Validity of Best-Estimate Methodology in Assessing Psychosocial Risk Factors and Making Psychiatric Diagnoses in Hong Kong Chinese Who Attempt Suicide

Article excerpt


Objectives: To examine the validity of the best-estimate method for making psychiatric diagnoses and determine potential psychosocial risk factors in a cohort of Hong Kong Chinese who attempted suicide.

Participants and Methods: Seventy-one persons attempting suicide and their proxy-informants were interviewed separately to ascertain each patient's diagnosis (according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition Axis I diagnosis), psychosocial profile, and life circumstances surrounding the index suicide attempt.

Results: There were substantial levels of agreement, high sensitivity and specificity on Axis I psychiatric diagnoses in subject-proxy pairs. Levels of subject-proxy agreement were substantial for other risk factor domains such as physical diagnoses, suicidal behaviour profile, social networking, and most life-event items. Fair-to-modest levels of agreement were observed in perceived well-being in general health, health service utilisation and life-events involving interpersonal conflicts within family and peer groups.

Conclusions: Results support the validity of the best-estimate methodology for assessing psychosocial risk factors and making psychiatric diagnosis among Hong Kong Chinese who attempt suicide.

Key words: Interview, psychological; Suicide, attempted


Suicide is a major public health problem affecting all nations. In 2000, the World Health Organization (WHO) reported that more than one million suicide deaths occurred worldwide. (1) Such a figure approximates to 14.5 suicide deaths per 100,000 people or one suicide death occurring every 40 seconds. From a public health perspective, suicide is the 13th leading cause of death worldwide and its detrimental effect has impacted all age-groups. To address the worldwide public health challenge posed by suicidal behaviour, there have been an increasing number of risk factor studies over the past few decades. The two main research approaches depend on community surveillance and psychological autopsy (PA). The PA approach, first developed and systematically described by Shneidman and Farberow in 1961, (2) involves enquiries into individual suicide cases, using in-depth psychosocial interviewing to gather information from proxy-informants. The psychosocial interview is designed to elicit detailed psychological profiles and psychosocial circumstances leading to suicide deaths. This method is generally regarded as the hallmark for risk factor research into suicide, despite its unavoidable limitations such as recall and attribution bias of the informants. (3) One indirect way to assess the validity of PA is by the 'best-estimate methodology among suicide attempters'. This method involves living subjects with risk factors that are otherwise comparable to those of deceased persons. Another interviewer conducts the same interview with a proxy-informant related to the living subjects. The degree of agreement between the information obtained from the two sources can be determined. Suicide attempters, despite their heterogeneous clinical profiles, may serve as the best 'proxy' study population to our reference population of suicide completers. (4)

A few studies have examined the reliability of proxy information in various elderly and adolescent psychiatric populations, (5,6) and suggest that informants under-report mood and substance abuse problems. Using the best-estimate method in late-life suicide attempters, Conner et al (4) reported satisfactory proxy-subject agreement on diagnoses of mood and substance dependency diagnoses; and some degree of agreement on behavioural profiles of suicide attempters and certain aspects of social support and stressful life events.

Suicide research in Chinese communities is challenged by social taboos. (7) The response rates for PA studies in Hong Kong Chinese were much lower than the rates encountered in Caucasian populations. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.