An information systems model of the determinants of electronic health record use.

OBJECTIVES: The prominence given to universal implementation of electronic health record (EHR) systems in U.S. health care reform, underscores the importance of devising reliable measures of factors that predict medical care providers' use of EHRs. This paper presents an easily administered provider survey instrument that includes measures corresponding to core dimensions of DeLone and McClean's (D & M) model of information system success. METHODS: Study data came from self-administered surveys completed by 460 primary care providers, who had recently begun using an EHR. RESULTS: Based upon assessment of psychometric properties of survey items, a revised D&M causal model was formulated that included four measures of the determinants of EHR use (system quality, IT support, ease of use, user satisfaction) and five indicators of provider beliefs about the impact on an individual's clinical practice. A structural equation model was estimated that demonstrated a high level of inter-correlation between the four scales measuring determinants of EHR use. All four variables had positive association with each of the five individual impact measures. Consistent with our revised D&M model, the association of system quality and IT support with the individual impact measures was entirely mediated by ease of use and user satisfaction. CONCLUSIONS: Survey research provides important insights into provider experiences with EHR. Additional studies are in progress to investigate how the variables constructed for this study are related to direct measures of EHR use.

How state counter-industry campaigns help prime perceptions of tobacco industry practices to promote reductions in youth smoking.

OBJECTIVE: This study assessed the impact of state media campaigns that prominently feature counter-industry messages on youth cigarette smoking, beyond the effects of price, secular trends, tobacco control efforts, and the national truth campaign. METHODS: Rates of youth smoking were compared in three groups of states: (1) those with long funded counter-industry campaigns (California, Florida, and Massachusetts); (2) states with more recently funded counter-industry media campaigns (Indiana, Minnesota, Mississippi, and New Jersey); and (3) other states. An analysis was performed for a series of national telephone surveys of 12-17 year olds between 1999 and 2002, controlling for differences in demographic background, the price of cigarettes, and exposure to the national truth campaign. RESULTS: Between 1999 and 2002, rates of current smoking and established smoking decreased significantly faster in states with established or more newly funded counter-industry campaigns than in other states. State counter-industry campaigns appear to prime, or make more salient, negative perceptions about tobacco industry practices. CONCLUSION: Results highlight the value of continued state counter-industry campaigns.

The impact of ancillary HIV services on engagement in medical care in New York City.

The advent of antiretroviral therapies in 1996 prompted an interest in the role played by ancillary services in improving access to and retention in medical care, particularly since the success of the new therapies is often contingent upon ongoing and appropriate primary medical care. Using self-reported survey data from a longitudinal representative sample of 577 HIV-positive adults in New York City, this paper explores the impact of such supportive services as drug treatment, case management, housing assistance, mental health treatment and transportation on engagement with medical care. The study's principal finding was that specific ancillary services were significantly associated with an increase in an individual's likelihood of entering medical care and maintaining appropriate medical care services for HIV, particularly when the services addressed a corresponding need.

Smoking among Asian American and Hawaiian/Pacific Islander youth: data from the 2000 National Youth Tobacco Survey.

OBJECTIVES: One of the first steps to reducing the disproportionate burden of tobacco on racial and ethnic minorities is to understand how tobacco differentially affects these populations. This paper, based on a nationally representative sample of Asian American youth and a smaller sample of Hawaiian/Pacific Islander youth, provides the tobacco control community with important information about the smoking behavior of these youth. METHODS: The National Youth Tobacco Survey conducted during the Spring of 2000 (NYTS 2000) provides recent national estimates of smoking behavior among Asian American youth. The data also permit a limited exploration of possible differences between self-described Asians and Hawaiian/Pacific Islander youth, two groups that have been combined into a single racial/ethnic category in earlier national studies. FINDINGS: This report provides estimates and 95 percent confidence intervals for current smoking, age of smoking initiation, use of menthol cigarettes and tobacco brand preferences. CONCLUSIONS: NYTS 2000 data indicate that during the last year of high school, one third of Asian American youth are smokers. Of these youth, 60% report that their usual brand of cigarettes is a menthol brand. Among female Hawaiian/Pacific Islander youth in middle school, more than 25% report having smoked during the past month.

Life stage, substance use and health decline in a community cohort of urban African Americans.

This longitudinal study analyzed chronic health effects, predominantly somatic, resulting from substance use in a homogeneously urban African American community cohort (n = 411). Individual and collective effects of eight licit and illicit substances on change in perceived health (index of reported health conditions formulated as regressed change scores) were tested as the cohort matured into young adulthood. Ten additional measures of health behavior, social attainment and social networks were introduced to control against spurious causal attribution. Gender differences in direct and intercorrelated drug effects and their timing processes were observed. Substance use explained a larger share of the variance in men's health than in women's. In absolute numbers of health problems, however, substance use had a greater effect on women's health.

Safer sex strategies for women: the hierarchical model in methadone treatment clinics.

Women clients of a methadone maintenance treatment clinic were targeted for an intervention aimed to reduce unsafe sex. The hierarchical model was the basis of the single intervention session, tested among 63 volunteers. This model requires the educator to discuss and demonstrate a full range of barriers that women might use for protection, ranking these in the order of their known efficacy. The model stresses that no one should go without protection. Two objections, both untested, have been voiced against the model. One is that, because of its complexity, women will have difficulty comprehending the message. The second is that, by demonstrating alternative strategies to the male condom, the educator is offering women a way out from persisting with the male condom, so that instead they will use an easier, but less effective, method of protection. The present research aimed at testing both objections in a high-risk and disadvantaged group of women. By comparing before and after performance on a knowledge test, it was established that, at least among these women, the complex message was well understood. By comparing baseline and follow-up reports of barriers used by sexually active women before and after intervention, a reduction in reports of unsafe sexual encounters was demonstrated. The reduction could be attributed directly to adoption of the female condom. Although some women who had used male condoms previously adopted the female condom, most of those who did so had not used the male condom previously. Since neither theoretical objection to the hierarchical model is sustained in this population, fresh weight is given to emphasizing choice of barriers, especially to women who are at high risk and relatively disempowered. As experience with the female condom grows and its unfamiliarity decreases, it would seem appropriate to encourage women who do not succeed with the male condom to try to use the female condom, over which they have more control.

Factors influencing participation in weekly support groups among women completing an HIV/STD intervention program.

Over the past three decades, the influence and importance of social support has been well documented and the findings have suggested a beneficial effect on stress-related situations, mental and physical health, and social functioning. More recently, small group/skills training behavioral interventions have demonstrated success in changing behaviors which affect the transmission of sexually transmitted diseases, including HIV among populations at risk for these diseases. Studies of support groups to date have been conducted exclusively in research settings where women are offered financial incentives for participation. Little is known about the willingness of women to participate in ongoing support groups after successfully completing a skills training intervention. The present study examines the factors that may influence participation among women in a weekly support group after completing a structured, six session HIV/STD intervention. Both quantitative and qualitative data are collected from 265 women in the intervention arm of a multi-site randomized controlled behavioral intervention trial. Results reveal that less than a quarter (22%) of women participated in at least one support group. Participation varied significantly by site, ranging from 34% to 15% (p = .008). Participation was also strongly linked to recent use of domestic violence services. Qualitative data indicated that although monetary incentives play some role in the woman's decision to participate, other factors are also important. These include program outreach, support group size, salience of the group content, consistency of group leadership from the intervention to the support group, and use of peer leaders along with professional facilitators. Implications for design of post-intervention support groups programs are discussed.

Effects of ZDV-based patient education on intentions toward ZDV use, HIV testing and reproduction among a US cohort of women.

This study examined the immediate effects of exposure to a patient education brochure concerning the risks and benefits of zidovudine (ZDV) therapy during pregnancy to reduce perinatal HIV transmission (protocol ACTG 076) on related knowledge, behavioural intentions and attitudes of women with and at-risk for HIV-infection. Self-reports were collected from 653 women of childbearing age from community family planning clinics and hospital-based HIV centres in 19 sites from nine US cities between May and November 1995. The intervention was a nine-page patient education brochure in Spanish, Creole and English versions, evently presenting the pros and cons of ZDV therapy to reduce perinatal HIV-transmission. Brochure exposure increased knowledge (p < 0.001) for all but one scale concerning ZDV resistance and increased the likelihood of women reporting intentions to take ZDV during pregnancy (p < 0.001) and to believe ZDV reduced transmission (p < 0.001). Brochure exposure had differential effects for some subpopulations. Intentions to have or terminate current or future pregnancies, knowledge about ZDV and attitudes toward ZDV varied mostly by ethnicity/race, language preference and HIV status. Pregnancy status, age, education and having an HIV-positive child had less impact on the brochure's effect, while income had no impact.

The application of meta-analysis in assessing racial differences in the effects of antihypertensive medication.

Meta-analysis is an important technique for synthesizing research findings. Although the statistical foundations of meta-analysis continue to be debated, few question its value as a rigorous framework for organizing literature reviews. In recent years, there has been increasing emphasis on the use of meta-analysis not only to summarize the central tendency of findings but also to explain variation between studies. This article reviews the major steps in a meta-analysis with an emphasis on comparative analysis of subgroups of studies. A meta-analysis of the antihypertensive efficacy of calcium channel blockers is used to illustrate how a comparative analysis can be applied to investigate racial variation in the effects of calcium channel blockers. A statistically significant trend is found between the proportion of African-American hypertensive subjects and the mean reduction in blood pressure. Meta-analytic techniques also are applied to explore possible confounders due to differences in research design and patient characteristics.

Delivering HIV/AIDS services: the professional care provider speaks out.

BACKGROUND: This study evaluates the stresses and satisfactions experienced by health care and social service providers working in HIV/AIDS service agencies in New York City. This study was part of the Ryan White Title I Evaluation in New York City. METHODS: This study is based on semi-structured interviews with 86 randomly sampled providers from a representative sample of 29 HIV/AIDS service agencies. Personal interviews were completed with a cross section of AIDS care providers. All staff interviewed were audiotaped to facilitate data analysis. Staff discussed their frustrations and their personal satisfaction at working in AIDS care. In addition, all staff completed the Maslach Burnout Inventory (MBI) to facilitate a structured comparison of their levels of burnout. RESULTS: Using the three subscales of the MBI, we found that interviewed AIDS care providers experienced lower than expected levels of burnout. Compared to national norms, health care and social service providers showed above-average levels of personal accomplishment, below-average levels of depersonalization, and average levels of emotional exhaustion. Interview transcripts were analyzed focusing on three broad themes: unique stressors of HIV/AIDS services, positive aspects of HIV/AIDS services, and effective provider supports. The study confirms that HIV/AIDS care providers feel a high level of personal commitment to working with HIV-positive clients. Personal commitment to HIV-positive clients may blunt some of the stresses associated with HIV/AIDS care.

Stabilizing the HIV/AIDS workforce: lessons from the New York City experience.

The Ryan White Title I Personnel Needs Study described here is an evaluation of the human resource needs of HIV/AIDS service agencies in New York City. The research presented here was conducted in collaboration with researchers from the Medical and Health Research Association and the New York City Department of Health and the Planning and Evaluation Committee of the New York City HIV Health and Human Services Planning Council. The assessment is divided into two components. The first component is a survey of 100 key informants from HIV/ AIDS service agencies. The second component is a detailed staffing survey of 70 personnel directors of HIV/AIDS service agencies. HIV/AIDS service directors perceived staff recruitment as a more difficult process than staff retention, regardless of agency type. Vacancy rates at the surveyed agencies varied by professional category; they were especially high in the category of nurse practitioners (13%) and outreach/education workers (15%). Agencies stressed that incentives that would positively affect recruitment and retention should be tailored to address the concerns of varied health care and social service professionals. In addition, agencies were not routinely able to provide some incentives deemed effective such as higher salaries, housing subsidies, and smaller caseloads. To reduce recruitment and retention problems, agency informants recommended a variety of incentive programs including malpractice insurance for physicians, flexible hours for full-time employees (including case managers, nurses, physician's assistants), smaller caseloads, and a decrease in the proportion of staff time devoted to direct client contact. Overall health care trends including truncated federal budgets, state Medicaid cutbacks, and the rapid conversion to managed care all affect the quality of patient care and of the work setting for health care and social service providers serving persons with AIDS. Medical Subject Headings (MeSH): AIDS, health personnel, employment supported, employee workload, staff attitude.

A balancing act: the tension between case-finding and primary prevention strategies in New York State's voluntary HIV counseling and testing program in women's health care settings.

This study sought (1) to identify factors that influence women's willingness to accept voluntary HIV counseling and testing at New York State Family Planning Programs (FPPs) and Prenatal Care Assistance Programs (PCAPs) and (2) to evaluate the effectiveness of such a voluntary counseling and testing program. Telephone interviews elicited organizational-level data from 136 agencies; a combination of telephone and face-to-face interviews was used to gather provider data from 98 HIV counselors; and client data were gathered from 354 women in face-to-face interviews at counseling sites. Slightly fewer than 60% of women agreed to be counseled, and, of those, under half consented to an HIV test at the counseling site. Approximately two thirds of the women who were tested returned for their results and posttest counseling. Clients' recall of pretest counseling content was relatively poor. Bivariate and regression analyses suggest that client, provider, and organizational factors are all associated with rates of pretest counseling and testing. The current voluntary counseling and testing program is achieving only moderate success. Although a substantial number of clients accept HIV counseling, many women remain reluctant to consent to HIV testing, and many who accept testing do not return for their results. Moreover, among those who receive pretest counseling, many do not recall important informational content, which suggests variation may exist in the quality of counseling or that one-time HIV counseling interventions are insufficient to communicate complex information. Medical Subject Headings (MeSH): AIDS, HIV serodiagnosis, women's health, patient education.

Choosing optimal support groups: a review and reformulation.

There is considerable empirical research on the division of supportive and caregiving services among different network members. Yet the principles that govern the configuration of social support networks are still not well understood. Cantor's (1979) and Shanas' (1979) hierarchical-compensatory model and Litwak's (1985) task-specific model are among the more prominent formulations for explaining the choice of support groups. We propose that the former model is a special case of the latter one. The major conceptual features of both models are summarized here, followed by a qualitative review of 16 key empirical studies coupled with a meta-analysis to demonstrate the broader application of the task-specific model.

The effect of video interventions on improving knowledge and treatment compliance in the sexually transmitted disease clinic setting. Lesson for HIV health education.

Eight intervention studies investigating patient education and treatment adherence in the sexually transmitted disease (STD) clinic setting are reviewed. Across the eight studies selected for analysis, meta-analytic procedures were applied to compare the impact of educational and prevention approaches. The effect of video was compared with that of other modes of health education on: 1) knowledge and attitudes about STDs and condoms and 2) treatment compliance, as measured by return for test of cure, drug compliance, premature resumption of sexual activity, and condom coupon redemption rates. The largest effects were those for video on knowledge and attitudes about STDs and condoms, followed by the effects of other non-video interventions on STD knowledge. Lower effects were found among video and non-video interventions targeting treatment compliance outcomes. These results are consistent with prior prevention studies that have demonstrated difficulty in achieving behavior change.

A public health model for cardiovascular risk reduction. Impact of cholesterol screening with brief nonphysician counseling.

To examine the impact of cholesterol screening with brief nonphysician counseling on cardiovascular risk factors, 886 employees at a large financial services firm underwent fingerstick screening followed by brief (3- to 5-minute) counseling by a registered nurse. At the 6-month follow-up, there were significant declines in total cholesterol levels (5.9 to 5.5 mmol/L [228 to 213 mg/dL]), weight (78 to 75 kg), blood pressure (119/78 to 115/75 mm Hg), and number of people reporting smoking (16.8% to 14.5%) among participants with a baseline cholesterol value of 5.2 mmol/L (200 mg/dL) or greater. A randomized experiment was conducted on 137 participants initially classified as having a "borderline-high blood cholesterol" level (5.2 to 6.2 mmol/L [200 to 239 mg/dL]) to test the impact of frequency of follow-up. Those receiving frequent follow-up (cholesterol measurement and brief counseling at 2, 4, and 6 months) reported significantly greater dietary change and demonstrated a trend toward greater declines in total cholesterol compared with those receiving follow-up at 6 months only. The results of this study support the feasibility and efficacy of cholesterol screening utilizing brief nonphysician counseling on multiple cardiovascular risk factors and suggest an enhanced effect when patients receive more frequent follow-up.

Differences in physician prevention practice patterns for white and minority patients.

A telephone survey of 120 randomly selected primary care physicians in New York City was completed in October, 1984 (response rate = 90%) concerning physicians' recommendations for health promotion and disease prevention. Responses from physicians with 50% or more Black and Hispanic patients were compared with responses from physicians with 50% or more White patients. The former were found to be less likely to follow guidelines from nationally recognized organizations for health promotion and disease prevention, although they were just as likely to value the importance of prevention in primary care. For example, physicians with predominantly Black and Hispanic patient populations were significantly less likely to recommend screening mammography (7% versus 23%) or recommend influenza vaccination for patients 65 or older (48% versus 74%) when compared with physicians with predominantly White patient populations. Factors that appeared to contribute to the difference in prevention practice patterns include physician training and education, the socioeconomic status of the patients, and the time physicians spend with patients. Differences in quality of preventive care provided to minority patients may be an additional factor in the disparity between the health status of White and non-White Americans.

Heroin availability and aggregate levels of use: secular trends in an urban black cohort.

The influence of heroin availability on the aggregate level of use of this drug was investigated for a normal Black cohort (born between 1952 and 1957) who grew up in Harlem (New York City). Data obtained on the second and third waves of a panel study were used to estimate annual rates of heroin initiation and cessation from the mid-1960s through 1983. The aggregated time-series variables indicated that initiation into heroin use was largely confined to adolescence and that cessation rates exhibited substantial year-to-year fluctuations with no apparent relationship to either chronological age or calendar year. Respondents born before 1955, however, had much higher rates of heroin use than those born in later years. Temporal trends in initiation and cessation were uncorrelated with changes in the purity of heroin sold in New York City between 1973 and 1983, suggesting that aggregate levels of heroin use in this sample were little affected by changes in supply. More speculatively, cohort differences in lifetime prevalence may reflect varying availability at the times younger and older cohorts entered adolescence. This possibility could not be directly tested because of the absence of reliable purity data going back sufficiently far in time.

Drugs, lifestyle, and health: a longitudinal study of urban black youth.

The association of drug use with indicators of health status was studied in a representative sample of urban Black youth interviewed first in adolescence and then six to seven later in young adulthood. Seven substances, with interaction terms as appropriate, were tested against a psychophysical health scale and also against a more inclusive general health scale. Both measures of health outcomes yielded similar findings. Earlier inhalants use and, under certain conditions, methadone were linked to health change in males. Heroin had a delayed relationship to females' health. Additionally, methadone was associated with poorer health among young Black women who also were heavy alcohol drinkers. The observed relationships between drug use and health persisted after introducing controls for associated lifestyle factors.