survey, to track selected health outcomes during the pandemic using the series RANDS during COVID-19 (RC-19). We examined access to preventive care among adults by chronic condition status, race, and Hispanic origin. METHODS: NORC at the University of Chicago conducted RC-19 among US adults in 3 rounds (June-July 2020 [round 1, N = 6800], August 2020 [round 2, N = 5981], and May-June 2021 [round 3, N = 5458]) via online survey and telephone. We evaluated reduced access to ≥1 type of preventive care due to the pandemic in the past 2 months for each round by using logistic regression analysis stratified by chronic condition status and race and Hispanic origin, adjusting for sociodemographic and health variables. RESULTS: Overall, 35.8% of US adults reported missing ≥1 type of preventive care in the previous 2 months in round 1, 26.0% in round 2, and 11.2% in round 3. Reduced access to preventive care was significantly higher among adults with ≥1 chronic condition (vs no chronic conditions) in rounds 1 and 2 (adjusted odds ratios [aOR)] = 1.5 and 1.4, respectively). Compared with non-Hispanic White adults, non-Hispanic Black adults reported significantly lower reduced access to preventive care in round 1 (aOR = 0.7), and non-Hispanic Other adults reported significantly higher reduced access to preventive care in round 2 (aOR = 1.5). CONCLUSIONS: Our findings may inform policies and programs for people at risk of reduced access to preventive care.

Multiple imputation of missing data with skip-pattern covariates: a comparison of alternative strategies.

Multiple imputation (MI) is a widely used approach to address missing data issues in surveys. Variables included in MI can have various distributional forms with different degrees of missingness. However, when variables with missing data contain skip patterns (i.e. questions not applicable to some survey participants are thus skipped), implementation of MI may not be straightforward. In this research, we compare two approaches for MI when skip-pattern covariates with missing values exist. One approach imputes missing values in the skip-pattern variables only among applicable subjects (denoted as imputation among applicable cases (IAAC)). The second approach imputes skip-pattern covariates among all subjects while using different recoding methods on the skip-pattern variables (denoted as imputation with recoded non-applicable cases (IWRNC)). A simulation study is conducted to compare these methods. Both approaches are applied to the 2015 and 2016 Research and Development Survey data from the National Center for Health Statistics.

Propensity-score Adjusted Estimates for Selected Health Outcomes From the Research and Development Survey.

To evaluate the quality of web surveys, the National Center for Health Statistics' Division of Research and Methodology has been conducting a series of studies with survey data from commercially recruited panels,referred to as the Research and Development Survey (RANDS). This report describes the propensity-score adjusted estimates from the second round of RANDS (RANDS 2) using the 2016 National Health Interview Survey (NHIS).

Variable inclusion strategies through directed acyclic graphs to adjust health surveys subject to selection bias for producing national estimates.

Along with the rapid emergence of web surveys to address time-sensitive priority topics, various propensity score (PS)-based adjustment methods have been developed to improve population representativeness for nonprobability- or probability-sampled web surveys subject to selection bias. Conventional PS-based methods construct pseudo-weights for web samples using a higher-quality reference probability sample. The bias reduction, however, depends on the outcome and variables collected in both web and reference samples. A central issue is identifying variables for inclusion in PS-adjustment. In this paper, directed acyclic graph (DAG), a common graphical tool for causal studies but largely under-utilized in survey research, is used to examine and elucidate how different types of variables in the causal pathways impact the performance of PS-adjustment. While past literature generally recommends including all variables, our research demonstrates that only certain types of variables are needed in PS-adjustment. Our research is illustrated by NCHS' Research and Development Survey, a probability-sampled web survey with potential selection bias, PS-adjusted to the National Health Interview Survey, to estimate U.S. asthma prevalence. Findings in this paper can be used by National Statistics Offices to design questionnaires with variables that improve web-samples' population representativeness and to release more timely and accurate estimates for priority topics.

Co-application of biochar and nitrogen fertilizer promotes rice performance, decreases cadmium availability, and shapes rhizosphere bacterial community in paddy soil.

Cadmium (Cd) contamination in soil has posed a great threat to crop safety and yield as well as soil quality. Biochar blended with nitrogen fertilizer have been reported to be effective in remediating Cd-contaminated soil. However, the influence of co-application of biochar and nitrogen fertilizer on the Cd bioavailability, rice yield and soil microbiome remains unclear. In this study, eight different treatments including control (CK), 5% biochar (B), 2.6, 3.5, 4.4 g/pot nitrogen fertilizers (N1, N2 and N3), and co-application of biochar and nitrogen fertilizers (BN1, BN2, BN3) were performed in a pot experiment with paddy soil for observations in an entire rice cycle growth period. Results showed single N increased soil available Cd content and Cd uptake in edible part of rice, while the soil available Cd content significantly decreased by 14.8% and 7.4%-11.1% under the B, BN treatments, and the Cd content in edible part of rice was significantly reduced by 35.1% and 18.5%-26.5%, respectively. Besides, B, N and BN treatments significantly increased the yield of rice by 14.3%-86.6% compared with CK, and the highest yield was gained under BN3 treatment. Soil bacterial diversity indices (Shannon, Chao1, observed species and PD whole tree index) under N2, N3 were generally improved. Cluster analysis indicated that bacterial community structures under BN treatments differed from those of CK and single N treatments. BN treatments enhanced the abundances of key bacterial phylum such as Acidobacteria, positively associated with yield, and increased the abundance of Spirochaetes, negatively correlated to soil available Cd and Cd uptake of rice. Furthermore, the regression path analysis (RPA) revealed that pH, organic matter (OM), alkaline hydrolysis of nitrogen (AHN) and available Cd were the major properties influencing Cd content in edible part of rice. Redundancy analysis (RDA) revealed that pH and available Cd played key role in shaping soil bacterial community. Thus, BN is a feasible practice for the improvements of rice growth and remediation of Cd-polluted soil.

A SEMIPARAMETRIC MULTIPLE IMPUTATION APPROACH TO FULLY SYNTHETIC DATA FOR COMPLEX SURVEYS.

Data synthesis is an effective statistical approach for reducing data disclosure risk. Generating fully synthetic data might minimize such risk, but its modeling and application can be difficult for data from large, complex surveys. This article extended the two-stage imputation to simultaneously impute item missing values and generate fully synthetic data. A new combining rule for making inferences using data generated in this manner was developed. Two semiparametric missing data imputation models were adapted to generate fully synthetic data for skewed continuous variable and sparse binary variable, respectively. The proposed approach was evaluated using simulated data and real longitudinal data from the Health and Retirement Study. The proposed approach was also compared with two existing synthesis approaches: (1) parametric regressions models as implemented in IVEware; and (2) nonparametric Classification and Regression Trees as implemented in synthpop package for R using real data. The results show that high data utility is maintained for a wide variety of descriptive and model-based statistics using the proposed strategy. The proposed strategy also performs better than existing methods for sophisticated analyses such as factor analysis.

Isobaric tags for relative and absolute quantitation (iTRAQ)-based proteomic analysis of mRNA splicing relevant proteins in aging HSPCs.

BACKGROUND: HSPC (hematopoietic stem/progenitor cell) aging was closely associated with the organism aging, senile diseases and hematopoietic related diseases. Therefore, study on HSPC aging is of great significance to further elucidate the mechanisms of aging and to treat hematopoietic disease resulting from HSPC aging. Little attention had been paid to mRNA splicing as a mechanism underlying HSPC senescence. RESULTS: We used our lab's patented in vitro aging model of HSPCs to analyze mRNA splicing relevant protein alterations with iTRAQ-based proteomic analysis. We found that not only the notable mRNA splicing genes such as SR, hnRNP, WBP11, Sf3b1, Ptbp1 and U2AF1 but also the scarcely reported mRNA splicing relevant genes such as Rbmxl1, Dhx16, Pcbp2, Pabpc1 were significantly down-regulated. We further verified their gene expressions by qRT-PCR. In addition, we reported the effect of Spliceostatin A (SSA), which inhibits mRNA splicing in vivo and in vitro, on HSPC aging. CONCLUSIONS: It was concluded that mRNA splicing emerged as an important factor for the vulnerability of HSPC aging. This study improved our understanding of the role of mRNA splicing in the HSPC aging process.

The National Center for Health Statistics' 2015 and 2016 Research and Development Surveys.

Objective: This report provides a general description of the background and operation of the first two rounds of the Research and Development Survey (RANDS), a series of cross-sectional surveys from probability-sampled commercial survey panels. The Division of Research and Methodology of the National Center for Health Statistics (NCHS) conducted the first two rounds of RANDS in 2015 and 2016. RANDS 1 and 2 are being used primarily for question design evaluation and for investigating statistical methodologies for estimation. Methods: NCHS contracted with Gallup, Inc. to conduct RANDS 1 in Fall 2015 and RANDS 2 in Spring 2016. RANDS 1 and 2 were conducted using a web survey mode and included survey questions from the National Health Interview Survey (NHIS) that were specifically chosen to provide comparison and evaluation of the survey methodology properties of web surveys and traditional household surveys. In this report, some demographic and health estimates are provided from both sources to describe the RANDS data. Results: In RANDS 1, 2,304 out of the original 9,809 invited panel members completed the survey, for a completion rate of 23.5%. In RANDS 2, 2,480 of the initial 8,231 invited respondents completed the survey, for a completion rate of 30.1%. RANDS 1 and 2 participants were similar to the quarterly NHIS participants with respect to sex, census region, and whether they had worked for pay in the previous week. Other characteristics varied, including age, race and ethnicity, and income. Most health estimates differed between RANDS and NHIS. Public-use versions of the RANDS data can be found at: https://www.cdc.gov/nchs/rands. Conclusion: RANDS is an ongoing platform for research to understand the properties of probability-sampled recruited panels of primarily web users, investigating and developing statistical methods for using such data in conjunction with large nationally representative health surveys, and for extending question-design evaluations.

A passive and inclusive strategy to impute missing values of a composite categorical variable with an application to determine HIV transmission categories.

PURPOSE: Multiple imputation (MI) is a widely acceptable approach to missing data problems in epidemiological studies. Composite variables are often used to summarize information from multiple, correlated items. This study aims to assess and compare different MI methods for handling missing categorical composite variables. METHODS: We investigate the problem in the context of a real application: estimating the prevalence of HIV transmission category, which is a composite variable generated by applying a hierarchical algorithm to a group of binary risk source variables from a national program data set. We use simulation studies to compare and assess the performance of alternative MI strategies. These methods include the active imputation, just another variable, and the passive imputation approaches. RESULTS: Our study suggests that the passive imputation approach performs better than the direct imputation approach and the inclusive and general imputation model (i.e. passive imputation with interactions) performs the best. There is no need to embed the information from the variable-combining algorithm in the passive imputation modeling. CONCLUSION: We recommend practitioners adopting an inclusive and general passive imputation modeling strategy.

A multiple imputation-based sensitivity analysis approach for data subject to missing not at random.

Missingness mechanism is in theory unverifiable based only on observed data. If there is a suspicion of missing not at random, researchers often perform a sensitivity analysis to evaluate the impact of various missingness mechanisms. In general, sensitivity analysis approaches require a full specification of the relationship between missing values and missingness probabilities. Such relationship can be specified based on a selection model, a pattern-mixture model or a shared parameter model. Under the selection modeling framework, we propose a sensitivity analysis approach using a nonparametric multiple imputation strategy. The proposed approach only requires specifying the correlation coefficient between missing values and selection (response) probabilities under a selection model. The correlation coefficient is a standardized measure and can be used as a natural sensitivity analysis parameter. The sensitivity analysis involves multiple imputations of missing values, yet the sensitivity parameter is only used to select imputing/donor sets. Hence, the proposed approach might be more robust against misspecifications of the sensitivity parameter. For illustration, the proposed approach is applied to incomplete measurements of level of preoperative Hemoglobin A1c, for patients who had high-grade carotid artery stenosisa and were scheduled for surgery. A simulation study is conducted to evaluate the performance of the proposed approach.