Direct-care paraprofessionals' recognition of psychopathology of varying severity in persons with mental retardation was evaluated. Factors that may influence paraprofessionals' decisions to initiate referrals for mental health services on behalf of individuals with mental retardation were also evaluated. Results suggest that staff members recognized and differentiated psychopathology of varying levels of clinical severity. Results also suggest that paraprofessionals are more likely to initiate making a referral when professionals are perceived as being competent in treating individuals with mental retardation, and when providers' interventions are consistent with the referring agency's philosophy.
Recent literature has documented that persons with mental retardation experience high rates of mental illness (Borthwick-Duffy, 1994; Borthwick-Duffy & Eyman, 1990; Butz, Bowling, & Bliss, 2000). Elevated co-morbidity rates between mental retardation and mental health difficulties mark the particular need for available mental health services appropriate to individuals with developmental disabilities. However, because adults with mental retardation are often unlikely to make self-referrals for needed mental health services (Edelstein & Glenwick, 1997, 2001), caregivers are often responsible for recognizing psychopathology in such individuals, and for initiating the referral-making process when treatment is thought to be necessary.
Direct-care paraprofessionals are frequently involved in providing residential supports for adults with mental retardation living in the community setting. Such paraprofessionals assume a variety of caregiver responsibilities, such as administering medications, assisting consumers with self-help needs, accompanying consumers to and from various community activities, and assisting consumers in the development of other daily adaptive behaviors. Nonetheless, direct-care paraprofessionals who frequently have the most consistent interaction with persons who have mental retardation may be ill-prepared to differentiate mental health disorders from nonclinical problems in adaptive functioning (Borthwick, 1988; Edelstein & Glenwick, 2001; Oliver, Leimkuhl, & Skillman, 2003). Indeed, many paraprofessionals report that they have not received training regarding mental illness in persons with mental retardation or feel that the training they received is inadequate (Oliver et al., 2003). Moreover, although research suggests that mental health treatments can be effective for persons with mental retardation (Rush & Frances, 2000), such treatments may be underutilized due to (a) limited awareness and recognition of mental health needs and (b) the perception that treatments may not benefit these individuals (Edelstein & Glenwick, 2001; Reiss, Levitan, & Szyszko, 1982). Yet, because persons who are referred for treatment are more likely to receive mental health services than those who are not, it is important to understand the factors influencing paraprofessionals' decisions to initiate mental health referrals.
In the research conducted to date, a number of factors have been found to influence paraprofessionals' ability to identify psychopathology and need for treatment for individuals with mental retardation. These factors include the type of problem displayed and the clients' level of mental retardation (Edelstein & Glenwick, 2001). Paraprofessionals' work setting and client gender were not found to be related to staff members' recognition of need for psychopathology or treatment (Edelstein & Glenwick, 2001). The current literature is, nonetheless, limited in that researchers have evaluated direct-care paraprofessionals' perceptions of only a narrow range of psychopathology, typically focused on externalized problem behavior. Moreover, the relationship between staff initiating the referral process and factors such as availability of on-site resources, attitudes toward psychological and medical treatments, and perceived access to professionals familiar with disability issues have not been evaluated. A greater understanding of direct-care paraprofessionals' interpretations of clinical disorders, recognition of need for treatment, and of their decision-making process in referring adults with mental retardation exhibiting psychopathology is needed to improve access to mental health services for this population.
Our purpose in this study was to survey direct-care paraprofessionals in terms of how they clinically interpret various vignettes of adults who have mental retardation and who are experiencing a broad range of mental health problems. In this study we address limitations to the current literature by systematically varying clinical vignettes on psychopathology severity and on level of mental retardation. We also evaluated potential factors that may be related to paraprofessionals' decisions to make referrals to mental health professionals.
We solicited paraprofessional participants who worked directly with adults who had mental retardation in either the residential or vocational setting. They were chosen from a potential pool of 375 eligible direct-care paraprofessionals working in four community-based agencies; 109 of the prospective respondents (29%) returned completed surveys, whereas individual agency response rates varied between 22% and 34%. Two of the agencies, which employed approximately 100 and 150 direct-care paraprofessionals, provided residential and vocational services to adults with borderline to profound mental retardation. Consumers receiving support services from these agencies lived in group home facilities, independent community residences, or other community-based living alternatives. Another agency, which employed 75 eligible paraprofessionals, provided only residential services to adults with borderline to severe mental retardation. The final agency, which employed 50 eligible paraprofessionals, provided community-based vocational and residential support services to adolescents and adults with physical and/or cognitive disabilities. Individuals with cognitive disabilities typically had moderate to profound mental retardation. Consistent with the other sites, only paraprofessionals working directly with adults who had mental retardation were surveyed. All four agencies were located in the same rural region of the Midwestern United States.
The survey used in the present study consisted of four sections. The first included demographic items (i.e., gender, ethnicity, age, and level of education) and items reflecting the participants' work history and current job description. The second included 12 randomly ordered vignettes, each containing potential mental/behavioral health situations of varying clinical severity. Nine of the 12 vignettes represented a broad scope (including both internalizing and externalizing disorders) of mental health problems ranging in clinical severity (e.g., high, moderate, low); 3 of the vignettes were nonclinical in nature. Each clinical vignette contained a cluster of disorder-specific symptoms selected primarily from the Reiss Screen for Maladaptive Behavior (Reiss, 1994), an instrument designed to evaluate psychopathology in persons with mental retardation.
The definitions of severity used for the hypothetical scenarios were adapted from descriptions used in Edelstein and Glenwick's (2001) article in which they evaluated paraprofessionals' attributions of psychopathology in adults with mental retardation. High clinical severity indicated that immediate stabilizing medication may be indicated and that an inpatient or controlled environment was recommended; moderate severity, that maintenance medication may be indicated and that outpatient treatment was recommended; low clinical severity, that medication was not likely indicated but outpatient treatment was recommended; and nonclinical scenarios were described as requiring no intervention beyond general and ordinary support.
To ensure vignette clinical severity validity, we conducted survey piloting by having 10 licensed psychologists (5 working in the inpatient setting and 5, in the outpatient setting) rank 14 potential vignettes for severity. Twelve of the vignettes were selected because they maintained an agreement coefficient of .80 or higher across the 10 raters (see Appendix A). In order to evaluate possible differences in clinical severity between levels of mental retardation, we created two versions of the survey where vignettes were alternated by level of functioning. Thus, across the two survey forms, each vignette was rated for adults with borderline/mild mental retardation and for adults with moderate/ severe mental retardation.
In the third section of the survey, paraprofessionals were asked to rate how important a variety of potential factors would be in their decisions to initiate a referral for mental health services outside their agency. Prospective survey factors were identified from the literature and from a focus group of eight staff conducted at one of the participating agencies. Sixteen of the most common factors were conceptually summarized into four factor categories: consumer behavior, staff concerns, program/provider philosophy, and resources. An additional factor category was included in this section to inquire whether having an onsite professional (i.e., psychiatrist, psychologist, or behavior specialist) would influence paraprofessionals' decision to initiate an off-site referral. The importance of each factor was rated on a 4-point Likert scale, ranging from 0 (unimportant) to 3 (highly important).
In the final survey section, staff members were asked to rate how beneficial therapy and medications may be in treating the mental health needs of adults with varying levels of mental retardation. Moreover, paraprofessionals were asked whether they had received any formal training specific to recognizing mental illness and initiating mental health service referrals.
Program directors of each prospective agency were contacted and invited to have their agencies and direct-care staff participate in this study. Once permission was obtained, program directors were asked how many direct-care paraprofessionals worked with adults who had mental retardation. Accordingly, the authors sent agencies a survey packet for each eligible staff member. Because two survey versions were used to evaluate vignette severity ratings by consumer level of functioning, a roughly equivalent number of survey packet forms were randomly distributed to individual paraprofessionals by staff managers, who were also instructed to ask participating direct-care paraprofessionals to return the completed survey within 2 weeks of receipt. Paraprofessionals were also informed in a cover letter that participation was voluntary, anonymous, and that they could return their survey directly to the researchers in the provided self-addressed, stamped envelope.
To determine whether combining the samples of staff supporting persons with mental retardation would be appropriate for the primary analyses, we conducted a preliminary analysis evaluating possible demographic differences among the four agencies. Although agencies differed in participant work setting (i.e., residential, vocational), χ2(1, 6, N = 108) = 45.00, p < .001, participant age, F(3, 103) = 7.05, p < .001, and overall work experience with persons who have mental retardation, F(3, 104) = 11.77, p < .001, they did not differ in terms of gender, race, or education level. In a follow-up analysis, we found that paraprofessionals' ratings of referral factor categories and severity ratings did not differ statistically by participant work setting, and work experience was not significantly correlated with paraprofessionals' ratings of referral factor categories. However, age and work experience, which were correlated (.56) at the .05 level, were found to be related to paraprofessionals' ratings of psychopathology severity. That is, age was correlated with responses for moderate severity, r = .25, and high severity, r = .27, vignettes, whereas work experience was correlated with responses for nonclinical, −.31, and low severity, −.22, vignettes at the .05 level. Because participant age and work experience were conceptually and metrically related, and because work experience was slightly more related to severity ratings than to age, we considered the agency samples collectively in the following analyses where, when relevant, work experience was used as a covariate. Table 1 shows relevant demographic information for the overall sample.
Responses to Clinical Vignettes
To evaluate how direct-care paraprofessionals rated clinical vignettes ranging in severity (i.e., nonclinical to high clinical severity) for persons with borderline to mild mental retardation or moderate to severe mental retardation, we computed a 2 (levels of functioning) × 4 (levels of severity) mixed-subjects factorial analysis of covariance using the general linear method, where participant work experience (in months) was used as a covariate. Results indicate that severity ratings did not differ by level of functioning (i.e., whether the individual was described as having borderline to mild mental retardation or moderate to severe mental retardation). Furthermore, the main effect for level of functioning was not significant. However, the effect for clinical severity, after controlling for work experience, was found to be significant, Wilks' λ = .842, F(3, 95) = 5.93, p < .001. Pairwise comparisons indicate that all severity categories (e.g., high, M = 3.56, standard deviation [SD] = .36; moderate, M = 2.84, SD = .38; low, M = 2.05, SD = .54) and nonclinical (M = 1.30, SD = .42) differed from each other beyond the .001 level. Means and SDs for each vignette by severity are shown in Table 2.
In addition to evaluating the mean differences across severity ratings, it is helpful to evaluate the frequencies of the severity ratings for the a priori severity categories. That is, it may be beneficial to evaluate how often staff members endorsed the expected severity categories as established by licensed mental health professionals. Table 3 presents a response frequency table with cell percentages. Findings suggest that although paraprofessionals most frequently rated clinical severity as expected, there was substantial variance. For example, although the majority of paraprofessionals appropriately viewed the vignette of a consumer looking forward to smoking during a work break as nonclinical, approximately 31% of respondents viewed this scenario as warranting clinical attention. Moreover, there appeared to be substantial variability in paraprofessionals' ratings of treatment need for low-severity vignettes, such as dependent personality traits where approximately half the sample viewed the vignette as nonclinical.
Referral Initiation Factors
In the next analysis we evaluated the perceived importance of various factors on how much they were likely to influence paraprofessionals' decisions when determining whether they should initiate making a referral outside their organization for some type of mental health treatment (e.g., behavioral support, psychotherapy, medication). A mean ranking for referral factor categories indicated that accessible resources (M = 2.45, SD = .53), program/ provider philosophy (M = 2.25, SD = .56), consumer behavior (M = 2.04, SD = .55), staff concerns (M = 2.03, SD = .60), and on-site professionals (M = 1.75, SD = .86) were all at least somewhat important in making referral decisions. The means and SDs of individual factors were ranked by their perceived importance (see Table 4).
We used an exploratory analysis to evaluate the perceived benefits of medications and psychotherapy offered as treatments to persons with varying levels of mental retardation. Staff generally viewed both medications (M = 2.27, SD = .97) and psychotherapy (M = 2.17, SD = .79) as being helpful treatments (as measured on a 6-point scale ranging from 1, highly helpful, to 6, highly unhelpful) for adults with borderline to mild mental retardation and mental illness. Similarly, paraprofessionals generally viewed medications (M = 1.94, SD = .95) and psychotherapy (M = 2.68, SD = 1.09) as potentially being helpful to somewhat helpful in treating adults with moderate to severe mental retardation and mental illness. Because each participant rated the relative value of the two treatments for both groups of higher and lower functioning individuals, we ran paired sample t tests to evaluate whether participants differentially viewed the benefits of medications and psychotherapy. Both t tests were significant, indicating that participants viewed psychotherapy as being more beneficial for adults with borderline to mild mental retardation, t(1, 104) = −4.36, p < .05, than for those with moderate to severe mental retardation, but viewed medications as being more beneficial for adults with moderate to severe mental retardation, t(1, 106) = 3.24, p < .05, than for adults with borderline to mild mental retardation. When asked whether paraprofessionals had ever been involved in referring a person with mental retardation and mental illness to receive mental health services, approximately 32% of responding staff members answered affirmatively. They were also asked about the level of training received from their agency on recognizing signs of mental illness or how to initiate a mental health treatment referral. Approximately 42% of respondents reported receiving training on recognizing mental health problems, whereas only about 20% reported receiving training on initiating some type of referral process for external services.
Understanding how paraprofessionals recognize and handle mental illness for individuals with mental retardation is important because paraprofessionals frequently serve as caregivers assuming the primary responsibility for consumers' welfare. In this study we evaluated whether direct-care paraprofessionals would be able to identify varying levels of psychopathology in adults with mental retardation when presented with a series of vignettes. Further, we explored whether they would differentially rate various factors as being influential in their decision to initiate referrals for mental health treatment for persons with mental retardation. Finally, we evaluated whether direct-care workers would display attitudinal differences toward the effectiveness of medicinal versus psychological treatments when used in the mental health treatment of persons with varying levels of mental retardation.
The results of this study supported the hypothesis that paraprofessionals working with individuals who had mental retardation could recognize and differentiate varying levels of clinical severity. However, in contrast to an earlier study (Edelstein & Glenwick, 2001), the current findings did not support the notion that paraprofessionals differentiate clinical severity of mental health difficulties by the severity of mental retardation, although they tended to differentially view the benefits of psychotherapy and medication for treating mental illness in persons with different levels of mental retardation. These results suggest that although staff were generally able to identify the level of clinical severity of a scenario in relation to treatment need, it is not clear to what extent they consider the functioning level of the individual. One issue that should be considered in light of the nonsignificant interaction between level of functioning and severity ratings is that participating agencies supported somewhat different populations, and it is unclear whether all participants had direct exposure to consumers with borderline to severe mental retardation. Future researchers would do well to further explore this potential moderating relationship.
In this study we found that there is substantial variability across staff responses for various types of problems, which suggests that paraprofessionals may be prone to under recognize or over pathologize mental health difficulties among persons with mental retardation. Such findings highlight the need for further paraprofessional training in which the following issues are specifically addressed: the manifestations of psychopathology in persons with mental retardation, the differences in such manifestations across differing mental retardation functioning levels, and strategies geared to help staff recognize basic behavioral benchmarks that indicate additional treatment may be necessary.
Results also suggest that a variety of factors are likely to be important to paraprofessionals' decision to make a referral for mental health services. Such concerns include the competence of available professionals, extreme problem behavior, staffing resources, the accessibility of provider facilities, the providers' experience with mental retardation, and whether staff members recognize the treatment need. Given that the perceived competence of paraprofessionals and limited resources (i.e., staff time) were ranked among the highest potential barriers, it would be beneficial for agency administrators to network and educate providers about mental retardation as well as make accommodations such that staffing time or transportation do not keep consumers from receiving needed services.
Finally, the results suggest that training on recognizing mental/behavioral health problems and initiating treatment referrals is not formally done on a routine basis, and most paraprofessionals have not been directly involved in initiating a mental health referral. It would be helpful if agency training programs incorporated instruction as to what paraprofessionals should do if they suspect that mental health services may be appropriate for a particular consumer. Consistent with the observations of Edelstein and Glenwick (2001), results suggest that training programs that help staff members recognize the signs of internalized disorders are particularly needed because they were observed to underappreciate the clinical significance in such scenarios.
The results of this study should be considered in the context of a number of limitations. First, although the 29% return rate of paraprofessionals supporting persons with mental retardation is comparable to similar field-based survey research, it is unclear how generalizable the results are to other paraprofessionals working in residential and vocational settings. It is possible that the sample is biased in the direction of staff who are more interested in, concerned about, or who have experienced mental illness. Future researchers should evaluate the potential impact that paraprofessionals' exposure to mental health problems and treatment services has on their referral practices. Second, because the data were collected through a survey in which we used hypothetical clinical vignettes, it should not be assumed that the findings necessarily represent paraprofessionals' direct experience with these potential scenarios. Finally, although vignette severity scaling was rated and established by a panel of licensed psychologists, vignettes were written in such a way that their severity could be ambiguous. For example, one vignette showed a consumer who “worried excessively” and manifested somatic symptoms that, although rated as mildly severe, could potentially be rated as highly severe if more information were available and revealed that the level of impairment indicated inpatient treatment.
Although direct-care paraprofessionals appear to recognize a broad variety of clinical problems of varying severity, there is a need for further attention addressing how such paraprofessionals recognize and handle potential mental health problems of the consumers they support. Researchers exploring how agencies can minimize service barriers and better ensure adequate access to mental health services for consumers is also needed. Such research would be particularly helpful in guiding training policy development in agencies supporting persons with mental retardation and mental illness.
This research was supported in part by grant funding from the South Dakota Center for Disabilities.
Authors: Matthew N. I. Oliver, MA, Doctoral Student, Trisha T. Miller, MA, Doctoral Student, and Gemma D. Skillman, PhD, Associate Professor, Department of Psychology, The University of South Dakota, 414 E. Clark St., Vermillion, SD 57069. email@example.com