Psychiatric morbidity among adult patients in a semi-urban primary care setting in Malaysia

Most of the psychiatric morbidity in the community is seen at the primary care level [1‐5]. Studies using screening instruments have reported prevalence rates ranging from 16 to 43% of general practice attenders [1, 6‐8]. A study of psychiatric morbidity using the General Health Questionnaire (GHQ) reported rates up to 30% [1]. In a Malaysian primary health setting the prevalence of psychiatric morbidity using GHQ-30 item version was 26.7% [9]. In the general population, as measured by GHQ-12 in the Malaysian National Health and Morbidity Survey, the prevalence of psychiatric disorder was 11% [10].

It has been reported that primary care practitioners miss about one third of psychiatrically ill people [4, 11, 12]. A number of reasons have been adduced for this. Patients seeing their primary care doctors tend to somatize their emotional distress, presenting with physical symptoms rather than overt psychological symptoms [13]. Medical history is often taken in conditions of little privacy thereby discouraging patients from sharing sensitive aspects of their distress [6]. Primary care practitioners (PCP) may also not be confident in diagnosing and treating psychiatric disorders [14, 15] Moreover PCP tend to have limited time in which to obtain a psychiatric history [15].

The consequences of psychiatric morbidity such as depression when it is not identified and treated can be severe. These include suicide, loss of jobs and relationships, and deterioration in physical health including higher risk of myocardial infarction [16‐18]. Early detection is important because it would reduce not only the above mentioned consequences but also unnecessary suffering for the patients [19, 20]. In view of the serious consequences of psychiatric morbidity it would be useful for us to explore the prevalence of psychiatric morbidity in our own local setting.

The use of self-report screening tools in primary care settings is potentially helpful in view of the impediments to recognition and treatment of mental disorder in primary care. A number of screening tools exist, including GHQ, Self-rating Questionnaire (SRQ) and the Patient Health Questionnaire (PHQ) [21‐23]. The aim of this paper is to report the prevalence of psychiatric disorder in a semi-urban primary care setting using the PHQ.

This is a cross-sectional study carried out in a semi-urban primary healthcare center located south of Kuala Lumpur. The center is a government run healthcare facility which is easily accessible by public transport. The center caters for the health needs of a population of 603,800 people of whom a majority are from the lower middle socioeconomic group. The center provides general medical outpatient care, emergency services and special clinics for antenatal, diabetes, hypertension, psychiatric illnesses and HIV-related diseases. The average attendance in a month is about 5,000 cases (100 to 150 patients per day), a majority of whom are antenatal cases seen by nurses and midwives. Other facilities include laboratory investigations, ECG, radiography and ultrasound. The clinics are run by a primary healthcare physician, one medical officer, eight nurses, two midwives, two health attendants, two assistant pharmacists and one laboratory technician.

Patients aged from 18–70 years were selected based on their registration number and subjected to a systematic random sampling. Exclusion criteria were patients who did not consent and those who were not literate.

The sociodemographic characteristics of the sample are shown in Table 1. A total of 267 subjects returned questionnaires during the study period. Four questionnaires were incomplete and were excluded from analysis. The sample consisted of 97 (36.9%) young adults aged 18 to 29 years, 91 (34.6%) aged 30 to 39 years, and 75 (28.5%) aged 40 to 70 years.

Of the 263 respondents who returned completed questionnaires, 233 (88.6%) were married. The majority of the respondents were female (n = 188, 71.7%). Malays formed the majority ethnic group (n = 199, 75.7%), followed by Indians (n = 32, 12%), Chinese (n = 29, 11%) and other races (n = 3, 2%).

Most of the respondents had at least secondary school education (64.6%) and 69% were employed. Almost half (46%) were from the lower income group.

About twenty percent had a previously known medical illness (including diabetes mellitus, hypertension, coronary heart disease or bronchial asthma). One respondent had previous psychiatric illness while another reported a family history of psychiatric illness (0.4%). Only 6.8% reported having a history of stress or losses in the previous four weeks.

It was found that 24.7% of respondents had at least one PHQ positive diagnosis. Some had more than one diagnosis. The diagnoses included depressive illness (n = 38,14.4%), somatoform disorder (n = 32, 12.2%), panic and anxiety disorders (n = 17, 6.5%), binge eating disorder (n = 9,3.4%), alcohol abuse (n = 6, 2.3%) (Table 2).

To analyze the association between PHQ positive status and sociodemographic and clinical characteristics, chi-square test was done with p < 0.05 taken as statistically significant.

Table 3 shows that age group was the only sociodemographic factor significantly associated with positive PHQ. It was significantly associated with young age group between 18 to 29 years. (p = 0.036)

Having a history of stressors in the last four weeks was found to be significantly associated with PHQ positive (p = 0.044)

Screening for psychiatric disorders in primary care is an important step to improve services. It should prompt physicians to consider further full diagnostic interview and referral to specialized psychiatric services whenever necessary [5]

An earlier study [25] used the GHQ-30 in detecting psychiatric morbidity in an urban Malaysian primary care population. In this study the PHQ is used because it is designed to provide diagnoses of mental disorders in primary care [15, 26]

In this study it was found that the prevalence of psychiatric disorder among adult patients attending a primary care center was 24.7%. This is similar to studies done in other countries which reported prevalence between 25% and 35% [2, 27‐29]. A study done in Taiwan showed a higher prevalence of 38.2% [30]

A previous Malaysian study of probable mental disorder in primary care, using GHQ, and carried out in an urban population attending private outpatient clinics in Kuala Lumpur, reported a prevalence of 29.9% [25].

The overall prevalence of psychiatric disorder (24.7%) and the prevalence of depressive illness (14.4%) are within the expected range in a primary care setting [2, 27‐29]. The prevalence of somatoform disorder (12.2%) and the lower than expected prevalence of anxiety disorders (6.5%) perhaps reflect the tendency of Asian patients to present their psychological distress in somatic symptoms [31‐33]. An interesting finding from this study is a finding of the higher than expected prevalence of binge eating disorder (3.4%) which is, to the best of our knowledge, the first estimate of prevalence of this disorder in a primary care setting in Malaysia.

Other studies found that being female, unemployed, separated or divorced is associated with a higher probability of psychiatric disorder [1, 2, 4, 10, 29, 34‐36]. However in this study, the only two factors found to be associated with psychiatric morbidity were age between 18 to 29 and a history of recent stressors.

Inferences from this study should be drawn with caution because of several limitations. The questionnaire was validated against the GHQ and not against a structured clinical interview which would have given better data on psychometric properties of the PHQ. There was a preponderance of Malays and women in this sample rendering the findings not applicable to primary care settings with different ethnic composition. It is more representative of semi-urban primary care centers with a preponderance of Malay attenders from the lower and lower middle income groups.

Systematic random sampling was carried out thereby reducing sampling bias. The sample size was adequate.

The healthcare centre in which the study was carried out shares common features with many other centres in rural Malaysia. Hence the study is potentially useful in estimating the prevalence of psychiatric disorder in similar settings.