- Email: [email protected]
Hospitalization in a pediatric ward of adolescent suicide attempters admitted to general hospitals
JOURNAL
OF ADOLESCENT
HEALTH
1994;15:416-422
INTERNATIONAL SECTION
Hospitalization in a Pediatric Ward of Adolescent Suicide Attempters Admitted to General ‘Hospitals ISABELLE GASQUET, M.D. AND MARIE CHOQUET, Ph.D.
firpose: Considering the growing interest of pediatricians in the adolescent patient, one miF.ht hypothesize that adolescent suicide attempters might receive superior treatment in a pediatric unit than in other in-patient settings. Methods: In order to examine this hypothesis, we compared sociodemographic and psychosocial characteristics of patients admitted to pediatric wards (n = 174) with those admitted to other departments (n = 2511, based on data from a survey of 430 adolescents aged 13-19 years hospitalized following a suicide attempt. Results: Patients admitted to pediatric units tended to be younger and were more often described as manifesting depression and sleep disturbances. However, no differences were found between the two groups for the other factors investigated (sex, socio-professional category of parents, academic level, family situation, psychiatric problems, social problems, etc.). We then examined differences between the two groups with respect to hospital care (length of stay, psychiatric consultations) and follow-up CO~?SC :coordinaiion with the family doctor and/or social services, recommended after-care). Our findings showed that for a similar psychosocial profile, pediatric units contact the family doctors more often and proposed a more extensive follow-up. The role in patient management of depression, family, social and school problems is also discussed. CS~?Jusion: Compared with other departments to which adolescent suicide attempters could be admitted, pediatric wards appear to show better performance.
From fhe National Institute of Health and Medical Research UNSERM),Villejuif, France Address reprintrequests to: lsabelleGasquet, M.D., INSERM U. 169, 16, avenue Paul Vaiilarrl Couturier,F-94807 VillejuifCedex
France. Manuscript accepted November 28, 2993. 416
KEY WORDS:
Attempted suicide Adolescence Hospital care Pediatrics Psychosocial factors Epidemiology
ktroduction Among the many epidemiological studies on adolescent suicide attempters (1,2,3,4), few evaluate ho:ipital care and follow-up course (5,6,7,8,9). While the)-? is consensus on the need for systematic hospitaliqation for at least 24 hours under the care of a multi-disciplinary team (lo,1 1,12,13,14,15) to allow psychiatric assessment (5,6,14), interview of relatives and friends (21, and social assessment (16,171, possible discrepancies between recommendations and institutional practices remain to be explored. Of the different hospital departments to which an adolescent suicide attempter may be admitted, pediatrics is often considered the most suitable environment. First, because teenagers who have attempted suicide firquently manifest a high degree of social dependence and emotional attachment to their parents (18). Second, blecause the decline in childhood diseases has led hospital pediatricians to increasingly devote attention to adolescent patients-particularly suicide attempters-who are frequently referred to them for treatment (19). Our study therefore sought to ascertain whether these patients did indeed receive superior treatment when placed in a pediatric unit. Based on the findings of a multi-hospital survey among French general hospitals, we attempted to identify the specific role played by the pediatric
0 Society for Adolescent Medicine, 1994 _. Published by Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010
July 1994
ward in the care of hospitalized adolescent suicide attempters, by: 1) comparing sociodemographic and psychosocial characteristics of patients admitted to pediatric wards and other departments; 2) examining certain aspects of the care provided, as well as the factors influencing that care (unit of hospitalization, patient’s psychosocial profile).
Materials and Methods Population All units of the 164 general hospitals located in 64 French dbpartements agreed to participate in a cross-sectional study involving the collection of data over a six-month period on all 13- to 1Pyear olds hospitalized at some point during that period. Only one record was used for each patient of a given hospital, regardless of the number of times that patient had been admitted. The contributing hospital departments included 80 emergency units, 47 pediatric, 63 internal medicine, 26 psychiatric, 155 surgical, and 168 other. Of 11,242 records collected between December 1988 and March 1990,10,125 (or 91.4%), were used. 81.6% (n = 8,264) of these indicated the reason for hospitalization, 5.4% (n = 430) being for attempted suicide. Methods Data were compiled from a two-part questionnaire comprising 61 closed preceded multiple choice questions: 1) the first part, completed on admission by the hospital admissions staff, concerned the patient’s family and school situation. 2) the second part, completed by the attending physician, included the reason(s) for hospitalization, hospital treatment received, recommended folloFvup course, if any, personal history, and any related family or personal problems. Total questionnaire completion time was approximately 30 minutes. Statistical Analysis The 430 hospitalized suicide attempters were divided into two groups: those admitted to a pediatric ward (PSA = 40%), and those hospitalized in other departments (ASA = 58%), including psychiatry. Five patients for whom no assignment was mentioned were not included in the analysis. Statistical Methods 1) Comparison between PSA and ASA groups was performed by chi-square tests (qualitative vari-
ADOLESCENT SUICIDE ATTEMPTERS
417
ables), and Student’s t test (quantitative variables). To avoid bias introduced by age differences, chisquare tests were adjusted for age (Mantel-Haenszel test). 2) Stepwise logistic regression was used to measure the relationship between the unit of referral and patient management. Included in the model as explicative variables were: depression (yes/no), family conflicts (yes/no), school problems (yes/no), social problems, e.g., parental unemr$oyment, poverty (yes/no), and the department ::o which the adolescent had been admitted (pc;:liatric ward: yes/no). Data were computerized at INSERWs calculation center (SAS) for chi-square and Studeiyt’s t test, and BMDP for logistic regression.
Results Description
of the Suicide Attempter Population
80.5% (n = 346) o$ Lhe suicide attc:r,pters were females, average age$..vas 16.5 years (SK,: 1.7 years), 93.7% were French, and 45.9% came ‘room broken homes (divorce or death of a parent). i“he average number of children per family was 2.7; 7.3% of the fathers were unemployed versus 6.3% of the mothers. 11.9% of the attempters’ fathers were of high socio-economic status (executives, pn~fessions), 39.8% were middle class (middle manai;ers, office workers, businessmen), 36.5% were bl,ue collar workers, and 2% farmers. 75.4% of the attempters were still at school, 14.9% were working or employed as trainees, and 9.7% were neithes employed nor at school. Average hospital stay was 5.2 days (SD = 10.3). Most youngsters were first admitted tc the emergency room (83.3%) and then transferred to an inpatient department (82.6%): 41.4% to a pediatric ward, the others (n = 251) to a variety of medical units-both general and specialized (e.g., hematology, nephrology, dermatology)-(27.5%), in-patient emergency wards (17.4%), and psychiatry (9.3%).
Comparative Hospitalized Departments
Profile of Suicide Attempters in a Pediatric Ward and in Other
So&demographic characteristics (‘To; k 2). The suicide attempters referred to a pediatric ward were generally the youngest pa&nts (under 16 years). Nevertheless, 47% exceeded the “legal” cutrrff age. For the other socio-demographic variables investigated (sex ratio, school attendance, parents’ profes-
418
JOURNAL OF ADOLESCENT HEALTH Vol. 15, No. 5
CASQUET AND CHOQUET
Table 1. Socio-demographic Attempters
by Department
Sex ratio (Girls/Boys) Age: 115 years Still at school Father employed Mother not working outside the home 2-parent household Living with father and/or mother
Characteristics of Hospitalization
Table 3. Hospital Care and Follow-up Course, by Unit of Hospitalization
of Suicide
PSA (1) (n = 1741, %
ASA (2) (n = 251), %
P (3)
4 47 88 78
5 a5 66 71
IS .OOl ns (3) ns (3)
35 57
32 52
ns (3) ns (3)
77
74
ns (3)
(Pediatrics: PSA; Others: ASA) (1) Hospitalized in pediatric ward; (2) Hospitalized in an adult unit; (3) After adjustment for two age groups (I5 and under; 16 and over).
sional status, family structure), no differences were found between PSA and ASA. Psychosocial characteristics (Table 2). Psychosocial problems among suicide attempters are frequently identified by hospital physicians in all types of settings. Of the total number of attempters studiedwhether in pediatrics or elsewhere-more than twothirds had family problems, nearly one-half had school problems, and one-third had psychiatric or social problems. For patients of similar age, depression (p I 0.01) and sleep disturbances (p 5 0.05) were reported more frequently by pediatric units than by other departments. This was not the case, however, for
psychiatric disorders and environmental (family, school, or social).
problems
Table 2. Psychosocial Characteristics of Suicide Attempters, by Unit of Hospitalization I=% (1) Physician report ok
(n = 174), %
ASA (2) (n = 2511, %
Family problems Social problems Depression Sleep disturbances Psychiatric problems School problems
71 28 63 35 26 56
68 30 53 30 31 40
(I 1Suicide attempters admitted (2) Suicide attempters admitted (3) After adjustment for 2 age older); (4) Significant difference before non-significant after adjustment.
to a pediatric ward; to an adult unit; groups (15 and under; adjustment
P (3)
.-
ns
--
.: .05 ns ns (4)
16 and
for age (p ~0.001).
Hospital
ASA (2) (n = 251)
P (3)
5 days 80
5 days 74
ns ns
59 3 20 87
47 5 17 63
.05 ns
we
Length of stay Psychiatric consultation Follow-up
I’-% (1) (n = 174)
course
Contact with family Contact with school Contact with social Proposed follow-up
doctor doctor services care
.Gl
(1) Suicide attempters admitted to a pediatric ward; (2) Suicide attempters admitted to an adult unit; (3)After adjustment for two age groups (age 15 and under; age 16 and c,ver).
Hospital and post-hospital care fTuMe 3). No differences were observed between the two groups with respect to duration of stay (about 5 days for both groups) or psychiatric consultations (3 out of 4 adolescents were seen by a psychiatrist during their hospital stay), but recommended follow-up did differ:
The family doctor was contacted for 59% of the PSA versus 47% of the ASA (p 5 0.05) group; After-care was proposed for 87% of the PSA, versus only 63% of the ASA (p I 0.001). However, no differences were observed with respect to contacting the school doctor or social services. Comparisorr of ‘psychiatrics” and “pediatrics” patients. As adolescent suicide attempters hospitalized
in a psychiatric unit form a specific sub-group within the category of non-pediatric patient, we sought to determine whether-despite a small sample size (n = 40)-differences could be found between “psychiatric” and “pediatric” patients. The only significant differences noted were age (“psychiatric” patients tended to be older: 17.9 years) and sex (more males than females: sex ratio F/M = 0.7). As for the other psychosocial variables, findings for the two groups were similar, as follows: depression (62% for patients in psychiatry vs. 63% in pediatrics), sleep disturbances (32% vs. 35%), social problems (31% vs. 28%), family problems (63% vs. 71%). After-care proposed appears similar in both departments (78% in psychiatry vs. 87% in pediatrics), as was coordination with social services (24% vs. 20%), although contacting the family doctor was less common in psychiatry than in pediatrics (30% vs. 59% in pediatrics p I 0.01).
July 1994
ADOLESCENT SUICIDE ATTEMPTERS
Discussiojq
Table 4. Contact with Family Doctor (Dependent Variable) itI Relation to the Type of Service and Psychosocial Problems (Independent Variable) B
Variables Pediatric ward: yes Social problems:
yes
0.4779 -0.494
B/SE 2.00 -2.15
419
OR=
r
1.61 (1.04-2.50) 0.61 (0.38-O 99)
.03 .05
B, beta coefficient; SE, standard error; OR, odds ratio. BEstimate from the multiple regression model. Adjustment according to the variables: depression, family problems, social problems, school problems, and unit of hospitalization (pediatrics vs. others). Results of Logistic Regression Analysis (Stepwise logistic regression: entry limit = 0.15; r~..~oval limit = 0.1). n = 349; Rreakdown: yes = 184, no = 165.
Factors Influencing Management of Suicide Attempters Since pediatric units contacted family doctors and proposed follow-up with greater frequency than other departments, and also reported more depression-related symptoms and sleep disturbances, it was important to determine whether the differences in after-care recommendations were symptomdriven or unit-driven, independently of symptoms. For this purpose, we used a logistic regression model, taking successively as dependent variables 1) decision lo contact the family doctor, and 2) recommendation of after-Cal-e. The independent variables concerned problems observed by the attending physician and the department to which the patient was referred (see “l~ethodology”). Our results showed a .&alionship between the department of hospitalization and follow-up course proposed, independently of the patient’s psychosocial profile, with pediatrj c departments tending more frequelsz.ly than others to contact the patient’s family doctor (OR = 1.6, p 5 0.05) (Table 41, and propose post-hospital follow-up care (OR = 3.3, p I .OOOl(Table 5). However, certain psychosocial factors did affect patient management, independently of the department of referral or of other variables introduced into the logistic regression. . The decision to contact the family doctor was less frequent in cases where social problems were detected (OR = 1.6, p I 0.5) (Table 41,‘but it was not influenced by depression-related symptoms or by school and family problems. e Follow-up care was more frequently prescribed for patients with depression-related symptoms (OR = 4.0, p I .OOOl) and family problems (OR = 3.3, p 5 .Ol) (Table 5) but vias not influenced by school or social problems.
This was the first multi-hospital study in France to investigate the care provided to all hospitalized adolescents, without regard to the unit in which they were placed or the reasons for hospitalization. AS it was based on voluntary participation of the hospitals, validation of the sample of suicide attempters was perfcnrmed (20). Suicide attempters hospitali,zed in a pediatric ward were compared to those admitted to other departments. Included in the latter heterogeneous populatison were referrals to psychiatric wards, who were probably severely disturbed at the time of their suicide attempt (6). Although this statistical grouping may have minimized the differences observed between the two groups, it should be emphasized that “psychiatric” patients only accounted for 15.9% (40 patients) of the ASA group. Nevertheless, but? au the small sample size, it was not possible to identify the specific role of psychiatry or assess its performance in patient evaluation or recommended after-care. The sociodemographic profile of the suicide attempterr, in our study resembled that found elsewhere ir:’ the literature (211, i.e., a larger percentage of females, foreigners, school dropouts and broken homes than in the French adolescent population in general. Our results suggest that hospital pediatric units play ,:‘n important role in the management of adolesceri;: suicide attempters. In France, nearly half of the a&lescents hospitalized following a suicide attempt
Table 5. Post-hospital Follow-up (Dependent Variable) in Relation to Unit of Hospitalization and Type of Psychosocial Problems (Independent Variable) -Variables Depression: yes Pediatric ward: yes Familv uroblems: ves
6 1.079 1.315 0.9072
B/SE
ORa
P
4.04 4.32 3.34
2.94 (1.74-4.98) 3.73 (2.05-6.78) 2.48 (1.454.23)
.OOOl .OOOl .OOl
B, brta coefficient; SE, standard error; OR, odds ratio. aEstix~dte from tlqe multiple regression model. Adjustment according b the variables depression, family problems, social problems, schools problems, and unit of hospitalization (pediatrics vs. others). n = 347 breakdown: yes = 255, no = 92. Results of Logistic Regression Analysis (Stepwise logistic regression: entry limit = 0.15; removal limit = 0.1).
420
GASQUET AND CHOQUET
structure (see “Appendix”) but also appears to be the trend in other countries (6,10,14,15,25)Several authors consider a pediatric unit to be more appropriate than other departments for fragile adolescents who have attempted suicide. First of a& because it is more likely to provide a nurturing environment-f major importance for patients just leaving childhood (12,16,26). Secondly, because of its mode of functioning-integration of parents in care, multi-discipline healthcare teams, greater focus on social aspects-and ability to address these patients’ medical and psychosocial needs. Finally, because it offers a more global approach, better adapted to adolescent pathology than other departments (22,27). As concerns the assignment of adolescent suicide attempters to non-pediatric units, while this may have been governed by the patient’s primary condition (e.g., patients with kidney failure admitted to nephrology) it may also simply have been the result of bed availability at the time. Although attempters hospitalized in a pediatric ward were generally younger than those referred to other departments, almost 50% were 16 years or older. This is of interest in that French law does not require pediatric, departments to admit adolescents over 15 years, except in the rare event of previous admission of a youngster to the ward. A partial explanation for this orientation may be the current interest of pediatricians in this particular patient group. The upper age limit in admission to a pediatric ward varies considerably from one country to another: it is 21 years in the United States compared with 15 years in the UK, for example. The data suggest that pediatric units identify depression-related symptoms and sleep disorders in adolescent suicide attempters more often than other departments. The literature estimates that the proportion of adolescent suicide attempters with depression-related symptoms is between 30 and 50% in psychiatric wards (25,28), 60 to 80% in emergency departments (29,30), and less than 30% in general medicine (31). Our finding of 63% in pediatric wards appears rather high compared to what has been reported in other studies conducted in general hospitals. This may be linked to the particular emphasis placed on adolescent behavior in the latter setting. Epidemiological studies have shown that depression is a risk factor not only for initial suicide attempts WW, but also for recurrence (6,34) and completion (35). A diagnosis of depression should therefore result in the provision for after-care, inde-
JOURNAL OF ADOLESCENT HEALTH Vol. 15, No. 5
pendently of the department of hospitalization or of any other symptomatology. A study of depressed adolescents in a pediatric ward (5) showed an enhanced frequency of psychiatric consultations during hospitalization and of subsequent referral to a psychiatric unit. Family conflicts are often associated with adolescent suicide attempts (25,36), a dysfunctional family being a risk factor for both suicidal gestures (30) and recurrent attempts (34,37), as well as for poor compliance with follow-up recommendations (6). It is therefore important to allow for the family situation when organizing care. Our findings suggest that this is currently the trend. Social problems are also considered as risk factors for suicide attempts, because of the resulting personal isolation and the close association observed between problems of social adjustment and depression (38). Whereas social problems are only found in a minority of suicide attempters (15% of cases for Tischler-1981, 25% for Hawton-1982b, and 29% in our study), our results showed that they decrease the probability of coordination with the family doctor. Several explanations for this situation are possible: either adolescents with social problems are less apt to have a family doctor than other youngsters, or hospital physicians give a higher priority to the social aspect over the strictly medical one. In addition to psychosocial factors, considered as major (4) and ih.erefore worthy of consideration when determining the type of care to be provided, our results suggest that the hospital department itself plays an important role in the management of care. Pediatric wards establish more frequent contact with the family doctors than adult units-a procedure that should be instituted routinely (ll)-and usually recommend more extensive followup. Although there is general consensus that short term after-care is a necessary minimum, because of the high risk of recurrence during the three months following a suicide attempt (14,39), this is not always put into practice. From the information available, which needs to be complemented by further studies, we can conclude that on the whole pediatric units in French hospitals offer satisfactory care to adolescents who have attempted suicide. The duration of hospitalization is consistent with current recommendations, a psychiatrist is generally consulted, the family doctor frequently contacted and follow-up care organized. Compared with other departments to which adolescent suicide attempters could be admitted, pediatric wards appear to per-
July 1994
ADOLESCENT SUICIDE ATTEMPTERS
form better. Nevertheless, further investigation is required, and certain aspects of patient management in this particular setting need to be studied in greater detail.
cent suicide attempters hospitalized on a general floor. J Adolesc Health Care 1988;9:491-4.
421
pediatrics
9. Gasquet I, Choquet M. Gender role in adolescent suicidal behavior: observations and therapeutic implications. Acta Psychiatr Stand 1993;87:59-65. 10. Marks A. Management of the suicidal adolescent on a nonpsychiatric adolescent unit. J Pediatrics 1979;95:305-8.
Appendix The French public hospital system consists mainly of two types of structure. The first is the general hospital-geared for short stays, i.e. <3 months, which includes an emergency room, where suicide attempters are initially received. If hospitalization is required, the patient is then transferred to an inpatient unit: medicine (general or specialized), surgery, or psychiatry. Alternatively, the attempter may be transferred to a second type of facility-the specialized institution-such as a psychiatric hospital or rehabilitation center, geared for longer stays (< 3 years). French law requires pediatric departments in general hospitals to receive suicide attempters below the age of 16 years, provided that beds are available. The physician in charge may, however, admit older adolescents but is under no obligation to do so. Public psychiatric facilities are found both in general hospitals and specialized institutions. Unlike psychiatric wards in general hospitals, the latter are sectors-orientated, with patients being admitted according to their place of residence. Private clinics are mainly geared to surgery and do not have emergency facilities.
References 1. Shafer D, Fischer P. The epidemiology of suicide in children and young adolescents. ] Am Acad Child Psychiatr 1981;20:545-65.
11. Hawton K. Annotation, attempted suicide in children adolescents. J Child Psycho1 Psychiatr 1982;23:497-503.
and
12. Canoui P, Basquin M. Prise en charge des tentatives de SUItide de l’adolescent de 13 a 16 ans. Entretiens de Bichats Psychiatric 1985227-231. 13. Blum R, Smith M. Training of health professionals in adolescent health care. J Adolesc Health Care 1988;9:46-50. 14. American Academy of Pediatrics, Committee on Adolescence. Suicide and suicide attempts in adolescents and young adults. Pediatrics 1988;81:322-4. 15. Jay MS, Graham CJ, Flowers C, et al. Adolescent suicide attempters presenting to a pediatric facility. Adolescence 1989;24:467-72. 16. Hawton K, Gath D, Smith E. Management of attempted tide in Oxford. Br J Psychiatr 1979;2:1040-2.
SUI-
17. Deykin EY, Hsieh CC, Joshi N, et al. Adolescent suicidal and self-destructive behavior: results of an intervention study. J Adolesc Health Care 1986;7:88-95. 18. Rutter M, Graham P, Chadwick OFD, et al. Adolescent turmoil: fact or fiction? J Child Psycho1 Psychiatr 1976;17:35-56. 19. Kozak LJ, Norton C, McManus M, et al. Hospital use patterns for children in the United States, 1983 and 1984. Pediatrics 1987;80:481-90. 20. Gasquet I, Choquet M, Lellouch J. Problemes poses par la prise en charge medico-psychologiqoe des adolescents hospitalises 1 la suite d’une tentative de suicide, oersoectives euidemiologiques. Memoire dc DEA 1991, LJni\;ersi& Paris Vii. 21. Davidson F, Philippe A. Suicide et tentative de suicide aujourd’hui, etude epidemiologique. 1986 Paris; ESF. 22. Marquis F, Deschamps JP, Legras B, et al. L’hospitalisafion des adolescents dans un centre hospitalier regional. Arch Fr Pediatr 1985;42:459-64. 23. Herve C, Gaillard M, Vinllard M, et al. L’hospitalisation des adolescents dans le departemcnt du Val-de-Marne. Ann Pediatr 1988;35:171-8. 24. Hodgman CH, Roberts EN. Adolescent atrician. J Pediatrics 1982;101:118-123.
suicide and the pedi-
25. Khan AU. Heterogeneity of suicidal adolescents. Child Adolcsc Psychiatr 1987i2692-6.
J Am Acad
2. Hawton K, O’Grady j, Osborn M, et al. Adolescents who take overdoses: their characteristics, problems, and contacts with helping agencies. Br J Psychiatr 1982a;l40:118-23.
26. Combes JC. L’adolescent
3. Gispert M, Wheeler K, Marsh L, et al. Suicidal adolescents: factors in evaluation. Adolescence 1985;20:753-61.
27. Levy-Leblord E, Amoux D, Cau D, et al. Une medecine pour adolescents: Qui soigne quoi? Arch i+ Pediatr 1982;39:771-6.
4. Spirit0 A, Brown L, Overholser J, et al. Attempted suicide in adolescence: a review and critique of the literature. Clin Psycho1 Rev 1989;9:335-63.
28. Carlson GA, Cantwell DP. Suicidal behavior and depression in children and adolescents. J Am Acad Child Psychiat I982;21:361-8.
5. Schreiber TJ, Johnson RL. The evaluation and treatment of adolescent overdoses in an adolescent medical service. J Nat1 Med Ass 1986;78:101-8.
29. Garfinkel BD, Froese A, Hood J. Suicide attempts in children and adolescents. Am J Psychiat 1982;139:1257-61.
6. Spirit0 A, Stark LJ, Fristad M, et al. Adolescent suicide attempters hospitalised on a general pediatrics floor. J Pediatric Psycho1 1987;12:171-89. 7. Kotila L, Lonnqvist ferences predicting 264-70.
J. Adolescent suicide attempts: suicide. Acta Psychiatr Stand
8. Spirit0 A, Stark LJ, Hart KJ, et al. Overt behavior
sex dif1988;77: of adoies-
suicidant.
Rev Ped 1988;24:294-300.
30. Christoffel KK, Marcus D, Sagerman S, et al. Adolescent suicide and suicide attempts: a pppulation study. Pediatric Emergency 1988;4:3240. 31. Taylor EA, Stansfeld SA. Children who poison themselves 11. Predictian of attendance for treatment. Br ] Psychiat 1?84b;145:132-5. 32. Garrison CZ, Jackson KL, Addy CL, et al. Suicidal behaviors in young adolescents, Am ) Epidemiol 1991;133:101I!~-14.
422
GASQUET AND CHOQUET
JOURNAL OF ADOLESCENT HEALTH Vol. 15, No. 5
33. Andrews JA. Suicidal attempts among older adolescents: prevalence and cooccurrence with psychiatric disorders. J Am Acad ChiId Ado&c Psychiat 1992;31:655-62. 34. Choquet M, Facy F, Dadidson F, et al. Suicide and attempted suicide among adolescents in France. In: R Fanner, S Hirsch, eds. The suicide syndrome. London: Croom Helm, 198lI73-89. 35. Brent DA, Perper JA, Allman CJ. Correlates of the medical lethality of suicide attempts in children and adolescents. J Am Acad Child Adolesc Psychiat 1987;26:87-91. 36. Tischler CL, McHenry PC. Adolescent
suicide attempts;
some
significant
factors. Suicide Life-Threaten
Behav 1981;11:86-92.
37. Hawton K, O’Grady J, Osborn M, et al. Classification of adolescents who take overdoses. Br J Psychiat 1982b;140: 124-31. 38. Spirit0 A, Hart K, Overholser J, et al. Social skills and depression in adolescent suicide attempters. Adolescence 1990; X543-52. 39. Goldacre M, Hawton K. Repetition of self-poisoning subsequent death in adolescents who take overdoses. Psychiat 1985;1&395-8.
and Br J
OF ADOLESCENT
HEALTH
1994;15:416-422
INTERNATIONAL SECTION
Hospitalization in a Pediatric Ward of Adolescent Suicide Attempters Admitted to General ‘Hospitals ISABELLE GASQUET, M.D. AND MARIE CHOQUET, Ph.D.
firpose: Considering the growing interest of pediatricians in the adolescent patient, one miF.ht hypothesize that adolescent suicide attempters might receive superior treatment in a pediatric unit than in other in-patient settings. Methods: In order to examine this hypothesis, we compared sociodemographic and psychosocial characteristics of patients admitted to pediatric wards (n = 174) with those admitted to other departments (n = 2511, based on data from a survey of 430 adolescents aged 13-19 years hospitalized following a suicide attempt. Results: Patients admitted to pediatric units tended to be younger and were more often described as manifesting depression and sleep disturbances. However, no differences were found between the two groups for the other factors investigated (sex, socio-professional category of parents, academic level, family situation, psychiatric problems, social problems, etc.). We then examined differences between the two groups with respect to hospital care (length of stay, psychiatric consultations) and follow-up CO~?SC :coordinaiion with the family doctor and/or social services, recommended after-care). Our findings showed that for a similar psychosocial profile, pediatric units contact the family doctors more often and proposed a more extensive follow-up. The role in patient management of depression, family, social and school problems is also discussed. CS~?Jusion: Compared with other departments to which adolescent suicide attempters could be admitted, pediatric wards appear to show better performance.
From fhe National Institute of Health and Medical Research UNSERM),Villejuif, France Address reprintrequests to: lsabelleGasquet, M.D., INSERM U. 169, 16, avenue Paul Vaiilarrl Couturier,F-94807 VillejuifCedex
France. Manuscript accepted November 28, 2993. 416
KEY WORDS:
Attempted suicide Adolescence Hospital care Pediatrics Psychosocial factors Epidemiology
ktroduction Among the many epidemiological studies on adolescent suicide attempters (1,2,3,4), few evaluate ho:ipital care and follow-up course (5,6,7,8,9). While the)-? is consensus on the need for systematic hospitaliqation for at least 24 hours under the care of a multi-disciplinary team (lo,1 1,12,13,14,15) to allow psychiatric assessment (5,6,14), interview of relatives and friends (21, and social assessment (16,171, possible discrepancies between recommendations and institutional practices remain to be explored. Of the different hospital departments to which an adolescent suicide attempter may be admitted, pediatrics is often considered the most suitable environment. First, because teenagers who have attempted suicide firquently manifest a high degree of social dependence and emotional attachment to their parents (18). Second, blecause the decline in childhood diseases has led hospital pediatricians to increasingly devote attention to adolescent patients-particularly suicide attempters-who are frequently referred to them for treatment (19). Our study therefore sought to ascertain whether these patients did indeed receive superior treatment when placed in a pediatric unit. Based on the findings of a multi-hospital survey among French general hospitals, we attempted to identify the specific role played by the pediatric
0 Society for Adolescent Medicine, 1994 _. Published by Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010
July 1994
ward in the care of hospitalized adolescent suicide attempters, by: 1) comparing sociodemographic and psychosocial characteristics of patients admitted to pediatric wards and other departments; 2) examining certain aspects of the care provided, as well as the factors influencing that care (unit of hospitalization, patient’s psychosocial profile).
Materials and Methods Population All units of the 164 general hospitals located in 64 French dbpartements agreed to participate in a cross-sectional study involving the collection of data over a six-month period on all 13- to 1Pyear olds hospitalized at some point during that period. Only one record was used for each patient of a given hospital, regardless of the number of times that patient had been admitted. The contributing hospital departments included 80 emergency units, 47 pediatric, 63 internal medicine, 26 psychiatric, 155 surgical, and 168 other. Of 11,242 records collected between December 1988 and March 1990,10,125 (or 91.4%), were used. 81.6% (n = 8,264) of these indicated the reason for hospitalization, 5.4% (n = 430) being for attempted suicide. Methods Data were compiled from a two-part questionnaire comprising 61 closed preceded multiple choice questions: 1) the first part, completed on admission by the hospital admissions staff, concerned the patient’s family and school situation. 2) the second part, completed by the attending physician, included the reason(s) for hospitalization, hospital treatment received, recommended folloFvup course, if any, personal history, and any related family or personal problems. Total questionnaire completion time was approximately 30 minutes. Statistical Analysis The 430 hospitalized suicide attempters were divided into two groups: those admitted to a pediatric ward (PSA = 40%), and those hospitalized in other departments (ASA = 58%), including psychiatry. Five patients for whom no assignment was mentioned were not included in the analysis. Statistical Methods 1) Comparison between PSA and ASA groups was performed by chi-square tests (qualitative vari-
ADOLESCENT SUICIDE ATTEMPTERS
417
ables), and Student’s t test (quantitative variables). To avoid bias introduced by age differences, chisquare tests were adjusted for age (Mantel-Haenszel test). 2) Stepwise logistic regression was used to measure the relationship between the unit of referral and patient management. Included in the model as explicative variables were: depression (yes/no), family conflicts (yes/no), school problems (yes/no), social problems, e.g., parental unemr$oyment, poverty (yes/no), and the department ::o which the adolescent had been admitted (pc;:liatric ward: yes/no). Data were computerized at INSERWs calculation center (SAS) for chi-square and Studeiyt’s t test, and BMDP for logistic regression.
Results Description
of the Suicide Attempter Population
80.5% (n = 346) o$ Lhe suicide attc:r,pters were females, average age$..vas 16.5 years (SK,: 1.7 years), 93.7% were French, and 45.9% came ‘room broken homes (divorce or death of a parent). i“he average number of children per family was 2.7; 7.3% of the fathers were unemployed versus 6.3% of the mothers. 11.9% of the attempters’ fathers were of high socio-economic status (executives, pn~fessions), 39.8% were middle class (middle manai;ers, office workers, businessmen), 36.5% were bl,ue collar workers, and 2% farmers. 75.4% of the attempters were still at school, 14.9% were working or employed as trainees, and 9.7% were neithes employed nor at school. Average hospital stay was 5.2 days (SD = 10.3). Most youngsters were first admitted tc the emergency room (83.3%) and then transferred to an inpatient department (82.6%): 41.4% to a pediatric ward, the others (n = 251) to a variety of medical units-both general and specialized (e.g., hematology, nephrology, dermatology)-(27.5%), in-patient emergency wards (17.4%), and psychiatry (9.3%).
Comparative Hospitalized Departments
Profile of Suicide Attempters in a Pediatric Ward and in Other
So&demographic characteristics (‘To; k 2). The suicide attempters referred to a pediatric ward were generally the youngest pa&nts (under 16 years). Nevertheless, 47% exceeded the “legal” cutrrff age. For the other socio-demographic variables investigated (sex ratio, school attendance, parents’ profes-
418
JOURNAL OF ADOLESCENT HEALTH Vol. 15, No. 5
CASQUET AND CHOQUET
Table 1. Socio-demographic Attempters
by Department
Sex ratio (Girls/Boys) Age: 115 years Still at school Father employed Mother not working outside the home 2-parent household Living with father and/or mother
Characteristics of Hospitalization
Table 3. Hospital Care and Follow-up Course, by Unit of Hospitalization
of Suicide
PSA (1) (n = 1741, %
ASA (2) (n = 251), %
P (3)
4 47 88 78
5 a5 66 71
IS .OOl ns (3) ns (3)
35 57
32 52
ns (3) ns (3)
77
74
ns (3)
(Pediatrics: PSA; Others: ASA) (1) Hospitalized in pediatric ward; (2) Hospitalized in an adult unit; (3) After adjustment for two age groups (I5 and under; 16 and over).
sional status, family structure), no differences were found between PSA and ASA. Psychosocial characteristics (Table 2). Psychosocial problems among suicide attempters are frequently identified by hospital physicians in all types of settings. Of the total number of attempters studiedwhether in pediatrics or elsewhere-more than twothirds had family problems, nearly one-half had school problems, and one-third had psychiatric or social problems. For patients of similar age, depression (p I 0.01) and sleep disturbances (p 5 0.05) were reported more frequently by pediatric units than by other departments. This was not the case, however, for
psychiatric disorders and environmental (family, school, or social).
problems
Table 2. Psychosocial Characteristics of Suicide Attempters, by Unit of Hospitalization I=% (1) Physician report ok
(n = 174), %
ASA (2) (n = 2511, %
Family problems Social problems Depression Sleep disturbances Psychiatric problems School problems
71 28 63 35 26 56
68 30 53 30 31 40
(I 1Suicide attempters admitted (2) Suicide attempters admitted (3) After adjustment for 2 age older); (4) Significant difference before non-significant after adjustment.
to a pediatric ward; to an adult unit; groups (15 and under; adjustment
P (3)
.-
ns
--
.: .05 ns ns (4)
16 and
for age (p ~0.001).
Hospital
ASA (2) (n = 251)
P (3)
5 days 80
5 days 74
ns ns
59 3 20 87
47 5 17 63
.05 ns
we
Length of stay Psychiatric consultation Follow-up
I’-% (1) (n = 174)
course
Contact with family Contact with school Contact with social Proposed follow-up
doctor doctor services care
.Gl
(1) Suicide attempters admitted to a pediatric ward; (2) Suicide attempters admitted to an adult unit; (3)After adjustment for two age groups (age 15 and under; age 16 and c,ver).
Hospital and post-hospital care fTuMe 3). No differences were observed between the two groups with respect to duration of stay (about 5 days for both groups) or psychiatric consultations (3 out of 4 adolescents were seen by a psychiatrist during their hospital stay), but recommended follow-up did differ:
The family doctor was contacted for 59% of the PSA versus 47% of the ASA (p 5 0.05) group; After-care was proposed for 87% of the PSA, versus only 63% of the ASA (p I 0.001). However, no differences were observed with respect to contacting the school doctor or social services. Comparisorr of ‘psychiatrics” and “pediatrics” patients. As adolescent suicide attempters hospitalized
in a psychiatric unit form a specific sub-group within the category of non-pediatric patient, we sought to determine whether-despite a small sample size (n = 40)-differences could be found between “psychiatric” and “pediatric” patients. The only significant differences noted were age (“psychiatric” patients tended to be older: 17.9 years) and sex (more males than females: sex ratio F/M = 0.7). As for the other psychosocial variables, findings for the two groups were similar, as follows: depression (62% for patients in psychiatry vs. 63% in pediatrics), sleep disturbances (32% vs. 35%), social problems (31% vs. 28%), family problems (63% vs. 71%). After-care proposed appears similar in both departments (78% in psychiatry vs. 87% in pediatrics), as was coordination with social services (24% vs. 20%), although contacting the family doctor was less common in psychiatry than in pediatrics (30% vs. 59% in pediatrics p I 0.01).
July 1994
ADOLESCENT SUICIDE ATTEMPTERS
Discussiojq
Table 4. Contact with Family Doctor (Dependent Variable) itI Relation to the Type of Service and Psychosocial Problems (Independent Variable) B
Variables Pediatric ward: yes Social problems:
yes
0.4779 -0.494
B/SE 2.00 -2.15
419
OR=
r
1.61 (1.04-2.50) 0.61 (0.38-O 99)
.03 .05
B, beta coefficient; SE, standard error; OR, odds ratio. BEstimate from the multiple regression model. Adjustment according to the variables: depression, family problems, social problems, school problems, and unit of hospitalization (pediatrics vs. others). Results of Logistic Regression Analysis (Stepwise logistic regression: entry limit = 0.15; r~..~oval limit = 0.1). n = 349; Rreakdown: yes = 184, no = 165.
Factors Influencing Management of Suicide Attempters Since pediatric units contacted family doctors and proposed follow-up with greater frequency than other departments, and also reported more depression-related symptoms and sleep disturbances, it was important to determine whether the differences in after-care recommendations were symptomdriven or unit-driven, independently of symptoms. For this purpose, we used a logistic regression model, taking successively as dependent variables 1) decision lo contact the family doctor, and 2) recommendation of after-Cal-e. The independent variables concerned problems observed by the attending physician and the department to which the patient was referred (see “l~ethodology”). Our results showed a .&alionship between the department of hospitalization and follow-up course proposed, independently of the patient’s psychosocial profile, with pediatrj c departments tending more frequelsz.ly than others to contact the patient’s family doctor (OR = 1.6, p 5 0.05) (Table 41, and propose post-hospital follow-up care (OR = 3.3, p I .OOOl(Table 5). However, certain psychosocial factors did affect patient management, independently of the department of referral or of other variables introduced into the logistic regression. . The decision to contact the family doctor was less frequent in cases where social problems were detected (OR = 1.6, p I 0.5) (Table 41,‘but it was not influenced by depression-related symptoms or by school and family problems. e Follow-up care was more frequently prescribed for patients with depression-related symptoms (OR = 4.0, p I .OOOl) and family problems (OR = 3.3, p 5 .Ol) (Table 5) but vias not influenced by school or social problems.
This was the first multi-hospital study in France to investigate the care provided to all hospitalized adolescents, without regard to the unit in which they were placed or the reasons for hospitalization. AS it was based on voluntary participation of the hospitals, validation of the sample of suicide attempters was perfcnrmed (20). Suicide attempters hospitali,zed in a pediatric ward were compared to those admitted to other departments. Included in the latter heterogeneous populatison were referrals to psychiatric wards, who were probably severely disturbed at the time of their suicide attempt (6). Although this statistical grouping may have minimized the differences observed between the two groups, it should be emphasized that “psychiatric” patients only accounted for 15.9% (40 patients) of the ASA group. Nevertheless, but? au the small sample size, it was not possible to identify the specific role of psychiatry or assess its performance in patient evaluation or recommended after-care. The sociodemographic profile of the suicide attempterr, in our study resembled that found elsewhere ir:’ the literature (211, i.e., a larger percentage of females, foreigners, school dropouts and broken homes than in the French adolescent population in general. Our results suggest that hospital pediatric units play ,:‘n important role in the management of adolesceri;: suicide attempters. In France, nearly half of the a&lescents hospitalized following a suicide attempt
Table 5. Post-hospital Follow-up (Dependent Variable) in Relation to Unit of Hospitalization and Type of Psychosocial Problems (Independent Variable) -Variables Depression: yes Pediatric ward: yes Familv uroblems: ves
6 1.079 1.315 0.9072
B/SE
ORa
P
4.04 4.32 3.34
2.94 (1.74-4.98) 3.73 (2.05-6.78) 2.48 (1.454.23)
.OOOl .OOOl .OOl
B, brta coefficient; SE, standard error; OR, odds ratio. aEstix~dte from tlqe multiple regression model. Adjustment according b the variables depression, family problems, social problems, schools problems, and unit of hospitalization (pediatrics vs. others). n = 347 breakdown: yes = 255, no = 92. Results of Logistic Regression Analysis (Stepwise logistic regression: entry limit = 0.15; removal limit = 0.1).
420
GASQUET AND CHOQUET
structure (see “Appendix”) but also appears to be the trend in other countries (6,10,14,15,25)Several authors consider a pediatric unit to be more appropriate than other departments for fragile adolescents who have attempted suicide. First of a& because it is more likely to provide a nurturing environment-f major importance for patients just leaving childhood (12,16,26). Secondly, because of its mode of functioning-integration of parents in care, multi-discipline healthcare teams, greater focus on social aspects-and ability to address these patients’ medical and psychosocial needs. Finally, because it offers a more global approach, better adapted to adolescent pathology than other departments (22,27). As concerns the assignment of adolescent suicide attempters to non-pediatric units, while this may have been governed by the patient’s primary condition (e.g., patients with kidney failure admitted to nephrology) it may also simply have been the result of bed availability at the time. Although attempters hospitalized in a pediatric ward were generally younger than those referred to other departments, almost 50% were 16 years or older. This is of interest in that French law does not require pediatric, departments to admit adolescents over 15 years, except in the rare event of previous admission of a youngster to the ward. A partial explanation for this orientation may be the current interest of pediatricians in this particular patient group. The upper age limit in admission to a pediatric ward varies considerably from one country to another: it is 21 years in the United States compared with 15 years in the UK, for example. The data suggest that pediatric units identify depression-related symptoms and sleep disorders in adolescent suicide attempters more often than other departments. The literature estimates that the proportion of adolescent suicide attempters with depression-related symptoms is between 30 and 50% in psychiatric wards (25,28), 60 to 80% in emergency departments (29,30), and less than 30% in general medicine (31). Our finding of 63% in pediatric wards appears rather high compared to what has been reported in other studies conducted in general hospitals. This may be linked to the particular emphasis placed on adolescent behavior in the latter setting. Epidemiological studies have shown that depression is a risk factor not only for initial suicide attempts WW, but also for recurrence (6,34) and completion (35). A diagnosis of depression should therefore result in the provision for after-care, inde-
JOURNAL OF ADOLESCENT HEALTH Vol. 15, No. 5
pendently of the department of hospitalization or of any other symptomatology. A study of depressed adolescents in a pediatric ward (5) showed an enhanced frequency of psychiatric consultations during hospitalization and of subsequent referral to a psychiatric unit. Family conflicts are often associated with adolescent suicide attempts (25,36), a dysfunctional family being a risk factor for both suicidal gestures (30) and recurrent attempts (34,37), as well as for poor compliance with follow-up recommendations (6). It is therefore important to allow for the family situation when organizing care. Our findings suggest that this is currently the trend. Social problems are also considered as risk factors for suicide attempts, because of the resulting personal isolation and the close association observed between problems of social adjustment and depression (38). Whereas social problems are only found in a minority of suicide attempters (15% of cases for Tischler-1981, 25% for Hawton-1982b, and 29% in our study), our results showed that they decrease the probability of coordination with the family doctor. Several explanations for this situation are possible: either adolescents with social problems are less apt to have a family doctor than other youngsters, or hospital physicians give a higher priority to the social aspect over the strictly medical one. In addition to psychosocial factors, considered as major (4) and ih.erefore worthy of consideration when determining the type of care to be provided, our results suggest that the hospital department itself plays an important role in the management of care. Pediatric wards establish more frequent contact with the family doctors than adult units-a procedure that should be instituted routinely (ll)-and usually recommend more extensive followup. Although there is general consensus that short term after-care is a necessary minimum, because of the high risk of recurrence during the three months following a suicide attempt (14,39), this is not always put into practice. From the information available, which needs to be complemented by further studies, we can conclude that on the whole pediatric units in French hospitals offer satisfactory care to adolescents who have attempted suicide. The duration of hospitalization is consistent with current recommendations, a psychiatrist is generally consulted, the family doctor frequently contacted and follow-up care organized. Compared with other departments to which adolescent suicide attempters could be admitted, pediatric wards appear to per-
July 1994
ADOLESCENT SUICIDE ATTEMPTERS
form better. Nevertheless, further investigation is required, and certain aspects of patient management in this particular setting need to be studied in greater detail.
cent suicide attempters hospitalized on a general floor. J Adolesc Health Care 1988;9:491-4.
421
pediatrics
9. Gasquet I, Choquet M. Gender role in adolescent suicidal behavior: observations and therapeutic implications. Acta Psychiatr Stand 1993;87:59-65. 10. Marks A. Management of the suicidal adolescent on a nonpsychiatric adolescent unit. J Pediatrics 1979;95:305-8.
Appendix The French public hospital system consists mainly of two types of structure. The first is the general hospital-geared for short stays, i.e. <3 months, which includes an emergency room, where suicide attempters are initially received. If hospitalization is required, the patient is then transferred to an inpatient unit: medicine (general or specialized), surgery, or psychiatry. Alternatively, the attempter may be transferred to a second type of facility-the specialized institution-such as a psychiatric hospital or rehabilitation center, geared for longer stays (< 3 years). French law requires pediatric departments in general hospitals to receive suicide attempters below the age of 16 years, provided that beds are available. The physician in charge may, however, admit older adolescents but is under no obligation to do so. Public psychiatric facilities are found both in general hospitals and specialized institutions. Unlike psychiatric wards in general hospitals, the latter are sectors-orientated, with patients being admitted according to their place of residence. Private clinics are mainly geared to surgery and do not have emergency facilities.
References 1. Shafer D, Fischer P. The epidemiology of suicide in children and young adolescents. ] Am Acad Child Psychiatr 1981;20:545-65.
11. Hawton K. Annotation, attempted suicide in children adolescents. J Child Psycho1 Psychiatr 1982;23:497-503.
and
12. Canoui P, Basquin M. Prise en charge des tentatives de SUItide de l’adolescent de 13 a 16 ans. Entretiens de Bichats Psychiatric 1985227-231. 13. Blum R, Smith M. Training of health professionals in adolescent health care. J Adolesc Health Care 1988;9:46-50. 14. American Academy of Pediatrics, Committee on Adolescence. Suicide and suicide attempts in adolescents and young adults. Pediatrics 1988;81:322-4. 15. Jay MS, Graham CJ, Flowers C, et al. Adolescent suicide attempters presenting to a pediatric facility. Adolescence 1989;24:467-72. 16. Hawton K, Gath D, Smith E. Management of attempted tide in Oxford. Br J Psychiatr 1979;2:1040-2.
SUI-
17. Deykin EY, Hsieh CC, Joshi N, et al. Adolescent suicidal and self-destructive behavior: results of an intervention study. J Adolesc Health Care 1986;7:88-95. 18. Rutter M, Graham P, Chadwick OFD, et al. Adolescent turmoil: fact or fiction? J Child Psycho1 Psychiatr 1976;17:35-56. 19. Kozak LJ, Norton C, McManus M, et al. Hospital use patterns for children in the United States, 1983 and 1984. Pediatrics 1987;80:481-90. 20. Gasquet I, Choquet M, Lellouch J. Problemes poses par la prise en charge medico-psychologiqoe des adolescents hospitalises 1 la suite d’une tentative de suicide, oersoectives euidemiologiques. Memoire dc DEA 1991, LJni\;ersi& Paris Vii. 21. Davidson F, Philippe A. Suicide et tentative de suicide aujourd’hui, etude epidemiologique. 1986 Paris; ESF. 22. Marquis F, Deschamps JP, Legras B, et al. L’hospitalisafion des adolescents dans un centre hospitalier regional. Arch Fr Pediatr 1985;42:459-64. 23. Herve C, Gaillard M, Vinllard M, et al. L’hospitalisation des adolescents dans le departemcnt du Val-de-Marne. Ann Pediatr 1988;35:171-8. 24. Hodgman CH, Roberts EN. Adolescent atrician. J Pediatrics 1982;101:118-123.
suicide and the pedi-
25. Khan AU. Heterogeneity of suicidal adolescents. Child Adolcsc Psychiatr 1987i2692-6.
J Am Acad
2. Hawton K, O’Grady j, Osborn M, et al. Adolescents who take overdoses: their characteristics, problems, and contacts with helping agencies. Br J Psychiatr 1982a;l40:118-23.
26. Combes JC. L’adolescent
3. Gispert M, Wheeler K, Marsh L, et al. Suicidal adolescents: factors in evaluation. Adolescence 1985;20:753-61.
27. Levy-Leblord E, Amoux D, Cau D, et al. Une medecine pour adolescents: Qui soigne quoi? Arch i+ Pediatr 1982;39:771-6.
4. Spirit0 A, Brown L, Overholser J, et al. Attempted suicide in adolescence: a review and critique of the literature. Clin Psycho1 Rev 1989;9:335-63.
28. Carlson GA, Cantwell DP. Suicidal behavior and depression in children and adolescents. J Am Acad Child Psychiat I982;21:361-8.
5. Schreiber TJ, Johnson RL. The evaluation and treatment of adolescent overdoses in an adolescent medical service. J Nat1 Med Ass 1986;78:101-8.
29. Garfinkel BD, Froese A, Hood J. Suicide attempts in children and adolescents. Am J Psychiat 1982;139:1257-61.
6. Spirit0 A, Stark LJ, Fristad M, et al. Adolescent suicide attempters hospitalised on a general pediatrics floor. J Pediatric Psycho1 1987;12:171-89. 7. Kotila L, Lonnqvist ferences predicting 264-70.
J. Adolescent suicide attempts: suicide. Acta Psychiatr Stand
8. Spirit0 A, Stark LJ, Hart KJ, et al. Overt behavior
sex dif1988;77: of adoies-
suicidant.
Rev Ped 1988;24:294-300.
30. Christoffel KK, Marcus D, Sagerman S, et al. Adolescent suicide and suicide attempts: a pppulation study. Pediatric Emergency 1988;4:3240. 31. Taylor EA, Stansfeld SA. Children who poison themselves 11. Predictian of attendance for treatment. Br ] Psychiat 1?84b;145:132-5. 32. Garrison CZ, Jackson KL, Addy CL, et al. Suicidal behaviors in young adolescents, Am ) Epidemiol 1991;133:101I!~-14.
422
GASQUET AND CHOQUET
JOURNAL OF ADOLESCENT HEALTH Vol. 15, No. 5
33. Andrews JA. Suicidal attempts among older adolescents: prevalence and cooccurrence with psychiatric disorders. J Am Acad ChiId Ado&c Psychiat 1992;31:655-62. 34. Choquet M, Facy F, Dadidson F, et al. Suicide and attempted suicide among adolescents in France. In: R Fanner, S Hirsch, eds. The suicide syndrome. London: Croom Helm, 198lI73-89. 35. Brent DA, Perper JA, Allman CJ. Correlates of the medical lethality of suicide attempts in children and adolescents. J Am Acad Child Adolesc Psychiat 1987;26:87-91. 36. Tischler CL, McHenry PC. Adolescent
suicide attempts;
some
significant
factors. Suicide Life-Threaten
Behav 1981;11:86-92.
37. Hawton K, O’Grady J, Osborn M, et al. Classification of adolescents who take overdoses. Br J Psychiat 1982b;140: 124-31. 38. Spirit0 A, Hart K, Overholser J, et al. Social skills and depression in adolescent suicide attempters. Adolescence 1990; X543-52. 39. Goldacre M, Hawton K. Repetition of self-poisoning subsequent death in adolescents who take overdoses. Psychiat 1985;1&395-8.
and Br J