Health Risks and Changes in Self-Efficacy Following Community Health Screening of Adults with Serious Mental Illnesses

Physical health screenings were conducted by researchers and peer wellness specialists for adults attending publicly-funded community mental health programs. A total of 457 adults with serious mental illnesses attended health fairs in 4 U.S. states and were screened for 8 common medical co-morbidities and health risk factors. Also assessed were self-reported health competencies, medical conditions, and health service utilization. Compared to non-institutionalized U.S. adults, markedly higher proportions screened positive for obesity (60%), hypertension (32%), diabetes (14%), smoking (44%), nicotine dependence (62%), alcohol abuse (17%), drug abuse (11%), and coronary heart disease (10%). A lower proportion screened positive for hyperlipidemia (7%). Multivariable random regression analysis found significant pre- to post-screening increases in participants’ self-rated abilities for health practices, competence for health maintenance, and health locus of control. Screening identified 82 instances of undiagnosed diabetes, hypertension or hyperlipidemia, and 76 instances where these disorders were treated but uncontrolled. These results are discussed in the context of how this global public health approach holds promise for furthering the goal of integrating health and mental health care.


Introduction
People with serious mental illnesses experience multiple health disparities [1], higher medical morbidity [2], and lifespans 10-30 years shorter on average than the general population [3]. These disparities exist in countries with universal access to healthcare and in countries with healthcare systems that are highly regarded [4]. Yet we have little detailed knowledge about the prevalence of medical co-morbidities in subgroups of this population, such as those receiving public mental health services in outpatient settings [5][6]. We report on a project conducted collaboratively by a university research center and a peer health and wellness promotion general U.S. population of non-institutionalized adults; and 2) that health self-efficacy would improve following screening.

Study population and screening procedures
Eligibility criteria included serious mental illness as defined by U.S. federal Public Law 102-321 [59] to include a DSM-IV-TR [60] diagnosis of mental illness accompanied by moderate to severe functional impairment; age 18 years or older; status as a client of the participating community mental health program; and ability to provide informed consent. The study used a "controlled intervention" design common in health screening research [30] that targeted specific communities, in this case defined as clients of particular mental health organizations served in programs for people with serious mental illness. The first screening was held at a peer-operated self-help program in New Brunswick, NJ, and attended by members of peer-run centers across the state. The second took place in the gymnasium of a university in Chicago, IL, and was attended by clients of a psychiatric rehabilitation agency located city-wide. The third was held in an Elk's Club Lodge and a church auditorium, and attended by clients of mental health agencies in Frederick and Rockville, MD (respectively). The fourth occurred at a large community mental health agency located in Smyrna, GA. Participants were recruited by program staff through clinical referral, flyers, announcements at membership meetings, waiting room posters, and word-of-mouth. While clinicians helped with recruitment, participation was strictly voluntary, and clinicians were not informed about whether their clients attended the health fairs. The University of Illinois at Chicago (UIC) Institutional Review Board (IRB) and the Sheppard Pratt Health System IRB approved this study including its informed consent procedures. Written informed consent was obtained from participants on the day of the screening by trained researchers, using procedures approved by the UIC and Sheppard Pratt IRBs. Capacity to provide informed consent was assessed by trained researchers with questions that assessed the person's understanding of the purpose of the research and foreseeable risks and anticipated benefits of study participation. No surrogate consent procedure was used. Only 2% (7 of 464) of those attending the screenings refused participation.
Following informed consent, participants received a number-coded "health passport" in which their test results were recorded, along with a wrist band containing a corresponding number. This allowed individuals to be matched with test results without collecting their names. Next, participants completed a pre-test interview, as described below. They then visited each of the testing stations, at which screening staff performed tests and recorded results in the health passport, and on a coded sheet maintained at each station.
After testing, participants met with trained peer health and wellness specialists who reviewed their test results, provided health education, and offered peer support. Following this meeting, participants completed a post-test interview while the screening results from their health passports were being entered into a data base. These results were later verified against those maintained at testing stations. At check-out, health passports were returned to participants by matching the numeric code on their wrist bands. They received gift cards from local stores and gift bags containing wellness products to thank them for their time.
Given prior studies showing that health fair participants increased their knowledge about health risks and adopted new health behaviors, fairs were designed to promote these processes. Screening activities allowed for dialogue between staff and participants at each station, including immediate feedback about test results, education regarding their meaning, chances to ask questions, and one-page handouts for later reference. At the final station, they met with peer health and wellness specialists who reviewed test results in their entirety, answered questions, and engaged in dialogues about potential strategies for addressing health risks. Peer staff focused on what participants could do in the immediate future, using a handout containing potential actions for different health risks. Finally, participants received a list of sources of free health care and further testing in their local communities, along with encouragement to follow up.
Screening staff included medical students, psychiatry residents, researchers, and peer specialists, each of whom completed a minimum of 6 hours of education for each test administered, including observation and corrective feedback. All serologic testing was conducted by licensed nurses or other medically trained staff. Throughout the screening process and afterwards, support and reassurance was available from the peer health and wellness specialists. To further increase participants' comfort levels, additional peers from each local agency were recruited, trained, and paid to work as health fair greeters and logistics coordinators, answering questions and guiding participants between screening stations.

Measures
Attitudinal measures. A standardized research protocol was administered at pre-and post-screening by trained interviewers to collect demographic information, medical history, health attitudes, and use of health care services. Ability to engage in health maintenance practices was assessed by the Responsible Health Practices subscale of the Self-Rated Abilities for Health Practices Scale [61]. Participants responded to 7 items such as "I recognize what symptoms should be reported to a doctor or nurse" using a 5-point Likert scale ranging from 0 (not at all) to 4 (completely) and referring to the present time. Inter-item reliability for this measure was good with alpha = .82 for pretest and .84 for post-test. Competence in health self-management was assessed using the Perceived Competence for Health Scale [62], adapted from a scale measuring self-management of diabetes. Participants responded to 4 items such as "I am capable of handling my health needs now" using a 7-point Likert scale ranging from 1 (not at all) to 7 (very true) and referring to the present time. Inter-item reliability for this measure was good with alphas of .82 for pre-test and .86 for post-test. The Multidimensional Health Locus of Control Scale (HLOC) [63] is comprised of 3 subscales that measure the degree to which individuals feel they can affect their health and health-related behaviors. The Internal subscale assesses respondents' degree of perceived control over their own health and consists of 6 items such as "I am directly responsible for my own health" using a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree) and referring to the present time. The Powerful Others subscale assesses the degree to which respondents believe their health is controlled by other individuals and includes 6 items such as "Health professionals control my health" rated using the same 6-point scale and referring to the present. The Chance subscale assesses the degree to which respondents believe their health is due to chance factors and includes 6 items such as "Most things that affect my health happen to me accidentally" using the same 6-point scale and referring to the present time. Pre-and post-test alphas for these subscales ranged from .60 to .74.
Prevalence of physical health conditions was assessed using items from the National Health Interview Survey (NHIS) [64] and the National Health and Nutrition Examination Survey (NHANES) [65]. Respondents were asked whether they had ever been told by a doctor or other health professional that they had a series of medical conditions and, if so, whether they still had the condition and were in treatment for it.
Health risk assessments. Measures of health risk included body mass index (BMI) based on height and weight; blood pressure; blood glucose; and non-fasting lipid profile including total cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL) and triglycerides. Glucose and lipid profiles were obtained using the Cholestech LDX system (Inverness Medical, Hayward, CA). Risk for alcohol abuse/dependence was assessed using the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) [66]; risk for substance abuse/ dependence using the Drug Abuse Screening Test (DAST) [67]; and nicotine dependence using the Fagerstrom Test for Nicotine Dependence [68]. Estimated 10-year risk of coronary heart disease was determined using the Framingham risk model [69] and calculated at the Cybermed website (Cybermed, 2000) based on the participant's age, sex, systolic blood pressure, total and HDL cholesterol, diabetes mellitus, and current smoking status. Using the tertiles of their Framingham risk score, participants were classified as being at "high," "intermediate," or "low" risk for heart disease.
Background characteristics. Participants self-reported demographic and other personal characteristics during the interviews including age, race/ethnicity, education, psychiatric diagnosis, and health insurance status.

Statistical analysis
Data were downloaded into SPSS [70] and descriptive statistics (means, medians) were calculated to create prevalence scores. Risk for each condition was defined according to standardized indicators and cut-off values contained in evidence-based practice guidelines from the National Institutes of Health (NIH) and Centers for Disease Control and Prevention (CDC). Random regression analysis using SuperMix [71] tested for pre-to post-screening changes in participants' perceived health efficacy and competency, controlling for person-level factors of sex, race, age, and education, as well as study site. Table 1 presents participants' demographic and clinical characteristics. Comparison to adult populations at each agency revealed no significant differences by sex, race, Hispanic/Latino ethnicity, education, age, diagnosis, and health insurance status with two exceptions. In NJ, a smaller proportion of males were screened (46%) than in the agency population (59%) (Z = -2.1, p<.05). In GA, more screening participants reported schizophrenia (40%) than the agency population (25%) (Z = 3.3, p<.001). Otherwise, screening participants were representative of each population targeted.
Compared to prevalence reported for the non-institutionalized adult U.S. population, proportions were higher for all but one of the conditions screened. Sixty percent of outpatients with mental illnesses were obese compared to 36% in the general population [72]. Fourteen percent had HbA1c values indicating diabetes compared to 2% of the general population [73]. Thirtytwo percent had blood pressure indicating hypertension compared to 29% in the general population [74]. Forty-four percent were current smokers compared to 19% in the general population [75], and nicotine dependence was medium to high in 62% compared to 57% in the general population [76]. Seventeen percent screened positive for alcohol abuse based on their current consumption compared to 7% prevalence in the general population [77]. Eleven percent screened positive for substance abuse compared to 3% prevalence in the general population [77]. Finally, 10% of participants screened high for coronary heart disease risk compared to their age and sex cohorts, while 3% of the general population was found to be at risk [78]. The one exception to this pattern was total cholesterol, with hyperlipidemia being less prevalent among participants with serious mental illnesses (7%) compared to the general U.S. population (13%) [79].

Undiagnosed, untreated, and unsuccessfully treated conditions
Using items from the NHANES III [80], we determined the proportion of participants ever diagnosed with three of the medical conditions for which they were screened, as well as the proportion currently being treated. Among the 62 participants with HbA1c levels indicating diabetes (6.5), 24% (n = 15) said they had not received a diagnosis of diabetes. Among the 96 participants reporting a diabetes diagnosis, 10% (n = 10) did not report current treatment. Among the 86 reporting current treatment, HbA1c levels indicated diabetes in 50% (n = 43). Thus, the screening identified 15 individuals with undiagnosed diabetes, and 43 with treated but uncontrolled diabetes.
Among the 145 who screened positive for high blood pressure (BP140/90), 43% (n = 60) said they had not received a diagnosis of hypertension. Among the 197 participants reporting a diagnosis of hypertension, 24% (n = 47) did not report current treatment. Among the 150 individuals reporting current treatment, high blood pressure was detected among 41% (n = 61). Thus, 60 individuals had undiagnosed hypertension, and 61 had treated but uncontrolled hypertension.
Among the 28 participants who screened positive for high cholesterol (240 mg/dL), 25% (n = 7) said that they had not been diagnosed with hyperlipidemia. Among the 198 participants reporting a diagnosis of high cholesterol, 32% (n = 64) did not report receiving current treatment. Among those reporting current treatment and with a reactive test, 9% (n = 12) had high cholesterol levels. Thus, screening identified 7 individuals with undiagnosed high cholesterol, and 12 with treated but uncontrolled high cholesterol.

Changes in health self-efficacy following screening
We tested for changes in health attitudes from pre-to post-screening. Multivariable random regression analysis controlling for age, sex, race, education, and study site revealed small but statistically significant increases in self-rated abilities for health practices, and in perceived competence for health maintenance (Table 3). Significant increases also were observed in two of the three health locus of control subscales, namely internal control, and powerful others. There were no significant changes in participants' perceptions of their health being controlled by chance factors.

Discussion
To our knowledge, this is the first study to screen for a comprehensive set of health risks among people serious mental illness receiving outpatient treatment in multiple regions of the U.S., providing same-day results to participants along with health education and peer support. Ours is also the first to involve peer health and wellness specialists in the design and implementation of screening activities, and the first to demonstrate changes in health self-efficacy among participants from pre-to post-screening. Compared to prevalence reported for the non-institutionalized U.S. adult population, we found that higher proportions of participants screened positive for obesity, diabetes, hypertension, smoking and nicotine dependence, alcohol abuse, drug abuse, and coronary heart disease. Only in the case of hyperlipidemia was the proportion lower in the participant group than in the general population. Our results are also highly similar to those found in three prior screening studies with this population [48][49][50].
As hypothesized, positive changes occurred in self-perceived health efficacy following participation in health screening. At post-test, participants gave higher ratings to their degree of competence in performing health practices and to their ability to engage in health maintenance activities. In addition, they reported an enhanced degree of internal control over their own health. However, they also showed an increase in the degree to which they felt that "powerful others" exerted control over their health. The latter finding may be due to their recent interactions with our screening staff who they may have perceived as "powerful" health experts. It may also be due to feelings of powerlessness and low self-efficacy in general that have been found in prior studies of this population [81][82][83].
Given that participants spent a modest amount of time engaging in health fair activities, totaling 60-90 minutes on average, even small increases in positive health attitudes and self-rated competencies are noteworthy. The changes we observed suggest that health fair follow-up activities might be beneficial in enhancing any gains that participants experienced [43,84]. For example, a study of individuals attending the Indiana Black and Minority Health Fair [85] found that those who received health counseling sessions afterwards were more likely than controls to report improvement in general health status and healthy behaviors at 15-month follow-up.
Health screening also was successful at detecting serious medical problems of which participants were either unaware or for which they were receiving treatment but did not demonstrate values in the expected normal ranges. Screening identified 82 instances of undiagnosed diabetes, hypertension or hyperlipidemia, and 76 instances where these disorders were treated but uncontrolled. This noteworthy level of unmet needs suggest that many would benefit from enhanced coordination with medical providers to address health risks.
One advantage of this approach to screening is that participants with newly identified health risks and enhanced health self-efficacy can immediately be provided with information about the nature and treatment of each condition, along with peer support for decision-making about next steps. Peer health and wellness specialists assessed participants' readiness for change in order to empower and encourage people, no matter what their planned course of action or inaction. For those expressing a desire to act, peer staff emphasized simple lifestyle alterations that participants could make in the near future, using a handout that cross-walked different health risks with potential life changes. For example, participants who expressed readiness to address high cholesterol were encouraged to add fiber to their diets, eat frequent small portions instead of three big meals, reduce or eliminate sweets and alcohol, and record what they ate in a food diary. Those with high nicotine dependence learned about new medications and psychosocial interventions for smoking cessation. Those not ready to quit were informed about the health advantages of reducing their smoking, even without complete cessation. Given documented health disparities in this population, our finding of lower prevalence for one chronic condition among participants than in the general population is noteworthy. This was hyperlipidemia, where 7% of our participants screened positive for high serum total cholesterol compared to 14% in the NHANES [79]. This may be due to the absence of a nonfasting test that may have under-estimated lipid abnormalities. Another possibility is that hyperlipidemia is more easily managed in concert with serious mental illness than are other co-occurring medical conditions. Our participants are among the one-in-five Americans managing multiple chronic conditions [86], and research suggests that certain clusters of co-morbid illnesses are more appropriately treated (e.g., depression and hyperlipidemia) [87] than others (e.g., psychosis and arthritis) [88]. Whatever the reason, this finding underscores the potential for those in the mental health field to partner with peer health specialists and primary care providers in efforts to screen for and successfully control co-morbid medical risks in this vulnerable population.
A number of caveats apply to our study findings. First, study participants were not nationally representative of adults with serious mental illness since they were recruited from mental health programs in only four U.S. states. Second, without a control group, we are unable to attribute the pre-post changes in health attitudes we observed to health fair participation itself. Third, we were unable to conduct follow-up testing to confirm our screening results and, thus, we do not know the proportion of "false negative" or "false positive" results that our testing yielded. Fourth the self-report nature of the NHANES questions introduces potential biases for operationalizing prevalence and treatment statuses given that respondents may not recall or may incorrectly report certain medical conditions and services. Fifth, administering the posttest assessment immediately after the health fair did not allow participants time for changes in health behaviors that might have been reflected in the attitudinal measures. Finally, the mental health agencies that participated may be more "health-conscious" than typical community mental health programs and, thus, their clients may be more empowered to take care of their health and healthier as a result.
Anecdotal observations shared with us by agency staff indicated that, after the screenings, many participants brought their health passports to meetings with their mental health clinicians and primary care providers to review test results and discuss future courses of action. Both during and after the screenings, participants expressed positive reactions to their contacts with peer health and wellness specialists who staffed the events. For many, this was their first exposure to peers in this role and they inquired about how to receive similar training. In a few cases, the local agency peers we hired and trained as greeters and logistics coordinators went on to complete their state's peer specialist certification or other health-related training. Another project outcome was the creation of a health screening manual for people in community mental health programs called "Promoting Wellness for People in Mental Health Recovery" that is available for free download at http://www.cmhsrp.uic.edu/health/designing_health_ screening.asp.
In conclusion, our results suggest that collaborative health risk screening involving peers can help to answer epidemiologic questions, provide targeted health education, and empower participants to better manage their medical needs. Given the global use of community screening as a tool for addressing health disparities, the fields of psychiatry and psychology, medicine, nursing, social work, occupational therapy, and public health have much to offer these efforts. This might include helping to organize screening events with community partners; encouraging the involvement of medical students, residents and interns; staffing health fair stations; and supporting patients' participation in screening activities. Through this public health approach, the integration of health and mental health care can be both practiced and promoted to reduce the high levels of morbidity and mortality that impede mental health recovery.