Comparison of Imputation Strategies for Incomplete Longitudinal Data in Life-Course Epidemiology

Incomplete longitudinal data are common in life-course epidemiology and may induce bias leading to incorrect inference. Multiple imputation (MI) is increasingly preferred for handling missing data, but few studies explore MI-method performance and feasibility in real-data settings. We compared 3 MI methods using real data under 9 missing-data scenarios, representing combinations of 10%, 20%, and 30% missingness and missing completely at random, at random, and not at random. Using data from Health and Retirement Study (HRS) participants, we introduced record-level missingness to a sample of participants with complete data on depressive symptoms (1998-2008), mortality (2008-2018), and relevant covariates. We then imputed missing data using 3 MI methods (normal linear regression, predictive mean matching, variable-tailored specification), and fitted Cox proportional hazards models to estimate effects of 4 operationalizations of longitudinal depressive symptoms on mortality. We compared bias in hazard ratios, root mean square error, and computation time for each method. Bias was similar across MI methods, and results were consistent across operationalizations of the longitudinal exposure variable. However, our results suggest that predictive mean matching may be an appealing strategy for imputing life-course exposure data, given consistently low root mean square error, competitive computation times, and few implementation challenges.

Patient-reported outcomes before treatment for localized prostate cancer: are there differences among countries? Data from the True North Global Registry

INTRODUCTION

Similar Patient-Reported Outcomes (PROs) at diagnosis for localized prostate cancer among countries may indicate that different treatments are recommended to the same profile of patients, regardless the context characteristics (health systems, medical schools, culture, preferences…). The aim of this study was to assess such comparison.

METHODS

We analyzed the EPIC-26 results before the primary treatment of men diagnosed of localized prostate cancer from January 2017 onwards (revised data available up to September 2019), from a multicenter prospective international cohort including seven regions: Australia/New Zealand, Canada, Central Europe (Austria / Czech Republic / Germany), United Kingdom, Italy, Spain, and the United States. The EPIC-26 domain scores and pattern of three selected items were compared across regions (with Central Europe as reference). All comparisons were made stratifying by treatment: radical prostatectomy, external radiotherapy, brachytherapy, and active surveillance.

RESULTS

The sample included a total of 13,483 men with clinically localized or locally advanced prostate cancer. PROs showed different domain patterns before treatment across countries. The sexual domain was the most impaired, and the one with the highest dispersion within countries and with the greatest medians' differences across countries. The urinary incontinence domain, together with the bowel and hormonal domains, presented the highest scores (better outcomes) for all treatment groups, and homogeneity across regions.

CONCLUSIONS

Patients with localized or locally advanced prostate cancer undergoing radical prostatectomy, EBRT, brachytherapy, or active surveillance presented mainly negligible or small differences in the EPIC-26 domains before treatment across countries. The results on urinary incontinence or bowel domains, in which almost all patients presented the best possible score, may downplay the baseline data role for evaluating treatments' effects. However, the heterogeneity within countries and the magnitude of the differences found across countries in other domains, especially sexual, support the need of implementing the PRO measurement from diagnosis.

Randomized Comparative Effectiveness Trial of 2 Federally Recommended Strategies to Reduce Excess Body Fat in Overweight, Low-Income Patients: MyPlate.gov vs Calorie Counting

PURPOSE

Since 2011, US authorities have supported the following 2 approaches to healthier body fat composition: the Centers for Disease Control and Prevention National Diabetes Prevention Program's calorie counting (CC) approach and the US Department of Agriculture's MyPlate (adherence to federal nutrition guidelines). The purpose of this study was to compare the effect of CC vs MyPlate approaches on satiety/satiation and on achieving healthier body fat composition among primary care patients.

METHODS

We conducted a randomized controlled trial comparing the CC and MyPlate approaches from 2015 to 2017. The adult participants were overweight, of low income, and were mostly Latine (n = 261). For both approaches, community health workers conducted 2 home education visits, 2 group education sessions, and 7 telephone coaching calls over a period of 6 months. Satiation and satiety were the primary patient-centered outcome measures. Waist circumference and body weight were the primary anthropometric measures. Measures were assessed at baseline, 6 months, and 12 months.

RESULTS

Satiation and satiety scores increased for both groups. Waist circumference was significantly decreased in both groups. MyPlate, but not CC, resulted in lower systolic blood pressure at 6 months but not at 12 months. Participants for both MyPlate and CC reported greater quality of life and emotional well-being and high satisfaction with their assigned weight-loss program. The most acculturated participants showed the greatest decreases in waist circumference.

CONCLUSIONS

A MyPlate-based intervention might be a practical alternative to the more traditional CC approach to promoting satiety and facilitating decreases in central adiposity among low-income, mostly Latine primary care patients.

Improving Care Linkage for Racial-Ethnic Minority Youths Receiving Emergency Department Treatment for Suicidality: SAFETY-A

OBJECTIVE

This study evaluated the effects of Safe Alternatives for Teens and Youths-Acute (SAFETY-A), a brief strengths-based, cognitive-behavioral family intervention, on racial-ethnic minority youths receiving emergency department (ED) treatment for suicidal episodes.

METHODS

Participants were 105 racial-ethnic minority youths enrolled in a randomized controlled trial evaluating SAFETY-A versus enhanced usual care for youths receiving ED treatment for suicidal episodes. Analyses examined group effects on care linkage after discharge and adequate treatment dose. A sample of 55 White youths was included for comparison.

RESULTS

Racial-ethnic minority youths who received SAFETY-A had higher treatment linkage rates than those receiving usual care. Adequate treatment dose rates did not differ by group.

CONCLUSIONS

Racial-ethnic minority youths receiving SAFETY-A had higher treatment linkage rates after discharge than those receiving usual care. SAFETY-A is a promising approach to enhance care continuity and mental health equity for racial-ethnic minority youths at risk for suicide.

Bayesian finite-population inference with spatially correlated measurements

Community-based public health interventions often rely on representative, spatially referenced outcome data to draw conclusions about a finite population. To estimate finite-population parameters, we are posed with two challenges: to correctly account for spatial association among the sampled and nonsampled participants and to correctly model missingness in key covariates, which may be also spatially associated. To accomplish this, we take inspiration from the preferential sampling literature and develop a general Bayesian framework that can specifically account for preferential non-response. This framework is first applied to three missing data scenarios in a simulation study. It is then used to account for missing data patterns seen in reported annual household income in a corner-store intervention project. Through this, we are able to construct finite-population estimates of the percent of income spent on fruits and vegetables. Such a framework provides a flexible way to account for spatial association and complex missing data structures in finite populations.

Assessing Alternative Imputation Strategies for Infrequently Missing Items on Multi-item Scales

Health-science researchers often measure psychological constructs using multi-item scales and encounter missing items on some participants. Multiple imputation (MI) has emerged as an alternative to ad-hoc methods (e.g., mean substitution) for handling incomplete data on multi-item scales, appealingly reflecting available information while accounting for uncertainty due to missing values in a unified inferential framework. However, MI can be implemented in a variety of ways. When the number of variables to impute gets large, some strategies yield unstable estimates of quantities of interest while others are not technically feasible to implement. These considerations raise pragmatic questions about the extent to which ad-hoc procedures would yield statistical properties that are competitive with theoretically motivated methods. Drawing on an HIV study where depression and anxiety symptoms are measured with multi-item scales, this empirical investigation contrasts ad-hoc methods for handling missing items with various MI implementations that differ as to whether imputation is at the item-level or scale-level and how auxiliary variables are incorporated. While the findings are consistent with previous reports favoring item-level imputation when feasible to implement, we found only subtle differences in statistical properties across procedures, suggesting that weaknesses of ad-hoc procedures may be muted when missing data percentages are modest.

Male-Female Disparities in Years of Potential Life Lost Attributable to COVID-19 in the United States: A State-by-State Analysis

Males are at higher risk relative to females of severe outcomes following COVID-19 infection. Focusing on COVID-19-attributable mortality in the United States (U.S.), we quantify and contrast years of potential life lost (YPLL) attributable to COVID-19 by sex based on data from the U.S. National Center for Health Statistics as of 31 March 2021, specifically by contrasting male and female percentages of total YPLL with their respective percent population shares and calculating age-adjusted male-to-female YPLL rate ratios both nationally and for each of the 50 states and the District of Columbia. Using YPLL before age 75 to anchor comparisons between males and females and a novel Monte Carlo simulation procedure to perform estimation and uncertainty quantification, our results reveal a near-universal pattern across states of higher COVID-19-attributable YPLL among males compared to females. Furthermore, the disproportionately high COVID-19 mortality burden among males is generally more pronounced when measuring mortality in terms of YPLL compared to age-irrespective death counts, reflecting dual phenomena of males dying from COVID-19 at higher rates and at systematically younger ages relative to females. The U.S. COVID-19 epidemic also offers lessons underscoring the importance of a public health environment that recognizes sex-specific needs as well as different patterns in risk factors, health behaviors, and responses to interventions between men and women. Public health strategies incorporating focused efforts to increase COVID-19 vaccinations among men are particularly urged.

Comparative effectiveness trial comparing MyPlate to calorie counting for mostly low-income Latino primary care patients of a federally qualified community health center: study design, baseline characteristics

BACKGROUND

Primary care-based behavior change obesity treatment has long featured the Calorie restriction (CC), portion control approach. By contrast, the MyPlate-based obesity treatment approach encourages eating more high-satiety/high-satiation foods and requires no calorie-counting. This report describes study methods of a comparative effectiveness trial of CC versus MyPlate. It also describes baseline findings involving demographic characteristics and their associations with primary outcome measures and covariates, including satiety/satiation, dietary quality and acculturation.

METHODS

A comparative effectiveness trial was designed to compare the CC approach (n = 130) versus a MyPlate-based approach (n = 131) to treating patient overweight. Intervenors were trained community health workers. The 11 intervention sessions included two in-home health education sessions, two group education sessions, and seven telephone coaching sessions. Questionnaire and anthropometric assessments occurred at baseline, 6- and 12 months; food frequency questionnaires were administered at baseline and 12 months. Participants were overweight adult primary care patients of a federally qualified health center in Long Beach, California. Two measures of satiety/satiation and one measure of post-meal hunger comprised the primary outcome measures. Secondary outcomes included weight, waist circumference, blood pressure, dietary quality, sugary beverage intake, water intake, fruit and vegetable fiber intake, mental health and health-related quality of life. Covariates included age, gender, nativity status (U.S.-born, not U.S.-born), race/ethnicity, education, and acculturation.

ANALYSIS

Baseline characteristics were compared using chi square tests. Associations between covariates and outcome measures were evaluated using multiple regression and logistic regression.

RESULTS

Two thousand eighty-six adult patients were screened, yielding 261 enrollees who were 86% Latino, 8% African American, 4% White and 2% Other. Women predominated (95%). Mean age was 42 years. Most (82%) were foreign-born; 74% chose the Spanish language option. Mean BMI was 33.3 kg/m²; mean weight was 82 kg; mean waist circumference was 102 cm. Mean blood pressure was 122/77 mm. Study arms on key baseline measures did not differ except on dietary quality and sugary beverage intake. Nativity status was significantly associated with dietary quality.

CONCLUSIONS

The two treatment arms were well-balanced demographically at baseline. Nativity status is inversely related to dietary quality.

TRIAL REGISTRATION

NCT02514889 , posted on 8/4/2015.

Benefits of Treatment Completion Over Premature Termination: Findings from the National Child Traumatic Stress Network

Objective: To evaluate potential differences in therapeutic outcomes between youths who completed a full course of treatment as planned compared to youths who terminated treatment prematurely. Method: Using longitudinal data from the National Child Traumatic Stress Network (NCTSN) Core Data Set, the present study examined demographic characteristics, trauma history, scores on standardized measures, and ratings of functional impairment and behavior problems in a large clinical sample of children and adolescents exposed to trauma who received treatment at NCTSN centers across the United States. Baseline and follow-up data were used to compare treatment completers (n= 3,108) and noncompleters (n = 4,029). Results: Both treatment completers and noncompleters received benefits from treatment by NCTSN mental health providers in that both groups showed significant decreases in mean scores from baseline to follow-up on all standardized measures. However, compared to noncompleters, treatment completers showed three types of significantly greater benefit at follow-up. These included: (a) greater rates of decline (i.e., steeper slopes) on all outcome measures; (b) greater reductions in the odds of falling within the clinical range on standardized measures; and (c) greater reductions in the odds of exhibiting functional impairment and behavior problems at follow-up. In contrast, compared to treatment completers, noncompleters reported significantly higher rates of lifetime exposure to community violence, psychological maltreatment, physical abuse, neglect, sexual abuse, and sexual assault. Conclusion: These findings underscore the value of incorporating engagement and retention strategies in treatments for traumatized youths to maximize therapeutic benefit and raise the standard of care.

Maintaining Internal Validity in Community Partnered Participatory Research: Experience from the Community Partners in Care Study

Objective

With internal validity being a central goal of designed experiments, we seek to elucidate how community partnered participatory research (CPPR) impacts the internal validity of public health comparative-effectiveness research.

Methods

Community Partners in Care (CPIC), a study comparing a community-coalition intervention to direct technical assistance for disseminating depression care to vulnerable populations, is used to illustrate design choices developed with attention to core CPPR principles. The study-design process is reviewed retrospectively and evaluated based on the resulting covariate balance across intervention arms and on broader peer-review assessments. Contributions of the CPIC Council and the study's design committee are highlighted.

Results

CPPR principles contributed to building consensus around the use of randomization, creating a sampling frame, specifying geographic boundaries delimiting the scope of the investigation, grouping similar programs into pairs or other small blocks of units, collaboratively choosing random-number-generator seeds to determine randomized intervention assignments, and addressing logistical constraints in field operations. Study protocols yielded samples that were well-balanced on background characteristics across intervention arms. CPIC has been recognized for scientific merit, has drawn attention from policymakers, and has fueled ongoing research collaborations.

Conclusions

Creative and collaborative fulfillment of CPPR principles reinforced the internal validity of CPIC, strengthening the study's scientific rigor by engaging complementary areas of knowledge and expertise among members of the investigative team.

Methamphetamine use and oral health-related quality of life

PURPOSE

Methamphetamine (MA) is associated with adverse health effects, including the rampant tooth decay condition called "Meth Mouth." However, the impact of MA use on oral health-related quality of life (OHRQOL) is unknown. This study assessed the relationship between MA use and self-reported OHRQOL.

METHODS

This cross-sectional study uses information from 545 MA-using participants recruited from Los Angeles County, California. Dental examinations were performed by three calibrated dentists using National Health and Nutrition Examination Survey (NHANES) protocols. Data on socio-demographic, behavioral, and drug-use history were recorded using questionnaires. Participants were categorized as 'light' or 'moderate/heavy' users based on reported frequency of MA use in the past 30 days. Route of MA administration was categorized as 'smoking' or 'other.' Self-reported OHRQOL was based on the Oral Health Impact Profile scale.

RESULTS

Majority of the participants were male (80.9%). Median age was 45.0 years (IQR-13.0). Median number of days of MA use was 10.0 (IQR-12.0). Smoking was the preferred route of MA use (70.2%). Root caries in ≥ 3 teeth were reported in 78% of MA users. More than half of the participants reported having painful aching in mouth, avoidance of particular food items, feeling embarrassed, and discomfort while eating in the last 12 months. In unadjusted logistic models, moderate/heavy MA users were more likely to report an affected sense of taste [OR = 1.58, 95% CI (1.10-2.27)] and avoidance of particular foods [OR = 1.45, 95% CI (1.02-2.01)] than light users. Among individuals preferring other MA administration routes, moderate/heavy MA users were 3.09 times as likely to report an affected sense of taste than light users [OR = 3.09, 95% CI (1.52-6.27)].

CONCLUSION

Oral health and OHRQOL appear to be worse among Methamphetamine users than in the US general population.

Periodontal status of current methamphetamine users

BACKGROUND

Methamphetamine (MA) use is associated with extensive dental caries. The purpose of this study was to determine the prevalence and severity of periodontal disease in a convenience sample of MA users.

METHODS

In this cross-sectional survey, MA users were recruited with a combination of snowball sampling and street outreach techniques. Three dentists, trained and calibrated to the oral assessments used in the National Health and Nutrition Survey, measured and recorded the participants' attachment loss, probing depth, and gingival recession. Concomitant interviews elicited psychological, substance use, medication, and dietary habits associated with MA use.

RESULTS

Periodontal assessments were completed on 546 adults. More than 69% were cigarette smokers, and more than 55% were medium to high MA users. Classifying prevalence by means of the Centers for Disease Control and Prevention and the American Academy of Periodontology definitions, cigarette smokers and medium to high MA users had a high prevalence of periodontal disease. The defining features of the participants were being 30 years and older (average, 42.2 years) and having severe and moderate periodontitis.

CONCLUSION

This is the first study to the authors' knowledge to systematically examine periodontal disease in a large population of current MA users. MA users in a Los Angeles urban setting had a high prevalence and severity of destructive periodontal disease. The frequency of MA use had a minimal impact on the severity of periodontal disease.

PRACTICAL IMPLICATIONS

An MA user can be at high risk of developing periodontal disease. Knowing that behavioral factors, such as smoking and consuming sugary beverages, are more important than MA use will assist the clinician in managing the treatment of MA users.

A Comparison of Methamphetamine Users to a Matched NHANES Cohort: Propensity Score Analyses for Oral Health Care and Dental Service Need

Dental problems are among the most frequently reported health issues of drug users. This study describes, among the largest population of methamphetamine (MA) users to date (N = 459, including both HIV-negative and HIV-positive participants) oral hygiene practice, dental care access, and dental quality of life. A matched control group from the Third National Health and Nutrition Examination Survey was utilized. Findings conclusively establish that MA users have severe oral health deficits compared to the general population: they are 3.5 times more likely to experience painful toothaches, 6.6 times to experience difficulty eating, and 8.6 times to be self-conscious due to dental appearance. HIV-positive users were more likely to have regular dental visits than HIV-negative users. Severity of use (both high-frequency use as well as injection as the method) was associated with poorer oral health care. Despite the magnitude of the need, few MA users receive the needed care.

A Community-Partnered, Participatory, Cluster-Randomized Study of Depression Care Quality Improvement: Three-Year Outcomes

OBJECTIVE

Community Partners in Care, a community-partnered, cluster-randomized trial with depressed clients from 93 Los Angeles health and community programs, examined the added value of a community coalition approach (Community Engagement and Planning [CEP]) versus individual program technical assistance (Resources for Services [RS]) for implementing depression quality improvement in underserved communities. CEP was more effective than RS in improving mental health-related quality of life, reducing behavioral health hospitalizations, and shifting services toward community-based programs at six months. At 12 months, continued evidence of improvement was found. This study examined three-year outcomes.

METHODS

Among 1,004 participants with depression who were eligible for three-year follow-up, 600 participants from 89 programs completed surveys. Multiple regression analyses estimated intervention effects on poor mental health-related quality of life and depression, physical health-related quality of life, behavioral health hospital nights, and use of services.

RESULTS

At three years, no differences were found in the effects of CEP versus RS on depression or mental health-related quality of life, but CEP had modest effects in improving physical health-related quality of life and reducing behavioral health hospital nights, and CEP participants had more social- and community-sector depression visits and greater use of mood stabilizers. Sensitivity analyses with longitudinal modeling reproduced these findings but found no significant differences between groups in change from baseline to three years.

CONCLUSIONS

At three years, CEP and RS did not have differential effects on primary mental health outcomes, but CEP participants had modest improvements in physical health and fewer behavioral health hospital nights.

Comparative Effectiveness of Two Models of Depression Services Quality Improvement in Health and Community Sectors

OBJECTIVE

The effectiveness of community coalition building and program technical assistance was compared in implementation of collaborative care for depression among health care and community sector clients.

METHODS

In under-resourced communities, within 93 programs randomly assigned to coalition building (Community Engagement and Planning) or program technical assistance (Resources for Services) models, 1,018 clients completed surveys at baseline and at six, 12, or 36 months. Regression analysis was used to estimate intervention effects and intervention-by-sector interaction effects on depression, mental health-related quality of life, and community-prioritized outcomes and on services use.

RESULTS

For outcomes, there were few significant intervention-by-sector interactions, and stratified findings suggested benefits of coalition building in both sectors. For services use, at 36 months, increases were found for coalition building in primary care visits, self-help visits, and appropriate treatment for community clients and in community-based services use for health care clients.

CONCLUSIONS

Relative to program technical assistance, community coalition building benefited clients across sectors and shifted long-term utilization across sectors.

Long-Term Effects of Pre-Placement Risk Factors on Children's Psychological Symptoms and Parenting Stress Among Families Adopting Children From Foster Care

This exploratory longitudinal study examined behavioral outcomes and parenting stress among families with children adopted from foster care, taking into account environmental and biological risk factors. Child internalizing and externalizing problems and parenting stress were assessed in 82 adopted children and their families at 2 months post-placement, 12 months post-placement, and then yearly until 5 years post-placement. A history of abuse/neglect predicted significantly higher externalizing and internalizing problems at a borderline level of statistical significance. In the initial stages after placement, externalizing problems were significantly higher among children who were 4 years or older at placement versus those who were younger than 4, although differences were no longer significant 5 years post-placement. Statistical trends in parenting stress reflected reduced stress in the first 12 months followed by a plateau for parents who adopted older children and greater stress for parents who adopted younger children. Familiar limitations for observational cohort data apply. Nonetheless, the availability of longitudinal follow-up on a sizable sample of children adopted from foster care adds insight to the psychological dynamics for adoptive families and suggests that families of children adopted from the foster care system may have unique needs for ongoing support around behavioral issues.

Mechanisms underlying methamphetamine-related dental disease

BACKGROUND

The authors clarified the causal mechanisms underlying the high prevalence of dental disease encountered in people who habitually use methamphetamine (meth).

METHODS

Using a stratified sampling approach, the authors conducted comprehensive oral examinations and psychosocial assessments for 571 study participants who used meth. Three calibrated dentists, who used National Health and Nutrition Examination Survey (NHANES) protocols, characterized the study participants' dental disease. The authors also collected data related to study participants' history of meth use and other attributes linked to dental disease.

RESULTS

Study participants who used meth manifested higher rates of xerostomia and caries experience compared with NHANES control participants. Participants who used meth had a higher level of daily consumption of sugary beverages compared with NHANES control participants. Smoking meth did not increase caries experience over other modes of intake. Dental hygiene was a significant determinant of dental health outcomes.

CONCLUSIONS

Mode of intake and frequency of meth use have a minimal impact on dental health outcomes. Behaviors, such as sugary beverage consumption and poor oral hygiene, better explain dental health outcomes.

PRACTICAL IMPLICATIONS

Having a better understanding of the causal mechanisms of "meth mouth" sets the stage for clinicians to provide more personalized interventions and management of dental disease in people who use meth.

Comparing Lay Community and Academic Survey Center Interviewers in Conducting Household Interviews in Latino Communities

BACKGROUND

The employment of professional interviewers from academic survey centers to conduct surveys has been standard practice. Because one goal of community-engaged research is to provide professional skills to community residents, this paper considers whether employing locally trained lay interviewers from within the community may be as effective as employing interviewers from an academic survey center with regard to unit and item nonresponse rates and cost.

METHODS

To study a nutrition-focused intervention, 1035 in-person household interviews were conducted in East Los Angeles and Boyle Heights, 503 of which were completed by lay community interviewers. A chi-square test was used to assess differences in unit nonresponse rates between professional and community interviewers and Welch's t tests were used to assess differences in item nonresponse rates. A cost comparison analysis between the two interviewer groups was also conducted.

RESULTS

Interviewers from the academic survey center had lower unit nonresponse rates than the lay community interviewers (16.2% vs. 23.3%; p < 0.01). However, the item nonresponse rates were lower for the community interviewers than the professional interviewers (1.4% vs. 3.3%; p < 0.01). Community interviewers cost approximately $415.38 per survey whereas professional interviewers cost approximately $537.29 per survey.

CONCLUSIONS

With a lower cost per completed survey and lower item nonresponse rates, lay community interviewers are a viable alternative to professional interviewers for fieldwork in community-based research. Additional research is needed to assess other important aspects of data quality interviewer such as interviewer effects and response error.

Multiple imputation with non-additively related variables: Joint-modeling and approximations

This paper investigates multiple imputation methods for regression models with interacting continuous and binary predictors when continuous variable may be missing. Usual implementations for parametric multiple imputation assume a multivariate normal structure for the variables, which is not satisfied for a binary variable nor its interaction with a continuous variable. To accommodate interactions, missing covariates are multiply imputed from conditional distribution in a manner consistent with the joint model. Alternative imputation methods under multivariate normal assumptions are also considered as candidate approximations and evaluated in a simulation study. The results suggest that the joint modeling procedure performs generally well across a wide range of scenarios and so do the approximation methods that incorporate interactions in the model appropriately by stratification. It is critical to include interactions in the imputation model as failure to do so may result in low coverage and bias. We apply the joint modeling approach and approximation methods in the study of childhood trauma with gender × trauma interaction.

Methamphetamine Users Have Increased Dental Disease: A Propensity Score Analysis

Methamphetamine (MA) users are assumed to have a high burden of tooth decay. Less clear is how the distribution and severity of dental caries in MA users differ from the general population. Using a covariate-balancing propensity score strategy, we investigated the differential effects of MA use on dental caries by comparing the patterns of decayed, missing, and filled teeth in a community sample of 571 MA users with a subset of 2,755 demographically similar control individuals selected from a National Health and Nutrition Examination Survey (NHANES) cohort. Recruited over a 2-y period with a stratified sampling protocol, the MA users underwent comprehensive dental examinations by 3 trained and calibrated dentists using NHANES protocols. Propensity scores were estimated with logistic regression based on background characteristics, and a subset of closely matched subjects was stratified into quintiles for comparisons. MA users were twice as likely to have untreated caries (odds ratio [OR] = 2.08; 95% confidence interval [95% CI]: 1.55 to 2.78) and 4 times more likely to have caries experience (OR = 4.06; 95% CI: 2.24 to 7.34) than the control group of NHANES participants. Additionally, MA users were twice as likely to have 2 more decayed, missing, or filled teeth (OR = 2.08; 95% CI: 1.29 to 2.79) than the NHANES participants. The differential involvement of the teeth surfaces in MA users was quite distinctive, with carious surface involvement being highest for the maxillary central incisors, followed by maxillary posterior premolars and molars. Users injecting MA had significantly higher rates of tooth decay compared with noninjectors (P = 0.04). Although MA users experienced decayed and missing dental surfaces more frequently than NHANES participants, NHANES participants had more restored surfaces, especially on molars. The high rates and distinctive patterns of dental caries observed could be used 1) to alert dentists to covert MA use in their patients and 2) as the basis for comprehensive management strategies.

The association between self-rated eating habits and dietary behavior in two Latino neighborhoods: Findings from Proyecto MercadoFRESCO

OBJECTIVE

Latinos are the largest racial and ethnic minority group in the United States and bear a disproportionate burden of obesity related chronic disease. Despite national efforts to improve dietary habits and prevent obesity among Latinos, obesity rates remain high. The objective of this study is to explore the relationship between self-rated dietary quality and dietary behavior among Latinos and how this may vary by socio-demographics to help inform future public health efforts aiming to improve eating habits and obesity rates.

DESIGN

Cross-sectional study using a series of chi-square tests, the non-parametric Wilcoxon-Mann-Whitney test and logistic regression to explore self-rated eating habits.

SETTING

Two urban, low-income, predominantly Latino neighborhoods in Los Angeles County.

SUBJECTS

1000 adults who self-identified as their household's primary food purchaser and preparer were interviewed from 2012 to 2013. Households were randomly selected based on their proximity to corner stores participating in a project to improve the food environment.

RESULTS

Most respondents (59%) report "good" eating habits. Significant associations between "good" eating habits and overall health, fruit and vegetable consumption were observed (p < 0.001). Despite these promising findings, we also find high levels of regular soda and energy-dense food consumption.

CONCLUSION

This study revealed a general understanding that healthy dietary habits are associated with fruit and vegetable consumption among Latinos in two urban neighborhoods. However, there is a need for more targeted health promotion and nutrition education efforts on the risks associated with soda and energy-dense food consumption to help improve dietary habits and obesity levels in low-income Latino communities.

Substantial improvements not seen in health behaviors following corner store conversions in two Latino food swamps

Background

The effectiveness of food retail interventions is largely undetermined, yet substantial investments have been made to improve access to healthy foods in food deserts and swamps via grocery and corner store interventions. This study evaluated the effects of corner store conversions in East Los Angeles and Boyle Heights, California on perceived accessibility of healthy foods, perceptions of corner stores, store patronage, food purchasing, and eating behaviors.

Methods

Household data (n = 1686) were collected at baseline and 12- to 24-months post-intervention among residents surrounding eight stores, three of which implemented a multi-faceted intervention and five of which were comparisons. Bivariate analyses and logistic and linear regressions were employed to assess differences in time, treatment, and the interaction between time and treatment to determine the effectiveness of this intervention.

Results

Improvements were found in perceived healthy food accessibility and perceptions of corner stores. No changes were found, however, in store patronage, purchasing, or consumption of fruits and vegetables.

Conclusions

Results suggest limited effectiveness of food retail interventions on improving health behaviors. Future research should focus on other strategies to reduce community-level obesity.

Electronic supplementary material

The online version of this article (doi:10.1186/s12889-016-3074-1) contains supplementary material, which is available to authorized users.

Dental disease patterns in methamphetamine users: Findings in a large urban sample

BACKGROUND

The authors used a large community sample of methamphetamine (MA) users to verify the patterns and severity of dental disease and establish a hierarchy of caries susceptibility by tooth type and tooth surface.

METHODS

Using a stratified sampling approach, 571 MA users received comprehensive oral examinations and psychosocial assessments. Three calibrated dentists characterized dental and periodontal disease by using National Health and Nutrition Examination Survey protocols. The authors also collected data on substance use history and other attributes linked to dental disease.

RESULTS

On all dental outcome measures, MA users evidenced high dental and periodontal disease, with older (≥ 30 years) and moderate or heavy MA users disproportionately affected. Women had higher rates of tooth loss and caries, as well as a greater prevalence of anterior caries. Current cigarette smokers were more likely to manifest 5 or more anterior surfaces with untreated caries and 3 or more teeth with root caries. Nearly 3% were edentulous, and a significant percentage (40%) indicated embarrassment with their dental appearance.

CONCLUSIONS

MA users have high rates of dental and periodontal disease and manifest a dose-response relationship, with greater levels of MA use associated with higher rates of dental disease. Women and current cigarette smokers are affected disproportionately. The intraoral patterns and hierarchy of caries susceptibility in MA users are distinctive.

PRACTICAL IMPLICATIONS

The prevalence and patterns of dental and periodontal disease could be used to alert dentists to possible covert MA use and to plan treatment. Concerns about dental appearance have potential as triggers for behavioral interventions.

Sampling and Recruiting Community-Based Programs Using Community-Partnered Participation Research

The inclusion of community partners in participatory leadership roles around statistical design issues like sampling and randomization has raised concerns about scientific integrity. This article presents a case study of a community-partnered, participatory research (CPPR) cluster-randomized, comparative effectiveness trial to examine implications for study validity and community relevance. Using study administrative data, we describe a CPPR-based design and implementation process for agency/program sampling, recruitment, and randomization for depression interventions. We calculated participation rates and used cross-tabulation to examine balance by intervention status on service sector, location, and program size and assessed differences in potential populations served. We achieved 51.5% agency and 89.6% program participation rates. Programs in different intervention arms were not significantly different on service sector, location, or program size. Participating programs were not significantly different from eligible, nonparticipating programs on community characteristics. We reject claims that including community members in research design decisions compromises scientific integrity. This case study suggests that a CPPR process can improve implementation of a community-grounded, rigorous randomized comparative effectiveness trial.

Evaluating model-based imputation methods for missing covariates in regression models with interactions

Imputation strategies are widely used in settings that involve inference with incomplete data. However, implementation of a particular approach always rests on assumptions, and subtle distinctions between methods can have an impact on subsequent analyses. In this research article, we are concerned with regression models in which the true underlying relationship includes interaction terms. We focus in particular on a linear model with one fully observed continuous predictor, a second partially observed continuous predictor, and their interaction. We derive the conditional distribution of the missing covariate and interaction term given the observed covariate and the outcome variable, and examine the performance of a multiple imputation procedure based on this distribution. We also investigate several alternative procedures that can be implemented by adapting multivariate normal multiple imputation software in ways that might be expected to perform well despite incompatibilities between model assumptions and true underlying relationships among the variables. The methods are compared in terms of bias, coverage, and CI width. As expected, the procedure based on the correct conditional distribution performs well across all scenarios. Just as importantly for general practitioners, several of the approaches based on multivariate normality perform comparably with the correct conditional distribution in a number of circumstances, although interestingly, procedures that seek to preserve the multiplicative relationship between the interaction term and the main-effects are found to be substantially less reliable. For illustration, the various procedures are applied to an analysis of post-traumatic stress disorder symptoms in a study of childhood trauma.

Methods for measuring diffusion of a social media-based health intervention

This study evaluated the feasibility of measuring diffusion from a social networking community-level intervention. One year after completion of a randomized controlled HIV prevention trial on Facebook, 112 minority men who have sex with men (MSM) were asked to refer African-American and/or Latino sex partners to complete a survey. Results suggest that, compared to non-referrers, referrers spent more time online, controlling for age, race, education, and condition. Over 60% of referrals reported hearing about the intervention, and over half reported that the referrer talked to them about changing health behaviors. Results provide support and initial feasibility of using social networking for diffusing community-based HIV interventions.

12-month outcomes of community engagement versus technical assistance to implement depression collaborative care: a partnered, cluster, randomized, comparative effectiveness trial

BACKGROUND

Depression collaborative care implementation using community engagement and planning (CEP) across programs improves 6-month client outcomes in minority communities, compared with technical assistance to individual programs (resources for services [RS]). However, 12-month outcomes are unknown.

OBJECTIVE

To compare effects of CEP and RS on mental health-related quality of life (MHRQL) and use of services among depressed clients at 12 months.

DESIGN

Matched health and community programs (n = 93) in 2 communities randomly assigned to receive CEP or RS. (ClinicalTrials.gov: NCT01699789).

MEASUREMENTS

Self-reported MHRQL and services use at baseline, 6 months, and 12 months.

SETTING

Los Angeles, California.

PATIENTS

1018 adults with depressive symptoms (8-item Patient Health Questionnaire score ≥10), 88% of whom were an ethnic minority.

INTERVENTION

CEP and RS to implement depression collaborative care.

MEASUREMENTS

The primary outcome was poor MHRQL (12-item mental health composite score ≤40) at baseline, 6 months, and 12 months; the secondary outcome was use of services at 12 months.

RESULTS

At 6 months, the finding that CEP outperformed RS to reduce poor MHRQL was significant but sensitive to underlying statistical assumptions. At 12 months, some analyses suggested that CEP was advantageous to MHRQL, whereas others did not confirm a significant difference favoring CEP. The finding that CEP reduced behavioral health hospitalizations at 6 months was less evident at 12 months and was sensitive to underlying statistical assumptions. Other services use did not significantly differ between interventions at 12 months.

LIMITATION

Data are self-reported, and findings are sensitive to modeling assumptions.

CONCLUSION

In contrast to 6-month results, no consistent effects of CEP on reducing the likelihood of poor MHRQL and behavioral health hospitalizations were found at 12 months. Still, given the needs of underresourced communities, the favorable profile of CEP, and the lack of evidence-based alternatives, CEP remains a viable strategy for policymakers and communities to consider.

PRIMARY FUNDING SOURCE

National Institute of Mental Health, Robert Wood Johnson Foundation, California Community Foundation, National Library of Medicine, and National Institutes of Health/National Center for Advancing Translational Science for the UCLA Clinical and Translational Science Institute.

Cognitive function after the initiation of adjuvant endocrine therapy in early-stage breast cancer: an observational cohort study

PURPOSE

This report examines cognitive complaints and neuropsychological (NP) testing outcomes in patients with early-stage breast cancer after the initiation of endocrine therapy (ET) to determine whether this therapy plays any role in post-treatment cognitive complaints.

PATIENTS AND METHODS

One hundred seventy-three participants from the Mind Body Study (MBS) observational cohort provided data from self-report questionnaires and NP testing obtained at enrollment (T1, before initiation of ET), and 6 months later (T2). Bivariate analyses compared demographic and treatment variables, cognitive complaints, depressive symptoms, quality of life, and NP functioning between those who received ET versus not. Multivariable linear regression models examined predictors of cognitive complaints at T2, including selected demographic variables, depressive symptoms, ET use, and other medical variables, along with NP domains that were identified in bivariate analyses.

RESULTS

Seventy percent of the 173 MBS participants initiated ET, evenly distributed between tamoxifen or aromatase inhibitors. ET-treated participants reported significantly increased language and communication (LC) cognitive complaints at T2 (P = .003), but no significant differences in NP test performance. Multivariable regression on LC at T2 found higher LC complaints significantly associated with T1 LC score (P < .001), ET at T2 (P = .004), interaction between ET and past hormone therapy (HT) (P < .001), and diminished improvement in NP psychomotor function (P = .05). Depressive symptoms were not significant (P = .10).

CONCLUSION

Higher LC complaints are significantly associated with ET 6 months after starting treatment and reflect diminished improvements in some NP tests. Past HT is a significant predictor of higher LC complaints after initiation of ET.

The association between nutrition facts label utilization and comprehension among Latinos in two east Los Angeles neighborhoods

BACKGROUND

The Nutrition Facts label can facilitate healthy dietary practices. There is a dearth of research on Latinos' utilization and comprehension of the Nutrition Facts label.

OBJECTIVE

To measure use and comprehension of the Nutrition Facts label and to identify correlates among Latinos in East Los Angeles, CA.

DESIGN

Cross-sectional interviewer-administered survey using computer-assisted personal interview software, conducted in either English or Spanish in the participant's home.

PARTICIPANTS/SETTING

Eligibility criteria were: living in a household within the block clusters identified, being age 18 years or older, speaking English or Spanish, identifying as Latino and as the household's main food purchaser and preparer. Analyses were based on 269 eligible respondents.

STATISTICAL ANALYSES PERFORMED

χ(2) test and multivariate logistic regression analysis assessed the associations among the main outcomes and demographics. Multiple imputations addressed missing data.

RESULTS

Sixty percent reported using the label; only 13% showed adequate comprehension of the label. Utilization was associated with being female, speaking Spanish, and being below the poverty line. Comprehension was associated with younger age, not being married, and higher education. Utilization was not associated with comprehension.

CONCLUSIONS

Latinos who are using the Nutrition Facts label are not correctly interpreting the available information. Targeted education is needed to improve use and comprehension of the Nutrition Facts label to directly improve diet, particularly among males, older Latinos, and those with less than a high school education.

Community-partnered cluster-randomized comparative effectiveness trial of community engagement and planning or resources for services to address depression disparities

BACKGROUND

Depression contributes to disability and there are ethnic/racial disparities in access and outcomes of care. Quality improvement (QI) programs for depression in primary care improve outcomes relative to usual care, but health, social and other community-based service sectors also support clients in under-resourced communities. Little is known about effects on client outcomes of strategies to implement depression QI across diverse sectors.

OBJECTIVE

To compare the effectiveness of Community Engagement and Planning (CEP) and Resources for Services (RS) to implement depression QI on clients' mental health-related quality of life (HRQL) and services use.

DESIGN

Matched programs from health, social and other service sectors were randomized to community engagement and planning (promoting inter-agency collaboration) or resources for services (individual program technical assistance plus outreach) to implement depression QI toolkits in Hollywood-Metro and South Los Angeles.

PARTICIPANTS

From 93 randomized programs, 4,440 clients were screened and of 1,322 depressed by the 8-item Patient Health Questionnaire (PHQ-8) and providing contact information, 1,246 enrolled and 1,018 in 90 programs completed baseline or 6-month follow-up.

MEASURES

Self-reported mental HRQL and probable depression (primary), physical activity, employment, homelessness risk factors (secondary) and services use.

RESULTS

CEP was more effective than RS at improving mental HRQL, increasing physical activity and reducing homelessness risk factors, rate of behavioral health hospitalization and medication visits among specialty care users (i.e. psychiatrists, mental health providers) while increasing depression visits among users of primary care/public health for depression and users of faith-based and park programs (each p < 0.05). Employment, use of antidepressants, and total contacts were not significantly affected (each p > 0.05).

CONCLUSION

Community engagement to build a collaborative approach to implementing depression QI across diverse programs was more effective than resources for services for individual programs in improving mental HRQL, physical activity and homelessness risk factors, and shifted utilization away from hospitalizations and specialty medication visits toward primary care and other sectors, offering an expanded health-home model to address multiple disparities for depressed safety-net clients.

Community-partnered evaluation of depression services for clients of community-based agencies in under-resourced communities in Los Angeles

BACKGROUND

As medical homes are developing under health reform, little is known regarding depression services need and use by diverse safety-net populations in under-resourced communities. For chronic conditions like depression, primary care services may face new opportunities to partner with diverse community service providers, such as those in social service and substance abuse centers, to support a collaborative care model of treating depression.

OBJECTIVE

To understand the distribution of need and current burden of services for depression in under-resourced, diverse communities in Los Angeles.

DESIGN

Baseline phase of a participatory trial to improve depression services with data from client screening and follow-up surveys.

PARTICIPANTS

Of 4,440 clients screened from 93 programs (primary care, mental health, substance abuse, homeless, social and other community services) in 50 agencies, 1,322 were depressed according to an eight-item Patient Health Questionnaire (PHQ-8) and gave contact information; 1,246 enrolled and 981 completed surveys. Ninety-three programs, including 17 primary care/public health, 18 mental health, 20 substance abuse, ten homeless services, and 28 social/other community services, participated.

MAIN MEASURES

Comparisons by setting in 6-month retrospective recall of depression services use.

KEY RESULTS

Depression prevalence ranged from 51.9 % in mental health to 17.2 % in social-community programs. Depressed clients used two settings on average to receive depression services; 82 % used any setting. More clients preferred counseling over medication for depression treatment.

CONCLUSIONS

Need for depression care was high, and a broad range of agencies provide depression care. Although most participants had contact with primary care, most depression services occurred outside of primary care settings, emphasizing the need to coordinate and support the quality of community-based services across diverse community settings.

Cardiovascular health: associations with race-ethnicity, nativity, and education in a diverse, population-based sample of Californians

PURPOSE

This study examined how race-ethnicity, nativity, and education interact to influence disparities in cardiovascular (CV) health, a new concept defined by the American Heart Association. We assessed whether race-ethnicity and nativity disparities in CV health vary by education and whether the foreign-born differ in CV health from their U.S.-born race-ethnic counterparts with comparable education.

METHODS

We used data from the 2009 California Health Interview Survey to determine the prevalence of optimal CV health metrics (based on selected American Heart Association guidelines) among adults ages 25 and older (n = 42,014). We examined the interaction between education and ethnicity-nativity, comparing predicted probabilities of each CV health measure between U.S.-born and foreign-born White, Asian, and Latino respondents.

RESULTS

All groups were at high risk of suboptimal physical activity levels, fruit and vegetable and fast food consumption, and overweight/obesity. Those with greater education were generally better off except among Asian respondents. Ethnicity-nativity differences were more pronounced among those with less than a college degree. The foreign-born respondents exhibited both advantages and disadvantages in CV health compared with their U.S.-born counterparts that varied by ethnicity-nativity.

CONCLUSIONS

Education influences ethnicity-nativity disparities in CV health, with most race-ethnic and nativity differences occurring among the less educated. Studies of nativity differences in CV health should stratify by education in order to adequately address SES differences.

Cognitive complaints after breast cancer treatments: examining the relationship with neuropsychological test performance

BACKGROUND

Cognitive complaints are reported frequently after breast cancer treatments. Their association with neuropsychological (NP) test performance is not well-established.

METHODS

Early-stage, posttreatment breast cancer patients were enrolled in a prospective, longitudinal, cohort study prior to starting endocrine therapy. Evaluation included an NP test battery and self-report questionnaires assessing symptoms, including cognitive complaints. Multivariable regression models assessed associations among cognitive complaints, mood, treatment exposures, and NP test performance.

RESULTS

One hundred eighty-nine breast cancer patients, aged 21-65 years, completed the evaluation; 23.3% endorsed higher memory complaints and 19.0% reported higher executive function complaints (>1 SD above the mean for healthy control sample). Regression modeling demonstrated a statistically significant association of higher memory complaints with combined chemotherapy and radiation treatments (P = .01), poorer NP verbal memory performance (P = .02), and higher depressive symptoms (P < .001), controlling for age and IQ. For executive functioning complaints, multivariable modeling controlling for age, IQ, and other confounds demonstrated statistically significant associations with better NP visual memory performance (P = .03) and higher depressive symptoms (P < .001), whereas combined chemotherapy and radiation treatment (P = .05) approached statistical significance.

CONCLUSIONS

About one in five post-adjuvant treatment breast cancer patients had elevated memory and/or executive function complaints that were statistically significantly associated with domain-specific NP test performances and depressive symptoms; combined chemotherapy and radiation treatment was also statistically significantly associated with memory complaints. These results and other emerging studies suggest that subjective cognitive complaints in part reflect objective NP performance, although their etiology and biology appear to be multifactorial, motivating further transdisciplinary research.

A Bayesian approach to improved estimation of causal effect predictiveness for a principal surrogate endpoint

The literature on potential outcomes has shown that traditional methods for characterizing surrogate endpoints in clinical trials based only on observed quantities can fail to capture causal relationships between treatments, surrogates, and outcomes. Building on the potential-outcomes formulation of a principal surrogate, we introduce a Bayesian method to estimate the causal effect predictiveness (CEP) surface and quantify a candidate surrogate's utility for reliably predicting clinical outcomes. In considering the full joint distribution of all potentially observable quantities, our Bayesian approach has the following features. First, our approach illuminates implicit assumptions embedded in previously-used estimation strategies that have been shown to result in poor performance. Second, our approach provides tools for making explicit and scientifically-interpretable assumptions regarding associations about which observed data are not informative. Through simulations based on an HIV vaccine trial, we found that the Bayesian approach can produce estimates of the CEP surface with improved performance compared to previous methods. Third, our approach can extend principal-surrogate estimation beyond the previously considered setting of a vaccine trial where the candidate surrogate is constant in one arm of the study. We illustrate this extension through an application to an AIDS therapy trial where the candidate surrogate varies in both treatment arms.

THE POTENTIAL FOR BIAS IN PRINCIPAL CAUSAL EFFECT ESTIMATION WHEN TREATMENT RECEIVED DEPENDS ON A KEY COVARIATE

Motivated by a potential-outcomes perspective, the idea of principal stratification has been widely recognized for its relevance in settings susceptible to posttreatment selection bias such as randomized clinical trials where treatment received can differ from treatment assigned. In one such setting, we address subtleties involved in inference for causal effects when using a key covariate to predict membership in latent principal strata. We show that when treatment received can differ from treatment assigned in both study arms, incorporating a stratum-predictive covariate can make estimates of the "complier average causal effect" (CACE) derive from observations in the two treatment arms with different covariate distributions. Adopting a Bayesian perspective and using Markov chain Monte Carlo for computation, we develop posterior checks that characterize the extent to which incorporating the pretreatment covariate endangers estimation of the CACE. We apply the method to analyze a clinical trial comparing two treatments for jaw fractures in which the study protocol allowed surgeons to overrule both possible randomized treatment assignments based on their clinical judgment and the data contained a key covariate (injury severity) predictive of treatment received.

Physical and psychosocial recovery in the year after primary treatment of breast cancer

PURPOSE

The 2000 National Institutes of Health Consensus Conference on Adjuvant Therapy of Breast Cancer recommended chemotherapy for all women with invasive cancer greater than 1 centimeter. Studies of long-term breast cancer survivors have found poorer quality of life (QOL) in women who received adjuvant chemotherapy. The aim of this article is to characterize physical and psychosocial recovery as a function of chemotherapy receipt in the year after medical treatment completion.

PATIENTS AND METHODS

Prospective longitudinal survey data (RAND SF-36 and Breast Cancer Prevention Trial [BCPT] Symptom Scales) collected from 558 women with breast cancer enrolled on the Moving Beyond Cancer (MBC) psychoeducational intervention trial were compared according to receipt of chemotherapy. MBC study enrollment occurred within 4 weeks after the end of primary treatment (eg, surgery, chemotherapy, radiation). Self-report questionnaire data collected at enrollment and at 2, 6, and 12 months thereafter were examined, controlling for intervention and with propensity score adjustment for imbalance of covariates. Outcome analyses were carried out by fitting linear mixed models by using SAS PROC MIXED.

RESULTS

Longitudinal SF-36 scale scores did not differ by chemotherapy treatment exposure, and both groups improved significantly (P < .01) in the year after primary treatment ended. However, adjuvant chemotherapy treatment was associated with significantly more severe physical symptoms, including musculoskeletal pain (P = .01), vaginal problems (P < .01), weight problems (P = .01), and nausea (P = .03).

CONCLUSION

Physical and psychosocial functioning improved significantly after breast cancer treatment, independent of receipt of adjuvant chemotherapy. Women who received chemotherapy experienced more severe and persistent physical symptoms that should be more effectively managed as part of survivorship care.

Predictors of risk tolerance among oral surgery patients

PURPOSE

This study attempts to provide insight on how the treatment preference for a mandible fracture and treatment received and its consequences are related to the patient's risk tolerance, as measured by the Standard Gamble (SG).

PATIENTS AND METHODS

Data from a prospective cohort study of 203 subjects receiving treatment at the former King/Drew Medical Center in Los Angeles, CA, for either a mandible fracture (n = 98) or third molar removal (n = 105) were examined. Subjects were interviewed at 4 time points: on admission to the medical center and at 3 monthly follow-up visits. Risk tolerance for hypothetical treatment scenarios is measured by use of the SG, a health-value utility measure assessing the tradeoff between good outcomes and serious complications associated with treatment. Separate regression analyses with subsets of predictors (sociodemographic, psychosocial health, and clinical characteristics) were conducted and then synthesized by use of the significant predictors in separate analyses.

RESULTS

For fracture subjects, there was a noticeable rise in the SG reports from admission to the 1-month follow-up. Their greater risk tolerance was associated with being older, receiving surgery, having a lower post-traumatic stress disorder score, and having a swollen jaw or face. For third molar subjects, SG did not change substantively over the course of the study. Predictors of greater risk tolerance for third molar subjects included the jaw or face being swollen and having to use less pain medication.

CONCLUSIONS

Findings from this study show a preference for less invasive treatment, with the majority of both groups preferring wiring, and support the theory that treatment choices differ between subjects with different health states. Factors associated with risk tolerance include the patient's age, treatment received, psychosocial health state, experience with previous treatment, and value for oral health quality of life.

Predictors of treatment preference for mandibular fracture

BACKGROUND

Patient treatment preferences do not necessarily remain stable over time.

OBJECTIVE

This study focuses on predictors of patient treatment choice and on the extent to which patients are willing to take risks by choosing surgical versus nonsurgical treatment for mandibular fracture.

METHODS

Surveys of African-American and Hispanic adults receiving treatment at King/Drew Medical Center for either a mandibular fracture (n = 98) or third-molar removal (n = 105) were used to investigate patterns of patient preference over the course of a 4-month study period using generalized estimating equations controlling for age, gender, income, and fracture versus third-molar patient. The study examined the effects of symptom rating and a"standard gamble" measure reflecting a patient's willingness to accept scarring or nerve damage. This analysis is based on 169 patients who participated in four waves of data collection.

RESULTS

The most salient predictor of patient treatment was the standard gamble measure at 1-month follow-up. Subjects with higher risk tolerance were more likely to select surgery versus jaw wiring. A higher likelihood of choosing surgery was associated with higher income and greater symptom severity. Fracture patients were more likely to select surgery compared with third-molar patients.

CONCLUSIONS

The significance of symptom severity 1-month post-surgery raises an important issue regarding the healing process. Moreover, the significance of standard gamble as a predictor of treatment choice for mandibular fracture should encourage other researchers to use this measure of willingness to accept risk when studying acute conditions such as jaw fracture.

Exploring nonresponse bias in a health survey using neighborhood characteristics

OBJECTIVES

We examined potential nonresponse bias in a large-scale, population-based, random-digit-dialed telephone survey in California and its association with the response rate.

METHODS

We used California Health Interview Survey (CHIS) data and US Census data and linked the two data sets at the census tract level. We compared a broad range of neighborhood characteristics of respondents and nonrespondents to CHIS. We projected individual-level nonresponse bias using the neighborhood characteristics.

RESULTS

We found little to no substantial difference in neighborhood characteristics between respondents and nonrespondents. The response propensity of the CHIS sample was similarly distributed across these characteristics. The projected nonresponse bias appeared very small.

CONCLUSIONS

The response rate in CHIS did not result in significant nonresponse bias and did not substantially affect the level of data representativeness, and it is not valid to focus on response rates alone in determining the quality of survey data.

Inflammatory biomarkers and fatigue during radiation therapy for breast and prostate cancer

PURPOSE

Biomarkers of radiation-induced behavioral symptoms, such as fatigue, have not been identified. Studies linking inflammatory processes to fatigue in cancer survivors led us to test the hypothesis that activation of the proinflammatory cytokine network is associated with fatigue symptoms during radiation therapy for breast and prostate cancer.

EXPERIMENTAL DESIGN

Individuals with early-stage breast (n = 28) and prostate cancer (n = 20) completed questionnaires and provided blood samples for determination of serum levels of interleukin 1beta (IL-1beta) and IL-6 at assessments conducted before, during, and after a course of radiation therapy. Serum markers of proinflammatory cytokine activity, including IL-1 receptor antagonist and C-reactive protein, were examined in a subset of participants. Random coefficient models were used to evaluate the association between changes in cytokine levels and fatigue.

RESULTS

As expected, there was a significant increase in fatigue during radiation treatment. Changes in serum levels of inflammatory markers C-reactive protein and IL-1 receptor antagonist were positively associated with increases in fatigue symptoms (Ps < 0.05), although serum levels of IL-1beta and IL-6 were not associated with fatigue. These effects remained significant (Ps < 0.05) in analyses controlling for potential biobehavioral confounding factors, including age, body mass index, hormone therapy, depression, and sleep disturbance.

CONCLUSIONS

Results suggest that activation of the proinflammatory cytokine network and associated increases in downstream biomarkers of proinflammatory cytokine activity are associated with fatigue during radiation therapy for breast and prostate cancer.

Addressing intimacy and partner communication after breast cancer: a randomized controlled group intervention

While quality of life for most breast cancer survivors (BCS) returns to normal by 1 year post-treatment, problems in sexual function and intimacy often persist. The present study tested the efficacy of a 6-week psycho-educational group intervention in improving BCS's sexual well-being. We conducted a mailed survey of BCS 1-5 years post-diagnosis to identify a sample of women who reported moderately severe problems in body image, sexual function or partner communication, and were deemed eligible for the randomized intervention trial. Using a pre-randomized design, 70% (n = 284) were assigned to a 6-week psycho-educational group intervention and 30% (n = 127) were assigned to a control condition (print material only); however, only 83 BCS agreed to participate in the intervention. Four months post-intervention, the intervention and control groups were not significantly different on the primary outcome of emotional functioning; however, BCS randomized to the intervention group were more likely to report improvements in relationship adjustment and communication as well as increased satisfaction with sex compared to controls. Members of the intervention group who were the least satisfied with their sexual relationship appeared to improve the most. Although modest in its effects, this intervention can be delivered in standard clinical settings. Having an identified treatment may help reduce physician reluctance to ask BCS about problems in intimacy and as appropriate, refer them for timely help.

Applying permutation tests with adjustment for covariates and attrition weights to randomized trials of health-services interventions

Using a health-services study as an illustrative example of longitudinal randomized field research with the potential for participants to be lost to follow-up, we apply a permutation test where the treatment indicator variable is randomly permuted in the context of regression models with covariates and attrition weighting. The test is applied to a multi-site randomized intervention trial of a quality-improvement program for adolescent depression treatment in primary-care settings, in which regression models were used to assess intervention effects with weights used to adjust for attrition bias. The foundation and motivation for this approach to the analysis are considered with attention to the demands associated with implementing such a strategy. The results from the permutation tests were qualitatively similar to the results obtained from conventional parametric models, and in fact suggested that the significance level from the conventional t-test was understated in this application.

Using an Approximate Bayesian Bootstrap to Multiply Impute Nonignorable Missing Data

An Approximate Bayesian Bootstrap (ABB) offers advantages in incorporating appropriate uncertainty when imputing missing data, but most implementations of the ABB have lacked the ability to handle nonignorable missing data where the probability of missingness depends on unobserved values. This paper outlines a strategy for using an ABB to multiply impute nonignorable missing data. The method allows the user to draw inferences and perform sensitivity analyses when the missing data mechanism cannot automatically be assumed to be ignorable. Results from imputing missing values in a longitudinal depression treatment trial as well as a simulation study are presented to demonstrate the method's performance. We show that a procedure that uses a different type of ABB for each imputed data set accounts for appropriate uncertainty and provides nominal coverage.

Do the benefits of rigid internal fixation of mandible fractures justify the added costs? Results from a randomized controlled trial

PURPOSE

Owing to its putative advantages over conventional maxillomandibular fixation (MMF), open-reduction and rigid internal fixation (ORIF) is used frequently to treat mandible fractures, particularly in noncompliant patients. The resource-intensive nature of ORIF, the large variation in its use, and the lack of systematic studies substantiating ORIF attributed benefits compel a randomized controlled investigation comparing ORIF to MMF treatment. The objective of this study was to determine whether ORIF provides better clinical and functional outcomes than MMF in noncomplying type of patients with a similar range of mandible fracture severity.

PATIENTS AND METHODS

From a total of 336 patients who sought treatment for mandible fractures, 142 patients with moderately severe mandible fractures were assigned randomly to receive MMF or ORIF and followed prospectively for 12 months. A variety of clinician and patient-reported measures were used to assess outcomes at the 1, 6, and 12 months follow-up visits. These measures included clinician-reported number of surgical complications, patient-reported number of complaints, as well as cumulative costs of treatment. Pain intensity was measured on a 10-point scale and the 12-item General Oral Health Assessment Index was used to assess the patients' oral health-related quality of life. Because the protocol allowed clinical judgment to overrule the randomly assigned treatment, outcomes were compared on an "intent-to-treat" basis as well as in terms of actual treatment received.

RESULTS

The sociodemographic and clinical characteristics of the injury did not differ among the 2 groups. On an intent-to-treat basis, the difference in complication rates was not significant but favored MMF; 8.1% of patients developed complications with MMF versus 12.5% with ORIF. Differences in the rate of patient complaints were not significant on an intent-to-treat basis, but a significant between-group difference (P = .012) favoring MMF was noted on an as-treated basis at the 1 month recall, with 40% of ORIF patients reporting greater than 1 complaint versus 18.8% of MMF patients. No significant differences were detected between the 2 treatment groups at any time point with respect to oral health-related quality of life reflected by the General Oral Health Assessment Index scores. In-patient days and total costs did not differ significantly on an intent-to-treat basis, but on an as-treated basis, patients treated with MMF had fewer in-patient days on average (1.64 vs 5.50 for ORIF) and lower average costs of treatment ($7,206 vs $26,089 for ORIF). In the intent-to-treat analyses, patients receiving MMF treatment had significantly lower (P = .05) pain scores at the 12-month recall (mean = 0.58, SE = 0.30) compared with patients assigned to ORIF (mean = 1.78, SE = 0.52).

CONCLUSION

Our study did not show a clear overall benefit of the resource-intensive ORIF over conventional MMF treatment in the management of moderately severe mandible fractures in at-risk patients; our data instead suggest some cost as well as oral health quality-of-life advantages for the use of MMF in this patient population.