Faculty Bibliography

AIM/OBJECTIVE:To illustrate with the example of a secondary data analysis study the use of the multiple imputation method to replace missing data. BACKGROUND:Most large public datasets have missing data, which need to be handled by researchers conducting secondary data analysis studies. Multiple imputation is a technique widely used to replace missing values while preserving the sample size and sampling variability of the data. DATA SOURCE/METHODS:The 2004 National Sample Survey of Registered Nurses. REVIEW METHODS/METHODS:The authors created a model to impute missing values using the chained equation method. They used imputation diagnostics procedures and conducted regression analysis of imputed data to determine the differences between the log hourly wages of internationally educated and US-educated registered nurses. DISCUSSION/CONCLUSIONS:The authors used multiple imputation procedures to replace missing values in a large dataset with 29,059 observations. Five multiple imputed datasets were created. Imputation diagnostics using time series and density plots showed that imputation was successful. The authors also present an example of the use of multiple imputed datasets to conduct regression analysis to answer a substantive research question. CONCLUSION/CONCLUSIONS:Multiple imputation is a powerful technique for imputing missing values in large datasets while preserving the sample size and variance of the data. Even though the chained equation method involves complex statistical computations, recent innovations in software and computation have made it possible for researchers to conduct this technique on large datasets. IMPLICATIONS FOR RESEARCH/PRACTICE/CONCLUSIONS:The authors recommend nurse researchers use multiple imputation methods for handling missing data to improve the statistical power and external validity of their studies.

BACKGROUND: In light of the current sentencing disparity (18:1) between crack and powder cocaine possession in the United States, we examined socioeconomic correlates of use of each, and relations between use and arrest, to determine who may be at highest risk for arrest and imprisonment. METHODS: We conducted secondary data analyses on the National Survey on Drug Use and Health, 2009-2012. Data were analyzed for adults age >/=18 to determine associations between use and arrest. Socioeconomic correlates of lifetime and annual use of powder cocaine and of crack were delineated using multivariable logistic regression and correlates of frequency of recent use were examined using generalized negative binomial regression. RESULTS: Crack users were at higher risk than powder cocaine users for reporting a lifetime arrest or multiple recent arrests. Racial minorities were at low risk for powder cocaine use and Hispanics were at low risk for crack use. Blacks were at increased risk for lifetime and recent crack use, but not when controlling for other socioeconomic variables. However, blacks who did use either powder cocaine or crack tended to use at higher frequencies. Higher education and higher family income were negatively associated with crack use although these factors were sometimes risk factors for powder cocaine use. CONCLUSIONS: Crack users are at higher risk of arrest and tend to be of lower socioeconomic status compared to powder cocaine users. These findings can inform US Congress as they review bills (e.g., The Smarter Sentencing Act), which would help eliminate cocaine-related sentencing disparities.

BACKGROUND AND OBJECTIVES: Promoted globally as an evidence-based intervention in the prevention of HIV and treatment of heroin addiction among people who inject drugs (PWID), opioid substitution treatment (OST) can help control emerging HIV epidemics among PWID. With implementation in December 2014, Kenya is the third Sub-Saharan African country to have introduced OST. We combine dynamic mathematical modelling with qualitative sociological research to examine the 'promise of methadone' to Kenya. METHODS, SETTING AND PARTICIPANTS: We model the HIV prevention impact of OST in Nairobi, Kenya, at different levels of intervention coverage. We draw on thematic analyses of 109 qualitative interviews with PWID, and 43 with stakeholders, to chart their narratives of expectation in relation to the promise of methadone. RESULTS: The modelled impact of OST shows relatively slight reductions in HIV incidence (5-10%) and prevalence (2-4%) over 5 years at coverage levels (around 10%) anticipated in the planned roll-out of OST. However, there is a higher impact with increased coverage, with 40% coverage producing a 20% reduction in HIV incidence, even when accounting for relatively high sexual transmissions. Qualitative findings emphasise a culture of 'rationed expectation' in relation to access to care and a 'poverty of drug treatment opportunity'. In this context, the promise of methadone may be narrated as a symbol of hope-both for individuals and community-in relation to addiction recovery. CONCLUSIONS: Methadone offers HIV prevention potential, but there is a need to better model the effects of sexual HIV transmission in mediating the impact of OST among PWID in settings characterised by a combination of generalised and concentrated epidemics. We find that individual and community narratives of methadone as hope for recovery coexist with policy narratives positioning methadone primarily in relation to HIV prevention. Our analyses show the value of mixed methods approaches to investigating newly-introduced interventions.

Aims: Implementation of substance use screening in general medical settings is hindered by the lack of a brief yet precise and comprehensive screening tool that is compatible with clinical workflows. To address this need, we developed the Substance Use Brief Screen (SUBS); a 4-item screener for tobacco, alcohol, and drug use (illicit and prescription) that is self-administered and may be easily integrated with electronic health records. Methods: Adult patients were recruited consecutively in the waiting area of an urban safety net primary care clinic. The SUBS was self-administered in English on touchscreen tablet computers. Reference standard measures of unhealthy substance use and substance use disorders were then administered, including self reported measures and saliva drug tests. The SUBS was compared against the reference standards to determine its sensitivity, specificity, and area under the curve (AUC) for each substance class. Results: Among the 390 participants, rates of past year use reported on the SUBS were 37% tobacco, 43% alcohol (4+ drinks/day), 20% illicit drugs, and 12% prescription drugs. Sensitivity and specificity of the SUBS for detecting past year unhealthy use were: tobacco 99% and 91% (AUC = .95); alcohol 94% and 68% (AUC = .81); drugs (illicit or prescription) 84% and 89% (AUC = .86). Sensitivity was lower for prescription drugs (57%) than for illicit drugs (78%). For detecting a substance use disorder, sensitivity and specificity were: tobacco 100% and 73% (AUC = .87); alcohol 93% and 64% (AUC = .79); drugs 85% and 82% (AUC = .84). Conclusions: The SUBS accurately identified unhealthy tobacco, alcohol, and drug use in this primary care sample, and had high sensitivity but lower specificity for identifying substance use disorders. Individuals screening positive on the SUBS should receive further assessment. Our findings support use of the SUBS for substance use screening in primary care, but additional tools may be needed for prescription drugs

Using latent class analysis, this study examined the pattern of non-opioid substance misuse among 19,101 enrollees into 85 opioid treatment programs. The most frequent non-opioid drugs were cannabis, anti-anxiety medications, and cocaine. Four non-opioid drug use latent classes were identified: low-use (73%), prescription drug use (16%), marijuana and cocaine use (8.5%), and poly-drug use (2.5%). Compared to the low-use class, participants in the other classes were more likely to be female, Caucasian, use tobacco, have chronic pain, and use prescription opioids either with or without heroin. Recognition of characteristics derived from these classes can improve opioid treatment program services.

There is growing awareness that students' experiences of stress may impede academic success, compromise mental health, and promote substance use. We examined these factors in an under-studied population, private/independent high school students, using a multi-method (qualitative and quantitative), iterative data collection and analytic process. We first conducted qualitative interviews with faculty and staff at a number of highly competitive private schools, followed by an anonymous quantitative survey with 128 11th grade students from two of these settings. We then conducted a qualitative exploration of the quantitative results with a subset of students. Next, a set of Expert Panel members participated in qualitative interviews to reflect on and interpret study findings. Overall, we found students experienced high levels of chronic stress, particularly in relation to academic performance and the college admissions process. While students described a range of effective, adaptive coping strategies, they also commonly internalized these serious pressures and turned to alcohol and drugs to cope with chronic stress, although not typically at problematic levels. We discuss study implications for both schools and families derived from the Expert Panel.

Breast cancer-related lymphedema is a syndrome of abnormal swelling coupled with multiple symptoms resulting from obstruction or disruption of the lymphatic system associated with cancer treatment. Research has demonstrated that with increased number of symptoms reported, breast cancer survivors' limb volume increased. Lymphedema symptoms in the affected limb may indicate a latent stage of lymphedema in which changes cannot be detected by objective measures. The latent stage of lymphedema may exist months or years before overt swelling occurs. Symptom report may play an important role in detecting lymphedema in clinical practice. The purposes of this study were to: 1) examine the validity, sensitivity, and specificity of symptoms for detecting breast cancer-related lymphedema and 2) determine the best clinical cutoff point for the count of symptoms that maximized the sum of sensitivity and specificity. Data were collected from 250 women, including healthy female adults, breast cancer survivors with lymphedema, and those at risk for lymphedema. Lymphedema symptoms were assessed using a reliable and valid instrument. Validity, sensitivity, and specificity were evaluated using logistic regression, analysis of variance, and areas under receiver operating characteristic curves. Count of lymphedema symptoms was able to differentiate healthy adults from breast cancer survivors with lymphedema and those at risk for lymphedema. A diagnostic cutoff of three symptoms discriminated breast cancer survivors with lymphedema from healthy women with a sensitivity of 94% and a specificity of 97% (area under the curve =0.98). A diagnostic cutoff of nine symptoms discriminated at-risk survivors from survivors with lymphedema with a sensitivity of 64% and a specificity of 80% (area under the curve =0.72). In the absence of objective measurements capable of detecting latent stages of lymphedema, count of symptoms may be a cost-effective initial screening tool for detecting lymphedema.

African American/Black and Hispanic persons living with HIV/AIDS ("AABH-PLHA") are under-represented in HIV/AIDS medical studies (HAMS). This paper evaluates the efficacy of a social/behavioral intervention to increase rates of screening for and enrollment into HAMS in these populations. Participants (N = 540) were enrolled into a cluster randomized controlled trial of an intervention designed to overcome multi-level barriers to HAMS. Primary endpoints were rates of screening for and enrollment into therapeutic/treatment-oriented and observational studies. Intervention arm participants were 30 times more likely to be screened than controls (49.3 % vs. 3.7 %; p < .001). Half (55.5 %) of those screened were eligible for HAMS, primarily observational studies. Nine out of ten found eligible enrolled (91.7 %), almost all into observational studies (95.2 %), compared to no enrollments among controls. Achieving appropriate representation of AABH-PLHA in HAMS necessitates modification of study inclusion criteria to increase the proportion found eligible for therapeutic HAMS, in addition to social/behavioral interventions.

Abstract Objectives: There is debate about whether marijuana (cannabis) use is more dangerous than alcohol use. Although difficult to make objective comparisons, research is needed to compare relative dangers in order to help inform preventive efforts and policy. Methods: Data were analyzed from a nationally representative sample of high school seniors in the Monitoring the Future study (2007-2011; Weighted n = 7437; modal age: 18) who reported lifetime use of alcohol or marijuana. Students were asked to indicate whether they experienced various adverse psychosocial outcomes resulting from use of each substance. We examined which outcomes were more prevalent for each substance. Results: Compared to alcohol use, marijuana use was more commonly reported to compromise relationships with teachers or supervisors, result in less energy or interest, and result in lower school or job performance. Compared to marijuana use, alcohol was more commonly reported to compromise relationships with friends and significant others; it was also reported to lead to more regret (particularly among females), and driving unsafely. Marijuana users were more likely to report no adverse outcomes. Females and white students were more likely to report various adverse outcomes and higher frequency use of each substance also increased occurrences of reported adverse outcomes. Conclusions: Marijuana and alcohol are associated with unique adverse psychosocial outcomes. Outcomes differ by sex and race/ethnicity, and perception or experience of outcomes may also be related to legal status and associated stigma. Public health interventions may be more effective by focusing on harm reduction strategies for these drug-specific outcomes.

Abstract Intravaginal practices (IVP) are common among African women and are associated with HIV acquisition. A behavioral intervention to reduce IVP is a potential new HIV risk-reduction strategy. Fifty-eight HIV-1-uninfected Kenyan women reporting IVP and 42 women who denied IVP were followed for 3 months. Women using IVP attended a skill-building, theory-based group intervention occurring weekly for 3 weeks to encourage IVP cessation. Vaginal swabs at each visit were used to detect yeast, to detect bacterial vaginosis, and to characterize the vaginal microbiota. Intravaginal insertion of soapy water (59%) and lemon juice (45%) was most common among 58 IVP women. The group-counseling intervention led to a decrease in IVP from 95% (54/58) at baseline to 0% (0/39) at month 3 (p=0.001). After 3 months of cessation, there was a reduction in yeast on vaginal wet preparation (22% to 7%, p=0.011). Women in the IVP group were more likely to have a Lactobacillus iners-dominated vaginal microbiota at baseline compared to controls [odds ratio (OR), 6.4, p=0.006] without significant change in the microbiota after IVP cessation. The group counseling intervention was effective in reducing IVP for 3 months. Reducing IVP may be important in itself, as well as to support effective use of vaginal microbicides, to prevent HIV acquisition.