Prevention case management improves socioeconomic standing and reduces symptoms of psychological and emotional distress among transgender women

Abstract

Urban transgender women often experience several sociocultural co-factors which contribute to their risk of HIV infection. A transgender-specific HIV Prevention Case Management (PCM) intervention was implemented in a community HIV prevention setting and targeted reducing sex work and homelessness, increasing legal employment and income, and reducing psychological and emotional distress symptoms. Sixty high-risk transgender women were enrolled in the ten-session PCM intervention. Participants completed approximately nine out of the ten sessions (M=8.7; SD=2.6) and six-month follow-up evaluations were completed with 97% of the participants. Findings from baseline to follow-up evaluations demonstrated a decrease in homelessness (31.0% vs. 10.3%, p<0.01), less reliance on exchange sex as a primary source of income (41.4% vs. 22.4%, p<0.05), and significant decreases in symptom complaints across multiple Brief Symptom Inventory sub-scales, including depression, hostility, phobic anxiety, and psychoticism (all significant at p<0.05). Further, socio-economic improvements following the intervention were significantly associated with psychological and emotional gains. The study suggests that adding a culturally appropriate PCM intervention in a community setting is beneficial in addressing co-factors for HIV infection as well as psychological and emotional distress symptoms among this extremely high-risk population.

Introduction

Many male-to-female (MTF) transgender women live under marginal socioeconomic conditions marked by low income, high unemployment, and unstable housing (Nemoto, Luke, Mamo, Ching, & Patria, 1999; Reback, Simon, Bemis, & Gatson, 2001; Wilson, 2007). Many transgender women report monthly incomes of less than $1000 (Clements-Nolle, Marx, Guzman, & Katz, 2001; Nemoto, Operario, Keatley, Han, & Soma, 2004; Sevelius, Reznick, Hart, & Schwarcz, 2009; Wilson, Garofalo, Harris, & Belzer, 2010) and inability to purchase daily necessities such as living expenses (Sanchez, Sanchez, & Danoff, 2009). Researchers have found high prevalence of temporary housing, unstable housing, and homelessness among this population, with some estimates reaching as high as 47% (Clements-Nolle et al., 2001; Harawa & Bingham, 2009; Nemoto et al., 2004; Sanchez et al., 2009; Sevelius et al., 2009; Simon, Reback, & Bemis, 2000; Wilson et al., 2010). Even within transgender populations, MTF transgenders are more likely than their female-to-male (FTM) counterparts to report unstable housing and low monthly income (Clements-Nolle et al., 2001).

Transgender women are often compelled by their socioeconomic circumstances to engage in street economy to earn a living. Research suggests that transgender women engage in high levels of sex work (Bowers, Branson, Fletcher, & Reback, 2011; Harawa & Bingham, 2009; Reback, Lombardi, Simon, & Frye, 2005; Wilson et al., 2009) and are more likely to turn to sex work than men who have sex with men (Nemoto et al., 1999; Sanchez, Finlayson, Murrill, Guilin, & Dean, 2010). Survival sex by transgender women makes them vulnerable to HIV, often by engaging in unsafe sexual behaviors with potentially high-risk partners (De Santis, 2009; Herbst et al., 2008). The potential for harm by such behaviors may be discounted by transgender women due to the desire to affirm one’s gender identity via high-risk exchange sex encounters (Bockting, Robinson, & Rosser, 1998; Kammerer, Mason, Connors, & Durkee, 2001; Melendez & Pinto, 2007; Nemoto, Keatley, Operario, & Soma, 2002; Nemoto, et al., 2004; Reisner et al., 2009).

Given their relatively low socioeconomic standing and high rates of sex work, it is not surprising that transgender women frequently report significant symptoms of psychological and emotional distress including clinical depression (Bockting, Huang, Ding, Robinson, & Rosser, 2005; Carson, 2009; Clements-Nolle et al., 2001; Operario & Nemoto, 2005; Sanchez et al., 2010), anxiety (Bockting, Knudson, & Goldberg, 2006; Carson, 2009), and suicidal ideation (Herbst et al., 2008; Mathy, 2002; Xavier & Simmons, 2001). Moreover, transgender women report elevated rates of a lifetime history of hospitalization for mental health issues (Clements-Nolle et al., 2001; Mathy, 2002; Reisner et al., 2009; Wilson, et al., 2009) and express a desire for mental health counseling, particularly to address transgender-related issues (Herbst et al., 2008).

Researchers have called for increased participation of transgender women in active HIV prevention services, particularly client-centered counseling and small-group interventions that focus on building safe sex skills (Harawa & Bingham, 2009). Prevention Case Management (PCM), similarly referred to as Comprehensive Risk Counseling and Services (CRCS), is a client-centered intervention for individuals at risk for HIV infection or transmission (Centers for Disease Control and Prevention [CDC], 2006). Through a combination of intensive individualized counseling and case management, PCM offers critical services to persons whose life circumstances (e.g., socioeconomic, psychosocial) have been prioritized above their risk reduction efforts (CDC, 1997). Since its development in the early 1990s, PCM has been utilized with diverse populations to target HIV injection risks, substance abuse, and unprotected sexual behaviors. This combined approach to prevention has demonstrated successful reductions in needle sharing (Gasiorowicz et al., 2005; Robles et al., 2004), drug injection (Robles et al., 2004), substance use (Myers et al., 2005; Wong et al., 2008), and sexual risk behaviors (Gasiorowicz et al., 2005; Myers et al., 2005; The Healthy Living Project Team, 2007) among individuals at risk for HIV infection or transmission. PCM interventions have also led to reductions in the number of unmet service needs (i.e., long-term housing, supportive mental health counseling, other basic needs; Thompson et al., 1998) as well as positive changes in self-efficacy (for condom use, injection drug, and substance use risk) and intentions to use condoms (Bauserman et al., 2003).

Given the existing evidence supporting the use of PCM, this study sought to evaluate the impact of a PCM intervention tailored for high-risk transgender women. Specifically, the intervention sought to reduce sex work by facilitating legitimate employment and reducing homelessness by helping participants obtain stable and affordable housing. Furthermore, it was hypothesized that reductions in sex work and unstable housing would be accompanied by concomitant reductions in symptoms of psychological and emotional distress.

Methods

Participants

Participants were recruited from a community-based, low-intensity, health education/risk reduction HIV prevention program serving male-to-female transgender women in the Hollywood/West Hollywood area of Los Angeles County. Oversight of all study activities was provided by the research institute’s Institutional Review Board.

Potential participants were deemed eligible for the study if they were a self-identified male-to-female transgender woman, between the ages of18 and 65 years, willing to provide informed consent, and willing to comply with study procedures. For the purpose of this study, transgender was defined as any woman who believed her male sex assigned at birth was in conflict with her gender identity. All transgender women were eligible for participation regardless of their stage of gender transition. Individuals were excluded if they did not meet all criteria, identified as cross-dressers, transvestites, or drag queens, that is, individuals who wear clothing of the opposite gender but do not believe their biological sex is different from their gender identity.

Procedure

Potential participants for the PCM intervention completed a minimum of three sessions in a low-intensity, HIV prevention health education/risk reduction program, which served as a behavioral screening procedure to demonstrate likelihood of being able to engage a multiple 10-session intervention, including a follow-up assessment. Those able to complete three sessions in the health education/risk reduction program were invited to participate in the 10-session PCM intervention. Following provision of informed consent, evaluations were conducted to verify eligibility and to collect baseline data including demographics, recent substance use and sexual risk behaviors, and levels of psychological and emotional distress. Participants received $20 on completion of the baseline assessment during program enrollment. After completing baseline assessments, participants met with the counselor to assess areas of behavioral change and to create a participant-centered prevention plan. To encourage attendance at all 10 sessions, participants were compensated for their time with a $10 incentive per session and a $25 bonus after the fifth and tenth sessions. Participants also received $50 for completing a six-month follow-up evaluation. The maximum participants could earn was $220. Incentives were not associated with behavioral change but rather served as a motivation for attending the scheduled PCM appointments. All intervention staff were post-operative, male-to-female transgender women.

Intervention

Prevention Case Management works to change behaviors that put a participant at increased risk for HIV transmission and acquisition, focusing particularly on participants that present with multiple and complex risk behaviors, and who have been unable to initiate or sustain changes in those behaviors with standard, low-intensity prevention interventions. PCM provides multiple, one-on-one counseling sessions to help the participant assess her own personal risk, and create an individualized plan to initiate and sustain behavior change. Given that the intervention was targeting transgender women, the PCM sessions served to identify and then reduce co-factors for HIV transmission that are specific to transgender women. The PCM sessions were 60 minutes in length, were conducted weekly, and all 10 sessions needed to be completed within a six-month timeframe.

Measures

Assessments of various lifestyle domains were administered during enrollment and at six-month follow-up evaluations.

Transgender health survey

This assessment includes collection of descriptive information specific to transgender women regarding health disparities, HIV-risk behaviors, medical and psychiatric information, housing and employment status, and legal system involvement (Reback et al., 2001).

Brief Symptom Inventory (BSI): (Derogatis & Melisaratos, 1983)

The Brief Symptom Inventory was administered as an indicator of state psychological (including psychosis) and emotional functioning.

Statistical analysis

Analyses were done using Stata v10SE (StataCorp, 2007). To evaluate changes in socioeconomic, behavioral, and BSI indicators over time, within subject approaches were used including paired t-tests for continuous variables and chi-square or Fisher’s exact tests for categorical variables. Additionally, a random effects seemingly unrelated regression analysis (SURE) was conducted to examine potential associations between changes in sociodemographic characteristics (i.e., legal employment, income, reliance on sex work, and homelessness) and changes in BSI symptom complaints. SURE analyses are appropriate when dealing with multiple oblique outcome variables, as is the case with the nine sub-scales of the BSI. Random effects SURE analyses are designed to compensate for the autocorrelation arising from repeated measurements of the same individual across time, and reduce the probability of committing a Type-I hypothesis testing error arising from multiple hypothesis tests on correlated outcome variables (Nguyen, 2008). Coefficients of a SURE analysis are similar to those found in more standard multivariate regression techniques; each coefficient describes the change in the specific BSI subscale score for every one unit increase in the predictor variable. In practice, coefficients for 0/1 dichotomous predictors (i.e., legal employment, sex work, and homelessness) can be compared in terms of magnitude, and describe the change in BSI subscales scores as a person acquires legal employment, engages in sex work, or becomes homeless. Coefficient estimates for income describe the change in BSI subscale scores as a participant moves from one ordinal income category to the next. All findings are considered significant at p<0.05, and because the SURE analysis predicts directional relationships, all SURE significance tests are one-tailed.

Results

Intervention retention

Participants completed on average approximately nine out of the ten sessions (M=8.7; SD=2.6). A total of45 participants (75.0%) completed all 10 sessions. The mean incentive payout was $196 (SD=$47) and 58/60 (96.7%) completed a six-month follow-up evaluation.

Sociodemographic characteristics

Participants ranged in age from 20 to 64 years (M=38.3) and were predominantly transgender women of color (n=43; 71.7%). Most identified themselves as heterosexual (n=30; 50.0%) or gay (n=14; 23.3%). Almost three-quarters (n=43; 71.7%) reported their gender identity as transgender or transsexual. Educational attainment was low with a mean of11.5 years (SD=2.9) of formal schooling. Seventeen participants (28.3%) self-reported an HIV-positive serostatus. In addition, 19 participants (31.7%) reported at least one sexually transmitted infection in their lifetime. An overwhelming majority of the sample reported at least one instance of incarceration in their lifetime (n=53; 88.3%). Twenty-five participants (41.7%) indicated having had at least one episode of substance abuse treatment (Table 1).

Table 1.

Demographic and socioeconomic characteristics.

	Mean (SD) or (N%)
Age	38.32 (10.2)
Education	11.5 (2.9)
<12 Years	20 (33.3%)
≥12 Years	40 (66.7%)
Race/ethnicity
Caucasian	17 (28.3%)
Hispanic/Latina	18 (30%)
African-American	13 (21.7%)
Multi/other	12 (20%)
Sexual identity
Heterosexual	30 (50%)
Bisexual	11 (18.3%)
Gay	14 (23.3%)
Other	5 (8.3%)
Gender identity
Woman	17 (28.3%)
Transgender/transsexual	43 (71.7%)
Reported HIV status
HIV-positive	17 (28.3%)
HIV-negative	39 (65%)
DK	4 (6.7%)
Sexually transmitted infection
Lifetime	19 (31.67%)
Incarceration
Lifetime	53 (88.33%)
Substance abuse Tx episode
Lifetime	25 (41.67%)

Participants showed significant changes in self-reported monthly earnings, with the largest category reduction occurring in the $51–$249 income bracket, and the largest category increases in the $500–$999 and $1000–$2999 income brackets. Fewer participants reported sex work as their primary source of income at follow-up evaluations (41.4% vs. 22.4%; p<0.05). Significant changes in self-reported residential status were observed, with the largest category reduction occurring in reports of living on the streets, and the largest category increase occurring in reports of living in a home. Subsequent t-tests for differences in proportions reveal the reduction in self-reported homelessness to be significant (31% vs. 10.3%; p<0.01).

Psychological and emotional levels of distress

Fewer symptoms of psychological and emotional distress were reported at follow-up evaluations across all BSI indicators, with significant (all minimum p<0.05) improvements being made in overall psychological distress (Global Severity Index) (1.01 vs. 0.82), positive symptom distress (PSDI) (1.76 vs. 1.58), interpersonal sensitivity (1.10 vs. 0.82), depression (1.11 vs. 0.88), hostility (0.64 vs. 0.46), phobic anxiety (0.98 vs. 0.66), and psychoticism (0.99 vs. 0.74) (see Table 2). Additionally, observed reductions in obsessive-compulsive symptoms were trending toward significance (1.25 vs. 1.06; p ≤ 0.1).

Table 2.

Changes in employment, residential status, and BSI symptom complaints from baseline to follow-up (N=58).

	Baseline N(%) or mean (SD)	Follow-up N(%) or mean (SD)	Significance
Income (past 30 days)			***
Less than $50	3 (5.2%)	5 (8.6%)
$51–$249	14 (24.1%)	6 (10.3%)
$250–$499	11 (19%)	9 (15.5%)
$500–$999	20 (34.5%)	23 (39.7%)
$1000–$2999	8 (13.8%)	13 (22.4%)
$3000–$4999	2 (3.5%)	2 (3.5%)
Income (recent sources)
Employment (part-/full-time)	11 (19%)	13 (22.4%)	ns
Sex work	24 (41.4%)	13 (22.4%)	*
Current residential status			***
House	14 (24.1%)	22 (37.9%)
Family’s house	3 (5.2%)	2 (3.5%)
Friend’s house	6 (10.3%)	8 (13.8%)
Partner’s house	0 (0%)	2 (3.5%)
Hotel/motel	9 (15.5%)	9 (15.5%)
Halfway house	4 (6.9%)	4 (6.9%)
Homeless shelter	4 (6.9%)	3 (5.2%)
Streets	18 (31%)	6 (10.3%)
Jail	0 (0%)	1 (1.7%)
Other	0 (0%)	1 (1.7%)
BSI symptom complaints
Global severity index	1.01 (0.64)	0.82 (0.67)	*
Positive symptom distress index	1.76 (0.56)	1.58 (0.57)	**
Somatization	0.81 (0.70)	0.67 (0.62)	ns
Obsessive-compulsive symptoms	1.25 (0.90)	1.06 (0.86)	ns
Interpersonal sensitivity	1.10 (0.78)	0.82 (0.87)	*
Depression	1.11 (0.90)	0.88 (0.88)	*
Anxiety	1.02 (0.93)	0.89 (0.89)	ns
Hostility	0.64 (0.57)	0.46 (0.53)	*
Phobic anxiety	0.98 (0.85)	0.66 (0.82)	**
Paranoid ideation	1.27 (0.80)	1.15 (0.89)	ns
Psychoticism	0.99 (0.74)	0.74 (0.72)	**

^*p ≤ 0.05;

^**p ≤ 0.01;

^***p ≤ 0.001.

As shown in Table 3, acquisition of legal employment was significantly negatively associated with symptom complaints on three of the nine BSI indicators: somatization, obsessive-compulsive disorder, and anxiety (all p<0.05), while observed reductions in hostility and paranoid ideation were trending toward significance (both p<0.1). Increases in monthly income were significantly negatively associated with symptom complaints on four of the nine BSI indicators (i.e., somatization, interpersonal sensitivity, phobic anxiety, and psychoticism, all p<0.05), and were associated with reductions of marginal significance (p<0.1) in another three (i.e., obsessive-compulsive disorder, depression, and anxiety). Engagement in sex work was significantly positively associated with depression scores. Last, being or becoming homeless was significantly positively associated with hostility and paranoid ideation scores.

Table 3.

Seemingly unrelated regression analyses (N = 58) of BSI symptom complaints on employment, sex work, income, and homelessness.

Covariate	Somatization	BSI index
		Obsessive- compulsive disorder	Interpersonal sensitivity	Depression	Anxiety	Hostility	Phobic anxiety	Paranoid ideation	Psychoticism
	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)	Coefficient (SE)
Legal employment	−0.276* (0.14)	−0.275* (0.16)	−0.043 (0.16)	0.03 (0.16)	−0.292* (0.17)	−0.153 (0.11)	−0.106 (0.17)	−0.184 (0.14)	−0.058 (0.13)
Sex work	0.036 (0.16)	−0.045 (0.2)	0.198 (0.19)	0.334* (0.19)	−0.042 (0.21)	0.06 (0.13)	0.159 (0.21)	−0.042 (0.18)	0.25 (0.16)
Income	−0.08* (0.04	−0.083 (0.053)	−0.125** (0.052)	−0.076 (0.05)	−0.084 (0.06)	0.02 (0.03)	−0.127** (0.06)	−0.027 (0.05)	−0.108** (0.04)
Homelessness	0.005 (0.18)	−0.162 (0.21)	0.134 (0.21)	−0.321 (0.2)	0.101 (0.22)	0.275* (0.14	−0.054 (0.23)	0.407* (0.19)	0.006 (0.18)

Note: Statistical controls: race, age, and education.

^*p ≤ 0.05;

^**p≤0.01.

Discussion

This study evaluated the impact of developing and implementing a high-intensity, culturally specific PCM intervention in a community HIV prevention setting specifically for high-risk transgender women. Most of the participants in the study were socio-economically disadvantaged with a low percentage reporting legal employment, and many reporting unstable housing and low monthly incomes. Additionally, 41.4% reported sex work as their primary source of income at baseline. Socioeconomic conditions improved following the PCM intervention: at follow-up, participants reported a higher monthly income and less reliance on sex work as a primary source of income. The reductions in exchange sex and increases in monthly income were encouraging; however, a majority of the participants still reported earning less than $1000 per month at follow-up, which is consistent with previous studies of this population (Clements-Nolle et al., 2001; Nemoto et al., 2004; Sevelius et al., 2009; Wilson et al., 2010). Additionally, there were significant changes in housing status such that nearly 60% of participants at follow-up were living in stable housing (i.e., on their own or with a friend or loved one), an increase of almost 20%. Nonetheless, unstable housing was still prominent among this sample (approximately 41%), which supports previous estimates among this population (41=51%; Clements-Nolle et al., 2001; Nemoto et al., 2004; Sanchez et al., 2009; Sevelius et al., 2009). Improvements in housing status are usually contingent upon financial resources and, in some cases, having a social support system. Therefore, participants may require a longer PCM intervention or a more intensive program in order to fully address barriers to acquiring a stable residence.

Psychological and emotional distress variables were markedly improved at follow-up evaluations with participants reporting fewer symptoms across all BSI indicators. Further, the progress that participants made in regards to income and housing, as well as their reduced reliance on sex work were each associated with psychological and emotional gains. Furthermore, the SURE analysis revealed far more BSI improvement in association with positive outcomes (i.e., legal employment and monthly income) than to decreases in negative outcomes (i.e., reliance on sex work as a primary source of income and homelessness). Thus, while it appears that culturally tailored PCM interventions can reduce negative outcomes and increase positive outcomes among transgender women, providers working with transgender women may find their participants better able to address job development and obtaining legal employment as treatment goals rather than focusing directly on sex work and transient housing. It seems reasonable that reductions in sex work and unstable housing are more likely to be achieved as treatment goals in the context of sustained legal employment.

Because PCM services utilize an individual counseling approach, participants were able to explore with the counselor the myriad of co-factors for HIV transmission that are unique to transgender women. For transgender women, the literature has unequivocally pointed out that HIV risks are intertwined with socioeconomic co-factors – many of which stem from stigma and discrimination regarding their gender identity and presentation (see De Santis, 2009 for review). To counter these effects, many transgender women engage in sex work which provides a source of income and can affirm their gender identity. However, this may create a situation whereby high-risk sexual behaviors are introduced as clients offer more money in exchange for not using condoms. As previous researchers have shown, continued involvement in sex work not only increases HIV risk but can affect a person’s self-esteem negatively (Clements-Nolle, Guzman, & Harris, 2008).

The results of this study are encouraging and reinforce the need to implement interventions that consider the complexity of co-factors faced by transgender women. These findings implicate a process by which transgender women make decisions about the activities in which they engage. Homelessness is a potent motivator and these data indicate that transgender women will engage either legal or illegal behaviors to secure safe housing and basic safety needs prior to addressing other, very important concerns such as reduction of drug use, psychological distress, and sexual risks. Using tailored interventions to increase the ability of transgender women to provide for basic safety needs via increased monthly income and access to available housing appears to have value in decreasing behaviors that confer vulnerability, such as high-risk sexual behaviors and sex work that are otherwise used to meet this objective.

These findings must be interpreted within the context of the study’s limitations. First, without a comparison or control arm the observed patterns of reduction in symptoms of psychological and emotional distress are associational. Second, the data were restricted to a small sample of urban transgender women capable of receiving HIV prevention services through a low-intensity health education/risk reduction program. As such, the results cannot be generalized to broad transgender populations, particularly to either more chaotic or more assimilated transgender populations. Researchers and service providers interested in utilizing PCM who do not already offer HIV prevention services to transgender women may need to develop culturally specific outreach strategies such as providing cosmetics as an outreach gift to attract and engage a potential participant. Employing transgender women as outreach workers, health educators, and counselors will, undoubtedly, increase an organization’s accessibility to this population (Nemoto Operario, Keatley, Nguyen, & Sugano, 2005). Recruitment and retention with this type of client-centered intervention is likely to hinge on an organization’s relationship with its community advisory board, which should, ideally, consist of members from the target population who can offer timely and dynamic suggestions for making the program appealing to both current and potential participants. Third, participants were provided incentives to attend sessions, which may have exaggerated behavior changes observed in the absence of same. Finally, all data were self-reported and, thus, are subject to underreporting of sensitive behaviors.

Conclusion

Despite these limitations, this study has demonstrated the benefits of a culturally appropriate PCM intervention for transgender women. Both moderate and significant socioeconomic behavior changes were demonstrated at six-month follow-up evaluations including an increase in reported monthly income, less reliance on sex work as a primary source of income, and a decrease in homelessness. Coinciding with the changes, participants reported significantly improved levels of psychological and emotional distress. The feasibility of this intervention was demonstrated by the high rate of retention during the intervention and at six-month follow-up evaluations. Participants completed approximately nine out of the ten required PCM sessions demonstrating the acceptability of the intervention by the target population. Given the high rates of HIV infection, multiple and complex co-factors, and levels of comorbidity among transgender women, identifying effective interventions for this population is an urgent need.