Absence of conflict of interest.
Citation
Highlights
-
The study’s objective was to examine the impact of a counseling intervention – Making Employment Needs (MEN) Count – on employment.
-
The study compared employment for adults who received the MEN Count intervention to a comparison group of adults who did not receive the intervention. Data was collected from study participants at baseline, and again at six- and twelve-month follow-ups.
-
The study suggested that the invention increased participants’ odds of being employed full-time relative to being unemployed.
-
The quality of causal evidence presented in this report is low because the study’s statistical analysis does not meet CLEAR’s standards. This means we are not confident that the estimated effects are attributable to MEN Count; other factors are likely to have contributed.
Intervention Examined
Making Employment Needs (MEN) Count
Features of the Intervention
MEN Count is a race- and gender-tailored counseling intervention focused on HIV (human immunodeficiency virus)/STI (sexually-transmitted infection) prevention for Black heterosexual men in Washington, DC. The intervention was designed to promote safer sex, reduce STIs, and help support stable housing and employment, as these structural factors contribute to sexual risk. MEN Count involved three one-hour, one-on-one counseling sessions delivered by a male peer case manager over a 60-to-90-day period. Sessions were focused on sexual risk reduction, asking clients to build action plans to change behavior and achieve goals related to lower risk for HIV/STIs, integrated with housing and employment case management as needed.
Eligible participants were self-identified Black adult men who reported heterosexual risk for HIV/STI (defined as unprotected sex with a woman and at least two female sex partners in the past year) and who were either currently unemployed or had experienced homelessness in the past six months.
Features of the Study
The study compared employment for adults in the MEN Count group versus a comparison group at baseline, six-, and twelve-month follow-ups. Participants were recruited simultaneously for both groups from a large public STI clinic, community and street outreach, flyers, Craigslist, and referrals. Participants were then assigned to the intervention or comparison condition based on case manager availability at the time of recruitment. The treatment group had access to the intervention. Participants in the comparison group had access to a case manager–delivered program of similar structure and length that was instead focused on stress management.
The total sample of 454 men was evenly divided between the intervention and comparison groups (227 in each). About half of participants had been homeless in the past 90 days and about two-thirds were unemployed at baseline. Around 30 percent of participants were between ages 18-24, 28 percent were 25-29, and 26 percent were 30-39, with the remainder being 40 years old or older. Of the 277 participants in each group, 95 treatment group participants completed the six-month follow-up and 120 completed the twelve-month follow-up while 105 comparison group participants completed the six-month follow-up and 123 completed the twelve-month follow-up.
The authors used a statistical model to compare self-reported employment at the three time points for the treatment versus comparison groups. The statistical model included a random intercept for individuals to account for repeated measures over time.
Findings
Employment
- The study suggested that the MEN Count intervention increased the odds of being employed full-time versus unemployed but did not change the odds of being employed full-time versus employed part-time.
Considerations for Interpreting the Findings
The study’s analysis model included individual-level random effects, which can produce biased estimates if unobserved individual characteristics are correlated with variables included in the statistical model. CLEAR standards thus require authors to present the results of a specification test that justifies the use of random effects in any statistical analysis. Because the study authors do not do so, this study cannot meet CLEAR standards for a moderate evidence rating, the highest rating available for nonexperimental studies.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because the study’s statistical analysis does not meet CLEAR’s standards. This means we are not confident that the estimated effects are attributable to MEN Count; other factors are likely to have contributed.