John D. Rogers

A 10-year national trend study of alcohol consumption, 1984-1995: is the period of declining drinking over?

OBJECTIVES Data from the 1984, 1990, and 1995 National Alcohol Surveys were used to investigate whether declines shown previously in drinking and heavy drinking across many demographic subgroups have continued. METHODS Three alcohol consumption indicators--current drinking (vs abstaining), weekly drinking, and weekly heavy drinking (5 or more drinks in a day)--were assessed for the total US population and for demographic subgroups. RESULTS Rates of current drinking, weekly drinking, and frequent heavy drinking, previously reported to have decreased between the 1984 and 1990 surveys, remained unchanged between 1990 and 1995. Separate analyses for each beverage type (beer, wine, and spirits) and most demographic subgroups revealed similar temporal patterns. CONCLUSIONS Alcohol consumption levels, declining since the early 1980s, may reach a minimum by the 21st century. Consumption levels should be monitored carefully over the next few years in the event that long-term alcohol consumption trends may be shifting.

Face-to-Face versus Telephone Interviews: Using Cognitive Methods to Assess Alcohol Survey Questions:

This research compares retrieval strategies and comprehension problems encountered by 30 heavier drinking respondents using protocol analysis on three alcohol use items: In the last 12 months, maximum quantity consumed in one day, frequency of drinking only one to two drinks, and frequency of drunkenness. We compare two modes of administering the questionnaire: face-to-face and telephone interviews. Results indicate that respondents in both modes used similar retrieval strategies. Anchoring (giving a response immediately) and context (describing a context in or around which drinking occurs) were most often used by respondents. Comprehension problems varied by question rather than by interview mode. The maximum quantity question was more likely to invoke issues around time frame; the one to two drinks item elicited problems concerning misinterpretation (“only” versus “at least”), and the drunkenness item was more likely to stimulate questions of definition. Overall, the findings from this study lend some support for moving from face-to-face surveys to telephone surveys in the next U.S. national alcohol survey and provide some direction for strengthening the transition.

HIV Risk Behaviors Among Three Classes of Female Sex Workers in Vietnam

Increases in the rate of HIV infection in Vietnam among female sex workers and their clients require more effective preventive interventions based on a better understanding of this population and important subgroups within it. Because little was known about women in the hospitality sex industry, this study compared demographic and work characteristics, history of sexually transmitted infections (STIs), and condom use among a sample of 310 female sex workers in low-, middle-, and high-class hospitality businesses in Ho Chi Minh City (mean age = 26.77 years, SD = 6.17). Data were collected through individual, face-to-face, semistructured interviews. Logistic regression models were used to identify demographic predictors of the incidence of STIs and reported condom use. Low occupational status was most strongly associated with a history of STIs, and use of condoms with new and regular clients also differed by occupational class. Implications are discussed.

Health Effects from Drinking: Type, Severity, and Associated Drinking Patterns Based on Qualitative and Quantitative Questions in a Methodological Survey

The specific content of reported health harms are explored in relation to heavy episodic drinking and alcohol dependence symptoms. A national telephone computer-assisted interview of U.S. adults (N = 635), was conducted in 1994. A modified National Alcohol Survey instrument assessed self-reported health harms and related economic costs and work missed. Among the 579 lifetime drinkers, 26% reported health harms from drinking. In multivariate logistic regression harms relating to intoxication, injuries or internal organs were associated with monthly heavy drinking (5+ drinks) and past dependence symptoms. Study limitations are noted. Dependence symptoms should be assessed in screening for those most at risk for alcohol-related health harms.

Factors Associated with the Integrated Educational Placement of Students with Severe Disabilities

A study, using survey methodology, was conducted to identify factors associated with the integrated educational placement of students with severe disabilities. Questionnaires were completed by the families and teachers of 1,009 students from five participating states. The questionnaires were designed to measure 19 variables that integration literature suggests are predictive of integrated placement. Logistical regression procedures were utilized to determine the probability of integrated placement as a function of the identified variables, which included characteristics of the student, family, school program, and community. A stepwise procedure was employed to examine three theoretical models of variables associated with integrated placement, representing integration advocacy, socioeconomic status, and program/facilities characteristics. Additionally, a combined model was estimated using the strongest predictors from the three theoretical models. Factors identified within each of the models are discussed in terms of their implications for increasing the integrated placement of students with severe disabilities.

Are Estimates of the Concentration of Alcohol Consumption Affected by Undercoverage? Evidence from Five Pooled U.S. Surveys

Conclusions concerning the properties of alcohol consumption distributions are often based on surveys that underestimate the amount consumed by sampled populations. Relationships between underreporting of alcohol consumption and the concentration of heavy drinking could present problems for several important research findings. Using data from five US telephone surveys, we compared state-level survey estimates of alcohol consumption and concentration with published estimates based on sales and taxes. Demographic factors predicting undercoverage were assessed in a multilevel analysis. Concentration was not significantly related to coverage, but it was related to rates of any drinking. State-level undercoverage was predicted by the proportion of males, younger age, political conservatism, Hispanic ethnicity, and the proportion of ethanol sold as spirits. Additional research is needed to determine the consequences ofthe relationships between these factors and undercoverage, but national estimates ofconsumption distribution properties are unlikely to be affected greatly by low coverage rates.