Search

Background: The American Cancer Society (ACS) and the NCI collaborate every 5 to 8 years to update the methods for estimating the numbers of new cancer cases and deaths in the current year for the U.S. and individual states. Herein, we compare our current projection methodology with the next generation of statistical models. Methods: A validation study was conducted comparing current projection methods (vector autoregression for incidence; Joinpoint regression for mortality) with the Bayes state-space method and novel Joinpoint algorithms. Incidence data from 1996–2010 were projected to 2014 using two inputs: modeled data and observed data with modeled where observed were missing. For mortality, observed data from 1995 to 2009, 1996 to 2010, 1997 to 2011, and 1998 to 2012, each projected 3 years forward to 2012 to 2015. Projection methods were evaluated using the average absolute relative deviation (AARD) between observed counts (2014 for incidence, 2012–2015 for mortality) and estimates for 47 cancer sites nationally and 21 sites by state. Results: A novel Joinpoint model provided a good fit for both incidence and mortality, particularly for the most common cancers in the U.S. Notably, the AARD for cancers with cases in 2014 exceeding 49,000 for this model was 3.4%, nearly half that of the current method (6.3%). Conclusions: A data-driven Joinpoint algorithm had versatile performance at the national and state levels and will replace the ACS's current methods. Impact: This methodology provides estimates of cancer data that are not available for the current year, thus continuing to fill an important gap for advocacy, research, and public health planning.

Joinpoint regression can model trends in time-specific estimates from aggregated data. These methods have been developed mainly for nonsurvey data such as cancer registry data assuming that the time-specific estimates are uncorrelated from time point to time point. This independence assumption can be violated for trends in time-specific estimates from complex survey samples due to using the same primary sampling units across time and, therefore, the full variance–covariance matrix of the time-specific estimates should be incorporated into the regression model fitting. This article extends these joinpoint methods for analyzing complex survey data within the National Cancer Institute’s Joinpoint software and empirically compares the extended method to existing methods for analyses of time trends in three surveys.

Background Little is known about the real-world care of young adult (YA) females (aged 20-39 years) with breast cancer. This study describes factors associated with the receipt of guideline-concordant care (GCC) among YAs. Methods The authors identified 1259 YA women with invasive breast cancer diagnosed in 2013 in the National Cancer Institute's Patterns of Care study. Hospital records were re-abstracted, and treatment was verified. Using the National Comprehensive Cancer Network's 2013 breast cancer guidelines, the authors assessed the receipt of GCC by cancer subtype among a subset of YAs (n = 952). Associations between sociodemographic and clinical factors and GCC receipt were examined. Results Most YAs were 35 to 39 years old (51.2%) and partnered (56.4%); half had hormone receptor-positive (HR+)/human epidermal growth factor receptor 2-negative (HER2-) tumors. GCC was found for 81.7% of YAs. Relationships between sociodemographic and clinical factors and GCC receipt differed by subtype. Stage was the only significant predictor of GCC receipt for all subtypes (stage II vs III: odds ratio [OR] for HR+/HER2+, 0.20; 95% confidence interval [CI], 0.08-0.50; OR for HR-/HER2+, 0.13; 95% CI, 0.07-0.25; OR for HR-/HER2-, 3.86; 95% CI, 1.55-9.62; OR for HR+/HER2-, 2.81; 95% CI, 1.63-5.80). Conclusions GCC is high among YAs with breast cancer. The effects of sociodemographic factors and treatment facility size on GCC differ by subtype. Consistent with recommendations, tumor biology, not age, is associated with GCC for all subtypes. Future studies should assess the effect of GCC on survival among YAs.

This chapter explores whether Californians in same-sex legal marriages and partnerships reported lower levels of psychological distress than other adult Californians after the 2008 California Supreme Court Decision that legalized same-sex marriage. We pooled 10 years of California Health Interview Survey (CHIS) data and employ a T1-T2 design to approximate a time series design. Dependent variables include overall self-related health, psychological distress, and household income. Independent variables include sexual identity and same-sex spouse. Bi-variate analyses compared self-reported mental and physical health between the two periods. We found decreased reports of poorer health and increased reports of very good health among gay men and lesbian women with legal spouses. Psychological distress decreased for legally coupled gay men and lesbians while increased slightly among unpartnered lesbian women and gay men. Household income increased among coupled lesbian women and gay men and decreased among others. Our project demonstrated positive health influences for Californians with legal same-sex spouses. We recommend future research projects that explore whether and how same- and opposite-sex marriage benefits health, well-being, and prosperity, and for marital status survey questions that are inclusive of sexual and gender identities and elicit the sex/gender of a respondent’s spouse.

Background: The American Cancer Society (ACS) and the NCI collaborate every 5–8 years to update the methods for estimating numbers of new cancer cases and deaths in the current year in the United States and in every state and the District of Columbia. In this article, we reevaluate the statistical method for estimating unavailable historical incident cases which are needed for projecting the current year counts. Methods: We compared the current county-level model developed in 2012 (M0) with three new models, including a state-level mixed effect model (M1) and two state-level hierarchical Bayes models with varying random effects (M2 and M3). We used 1996–2014 incidence data for 16 sex-specific cancer sites to fit the models. An average absolute relative deviation (AARD) comparing the observed with the model-specific predicted counts was calculated for each site. Models were also cross-validated for six selected sex-specific cancer sites. Results: For the cross-validation, the AARD ranged from 2.8% to 33.0% for M0, 3.3% to 31.1% for M1, 6.6% to 30.5% for M2, and 10.4% to 393.2% for M3. M1 encountered the least technical issues in terms of model convergence and running time. Conclusions: The state-level mixed effect model (M1) was overall superior in accuracy and computational efficiency and will be the new model for the ACS current year projection project. Impact: In addition to predicting the unavailable state-level historical incidence counts for cancer surveillance, the updated algorithms have broad applicability for disease mapping and other activities of public health planning, advocacy, and research.

Background Life expectancy is increasingly incorporated in evidence-based screening and treatment guidelines to facilitate patient-centered clinical decision-making. However, life expectancy estimates from standard life tables do not account for health status, an important prognostic factor for premature death. This study aims to address this research gap and develop life tables incorporating the health status of adults in the United States. Methods Data from the National Health Interview Survey (1986–2004) linked to mortality follow-up through to 2006 (age ≥ 40, n = 729,531) were used to develop life tables. The impact of self-rated health (excellent, very good, good, fair, poor) on survival was quantified in 5-year age groups, incorporating complex survey design and weights. Life expectancies were estimated by extrapolating the modeled survival probabilities. Results Life expectancies incorporating health status differed substantially from standard US life tables and by health status. Poor self-rated health more significantly affected the survival of younger compared to older individuals, resulting in substantial decreases in life expectancy. At age 40 years, hazards of dying for white men who reported poor vs. excellent health was 8.5 (95% CI: 7.0,10.3) times greater, resulting in a 23-year difference in life expectancy (poor vs. excellent: 22 vs. 45), while at age 80 years, the hazards ratio was 2.4 (95% CI: 2.1, 2.8) and life expectancy difference was 5 years (5 vs. 10). Relative to the US general population, life expectancies of adults (age Conclusions Considerable shortage in life expectancy due to poor self-rated health existed. The life table developed can be helpful by including a patient perspective on their health and be used in conjunction with other predictive models in clinical decision making, particularly for younger adults in poor health, for whom life tables including comorbid conditions are limited.

Objective There has been steady progress in reducing cancer mortality in the United States; however, this progress hasn't been evenly distributed across regions. This paper assesses trends in cancer mortality salience (CMS), that is, agreeing that getting cancer is a death sentence, over time in the United States and examines correlates of CMS. Methods Data from three administrations of the Health Information National Trends Survey (HINTS), gathered in 2008, 2013, and 2017, were merged, resulting in a total sample of 10,063 respondents. Trends in changes in CMS over time were examined as well as maps of the distribution of CMS in the United States. A logistic regression model was conducted, regressing CMS on a set of sociodemographic, psychological, health-related, and environmental predictors. Results The aggregated percentage of US adults who agreed with the CMS statement changed over time and was modified by age. Geographic distribution of agreement with CMS was inconsistent across the United States. In the adjusted logistic model, perceived health (worse health), cancer prevention, fatalism, and confusion, and cancer status (no cancer) were all significantly associated with CMS. There was also a significant interaction between survey year and age. Conclusion Despite recent information that cancer mortality rates are decreasing, most US adults still see cancer as a death sentence and this is especially an issue in certain subgroups. These findings have ramifications for groups of people who may be at risk for developing cancer given their attitudes and beliefs that there isn't much they can do to prevent or control it.

Introduction The workplace and home are sources of exposure to secondhand smoke, a serious health hazard for nonsmoking adults and children. Smoke-free workplace policies and home rules protect nonsmoking individuals from secondhand smoke and help individuals who smoke to quit smoking. However, estimated population coverages of smoke-free workplace policies and home rules are not typically available at small geographic levels such as counties. Model-based small-area estimation techniques are needed to produce such estimates. Methods Self-reported smoke-free workplace policies and home rules data came from the 2014–2015 Tobacco Use Supplement to the Current Population Survey. County-level design-based estimates of the two measures were computed and linked to county-level relevant covariates obtained from external sources. Hierarchical Bayesian models were then built and implemented through Markov Chain Monte Carlo methods. Results Model-based estimates of smoke-free workplace policies and home rules were produced for 3134 (of 3143) US counties. In 2014–2015, nearly 80% of US adult workers were covered by smoke-free workplace policies, and more than 85% of US adults were covered by smoke-free home rules. We found large variations within and between states in the coverage of smoke-free workplace policies and home rules. Conclusions The small-area modeling approach efficiently reduced the variability that was attributable to small sample size in the direct estimates for counties with data and predicted estimates for counties without data by borrowing strength from covariates and other counties with similar profiles. The county-level modeled estimates can serve as a useful resource for tobacco control research and intervention. Implications Detailed county- and state-level estimates of smoke-free workplace policies and home rules can help identify coverage disparities and differential impact of smoke-free legislation and related social norms. Moreover, this estimation framework can be useful for modeling different tobacco control variables and applied elsewhere, for example, to other behavioral, policy, or health related topics.

There is increased interest in eliminating health disparities in the United States and worldwide. Broadly defined, health disparities refer to preventable inequalities in health status, such as cancer to ethnicity, socioeconomic status, gender, education, environment, and geographic locations. To make informed health policy decisions, it is essential to precisely measure the magnitude of disparities and assess trends over time. The Health Disparities Calculator (HD*Calc) is free statistical software that calculates 11 commonly used measures of health disparities and provides corresponding 95% CIs for the 11 measures using either an analytic method or a Monte Carlo simulation–based method; however, the derivation of SEs and coverage properties of the CIs have not been formally evaluated. We used simulation studies to assess the coverage properties of these CIs. We have also conducted bias analyses for measures implemented in HD*Calc using age-adjusted cancer incidence rates from national, state, and county level SEER data. The results of these analyses indicate that HD*Calc should be used with caution to construct CIs for some health disparity measures when the proportion of zero event counts is greater than 25%.

Cancer screening prevalence from the Behavioral Risk Factor Surveillance System (BRFSS), designed to provide state-level estimates, and the National Health Interview Survey (NHIS), designed to provide national estimates, are used to measure progress in cancer control. A detailed description of the extent to which recent cancer screening estimates vary by key demographic characteristics has not been previously described. We examined national prevalence estimates for recommended breast, cervical, and colorectal cancer screening using data from the 2012 and 2014 BRFSS and the 2010 and 2013 NHIS. Treating the NHIS estimates as the reference, direct differences (DD) were calculated by subtracting NHIS estimates from BRFSS estimates. Relative differences were computed by dividing the DD by the NHIS estimates. Two-sample t-tests (2-tails), were performed to test for statistically significant differences. BRFSS screening estimates were higher than those from NHIS for breast (78.4% versus 72.5%; DD=5.9%, p

Breast and cervical cancer incidence vary by urbanicity, and several ecological factors could contribute to these patterns. In particular, cancer screening or other sociodemographic and health care system variables could explain geographic disparities in cancer incidence.Governmental and research sources provided data on 612 counties in the Surveillance, Epidemiology, and End Results program for rural-urban continuum code, socioeconomic status (SES) quintile, percent non-Hispanic White residents, density of primary care physicians, cancer screening, and breast and cervical cancer incidence rates (2009-2013). Ecological mediation analyses used weighted least squares regression to examine whether candidate mediators explained the relationship between urbanicity and cancer incidence.As urbanicity increased, so did breast cancer incidence (βˆ = 0.23; p .001). SES quintile and density of primary care physicians mediated this relationship, whereas percent non-Hispanic White suppressed it (all p .05); county-level mammography levels did not contribute to the relationship. After controlling for these variables, urbanicity and breast cancer incidence were no longer associated (βˆ = 0.11; p .05). In contrast, as urbanicity increased, cervical cancer incidence decreased (βˆ = -0.33; p .001). SES quintile and density of primary care physicians mediated this relationship (both p .05); percent non-Hispanic White and Pap screening levels did not contribute to the relationship. After controlling for these variables, the relationship between urbanicity and cervical cancer incidence remained significant (βˆ = -0.13; p .05).County-level SES and density of primary care physicians explained the relationships between urbanicity and breast and cervical cancer incidence. Improving these factors in more rural counties could ameliorate geographic disparities in breast and cervical cancer incidence.

Unit-level logistic regression models with mixed effects have been used for estimating small area proportions in the literature. Normality is commonly assumed for the random effects. Nonetheless, real data often show significant departures from normality assumptions of the random effects. To reduce the risk of model misspecification, we propose an adaptive hierarchical Bayes estimation approach in which the distribution of the random effect is chosen adaptively from the exponential power class of probability distributions. The richness of the exponential power class ensures the robustness of our hierarchical Bayes approach against departure from normality. We demonstrate the robustness of our proposed model using both simulated and real data. The results suggest that the proposed model works reasonably well to incorporate potential kurtosis of the random effects distribution.

Background National health surveys, such as the National Health Interview Survey (NHIS) and the Behavioral Risk Factor Surveillance System (BRFSS), collect data on cancer screening and smoking-related measures in the US noninstitutionalized population. These surveys are designed to produce reliable estimates at the national and state levels. However, county-level data are often needed for cancer surveillance and related research. Methods To use the large sample sizes of BRFSS and the high response rates and better coverage of NHIS, we applied multilevel models that combined information from both surveys. We also used relevant sources such as census and administrative records. By using these methods, we generated estimates for several cancer risk factors and screening behaviors that are more precise than design-based estimates. Results We produced reliable, modeled estimates for 11 outcomes related to smoking and to screening for female breast cancer, cervical cancer, and colorectal cancer. The estimates were produced for 3,112 counties in the United States for the data period from 2008 through 2010. Conclusion The modeled estimates corrected for potential noncoverage bias and nonresponse bias in the BRFSS and reduced the variability in NHIS estimates that is attributable to small sample size. The small area estimates produced in this study can serve as a useful resource to the cancer surveillance community.

Preventive behaviors established during adolescence can reduce cancer throughout the life span. Understanding the combinations of multiple behaviors, and how these behaviors vary across states, is important for identifying where additional interventions are needed. Using data on 2011-2015 vaccination, energy balance, and substance use from national surveys, we created state-level composite scores for adolescent cancer prevention. Hierarchical Bayesian linear mixed models were used to predict estimates for states with no data on select behaviors. We used a Monte Carlo procedure with 100,000 simulations to generate states’ ranks and 95% confidence intervals. Across states, hepatitis B vaccination was 84.3% to 97.1%, and human papillomavirus vaccination was 41.8% to 78.0% for girls and 19.0% to 59.3% for boys. For energy balance, 20.2% to 34.6% of adolescents met guidelines for physical activity, 4.1% to 15.8% for fruit and vegetable consumption, and 66.4% to 82.0% for healthy weight. For substance use, 82.5% to 93.5% reported abstaining from binge alcohol use, 84.3% to 95.4% from cigarette smoking, and 62.9% to 92.8% from marijuana use. (1) Rhode Island, (2) Colorado, (4) Hawaii and New Hampshire (tied), and (5) Vermont performed the best for adolescent cancer prevention, and (47) Missouri, (48) Arkansas, Mississippi, and South Carolina (tied), and (51) Kentucky performed the worst. However, 95% CIs around ranks often overlapped, indicating lack of statistical differences. Adolescent cancer prevention behaviors clustered into a composite index. States varied on their performance on this index, especially for states at the high and low extremes, but most states did not differ statistically. These findings can inform decision makers about where and how to intervene to improve cancer prevention among adolescents.

The Physical Activity Monitor component was introduced into the 2003-2004 National Health and Nutrition Examination Survey (NHANES) to collect objective information on physical activity including both movement intensity counts and ambulatory steps. Because of an error in the accelerometer device initialization process, the steps data were missing for all participants in several primary sampling units, typically a single county or group of contiguous counties, who had intensity count data from their accelerometers. To avoid potential bias and loss in efficiency in estimation and inference involving the steps data, we considered methods to accurately impute the missing values for steps collected in the 2003-2004 NHANES. The objective was to come up with an efficient imputation method that minimized model-based assumptions. We adopted a multiple imputation approach based on additive regression, bootstrapping and predictive mean matching methods. This method fits alternative conditional expectation (ace) models, which use an automated procedure to estimate optimal transformations for both the predictor and response variables. This paper describes the approaches used in this imputation and evaluates the methods by comparing the distributions of the original and the imputed data. A simulation study using the observed data is also conducted as part of the model diagnostics. Finally, some real data analyses are performed to compare the before and after imputation results. Published 2016. This article is a U.S. Government work and is in the public domain in the USA.

Previous research describing how informal cancer caregiving impacts employment has been conducted in small samples or a single disease site. This paper provides population-based estimates of the effect of informal cancer caregiving on employment and characterizes employment changes made by caregivers. The samples included cancer survivors with a friend or family caregiver, participating in either the Medical Expenditure Panel Survey Experiences with Cancer Survivorship Survey (ECSS) (n = 458) or the LIVESTRONG 2012 Survey for People Affected by Cancer (SPAC) (n = 4706). Descriptive statistics characterized the sample of survivors and their caregivers’ employment changes. Multivariable logistic regression identified predictors of caregivers’ extended employment changes, comprising time off and changes to hours, duties, or employment status. Among survivors with an informal caregiver, 25 % from the ECSS and 29 % from the SPAC reported that their caregivers made extended employment changes. Approximately 8 % of survivors had caregivers who took time off from work lasting ≥2 months. Caregivers who made extended employment changes were more likely to care for survivors: treated with chemotherapy or transplant; closer to diagnosis or end of treatment; who experienced functional limitations; and made work changes due to cancer themselves compared to caregivers who did not make extended employment changes. Many informal cancer caregivers make employment changes to provide care during survivors’ treatment and recovery. This study describes cancer caregiving in a prevalent sample of cancer survivors, thereby reflecting the experiences of individuals with many different cancer types and places in the cancer treatment trajectory.

Supplemental material, HEB839111_Supplemental_Material for Adolescent Behavioral Cancer Prevention in the United States: Creating a Composite Variable and Ranking States’ Performance by Jennifer L. Moss, Benmei Liu and Li Zhu in Health Education & Behavior

Background Having HIV/AIDS has been associated with a higher prevalence of smoking. Moreover, evidence suggests that people with HIV/AIDS who smoke have poorer treatment and survival outcomes. The HIV–smoking relationship is understudied in sub-Saharan Africa, where tobacco use patterns and HIV prevalence differ greatly from other world regions. Methods Cross-sectional data from the Demographic Health Surveys and AIDS Indicator Surveys, representing 25 sub-Saharan African countries, were pooled for analysis (n = 286850). The association between cigarette smoking and HIV status was analyzed through hierarchical logistic regression models. This study also examined the relationship between smokeless tobacco (SLT) use and HIV status. Results Smoking prevalence was significantly higher among men who had HIV/AIDS than among men who did not (25.90% vs 16.09%; p < .0001), as was smoking prevalence among women who had HIV/AIDS compared with women who did not (1.15% vs 0.73%; p < .001). Multivariate logistic regression revealed that the odds of smoking among people who had HIV/AIDS was 1.12 times greater than among people who did not when adjusting for socioeconomic, demographic, and sexual risk factors (adjusted OR = 1.12, 95% CI = 1.04% to 1.21%; p < .001). Similarly, multivariate logistic regression revealed that HIV-positive individuals were 34% more likely to use SLT than HIV-negative individuals (adjusted OR = 1.34, 95% CI = 1.17% to 1.53%). Conclusion Having HIV was associated with a greater likelihood of smoking cigarettes as well as with using SLT in sub-Saharan Africa. These tobacco use modalities were also associated with male sex and lower socioeconomic status. Implications This study shows that in sub-Saharan Africa, as in more studied world regions, having HIV/AIDS is associated with a higher likelihood of smoking cigarettes when adjusting for demographic, socioeconomic, and sexual risk factors. This study also supports the literature stating that cigarette smoking is inversely associated with socioeconomic status, as evidenced by higher smoking prevalence among poorer individuals, less educated individuals, and manual and agricultural laborers.

Introduction This study statistically ranked states’ performance on adolescent substance use related to cancer risk (past-month cigarette smoking, binge alcohol drinking, and marijuana use). Methods Data came from 69,200 adolescent participants (50 states and the District of Columbia) in the National Survey on Drug Use and Health (NSDUH) and 450,050 adolescent participants (47 states) in the Youth Risk Behavior Surveillance System (YRBSS). Adolescents were aged 14 to 17 years. For 2011–2015, we estimated and ranked states’ prevalence of adolescent substance use. We calculated the ranks’ 95% confidence intervals (CIs) using a Monte Carlo method with 100,000 simulations. Spearman correlations examined consistency of ranks. Results Across states, the prevalence of cigarette smoking was 4.5% to 14.3% in NSDUH and 4.7% to 18.5% in YRBSS. Utah had the lowest prevalence (NSDUH: rank = 51 [95% CI, 47–51]; YRBSS: rank = 47 [95% CI, 46–47]), and states’ ranks across surveys were correlated (r = 0.66, P < .001). The prevalence of binge alcohol drinking was 5.9% to 14.3% (NSDUH) and 7.1% to 21.7% (YRBSS). Utah had the lowest prevalence (NSDUH: rank = 50 [95% CI, 40–51]; YRBSS: rank = 47 [95% CI, 47–47]), but ranks across surveys were weakly correlated (r = 0.38, P = .01). The prevalence of marijuana use was 6.3% to 18.7% (NSDUH) and 8.2% to 27.1% (YRBSS). Utah had the lowest prevalence of marijuana use (NSDUH: rank = 50 [95% CI = 33–51]; YRBSS: rank= 46 [95% CI, 46–46]), and ranks across surveys were correlated (r = 0.70, P < .001). Wide CIs for states ranked in the middle of each distribution obscured statistical differences among them. Conclusion Variability emerged across adolescent substance use behaviors and surveys (perhaps because of administration differences). Most states showed statistically equivalent performance on adolescent substance use. Adolescents in all states would benefit from efforts to reduce substance use, to prevent against lifelong morbidity.

BACKGROUND AND OBJECTIVES: In 2016, with the discontinuation of the Collaborative Staging system, the cancer surveillance community planned to rely on physician-assigned TNM stage documented in the medical record. The objectives of this study were to describe how often physician-assigned staging components were documented in the medical records accessible to the registrar and to assess the agreement of these physician-assigned components with registrar-assigned values. METHODS: Medical record documents for 282 routine cases from 5 cancer sites were selected from the Surveillance, Epidemiology, and End Results registries. First, the documents were evaluated to determine how often they contained the TNM staging components. Next, the available components were compared with values assigned by a panel of experienced cancer registrars. The agreement for each type of source document was estimated among 100 cases. RESULTS: Overall, the physician-assigned TNM components and stage groups were not often found in the medical record. Pathologic T and N were found most frequently (65% and 64%, respectively). Agreement between physician-assigned and registrar-assigned TNM components varied (cT = 57%, cN = 72%, pT = 83%, pN = 89%). For stage group, agreement was better when the stage group was documented more than once (clinical, 71%; pathologic, 67%). Path reports included valid pT and pN in 79% and 89% of cases, respectively. Oncology consultation notes provided valid cT for 83% of cases. Validity was lower for other document sources. CONCLUSIONS: The physician-assigned TNM components will rarely be documented in the medical record and available to the registrar. Collection of accurate stage information for cancer surveillance requires cancer registrars to review the full medical record and assign the TNM components required for stage.

The relative concentration index (RCI) and the absolute concentration index (ACI) have been widely used for monitoring health disparities with ranked health determinants. The RCI has been extended to allow value judgments about inequality aversion by Pereira in 1998 and by Wagstaff in 2002. Previous studies of the extended RCI have focused on survey sample data. This paper adapts the extended RCI for use with directly standardized rates (DSRs) calculated from population-based surveillance data. A Taylor series linearization (TL)-based variance estimator is developed and evaluated using simulations. A simulation-based Monte Carlo (MC) variance estimator is also evaluated as a comparison. Following Wagstaff's approach in 1991, we extend the ACI for use with DSRs. In all simulations, both the TL and MC methods produce valid variance estimates. The TL variance estimator has a simple, closed form that is attractive to users without sophisticated programming skills. The TL and MC estimators have been incorporated into a beta version of the National Cancer Institute's Health Disparities Calculator, a free statistical software tool that enables the estimation of 11 commonly used summary measures of health disparities for DSRs.

Background: Although screening mammography may contribute to decreases in breast cancer mortality in a population, it may also increase the risk of false positives, anxiety, and unnecessary and costly medical procedures in individuals. We report trends in self-reported non-normal screening mammography results, lumpectomies, and breast cancer in a representative sample of California women. Methods: Data were obtained from the 2001, 2005, and 2009 cross-sectional California Health Interview Surveys (CHIS) and weighted to the California population. CHIS employed a multistage sampling design to administer telephone surveys in 6 languages. Our study sample was restricted to women 40 years and older who reported a screening mammogram in the past 2 years. Sample sizes were 13,974 in 2001, 12,069 in 2005, and 15,552 in 2009. Women reporting non-normal results were asked whether they had an operation to remove the lump and, if so, whether the lump was confirmed as malignant. Findings: Between 2001 and 2009, the percent of California women who reported having been diagnosed with breast cancer was relatively stable. For each of the three age groups studied, the percentage of non-normal mammography results increased and the percentages of lumpectomies decreased and, for every woman reporting a diagnosis of breast cancer, three women reported a lumpectomy that turned out not to be cancer. This ratio was greater for younger women and less for older women. Conclusions: Despite relatively constant rates of breast cancer diagnosis from 2001 to 2009, the percentage of nonnormal mammography results increased and lumpectomies declined.

While it is clear that self-reported racial/ethnic discrimination is related to illness, there are challenges in measuring self-reported discrimination or unfair treatment. In the present study, we evaluate the psychometric properties of a self-reported instrument across racial/ethnic groups in a population-based sample, and we test and interpret findings from applying two different widely-used approaches to asking about discrimination and unfair treatment. Even though we found that the subset of items we tested tap into a single underlying concept, we also found that different groups are more likely to report on different aspects of discrimination. Whether race is mentioned in the survey question affects both frequency and mean scores of reports of racial/ethnic discrimination. Our findings suggest caution to researchers when comparing studies that have used different approaches to measure racial/ethnic discrimination and allow us to suggest practical empirical guidelines for measuring and analyzing racial/ethnic discrimination. No less important, we have developed a self-reported measure of recent racial/ethnic discrimination that functions well in a range of different racial/ethnic groups and makes it possible to compare how racial/ethnic discrimination is associated with health disparities among multiple racial/ethnic groups.

INTRODUCTION: The workplace is a major source of exposure to secondhand smoke from combustible tobacco products. Smokefree workplace policies protect nonsmoking workers from secondhand smoke and help workers who smoke quit. This study examined changes in self-reported smokefree workplace policy coverage among U.S. workers from 2003 to 2010–2011. METHODS: Data came from the 2003 (n = 74,728) and 2010–2011 (n = 70,749) waves of the Tobacco Use Supplement to the Current Population Survey. Among employed adults working indoors, a smokefree workplace policy was defined as a self-reported policy at the respondent’s workplace that did not allow smoking in work areas and public/common areas. Descriptive statistics were used to assess smokefree workplace policy coverage at two timepoints overall, by occupation, and by state. RESULTS: The proportion of U.S. workers covered by smokefree workplace policies increased from 77.7% in 2003 to 82.8% in 2010–2011 (p < .00001). The proportion of workers reporting smokefree workplace policy coverage increased in 21 states (p < .001) and decreased in two states (p < .001) over this period. In 2010–2011, by occupation, this proportion ranged from 74.3% for blue collar workers to 84.9% for white collar workers; by state, it ranged from 63.3% in Nevada to 92.6% in Montana. CONCLUSIONS: From 2003 to 2010–2011, self-reported smokefree workplace policy coverage among indoor adult workers increased nationally, and occupational coverage disparities narrowed. However, coverage remained unchanged in half of states, and disparities persisted across occupations and states. Accelerated efforts are warranted to ensure that all workers are protected by smokefree workplace policies. IMPLICATIONS: This study assessed changes in the proportion of indoor workers reporting being covered by smokefree workplace policies from 2003 to 2010–2011 overall and by occupation and by state, using data from the Tobacco Use Supplement to the Current Population Survey. The findings indicate that smokefree workplace policy coverage among U.S. indoor workers has increased nationally, with occupational coverage disparities narrowing. However, coverage remained unchanged in half of states, and disparities persisted across occupations and states. Accelerated efforts are warranted to ensure that all workers are protected by smokefree workplace policies.

Objectives: Identifying the best and worst states for coverage of cancer-preventing vaccines (hepatitis B [HepB] and human papillomavirus [HPV]) may guide public health officials in developing programs, such as promotion campaigns. However, acknowledging the imprecision of coverage and ranks is important for avoiding overinterpretation. The objective of this study was to examine states’ vaccination coverage and ranks, as well as the imprecision of these estimates, to inform public health decision making. Methods: We used data on coverage of HepB and HPV vaccines among adolescents aged 13-17 from the 2011-2015 National Immunization Survey-Teen (n = 103 729 from 50 US states and Washington, DC). We calculated coverage, 95% confidence intervals (CIs), and ranks for vaccination coverage in each state, and we generated simultaneous 95% CIs for ranks using a Monte Carlo method with 100 000 simulations. Results: Across years, HepB vaccination coverage was 92.2% (95% CI, 91.8%-92.5%; states’ range, 84.3% in West Virginia to 97.0% in Connecticut). HPV vaccination coverage was 57.4% (95% CI, 56.6%-58.2%; range, 41.8% in Kansas to 78.0% in Rhode Island) for girls and 31.0% (95% CI, 30.3%-31.8%; range, 19.0% in Utah to 59.3% in Rhode Island) for boys. States with the highest and lowest ranks generally had narrow 95% CIs; for example, Rhode Island was ranked first (95% CI, 1-1) and Kansas was ranked 51st (95% CI, 49-51) for girls’ HPV vaccination. However, states with intermediate ranks had wider and more imprecise 95% CIs; for example, New York was 26th for girls’ HPV vaccination coverage, but its 95% CI included ranks 18-35. Conclusions: States’ ranks of coverage of cancer-preventing vaccines were imprecise, especially for states in the middle of the range; thus, performance rankings presented without measures of imprecision could be overinterpreted. However, ranks can highlight high-performing and low-performing states to target for further research and vaccination promotion programming.

Understanding statistical differences in states’ percentages and ranks of adolescents meeting health behavior guidelines can guide policymaking. Data came from 531,777 adolescents (grades 9–12) who completed the Youth Risk Behavior Surveillance System survey in 2011, 2013, or 2015. We measured the percentage of adolescents in each state that met guidelines for physical activity, fruit and vegetable (F&V) consumption, and healthy weight status. Then we ranked states and calculated the ranks’ 95% CI’s using a Monte Carlo method with 100,000 simulations. We repeated these analyses stratified by sex (female or male) or race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic/Latino, or other). Pearson’s and Spearman’s correlation coefficients examined consistency in the percentages and ranks (respectively) across behaviors and subgroups. Meeting the physical activity and F&V consumption guidelines was relatively rare among adolescents (25.8% [95% CI=25.2%–26.4%] and 8.0% [95% CI=7.6%–8.3%], respectively), while meeting the healthy weight guideline was common (71.5% [95% CI=70.7%–72.3%]). At the state level, percentages of adolescents meeting these guidelines were statistically similar; states’ ranks had wide CI’s, resulting in considerable overlap (i.e., statistical equivalence). For each behavior, states’ percentages and ranks were moderately to highly correlated across adolescent subgroups (Pearson’s r=0.33–0.96; Spearman’s r=0.42–0.96), but across behaviors, only F&V consumption and healthy weight were correlated (Pearson’s r=0.34; Spearman’s r=0.37). Adolescents in all states could benefit from initiatives to support cancer prevention behaviors, especially physical activity and F&V consumption. Programs in states that ranked highly on all assessed health behaviors could be adapted for dissemination in lower-performing states.

Cross-cultural variability in respondent processing of survey questions may bias results from multiethnic samples. We analyzed behavior codes, which identify difficulties in the interactions of respondents and interviewers, from a discrimination module contained within a field test of the 2007 California Health Interview Survey. In all, 553 (English) telephone interviews yielded 13,999 interactions involving 22 items. Multilevel logistic regression modeling revealed that respondent age and several item characteristics (response format, customized questions, length, and first item with new response format), but not race/ethnicity, were associated with interactional problems. These findings suggest that item function within a multi-cultural, albeit English language, survey may be largely influenced by question features, as opposed to respondent characteristics such as race/ethnicity.

The National Cancer Institute's Surveillance, Epidemiology, and End Results Program releases research files of cancer registry data. These files include geographic information at the county level, but no finer. Access to finer geography, such as census tract identifiers, would enable richer analyses-for example, examination of health disparities across neighborhoods. To date, tract identifiers have been left off the research files because they could compromise the confidentiality of patients' identities. We present an approach to inclusion of tract identifiers based on multiply imputed, synthetic data. The idea is to build a predictive model of tract locations, given patient and tumor characteristics, and randomly simulate the tract of each patient by sampling from this model. For the predictive model, we use multivariate regression trees fitted to the latitude and longitude of the population centroid of each tract. We implement the approach in the registry data from California. The method results in synthetic data that reproduce a wide range (but not all) of analyses of census tract socioeconomic cancer disparities and have relatively low disclosure risks, which we assess by comparing individual patients' actual and synthetic tract locations. We conclude with a discussion of how synthetic data sets can be used by researchers with cancer registry data.

Objective To assess primary care physicians' (PCPs) knowledge of energy balance related guidelines and the association with sociodemographic characteristics and clinical care practices. Method As part of the 2008 U.S. nationally representative National Survey of Energy Balance Related Care among Primary Care Physicians (EB-PCP), 1776 PCPs from four specialties who treated adults ( n = 1060) or children and adolescents ( n = 716) completed surveys on sociodemographic information, knowledge of energy balance guidelines, and clinical care practices. Results EB-PCP response rate was 64.5%. For PCPs treating children, knowledge of guidelines for healthy BMI percentile, physical activity, and fruit and vegetables intake was 36.5%, 27.0%, and 62.9%, respectively. For PCPs treating adults, knowledge of guidelines for overweight, obesity, physical activity, and fruit and vegetables intake was 81.4%, 81.3%, 70.9%, and 63.5%, respectively. Generally, younger, female physicians were more likely to exhibit correct knowledge. Knowledge of weight-related guidelines was associated with assessment of body mass index (BMI) and use of BMI-for-age growth charts. Conclusion Knowledge of energy balance guidelines among PCPs treating children is low, among PCPs treating adults it appeared high for overweight and obesity-related clinical guidelines and moderate for physical activity and diet, and was mostly unrelated to clinical practices among all PCPs.

INTRODUCTION This study examined the reliability of self-reported smoking history measures. The key measures of interest were time since completely quitting smoking among former smokers; age at which fairly regular smoking was initiated among former and current smokers; the number of cigarettes smoked per day and the number of years of daily smoking among former smokers; and never smoking. Another goal was to examine sociodemographic factors and interview method as potential predictors of the odds of strict agreement in responses. METHODS Data from the 2002-2003 Tobacco Use Supplement to the Current Population Survey were examined. Descriptive analysis was performed to detect discrepant data patterns, and intraclass and Pearson correlations and kappa coefficients were used to assess reporting consistency over the 12-month interval. Multiple logistic regression models with replicate weights were built and fitted to identify factors influencing the logit of agreement for each measure of interest. RESULTS All measures revealed at least moderate levels of overall agreement. However, upon closer examination, a few measures also showed some considerable differences in absolute value. The highest percentage of these differences was observed for former smokers' reports of the number of years smoking every day. CONCLUSIONS Overall, the data suggest that self-reported smoking history characteristics are reliable. The logit of agreement over a 12-month period is shown to depend on a few sociodemographic characteristics as well as their interactions with each other and with interview method.

Background The National Institutes of Health (NIH) Health Care Systems Research Collaboratory (NIH Collaboratory) seeks to produce generalizable knowledge about the conduct of pragmatic research in health systems. This analysis applied the PRECIS-2 pragmatic trial criteria to five NIH Collaboratory pragmatic trials to better understand 1) the pragmatic aspects of the design and implementation of treatments delivered in real world settings and 2) the usability of the PRECIS-2 criteria for assessing pragmatic features across studies and across time. Methods/Design Using the PRECIS-2 criteria, five pragmatic trials were each rated by eight raters. For each trial, we reviewed the original grant application and a required progress report written at the end of a 1-year planning period that included changes to the protocol or implementation approach. We calculated median scores and interrater reliability for each PRECIS domain and for the overall trial at both time points, as well as the differences in scores between the two time points. We also reviewed the rater comments associated with the scores. Results All five trials were rated to be more pragmatic than explanatory, with comments indicating that raters generally perceived them to closely mirror routine clinical care across multiple domains. The PRECIS-2 domains for which the trials were, on average, rated as most pragmatic on the 1 to 5 scale at the conclusion of the planning period included primary analysis (mean = 4.7 (range = 4.5 to 4.9)), recruitment (4.3 (3.6 to 4.8)), eligibility (4.1 (3.4 to 4.8)), setting (4.1 (4.0 to 4.4)), follow-up (4.1 (3.4 to 4.9)), and primary outcome (4.1 (3.5 to 4.9)). On average, the less pragmatic domains were organization (3.3 (2.6 to 4.4)), flexibility of intervention delivery (3.5 (2.1-4.5)), and flexibility of intervention adherence (3.8 (2.8-4.5)). Interrater agreement was modest but statistically significant for four trials (Gwet’s AC1 statistic range 0.23 to 0.40) and the intraclass correlation coefficient ranged from 0.05 to 0.31. Rating challenges included assigning a single score for domains that may relate to both patients and care settings (that is, eligibility or recruitment) and determining to what extent aspects of complex research interventions differ from usual care. Conclusions These five trials in diverse healthcare settings were rated as highly pragmatic using the PRECIS-2 criteria. Applying the tool generated insightful discussion about real-world design decisions but also highlighted challenges using the tool. PRECIS-2 raters would benefit from additional guidance about how to rate the interwoven patient and practice-level considerations that arise in pragmatic trials. Trial registrations Clinicaltrials.gov trial registrations: NCT02019225, NCT01742065, NCT02015455, NCT02113592, NCT02063867.

The Physical Activity Monitor component was introduced into the 2003-2004 National Health and Nutrition Examination Survey (NHANES) to collect objective information on physical activity including both movement intensity counts and ambulatory steps. Because of an error in the accelerometer device initialization process, the steps data were missing for all participants in several primary sampling units, typically a single county or group of contiguous counties, who had intensity count data from their accelerometers. To avoid potential bias and loss in efficiency in estimation and inference involving the steps data, we considered methods to accurately impute the missing values for steps collected in the 2003-2004 NHANES. The objective was to come up with an efficient imputation method that minimized model-based assumptions. We adopted a multiple imputation approach based on additive regression, bootstrapping and predictive mean matching methods. This method fits alternative conditional expectation (ace) models, which use an automated procedure to estimate optimal transformations for both the predictor and response variables. This paper describes the approaches used in this imputation and evaluates the methods by comparing the distributions of the original and the imputed data. A simulation study using the observed data is also conducted as part of the model diagnostics. Finally, some real data analyses are performed to compare the before and after imputation results. Published 2016. This article is a U.S. Government work and is in the public domain in the USA.

In 2016, the cancer registry community will directly assign T, N and M components of stage. The Surveillance, Epidemiology, and End Results program implemented a field study to determine how often T, N and M were not available in the medical record, requiring the registrar to directly assign clinical or pathologic TNM stage components. The field study also identified specific training needs.T, N and M status were collected from multiple sources within medical records for a total of 280 cases, 56 each from breast, prostate, colon, lung, and ovarian cancer. TNM data elements were also directly assigned by a series of reviewers and by study participants using the medical records with TNM information redacted. Availability of physician-assigned TNM was estimated from the medical record. Also, participant responses were compared to preferred answers.Pathologic T, N and M were available more often in the medical records than were clinical values and varied by site. Pathologic T and N were available for about two-thirds of the cases, but the clinical elements were available for only about 20% of cases. The agreement between participant responses and review panel assignments varied by data element and cancer site. Agreement was modest for most data elements and cancer sites, ranging from 54% for clinical T to 92% for clinical M for all cancer sites combined.The data elements for TNM staging and stage group were often missing from the medical records, so registrars in the field will need to assign TNM frequently. Furthermore, the results of this study strongly suggest that more training is required, even among those who currently assign TNM.

Objectives. We examined patterns of cervical and breast cancer screening among Asian American women in California and assessed their screening trends over time. Methods. We pooled weighted data from 5 cycles of the California Health Interview Survey (2001, 2003, 2005, 2007, 2009) to examine breast and cervical cancer screening trends and predictors among 6 Asian nationalities. We calculated descriptive statistics, bivariate associations, multivariate logistic regressions, predictive margins, and 95% confidence intervals. Results. Multivariate analyses indicated that Papanicolaou test rates did not significantly change over time (77.9% in 2001 vs 81.2% in 2007), but mammography receipt increased among Asian American women overall (75.6% in 2001 vs 81.8% in 2009). Length of time in the United States was associated with increased breast and cervical cancer screening among all nationalities. Sociodemographic and health care access factors had varied effects, with education and insurance coverage significantly predicting screening for certain groups. Overall, we observed striking variation by nationality. Conclusions. Our results underscore the need for intervention and policy efforts that are targeted to specific Asian nationalities, recent immigrants, and individuals without health care access to increase screening rates among Asian women in California.

Importance The emergency department (ED) is used to manage cancer-related complications among the 15.5 million people living with cancer in the United States. However, ED utilization patterns by the population of US adults with cancer have not been previously evaluated or described in published literature. Objective To estimate the proportion of US ED visits made by adults with a cancer diagnosis, understand the clinical presentation of adult patients with cancer in the ED, and examine factors related to inpatient admission within this population. Design, Setting, and Participants Nationally representative data comprised of 7 survey cycles (January 2006-December 2012) from the Nationwide Emergency Department Sample were analyzed. Identification of adult (age ≥18 years) cancer-related visits was based on Clinical Classifications Software diagnoses documented during the ED visit. Weighted frequencies and proportions of ED visits among adult patients with cancer by demographic, geographic, and clinical characteristics were calculated. Weighted multivariable logistic regression was used to examine the associations between inpatient admission and key demographic and clinical variables for adult cancer–related ED visits. Main Outcomes and Measures Adult cancer–related ED utilization patterns; identification of primary reason for ED visit; patient-related factors associated with inpatient admission from the ED. Results Among an estimated 696 million weighted adult ED visits from January 2006 to December 2012, 29.5 million (4.2%) were made by a patient with a cancer diagnosis. The most common cancers associated with an ED visit were breast, prostate, and lung cancer, and most common primary reasons for visit were pneumonia (4.5%), nonspecific chest pain (3.7%), and urinary tract infection (3.2%). Adult cancer–related ED visits resulted in inpatient admissions more frequently (59.7%) than non–cancer-related visits (16.3%) ( P Conclusions and Relevance Consistent with national prevalence statistics among adults, breast, prostate, and lung cancer were the most common cancer diagnoses presenting to the ED. Pneumonia was the most common reason for adult cancer–related ED visits with an associated high inpatient admission rate. This analysis highlights cancer-specific ED clinical presentations and the opportunity to inform patient and system-directed prevention and management strategies.

Summary Background Overweight and obesity are major health risks in the United States (US) and primary care physicians (PCPs) are uniquely positioned to address them. However, their personal beliefs about weight-related care may influence their delivery of care. Methods A nationally representative sample of 2022 physicians completed the National Survey of Energy Balance-Related Care among Primary Care Physicians. Physicians responded to questions regarding their beliefs and clinical practices associated with weight control including assessment, counselling, referral and follow-up for diet, physical activity, and weight. Multivariate logistic regression was used to examine associations between physician characteristics and personal beliefs, and associations between personal beliefs and care delivery, adjusting for specialty, age, gender, race, region, urban/rural location, and patient population. Results Most physicians feel a responsibility (97%) to promote weight-related care, but over half (53%) have concerns about their effectiveness and almost two-thirds feel they lack effective strategies to help patients (63%). Demographics and medical specialty were associated with beliefs (female, Asian-American, Midwest and Southern location, and internal medicine physicians were more likely to have stronger positive beliefs about weight-related care). Personal beliefs about weight-related care were associated with the likelihood of its delivery. However, two practices, regular BMI assessment and referring patients for further evaluation and management, were less related to PCP beliefs than were other care practices. Conclusions PCPs’ beliefs may be important to their practice of weight-related care. Training in behavioural counselling, and providing physician's tools and resources may help to address their concerns about helping patients with weight-related care.

Purpose. Primary care physicians (PCPs) may not adequately counsel or monitor patients regarding diet, physical activity, and weight control (i.e., provide energy balance care). We assessed the organization of PCPs' practices for providing this care. Design. The study design was a nationally representative survey conducted in 2008. Setting. The study setting was U.S. primary care practices. Subjects. A total of 1740 PCPs completed two sequential questionnaires (response rate, 55.5%). Measures. The study measured PCPs' reports of practice resources, and the frequency of body mass index assessment, counseling, referral for further evaluation/management, and monitoring of patients for energy balance care. Analysis. Descriptive statistics and logistic regression modeling were used. Results. More than 80% of PCPs reported having information resources on diet, physical activity, or weight control available in waiting/exam rooms, but fewer billed (45%), used reminder systems (< 30%), or received incentive payments (3%) for energy balance care. A total of 26% reported regularly assessing body mass index and always/often providing counseling as well as tracking patients for progress related to energy balance. In multivariate analyses, PCPs in practices with full electronic health records or those that bill for energy balance care provided this care more often and more comprehensively. There were strong specialty differences, with pediatricians more likely (odds ratio, 1.78; 95% confidence interval, 1.26–2.51) and obstetrician/gynecologists less likely (odds ratio, 0.28; 95% confidence interval, 0.17–0.44) than others to provide energy balance care. Conclusion. PCPs' practices are not well organized for providing energy balance care. Further research is needed to understand PCP care-related specialty differences.