Patient health questionnaire for screening psychiatric disorders in secondary healthcare.

BACKGROUND: The adult population often suffers from a number of physical and mental problems. This study was conducted to determine the proportion of mental illnesses in adult population visiting the outpatient departments at Dr. TMA Pai Rotary Hospital, Karkala and to study the socio-demographic correlates of psychiatric disorders.
MATERIALS AND METHODS: A cross-sectional study was conducted during March 2004 among 193 adult individuals of 18 years and above at Dr. TMA Pai Rotary Hospital, Karkala, Karnataka. Data was analyzed by the statistical package for social sciences version 10.0 for windows and results were expressed in terms of proportions and their 95% confidence intervals (CI). Chi-square test, multiple logistic regression with adjusted odds ratio and its 95% CI.
RESULTS: The proportion of psychiatric disorders in adult population was determined to be 39.9%. Proportion of psychiatric morbidity among males and females were 36.2 and 42.2%, respectively.
CONCLUSION: This study revealed that socio-demographic correlates like age group of 50 years and above, unemployed or housewives, living alone, and a history of psychiatric illness in the family were independently associated with psychiatric disorders in adult population.

INTRODUCTION

Today, the adult population does suffer from a number of physical and mental morbidities along with psychosocial problems. An estimated 500 million people worldwide are believed to suffer from neurotic, stress-related, and somatoform disorders. A further 200 million suffer from mood disorders, such as chronic and manic depression.[1] Nearly 400 million suffer from mental or neurological disorders or from psychosocial problems such as those related to alcohol and drug abuse. It is estimated that five out of the ten most disabling disorders are psychiatric in nature. Unipolar major depression, alcohol abuse, bipolar affective disorder (manic depression), schizophrenia and obsessive-compulsive disorder (OCD) are among the 10 leading causes of disability worldwide in 1990.[2]

Though psychiatric problems impose a heavy burden of morbidity and disability on the society, management of mental ill health remains at the bottom of medical packing order. Only the most severe cases, such as schizophrenia or manic depression, receive what minimal care there is, even in developed countries. Coping with mental health problems have become more difficult because we do not have information regarding people not getting the help they need; help that is available; help that can be obtained at no great cost.[2]

Kessler et al.,[3] employed a revised version of composite international diagnostic interview (CIDI) with a national probability sample of 8,098 aged between 15-54 years’ non-institutionalized civilian population in the US. Forty-eight percentages of respondents reported at least one lifetime disorder, and 29.5% reported at least one DSM III R (Diagnostic and Statistical Manual of Mental Disorders, Third Edition) disorder over a 12-month period.

Since 1960, epidemiological studies of psychiatric morbidity in different samples of the Indian population have been conducted. The prevalence rate of mental disorders differed from time to time in different studies under different settings. The Bhore Committee (Govt. of India, 1946) extrapolated from rates in UK and USA and concluded that mental patients requiring hospitalization in India be taken as 2 per 1,000. Again in 1966, the Mental Health Advisory Committee of the Govt. of India suggested a probable prevalence of mental illness of 20 per 1,000 population in general, 18 per 1,000 for semi-rural and 14 per 1000 for rural areas.[45]

A meta-analysis of 13 epidemiological studies consisting of 33,572 people, belonging to 6,550 Indian families was attempted by Reddy and Chandrashekar.[6] The estimated prevalence rate is 58.2/1,000. Allowing an error of 20%, as several studies did not include neurotic disorders and drug addiction. They calculated the prevalence rate to range between 46.6 and 69.8 per 1000 population.[6]

The national prevalence rates for all mental disorders as estimated by Ganguli[4] after analyzing 15 epidemiological studies on psychiatric morbidities in India were 70.5 (rural), 73 (urban), and 73 (rural + urban) per 1,000 population.

Carstairs and Kapur[7] defined a case as “one who had one or more symptoms” of 124 psychiatric symptoms included in the Indian Psychiatric Survey Schedule developed and standardized by themselves. By this definition, they reported a case rate of 370 per 1,000 population.

Though mental health problems are common in adult age, few studies had been conducted in India, to understand the problem. A study on psychiatric morbidity in Kukundoor village of Karkala taluk, Karnataka by Ajay[8], using the SCAN 2.1 version (WHO, 1998) questionnaire, determined a 1-month point prevalence of 63.8% of mental disorders. The prevalence was estimated to be 75% for the migrant and 57.45% for the nonimmigrant population among adult community in the village. But no study had been conducted using Patient Health Questionnaire (PHQ) for screening mental health problems in adult population of Karkala.

MATERIALS AND METHODS

Karkala taluk belongs to Udupi district of Southern Karnataka. It has an area of 1091 km2, with a population of 180,453 living in about 50 villages. Its high literacy rate of 81.4%, birth rate of 16.9 per 1,000 population and a favorable sex ratio of 1,155 reflect the socioeconomic development of this region. The infant mortality rate was estimated at 28/1,000 live births in 1991. Majority of the population are Hindus (85.8%), followed by Christians (7.7%), Muslims (6.4%), and Jains (1%). Migrant population from Tamil Nadu working as laborers is also found in some pockets of this region. Dr. TMA Pai Rotary Hospital is a district level hospital that provides mainly maternal and child health services along with ophthalmic care to the rural population of Karkala.

A cross-sectional study was conducted for 1 month (1st to 31st March 2004) involving 193 adult individuals in the age group of 18 years and above to determine the proportion of mental illnesses in adult population visiting the outpatient departments (OPDs) at Dr. TMA Pai Rotary Hospital, Karkala and to study the sociodemographic correlates of psychiatric disorders. Due to feasibility constraints, purposive sampling method was applied to identify the study subjects.

Inclusion criteria

Those who were 18 years and above, visited the hospital during the study period, and gave a verbal informed consent to participate were included in this study.

Exclusion criteria

All those individuals who were previously diagnosed as mentally challenged by the psychiatrists and visited the psychiatric OPD for regular follow up, were excluded form this study

As it is a well-established fact that patients admitted in hospitals are more prone for developing psychiatric illnesses due to their diseased conditions and isolation from family, so all patients admitted in any inpatient ward of the hospital were also excluded.

Study instruments

A face sheet consisting of information regarding the household and family particulars of the respondent was used for data collection. Screening for psychiatric disorders was determined using the instrument PHQ. The data collection tools were translated from English into local language Kannada and Hindi, by a language expert not related to this study. They were later back-translated into English by another independent language expert, not acquainted with the original versions. The back-translation was subsequently compared with the original version by a psychiatrist for conceptual equivalence of the items. Necessary finer adjustments were made to convey the correct information to the participants.

Validation and utility of a self-report version of PRIME-MD

Primary Care Evaluation of Mental Disorders (PRIME-MD) consists of a one-page PHQ and a 12-page clinical evaluation guide for physicians to assess four groups of mental disorders (mood, anxiety, alcohol, and somatoform) and eating disorders. The PHQ is a self-administered instrument comprising of 26 yes or no questions about signs and symptoms present during the past 1 month and a question about the patient's overall health. Based on the answers, physicians determined which, if any, of the evaluation guide diagnostic modules should be used. The PRIME-MD arose as a screening instrument, but its clinical application has been restricted by its administration time. This criterion standard study was undertaken between May 1997 and November 1998 to determine whether the self-administered PRIME-MD PHQ is valid and useful for diagnosing mental disorders in primary care compared with the original clinician-administered PRIME-MD. Sixty-two primary care physicians (21 internal medicine, 41 family practice) assessed 3,000 adult patients drawn from eight US primary care clinics. Of these, 585 patients were assessed by a mental health professional within 48 h of completing the PHQ. Measures of outcome were PHQ diagnoses compared with diagnoses made independently by mental health practitioners, function status measures, disability days, health care utilization, and treatment/referral decisions. PHQ was used for diagnosis in 825 (28%) of the 3000 patients and 170 (29%) of the 585. As with the original PRIME-MD, agreement between PHQ diagnoses and those made by mental health practitioners was good (for diagnosis of any one or more PHQ disorder, kappa=0.65; overall accuracy=85%; sensitivity=75%; specificity=90%). Patients with PHQ diagnoses had an impairment that is more functional and more disability days, and used more health care resources than did those without PHQ diagnoses (for all group main effects, P<0.001). It took much less time, on average, for the physician to review the PHQ than to conduct the original PRIME-MD (<3 min for 85% vs. 16% of the cases). In spite of the fact that 80% of physicians reported that they would find regular use of the PHQ helpful, new management actions were enacted or planned for only 117 (32%) of the 363 patients with one or more PHQ diagnoses previously unidentified. This study suggests that the PHQ has diagnostic validity equivalent to the original clinician-administered PRIME-MD while being more efficient to use.[9]

Data collection procedure

All investigators were trained by psychiatrists on proper procedure to administer the questionnaire. The study instrument was pre-tested on a small group of individuals (n=10) to determine whether it optimally suited the study conditions on accounts of feasibility and acceptability. The results obtained from this data were kept separate and not included in the main study. After informed verbal consent was obtained, a designated respondent was administered a selected set of questionnaires by the investigators. Care was taken to ensure privacy and confidentiality of the interview as part of the study. A brief general health check-up of the respondent was conducted at the beginning to establish a good rapport with him and also to gain his confidence. On an average, it took nearly 10 minutes for the investigators to administer the PHQ questionnaire. The diagnoses generated by the screening instrument in this study were strictly kept confidential and were reconfirmed by consulting a psychiatrist before arriving at a final ICD-10 diagnosis for data analysis. Those who proved to be positive in this screening instrument were handed over a referral slip and confidentially requested to visit the Psychiatry OPD of Dr. TMA Pai Rotary Hospital, Karkala at the earliest for a free consultancy.

Data analysis

The collected data was tabulated and analyzed by using the statistical package SPSS (Statistical Package for Social Sciences) version 10.0 for Windows. Findings were described in terms of proportions and their 95% confidence intervals (CI). Chi-square test was applied to study the relationship between different sociodemographic variables and psychiatric morbidity. To determine the independent effect of various factors on psychiatric disorders, multiple logistic regression was performed and their significance was estimated in terms of adjusted odds ratio (OR) and its 95% CI. P value less than 0.05 was considered as significant.

RESULTS AND DISCUSSION

In this study, 193 individuals in the age group of 18 years and above were interviewed after obtaining a verbal informed consent. The baseline characteristics of the study population revealed that there were 43.5% males and 56.5% females. Majority (40.4%) belonged to the age group of <30 years while 72.0% were married. 69.9% of the participants were literates, which is comparable with the national literacy rate of India which is 65.38% (2001). Though 77.4% among males were literate, only 64.2% of females were literate. The female literacy rate among the participants was higher as compared to national female literacy rate of 54% (2001).[10] In this study, 23.8% of the individuals belonged to low socioeconomic status (below poverty level). Poverty level is defined as expenditure required for a daily calorie intake of 2,400 per person in rural areas and 2,100 in urban areas. This expenditure is officially estimated at Rs. 181.50 now revised as Rs. 272.00 per capita per month in rural areas and Rs. 209.50 now revised as Rs. 292.00 in urban areas at 1991-92, now updated with 2001-02 prices.[10]

In this study, out of 193 individuals interviewed, 77 (39.9%) were found to be having psychiatric disorders. This is consistent with the observations made by Carstairs and Kapur,[7] who reported a case rate of 370 per 1,000 population. However, this proportion is less than the study by Ajay,[8] who reported a 1-month point prevalence of mental disorders to be 63.8% in adult population of Kukundoor village of Karkala taluk. A comparatively low proportion (20.7%) of psychiatric disorders was previously reported by Nandi et al.[11]

Among those having psychiatric disorders, 77 (39.9%) in the present study, majority 27 (35.1%) were suffering from somatoform disorders, while 26 (33.8%) from major depressive disorder, 9 (11.7%) from other depressive syndromes, 22 (28.6%) from panic syndrome, 23 (29.9%) from other anxiety syndromes, 3 (3.9%) from bulimia nervosa (eating disorders), and 5 (6.5%) from alcohol abuse. Ajay[8] also observed a high proportion of mood disorders (32.6%) and anxiety disorders (20.8%) in adult population of Kukundoor village of Karkala taluk.[8]

Majority (53.2%) of these cases were suffering from only one, 44.2% from two and 2.6% from three types of psychiatric morbidities. More than one diagnoses (mean=1.5, SD=0.55) was attributed to majority of these cases. These observations are consistent with the findings by Kessler et al.,[3] who also reported major depressive episode, alcohol dependence, social phobia, and simple phobia as most common psychiatric morbidities in adult population. In his study, women had higher prevalence than men of affective disorders, anxiety disorders, and non-affective psychosis. Men had higher rates of substance use disorders and anti-social personality disorders.[3]

In this study, the proportion of psychiatric disorders was higher among females (42.2%) than males (36.9%), but this difference was not found to be statistically significant. Our study findings are consistent with observations by Verghese et al.,[12] Shah et al.,[13] Sen et al.,[14] and Shaji et al.,[15] who reported a significantly high proportion of psychiatric disorders among females. However, Ajay[8] reported a significantly high proportion of psychiatric morbidities among males.

The age of the respondents varied between 18 and 73 years, while the mean age was found to be 35.2 years (SD=10.8). The proportion of psychiatric disorders was highest (66.7%) in the age group of 50 years and above. The difference in proportion of psychiatric morbidity between different age groups was found to be statistically significant (χ2=10.97, df=3, P=0.012*). The proportion of psychiatric disorders showed a positive linear trend of increase with the progression of age, which was also found to be statistically significant. This is consistent with the findings by Shah et al.,[13] Sen et al.,[14] Shaji et al.,[15] and Ajay[8] who also reported a high proportion of psychiatric morbidities among the elderly. Barua[16] also reported a significantly high proportion of depressive disorders among the elderly with progression of age.

Proportion of psychiatric disorders was high (67.4%) among individuals belonging to low socioeconomic status (below poverty level) as compared to those above poverty level (31.3%) and the difference between these two groups was also found to be statistically significant. Sen et al.,[14] and Shaji et al.,[15] had also reported the same.

Though majority of the respondents were married (72.0%); but a majority of the unmarried, widowed, or divorced individuals were men (55.6%). Among the singles (28%), majority were widowed (53.7%) while 40.7% were unmarried and 5.6% divorced. Proportion of psychiatric disorders was higher among the unmarried, widowed, or divorced individuals (61.1%) as compared to their married counterparts (31.7%). This difference was found to be statistically significant. This is in contrast to the findings by Verghese et al.,[12] and Ajay[8] who reported a high proportion of psychiatric morbidities among married individuals.

In the category of unemployed or housewives, majority of the respondents were housewives (75.0%). Proportion of unemployed or housewives affected with psychiatric disorders was 68.4%. It was 40.6% and 13.5 among unskilled and skilled laborers, respectively. The difference was found to be statistically significant. This is consistent with the observations by Shah et al.,[13] who also reported a high proportion of mental illnesses among unemployed individuals.

We observed that proportion of psychiatric illness was higher (73.2%) among those who lived alone as compared to those who lived only with children or relatives (40.3%) and who lived with spouse (24.4%). This difference between the groups was found to be statistically significant. Our findings are similar to the observations by Kennedy et al., (1984-85)[17] who reported a significantly high proportion of depressive disorders among those who lived alone.

Proportion of psychiatric disorders among illiterates was higher (63.8%) as compared to literates (29.6%). The difference between the two groups was found to be statistically highly significant. Similar findings were reported by Verghese et al.,[12] Shah et al.,[13] and Ajay.[8]

Proportion of psychiatric disorders among the individuals who had a history of death in their family within the last 6 months was found to be 28.6%, but the difference between the groups was not found to be statistically significant. Our findings are in contrast with the observations by Barua[16] and Kennedy et al., (1984-85)[17] who also reported a significantly high proportion of depression among those who had a history of death in their family within the last 6 months.

Table 1 shows the findings of multiple logistic regression analysis performed after univariate analysis. Multiple logistic regressions revealed that age group of 50 years and above (adjusted odds ratio=4.82 and 95% CI=1.3-17.5), unemployed or housewives (adjusted odds ratio=4.76 and 95% CI=1.3-17.6), living alone (adjusted odds ratio=4.7 and 95% CI=1.3-17.7), and a history of psychiatric illness in the family (adjusted odds ratio=10.74 and 95% CI=1.3-91.0) had independent significant association with psychiatric disorders in adult population.

Table 1

Results of multiple logistic regression analysis

CONCLUSION

A cross-sectional study was conducted among 193 adult individuals of 18 years and above at Dr. TMA Pai Rotary Hospital, Karkala. The objectives were to determine the proportion of mental illnesses and to study the correlates of psychiatric disorders in adult population. In this study, the proportion of mental illnesses in adult population was determined to be 39.9%. Proportion of psychiatric morbidity among males and females were 36.2 and 42.2%, respectively. Multiple logistic regression analysis revealed that sociodemographic correlates like age group of 50 years and above, unemployed and housewives, living alone, and a history of psychiatric illness in the family were independently associated with psychiatric disorders in adult population.

Limitations

Due to feasibility and time constraints, the sample size could not be increased further. Inability to apply any probability sampling technique along with a small sample size restricted the generalization of these study findings

Majority of the patients were unaware of the drugs they had consumed in recent past. So, a psychiatric illness, as a result of any drug reaction, could not be confirmed

All diagnoses generated by this screening instrument needed clinical confirmation by psychiatrists before initiating appropriate management procedures.

Recommendations

More studies should be conducted using the PHQ instrument to confirm the findings of this study

It should be made mandatory to screen all patients for mental disorders wherever feasible in every outpatient department.