Tailoring disaster mental health services to diverse needs: an analysis of 36 crisis counseling projects.
Abstract: The federal Crisis Counseling Program (CCP) funds states' delivery of mental health services after disasters. These services are provided by social workers, other mental health professionals, and paraprofessionals from the local community. The present study examined whether CCP grant recipients that reported more tailoring of their interventions to the needs of diverse community segments achieved greater community penetration. The study reviewed archival records from 36 crisis counseling projects ending between 1996 and 2001. Numbers of clients and client ethnicity were determined through service logs. Tailoring of services was determined by content coding of projects' reports. Community demographics were determined from census data. Fifty-six percent of the projects reported using three or more tailoring strategies, suggesting a "precompetence" or greater stage of cultural competence. The proportion of members of racial or ethnic minority groups among program clients closely matched the proportion in grantees' communities. Projects that reported more types of tailored activities reached more clients and served more members of minority groups. These findings confirm that adapting crisis counseling services to diverse local needs is associated with greater community penetration of mental health services.

KEY WORDS: cultural competence; disaster; mental health; outreach strategies; racial/ethnic disparity
Article Type: Report
Subject: Social workers (Practice)
Psychiatric services (Services)
Psychiatric services (Social aspects)
Psychiatric services (Demographic aspects)
Emergency management (Management)
Cross-cultural counseling (Management)
Authors: Rosen, Craig S.
Greene, Carolyn J.
Young, Helena E.
Norris, Fran H.
Pub Date: 08/01/2010
Publication: Name: Health and Social Work Publisher: National Association of Social Workers Audience: Academic; Professional Format: Magazine/Journal Subject: Health; Sociology and social work Copyright: COPYRIGHT 2010 National Association of Social Workers ISSN: 0360-7283
Issue: Date: August, 2010 Source Volume: 35 Source Issue: 3
Topic: Event Code: 360 Services information; 290 Public affairs; 200 Management dynamics Canadian Subject Form: Cross cultural counselling Computer Subject: Company business management
Product: Product Code: 8000186 Mental Health Care; 9105250 Mental Health Programs NAICS Code: 62142 Outpatient Mental Health and Substance Abuse Centers; 92312 Administration of Public Health Programs
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 234998147
Full Text: The federal Crisis Counseling Program (CCP) funds states' delivery of mental health services to communities affected by disasters. The CCP is based on a social work and public health outreach model that stresses proactively bringing services to people wherever they are in the community. Guidance from the CCP to state grantees (U.S. Department of Health and Human Services [HHS], 2003) has long stressed several principles that are related to cultural competence. These include valuing cultural differences, recruiting disaster workers from all segments of the affected community (Cohen, 1984), involving community leaders as advisers and cultural brokers (Gould, 1988; Hernandez, Nesman, Mowery, Acevedo-Polakovich, & Callejas, 2009), encouraging clients to access culturally appropriate sources of support (Ida, 2007), and adapting services to fit the needs of different cultural and linguistic segments of the population (Hernandez, Nesman, Isaacs, Callejas, & Mowery, 2006; Sue, 2006). However, there has previously been little formal evaluation of the extent to which crisis counseling projects embrace these principles and no test of whether projects that strive for cultural competence are in fact more successful in serving the entire affected community.

CCP

Crisis counseling has become an integral component of federal assistance to communities recovering from disasters and other mass casualty events. In 1974, the Disaster Relief Act (P.L. 93-288) established the Federal Crisis Counseling Assistance and Training Program (that is, the CCP) to provide supplemental funding for crisis mental health services to U.S. states and territories affected by federally declared disasters. Administration of such grants is overseen by the Emergency Mental Health and Traumatic Stress Services Branch of the Center for Mental Health Services, within the Substance Abuse and Mental Health Services Administration (SAMHSA). Through this program, U.S. states or territories request federal funding for mental health assistance for an average of 10 to 12 disaster events annually (Norris et al., 2005).

The CCP supports a mix of psychoeducation and brief (typically one- to three-session) counseling services. These activities are intended to accelerate survivors' return to predisaster levels of functioning by educating them about common emotional reactions after disasters; normalizing and destigmatizing their emotional reactions; promoting self-care and positive coping; encouraging use of social supports; facilitating referrals for instrumental assistance for food, housing, medical, and financial needs; and, when needed, offering referral to mental health treatment (HHS, 2000).

The CCP has a strong outreach orientation and aims to bring services to people who might not otherwise seek out mental health assistance. Some outreach principles of the CCP model are providing services in community settings rather than in formal treatment settings, using nonstigmatizing language that does not connote "disorder" or "treatment," recruiting and training indigenous staff drawn from the local community, and hiring a mix of mental health professionals (often social workers) and paraprofessionals who have experience working with local community groups (HHS, 2000).

CULTURE AND RECOVERY AFTER DISASTER

Ethnicity and culture play key roles in how people recover after disasters (HHS, 2003). Although disasters can, and do, affect everyone, disadvantaged racial and ethnic communities are often more severely affected and have more difficulty recovering (Fothergill, Maestas, & Darlington, 1999).

One factor affecting people's ability to recover is economic. Members of economically or socially marginalized groups may have greater unmet needs after disasters (International Federation of Red Cross and Red Crescent Societies, 2007). They are more likely to live in vulnerable areas that may be hardest hit when disasters occur, and they are less able to absorb disaster-related economic losses.

In addition, members of ethnic and racial minority groups are often underserved by conventional mental health services (HHS, 2001). After both the 9/11 attacks in 2001 and Hurricane Katrina in 2005, African Americans with significant symptoms of anxiety or depression were less likely than white Americans to receive conventional mental health treatment (Boscarino, Adams, Stuber, & Galea, 2005; Wang et al., 2007). However, these ethnic disparities were somewhat mitigated by the availability of free, community-based crisis counseling. African Americans were as likely as or more likely than white Americans to receive federally funded crisis counseling services following 9/11 (Donohue, Covell, Foster, Felton, & Essock, 2006) and Hurricane Katrina (Norris & Bellamy, 2009).

Another important factor is that recovery occurs in a community context. People coming together, mobilizing social support, and making meaning of the event are important aspects of recovery after a disaster. These social processes are shaped by culture. Mental health services should therefore support and work in consonance with culturally sanctioned processes for recovery (Norris & Alegria, 2005).

PRINCIPLES OF CULTURAL COMPETENCE

Cultural competence is a complex construct that goes beyond mere awareness of and sensitivity to cultural differences. It also involves appreciating and valuing diversity, adapting services to meet client needs, and striving to ensure that service delivery is appropriate to those needs (Sue, 2006). Cross, Bazron, Dennis, and Isaacs (1989) outlined nine principles for culturally competent mental health programs: (1) Recognize the importance of culture and respect diversity; (2) maintain a current profile of the cultural composition of the community; (3) recruit disaster workers who are representative of the community or service area; (4) provide ongoing cultural competence training to disaster mental health staff, (5) ensure that services are accessible, appropriate, and equitable; (6) recognize the role of help-seeking behaviors, customs and traditions, and natural support networks; (7) involve community leaders and organizations representing diverse cultural groups as "cultural brokers"; (8) ensure that services and information are culturally and linguistically competent; and (9) assess and evaluate the program's level of cultural competence.

The CCP has long stressed values of inclusiveness, equity, and flexibility, but it has not always explicitly framed these in terms of cultural competence. SAMHSA has more recently developed a guidance document for developing cultural competence in disaster mental health programs (HHS, 2003). In that document, SAMHSA noted that disaster mental health programs can be viewed along Cross et al.'s (1989) continuum of stages of cultural competence: cultural destructiveness, cultural incapacity, cultural blindness, cultural precompetence, cultural competence, and (ideally) cultural proficiency. However, there are few data on how existing programs are arrayed along that continuum.

CULTURAL COMPETENCE IN CRISIS COUNSELING

Despite the large body of literature with recommendations for programs to improve their cultural competency, there has been only limited research evaluating the effectiveness of such efforts (Bhui, Warfa, Edonya, McKenzie, & Bhugra, 2007). Few studies have empirically examined the cultural competency practices of mental health services (see Stork, Scholle, Greeno, Copeland, & Kelleher, 2001, for an exception).

A qualitative survey of state-level directors of projects funded by federal crisis counseling grants, conducted in parallel with the present study, sought to capture directors' views on a wide range of lessons learned from previous disasters (Elrod, Hamblen, & Norris, 2006; Norris et al., 2005). Although the interviews did not focus specifically on cultural competence, several directors spontaneously discussed it. For example, one director said, "I think having people of the same culture, the same religion, really just gives you more credibility" (Norris et al., 2005, p. c41).

Another director gave an example of adapting to local help-seeking behaviors and customs: "Outreach in a Hispanic neighborhood will work, particularly if you have indigenous people. Door-to-door outreach in the Muslim neighborhood will not work; it's a different approach for different communities" (Elrod et al., 2006, p. 159). Another grantee discussed trial-and-error attempts to ensure that services were linguistically competent for members of a particular immigrant community: "So they developed a [translated] pamphlet ... they realized upon trying to distribute it, that there were very few in the whole population that could read their own language" (Elrod et al., 2006, p. 160).

GOALS OF THE PRESENT STUDY

The present study reviewed archival data from 36 crisis counseling projects to look at the association between the extent to which services were tailored to meet diverse community needs and the actual reach of services provided. This would enable us to test whether projects that reported taking actions to address diverse community needs were more successful in reaching all segments of the affected population.

We coded the narratives of grantees' final reports for mentions of four things: (1) training staff in cultural issues, (2) tailoring services for specific cultural subgroups, (3) adapting services for non-English speakers, and (4) adapting services for people with other specific needs. We created an overall index of tailoring of activities based on how many of these different types of activities projects reported. We then examined project records in terms of reach of services. We used these data to address four questions. First, what proportion of crisis counseling projects reported one or more of the four types of tailoring activities? Second, what community characteristics, types of disasters, and grant types and expenditures were associated with projects reporting more types of tailoring of their services? Third, did projects that reported more types of tailoring reach a higher proportion of clients from racial or ethnic minority groups? Fourth, controlling for expenditures, did projects that reported more types of tailoring reach more clients in total than did those that reported fewer types of tailoring?

METHOD

Sample

This study was part of a retrospective evaluation of CCPs and examined archival data from crisis counseling projects over a five-year period (Norris et al., 2005, p. c41). The three inclusion criteria for the study were as follows: (1) The project closed out between October 1,1996, and September 30, 2001; (2) the grantee was a U.S. state (territories such as Puerto Rico and Guam were not included); and (3) the final report contained information on both the project expenditures (or budget) and reach (either number of customers served or total number of contacts). States can potentially receive two grants for the same disaster. Multiple grants for the same event were treated as separate projects. Characteristics of the projects in the sample are shown in Table 1.

Variables

The present evaluation is informed by a logic model of the CCP (Rosen, Young, & Norris, 2006) that examines program outputs (units of service delivered) as a function of characteristics of the affected community, the nature of the disaster, resources available to the program, and program activities or strategies (see Figure 1). Community characteristics were determined from census data. Population density (log of the average number of people per square mile) was included as a measure of urbanicity. Per capita income was included as a potential indicator of community resources. Proportion of minority residents was the proportion of people in the total population who were African American, Asian American, Hispanic, or Native American. Type of disaster was coded into five categories: fires (10 percent), tornados (30 percent), hurricanes or named tropical storms (15 percent), freezes or winter storms (10 percent), and other floods or storms (35 percent).

Project inputs were tangible and intangible resources available to support service delivery. Grant type referred to whether the award was for a short-term Immediate Services Project (42 percent of projects, n = 15) or a longer Regular Services Project (58 percent, n = 21). Expenditures referred to the total amount of money spent. Because expenditures and budget were highly correlated ([beta] = .98, p < .001), we used regression modeling to impute expenditures for three projects that reported their budgets but not expenditures. The distribution of expenditures was positively skewed (many small projects and a few very big ones), so we used the log of expenditures in all analyses.

[FIGURE 1 OMITTED]

Project activities are the actions performed to realize program goals. In the present study, we focused on the extent to which projects tailored their outreach and counseling activities to diverse community needs. One of the authors, who was blind to data on program reach, coded grantees' final reports for mention of four types of activities: (1) training outreach and counseling staff in cultural issues; (2) adapting services to address the needs of members of specific cultural subgroups; (3) modifying services to accommodate people who do not speak English; and (4) taking steps to address people with other specific needs, such as older adults, children, farmers, and so forth. Each was coded 1 if present or 0 if not mentioned in the final report. These four items were sufficiently correlated (Cronbach's alpha = .72) to be combined into a single index ranging from 0 (mentioned no tailoring activities) to 4 (mentioned all four types of tailoring activities).

Two project outputs were included as dependent variables. The number of total clients who received counseling or educational outreach was determined from the grantees' final reports. Because the number of clients was highly correlated with total contacts (r = .90, p < .001), we imputed number of clients from total contacts for two of the projects. The log of total clients was used in all analyses. The proportion of project clients who were members of racial and ethnic minority groups was determined from program records. Information on client ethnicity (breakdown of African American, Asian American, Hispanic, and Native American clients) was available for 14 projects (39 percent). Four projects (11 percent) reported only the aggregated number of clients who were members of "minority" groups. Eighteen projects (50 percent) reported no information on clients' race or ethnicity. Projects that did and did not report client race or ethnicity were similar in their expenditures, total number of clients served, and proportion of community residents who were members of racial and ethnic minority groups [ts(35) = 0.09 to 0.55, ns].

Data Analysis

We first provided descriptive statistics on the proportions of programs that reported various tailoring activities. We used correlation and linear regression to examine whether community characteristics, type of disaster, grant type, and expenditures predicted the number of types of tailoring activities that programs reported. Variables with significant bivariate associations were included in the regression model predicting level of tailoring. We then used bivariate correlation to identify covariates to be included in our regression models of program reach. Our second regression analysis examined whether the number of types of tailoring activities predicted the proportion of clients who were members of ethnic minority groups, after controlling for significant covariates. The third regression model examined whether the number of types of tailoring activities predicted the total number of clients served by the program, after controlling for significant covariates.

RESULTS

Preliminary Analyses

Type of disaster, population density, and grant type (immediate versus regular services grant) were unrelated to any of the three dependent variables and were not included in further analyses. Scatter plots revealed a nonlinear relationship between tailoring of activities and the proportion of people in the community who were members of racial or ethnic minority groups. On the basis of these plots, the variable indicating the proportion of people in the community who were members of minority groups was dichotomized (<.30 versus [less than or equal to] .30) for some analyses.

Tailoring of Activities

Roughly two-third of projects (64 percent, n = 23) reported instances of adapting their activities to clients from a particular ethnic or cultural group. Such instances included hiring of staff from different racial or ethnic groups, including representatives of diverse ethnic groups on an "Unmet Needs" advisory group; partnering with community organizations such as faith-based groups that have strong community ties; and incorporating culturally sanctioned recovery practices (for example, having a priest bless homes to exorcise the "evil" of a hurricane). Over half of the projects (58 percent, n = 21) mentioned steps to accommodate people who did not speak English. These steps included hiring bilingual staff and translating written materials into clients' languages. Two-thirds (67 percent, n = 24) of projects mentioned steps to address specific needs of other types of clients. Examples of these strategies include engaging children and teenagers in art projects related to the disaster, offering stress management classes for emergency medical technicians or firefighters led by a peer counselor, outreaching to undocumented workers by hiring indigenous counselors and reiterating that services are anonymous, and engaging elders in compiling a book of their survival stories. Only 11 projects (31 percent) mentioned offering diversity-related training to their staff, such as training in cultural competence or in how to work with translators. Combining these results into a single index, six projects (17 percent) reported using all four types of tailoring, 14 (39 percent) reported three kinds of tailoring, five (14 percent) reported two kinds of tailoring, three (8 percent) reported one type of tailoring, and eight (22 percent) reported none of these types of tailoring.

Variables Associated with Greater Tailoring of Activities

Bivariate analyses showed that tailoring was greater in projects with higher expenditures and in projects serving communities where more than 30 percent of the population consisted of members of racial or ethnic minority groups (see Table 2). These variables were included in a hierarchical regression. The full model was significant [[R.sup.2] = .340, F(2, 33) = 8.5, p < .001]. The (dichotomized) proportion of minority group members in the community accounted for 22 percent of the variance in tailoring of activities [F(1, 34) = 9.8, p < .01]. Expenditures accounted for an additional 12 percent of the variance [F(1, 33) = 5.8, p < .01].

Reach to Members of Minority Groups

Among the 18 projects that provided information on client race or ethnicity, the median proportion of clients who were members of racial or ethnic minority groups was .30. This was only slightly lower than the median proportion of people in the community (.33) who were African American, Asian American, Hispanic, or Native American .The proportion of project clients who were members of minority groups was very strongly correlated (r = .88, p <. 01) with the proportion of community residents who were members of minority groups, and it was also correlated with higher expenditures (r = .63,1) <.0 l) and more tailoring of activities (r = .63, p <. 01) (see Table 2). Project expenditures did not add unique variance after controlling for the proportion of minorities in the community, so it was not included in the regression. The regression model explained 85 percent of the variance in the proportion of clients who were members of racial or ethnic minority groups [F(2, 15) = 43.2, p < .001]. The proportion of minority clients was strongly predicted by the proportion of minorities in the community [[R.sup.2] = .781, F(1,16) = 57.1, p < .01), with tailoring of services explaining an additional 7.1 percent of the variance [F(1, 15) = 7.2, p < .05].

These results are shown in Figure 2. Three projects that reported no tailoring were all below the regression line (that is, they reached fewer minorities than one would expect from the proportion of minorities in the community). Projects with high tailoring of activities tended to be at or above the regression line. The only project with high tailoring that fell below the regression line (over 90 percent of the affected population but only 60 percent of clients were minority group members) was a disaster that affected a large number of Native Americans living on tribal reservation lands. These people could only receive project services if the sovereign tribal government requested assistance from the state.

Total Clients Served

Projects in communities where more of the population consisted of members of minority groups, projects with larger expenditures, and projects with greater tailoring of activities tended to serve more clients (see Table 2). The proportion of community residents who were members of racial and ethnic minority groups explained 12 percent of the variance in total clients [F(1, 34) = 4.4, p < .05]. Program expenditures explained another 22 percent of the variance in number of clients served [F(1,33) = 10.8, p < .01]. The effect of minority group members in the community population became nonsignificant after controlling for expenditures. Tailoring of activities explained an additional 15 percent of the variance [F(1, 32) = 9.4, p < .01]. The full model explained nearly half the variance in total clients served [[R.sup.2] = .486, F(3, 32) = 10.0, p < .001].

[FIGURE 2 OMITTED]

These results are graphed in Figure 3. Seven out of eight projects (87 percent) with low tailoring were below the regression line, indicating that they reached fewer clients than would be expected from their expenditures. Only five out of 20 projects (25 percent) with high tailoring were below the regression line.

DISCUSSION

The 36 projects in this review spanned a five-year period and together provided services to roughly 160,000 people. Despite the breadth of these data, we were limited by working with program records that were not intended for research. Estimates of total clients served may be somewhat unreliable because projects may have differed in how they defined service contacts. Half the projects did not report client race or ethnicity, so we can draw no conclusions about their reach to members of racial or ethnic minority groups. Projects that reported client ethnicity based assessment of this on staff members' impressions, not on clients' self-definitions. Also, we were only able to examine reach to "members of minority groups" in aggregate, not through more nuanced questions about how effectively projects reached members of specific community segments.

Available program records also did not provide sufficient information for a thorough evaluation of projects' cultural competence. Our coding of program reports for mention of tailoring strategies was similar to the approach used by Stork et al. (2001) to analyze cultural competence requirements in Medicaid contracts. Nonetheless, it was an imperfect indicator of practices. We best understand our data as indicating that projects were at least at a "precompetent" stage of cultural competence (Cross et al., 1989). Precompetent organizations make some efforts to serve diverse community needs: They "hire staff from the cultures they serve, involve people of different cultures on their boards of directors or advisory committees, and provide at least rudimentary training in cultural differences" (HHS, 2003, p. 13). However, our data cannot show whether projects had evolved even further to reach cultural competence or proficiency. In these stages, organizations go beyond a few limited measures: They engage in ongoing cultural reassessment and strive to continually expand their cultural knowledge and capabilities.

[FIGURE 3 OMITTED]

Despite these limitations, our results provide important evidence that addressing cultural and other differences improves community penetration of mental health services. Projects that engaged in more types of tailoring activities reached more clients than did other projects with similar expenditures. This likely reflects a program value of working to create a more welcoming and open environment rather than any specific set of tactics.

Our second striking finding was grantees' success in reaching members of racial and ethnic minority groups. Although similar findings have been reported from the Project Liberty crisis counseling grant (Donohue et al., 2006), these are sharply divergent from a large body of literature showing that people who belong to racial or ethnic minority groups tend to underuse conventional mental health programs (HHS 2001; Wang et al., 2007). Several aspects of the crisis counseling model may contribute to its more effectively penetrating minority communities. Providing free services removes economic barriers to accessing services. Bringing services into the community, recruiting indigenous community members as counselors, and using normalizing and nonpathologizing language may also help make mental health services more consistent with cultural help-seeking norms (Norris & Alegria, 2005). More research is needed in nondisaster settings to evaluate whether such community-based models of service delivery can reduce racial and ethnic disparities in use of mental health care in other types of programs.

Finally, the present study examined only the reach of crisis counseling, not its quality or outcomes. We cannot assess the degree to which crisis counseling actually benefited clients from different cultural groups. Recent surveys of recipients of federally funded crisis counseling after Hurricane Katrina found that the projects generally got high marks for respect and cultural competence (Norris, Hamblen, & Rosen, 2009). These findings are encouraging but not a cause for complacency: Cultural competence requires continual self-evaluation and ongoing commitment to ensure the equity and cultural suitability of services.

REFERENCES

Bhui, K., Warfa, N., Edonya, R, McKenzie, K., & Bhugra, D. (2007). Cultural competence in mental health care: A review of model evaluations. BMC Health Services Research, 7(15). Retrieved from http://www. biomedcentral.com/1472-6963/7/15

Boscarino, J. A., Adams, R. E., Stuber, J., & Galea, S. (2005). Disparities in mental health treatment following the World Trade Center disaster: Implications for mental health care and health services research. Journal of Traumatic Stress, 18, 287-297.

Cohen, Y. (1984). Residential treatment as a holding environment. Residential Group Care and Treatment, 2, 33-43.

Cross, T. L., Bazron, B. J., Dennis, K. W., & Isaacs, M. R. (1989).Towards a culturally competent system of care: A monograph on effective services for minority children who are severely emotionally disturbed. Washington, DC: Georgetown University, Child Development Center.

Disaster Relief Act of 1974, P.L. 93-288, 88 Star. 143 (1974).

Donohue, S. A., Covell, N. H., Foster, M. J., Fehon, C. J., & Essock, S. M. (2006). Demographic characteristics of individuals who received Project Liberty crisis counseling services. Psychiatric Services, 57, 1261-1267.

Elrod, C. L., Hamblen, J. C., & Norris, F. H. (2006). Challenges in implementing disaster mental health programs: State program directors' perspectives. Annals of the American Academy of Political and Social Science, 604, 152-170.

Fothergill, A., Maestas, E.G.M., & Darlington, J. D. (1999). Race, ethnicity, and disasters in the United States: A review of the literature. Disasters, 23, 156-173.

Gould, K. H. (1988). Asian and Pacific Islanders: Myth and reality. Social Work, 33, 142-147.

Hernandez, M., Nesman, T., Isaacs, M., Callejas, L. M., & Mowery, D. (Eds.). (2006). Examining the research base supporting culturally competent children's mental health services (FMHI Publication No. 240-1).Tampa: University of South Florida, Louis de la Parte Florida Mental Health Institute, Research and Training Center for Children's Mental Health.

Hernandez, M., Nesman, T., Mowery, D., Acevedo-Polakovich, I. D., & Callejas, L. M. (2009). Cultural competence: A literature review and conceptual model for mental health services. Psychiatric Services, 60, 1046-1050.

Ida, D. J. (2007). Cultural competency and recovery within diverse populations. Psychiatric Rehabilitation Journal, 31, 49-53.

International Federation of Red Cross and Red Crescent Societies. (2007). World disasters report 2007--Focus on discrimination. Retrieved from http://www.ifrc.org/ publicat/wdr2007/index.asp

Norris, F. H., & Alegria, M. (2005). Mental health care for ethnic minority individuals and communities m the aftermath of disasters and mass violence. CNS Spectrums, 10, 132-140.

Norris, F. H., & Bellamy, N. (2009). Evaluation of a national effort to reach Hurricane Katrina survivors and evacuees:The Crisis Counseling Assistance and Training Program. Administration and Policy in Mental Health and Mental Health Services Research, 36, 165-175.

Norris, F. H., Hamblen, J. L., & Rosen, C. S. (2009). Service characteristics and counseling outcomes: Lessons from a cross-site evaluation of crisis counseling after Hurricanes Katrina, Rita and Wilma. Administration and Policy in Mental Health and Mental Health Services Research, 36, 176-185.

Norris, F. H., Rosen, C. S., Elrod, C. L., Young, H. E., Hamblen, J. L., & Gibson, L. (2005). Retrospective 5-year evaluation of the Crisis Counseling Program. White River Junction, VT: National Center for Posttraumatic Stress Disorder.

Rosen, C. S, Young, H. E., & Norris, F. H. (2006). On a road paved with good intentions, you still need a compass: Monitoring and evaluating disaster mental health services. In E. C. Ritchie, P. J. Watson, & M. J. Friedman (Eds.), Interventions following mass violence and disasters (pp. 206-226). New York: Guilford Press.

Stork, E., Scholle, S., Greeno, C., Copeland, V. C., & Kelleher, K. (2001). Monitoring and enforcing cultural competence in Medicaid managed behavioral health care. Mental Health Services Research, 3, 169-177.

Sue, S. (2006). Cultural competency: From philosophy to research and practice. Journal of Community Psychology, 34, 237-245.

U.S. Department of Health and Human Services. (2000). Training manual for mental health and human service workers in major disasters (DHHS Publication No. ADM 90-538). Rockville, MD: Center for Mental Health Services, Substance Abuse and Mental Health Services Administration.

U.S. Department of Health and Human Services. (2001). Mental health: Culture, race, and ethnicity--A supplement to Mental Health: A Report of the Surgeon General (DHHS Publication No. SMA-01-3613). Rockville, MD: U.S. Department of Health and Human Services, Public Health Service, Office of the Surgeon General.

U.S. Department of Health and Human Services. (2003). Developing cultural competence in disaster mental health programs: Guiding principles and recommendations (DHHS Publication No. SMA 3828). Rockville, MD: Center for Mental Health Services, Substance Abuse and Mental Health Services Administration.

Wang, P. S., Gruber, M.J., Powers, R. E., Schoenbaum, M., Speirer, A. H., Wells, K.B., & Kessler, R. E. (2007). Mental health service use among Hurricane Katrina survivors in the eight months after the disaster. Psychiatric Services, 58, 1403-1411.

Craig S. Rosen, PhD, is deputy director, Dissemination and Training Division, National Center for PTSD, Department of Veterans Affairs (VA) Palo Alto Health Care System, Palo Alto, CA, and associate professor, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA. Carolyn J. Greene, PhD, is a project director and clinical psychologist, Dissemination and Training Division, National Center for PTSD, VA Palo Alto Health Care System. Helena E. Young, PhD, is a clinical psychologist, VA Palo Alto Health Care System. Fran H. Norris, PhD, is a research associate, Executive Division, National Center for PTSD, White River Junction VA Medical Center, White River Junction, VT; director, National Center for Disaster Mental Health Research; and research professor, Departments of Psychiatry and of Community and Family Medicine, Dartmouth Medical School, Hanover, NH. Address correspondence, to Craig S. Rosen, 795 Willow Road (PTSD-334), Menlo Park, CA 94025; e-mail: craig. rosen@va.gov. This work was supported in part by Interagency Agreement #AMO3C8400A awarded to the National Center for PTSD by the Center for Mental Health Services, Substance Abuse and Mental Health Services Administration. The views expressed in this article are those of the authors and not necessarily those of the Department of Veterans Affairs or the Substance Abuse and Mental Health Services Administration.

Original manuscript received February 14, 2008

Final revision received November 20, 2009

Accepted April 15, 2010
Table 1: Community and Project Characteristics
(N = 36 Projects)

Variable                             Minimum     Maximum

Community
  Population of affected
    counties (thousands)              13         6,101
  Population density
    (people per square mile)           3         1,797
  Per capita income ($1,000)           6.3          28.5
  Minority population (%)              3.00         96.00
Input: Expenditures ($1,000)          12         2,304
Activity: Tailoring of activities      0.0           4.0
Output: Minority clients (%)           1            69
Total clients served                 204        28,137

Variable                                M          SD

Community                                          --
  Population of affected
    counties (thousands)             1,501      1,864
  Population density
    (people per square mile)           304        470
  Per capita income ($1,000)            19.3        4.5
  Minority population (%)                0.36       0.24
Input: Expenditures ($1,000)           411        561
Activity: Tailoring of activities        2.2        1.4
Output: Minority clients (%)            30         23
Total clients served                 4,497      5,693

Table 2: Correlations between Study Variables

Variable                     1         2          3         4

1. Tailored activities
2. Per capita income      .24       --
3. Minority population    .25       -.47 ***   --
4. Minority population
     (dichotomized)       .47 ***    .01       .77 ***   --
5. Expenditures (log)     .51 ***   -.08       .32 *     .43 ***
6. Minority clients (a)   .63 ***   -.18       .88 **    .90 ***
7. People served (log)    .64 ***    .28 *     .15       .34 **

Variable                     5         6          7

1. Tailored activities
2. Per capita income
3. Minority population
4. Minority population
     (dichotomized)
5. Expenditures (log)     --
6. Minority clients (a)   .63 ***   --
7. People served (log)    .57 ***   0.38       --

Note: Tailored activities = index (0 to 4) of degree to which program
tailored activities to specific community groups; Per capita income =
per capita income in affected counties; Minority population =
percentage of non-Whites and/or Latinos in the affected counties;
Minority population (dichotomized) = more or less than 30 percent
non-Whites and/or Latinos in the affected counties; Expenditures
(log) = log of funds spent by the project; Minority clients =
proportion of non-Whites and/or Latinos among program clients;
People served (log) = log of total number of clients served.

(a) n = 18.

* p < .10. 1 ** p < .05. *** p < .01.
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.