Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Nov 26.
Published in final edited form as: Am J Orthopsychiatry. 2005 Jan;75(1):54–62. doi: 10.1037/0002-9432.75.1.54

Failure to Seek Health Care Among the Mentally Ill

Hyeouk C Hahm 1,2, Steven P Segal 3
PMCID: PMC7690558  NIHMSID: NIHMS1639900  PMID: 15709850

Abstract

This study describes failure to seek health care among 673 new adult clients seeking mental health services in the San Francisco Bay area. Overall, 49% (n = 328) reported a failure to seek health care they believed was needed in the past year. People with dual diagnosis, severe depression, chronic physical illness, fear of coercive treatment, private insurance, and no insurance were more likely to fail to seek health care. Greater use of private physicians decreased the odds of failure to seek health care. These findings highlight the need to target groups at risk for failure to seek health care and the need to design nonthreatening programs to improve health access for people with mental illness.

There is substantial evidence that people with mental illness have high mortality and morbidity rates (Black, Warrack, & Winokur, 1985a, 1985b; Felker, Yazel, & Short, 1996; Korany, 1979; Wright & Weber, 1987). These phenomena are associated with high suicide rates, accident rates, and incidence of drug problems (Babigian & Odoroff, 1969; Eastwood, Stansy, & Meiser, 1982). Also of concern are reported high rates of unrecognized medical disorders and consequent neglect of physical problems (Black et al., 1985a, 1985b; Druss, Bradford, Roseheck, Radford, & Krumholz, 2001; Hall, Gardner, Popkin, LeCann, & Slickeny, 1981; Korany, 1979; Segal & Kotler, 1991). For instance, Korany (1979), in an investigation of mortality among 2,090 psychiatric patients, found that 1 in 2 had undetected major medical illnesses. Over 80% of the patients had physical illnesses at the time of their death. Farmer (1987) also found that although half of 110 severely mentally ill patients in Seattle, Washington, had undiagnosed medical problems and over one third of them had known medical problems during psychiatric treatment, 88% of these patients reported that they did not have a primary care physician. Furthermore, 68% reported that their last physical examination occurred during their last admission to a psychiatric hospital.

Although the use of mental health services by people with mental illness has been well documented (National Advisory Mental Health Council, 1993; Rupp, 1991), we know little about their use of medical services (Jeste & Unutzer, 2001; Kushel, Vittinghoff, & Haas, 2001). Health care utilization patterns among the mentally ill appear to contribute to patients’ increased health and mortality risks (Liberman, & Coburn, 1986). Enhanced opportunities for receiving needed health care among seriously mentally ill patients may therefore prevent exacerbation of illness and reduce the need for expensive care in hospital emergency rooms, a primary source of care for physical, mental, alcohol, and drug problems (Ball & Havassy, 1984; Berren, Santiago, Zent, & Carbone, 1999). Prior research has shown that health care service utilization among adults is influenced by patient predisposition to seek help (associated with patients’ race/ethnicity, gender, and/or education), level of need (i.e., severity of symptoms and diagnosis), and factors enabling access (e.g., insurance status, cost of copayment, and income; Anderson, 1995; Anderson & Newman, 1979; Millman, 1993).

In the case of people with mental illness, fear of civil commitment may be a disenabling factor that affects their health care use. Many myths and misunderstandings lead to the stigmatization of the severely mentally ill, and mentally disabled people tend to be mistrustful and suspicious of health care providers because of the fear of being forcefully committed to a psychiatric hospital (Howard, Cornille, Lyons, Vessey, & Saunders, 1996; Weissman, 2001).

The theoretical foundation of this study is an adaptation of Anderson’s (1995; Anderson & Newman, 1979) multivariate behavioral model for understanding health care use. This study explores failure to seek health care among mentally ill individuals within six county health care systems in the San Francisco Bay area.

Method

Participants and Settings

All project research procedures were approved by the University of California, Berkeley, Institutional Review Board on Research Involving Human Subjects. Between 1996 and 2000, 673 new clients seeking mental health agency assistance within six county health care systems (i.e., Sacramento, Alameda, San Francisco, Contra Costa, Santa Cruz, and Solano) were interviewed in the greater San Francisco Bay area of Northern California. Two types of mental health agencies serving the mental health needs of their catchment areas were source sites (i.e., community mental health agencies; CMHAs) and self-help agencies (SHAs).

New entering clients (n = 787) were asked to participate in the study, and a total of 86% (N = 673) agreed to participate in a comprehensive interview and assessment of their health, mental health, and social and economic situation. Participant refusal rates did not differ by county health system or by source site (CMHA vs. SHA, respectively, 15% and 14%). There were no significant differences in gender, ethnicity, or housing status. The detailed characteristics of this sample have been described in a previous article (Segal, Hardiman, & Hodges, 2002).

Statistical Analysis

Overall sample characteristics are reported along with statistics on those who failed to seek health care and those who sought health care in the past year. Thirty-one health problems were compared between those who failed to seek health care and their counterparts. Relative risk ratios were computed for each health problem.

A three-stage logistic regression was used to estimate the influence of predisposing (i.e., age, gender, African American ethnicity, education), need (i.e., dual diagnosis, Center for Epidemiological Studies–Depression scale [CES-D] score, Brief Psychiatric Rating Scale [BPRS] score, schizophrenia, schizoaffective disorder, major depressive disorder, bipolar disorder, other disorders, number of chronic illnesses, housing status), enabling (i.e., types of insurance, monthly income, fear of psychiatric hospitalization), and context factors (i.e., county health care system, mental health agency type, i.e., SHA vs. CMHA, extent of private office health care utilization in the county). Because we have a large number of African American participants and there is a concern about inadequate care and limited health care utilization by African Americans (Department of Health and Human Services, 2001; Hueston & Hubbard, 2000), we compared African Americans with all other participants. Because a substantial number of participants did not answer or did not know about their insurance status, we created a dummy variable for those with missing information on insurance status to increase statistical power. In the first stage of the logistic regression, we entered all the factors except for the context factors. The context factors (five of six county systems all contrasted with Sacramento; SHA vs. CMHA) were entered in the second stage. Finally, in the third stage, we considered interactions among predisposing, need, enabling, and context factors.

Measurements and Coding

Predisposing factors included age, gender (female = 1, male = 0), race (African American = 1, others = 0), and education (reported in years).

Need factors were as follows: (a) dual diagnosis (yes = 1, no = 0), defined as those who had lifetime substance abuse or dependence and lifetime psychiatric diagnoses given Diagnostic Interview Schedule (DIS; Robins, Helzer, Cottler, & Goldring, 1988) criteria, and (b) CES-D score (range 20–80, with higher scores indicating greater depression; Radloff, 1977). In addition, (c) the BPRS (Overall & Gorham, 1962) was used as a measure of psychological disability (Rhoades & Overall, 1988; Segal & Kotler, 1993). The BPRS involves interviewer ratings of symptom severity on 24 items scaled from 1 to 7 (range 7–168). Interrater reliability in this study was .90, and internal consistencies varied between .79 and .86 (Segal, Hardiman, & Hodges, 2002). (d) Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 1994) Axis I diagnoses were included (schizophrenia and schizoaffective disorder, major depressive disorder, bipolar disorder, and other disorders, e.g., obsessive–compulsive disorder or anxiety disorder). (e) Number of chronic illnesses was measured as the sum of positive responses to health problems (see list in Table 1). (f) Housing status was assessed as stable (reports of living in a house or an apartment), marginal (reports of living in a house or an apartment but having a history of homelessness for the last 5 years; or reports of living in a hotel, transitional housing, or board and care home), and literally homeless (reports of living in the street, a car, or in a shelter).

Table 1.

Reported Health Problems and Relative Risk of Failure to Seek Health Care

Failed to seek health care in past year (n = 328) Sought health care in past year (n = 338)
Relative risk ratio (failed vs. received)
Health problem n % n % χ2 P
Tuberculosis 12 3.7 4 1.2 4.34 .04 3.1
Hepatitis 37 11.3 14 4.1 12.00 .001 2.8
Bowel/urination control problems last 6 months 43 13.2 21 6.2 9.11 .003 2.1
Skin problem 123 37.8 60 17.8 12.70 .000 2.1
Cancer 8 2.4 4 1.2 1.48 .22 2.0
Hernia or rupture 18 5.5 9 2.7 3.41 .06 2.0
Heart or chest pain 92 28.0 48 14.2 19.20 .000 2.0
Diabetes 25 7.7 14 4.1 3.65 .05 1.9
Bronchitis 63 19.3 44 13.0 4.73 .03 1.9
Trouble with teeth or gums 184 56.3 103 30.5 44.50 .000 1.9
Joint/muscle stiff/ache 161 49.1 89 26.3 36.80 .000 1.9
Fainting or loss of consciousness 44 13.5 25 7.4 6.49 .011 1.8
Breathing trouble 136 41.6 77 22.8 26.70 .000 1.8
Asthma 85 26.0 50 14.8 12.70 .000 1.8
Tiredness 177 54.1 103 30.5 37.70 .000 1.8
Stomach or duodenal ulcer 46 14.2 28 8.3 5.55 .02 1.7
High blood pressure 72 22.1 43 12.7 9.95 .002 1.7
Back pain 185 56.4 110 32.5 38.40 .000 1.7
Coughing or frequent chest colds 80 24.4 49 14.5 10.40 .001 1.7
Active AIDS 7 1.0 2 0.6 2.97 .08 1.7
Swollen ankles 58 17.7 36 10.7 6.79 .009 1.7
Paralysis 17 5.2 11 3.3 1.53 .21 1.6
Arthritis or rheumatism 121 37.0 76 22.5 16.60 .000 1.6
Leg cramps 112 34.1 71 21.0 14.40 .000 1.6
Other 86 26.5 57 17.0 6.40 .01 1.6
Epilepsy, fits, or seizures 26 7.9 18 5.3 1.83 .17 1.5
Anemia 38 11.9 27 8.0 2.44 .12 1.5
Stomach pain 86 26.2 62 18.3 5.97 .02 1.4
HIV infection 7 2.1 6 1.8 0.11 .74 1.2
Stroke 2 0.6 4 1.2 0.61 .43 0.5
Parkinson’s disease 1 0.3 2 0.6 0.33 .56 0.5
Open in a new tab

Note. Health problems are ordered by relative risk ratio.

Enabling factors were as follows: (a) Insurance status was assessed as government insurance (individuals in receipt of Medicaid, Medicare, or Veteran’s Administration insurance); criminal justice health care (health care received in prison or jail system); no insurance (reports of self- and/or family pay) and/or county paid (i.e., county ends up paying for the medical bill for those who do not have insurance at the time of seeking care); and private insurance (those who reported having private insurance). (b) Income in the past month was measured as the sum of all income sources, including supplemental security income, Social Security Disability Income, Government Assistance, Temporary Assistance for Needy Families, food stamps, any retirement or pension, unemployment, family or friend contributions, work, Veterans Affairs, or any other sources of income (including illegal activities). (c) Fear of psychiatric hospitalization was assessed with the participant’s graded response to the question, “Before you came to the agency, how worried were you that the counselors/staff would have you hospitalized for psychiatric reasons?” (not at all worried = 1, extremely worried = 5).

Context factors were as follows: (a) county health system (Sacramento, Santa Cruz, San Francisco, Alameda, Contra Costa, and Solano), (b) mental health agency type (CMHAs = 1, SHAs = 0), and (c) extent of private office care, measured as the proportion of people in each county health system who received their health care in a private office.

Measure of the Dependent Variables

Failure to seek health care was measured by response to the question, “In the past year, have you had times when you felt you should have gone to a doctor or other health provider but did not?” Those who answered “yes” were considered as having failed to seek care (coded as 1), and those who answered “no” were considered as having sought care (coded as 0).

Results

Total Participants

The mean age of the participants was 39.2 years (SD = 9.8). As shown in Table 2, participants were predominantly White (White, 54%; African American, 29%; Latino, 9.8%; Asian 2.5%; Native Americans, 2.2%; other ethnicities, 2.5%). The mean years of education was 12; over one third of the participants had not graduated from high school, and only 10% had a bachelor’s degree or more. The majority of participants did not have stable housing, and their average monthly income was $512. One fourth reported not having health insurance. Among the total sample of 673, 49.2% (n = 328) reported a failure to seek care in the past year. More than half of African Americans, Native Americans, and those of other ethnicities reported a failure to seek health care. In general, the more severe the mental illness and/or physical illness a participant had, the more likely he or she was to report having failed to seek health care (see Table 2). Finally, the rates of failure to seek health care among the participants in the Sacramento County health care system were substantially lower (19.7%), whereas about half or more of the participants reported failure to seek health care in the rest of the county health systems.

Table 2.

Characteristics of Participants Reporting Seeking or Failing to Seek Health Care in the Past Year

Total samplt (n = 673) Those who failed to seek health care (n = 328) Those who sought health care (n = 345)
Sample characteristics n % M SD n % M SD n % M SD
Gender
 Male 361 54.0 176 48.8 185 51.3
 Female 312 46.0 152 48.7 160 51.3
Ethnicity
 African American 183 29.0 103 56.3 80 43.7
 White 339 53.5 146 43.7 193 56.3
 Latino 62 9.8 34 48.8 28 45.2
 Native American 14 2.2 8 57.1 6 42.9
 Asian 16 2.5 4 25.0 12 75.0
 Other, declined to answer 16 2.5 13 81.3 3 18.8
Age 39.2 9.8 39.8 9.6 38.7 9.9
Education 12.3 2.6 12.1 2.8 12.5 2.5
CES-D 48.3 13.8 51.9 12.7 44.7 14.0
BPRS 37.8 11.55 39.8 12.5 35.7 10.2
Chronic illness 2.5 2.5 3.3 2.51 1.8 2.4
Income 512.6 395.3 504.1 420.7 517.8 364.5
Degree of worry provider would hospitalize care seeker for psychiatric reasons 1.73 1.16 2.0 1.3 1.5 1.0
DSM-IV diagnosis Major depressive disorder 421 63.6 222 52.7 199 47.3
 Schizophrenia or schizoaffective disorder 106 16.0 50 47.2 56 52.8
 Bipolar disorder 29 4.4 13 44.8 16 55.2
 Other disorder 106 16.0 42 39.6 64 60.3
Type of insurance Government insurance 320 47.5 157 47.9 163 48.2
 Private insurance 38 5.6 22 57.0 16 42.0
 Criminal justice health care 13 1.9 5 38.0 8 61.0
 No insurance at the time of treatment 173 25.7 85 52.0 77 42.0
 I don’t know/missing 150 22.3 67 46 78 54.0
Study participants’ county health system
 Alameda 267 39.7 142 53.3 124 46.6
 Contra Costa 22 3.3 10 47.6 11 52.4
 Sacramento 134 19.9 24 19.7 11 82.1
 San Francisco 54 8.0 32 59.3 22 40.7
 Santa Cruz 71 10.5 41 58.6 29 41.4
 Solano 125 18.6 79 65.3 42 34.7
Degree of private office care in study participant’s health system
 Sacramento 44 54.3 3 6.8 41 93.2
 Solano 50 47.2 30 62.5 18 37.5
 Santa Cruz 20 34.5 12 60.0 8 40.0
 Contra Costa 6 33.3 4 80.0 1 20.0
 San Francisco 9 21.4 5 55.6 4 44.4
 Alameda 36 16.6 17 48.6 18 51.4
Study participant’s mental health service agency
 SHA 226 33.6 109 48.2 117 51.8
 CMHA 447 66.4 219 49.8 221 50.2
Open in a new tab

Note. CES-D = Center for Epidemiological Studies–Depression scale; BPRS = Brief Psychiatric Rating Scale; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders (4th ed.); SHA = self-help agency; CMHA = community mental health agency.

Health Problems of Participants

The health problems reported in Table 1 are listed in descending order on the basis of the size of relative risk ratios comparing those who failed to seek care with those who sought health care. Overall, those who failed to seek health care were significantly more likely to suffer from nearly all the health problems listed. Respondents reported a very high prevalence of serious infectious diseases, such as tuberculosis and hepatitis. Participants who failed to seek health care were 3 times more likely to have tuberculosis and 1.7 times more likely to report AIDS than those who sought care (no prevalence differences by race or gender were observed). Those who reported failure to seek health care were twice as likely to suffer from life-threatening diseases, such as cancer, hernia rupture, and diabetes, and almost twice as likely to suffer the most commonly reported health problems: trouble with teeth or gums (56.3%, n = 184), back pain (54.1%, n = 185), and tiredness (54.1%, n = 177).

Association With Failed Health Care

A three-stage logistic regression was used to model the associations between failed health care and predisposing, need, enabling, and context factors (see Table 3). In the first stage of the model, after adjustment for all other variables, African Americans were 1.7 times more likely to report a failure to seek health care than non-African Americans; in addition, those with dual diagnosis, higher CES-D scores, and a higher number of chronic illnesses were more likely to report a failure to seek health care. These findings suggest that serious mental and physical problems were independently associated with failures to seek health care. Respondents who had a stable home were less likely to report a failure to seek care. Compared with those who received health care in the criminal justice system, those who had private insurance were 3 times more likely, and those who did not have insurance were about 2.6 times more likely, to report a failure to seek health care. In addition, the greater a participant’s fear of being hospitalized for psychiatric reasons was, the more likely he or she was to report a failure to seek health care. For each one-unit increase in this scale, there was a 30% increase in the odds of failure to seek health care.

Table 3.

Adjusted Odds Ratios for Report of Failure to Seek Health Care Within the Past Year by Predisposing, Need, Enabling, and Context Factors (N = 570)

Single stage model Two-stage model Two-stage revised model
Predictor variable OR CI P OR CI P OR CI P
Predisposing factors
 Age (years) 1.19 0.96–1.46 1.18 0.96–1.47 1.18 0.96–1.47
 Female (vs. male) 1.01 0.68–1.50 1.08 0.72–1.62 1.01 0.72–1.63
 African American (vs. others) 1.73 1.13–2.64 .012* 1.45 0.90–2.33 1.45 0.91–2.33
 Education 0.98 0.91–1.06 0.97 0.89–1.05 0.97 0.89–1.05
Need factors
 Dual diagnosis (vs nondual diagnosis) 2.01 1.34–3.02 .000*** 2.09 1.38–3.17 .001** 2.09 1.34–3.17 .001**
 CES-D 1.04 1.02–1.05 .000*** 1.04 1.02–1.06 .000*** 1.04 1.02–1.06 .000***
 BPRS 1.00 0.99–1.02 1.00 0.98–1.02 1.00 0.97–1.02
 DSM-IV diagnoses (vs. other diagnosis)
 Major depression 1.12 0.68–2.09 1.37 0.77–2.45 1.37 0.77–2.45
 Schizophrenia and schizoaffective disorder 0.94 0.47–1.85 1.38 0.67–2.84 1.38 0.67–2.84
 Bipolar disorder 1.42 0.51–4.00 1.63 0.58–4.74 1.64 0.58–4.63
 No. chronic illnesses 1.18 1.10–1.30 .000*** 1.12 1.03–1.23 .014* 1.12 1.03–1.23 .014*
 Stable home (vs. marginally homeless or homeless) 0.53 0.35–0.82 .004** 0.67 0.41–1.07 0.67 0.41–1.06
Enabling factors Insurance status (vs. criminal justice health care)
 Government insurance 2.17 0.91–4.91 2.02 0.84–4.81 2.01 0.84–4.82
 Private insurance 3.42 1.15–10.20 .027* 3.41 1.10–10.60 .033* 3.42 1.10–10.60 .033*
 No insurance 2.60 1.16–5.87 .021* 2.52 1.09–5.83 .031* 2.51 1.07–5.83 .031*
 Insurance missing 2.02 0.78–5.15 2.14 0.83–5.60 2.14 0.82–5.60
 Income 0.98 0.94–1.03 0.98 0.92–1.02 0.97 0.92–1.02
Worried staff would hospitalize him or her for psychiatric reasons 1.30 1.09–1.55 .003** 1.23 1.02–1.47 .03* 1.23 1.02–1.47 .03*
Context factors County health system (vs. Sacramento)
 Solano 6.23 3.04–12.70 .000*** 3.81 2.07–7.01 .000***
 Alameda 4.17 1.95–8.91 .000*** 0.31 0.08–1.11
 San Francisco 3.97 1.58–10.40 .004** 0.41 0.12–1.39
 Santa Cruz 3.94 1.70–9.20 .001** 0.91 0.29–2.87
 Contra Costa 4.88 1.47–16.10 .009** 1.14 0.33–3.88
County mental health agencies (vs. self-help agencies) 1.00 0.64–1.57 1.00 0.64–1.56
Degree of private office care 0.93 0.89–0.97 .001**
Open in a new tab

Note. Predisposing, need, and enabling factors were entered in the first stage of the two-stage models, and context factors were entered in the second stage. OR = odds ratios; CI = confidence interval; CES-D = Center for Epidemiological Studies–Depression scale; BPRS = Brief Psychiatric Rating Scale; DSM–IV = Diagnostic and Statistical Manual of Mental Disorders (4th ed.).

*

p < .05.

**

p < .01.

***

p < .001.

In the second stage of the model, after we controlled for health system and agency type, being African American and having stable housing were no longer significant factors. Compared with Sacramento County’s health system, all other county health systems showed significantly increased odds of failure to seek health care.

Exploring the latter finding in greater detail, we found that participants in the Sacramento County health care system were less likely to fail to seek health care. To interpret this phenomenon, we began by identifying variables that could be expected to account for this result. We then found that participants in the Sacramento County health system were significantly more likely than participants in other systems to use health care in private offices (54.3% vs. 27.4%), χ2(1, N = 522) = 22.8, p < .0001. They were also significantly less likely to obtain health care in emergency rooms, vans, and outpatient clinics. A greater tendency to visit a private office for health care was significantly associated with a decrease in the odds of failure to seek health care, even after all other variables were controlled. We also observed that the health systems differed quite dramatically in the percentage of individuals receiving health care in the private offices of physicians. To determine whether the health system context effect was accounted for by the latter finding, we reran the first two stages of the logistic model, including in the context stage a variable that reflected the proportion of cases receiving health care in private offices in each health system as well as the other context factors. In this revised model (see Table 3), after we controlled for the proportion of individuals receiving their health care in a private office, only Solano County’s health system remained significant. In addition, we found that after we took other factors into account, the greater the proportion of private office use for receipt of health care in a system was, the less likely a given client was to report a failure to seek health care. None of the interactions were statistically significant; therefore, the third stage is not reported in Table 3.

Discussion

Results indicate that every year, one in two study participants think that they should have gone to health care providers but did not. An important question in this context is whether the failed visits were indeed necessary in relation to improving the individuals’ health status. Our results show that those who failed to seek health care were significantly more likely to have severe mental illnesses and more chronic illnesses, including a high prevalence of infectious diseases (tuberculosis and hepatitis). Tuberculosis infection among the U.S. population in 2000 was about 5.8 cases per 100,000 (16,377 actual cases, 0.0058%; Centers for Disease Control & Prevention, 2002). If we accept the number of tuberculosis self-reports at face value, the rate among those who reported a failure to seek health care is 3,659 cases per 100,000 (630 times higher than the national rate as reported by the Centers for Disease Control & Prevention). Furthermore, those who reported a failure to seek health care were 11 times more likely to have hepatitis than the general U.S. population (Centers for Disease Control & Prevention, 2003). Our study indicated that, compared with those who received health care in the criminal justice system, those who did not have insurance and who had private insurance were about three times more likely to fail to seek health care. This result is consistent with McAlpine and Mechanic’s (2000) findings indicating that severely mentally ill patients under criminal justice health care were 3.8 times more likely to get specialty mental health care than those who had no insurance. We also speculate that the reason for failure to seek health care among those with private insurance has to do with the financial burden of copayments and undercoverage by health maintenance organizations. Results from a RAND study (Brook et al., 1983) demonstrated that receiving free health care significantly reduced elevated risks of hypertension, vision problems, and mortality among poor people.

Because, in general, people who get health care in private offices tend to have higher incomes and better insurance, we further analyzed whether there were differences in income and insurance status between participants in Sacramento County and their counterparts—that is, given the positive results associated with private office visits and their frequency in Sacramento. We found none. We also found that only 17.5% of participants in Sacramento who had government insurance reported failure to seek care, whereas 56% of participants who had government insurance in the other counties reported failure to seek health care, χ2(1, N = 328) = 38.6, p < .000. This result led us to investigate whether physicians in Sacramento accepted more government insurance compared with physicians in other counties. Because our data do not permit this analysis, we used data from the California Department of Health describing all the physicians who accept government insurance in every county in California. We found no substantial difference between Sacramento and other counties related to the acceptance of government insurance. We also found that when we take account of the health system in which the clients are served in our logistic model, traditional factors thought to be associated with failure to seek health care (African American ethnicity and lack of stable housing; Gelberg & Linn, 1988) disappear, leaving only mental disorder severity indicators and system of care to account for the observed differences.

This is a sample with major psychiatric problems. That such problems should predominate in the issue of failure to seek health care harkens back to previous observations in the mental health literature describing the poor health of the seriously mentally ill. Such findings suggest that mentally ill individuals may be less able to make responsible decisions about their own health status and may be in need of more extensive assistance to avoid unnecessary and costly decisions regarding their health status. Alternatively, it might be hypothesized that these individuals have a good sense of the severity of their mental health problems and know that approaching a clinic, with its more formal administrative procedures, will place them at greater risk of losing their freedom via civil commitment. They therefore avoid seeking needed health care except in the less threatening private physician’s office. Although a more certain conclusion to these alternative explanations must await future research, and both explanations may have validity for different patients, the need to explore ways to improve health care access for the severely mentally ill population, perhaps in less threatening settings, is paramount. If not for humanitarian reasons then at least because of the high prevalence of infectious diseases among the group, we must provide such access to protect the public’s health.

Our study has certain limitations. First, the findings from this research project are based solely on self-reports. Underreporting and overreporting of health conditions are not uncommon in self-reported data (Lindan et al., 1994). To overcome the shortcoming of underreports of sensitive health conditions, future studies should consider using standard clinical tests. A second limitation has to do with the directionality in the multivariate analyses. We measured the dependent variable, failure to seek health care, by asking about participants’ experience in the past year. However, some of the predictors, such as the BPRS, CES-D, and dual diagnosis, were measured at the end of the year. In other words, mental health symptoms were measured after the failure to seek medical care. This applies to the 6-month retrospective assessment of physical health as well. Because participants have already failed to seek health care, their mental or physical health status might have been worse at the time of the interview. Finally, a longitudinal study is needed to see what factors are related to failure to seek health care over time. Studies using nationally representative samples that contain long-term measures would generate more accurate estimates of failure to seek health care, though they would unfortunately lack motivational measures.

Despite these limitations, these findings have a number of important implications. From the viewpoint of national health policy, collective efforts should be made to address the urgent need for people with mental illnesses to gain access to the health care system. More outreach services are needed to identify and to educate poor mentally ill people with poor health who are eligible for government health insurance programs. As well, providing comprehensive and integrated medical and mental health services would be prudent. Adequate access to the health care system promotes prevention, screening, and adherence to treatment. In particular, for infectious disease, especially tuberculosis, adherence to treatment is essential to curing the disease and to prevent infection and disease in others (Galavotti, Salzman, Sauter, & Sumartojo, 1997). Our study observed that severely mentally ill individuals avoided heath care often for fear of civil commitment and coercive treatment. In fact, the only system of care associated with a reduction in failure to seek health care was care that was delivered in a private office (not a clinic, which holds greater threats of coercive care). The private office is less likely to have bureaucratic procedures and more likely to have an informal relationship between doctors and patients. Perhaps increasing access to health care via practitioners in private offices is a way not only to improve patients’ health status but to avoid the need to use more coercive interventions, promoting patient engagement rather than flight.

In conclusion, this study presents new insights into issues of health care access among the mentally ill. Targeting groups at risk for failure to seek health care and providing the services discussed above among people who are mentally ill may decrease emergency room visits and inpatient hospitalization, and, furthermore, it may eventually improve the health care and health status of this disabled population.

Acknowledgments

This study was supported by National Institute of Mental Health (NIMH) Grant RO1 MH-37310 and NIMH Training Grant MH-18828. We thank the fellows at the Center for Self-Help Research and Dr. Maureen Lahiff for reading and commenting on the manuscript.

Contributor Information

Hyeouk C. Hahm, University of California, Berkeley,; Center for Self-Help Research

Steven P. Segal, University of California, Berkeley.

References

  1. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed). Washington, DC: Author. [Google Scholar]
  2. Anderson RM (1995). Revisiting the behavioral model and access to medical care: Does it matter? Journal of Health and Social Behavior, 36, 1–10. [PubMed] [Google Scholar]
  3. Anderson RM, & Newman JF (1979). Societal and individual determinants of medical care utilization in the United States. Milbank Quarterly, 51, 95–124. [PubMed] [Google Scholar]
  4. Babigian HM, & Odoroff CL (1969). The mortality experience of a population with psychiatric illness. American Journal of Psychiatry, 126, 470–480. [DOI] [PubMed] [Google Scholar]
  5. Ball FJ, & Havassy BE (1984). A survey of the problems and needs of homeless consumers of acute psychiatric services. Hospital and Community Psychiatry, 35, 917–921. [DOI] [PubMed] [Google Scholar]
  6. Berren MR, Santiago JM, Zent MR, & Carbone CP (1999). Health care utilization by persons with severe and persistent mental illness. Psychiatric Services, 50, 559–561. [DOI] [PubMed] [Google Scholar]
  7. Black DW, Warrack G, & Winokur G (1985a). The Iowa record-linkage study: II. Excess mortality among patients with organic mental disorders. Archives of General Psychiatry, 42, 76–81. [DOI] [PubMed] [Google Scholar]
  8. Black DW, Warrack G, & Winokur G (1985b). The Iowa record-linkage study: III. Excess mortality among patients with “functional” disorders. Archives of General Psychiatry, 42, 82–88. [DOI] [PubMed] [Google Scholar]
  9. Brook R, Ware JE, Rogers WH, Keeler EB, Davies AR, & Donald CA (1983). Does free care improve adults’ health? New England Journal of Medicine, 309, 1426–1434. [DOI] [PubMed] [Google Scholar]
  10. Centers for Disease Control and Prevention. (2002). Tuberculosis morbidity among the U.S. born and foreign born population-United States. Morbidity and Mortality Weekly Report, 51(5), 101–104. [PubMed] [Google Scholar]
  11. Centers for Disease Control and Prevention. (2003). Viral hepatitis C. Retrieved January 13, 2003, from https://www.cdc.gov/ncidod/diseases/hepatitis/fact.htm
  12. Department of Health and Human Services. (2001). A report of the Surgeon General Rockville, MD: U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Mental Health Services, National Institute of Health, National Institute of Mental Health. [Google Scholar]
  13. Druss BG, Bradford D, Roseheck R, Radford MJ, & Krumholz HM (2001). Quality of medical care and excess mortality in older patients with mental disorders. Archives of General Psychiatry, 58, 565–572. [DOI] [PubMed] [Google Scholar]
  14. Eastwood MR, Stansy W, & Meiser HMR (1982). Mental illness and mortality. Comprehensive Psychiatry, 23, 377–385. [DOI] [PubMed] [Google Scholar]
  15. Farmer S (1987). Medical problems of chronic patients in a community support program. Psychiatric Services, 38, 745–749. [DOI] [PubMed] [Google Scholar]
  16. Felker B, Yazel JJ, & Short D (1996). Mortality and medical comorbidity among psychiatric patients: A review. Psychiatric Services, 47, 1356–1363. [DOI] [PubMed] [Google Scholar]
  17. Galavotti C, Salzman LE, Sauter SL, & Sumartojo E (1997). Behavioral science activities at the Centers for Disease Control and Prevention: A selected overview of exemplary programs. American Psychologist, 52, 154–166. [DOI] [PubMed] [Google Scholar]
  18. Gelberg L, & Linn LS (1988). Social and physical health of homeless adults previously treated for mental health problems. Hospital and Community Psychiatry, 39, 510–516. [DOI] [PubMed] [Google Scholar]
  19. Hall R, Gardner ER, Popkin MK, LeCann AF, & Slickeny SK (1981). Unrecognized physical illness promoting psychiatric admission: A prospective study. American Journal of Psychiatry, 138, 629–635. [DOI] [PubMed] [Google Scholar]
  20. Howard K, Cornille JS, Lyons JT, Vessey RJ, & Saunders SM (1996). Patterns of mental health service utilization. Archives of General Psychiatry, 53, 696–703. [DOI] [PubMed] [Google Scholar]
  21. Hueston W, & Hubbard ET (2000). Preventive services for rural and urban African American adults. Archives of Family Medicine, 9, 263–266. [DOI] [PubMed] [Google Scholar]
  22. Jeste D, & Unutzer J (2001). Improving the delivery of care to the seriously mentally ill. Medical Care, 39, 907–909. [DOI] [PubMed] [Google Scholar]
  23. Korany E (1979). Morbidity and rate of undiagnosed physical illness in a psychiatric population. Archives of General Psychiatry, 36, 414–419. [DOI] [PubMed] [Google Scholar]
  24. Kushel M, Vittinghoff E, & Haas JS (2001). Factors associated with the health care utilization of homeless persons. Journal of the American Medical Association, 285, 200–206. [DOI] [PubMed] [Google Scholar]
  25. Liberman A, & Coburn AF (1986). The health of the chronically mentally ill: A review of the literature. Community Mental Health Journal, 22, 104–116. [DOI] [PubMed] [Google Scholar]
  26. Lindan C, Avins AL, Woods WJ, Judes ES, Clark W, & Hully SB (1994). Levels of HIV testing and low validity of self-reported test results among alcoholics and drug users. AIDS, 8, 1149–1155. [DOI] [PubMed] [Google Scholar]
  27. McAlpine D, & Mechanic D (2000). Utilization of specialty mental health care among persons with severe mental illness: The roles of demographics, need, insurance, and risk. Health Services Research, 35, 1277–1293. [PMC free article] [PubMed] [Google Scholar]
  28. Millman M (1993). Access to health care in America. Washington, DC: National Academies Press. [PubMed] [Google Scholar]
  29. National Advisory Mental Health Council. (1993). Health care reform for Americans with severe mental illnesses. American Journal of Psychiatry, 150, 1447–1465. [DOI] [PubMed] [Google Scholar]
  30. Overall J, & Gorham D (1962). The Brief Psychiatric Rating Scale. Psychological Reports, 10, 799–812. [Google Scholar]
  31. Radloff L (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. [Google Scholar]
  32. Rhoades HM, & Overall JE (1988). The semi-structured Brief Psychiatric Rating Scale interview and rating guide. Psychopharmacology Bulletin, 24, 101–104. [PubMed] [Google Scholar]
  33. Robins L, Helzer J, Cottler L, & Goldring E (1988). NIHM Diagnostic Interview Schedule, Version III, revised St. Louis, MO: Washington University. [Google Scholar]
  34. Rupp A (1991). Underinsurance for severe mental illness. Schizophrenia Bulletin, 17, 401–405. [DOI] [PubMed] [Google Scholar]
  35. Segal SP, Hardiman ER, & Hodges JQ (2002). Characteristics of new clients at self-help and community mental health agencies and geographic proximity. Psychiatric Services, 53, 1145–1152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Segal SP, & Kotler PL (1991). A ten-year perspective of mortality risk among mentally ill patients in sheltered care. Hospital and Community Psychiatry, 42, 708–713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Segal SP, & Kotler PL (1993). Personal outcomes and sheltered care residence: Ten years later. American Journal of Orthopsychiatry, 63, 80–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Weissman M (2001). Stigma. Journal of the American Medical Association, 285, 261–262.11176821 [Google Scholar]
  39. Wright J, & Weber E (1987). Homelessness and health. New York: McGraw-Hill. [Google Scholar]

RESOURCES