A preliminary statistical model for identifying repeaters of parasuicide

Abstract

This paper presents a statistical model constructed using logistic regression to identify those at high-risk of repeating parasuicide. The subjects in the study are Cork city residents who exhibited parasuicidal behaviour between I January and 30 June 1995. Repetition of the behaviour within six months of the index episode distinguishes repeaters from non-repeaters. The model was designed so that it could be used by non-clinicians and hence does not require information relating to psychiatric diagnosis or use of psychiatric services. The proportion of subjects correctly classified remained stable across a range of cut-point probabilities (mean - 86%. range: 83.9-87.5%). Using a cut-point of 0.2. 96% of repeaters and 81% of non-repeaters were correctly classified. Using 0.45 led to the correct identification of 81% of repeaters and 90% of non-repeaters. If these high levels of sensitivity and specificity are maintained in validation tests on future cohorts in Cork city then the model could form the basis of an intervention programme designed to prevent the repetition of parasuicide.     

Predictive scales for parasuicide repetition. Further results.

A six-item predictive scale for parasuicide repetition was retested on a second sample of Italian parasuicides. All the patients firstever admitted after a deliberate self-harm to the psychiatric department of a general hospital during the period April 1973-March 1976 were interviewed and followed-up after an interval of 1 year. The scale showed the same predictive power found in the first sample (1970-73) and gave a range of probability of repetition within 1 year of 15% at a score of 0 up to 44.8% at the scores 3, 4, 5, and 6. Moreover it discriminated between repeaters and non-repeaters at a significant level. Only two items (diagnosis of sociopathy and change of domicile in the last year) of the five items which discriminated repeaters from non-repeaters in the first study were still significantly associated with repetition, while four additional items (previous in-patient psychiatric treatment; previous parasuicide resulting in hospital admission; unemployment; criminal record) appeared as new discriminating factors. In spite of these changes the rate of repetition remained constant over the years at around 26%.     

The profile of parasucide repeaters in Kuwait

A one-year cohort of 92 parasuicides was followed up at 6 monthly intervals for a period of 2 years since their index parasuicide to study the outcome and evaluate the level of social readjustment. One male patient died of suicide (1.1%) and 18 (19.6%) individuals repeated parasuicide using self-poisoning. The rate of repeated parasuicidal behaviour in this investigation was considerably higher than that reported in comparable studies from developing countries but similar to that from a number of western European cities. The data indicated that the probability of further episodes of parasuicidal behaviour increased in the few months after the index episode. Although the repeaters and non-repeaters were essentially similar in most of their sociodemographic characteristics, the former were a distinctive group in many respects. The factors found to be significantly associated with subsequent parasuicide included self-poisoning by prescribed drugs at the index parasuicide, previous parasuicidal behaviour prior to the index, occupational status of a housewife, past history of depression and/or dependence and precipitating life events in the family environment. Contrary to our hypothesized theory, none of the types of attitudes received by the patients from their family members at the index parasuicide related to outcome. The global level of social readjustment of non-repeaters was about three times higher than that of repeaters of parasuicide. The implications of these findings for future policy making were discussed with respect to prevention of parasuicide.     

A multivariate study of predictors of repeat parasuicide

OBJECTIVE: To identify variables which differentiate future repeaters of parasuicide from non-repeaters in a multivariate analysis. METHOD: Interviews were conducted with 507 parasuicide cases; data were collected on precipitating factors for the index parasuicide, psychiatric and medical history, stressful life events, prior history of parasuicide, hopelessness, anger, self-esteem and social adjustment. Individuals were followed for 1-2 years to determine if a repeat parasuicide occurred. RESULTS: A logistic regression model identified four significant predictors of repeat parasuicide: prior history of parasuicide, a history of depression, a history of schizophrenia and poor physical health. A risk factor scale constructed from these four variables showed that the risk of repeat parasuicide increases as the number of risk factors increases. CONCLUSION: This study identifies four key predictors of repeat parasuicide, and provides evidence that the risk of repeat parasuicide increases when multiple risk factors are present.     

Formal and informal help during the year after a suicide attempt: a one-year follow-up

AIMS: The aims of this study are threefold: to investigate formal and informal help with clinical and social needs that patients who attempted suicide received during the year after their attempt; to examine whether help from services was estimated to be adequate; and to look at whether patients who repeated suicide attempt(s) during follow-up differed from those who did not. METHODS: The Camberwell Assessment of Need instrument measuring 22 need areas was used in semi-structured interviews with 140 patients at 1 and 12 months after a suicide attempt. RESULTS: Help given from services was rated as high at both 1 and 12 months in health-related areas, but lower in areas related to social needs. Informal help was initially frequent, with some exceptions. The amount of help from services did not decrease in any need area during follow-up, neither in repeaters nor in non-repeaters. There were no changes in informal help in repeaters, while in non-repeaters informal help decreased in some areas. At both 1 and 12 months, repeaters and non-repeaters mostly found help from services to be adequate. However, in the areas of information, intimate relationships, psychotic symptoms and sexual expression about half of the patients in both groups did not consider that they had received the right type of help. CONCLUSIONS: The generally high level of formal help in health-related areas during the first year after a suicide attempt and the high rated adequacy of help given is satisfactory. However, in certain areas lack of formal help was evident. Informal caregivers contributed significantly to the help that patients received. The use of the Camberwell Assessment of Need instrument could improve observation on needs areas and whether help is available. Perhaps this kind of evaluation could be used earlier than after 1 year in such a vulnerable group as suicide attempters.     

Parasuicide: The value of suicidal intent and various motives as predictors of future suicidal behaviour

Abstract

The main aim of the study was to examine whether various aspects of suicidal intent or various motives for an index parasuicide can predict nonfatal or fatal repetition of suicidal behaviour. 776 parasuicide patients from 5 Nordic regions participating in the WHO/EURO Multicentre Study on Parasuicide were followed for one year. The Suicide Intent Scale (SIS) and the Motives for Parasuicide Questionnaire (MPQ) were used as predictor variables. A low level of suicidal intent predicted nonfatal repetition. The motive “make things easier for someone” differentiated female repeaters From male repeaters. The report of an unclear motive was the only significant predictor of fatal repetition. Suicidal intent and various motives for a parasuicide have some value in predicting repetition of suicidal behaviour. but should, perhaps. in future studies be combined with other variables in order to increase the predictive value.     

Parasuicide: The value of suicidal intent and various motives as predictors of future suicidal behaviour

The main aim of the study was to examine whether various aspects of suicidal intent or various motives for an index parasuicide can predict nonfatal or fatal repetition of suicidal behaviour. 776 parasuicide patients from 5 Nordic regions participating in the WHO/EURO Multicentre Study on Parasuicide were followed for one year. The Suicide Intent Scale (SIS) and the Motives for Parasuicide Questionnaire (MPQ) were used as predictor variables. A low level of suicidal intent predicted nonfatal repetition. The motive 'make things easier for someone' differentiated female repeaters from male repeaters. The report of an unclear motive was the only significant predictor of fatal repetition. Suicidal intent and various motives for a parasuicide have some value in predicting repetition of suicidal behaviour, but should, perhaps, in future studies be combined with other variables in order to increase the predictive value.     

Types of attempted suicide (parasuicide).

The clinical usefulness of a typology of parasuicide, reported in a previous paper, has been improved. A finer grained classification, using six instead of only three clusters, has proved itself to be clinically highly realistic. By a new method of cluster analysis which examines 80 attributes in each of 350 patients, the following types of parasuicide emerge: (1) operant, not alienated; (2) repeaters; (3) depressed with high life endangerment; (4) operant and alienated; (5) wristcutters; (6) undifferentiated. This more detailed typology may offer some advantages for clinical management and research.     

Repeated use of psychiatric emergency out-patient services among new patients: a 3-year follow-up study

The aim of this study was to examine repeated use of psychiatric emergency out-patient services during the second and third years after the first contact. A 1-year treated incidence cohort of 537 new patients was studied in the Department of Psychiatry in Oulu, Finland. Repeat users were defined as patients belonging to the upper 10th percentile of the emergency out-patient contacts. The median of emergency out-patient contacts among repeaters was 4. The repeaters constituted 8% of the cohort and they used 65% of the cohort's emergency contacts. They were more likely to be male and living alone, and they tended to have more serious diagnoses than non-repeaters. Having hospital admissions, planned out-patient contacts and repeated emergency out-patient contacts also during the first year of follow-up was associated with an increased probability of repeatedly using emergency services during the second and third years. Living alone and having hospital admissions during the follow-up period were associated with being a continuous repeat visitor during the whole follow-up period. It is concluded that the extended repeated use of emergency services is associated with inadequate social support and serious psychiatric problems.     

Repetition of deliberate self-harm by adolescents: the role of psychological factors.

The aim of this study was to examine the relationship between psychological variables and repetition of deliberate self-harm by adolescents (n=45) aged 13-18 years who had been admitted to a general hospital having taken overdoses. Standardized measures of depression, hopelessness, suicidal intent, impulsivity, trait and state anger, self-esteem and problem-solving (both self-report and observer-rated) were administered to the adolescents while still in the general hospital. Repetition was assessed on the basis of previous overdoses (n=14) and repetition of self-harm (self-poisoning and self-injury) during the subsequent year (n=9). Adolescents with a history of a previous overdose and/or who repeated self-harm during the following year (n=18) differed from non-repeaters in having higher scores for depression, hopelessness and trait anger, and lower scores for self-esteem, self-rated problem-solving and effectiveness of problem-solving rated on the basis of the Means End Problem Solving test, all measured at the initial assessment. These differences largely disappeared when level of depression was controlled for. Similarly, differences found between repeaters and non-repeaters in the year following the index overdoses for problem-solving were much reduced when account was taken of differences in depression scores.Depression is a key factor associated with risk of repetition of adolescent self-harm (and hence of suicide risk). In the management of adolescents who have harmed themselves, careful assessment of depression and appropriate management of those who are depressed is essential.     

Neuropsychological performance in psychiatric patients with or without documented brain dysfunction

The current study examined neuropsychological performance of schizophrenic and depressed patients with and without structural or EEG signs of brain dysfunctioning. The neuropsychological test battery was designed to sample intellectual functioning, psychomotor skills, nonverbal memory, and novel problem solving ability. Patients were classified into four groups: schizophrenics without signs of brain dysfunctioning, schizophrenics with signs of brain dysfunctioning, depressed without signs of brain dysfunctioning, and depressed with signs of brain dysfunctioning. The Trail Making Test--Parts A & B, the difference between these two components of the Trail Making Test, the World Fluency Test and a Laterality Index based on the age scale scores of the WAIS significantly discriminated between patients with and without brain dysfunction. Using these variables and a Linear Discriminant Function Analysis, we found that 84.5% of our subjects could be correctly classified. Substantially fewer patients could be correctly classified with respect to their psychiatric diagnosis (i.e., 66.7% correct classification). In fact, only the PIQ from the WAIS-R showed significant differences between the depressed and schizophrenic subjects. Poor neuropsychological performance was interpreted as showing more than the behavioral disorganization associated with psychiatric states and was felt to be related to the presence of objective signs of brain dysfunctioning.     

The Marke-Nyman Temperament (MNT) Scale in suicide attempters

A total of 89 suicide attempters were diagnosed according to the DSM-III-R, and categorized according to suicide method as ‘violent’ and ‘non-violent’ attempters, and also as those who had made previous suicide attempts (‘repeaters’) and those who had not (‘non-repeaters’). All of the subjects completed the MNT questionnaire. Our hypothesis was that the temperament patterns would reflect specific clinical characteristics, such as psychiatric diagnoses and type of suicidal behaviour. The ‘validity’ (ranging from fatiguableness to a high level of psychic energy) showed a bimodal distribution, suggesting the existence of two subgroups. The subjects with adjustment disorders displayed a significantly higher mean validity than the other patients. The lowest validity scores were found in subjects with bipolar depression. Men showed significantly higher 'stability’ (reflecting emotional distance from others) than women. Subjects with major depression and those who were ‘violent’ suicide attempters showed somewhat higher 'solidity’ (unchangeability, non-impulsiveness) compared with the other groups. Female ‘non-repeaters’ displayed significantly lower solidity than males or other females. In general, no clear temperament patterns were detected in patients categorized according to type of suicidal behaviour. In contrast to the findings of studies of normal populations, the intercorrelations between the dimensions were considerable, especially among the men. This prompted us to perform a cluster analysis procedure in which the subjects were allocated to five clusters according to the MNT dimensions. No distinct clinical characteristics appeared in these clusters, except in the cluster with the highest validity and the lowest stability, as 54% (P= 0.014) of the patients in this cluster had an adjustment disorder diagnosis. These temperament results reflect the clinical heterogeneity of patients who attempt suicide.     

Applicability of the CANTAB-PAL Computerized Memory Test in Identifying Amnestic Mild Cognitive Impairment and Alzheimer's Disease

Aim: To compare the diagnostic accuracy of a computerized test, the CANTAB paired associate learning (PAL) to that of an established and validated noncomputerized test, the CERAD Wordlist Learning task in differentiating between normal aging, aMCI and AD in a cross-sectional design. Methods: 58 participants were assessed (19 with mild probable AD, 17 aMCI, 22 healthy controls). Results: The variables found to best discriminate between the three groups were the CANTAB PAL total errors adjusted (p < 0.0001, 81.0% of the cases correctly classified), and CERAD Wordlist Learning Delayed Recall (p < 0.0001, 77.6% of the cases correctly classified). Using both PAL total errors adjusted and Wordlist Learning Delayed Recall, 84.5% of the cases were correctly classified. Discussion: The results suggest that the CANTAB could be used for screening of AD-typical memory impairment.     

Case-control study of unemployment and parasuicide

OBJECTIVE: The objective of the study was to estimate the odds ratio for the association between unemployment and parasuicide, controlling for potential confounding variables. METHODS: A case-control study was conducted involving 507 cases of parasuicide and 200 age- and sex-matched controls presenting to emergency departments in Edmonton, Canada. Subjects were interviewed using an extensive questionnaire developed for a World Health Organization study of parasuicide. The data were analyzed using conditional logistic regression. RESULTS: In a crude analysis, the odds ratio for the association between unemployment and parasuicide was 12.0 (95% confidence interval, 6.0-23.9). After controlling for sociodemographic factors, lifetime prevalence of several Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition, disorders, previous parasuicide, hopelessness, childhood experiences, and emotional support from family and friends, the odds ratio was 5.2 (95% confidence interval, 2.3-11.4). CONCLUSIONS: Those who are unemployed have an increased risk of parasuicide, and this association persists after accounting for a wide range of potential confounders.     

Parasuicide in a northern Swedish county 1989–1995 and its relation to suicide

The epidemiology of parasuicide and its relation to suicide during the years 1989–1995 was studied in a northern Swedish county. The total parasuicide rate was declining, especially among older women and among men within the 15–24 age group, and a decrease in the overall parasuicide/suicide ratio was found. Groups with the highest parasuicide rate revealed the lowest suicide rate, and reversed, groups with the highest suicide rate revealed the lowest parasuicide rate. Among the parasuicide population, 1.8% committed suicide within one year, and 3.3% within 3 years. Suicides among the monitored cases of parasuicide 1989–1995 constituted about 14.5% of the suicides in the area during 1989–1996, suggesting that even by successful treatment of current parasuiciders only a limited proportion of the suicides would be prevented. However, it is argued that parasuicide should be regarded as a public health problem in its own rights, since it is causing tremendous pain and suffering among the persons involved.     

Reliable discrimination of elderly depressed and demented patients by electroencephalographic sleep data

Using electroencephalographic sleep data from a sample of 235 elderly subjects, discriminant function analyses of sleep alterations in depression and dementia were performed. Overall, 80% of patients were correctly classified using a backward discriminant function analysis, and 81% with a general stepwise discriminant function analysis. Four measures contributed to the separation of depressed and demented patients: rapid eye movement (REM) sleep latency (lower in depressives); REM sleep percent (higher in depressives); indeterminate non-REM sleep percent (higher in demented patients, reflecting greater loss of spindles and K complexes); and early morning awakening (more marked in depressives). When both discriminant functions were subjected to cross-validation in independent subsamples, both procedures correctly identified 78% of patients. The classification functions derived from nondemented depressed and nondepressed demented patients were applied to a mixed-symptom group (n = 42). Overall, 27 patients (64%) with either depressive pseudodementia or dementia with depressive features were correctly classified using the same four predictor variables. These findings suggest that sleep physiological alterations of depression and dementia reflect between-group differences in sleep continuity, sleep architecture, and REM sleep temporal distribution, and that the differences are statistically reliable, in both diagnostically pure and mixed clinical presentations. These findings are discussed in the context of current hypotheses of sleep regulation and its mechanisms.     

Repetition of deliberate self-harm as seen in Uganda.

Described is the presentation of repetition of deliberate self-harm (DSH) as seen in an African population in urban Uganda. A Luganda version (local language in study area) of the modified European Parasuicide Interview Schedule I (EPSIS I) was used to collect the data. An univariate analysis was conducted. Results indicated that repeaters of DSH differed significantly from non-repeaters on several dimensions. The factors included were: more often single, less often had children, staying alone or with their parents, reported sexual problems as a precipitant of current DSH event, more negative life events in childhood and less negative life events in the last year. Furthermore, a multivariate analysis was conducted resulting in only sexual problems and the psychological factor of trait anger, which retained statistical significance. The significance of these findings is discussed.     

A review of the literature on the epidemiology of parasuicide in the general population

OBJECTIVE: The author reviewed recent literature on the epidemiology of parasuicide in the general population. Major risk factors are also discussed. METHODS: Parasuicide was defined as suicide attempts and deliberate self-harm inflicted with no intent to die. Articles in English on rates of parasuicide in the general population from 1970 through June 2000 were identified by keyword searches of the PsycINFO and MEDLINE databases. Studies that provided data from a representative population sample and provided incidence or prevalence rates were included in the review. Articles focusing on subsamples such as adolescents or psychiatric patients were excluded. Results and conclusions: Twenty studies were analyzed. Although methodological problems constitute a major limitation in interpretation of the results, the literature indicates that parasuicide is a serious public health problem. Reported annual rates of parasuicide in recent decades range from 2.6 to 1,100 per 100,000, and lifetime prevalence rates range from 720 to 5,930 per 100,000. The most important risk factors identified were younger age and female gender. Others included being single or divorced, being unemployed, having a recent change in living situation, having a mental disorder, and having a previous parasuicide incident.     

Suicide intent among parasuicide patients in Nicaragua: a surveillance and follow-up study.

This study examines suicide intent among parasuicide patients in a low-income country, Nicaragua, with special reference to gender patterns and future suicidal behavior. Using the Suicide Intent Scale (SIS), suicide intent was assessed in 204 persons presenting to hospital after parasuicide. Repetition was checked after a mean follow-up period of three years. The total SIS scores did not differ between women and men. However, a higher SIS score among women was significantly associated with older age, having children and use of pesticide as the parasuicide method. The overall method of suicide intent was low in Nicaragua compared to other countries, as was the nonfatal repetition rate (4.8% after three years). Subsequent suicides were found only in three men. Factor structures within the SIS disclosed supported the cross-cultural validity of the instrument. The level of suicide intent at the index attempt did not show any association with future suicidal behavior.     

Parasuicide in psychiatric inpatients: results of a controlled investigation

A group of 20 psychiatric inpatients who committed parasuicide during their hospital stay were compared with 58 randomly selected control inpatients and with 34 patients who were admitted following a parasuicide. The findings indicate: (1) The frequency of in-hospital parasuicide amounts to 3.2% of all admissions. (2) Apart from the more frequent past suicidal activity of the in-hospital parasuicide group, no really important differences were identified between these and control patients. The broad definition of parasuicide probably accounts for this result. (3) The patients who committed parasuicide before admission represent a diagnostically different, clinically less severely ill population. The differences between both groups of parasuicide patients support the notion of the heterogeneity of the parasuicide population.