Trent D. Buskirk

The rationale, design, and implementation of the American Cancer Society's studies of cancer survivors

The American Cancer Society (ACS) defines cancer survivorship as beginning at diagnosis with cancer and continuing for the balance of life and views quality of life (QOL) as a key outcome. In this article, the authors describe the rationale, methodology, and sample characteristics of the 2 ACS Studies of Cancer Survivors (SCS): 1) a longitudinal study identifying and surveying survivors approximately 1 year postdiagnosis that includes plans to resurvey the panel at 2 years, 7 years, and 12 years postdiagnosis to identify predictors of QOL; and 2) a cross‐sectional study of QOL among 3 separate cohorts of survivors who were approximately 3 years, 6 years, and 11 years postdiagnosis at the time of data collection. Survivors of prostate, breast, lung, colorectal, bladder, skin, kidney, ovarian, and uterine cancers and of non‐Hodgkin lymphoma were sampled from 25 different central cancer registries, with African‐American and Hispanic survivors over sampled. Survivors completed either mail or telephone surveys that described their physical, psychological, social, and spiritual functioning. The overall recruitment rate was 34.0%; 15411 participants completed surveys, of whom 40.1% had a high school education or less and 19.4% were racial/ethnic minorities. The SCS surveys provide a large diagnostically, geographically, and demographically diverse database on cancer survivorship that was designed to overcome some of the limitations of past research. Future reports will compare QOL of survivors at different well‐defined times postdiagnosis, investigate the issues of understudied populations and diagnostic groups, and describe survivor QOL at state levels. Insights valuable to those considering registry‐based studies are offered on issues of ascertainment, sampling, and recruitment. Cancer 2007.

Worksite Opportunities for Wellness (WOW): Effects on cardiovascular disease risk factors after 1 year

OBJECTIVE To evaluate the effectiveness of a worksite health promotion program on improving cardiovascular disease risk factors. METHODS In St Louis, Missouri from 2005 to 2006, 151 employees (134 F, 17 M, 81% overweight/obese) participated in a cohort-randomized trial comparing assessments + intervention (worksite A) with assessments only (worksite B) for 1 year. All participants received personal health reports containing their assessment results. The intervention was designed to promote physical activity and favorable dietary patterns using pedometers, healthy snack cart, WeightWatchers(R) meetings, group exercise classes, seminars, team competitions, and participation rewards. Outcomes included BMI, body composition, blood pressure, fitness, lipids, and Framingham 10-year coronary heart disease risk. RESULTS 123 participants, aged 45+/-9 yr, with BMI 32.9+/-8.8 kg/m(2) completed 1 year. Improvements (P< or =0.05) were observed at both worksites for fitness, blood pressure, and total-, HDL-, and LDL-cholesterol. Additional improvements occurred at worksite A in BMI, fat mass, Framingham risk score, and prevalence of the metabolic syndrome; only the changes in BMI and fat mass were different between worksites. CONCLUSION A multi-faceted worksite intervention promoted favorable changes in cardiovascular disease risk factors, but many of the improvements were achieved with worksite health assessments and personalized health reports in the absence of an intervention.

What makes cancer survivor stories work? An empirical study among African American women

IntroductionCancer survivors play a vital role in cancer control as messengers of hope and information, and advocates for prevention and screening. Understanding what makes survivor stories effective can enhance survivor-delivered programs and interventions.MethodsBy random assignment and using a cross-classified design, 200 African American women viewed videotaped stories (n = 300) from 36 African American breast cancer survivors. Analyses examined effects of story attributes (narrative quality, health message strength), participant characteristics (ways of knowing, experience with breast cancer) and identification with the survivor on women’s: (1) level of engagement in the story; (2) positive thoughts about the story; and, (3) remembering key messages about breast cancer and mammography in the story.ResultsParticipant characteristics were significant predictors of all three study outcomes, accounting for 27.8, 2.6 and 22.2% of their total variance, respectively. In comparison, the variability in these outcomes that could be attributed to differences in the stories was small (0.6, 1.1 and 2%, respectively). The effects of participant characteristics on level of engagement and positive thoughts were mediated by identification with the survivor.ConclusionsThe best predictor of a woman becoming engaged in a breast cancer survivor’s story and having positive thoughts about the story was whether she liked the survivor and viewed her as similar to herself (i.e., identification).Implications for cancer survivorsSurvivor stories may be most effective when audience members identify with the survivor. Finding key characteristics that can reliably match the two will advance cancer communication science and practice.

Using Text Messages in U.S. Mobile Phone Surveys

Attempts to interview the general population using cell phones have revealed underlying weaknesses. Noncoverage is severe because all persons who rely solely on fixedline telephones are excluded. Nonresponse stemming from the difficulty of contacting potential respondents and convincing them to participate also increases. Only the spread of mobile technology will remedy noncoverage. This article presents the results of an experiment that tests the effectiveness of sending text messages to improve non-response. The results give little support to our hypothesis that sending an advance text message would make contacting respondents easier. Other outcome rates do show positive effects. We also discovered that the act of sending the text message provided valuable data about the sample unit. This outside information, similar to the information traditional telephone surveys obtain from directories, indicates whether a mobile number is in service and allows researchers to adapt their calling rules to achieve greater efficiency.

Making Mobile Browser Surveys Smarter Results from a Randomized Experiment Comparing Online Surveys Completed via Computer or Smartphone

With nearly 50% of U.S. mobile phone subscribers using smartphones, survey researchers are beginning to explore their use as a data collection tool. The Got Healthy Apps Study (GHAS) conducted a randomized experiment to compare mode effects for a survey completed via iPhone mobile browser and online via desktop/laptop computer web browser. Mode effects were assessed for three types of outcomes: randomization/recruitment, survey process/completion, and survey items. In short, the distribution of survey completion times and the distribution of the number of apps owned were significantly different across survey mode after accounting for block group. Other key mode effects outcomes (including open-ended items, slider bar questions, and missing item rates) showed no significant differences across survey mode. Some interesting qualitative findings suggest that iPhone respondents enter more characters and omit fewer items than originally thought.

Disparities by Race and Ethnicity in Cancer Survivor Stories Available on the Web

Background The rapid growth of eHealth could have the unintended effect of deepening health disparities between population subgroups. Most concerns to date have focused on population differences in access to technology, but differences may also exist in the appropriateness of online health content for diverse populations. Objective This paper reports findings from the first descriptive study of online cancer survivor stories by race and ethnicity of the survivor. Methods Using the five highest-rated Internet search engines and a set of search terms that a layperson would use to find cancer survivor stories online, we identified 3738 distinct sites. Of these, 106 met study criteria and contained 7995 total stories, including 1670 with an accompanying photo or video image of the survivor. Characteristics of both websites and survivor stories were coded. Results All racial minority groups combined accounted for 9.8% of online cancer survivor stories, despite making up at least 16.3% of prevalent cancer cases. Also notably underrepresented were stories from people of Hispanic ethnicity (4.1%), men (35.7%), survivors of colon cancer (3.5%), and older adults. Conclusions Because racial/ethnic minority cancer survivors are underrepresented in survivor stories available online, it is unlikely that this eHealth resource in its current form will help eliminate the disproportionate burden of cancer experienced by these groups.

Effects of patient health literacy, patient engagement and a system-level health literacy attribute on patient-reported outcomes: a representative statewide survey.

BackgroundThe effects of health literacy are thought to be based on interactions between patients’ skill levels and health care system demands. Little health literacy research has focused on attributes of health care organizations. We examined whether the attribute of individuals’ experiences with front desk staff, patient engagement through bringing questions to a doctor visit, and health literacy skills were related to two patient-reported outcomes.MethodsWe administered a telephone survey with two sampling frames (i.e., household landline, cell phone numbers) to a randomly selected statewide sample of 3358 English-speaking adult residents of Missouri. We examined two patient-reported outcomes – whether or not respondents reported knowing more about their health and made better choices about their health following their last doctor visit. Multivariable logistic regression models were used to examine the independent contributions of predictor variables (i.e., front desk staff, bringing questions to a doctor visit, health literacy skills).ResultsControlling for self-reported health, having a personal doctor, time since last visit, number of chronic conditions, health insurance, and sociodemographic characteristics, respondents who had a good front desk experience were 2.65 times as likely (95% confidence interval [CI]: 2.13, 3.30) and those who brought questions were 1.73 times as likely (95% CI: 1.32, 2.27) to report knowing more about their health after seeing a doctor. In a second model, respondents who had a good front desk experience were 1.57 times as likely (95% CI: 1.26, 1.95) and those who brought questions were 1.66 times as likely (95% CI: 1.29, 2.14) to report making better choices about their health after seeing a doctor. Patients’ health literacy skills were not associated with either outcome.ConclusionsResults from this representative statewide survey may indicate that one attribute of a health care organization (i.e., having a respectful workforce) and patient engagement through question asking may be more important to patient knowledge and health behaviors than patients’ health literacy skills. Findings support focused research to examine the effects of organizational attributes on patient health outcomes and system-level interventions that might enhance patient health.

N the Network

Despite higher hit rates for cell phone samples, inefficiencies in processing calls to these numbers relative to landline numbers continue to be documented in the U.S. literature. In this study, we propose one method for using cell phone provider information and Internet resources for validating number status. Specifically, we describe how we used ‘‘in network’’ options available from three major providers’ web sites to determine the validity of cell phone numbers. We tested differences in working number rates (WNRs) among valid and nonvalid numbers against a normal processing control group and determined that the WNR among valid numbers was approximately 14 percentage points higher than the WNR of the comparison group. This process also shows promise in reducing the effort required to determine working status and may provide a basis for developing screening tools for cell phones that capitalize on resources that are unique to this technology.

An Introduction to Machine Learning Methods for Survey Researchers

Machine learning techniques comprise an array of computer-intensive methods that aim at discovering patterns in data using flexible, often nonparametric, methods for modeling and variable selection. These methods offer an expansion to the more traditional methods, such as OLS or logistic regression, which have been used by survey researchers and social scientists. Many of the machine learning methods do not require the distributional assumptions of the more traditional methods, and many do not require explicit model specification prior to estimation. Machine learning methods are beginning to be used for various aspects of survey research including responsive/adaptive designs, data processing and nonresponse adjustments and weighting. This special issue aims to familiarize survey researchers and social scientists with the basic concepts in machine learning and highlights five common methods. Specifically, articles in this issue will offer an accessible introduction to: LASSO models, support vector machines, neural networks, and classification and regression trees and random forests. In addition to a detailed description, each article will highlight how the respective method is being used in survey research along with an application of the method to a common example. The introductory article will provide an accessible introduction to some commonly used concepts and terms associated with machine learning modeling and evaluation. The introduction also provides a description of the data set that was used as the common application example for each of the five machine learning methods.

Assessing community-based injury prevention services in U.S. Children’s Hospitals

Objective Not-for-profit hospitals are required to meet federal reporting requirements detailing their community benefit activities, which support their tax-exempt status. Childrens hospitals have long provided community injury prevention (IP) programming and thus can inform public health outreach work in other areas. This work describes IP programming as a community service offered by childrens hospitals in the U.S. Methods The IP specialist at 232 US-based member institutions of the Childrens Hospital Association were invited to complete an assessment of their hospitals IP outreach programming. Results 47.7 percent of hospitals request financial data from IP programming for tax reporting purposes. Almost all offer injury prevention (IP) services; the majority are in the community (60.3%) and 34.5% are hospital-based. Most IP units are independent (60.3%) and 71.8% are responsible for their own budgets. Conclusions By integrating dissemination and implementation sciences and community health needs assessments, these findings can help advance community services provided by hospitals to impact public health.