A survey of Dutch GPs' attitudes towards help seeking and follow-up care for relatives bereaved by suicide.

BACKGROUND: Relatives who are bereaved by suicide likely consult their GP when they feel the need for professional help. GPs may play a key role in establishing who is at risk for adverse consequences of the loss as they are familiar with relatives' possible psychiatric vulnerabilities. The availability of evidence-based services for relatives of suicide victims is limited. Successful implementation of services needs analysis of key factors considered critical in the achievement of changes. We investigated GPs' management of help requests of relatives bereaved by suicide and examined determinants of GPs willingness to refer for evidence-based follow-up care.
METHODS: A cross-sectional survey among 488 GPs in the northern part of The Netherlands.
RESULTS: A 44% response was achieved (n = 214) during the last 3 years, 38 (18%) were exposed to suicide, 21 (10%) to help requests without being exposed to suicide and 52 (24%) to both suicide and help requests. Out of 106 requests, 69 (65%) were handled by the GP; 60 (57%) were either directly or additionally referred, principally for mental health care. Suicide exposure and female gender were associated with the doctor's perception that follow-up care following a loss through suicide is useful. The perception that help is useful increased the likelihood of GPs' referral for evidence-based follow-up care.
CONCLUSIONS: GPs support the availability of evidence-based follow-up care for relatives of suicide victims. To modify GPs' key role in referring relatives for it, GPs should be well informed of its usefulness and to whom.

Background

A loss through suicide is a dreadful experience to those who are left behind.1 These persons likely consult their GP when they feel the need for help. Little is known of how GPs deal with the bereaved families. Appropriate care meets an unmet need; 26–88% feel that help is required2,3 while evidence-based services are scarce.

Relatives of suicide victims are at increased risk of adverse mental health consequences of the loss, e.g. complicated grief (CG).3–5 CG is characterized by avoidance, disbelief, numbness, detachment and excessive irritability and anger. It is strongly associated with suicidal behaviour, physical morbidity and other mental morbidity.6,7 Moreover, suicidal behaviour8 and completed suicide9 cluster in families, posing relatives of suicide victims at increased suicide risk. Prevention of suicidal behaviours is therefore particularly relevant among relatives of suicide victims. It may be effectuated by referring these high-risk individuals for effective follow-up care.10

Family-based cognitive behaviour therapy (FGT) is more helpful to reduce feelings of guilt and maladaptive grief reactions than usual care.11 Among relatives with suicidal feelings, FGT seems to be more effective to decrease the risk of CG confirming the benefits of grief interventions for high-risk individuals.12

GPs’ active engagement in suicide prevention strategies is acknowledged.13 GPs are identified as key persons to initiate14and ensure follow-up care for people bereaved by suicide.15 GPs’ attitudes towards referral of suicide bereaved relatives for follow-up care are unknown. Knowledge of these factors is of interest since acceptance and implementation of effective interventions need careful analyses of key factors critical in the achievement of changes in practice.16

In order to study the proposed key role of GPs in referring relatives for follow-up care, we investigated the GPs’ attitudes and management of help requests following a suicide within the framework of experimentally implementing FGT. Factors associated with GPs’ willingness to refer relatives were examined. Among these factors was the GPs’ gender as female GPs have shown to be more attentive to psychosocial factors than male GPs.17 The findings from the present study may be of interest for policy makers and mental health care providers deliberating the availability of effective suicide prevention measures.

Methods

A cross-sectional, semi-structured questionnaire survey was sent to a random sample of 488 GPs of the total number of 895 in the northern part of The Netherlands18 GPs’ exposure to patient suicide, bereaved relatives’ help requests, management of help requests, preference as to the type of care and GPs’ contentment with their interventions were assessed. GPs’ involvement in patient suicide was assessed by asking ‘Have you been exposed to a patient's suicide and if so how many times?’ Exposure to help requests was assessed by the following item: ‘Have you been exposed to help requests for bereavement by suicide, not necessarily concerning suicide of one of your patients. If so, how many times?’ GPs were asked to report suicides and help requests over the past 3 years. Of at most the last two help requests, the time lag (in months) between the suicide and the request was assessed. GPs reported whether they counselled the relatives themselves and/or referred for additional help. In case of referral, GPs indicated the type of health care workers. Subsequently, it was assessed if GPs were content with their interventions. GPs could rate ‘yes’, ‘partly’, ‘no’ or ‘I don't know’. Contentment was established if GPs indicated yes. Further, it was assessed what kind of health care setting is generally preferred: mental health care, primary health care or other kinds of care. Multiple services could be indicated.

In the questionnaire, the aims, method, benefits and potential risks of FGT were briefly presented. GPs were asked if they think FGT is useful (yes, no, I don't know). Usefulness of FGT was established if GPs indicated yes. GPs were also asked if they would be willing to refer patients for FGT if it were available (yes, ‘conditionally’, no, I don't know). Willingness to refer relatives for FGT was established if GPs indicated ‘yes’ or ‘conditionally’. Space was left to explain the responses.

Data analyses

Characteristics of the GPs, their exposure to suicide and help requests, management of help requests, preferences as to types of care and GPs contentment with the outcome of their proceedings were descriptively analysed. Next, bivariate associations between GPs’ exposure to suicide, exposure to help requests, GPs’ gender, perceptions of usefulness and willingness to refer for FGT were calculated using logistic regression. Using multivariable logistic regression models, it was examined whether female gender and GPs’ exposure to suicide were independently associated with GPs’ exposure to help requests. Further, we investigated whether female gender, GPs’ exposure to suicide and exposure to help requests are independently associated with the perception that follow-up care would be useful. Finally, the associations between the aforementioned determinants as well as perception that help is useful were examined in independent association with GPs’ willingness to refer relatives for follow-up care. Results were expressed as odds ratios as measures of relative risk, indicating the magnitude of associations. Results are presented with a 95% confidence interval. The two-sided level of significance was set at 0.05. Analyses were conducted using SPSS version 14.

Results

Of the 488 GPs who were approached, 214 (44%) returned the questionnaire. The proportion of female responders was equal to the proportion in the study region (31%).18 Female doctors had fewer mean (SD) years of practice experience than male doctors (11 (7.9) years versus 17 (9.6) years). Sample characteristics and the exposure to suicide and help requests are shown in Table 1.

T

Characteristics, suicide and help request exposure of GPs representing the samplea

Total response	214 (100)
Female responders	66 (31)
Years of practice experience mean (SD)	15 (9.4)
Exposed to suicide	91 (43)
Median n of suicide (95% CI)	1 (1.4–1.7)
Exposed to help seeking	74 (35)
Median n of consultations (95% CI)	2 (1.6–2.0)
Exposed to both suicide and help seeking	52 (24)

Stated in n (%) unless indicated otherwise. CI, confidence interval.

Sixty-two per cent (n = 176) indicated that follow-up care for relatives following a suicide pertains to primary health care, 47 GPs (27%) to mental health care, 31 (22%) indicated that bereavement support is not a professional health care occupation; 9 (4%) did not know. Seventy-four GPs reported 128 help requests; 106 could be explored; 81 (76%) were presented within 6 months following the suicide; 13 (12%) were >1 year after the suicide. Table 2 shows how GPs dealt with the help requests (n = 106). In 61 cases (58%), the GP was content with the results of their interventions, 33 GPs (31%) indicated ‘partial’ and 7 GPs (7%) were not content with the outcomes.

T

GPs’ management of help requests (n = 106)

	n (%)
GP offered grief counselling	69 (65)
GR offered grief counselling + referred relatives for additional help	23 (22)
GP referred relatives directly for additional help	37 (35)
Services relatives were referred fora
Outpatient mental health care	35 (58)
Inpatient mental health care	2 (3)
Clinical psychologist	29 (50)
Mutual support group	15 (25)

Relatives were occasionally referred for multiple services.

One hundred thirty-nine GPs (66%) indicated that FGT might be useful; 146/214 (68%) stated that they would refer patients for FGT if it were actually available; 43 (21%) would conditionally refer patients. Most noted conditions were ‘when FGT has additional effects’ and ‘when the patient wants to’. Bivariate relationships between determinants of GPs’ management of help requests after a suicide are displayed in Table 3.

T

Bivariate associations between determinants of GPs’ management of help requests by relatives bereaved by suicidea

	GPs’ female gender	P	Exposure to patient suicide	P	Exposure to help requests	P	Perception that help is useful	P
Exposure to a patient suicide	1.1 (0.6–2.0)	0.80
Exposure to help requests	0.8 (0.4–1.5)	0.43	6.5 (3.5–12.1)	<0.001
Perception that help is useful	2.7 (1.3–5.7)	0.004	1.7 (1.0–3.0)	0.07	0.8 (0.4–1.4)	0.36
Willingness to refer relatives for follow-up care	0.4 (0.7–2.7)	0.30	0.9 (0.5–1.5)	0.62	0.5 (0.3–0.9)	0.05	4.2 (2.3–7.7)	<0.001

Values are univariable odds ratios with 95% confidence intervals.

Table 4 shows that GPs who were exposed to a patient suicide were six times more likely to be consulted by relatives. Female GPs were consulted equally frequent as male colleagues, irrespective of suicide exposure. More GPs who had recently been exposed to suicide believed that follow-up care for the bereaved would be useful than those who were not. Female gender was strongly associated with the perception that help is useful, independent of exposure to patient suicide and exposure to relative's help request. However, the perception that follow-up care is useful was less likely when GPs had been confronted with help requests in the previous 3 years. The willingness to refer patients for follow-up care did not depend on GPs’ gender or suicide exposure. When GPs were, perceiving that follow-up care is useful, referring patients for it was four times more likely. Adjusting the outcomes for the number of years of practice experience did not change the results.

T

Factors associated with GPs’ willingness to refer relatives bereaved by suicide for follow-up carea

	Exposed to help seeking	P	Perception of usefulness	P	Willingness to refer	P
	OR (95% CI)		OR (95% CI)		(OR 95% CI)
Female gender	1.23 (0.64–2.3)	0.54	2.70 (1.36–5.36)	0.004	1.05 (0.53–2.10)	0.89
Exposure to patient suicide	6.12 (3.29–11.39)	<0.001	2.18 (1.12–4.24)	0.022	0.87 (0.43–1.73)	0.68
Exposed to help seeking			0.51 (0.26–1.01)	0.055	0.54 (0.27–1.08)	0.080
Perception of usefulness					4.13 (2.17–7.86)	<0.001

aAdjustment for years of practice experience of the GP did not change the results. OR, odds ratio; CI, confidence interval.

Although not statistically significant, exposure to help requests was associated with a two times reduced likelihood of the perception that follow-up care is useful and GPs willingness to refer relatives for it.

Discussion

This study demonstrates that GPs think that effective follow-up care for relatives bereaved by suicide (e.g. FGT) is useful and they are willing to refer relatives if it were available. It is in line with Halligan and Corcoran (2001)19 who reported a 62% proportion who thinks that services to rely on in the aftermath of suicide would be helpful. Although it has been suggested that female GPs are more involved in patients who present with psychosocial problems than male colleagues,20 it apparently does not hold for help seeking following a suicide. The high impact of a patient suicide on GPs19 may be an overriding factor in this respect.

Bereaved families believe that the GPs should initiate contact with the family briefly after the suicide.14,15 In this study, it remained unclear whether relatives or GPs were the initiators. Therefore, the findings do not allow the suggestion that GPs meet relatives’ needs sufficiently, taking into account that 57% of the help requests were referred for mental health care (Table 2). Possibly, GPs recognize the psychiatric vulnerability of relatives of suicide victims,21 whereas an estimated 20% of the bereaved are actually at increased risk of psychiatric consequences. It implicates that the high rate of referral for mental health care may, in part, be unnecessary, but probably come along with relatives persisting need for help.2 Possibly, GPs may be inclined to refer relatives anyway due to GPs’ feelings of guilt or due to a disrupted relationship with the bereaved family14,19 or when the GP is blamed for the suicide.11 However, unnecessary referral for mental health care may stigmatize relatives and should be prevented.

Limitations

The response rate (44%) is low, but not uncommon in studies on subjects of sensitive nature (e.g.14,22,23). Probably, GPs who were recently involved in suicide might be somewhat over represented in the current sample. The explanation that higher exposure is causing higher response rates in this research area is supported by a previous finding of Hilligan and Corcoran (2001). They achieved a 79% response in the sample of GPs, of which 86% was confronted with a patient suicide over a 10 years period, whereas Dutch GPs are only four times involved in a patient suicide during their professional life.24 Responders in the current study may have been implicated in a randomised controlled trial (RCT) to the effectiveness of FGT.11 Possibly, this has affected their attitudes and proceedings regarding relatives’ help requests. This assumption is supported by the large proportion of GPs (65%) who counselled the bereaved themselves; more often than a ∼50% rate found in previous studies.25 Unfortunately, it remained unclear to what extent the sample represents GPs who were actually involved in the RCT. Overall, the extrapolation of the results should be applied cautiously; however, issues related to follow-up care for those bereaved by suicide principally concern GPs who were actually involved and this may, in turn, support the findings’ validity. In countries with self-employed doctors and a referral system, GPs are in a better position to provide psychosocial care.26 This may moderate the generalizabiliy of the findings to countries without a referral system.

Clinical implications and future research

GPs should be well informed of the effectiveness of follow-up care for bereaved relatives as when GPs are convinced of the benefits, they are more inclined to refer relatives for it, especially male GPs who previously dealt with help seeking relatives of suicide victims. However, GPs are pragmatic and cautious of change27; their sense of competence regarding the management of help requests of relatives bereaved by suicide may discourage the implementation of evidence-based services.16,28 Given the need for help and its effects for relatives who are liable to adverse health consequences and feelings of guilt, FGT should be available for vulnerable relatives. FGT might be provided in primary care, for instance by consultation of trained mental health nurses. This may give easier access to this kind of help than when provided in mental health care. The availability of FGT should be recommended in guidelines on the prevention of suicidal behaviours. In view of the key role for GPs in suicide prevention, especially in referring people bereaved by suicide for effective follow-up care, deliberations of GPs in the management of these help requests might be explored in future qualitative research.

Declaration

Funding: The IMPRESS study (Primary prevention of CG in the aftermath of suicides. Region-wide implementation of a brief, nurse-led, cognitive behavioural, family-oriented, primary care and evidence-based approach) was funded by ZON-MW (The Netherlands Organisation for Health Research and Development) (24200027); http://www.zonmw.nl/en.

Funding to pay the Open Access publication charges for this article was provided by the department of General Practice Medicine of the University Medical Center Groningen, The Netherlands.

Ethical approval: The study was approved by the University Medical Center Groningen ethics committee (METcUMCG 2002/137).

Conflicts of interest: None declared.