We discuss the findings in terms of topic addressed as well as comparisons by prior treatment use, gender or marital status. In this group of privately insured employees, the most commonly cited reasons for entering treatment were health concerns, difficulties at home, with family or friends, and difficulties at work. Although more than 41% indicated difficulties at work as a reason for entering treatment, far fewer indicated that they had received pressure or encouragement from work-related sources. This suggests that work-related coercion or encouragement, though evident for many individuals, is much less common than employees' perceptions of work problems as a factor in decision-making. It is possible that work-related problems were noticed and commented upon by others outside of the workplace, such as family members realizing that work problems stem from a substance use condition and urging treatment entry. It is also possible that for many people, work-related problems were not severe enough to provoke direct workplace involvement.
Respondents who reported prior use of substance abuse treatment were significantly more likely than first-time service users to report difficulties at home, with family or friends, financial concerns, and difficulties of work as reasons for treatment entry, and were similar in terms of other possible reasons. Experienced treatment users arrive in treatment with a greater burden of problems that precipitated treatment entry. They may need, be more likely to need, concrete help with financial and work problems as well as assistance with family and relationship issues related to substance use issues. Legal issues are present in this employed sample but were cited by less than one fifth of respondents.
Interestingly, women cited difficulties at home, with family or friends much more frequently as a reason for entering treatment than men. This may underline the particular importance of these relationships for women, and highlights the centrality of attending to the social system in treatment. The precise nature of the social connections, of course, is critical and may either support or hinder recovery. Our data do not address this issue but do disproportionately find difficulties in these relationships as a trigger for women's treatment seeking. The fact that unmarried clients were more likely than married clients to cite work in financial difficulties as a reason for entering treatment may reflect the potentially stabilizing influence of marriage-though as cited earlier, the literature has been mixed on this point.
There has been increasing concern about missed opportunities for identifying substance use conditions in medical settings and about the lack of integration or coordination between specialty behavioral health and general medical care, and there is hope that federal health reform will foster solutions to these problems . The study findings illustrate the challenge of this aspect of health care. Although nearly two thirds of the respondents cited health concerns as a reason they entered treatment, only 24.4% indicated that a physician or other health care provider had encouraged or pressured them to enter treatment. It is possible that with better screening and intervention, patients might have received help sooner. The finding that only 18.4% of respondents believed their specialty substance abuse treatment provider communicated with their primary care physician also raises a red flag. This suggests a major lack of direct communication among providers, even if the clients themselves may have relayed information about their treatment to both types of providers. All of, these findings underline the ongoing need for better behavioral health and general medical care coordination, an important goal for the field .
Family and friends were identified as the number one source of encouragement or pressure to enter treatment. This supports findings of prior research [3, 22]. About one third of the sample did not endorse any of the sources of encouragement or pressure that were asked about. It is likely that some of these might have received pressure from the legal/criminal justice system, while others in this group might not have received any specific encouragement or pressure to enter treatment.
It is interesting that despite the greater proportion of persons with prior treatment reporting several precipitating problems, there was no significant difference in the sources of encouragement or pressure. Thus it appears that persons with a treatment history are more likely to experience a number of precipitating problems in a subsequent treatment experience but are not more likely to be subject to pressure or encouragement. This seems surprising, but one possible explanation is that individuals with a treatment history are better able to recognize when they need help.
Most respondents found treatment helpful, and this was the case across prior service use, gender and marital status subgroups with no statistically significant differences. The apparently large difference in the helpfulness ratings of clients with versus without prior service use was not statistically significant. With the p value of less than .10 in this relatively small sample, however, this question merits further investigation in large samples. Clients with prior experience of treatment systems might have a better idea how treatment works and how to get the most out of it, they may feel more positively knowing that it has perhaps helped in the past, and they may have more realistic expectations regarding treatment.
About one quarter of respondents in this privately-insured group reported stopping treatment sooner or getting less treatment than they wanted. Many of these respondents felt that treatment was not working, that treatment time or location was inconvenient, or had other problems with service providers. These are helpful findings for providers seeking to better engage their clients. Half of the respondents reporting unmet need cited problems with coverage or costs as the reason. According to the National Household Survey on Drug Use and Health, the vast majority of people with unmet need for substance abuse treatment fail to recognize that they have a problem . However, among those who do recognize their problem and make an effort to get help, 9% indicate that they have health coverage but it did not cover treatment or cover the costs. The findings from the study presented here offer insight into the viewpoints of those with comprehensive, employer-sponsored coverage who have successfully accessed treatment. Without detailed clinical data it is impossible to ascertain whether further treatment was clinically indicated. Future work might specifically focus on this issue in a larger sample, and examine detailed treatment patterns in relation to perceived barriers. Another potential barrier, lack of sensitivity to cultural background such as race, ethnicity, religion, age, sexual orientation, or language preferences, proved to be rare in practice within this sample.
Reported work impairments were common, supporting prior evidence . The vast majority of respondents felt that treatment had improved work attendance and ability to perform responsibilities when at work, as well as helping to decrease substance use. The finding that prior service users were significantly more likely to report work impairments is similar to the findings that higher proportions of prior service users endorsed various problem areas as reasons for entering treatment. In this context, it is interesting, and heartening, that both experienced and first-time service users reported these positive outcomes in similar proportions. The fact that women were more likely than men to indicate that their work attendance had improved as a result of services received provides an interesting nuance, given that they had similar global ratings of how much treatment had helped.
There are several limitations to the study. The sample was drawn from a single MBHO's covered population, thus generalizability may be limited to the extent that treatment arrangements vary by MBHO. However, MHN's benefits are generally comprehensive in terms of a continuum of care and carry benefit levels similar to many other MBHOs. Generalizability is also limited by the demographics of the sample, including the small numbers of nonwhite or young employees. Another limitation is loss of sample due to unavailability of home telephone numbers, and substantial non-response. The relatively small sample size means that there was limited statistical power for some conditional analyses, although numerous observed differences did attain statistical significance. The study nonetheless provides a rare window into the experience of employed clients in substance abuse treatment through an MBHO, which can also inform future studies. This is similar to other studies with small samples that can illuminate under-studied topics . The measures derived from survey items have not been validated. These analyses and comparisons of experienced and first-time treatment users are essentially descriptive in nature. Future work could examine in more depth selected experience of care or outcomes measures in a multivariate context, to control for multiple confounders such as the fact that new users were more likely to have alcohol-only diagnoses. Detailed clinical data on factors such as severity of substance use conditions were not available. Respondents received different types and quantities of services so this study is a snapshot of clients recently in some form of treatment. Not all of the services received were specialty treatment services provided by MHN. However, the vast majority of services reported on were specialty substance abuse treatment services provided through respondents' employer-sponsored benefits, indicating MHN services. Despite these limitations, these data offer further insight into perceptions of substance abuse treatment in a privately insured, largely working population.