| | Increasing the appeal and utilization of services for alcohol and drug problems: What consumers and their social networks preferReceived 26 March 2007; received in revised form 31 October 2007; accepted 12 November 2007. published online 03 January 2008. Abstract BackgroundA large gap exists in the United States between population need and the utilization of treatment services for substance-related problems. Surveying consumer preferences may provide valuable information for developing more attractive services with greater reach and impact on population health. MethodsA state-level telephone survey using random digit dialling sampling methods assessed preferences for available professional, mutual help, and lay resources, as well as innovative computerized and self-help resources that enhance anonymity (N =439 households in Alabama). ResultsRespondents preferred help that involved personal contact compared to computerized help or self-help, but were indifferent whether personalized help was dispensed by professional or lay providers. Attractive service features included lower cost, insurance coverage, confidentiality, rapid and convenient appointments, and addressing functional problems and risks of substance misuse. Respondents in households with a member who misused substances rated services more negatively, especially if services had been used. ConclusionThe findings highlight the utility of viewing substance misusers and their social networks as consumers, and the implications for improving the system of care and for designing and marketing services that are responsive to user preferences are discussed. A large gap exists in the United States between population need and service utilization for substance use disorders, which affect about 10% of the population each year (Grant et al., 2004; Narrow, Regier, Rae, Mandersched, & Locke, 1993; Substance Abuse and Mental Health Administration (SAMHSA), 2005; Wang et al., 2005). For example, the National Comorbidity Survey Replication (Wang et al., 2005) indicated that less than 4 in 10 persons with alcohol or drug-related problems received help from any source during the past year, and fewer still received minimally adequate treatment from a qualified provider. Although rates of help-seeking have increased over time and are higher when help from informal sources (e.g., family, friends, and clergy) is assessed in addition to treatment and mutual help groups (Narrow et al., 1993), increasing the utilization of services remains a public health priority, including as an HIV/AIDS and hepatitis prevention strategy. This is especially relevant in the Southern United States, where HIV rates are increasing, particularly among people of colour (U.S. Centers for Disease Control & Prevention (CDC), 2004). Despite this public health need, stakeholders ranging from drug policy makers, treatment providers, and substance users agree that abstinence-based clinical treatments often are difficult to access and can be unappealing, even if effective in reducing substance use and related problems (Appel, Ellison, Jansky, & Oldak, 2004; Fountain, Strang, Griffiths, Powis, & Gossip, 2000; French, Homer, & Nielsen, 2006). In response, the international harm reduction movement has promoted expanding the continuum of care for substance misuse across community and clinical settings and has emphasized making services available on demand without barriers such as long waiting times or sobriety requirements for entry and retention (MacCoun & Reuter, 2001). Consumer preferences and needs have been considered in making these changes, in addition to evidence-based practices. The changes have had salutary effects on HIV/AIDS rates in countries that have adopted them, and they have helped reach more of the population with problems, including persons with milder problems who nevertheless contribute the bulk of harm and cost (Humphreys & Tucker, 2002). This consumer-responsive approach to treatment services is less well developed in the United States, which historically has pursued a criminal justice approach to drug control and continues to emphasize abstinence-oriented treatment and mutual help group involvement. Persons with problems are reluctant to seek help until forced to do so or until their problems become severe (Cunningham, Sobell, Sobell, Agrawal, & Toneatto, 1993). Although mutual help groups like Alcoholics Anonymous (AA) are widely available and do not require entry into the health care system, these groups are not appealing to everyone, and similar “low threshold” professional alternatives are generally not available. Persons with problems also use costly medical services that do not address their substance-related problems (Fortney, Booth, & Curran, 1999), or they attempt to solve their problems on their own. Such “natural” resolutions are more common than intervention-assisted resolutions (Sobell, Cunningham, & Sobell, 1996). In order to develop more attractive services with greater reach and impact, research is needed on public perceptions about available treatments, barriers to care, and consumer needs and preferences. Developing and marketing services that balance consumer interests with evidence-based practices is not a new idea in other areas of health promotion and disease prevention that have adopted a “social marketing” approach to behaviour change programs (Walsh, Rudd, Moeykens, & Mokoney, 1993). However, such consumer-centred approaches are uncommon in alcohol and drug treatment programs in the United States, other than smoking cessation programs. The scope of such consumer research should be broad and not limited only to persons with substance-related problems. Although they are the end point consumers of services, members of their social networks and other stakeholders (e.g., court system and employers) often influence patterns of care-seeking (Tucker & King, 1999), so their needs and preferences matter as well. Anecdotal reports about barriers to care and ways to make services more appealing are abundant in the natural recovery and harm reduction literatures (Tucker & King, 1999), but public views of alcohol and drug problems, how problems are resolved, and preferences for helping resources have not been well investigated. As a starting point, survey methods offer a way to investigate consumer preferences for existing services compared to innovative alternatives. This has the advantage of not prejudging which innovations may be attractive to consumers and their social networks and may help avoid costly mistakes in funding new programs that prove to be unappealing. The few surveys on services for substance use and mental health disorders have shown that consumer preferences do not always track findings about treatment efficacy; e.g., the Consumer Reports mental health survey of about 7000 readers (Seligman, 1995) showed that AA was rated as a preferred helping resource for drinking problems that resulted in greater improvement compared to treatment by mental health professionals. Other studies of treatment satisfaction have suggested that clients with substance-related problems value components related to treatment quality (e.g., counsellor skill, time with counsellor), privacy, and convenience (e.g., parking and location) (Fountain et al., 2000; Rohrer & Hilsenrath, 1999). Recent research on the appeal and use of telehealth services for alcohol and drug problems has yielded encouraging, but somewhat mixed findings (e.g., Copeland & Martin, 2004; Cunningham, Humphreys, Koski-Jannes, & Cordingly, 2005); e.g., computer-literate younger adults liked computerized options, but internet resources alone were insufficient to promote change without additional self-help materials (Cunningham et al., 2005). To investigate how conventional treatments for alcohol and drug problems could be supplemented with telehealth and self-help options to expand the continuum of care, the present study surveyed consumer preferences for a range of services using a state-wide sample of households in Alabama, a Deep South state with a rising HIV/AIDS rate. The telephone survey assessed preferences for professional treatments and mutual help groups, as well as lower threshold services that do not require entry into the health care system and are more private, anonymous, and less costly. Preferences for service features related and unrelated to substance use were assessed. Households that did and did not report having a member with a current or past alcohol or drug problem were surveyed to ascertain how such experiences affected preferences for care. Five hypotheses were advanced based on research on help-seeking: (1) preferences for services for alcohol and drug problems should be influenced by generic service features known to influence health care utilization (e.g., cost, insurance, and access). (2) Consistent with established incentives for help-seeking for substance-related problems, service features that involve addressing problems and risks associated with substance use should be highly preferred. (3) Preferences for treatment from medical or mental health professionals should be relatively greater for serious alcohol and drug problems compared to mild to moderate problems. (4) Respondent experiences with substance misuse and its treatment, either by oneself or a household member, should be associated with a more negative view of services. (5) Certain services and features should be of greater importance to select subgroups; e.g., computer help should appeal more to younger, educated respondents with higher computer use, and women should care more about childcare during appointments. Method  Sample recruitment The study was approved by the university Institutional Review Board. A sample of residential telephone numbers was surveyed between September 2002 and August 2003 using random digit dialling (RDD) methods. A list of randomly selected telephone numbers was purchased from a commercial provider that included listed and unlisted numbers with known unassigned, disconnected, or business phone numbers removed. The RDD method improves survey efficiency by reducing the number of calls made to phone numbers that will not yield an eligible respondent. The person who answered the phone was told the call would involve a 20-min survey of views about services for alcohol and drug problems and that the survey was not about personal experiences with substance misuse or its treatment. Participation was unpaid and voluntary. Respondents aged 19 years or older, the age of majority and legal consent in Alabama, gave verbal consent and were interviewed. The 72-item survey took an average of 24min to complete (S.D.=5.97). Interviews were completed with 439 households. The upper bound survey response rate was 38% (CDC, 2005), which represents the proportion of completed interviews to the number of completes, refusals, and terminations [439/(439+706+19)]. The 95% confidence interval for the survey was ±4.68%, indicating the true population values were estimated to be within ±4.68% of the values obtained in the survey (Creative Research Systems, 2006). Survey questionnaire The 72-item questionnaire was pre-tested for literacy level and was judged to be appropriate for the study population. Survey items concerned with consumer preferences for services and service features are described next and were derived from the literature summarized earlier. Participants made preference ratings using 10-point Likert scales because of the ease such ratings afford for telephone data collection when no written materials are involved. The 10-point scales were employed to promote dispersion (variability) in item response distributions. Additional items about perceived need and utilization of services will be reported elsewhere (Tucker, Foushee, & Simpson, 2007). Service availability and preferences Respondents were given a brief description of each of eight helping resources, and they indicated whether the resource was available in their community (yes, no, and do not know/unfamiliar). Five services involved face-to-face contact; two were professional and required entry into the health care system (formal treatment by medical or mental health professionals), and three did not (informal help from support groups, family and friends, or church-based services). Three services did not involve much personal contact, if any (self-help using books, videotapes, or computers; computerized help from medical or mental health professionals; computerized support groups). Participants rated each service from 1 (“I would not want to use this service if a family member or I had an alcohol or drug problem”) to 10 (“I would want to use this service if a family member or I had an alcohol or drug problem”). For alcohol and drug problems separately, participants indicated their first choice for getting help for a mild to moderate problem and for a substantial to severe problem. Substance-related service features Participants made identical ratings using 1–10 scales of 8 features related to substance abuse (sobriety not required for initial appointment, substance abuse services offered in general medical care setting, HIV testing/counselling available on-site, input into treatment goals, help with work and money problems related to substance abuse, help with interpersonal problems, help managing social situations involving substance use without using, help reducing risks of substance use such as catching HIV or hepatitis, driving under the influence). These item had good internal consistency (ά=.829). Designation of substance-positive households Households were classified as having or not having a member with substance-related problems (SP+ or SP−) based on respondent answers to the question, “Have you or anyone in your household had an alcohol or drug problem?” Follow-up questions to affirmative answers determined if the respondent or another household member had the problem and which substances were involved. Data analyses Because consumer preferences have practical significance for service organizations and providers, analyses were first conducted on the individual ratings of different services and features for the sample as a whole. Because some items had non-normal distributions, nonparametric Friedman tests for k-related samples examined whether preferences varied across services, substance-related features, and generic service features. Differences in service choices (formal or informal) as a function of alcohol and drug problem severity (mild or serious) were evaluated using the McNemar test for correlated proportions. To examine how preferences varied as a function of household and respondent characteristics, individual items in the services, substance-related feature, and generic feature domains were factor analysed to reduce the ratings to a smaller number of factors. Factor analyses were performed using principal component analysis as the extraction method and varimax with Kaiser normalization as the rotation method. Factors with eigenvalues greater than 1.0 were retained and rotated, and items with factor loadings >.45 after rotation were retained as indicators of interpretable factors. Because factor scores are derived from the sums of multiple items, they are less susceptible to any violations of normality (per the central limit theorem) and were examined using repeated-measures analyses of variance (ANOVAs). Raw scores on each significant factor, computed as the average of items that loaded on each factor, were compared in ANOVAs for the sample as a whole and as a function of household substance problem status (SP+/SP−), help-seeking status (HS+/HS−), respondent race (White/Black, excluding the <5% of respondents of other races) and gender (male/female). Results Sample characteristics Like many Deep South states in the United States, Alabama has a higher percentage of African Americans (26.4%), more abstainers from alcohol and illicit drugs (SAMHSA, 2005), and greater religious involvement (Kosmin & Mayer, 2001) compared to the U.S. as a whole. The sample was similar to 2000 U.S. Census data for Alabama, except that it included more women and was more highly educated than the Alabama general adult population; therefore, gender and education were used to weight variables for analysis, which also indirectly adjusted for income. Results were based on the weighted data unless otherwise noted (e.g., for sample description, analysis of gender effects). Table 1 presents the unweighted and weighted respondent demographic characteristics, and Table 2 presents the household characteristics related to substance use and help-seeking. Based on the unweighted data, 26.60% of households had a member with a substance-related problem (SP+ households), of which 36.84% were the respondents themselves (9.63% of the sample). Alcohol was the most frequently misused drug, followed by illicit drugs and misuse of prescription drugs. About two-thirds of SP+ households (65.50%) had sought help for substance misuse, with use of medical services (55.26%) and support groups like AA (47.4%) being most common. | | |  | Characteristic | N | % of sample | |
|---|
| | | Unweighted | Weighted | |
|---|
| Men | 148 | 33.7 | 48.3 | | | Women | 291 | 66.3 | 51.7 | | | White | 302 | 68.8 | 67.6 | | | African American | 117 | 26.7 | 26.7 | | | Other race/ethnicity | 19 | 4.6 | 5.7 | | | | | | Educational level | | | Kindergarten—grade 11 | 46 | 10.5 | 25.3 | | | High school graduate/some college | 230 | 52.5 | 45.5 | | | College graduate or beyond | 162 | 37.0 | 29.2 | | | | | | Annual household income ($) | | | <20,000 | 93 | 24.3 | 27.7 | | | 20,000–40,000 | 104 | 27.2 | 26.8 | | | 40,000–60,000 | 87 | 22.7 | 23.0 | | | 60,000–80,000 | 54 | 14.1 | 11.9 | | | >80,000 | 45 | 11.7 | 10.6 | | | | |
Preferences for helping resource Preference variations across resources As summarized in Table 3, the overall Friedman test was significant for helping resources, χ2=541.37, d.f.=7, p<.001, and follow-up comparisons were conducted using the Wilcoxon signed ranks test. Means in Table 3 with different letters were significantly different from one another (p<.05); means that share a letter did not differ significantly. Help from family and friends was most preferred, even though it was used infrequently, followed by other professional and lay services that involved personal contact. Self-help received intermediate ratings that differed significantly from all other forms of help. Computerized help of any kind received significantly lower ratings than all other types of help. Contrary to expectations, computer preferences were not significantly correlated with age, income, education, or hours spent per week using a personal computer. | | | | Variable | M | S.D. | |
|---|
| Formal and informal helping resources | | | Family and friends | 8.21 a | 2.71 | | | Medical professionals | 7.99 a,b | 2.88 | | | Support groups (e.g., AA and NA) | 7.90 a,b,c | 2.84 | | | Religious leaders/counsellors | 7.82 b,c | 2.90 | | | Mental health professionals | 7.71 c | 2.92 | | | Self-help (e.g. books and videos) | 6.57 d | 3.03 | | | Computerized professional help | 5.31 e | 3.29 | | | Computerized support groups | 5.15 f | 3.29 | | | | | | Service features related to substance abuse | | | Help resolving relationship problems | 8.90 a | 1.97 | | | Reducing risks of use (e.g., HIV and DUI) | 8.88 a | 2.03 | | | Help resolving work/money problems | 8.82 a | 1.95 | | | Help to not use when others use | 8.64 b | 2.34 | | | On-site HIV/Hepatitis C testing/counselling | 8.57 b | 2.36 | | | On-site medical care | 8.35 c | 2.37 | | | Input into substance-related goals | 7.83 d | 2.62 | | | Sobriety not required for initial appointment | 7.69 d | 2.84 | | | | | | Generic service features | | | Service cost covered by insurance | 9.44 a | 1.51 | | | Confidential/privacy protections | 9.41 a | 1.47 | | | Low cost or free services | 9.31 a | 1.66 | | | Quick appointments available | 8.97 b | 1.96 | | | Multiple appointments per week | 8.86 b | 1.84 | | | Evening/weekend appointments | 8.81 b,c | 2.16 | | | Short travel time | 8.62 c | 2.26 | | | Childcare available | 8.02 d | 2.99 | | | Low cost or free parking | 7.89 d | 2.81 | | | | |
Role of problem severity For alcohol problems, most respondents had consistent preferences for either formal (38.44%) or informal (27.46%) help regardless of problem severity. The remainder (34.10%) had preferences that varied significantly with severity, χ2=64.46, d.f.=1, p<.001. Most (83.22%) choose informal help for milder problems and formal treatment for serious problems. For drug problems, relatively more respondents had consistent preferences for formal treatment (55.7%) regardless of problem severity, and fewer had consistent preferences for informal help (19.18%). Among respondents (25.11%) who had preferences that varied significantly with drug problem severity, the great majority (85.71%) preferred medical care for milder problems, and their preferences for serious problems showed no discernable pattern, χ2=45.83, d.f.=1, p<.001. Role of household substance problem and help-seeking status Table 4 presents the factor analysis results for the whole sample and as a function of household substance problem (SP+/SP−) and help-seeking (HS+/HS−) status. Means are based on the average of items loading on significant factors. The factor analysis of helping resource preferences yielded a two-factor solution that accounted for 56.09% of the total variance. One factor reflected help from all five sources of help that involved personal contact. The second factor reflected more anonymous help from computerized sources or self-help. The sample as a whole preferred personalized help over anonymous help, F(1, 400)=292.66, p<.0001. As hypothesized, households with a substance-abusing member rated services lower than those without a member, F(1, 400)=4.03, p=.045, and, among SP+ households, past experiences with substance-related services was associated with lower overall ratings (M=5.95, S.D.=1.73) compared to no prior help-seeking (M=7.48, S.D.=1.67), F(1, 117)=21.00, p<.0001. | | | | Factor | % variance explained | Sample | SP+ | SP− | HS+ | HS− | |
|---|
| | | M | S.D. | M | S.D. | M | S.D. | M | S.D. | M | S.D. | |
|---|
| Helping resources | | | Personalized services (5) | 29.99 | 7.93 a | 2.02 | 7.65 | 2.01 | 8.05 | 2.02 | 7.22 | 1.92 | 8.40 | 1.97 | | | Anonymous services (3) | 26.13 | 5.69 b | 2.61 | 5.35 | 2.54 | 5.83 | 2.63 | 4.67 | 2.37 | 6.57 | 2.39 | | | | | | Service features | | | Substance-specific features (5) | 20.10 | 8.61 b | 1.69 | 8.59 | 1.80 | 8.62 | 1.64 | 8.67 | 1.43 | 8.44 | 2.30 | | | Other services/supports (5) | 16.91 | 8.28 c | 1.95 | 8.32 | 1.90 | 8.27 | 1.98 | 8.44 | 1.64 | 8.11 | 2.29 | | | Cost, convenience, coverage (5) | 15.38 | 9.17 a | 1.23 | 9.28 | 0.96 | 9.13 | 1.33 | 9.26 | 1.01 | 9.31 | 0.88 | | | Service on demand (2) | 9.25 | 8.30 c | 2.07 | 8.29 | 2.24 | 8.31 | 2.01 | 8.73 | 1.57 | 7.54 | 2.93 | | | | |
Respondent gender and race Women had significantly higher overall preferences for help (M=6.89, S.D.=1.93) than men (M=6.43, S.D.=1.96), F(1, 353)=3.96, p=.047, but gender did not influence relative preferences for personal and anonymous help. Although both races preferred personal help over anonymous help, a significant Race×Preference interaction, F(1, 385)=21.33, p<.0001, showed that Whites had relatively stronger preferences for personal (M=8.09, S.D.=1.79) than anonymous (M=5.50, S.D.=2.49) help compared to Blacks. Blacks rated personal (M=7.64, S.D.=2.52) and anonymous (M=6.25, S.D.=2.80) help more similarly. Preferences for substance-related and generic service features As summarized in Table 3, the Friedman tests were significant for substance-related, χ2=165.36, d.f.=7, p<.001, and generic, χ2=334.17, d.f.=8, p<.001, service features. A factor analysis was conducted to evaluate the relative value of substance-related and generic service features (17 items). As shown in Table 4, a four-factor solution was found that accounted for 61.74% of the total variance. In order of the variance explained, Factor 1 included features specific to substance-related problems and goals (e.g., help resolving problems, reducing risks of use). Factor 2 encompassed other services and supports for participating in substance-focused services (e.g., HIV/hepatitis C testing and counselling, on-site medical care, free parking, childcare, and short travel time). Factor 3 reflected generic service features and included convenient, low cost, insured, and confidential services. Factor 4 was a “service on demand” factor with high loadings for quick appointments and no sobriety requirement. A repeated-measures ANOVA showed that the mean preference ratings of items included in the four factors varied significantly for the sample as a whole, F(3, 466)=41.87, p<.0001. As shown in Table 4, pairwise contrasts using the LSD test indicated that the generic service features loading on Factor 3 were preferred significantly more so than all other features. Substance-specific features loading on Factor 1 were rated second highest. They were preferred significantly more than the “service on demand” Factor 4 and the “other services/supports” Factor 3, which did not differ significantly. Identical analyses limited to SP+ and HS+ households generally showed the same pattern of relative preferences. The only exception was that respondents in HS+ households had significantly higher scores on the service on demand factor reflecting quick initial appointments and no sobriety requirement compared to respondents in HS− households, F(3, 408)=4.73, p<.003. Men and women had similar preferences except that women rated service features as generally more important (M=8.73, S.D.=1.15) compared to men (M=8.41, S.D.=1.55), F(1, 407)=5.81, p=.016, and they had higher preferences for childcare (M=8.20, S.D.=2.86) compared to men (M=7.60, S.D.=3.24), F(1, 432)=2.00, p=.046. The preferences of Blacks and Whites followed the pattern for the whole sample, except that Blacks had higher preferences for having other services and supports available (M=8.98, S.D.=1.52) compared to Whites (M=8.05, S.D.=1.94), F(1, 406)=23.44, p<.001. Discussion In today's healthcare marketplace, the efficient and effective use of limited resources requires market research on the needs and preferences of potential consumers, including the extent to which current services satisfy the consumer base (Fountain et al., 2000). The present study is innovative in studying consumer preferences for services for alcohol and drug problems in the general population and in households and individuals directly affected by substance misuse. The scope of assessment proved important because, as hypothesized, respondents in households with a member with problems rated services more negatively, especially if services had been used. Most respondents were aware of services in their area, so preferences did not appear to reflect differences in service availability. The results suggest that consumer dissatisfaction with existing services, more so than lack of availability, impedes utilization. They also point to malleable features of services that mattered to consumers and that can be manipulated to enhance their appeal and use. Regardless of household help-seeking or substance problem status, respondents strongly preferred resources that involved personal contact compared to more anonymous computerized and self-help options, and this pattern was more pronounced among White than Black respondents. The result runs counter to the common notion that consumers prefer more anonymous interventions for stigmatized disorders like alcohol and drug problems. Even “anonymous” self-help groups like AA and NA involve frequent face-to-face contact and sharing personal information in group settings. Respondents of both races, however, were indifferent whether personalized services were dispensed by professional or lay providers. This is reminiscent of the Consumer Reports finding that consumers liked AA at least as much as professional mental health treatment (Seligman, 1995). Although this may be a regional preference in part, it was robust across the sample and cannot be attributed to variations in education, income, gender, or race. This finding has implications for service delivery. First, it supports supplementing professional services with help from trained paraprofessionals and community “lay helpers.” This approach has been used successfully for other chronic health-related problems, including diabetes management (Ingram, Gallegos, & Elenes, 2005), cancer risk reduction and control (Hinton, Downey, Lisovicz, Mayfield-Johnson, & White-Johnson, 2005), and cardiovascular health and weight management (Littleton et al., 2002). Peer outreach services aimed at reduced drug use and HIV prevention have been found to be effective (e.g., Latkin, Sherman, & Knowlton, 2003), and coordinating peer and professional services could increase service contact opportunities, fill a gap in the continuum of care, and increase client satisfaction, while containing costs. Second, personal attention appears to be important to consumers. This suggests the value of testing new telehealth and “self-administered” interventions that include interactive components, such as online interactions with a counsellor or coach. Such strategies may be better received if they include some “face time.” Respondents also highly valued service features that addressed functional problems and risks related to substance use. The appeal of services could be increased by including such features and by expanding treatment goals beyond the usual focus on reducing and eliminating substance use. Generic service features known to influence demand for health care (e.g., low cost, insurance coverage, convenience, and confidentiality) also were highly valued, more so than substance-related features. The latter finding may be due in part to sample characteristics; most respondents did not have alcohol or drug problems, so basic service features would be expected to be of more value to them. While generic features are often available in other areas of health care, they have been less emphasized in U.S. alcohol and drug treatment programs, which often have long waiting lists, abstinence requirements for treatment entry and retention, inconvenient scheduling, and limited insurance coverage (often with low lifetime limits). In addition to suggesting that substance-focused services be positioned within and responsive to these broader health system and economic factors, the results support viewing substance users as potential healthcare consumers who are sensitive to changes in both the direct (e.g., financial and time) and indirect (e.g., stigma and embarrassment) costs of receiving services. The factors that influence consumers of services for alcohol and drug problems appear to be similar to those that influence consumers of other medical care. The few gender differences indicated that men had less positive views of services and placed less value on childcare compared to women. Substance use disorders are more common in men, so their greater dissatisfaction deserves attention. The few observed racial differences may be relevant to understanding health disparities and slowing the spread of HIV in the Southern United States, which has disproportionately affected rural communities of colour. Offering HIV and hepatitis testing and counselling and on-site medical care as part of treatment for substance misuse appears important to this consumer group. Problem severity influenced service preferences among some, but not all respondents. Over 25% indicated that, if they had a drinking problem, they would not want formal help regardless of problem severity, which is consistent with population data showing that resolutions without treatment are common (Sobell et al., 1996). The untreated majority of problem drinkers who avoid or delay seeking treatment may be receptive to telehealth and other low threshold interventions that do not require entry into the healthcare system. A sizeable minority also indicated they would choose informal help for mild to moderate drinking problems and reserve formal treatment for serious problems. This consumer-selected market segmentation resembles a public health approach to resource allocation in systems of care (Humphreys & Tucker, 2002). By comparison, formal treatment tended to be preferred for drug problems regardless of problem severity. The study has limitations related to the state-level sample of adults who would respond to a telephone survey about services for substance-related problems. First, the sample differed from select population demographic characteristics and the population prevalence of substance-related problems and help-seeking. Although RDD would be expected to minimize sampling bias within the specified population, the sample included more educated and female respondents, which was addressed statistically by sample weighting. Compared to national samples (SAMHSA, 2005, Wang et al., 2005), the sample also included more substance users and reported greater help-seeking from all sources, but had similar patterns of substance misuse. The topic of substance abuse almost certainly contributed to these characteristics; some potential participants were likely repelled by the topic, while others were drawn in because they had relevant experiences. Our broader definition of help-seeking also probably contributed to the higher observed rates of help-seeking (cf. Narrow et al., 1993). Although no claim of sample representativeness can be made along these dimensions, over-sampling of SP+ and HS+ households was beneficial for examining their role in consumer preferences, which was a key study goal. Second, the survey response rate was lower than desired, even though it was well within typical rates for state-wide random digit dial surveys (Yarber, Milhausen, Crosby, & Torabi, 2005) and was considered satisfactory given caller ID and other impediments to telephone data collection. In addition, studies have demonstrated only small differences in estimates obtained from surveys conducted with a range of response rates (Berdie, 1991; Keeter, Miller, Kohut, Groves, & Presser, 2000) and when non-responsiveness was due to the survey topic having variable interest or relevance to respondents (Groves, Presser, & Dipko, 2004). This was the case in the present study; i.e., most refusals indicated their household would never need such services. A final potential limitation concerns the possibility that the RDD method may have under-sampled households with severe cases of alcohol or drug dependence due to the associated chaotic lifestyle or irregular phone access. Mitigating this concern are studies indicating that the method reduces bias by giving every household with a telephone a chance to participate and, further, that estimates based on telephone-only households generally have little bias compared to all households with and without land-line telephones (Anderson, Nelson, & Wilson, 1998; Blumberg, Luke, & Cynamon, 2006). In addition, in the present study it was possible to interview another household member even if the member with substance-related problems was not contacted (e.g., due to intoxication or incarceration). Of surveyed households who reported having a member with problems, the respondent was the person with problems in 37% of such households. Although the survey results generally concurred with research on help-seeking conducted with diverse samples, some regional preferences may have been operative as well (e.g., lack of interest in computerized help). Investigating both general and local forces that influence preferences for care seems important for drawing out persons with stigmatized disorders like alcohol and drug problems and HIV and should be addressed in future surveys using a range of samples. For example, preferences and potential marketing effectiveness may differ among potential consumers linked to the criminal justice system or who have various other concurrent health and life issues. This approach to understanding consumer demand in broad and niche markets is already present in mainstream health care, which spans general medical practice as well as specialty and “boutique” health services (e.g., cosmetic procedures, fitness services, “lifestyle” pharmaceuticals). With these qualifications, the study broke new ground by viewing substance misusers and their social networks as consumers in a competitive market with many options, including continued drug use (Tucker & Simpson, 2003). When considered in this manner, a necessary strategic goal is to shift consumer preferences in ways that increase the relative value of service utilization over continued drug use. This consumer-oriented perspective has not found much acceptance in the United States, even though it appears important for reducing substance-related problems and helping close the gap between need and service utilization. Scientific groups (e.g., Institute of Medicine, 2006) have endorsed the need for a more consumer-centric approach to care, but the reality remains that services for alcohol and drug problems are often unappealing, even if effective. Consumer preferences in the present study were tied to malleable features that can be modified, and other studies have shown that consumer-responsive treatments have higher client satisfaction rates, which appear to improve retention and long-term outcomes (Carlson & Gabriel, 2001; Fountain et al., 2000; Rohrer & Hisenrath, 1999). The challenge, then, is to determine how best to incorporate consumer preferences and needs in an evidence-based system of care that appeals to different market segments, particularly the majority that delays or avoids seeking clinical care. Acknowledgements This research was supported in part by funds provided by the Center for AIDS Research at the University of Alabama at Birmingham and by grant no. K02 AA00209 from the National Institute on Alcohol Abuse and Alcoholism. Portions of this research were presented at the annual meeting of the American Psychological Association, Honolulu, HI, August 2004. References Appel et al., 2004. 1.Appel PW, Ellison AA, Jansky HK, Oldak R. Barriers to enrollment in drug abuse treatment and suggestions for reducing them: Opinions of drug injecting street outreach clients and other system stakeholders. The American Journal of Drug and Alcohol Abuse. 2004;30:129–153. MEDLINE |
CrossRef
Anderson et al., 1998. 2.Anderson JE, Nelson DE, Wilson RW. Telephone coverage and measurement of health risk indicators: Data from the National Health Interview Survey. American Journal of Public Health. 1998;88:1392–1395. MEDLINE |
CrossRef
Berdie, 1991. 3.Berdie DR. Telephone survey response rates: How high is high enough?. Marketing Research. 1991;28:35–44. Blumberg et al., 2006. 4.Blumberg SJ, Luke JV, Cynamon ML. Telephone coverage and health survey estimates: Evaluating the need for concern about wireless substitution. American Journal of Public Health. 2006;96:926–931. MEDLINE |
CrossRef
Carlson and Gabriel, 2001. 5.Carlson MJ, Gabriel RM. Patient satisfaction, use of services, and one-year outcomes in publicly funded substance abuse treatment. Psychiatric Services. 2001;52:1230–1236. MEDLINE |
CrossRef
Copeland and Martin, 2004. 6.Copeland J, Martin G. Web-based interventions for substance use disorders: A qualitative review. Journal of Substance Abuse Treatment. 2004;26:109–116. Abstract | Full Text |
Full-Text PDF (163 KB)
|
CrossRef
CDC, 2004. 7.Centers for Disease Control and Prevention (2004). HIV/AIDS surveillance report. Atlanta, GA: U.S. Department of Health and Human Services. CDC, 2005. 8.Centers for Disease Control and Prevention (2005). Behavioral risk factor surveillance system: 2005 data quality reports. Retrieved February 8, 2007 from http://www.cdc.gov/brfss/technical_infodata/quality.htm. Creative Research Systems, 2006. 9.Creative Research Systems (2006). Sample size calculator. Retrieved August 23, 2006 from http://www.surveysystem.com/sscalc.htm. Cunningham et al., 1993. 10.Cunningham JA, Sobell LC, Sobell MB, Agrawal S, Toneatto T. Barriers to treatment: Why alcohol and drug abusers delay or never seek treatment. Addictive Behaviors. 1993;18:347–353. MEDLINE |
CrossRef
Cunningham et al., 2005. 11.Cunningham JA, Humphreys K, Koski-Jannes A, Cordingly J. Internet and paper self-help materials for problem drinking: Is there an additive effect?. Addictive Behaviors. 2005;30:1517–1523. MEDLINE |
CrossRef
Fortney et al., 1999. 12.Fortney JC, Booth BM, Curran GM. Do patients with alcohol dependence use more services? A comparative analysis with other chronic disorders. Alcoholism: Clinical and Experimental Research. 1999;23:127–133. MEDLINE |
CrossRef
Fountain et al., 2000. 13.Fountain J, Strang J, Griffiths P, Powis B, Gossop M. Measuring met and unmet need of drug misusers: Integration of quantitative and qualitative data. European Addiction Research. 2000;6:97–103. MEDLINE |
CrossRef
French et al., 2006. 14.French MT, Homer JF, Nielsen AL. Does America spend enough on addiction treatment? Results from public opinion surveys. Journal of Substance Abuse Treatment. 2006;31:245–254. Abstract | Full Text |
Full-Text PDF (211 KB)
|
CrossRef
Grant et al., 2004. 15.Grant BF, Dawson DA, Stinson FS, Chou SP, Dufour MC, Pickering RP. The 12-month prevalence and trends in DSM-IV alcohol abuse and dependence: United States, 1991–1992 and 2001–2002. Drug and Alcohol Dependence. 2004;74:223–234. Abstract | Full Text |
Full-Text PDF (125 KB)
|
CrossRef
Groves et al., 2004. 16.Groves RM, Presser S, Dipko S. The role of topic interest in survey participation decisions. Public Opinion Quarterly. 2004;68:2–31.
CrossRef
Hinton et al., 2005. 17.Hinton A, Downey J, Lisovicz N, Mayfield-Johnson S, White-Johnson F. The community health advisor program and the Deep South network for cancer control. Family Community Health. 2005;28:20–27. MEDLINE Humphreys and Tucker, 2002. 18.Humphreys K, Tucker JA. Introduction: Toward more responsive and effective intervention systems for alcohol-related problems. Addiction. 2002;97:126–132. MEDLINE |
CrossRef
Ingram et al., 2005. 19.Ingram M, Gallegos G, Elenes J. Diabetes is a community issue: The elements of a successful outreach and education model on the U.S.-Mexico border. Prevention of Chronic Disease. 2005;2:A15. Institute of Medicine, 2006. 20.Institute of Medicine . Improving the quality of health care for mental health and substance use conditions. Washington, D.C.: National Academy Press; 2006;. Keeter et al., 2000. 21.Keeter S, Miller C, Kohut A, Groves RM, Presser S. Consequences of reducing nonresponse in a national telephone survey. Public Opinion Quarterly. 2000;64:125–148.
CrossRef
Kosmin and Mayer, 2007. 22.Kosmin, B.A., & Mayer, E. (2001). American religious identification survey. The Graduate Center, City University of New York. Retrieved February 18, 2007 from http://www.gc.cuny.edu/faculty/research_briefs/aris/aris_index.htm. Latkin et al., 2003. 23.Latkin CA, Sherman S, Knowlton A. HIV prevention among drug users: Outcome of a network-oriented peer outreach intervention. Health Psychology. 2003;22:332–339. MEDLINE |
CrossRef
Littleton et al., 2002. 24.Littleton MA, Cornell CE, Dignan M, Brownstein N, Raczynski JM, Stalker V, et al. Lessons learned from the Uniontown Community Health Project. American Journal of Health Behavior. 2002;26:34–42. MEDLINE MacCoun and Reuter, 2001. 25.MacCoun RJ, Reuter P. Drug war heresies: Learning from other vices, times, and places. New York: Cambridge University Press; 2001;. Narrow et al., 1993. 26.Narrow WE, Regier DA, Rae DS, Mandersched RW, Locke BZ. Use of services by persons with mental and addictive disorders. Findings from the National Institute of Mental Health Epidemiologic Catchment Area Program. Archives of General Psychiatry. 1993;50:95–107. Rohrer and Hisenrath, 1999. 27.Rohrer JE, Hisenrath P. Client satisfaction with substance abuse treatment. Health Marketing Quarterly. 1999;17:31–42. MEDLINE |
CrossRef
SAMHSA, 2005. 28.Substance Abuse and Mental Health Administration (2005). Overview of findings of the 2004 National Survey on Drug Use and Health. Retrieved August 25, 2006 from http://www.oas.samhsa.gov/nsduh/2k4nsduh/2k4overview/2k4overview.pdf. Seligman, 1995. 29.Seligman MEP. The effectiveness of psychotherapy: The Consumer Reports survey. American Psychologist. 1995;50:965–974.
CrossRef
Sobell et al., 1996. 30.Sobell LC, Cunningham JA, Sobell MB. Recovery from alcohol problems with and without treatment: Prevalence in two population studies. American Journal of Public Health. 1996;86:966–972. MEDLINE |
CrossRef
Tucker et al., 2007. 31.Tucker, J. A., Foushee, H. R., & Simpson, C. A. (2007). Public perceptions of substance abuse and how problems are resolved: Implications for medical and public health services. Manuscript submitted for review. Tucker and King, 1999. 32.Tucker, J. A. & King, M. P. (1999). Resolving alcohol and drug problems: Influences on behavior change and help-seeking strategies. In J. A. Tucker, D. M. Donovan, D. M., & G. A. Marlatt, G. A. (Eds.), Changing addictive behavior: Bridging clinical and public health strategies (pp. 97–126). New York: Guilford Press. Tucker and Simpson, 2003. 33.Tucker JA, Simpson CA. Merging behavioral economic and public health approaches to the delivery of services for substance abuse: Concepts and applications. In: Vuchinich RE, Heather N editor. Choice, behavioural economics, and addiction. Amsterdam: Pergamon; 2003;p. 365–378. Walsh et al., 1993. 34.Walsh DC, Rudd RE, Moeykens BA, Moloney TW. Social marketing for public health. Health Affairs. 1993;12:104–119. MEDLINE |
CrossRef
Wang et al., 2005. 35.Wang PS, Lane M, Olfson M, Pincus HA, Wells KB, Kessler RC. Twelve-month use of mental health services in the United States. Archives of General Psychiatry. 2005;62:629–640.
CrossRef
Yarber et al., 2005. 36.Yarber WL, Milhausen RR, Crosby RA, Torabi MR. Public opinion about condoms for HIV and STD prevention: A midwestern state telephone survey. Perspectives on Sexual and Reproductive Health. 2005;37:148–154. MEDLINE |
CrossRef
Department of Health Behavior, School of Public Health, University of Alabama at Birmingham, 1665 University Boulevard, 227 RPHB, Birmingham, AL 35294, USA  Corresponding author. Tel.: +1 205 934 5256; fax: +1 205 934 9325.
PII: S0955-3959(07)00240-X doi:10.1016/j.drugpo.2007.11.004 © 2007 Elsevier B.V. All rights reserved. | |
|
FULL TEXT