Dr. Ng, Ms. Jin, and Dr. Ho are affiliated with the Department of Psychological Medicine, National University of Singapore, 5 Lower Kent Ridge Rd., Singapore 119074 (e-mail: firstname.lastname@example.org). Dr. Chua is with the Department of General Psychiatry, Institute of Mental Health, Singapore. Dr. Fones is with the Department of Psychiatry, Gleneagles Hospital, Singapore. Dr. Lim is with the Department of Behavioural Medicine, Singapore General Hospital, Singapore.
The high prevalence of psychiatric disorders worldwide indicates enormous needs for mental health services. Patients with mental disorders often do not seek professional help. It has been estimated that between one-third and two-thirds of serious cases of psychiatric disorders worldwide received no treatment (1).
Numerous studies have established that objective measures of need, such as the presence of a diagnosable psychiatric disorder and interference with life functioning, clearly determine professional help seeking. However, objective measures of need do not fully determine actual health behavior. The impetus for and continuity of help seeking often lie with self-assessment of mental health and perceived need for help. Health beliefs are attitudes, values, and knowledge that people have about their health and health services that influence their subsequent perceptions of need and use of health services (2). Health beliefs are widely cited explanations of why some people with mental health problems seek help and others do not. These reasons include self-perceived presence of a mental illness, self-rating of mental health, stigma and embarrassment, difficulty in discussing mental health problems, lack of trust in mental health professionals, and overestimation of one's coping abilities (3,4). Along with demographic and social background, these reasons constitute a set of factors that predispose, enable, or impede the use of services that is not well studied and is little understood.
In this report, using data from the National Mental Health Survey of Adults in Singapore, we assessed the prevalence of psychiatric disorders and whether people sought help from health care professionals. We investigated whether—apart from need factors (such as the presence of a psychiatric diagnosis and functional disability) and sociodemographic factors—the likelihood of help seeking was independently associated with health beliefs and family and social support.
The National Mental Health Survey of Adults in Singapore was conducted from February 15, 2003, to March 30, 2004, with a nationally representative random sample of Singaporean citizens and permanent resident adults of ages 20—59 years. Details of the sampling methodology and questionnaire interviews are published elsewhere (5). A total of 2,847 respondents were interviewed from a total of 3,875 eligible persons (73% response rate).
Respondents with recent psychiatric disorders were identified by the respondent's self-report of having received treatment for mental problems by a physician in the past year. In addition, an active two-stage case ascertainment used the 12-item General Health Questionnaire (6,7), followed by the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) (8), to derive DSM-IV diagnoses of depressive and anxiety disorders. Psychotic disorders, substance abuse, and other disorders were not considered in the survey because they were relatively uncommon.
Additional dimensions of objective needs assessment were elicited by questions about physical and social role functioning: "Have you had any of the following problems with your work or other daily activities as a result of any emotional problems (such as feeling depressed or anxious): accomplished less than you would like? cut down on the amount of time you spent on work or other activities?" (9). The number of medical problems (nonpsychiatric) co-occurring with psychiatric problems was quantified from self-reports from a list of common chronic medical illnesses.
Respondents were asked, "In the past 12 months, have you sought professional help for problems with mental health or emotion from a general practitioner, psychiatrist, psychologist, psychiatric nurse, social worker and mental health counselor, or others? (please specify)."
Self-rating of mental health status was assessed on a 5-point Likert scale (excellent, very good, good, fair, and poor). Respondents also were asked in a separate section of the questionnaire (and acknowledged) whether they had mental health problems. Along with the presence of a diagnosable mental disorder, a negative response to this question would suggest an attitude reflecting psychological states of denial or lack of insight.
The respondents' health beliefs, including perceptions or views about mental illness, professional care and medications, and one's own problem-solving skills, were assessed by statements about the curability of mental illness, embarrassment and stigma, ease of discussing mental problems, effectiveness and safety of treatment, and trust in professionals (Table 1).
Other variables included the extent of family and social support ("family or someone able to provide help when you needed it"), age, gender, education level, marital status, living arrangement, income, and employment status.
We performed statistical analysis for 2,801 respondents after excluding 46 respondents with missing data. Categorical data analyses included chi square tests of significance and logistic regression to evaluate the association of health belief variables with help seeking, with the presence of other need and predisposing and enabling variables accounted for. Because the sampling design involved disproportionate stratification by ethnic group, weighted analyses were performed with SPSS version 12 to derive estimates of proportions and odds ratios.
The mean±SD age of respondents was 41±9.6 years; 1,758 (63%) respondents were female; 101 (4%) had no education; 47 (2%) lived alone; 132 (5%) were unemployed; 355 (13%) had monthly income less than $1,000 ($635 U.S.); 13 (5%) lived in smaller-sized low-end public housing; and 193 (7%) were divorced, separated, or widowed.
The overall population-weighted prevalence estimate of depressive and anxiety disorders (N=181) in the past year was 10%. The most frequent recent disorders were major depressive disorder (6%) and generalized anxiety disorder (4%), and dysthymia (1%) and panic disorder without agoraphobia (1%), which included 2% who had co-occurring depressive and anxiety disorders. However, only 1,450 (5%) of all respondents and 61 (34%) respondents with depressive or anxiety disorders or both rated their mental health status as fair or poor. Only 71 respondents overall (3%) and 28 respondents (15%) with depressive or anxiety disorders stated that they had mental problems.
An estimated 6% of the population (N=157) sought help from professionals for mental and emotional problems. Among 181 respondents with depressive and anxiety disorders diagnosed with SCAN in the survey, an estimated 57 respondents (32%) had sought help.
We modeled the probabilities of individuals to seek help for mental and emotional problems in logistic regressions that included need, predisposing variables, and enabling variables in a full model that included all variables (Table 1). As expected, the presence of SCAN-indicated depressive and anxiety disorders and physical and social role limitations resulting from mental problems were significantly associated with the likelihood of help seeking, as was medical comorbidity. In addition, socioeconomic status was associated with help seeking. The relationship was nonlinear, however, with respondents living in smaller low-end housing or in larger high-end housing being significantly more likely to seek help.
In the presence of objective need and other variables, the respondent's acknowledgment of a mental health problem (odds ratio [OR]=5.22) and fair to poor self-rated mental health (OR=2.26) were additionally and independently associated with the likelihood of help seeking. In both crude and adjusted analyses, the following beliefs and attitudes were not associated with help seeking: mental illness can be cured, feeling embarrassed and ashamed about mental illness, feeling uncomfortable about discussing mental problems, the respondent's problems can be overcome without help, drugs do more harm than good, mental health professionals can do very little to help, and religion or spiritual beliefs are an important source of support. Paradoxically, those who felt the most uncomfortable about discussing mental problems were more likely to have sought help. A report of poor family support was not consistently associated with help seeking after we controlled for confounding variables.
The results of our survey confirm the common observation that a large proportion of adults with depressive and anxiety disorders do not seek help. Not surprisingly, the presence of a DSM-IV depressive or anxiety disorder and resultant functional disability were strong independent predictors of help seeking. However, objective needs factors did not completely explain help-seeking behavior for mental health needs. In Singapore's health care and social service system, which provides equitable and easy access to affordable care over a small geographical area (700 km2), there are no major structural barriers to mental health services. Patients with mental health problems receive affordable primary care treatments from general practitioners or subsidized specialist care in the public psychiatric services (one large psychiatric institute and four general hospital psychiatric units). The main determining factors for seeking help are not health service system factors, such as availability and access to primary and specialist care for mental illnesses, but personal and social factors. Health beliefs and perceptions about mental illness and health care are thus crucial factors that need to be investigated.
We found that self-rated mental health was an additional independent predictor of help seeking. In this study poor self-rated mental health was correlated with the presence of a psychiatric disorder (κ=.35, p<.001). Our study thus suggests that self-rated mental health status is a valid indicator of mental health status and use of health services.
Contrary to expectation, we could find no strong supporting evidence in this study that beliefs were associated with help seeking—for example, beliefs about stigma or embarrassment; difficulty in discussing mental health issues; beliefs about the curability of mental illnesses, the harmfulness of medications, or the effectiveness of mental health professionals; perceived ability to solve problems; and religious or spiritual beliefs. Paradoxically, those who felt the most uncomfortable about discussing mental problems were more likely to have sought help. A possible explanation is that perceptions of distress and suffering and the thresholds for seeking help vary among individuals with the same level of illness severity and vary within the same individual at different points in time. Such concerns and apprehensions are, nevertheless, surmountable. Hence, help-seeking behavior may not remain consonant with held beliefs over time, as a result of positive or negative experiences with subsequent service contact. In a cross-sectional study, it is difficult to disentangle these changing two-way interactions over time. To address this issue, we also reanalyzed the data after excluding respondents known to be in treatment; however, the results were not substantially altered. Our results thus suggest that if health belief was indeed a determining factor for help seeking, it was neither a consistent nor a robust factor.
This appeared to be true also with social factors. Family members or significant others within an individual's social network have been shown to play a major role in shaping help-seeking behavior (10). We were not able to demonstrate this relationship convincingly in our study, and perhaps there was even a suggestion that those with somewhat poorer family support appeared more likely to seek help. It is possible that in families with good psychosocial support, reluctance to have a member identified as having a mental illness may combine with active support to reduce the likelihood of seeking help.
Because of the large sample and high response rate in this study, we were able to examine simultaneously a large number of variables and to obtain robust findings. Respondents without a diagnosable disorder might include a small number with psychotic disorders or substance abuse and other disorders, resulting in some underestimation of effects. The limitations of a cross-sectional design were such that we were not able to assess earlier patterns of health beliefs and help-seeking behavior. It is difficult to unravel the complex personal beliefs and cultural influences in the path to care of people with mental illness. Qualitative research methods should better elucidate these belief and culture factors, especially in regard to the reasons why people do and do not seek help. Such a positive approach to understanding the personal factors influencing perceived need and use of mental health services could translate into public education that emphasizes psychological self-awareness and self-assessment of mental health as a tool for action-oriented programs.
The likelihood of help seeking was clearly predicted by objective measures of need based on diagnosis and poor functioning. Subjective measures based on self-rated mental health were strongly predictive of help seeking. We could not demonstrate that beliefs about mental illness and mental health services and family and social support were associated with help seeking, possibly because of limitations in measuring these factors and confounding by other unmeasured factors. If they were indeed determining factors for help seeking, their effects were probably neither consistent nor robust.
The research was supported by grant NMRC-0846-2002 from the National Medical Research Council, Ministry of Health, Singapore.
The authors report no competing interests.