Study on prevalence of suicidal ideation and risk factors of suicide among patients visiting psychiatric OPD at Shree Birendra Hospital, Kathmandu Nepal

Introduction Suicide is a global public health issue. Several environmental, psychosocial, behavioral factors along with physical, sexual, and emotional abuse have been associated with suicidal ideation and attempts. Childhood physical, sexual abuse, and health risk behaviors are also associated with suicidal attempts. The suicidal ideation prevalence varied from 1 to 20% and it varied with study population, geography, age group, gender, and other factors. The Beck suicidal ideation scale is an effective tool for assessing the major suicidal ideation with a six cut-off score. Materials and method 160 patients who met the inclusion criteria were enrolled into this cross-sectional study after random sampling among the patients visiting the Psychiatric OPD of Shree Birendra Hospital, Kathmandu, Nepal. The Semi-Structured Interview Schedule (SSIS), Beck Scale for Suicide Ideation (BSS), and Kuppuswamy’s Scale were used to collect the data from the patients enrolled in the study. The Chi-square test and binary logistic regression analyses were used to identify and differentiate the factors associated with high suicidal risk. Results Out of total 160 patients, 65% (n = 104) were female, 92.5% (n = 148) were married, 61.9% (n = 99) were residing in urban area, 93.1% (n = 148) were Hindus, 74.4% (n = 119) patients were living in the nuclear family, 5% (n = 8) patients had family history of psychiatric illness and 10.6% (n = 17) patients were using the substance of abuse. In the Beck scale for suicidal ideation questionnaire, 87.5% (n = 140) patients had moderate to strong wish to live, 89.4% (n = 143) patients responded as they would take precautions to save a life, 88.8% (n = 142) patients had such ideation/wish for brief, 96.3% (n = 154) had not considered for specificity/planning of contemplated suicidal attempt, 91.9% (n = 147) patients stated that they would not attempt active suicide because of a deterrent example from family, religion, irreversibility of the act and 98.1% (n = 157) patients had revealed ideas of deception/concealment of contemplated suicide openly. 16.9% (n = 27) of participants were categorized as high risk for suicide while 83.1% (n = 133) patients were as a low-risk category for suicide based on the Beck scale for suicidal ideation scoring. Conclusion In conclusion, this study found that most of the suicidal attempts were done as an act of impulse and it is higher among female and married individuals residing in the urban areas. This study did not establish any statistically significant association or differences among independent variables with the higher risk scoring in the Beck suicidal ideation scale.


Introduction
Suicide is a global public health issue. To date, several environmental, psychosocial and behavioral factors are associated with suicidal ideation, suicidal attempts, and suicide in studies conducted in various countries [1]. Physical, sexual, and emotional abuse has been associated with adolescent suicidal ideation and attempts [2]. In addition, childhood physical and sexual abuse appears to be risk factors for future suicide attempts. It is noted that suicidal behaviors often coexist with other health risk behaviors [1,2]. Recognizing health risk behaviors and other factors that influence suicidal behavior can improve understanding about those at high risk of suicide associated with suicidal behaviors in adolescents. Studies in other countries also revealed that parental support, family factors as well as psychological factors also influence suicidal behavior [3].
It is observed that period prevalence of suicidal ideation varies widely, ranging approximately from 1% to 20% [3]. Also, the prevalence varies with study population, geography, age group, sex, and influence by other factors as mentioned. A study among medical students in Nepal showed suicidal ideation among 10% of students while another study on adolescent students also showed similar findings [4,5]. In a study conducted among the community and health care seeking populations of five low-and middle-income countries including Nepal showed suicidal ideation in 10.3% [6]. Several risk factors for suicidal behavior have been identified, for example, low socioeconomic status, experienced child abuse, and mental disorders. One of the studies conducted in 2016 at Dharan showed a cut-off score of 6 on Beck's suicidal ideation scale which is a scale to measure suicidal ideation in this study [7]. Based on a hospital-based study among attempted suicide showed OP poisoning was the most common method [8].
The objective of this study was to study demographic profiles among patients included in the study, meeting the inclusion criteria, and to study the severity of suicidal ideation using the Beck scale among patients having suicidal ideation visiting the psychiatric outpatient department (OPD) of Shree Birendra Hospital (SBH).

Study design
The design of the study was a hospital-based cross-sectional study.

Sampling method
A simple random sampling method was used. Subjects were selected randomly by luck draw among patients visiting the OPD daily. Two number 0 and 1 mentioned in the paper was kept folded in a lucky draw box. Those who received 0 after the lucky draw was not enrolled in the study and those who drew 1 in the lucky draw were enrolled.

Study site and population
The study site was Psychiatric OPD of Shree Birendra Hospital located in Kathmandu district of Nepal. The study population included patients who visited the Psychiatric OPD of SBH.
• Psychiatric disorders diagnosed (specified in the study objectives) by consultant psychiatrist using ICD-10, DCR • Literate (with formal schooling/education) • Those who provided consent to be enrolled in the study

Exclusion criteria
• Neurodevelopmental Disorders • Acute Mental Disorders

Sample size
The sample size was calculated using a 95% confidence level with ±5% standard error for an unknown population considering a 10% period prevalence for suicidal ideation based on earlier studies in the Nepalese context [6]. The calculated sample size was 139. The study variables were age, sex, education, personality, socio-demographic profile, psychiatric diagnosis and suicidal Ideation.

Data collections techniques
Patients who visited the psychiatry OPD were recorded and selected in fixed numbers by lucky draw. Every selected patient was explained about the study and consent was taken in written form. Subsequently, patients were assessed through socio-demographic Pro-forma and self-rated Beck Suicidal Ideation Scale [9]. Data collection was done from July 7, 2020, to Jan 29, 2021.

Data collections tools
Data collection for the study was done using Semi-Structured interview Schedule (SSIS), Beck Scale for Suicide Ideation (BSS), and Kuppuswamy's Scale (available as S1 File).

Statistical methods
Data was collected from consented selected patients done using a pre-structured questionnaire, cleaned in Excel, and then imported and analyzed using STATA v.15. Simple descriptive analysis was performed and presented with appropriate frequency tables and cross-tabulation. Kuppuswamy scale, sex, residency of living, marital status and type of the family in which patients were living, family history of psychiatric illness, and psychiatric diagnosis of patients were taken as independent variables while the low or high risk in Beck scale for suicidal ideation taken as the dependent variable.
Logistic regression analysis was performed taking a risk in the Beck scale for suicidal ideation as the dependent variable and other determining socio-economic factors as the independent variable. Sum score of Beck scale for suicidal ideation of > 2 was taken as high-risk category and was the outcome of our interest so logistic regression analysis was run for the high risk for suicidal ideation (> 2) for low risk for suicidal ideation (� 2). Firstly, binary logistic regression analysis was used across risk for suicidal ideation for individual independent variables to estimate the unadjusted odds ratio (OR). Then, multiple logistic regression was used to estimate adjusted OR. For logistic regression analysis, high risk for suicidal ideation was labeled as 1 and low risk for suicidal ideation was labeled as 0, and odds of occurrence of high risk for suicidal ideation concerning low risk for suicidal ideation estimated.

Ethical consideration
The study was carried out after approval from the Institutional Review Committee of the Nepalese Army Institute of Health Sciences (NAIHS). All the participants meeting inclusion criteria consenting for study enrolled in data collection.

Results
In this study, a total of 160 patients meeting the inclusion criteria were included. There were 65% (n = 104) female and 35% (n = 56) male patients and among them, 92.5% (n = 148) were married. Among 160 patients, 61.9% (n = 99) were residing in urban area and 38.1% (n = 61) patients were in rural area. Out of 160 patients, 93.1% (n = 148) of patients were following the Hindu religion. In the study, the majority of the patients were Chhetri (48.1%) followed by Brahmin (20.6%) and the remaining others were as shown in Table 1. 74.4% (n = 119) patients were living in the nuclear family and only 5% (n = 8) patients had a family history of psychiatric illness. About 10.6% (n = 17) patients were using the substance of abuse either alcohol [5% (n = 8)], tobacco/cigarette [5% (n = 8)] or both [0.6% (n = 10)] as shown in Table 1. Among 13 different psychiatric diagnoses reported, anxiety disorder not otherwise specified was the commonest diagnosis [36.25% (n = 58)], followed by depression and psychosis [15.63% (n = 25)] each.
In the Beck scale for suicidal ideation questionnaire, 87.5% (n = 140) patients had moderate to strong wish to live whereas 10.6% (n = 17) had a weak wish and only 1.9% (n = 3) had no wish to live. In the question regarding the wish to die, 76.3% (n = 122) patients had no wish to die whereas 18.8% (n = 30) had a weak wish to die and only 5% (n = 8) had a moderate to strong wish to die. Among 160 patients, only 4.4% (n = 7) outweigh dying for living while 7.5% (n = 12) equaled the living and dying reason, and the remaining patients gave the reason as living outweigh the dying. About 3.1% (n = 5) had moderate to strong desire to make active suicide attempts and 8.1% (n = 13) had a weak desire to do so and the remaining had no desire on it. Regarding the passive suicidal desire, 89.4% (n = 143) patients responded as they would take precautions to save a life while 2.5%(n = 4) patients would avoid steps necessary to save or maintain life and remaining would leave life/death to chance. Regarding the duration of having suicide ideation/ wish, 88.8% (n = 142) patients had such ideation/wish for brief, fleeting periods. 85.0% (n = 136) patients had suicidal ideation on rarely, occasionally, 88.1% (n = 141) patients were rejecting attitude toward suicidal ideation/wish and 6.3% (n = 10) had ambivalent/indifferent attitude toward it. In addition, 93.8% (n = 150) patients had sense of control over suicidal action/acting-out wish, 5.6%(n = 9) had unsure regarding such control and 0.6% (n = 1) had no sense of control. 91.9% (n = 147) patients stated that they would not attempt active suicide because of a deterrent e.g. from family, religion, irreversibility of the act whereas 5.6%(n = 9) had some concern about deterrents and 2.5% (n = 4) had minimal or no concern about deterrents. 95.0% (n = 152) patients had contemplated the suicidal attempt to manipulate the environment or to get attention/revenge whereas 2.5% (n = 4) had done to escape, surcease, solve problems and remaining patients had contemplated for in-between reasons mentioned as above. 96.3% (n = 154) had not considered for specificity/planning of contemplated suicidal attempt and 93.8% (n = 150) patients had no courage or were a too weak or afraid or incompetent sense of "capability" to carry out the suicidal attempt. Only 1.3% (n = 2) patients had expectancy or anticipation of actual suicidal attempt while 96.3% (n = 154) patients did not expect or anticipate the actual attempt and 2.5% (n = 4) were unsure about it.
Only 0.6% (n = 1) had completed actual preparation for the contemplated attempt whereas 3.1% (n = 5) had partial actual preparation and the remaining patients had not prepared in actual. None of the patients under this study had left suicide notes while contemplating the attempt. 98.8% (n = 158) patients had not made definite plans or completed arrangements for final acts in anticipation of death (e.g., insurance, will) following the suicide attempt while 0.6% (n = 1) had thought about it and made the definite plan. 98.1% (n = 157) patients had revealed ideas of deception/concealment of contemplated suicide openly while 1.9% (n = 3) patients held back such ideas as shown in Table 2.
Out of the 160 patients, 66.3% (n = 106) patients belonged to upper-lower class in Kuppuswamy scale, 23.8% (n = 38) belonged to lower-middle class, 7.5% (n = 12) to upper-middle class and 2.5% (n = 4) belonged to lower class in Kuppuswamy scale as shown in Fig 1. In Beck scale for suicidal ideation, out of 160 patients, 83.1% (n = 133) patients had summation of Beck score of � 2 (low risk for suicidal ideation) and 16.9% (n = 27) had summation of Beck score of > 2 (high risk for suicidal ideation) as shown in Fig 2. The Chi-square test was used to check the association among the independent variables and dependent variables (risk for suicidal ideation). From the analysis association between psychiatric diagnosis and Beck scale-based suicide risk category was observed (p<0.05). However, no association was observed between the other independent variables and the dependent variables (Beck scale based suicide risk category) ( Table 3).
Binary logistic regression showed an increase in age lessens the chance of the Beck sum score being >2 (High risk), and diagnosis of adjustment disorder, and depression has higher odds of having Beck sum score being >2 (High risk) in comparison to anxiety disorder not otherwise specified. For the rest of the variables, there were no significant differences in the Beck score category and other independent variables ( Table 4). Adjusting the effect of one variable over another and rerunning the multinomial logistic regression analysis also showed the persistence of such relation.

Discussion
This study showed that most of the patients attempting suicide had suicidal ideations for a brief period and on rare occasions with an intact sense of control over the suicidal ideation. In this study, the suicidal attempt was higher among females, 65% of all patients. The study by  [5]. In this study, about two-thirds of the patients were from urban areas. Almost (92.5%) all patients enrolled in this study were married and this finding was comparable to the findings from the study by Jordans M et al 2017 in which 69.1% to 91.5% of participants were married [6]. In this study, 93.1% of the participants were Hindus which was higher than the finding from the study done by Menezes RG et al. with 75.7% of Hindu participants [5]. In this study, almost half (48.1%) of the participants were Chettri followed by Brahmin (20.6%) and it could be due to their higher predominance in the general population.
In this study, almost all patients (95%) were without a family history of psychiatric illness. When patients were enquired regarding health risk behaviors, only 10.6% of the patients were abusing the substance. In a study by Sharma B et al in 2015, the health risk behaviors among the participants ranged from 7.0% to 49.2% and were smokers, alcohol drinking, taking illicit drugs, and sexual intercourse [1]. This could be due to the difference among the type of study population.
The majority of the participants in our study wished to live following the act of suicidal attempt with only 5% wishing to die. The majority of the patients in this study were opting for the living with no desire to make active suicide and preferring to take precautions to save a life by avoiding passive suicidal desire. The most common cause for contemplating the suicidal attempt among the study patients was to manipulate the surrounding environment or to get attention or revenge and with secondary gain effect. So, the majority of the patients were having the suicidal ideation/wish for brief, fleeting time and rare or occasionally, lacking the  specificity or planning and availability or opportunity of the contemplated attempt, lacking courage, or felt too weak to carry out the attempt. There was a lack of actual preparation for the contemplated act in the majority of the patients and the majority of them did leave the suicide note and did not anticipate the death in the form of insurance, will. Almost all of the patients had concealed their idea of contemplating suicide before their attempt but most of them showed the attitude of rejecting the suicidal ideation/wish. The study by Kim J et al in Korea in 2015 found that the impulsive suicide attempters had shown higher desire to live, less frequency of having suicidal ideation/wish, less pre-occurrence of hopelessness, doing suicidal attempt with a hope to change [10]. This reflects that the majority of the patients enrolled in this study had impulsive suicidal attempts.
Our study showed an increase in age reduces the risk of suicide, however, diagnosis of adjustment disorder, depression, and GAD has higher odds of having the risk of suicide in comparison to anxiety disorder not otherwise specified. Though the statistical relation was observed the confidence intervals were wide due to the small sample of the study, and most of the patients performed suicidal attempts as the act of the impulse. However, our finding of the association of suicidal risk with depression is similar to another study done in eastern Nepal [7]. In the present study, we evaluated some facets of socio-economic, clinical, and behavioral factors for risk of suicide, and the risk of suicide was evaluated using the Beck scale. However, there might be other factors contributing to suicide. So, due to the small sample and limited exploration, further bigger studies exploring more details on the risk of suicide among vulnerable psychiatric patients in the context of Nepal are advised. Additionally, our study is a small size single-center study, so the findings of the present study may vary with others and this needs to be considered while generalizing the findings.

Conclusion
In conclusion, this study found that most of the suicidal attempts were done as an act of impulse and they were higher among female and married individuals residing in the urban areas. Our study showed an increase in age reduces the risk of suicide. However psychiatric comorbidities like adjustment disorder, depression, and GAD has higher odds of having the risk of suicide in comparison to anxiety disorder not otherwise specified. No statistically significant association or relationship was established among the Kuppuswammy scale, sex, residency, marital status, and type of family with the higher risk scoring in the Beck suicidal ideation scale.