Suicidal Behavior and Difficulty of Patients, as Perceived by Community Mental Health Nurses

Abstract

Background: 

Mental health professionals who work in community mental health services play an important role in treating patients after attempted suicide or deliberate self-injury. When such behaviors are interpreted negatively, patients may be seen as difficult, which may lead to ineffective treatment and mutual misunderstanding.

Objective: 

The goal of this study was to assess the association between the grading of suicidality and perceived difficulty. We hypothesized that a higher grading of suicidality is associated with increased perceived difficulty.

Methods: 

We analyzed cross-sectional data from 176 patients who participated in 2 cohort studies: 92 patients in the MATCH-cohort study and 84 patients in the Interpersonal Community Psychiatric Treatment (ICPT) study. The dependent variable was perceived difficulty, as measured by the Difficult Doctor-Patient Relationship Questionnaire (DDPRQ) and the Difficulty Single-item (DSI), a single item measuring the difficulty of the patient as perceived by the professional. Grading of suicidality was considered as the independent variable. Multiple linear and logistic regression was performed.

Results: 

We found a significant association between perceived difficulty (DDPRQ) and high gradings of suicidality (B: 3.96; SE: 1.44; β: 0.21; P=0.006), increasing age (B: 0.09; SE: 0.03; β: 0.22; P<0.003), sex (female) (B: 2.33; SE: 0.83; β: 0.20; P=0.006), and marital status (being unmarried) (B: 1.92; SE: 0.85; β: 0.17; P=0.025). A significant association was also found between the DSI and moderate (odds ratio: 3.04; 95% CI: 1.355-6.854; P=0.007) and high (odds ratio: 7.11; 95% CI: 1.8.43-24.435; P=0.005) gradings of suicidality.

Conclusion: 

In this study, we found that perceived difficulty was significantly associated with moderate and high gradings of suicidality, increasing age, female sex, and being unmarried.

Suicide is a major public health issue worldwide, with ∼800,000 people dying by suicide every year, accounting for 1.5% of all deaths.¹ For each death from suicide, there are 20 suicide attempts (the intention to die by self-injurious behavior), for a total of ∼16 million attempts. Furthermore, it is estimated that 160 million persons have suicidal thoughts annually worldwide.² Mental health professionals working in community mental health services frequently deal with patients who have just attempted suicide, show nonsuicidal self-injury, or have suicidal thoughts,³ and they play an important role in treating these patients.^4,5 Although treating these patients is part of their daily work, mental health professionals often find them challenging to manage and a source of distress.^3,6 Several factors, such as cultural, religious, and professional background and knowledge of and experience dealing with suicidality, influence the behaviors of these professionals,⁷ who may feel incompetent and avoid direct communication with suicidal patients.⁸

Furthermore, an eventual suicide of a patient may evoke feelings of guilt, sadness, and incompetence that are sometimes difficult to handle.⁷ This is more likely to occur when the patient is younger, or the professional has just started working, is still in training, has never experienced a suicide before, or has little support from colleagues.^7,9 A lack of knowledge and understanding of why people show suicidal or self-injurious behavior may contribute to professionals possibly developing the belief that patients are attention-seeking and manipulative.¹⁰ Such a belief could lead to a negative attitude and negative prejudices toward these patients.¹¹ Thus, patients may come to be seen by mental health professionals as “difficult,” a term referring to patients with severe mental illnesses and challenging and ambivalent behaviors who have received insufficient adequate treatment.¹² This perceived “difficulty” can result in mutual misunderstanding and ineffective treatment, when treatment lacks an empirical and theoretical base and clear treatment goals are absent.¹³ As a result the quality of care often becomes low, resulting in more symptoms and long-term and intensive care use and dependency.¹² To our knowledge, the direct association between perceived difficulty and suicidality has not received clear attention in the literature to date. Therefore, the goal of this study was to assess the association between the grading of suicidality and perceived difficulty. We hypothesized that a higher grading of suicidality is associated with increased perceived difficulty.

METHODS

Setting and Sample

To test our hypothesis, we combined data from 2 existing samples of patients who had participated in studies in secondary mental health services (specialist treatment and support provided by various health professionals for patients who were referred to them for specific expert care), in which identical instruments and questionnaires were used.

Sample 1: MATCH-cohort Study

The first sample was drawn from the MATCH-cohort study, a longitudinal study designed to examine the determinants and consequences of long-term use of health services and complex care situations in people with common mental disorders, described in more detail elsewhere.¹⁴ The original study enrolled 283 patients and their professionals from 3 large Dutch mental health services at baseline. The patients were between 18 and 65 years of age with a common mental disorder (eg, depression or anxiety disorder) and/or a personality disorder according to DSM-IV criteria [the Dutch version of DSM-5 was not available at the time the assessments were done in this study (2012-2016) and in the study described below (2014-2016)]. Patients with psychotic, bipolar I, or cognitive disorders as a primary diagnosis and patients who were unable to read and understand Dutch were excluded. Because the patients in the second sample (see below) received secondary mental health care, we only used data from the 92 MATCH patients who also received secondary mental health care at baseline for the present analysis. The other patients in the MATCH sample received other forms of mental health care and were therefore excluded from our analysis.

Sample 2: Interpersonal Community Psychiatric Treatment (ICPT) Cluster Randomized Trial

The second sample came from a multicenter, cluster, randomized controlled trial¹⁵ that studied the effects of ICPT versus care as usual. The inclusion criteria for patients in the ICPT study were identical to those in the MATCH-cohort study (18 to 65 y of age, common mental disorder, and/or a personality disorder). As in the MATCH-cohort study, patients with psychotic, bipolar I, or cognitive disorders as a primary diagnosis and patients who were unable to read and understand Dutch were excluded. The ICPT study included 93 patients, 84 of whom had completed the assessment of the outcome variable at baseline that was needed for our study.

Measures

All instruments and their psychometric properties used in this study were described in the 2015 publication describing the ICPT study protocol.¹⁶ These were the Mini Neuropsychiatric Interview (MINI Plus), the Difficult Doctor-Patient Relationship Questionnaire (DDPRQ), and a single item measuring the difficulty of the patient as perceived by the professional.

Sociodemographic Variables (Assessed by Researcher)

At baseline, questions concerning age, sex, marital status, ethnicity, working situation, education, and income were assessed.

Diagnoses and Grading of Suicidality (Assessed by Researcher)

DSM-IV Axis I disorders were assessed, with the Dutch Mini Neuropsychiatric Interview (MINI Plus), a structured diagnostic interview, administered at baseline. The MINI Plus is the briefest full psychiatric interview available and, depending on the number of disorders, takes between 15 and 45 minutes to administer.¹⁷ Overall, the validity and reliability of the MINI Plus are considered good.¹⁸

Included in the MINI Plus are 6 questions about suicidality, including: Q1 “Think that you would be better off dead or wish you were dead?” (1 point), Q2 “Want to harm yourself or to hurt or to injure yourself?” (2 points), Q3 “Think about suicide?” (3 points), Q4 “Plan or intend to hurt yourself, either passively or actively?” (4 points), Q5 “Take any active steps to prepare to injure yourself or to prepare for a suicide attempt in which you expected or intended to die?” (5 points), Q6 “Make a suicide attempt?” (6 points). The grading of the suicidality ranges from low (1 to 5 points in total), to moderate (6 to 9 points in total), to high (>9 points in total).

DSM-IV Axis II disorders were assessed with the Structured Interview for DSM-IV Personality (SIDP-IV), a structured clinical interview,¹⁹ if the patient screened positive on the 10-item Standardized Assessment of Personality—Abbreviated Scale-Self Report (SAPAS-SR).²⁰ The SIDP-IV is a widely used semistructured interview with good psychometric properties.²¹ The SAPAS-SR has been found to be one of the briefest, most sensitive, and specific screening instruments for personality disorders and is very useful in clinical populations.²²

Perceived Difficulty Outcome Variable (Assessed by Community Mental Health Nurse)

The dependent variable was measured using the DDPRQ, a 10-item instrument that assesses problems in the relationship between patient and professional and perceived difficulty (eg, “How ‘frustrating’ do you find this patient?” or “How at ease did you feel when you were with this patient today?”). The DDPRQ consists of a 6-point Likert response scale from “not at all” to “a great deal.” The sum score is based on 10 items, with a score of 30 or above a cutoff for a patient who is considered difficult and a possible total score of 60 points.²³ Overall, the DDPRQ has good to very good psychometric properties, with Cronbach α of 0.88.²⁴

The dependent variable was also measured using a “Difficulty Single-item” (DSI) question measuring the difficulty of the patient as perceived by the professional. This question was “To what extent do you rate this patient as difficult?” scored on a 7-point Likert scale ranging from “not at all difficult” to “very difficult,” with 7 the highest possible score.²⁵ This single question has not yet been validated.

Statistical Analysis

The sociodemographic variables working status and source of income were dichotomized before further analysis. On preliminary inspection, the outcomes of the DSI proved to be bimodally distributed. Therefore, for further analysis in logistic regression, we used a dichotomized variable with 2 values (“no perceived difficulty” for scores of 1 to 3 and “perceived difficulty” for scores of 4 to 7). Scores on the DDPRQ had a normal distribution and could be analyzed with linear regression.

Univariate linear regression analysis was performed to determine predictors of perceived difficulty based on the DDPRQ sum score and univariate logistic regression was used for the dichotomized single-item DSI score. A significant value of P≤0.20 was used to select variables (demographic and clinical variables and level of suicide risk) to be included in further analysis. Then, multivariate linear regression for the DDPRQ sum score and multivariate logistic regression for the DSI outcome variable were used to identify demographic and clinical variables (diagnoses and grading of suicidality) that were independently related to perceived difficulty as assessed by the DDPRQ and the DSI. All variables were entered in a backward stepwise manner, only retaining the variables that were statistically significant in the model. Significance was set at P value ≤0.05. All statistical analyses were performed using SPSS, version 26.

Ethical Approval

All procedures complied with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. The MATCH study was approved by a certified Medical Ethics Review Committee, The Clinical Research Centre Nijmegen (CRCN), in The Netherlands (Ref: NL41139.091.12), as was the ICPT study (Ref: NL44744.091.13 with NTR: 3988).

RESULTS

We analyzed cross-sectional data from 176 patients who participated in the 2 cohort studies: 92 patients from the MATCH-cohort study and 84 patients from the ICPT study. Perceived difficulty was measured by the DDPRQ and the DSI. The sociodemographic and clinical variables of the patients are shown in Table 1.

TABLE 1 - Sociodemographic Variables, Clinical Variables, and Perceived Difficulty

	ICPT and MATCH (N=176) [n (%)]
Sociodemographic variables
Age (mean±SD) (y)	38.7±12.6
Sex
Female	117 (66.5)
Male	59 (33.5)
Ethnicity
Dutch	164 (93.2)
Other	12 (6.8)
Marital status
Married	50 (28.4)
Unmarried	126 (71.6)
Working status
Employed	39 (22.2)
Incapacitated	70 (39.8)
Volunteer	28 (15.9)
Looking for job	8 (4.6)
Other	31 (17.6)
Education
Primary education	8 (4.6)
Secondary education	52 (29.6)
Tertiary education	116 (65.9)
Source of income
Salary	37 (21.0)
Social benefit	114 (64.8)
Student grant	8 (4.6)
Other	17 (9.7)
Clinical variables on the basis of the MINI Plus
Axis I disorders
Depressive disorder	51 (29.0)
Anxiety disorder	33 (18.8)
Alcohol abuse	16 (9.1)
Substance abuse	12 (6.8)
Axis II disorders
Paranoid PD	7 (4)
Schizoid PD	3 (1.7)
Schizotypal PD	3 (1.7)
Antisocial PD	2 (1.1)
Borderline PD	26 (14.8)
Histrionic PD	1 (0.6)
Narcissistic PD	1 (0.6)
Avoidant PD	28 (15.9)
Dependent PD	12 (6.8)
Obsessive-compulsive PD	20 (11.4)
Grading of suicidality
No suicidality	75 (42.6)
Low	48 (27.3)
Moderate	38 (21.6)
High	15 (8.5)
Perceived difficulty
DDPRQ (mean±SD)	27.3±5.1
DSI (mean±SD)	3.4±1.4

DDPRQ indicates Difficult Doctor-Patient Relationship Questionnaire^23,24; DSI, Difficulty Single-item, a single item measuring the difficulty of the patient as perceived by the professional; ICPT, Interpersonal Community Psychiatric Treatment Study^15,16; MATCH, MATCH-cohort study¹⁴; MINI, Mini Neuropsychiatric Interview; PD, personality disorder.

The sociodemographic and clinical variables were screened for their association with the outcome variable, as shown in Table 2 (DDPRQ sum score, univariate linear regression, and DSI, univariate logistic regression). The variables with a P value ≤0.20 for their association with the dependent variables were included in the multivariate analyses for DDPRQ sum score and DSI (Table 3). We found a significant association between perceived difficulty (DDPRQ) and high gradings of suicidality (B: 3.96; SE: 1.44; β: 0.21; P=0.006), increasing age (B: 0.09; SE: 0.03; β:0.22; P<0.003), sex (female) (B: 2.33; SE: 0.83; β: 0.20; P=0.006), and marital status (being unmarried) (B: 1.92; SE: 0.85; β: 0.17; P=0.025). We also found a significant association between the DSI and moderate (odds ratio: 3.04; 95% CI: 1.355-6.854; P=0.007) and high (odds ratio: 7.11; 95% CI: 1.843-27.435; P=0.005) gradings of suicidality.

TABLE 2 - Univariate Associations Between Sociodemographic and Clinical Variables and Grading of Suicidality and Perceived Difficulty (DDPRQ and DSI)

	Perceived difficulty (DDPRQ) ICPT and MATCH (N=176)				Perceived difficulty (DSI) ICPT and MATCH (N=176)
Variables	B	β	SE	P	B	Wald	95% CI	P
Sociodemographic variables
Age	<0.01	0.13	<0.01	0.077	0.00	0.00	0.977-1.024	0.980
Sex (male vs. female)	1.67	0.15	0.86	0.053	−0.05	0.02	0.487-1.847	0.877
Ethnicity (Dutch vs. other)	−1.28	−0.06	1.53	0.402	−0.52	0.68	0.172-2.048	0.409
Marital status (married vs. unmarried)	1.87	0.17	0.84	0.028	0.16	0.24	0.608-2.283	0.628
Working status (job vs. other)	0.22	0.02	0.79	0.782	0.13	0.10	0.527-2.439	0.749
Education
Primary education	REF	REF	REF	REF	REF	REF	REF	REF
Secondary education	−1.00	−0.09	0.86	0.259	0.18	0.26	0.603-2.383	0.606
Tertiary education	−1.45	−0.12	1.00	0.152	0.30	0.59	0.622-2.965	0.442
Income	0.79	0.06	0.94	0.399	−0.18	0.23	0.404-1.730	0.630
Clinical variables MINI Plus
Axis I disorders	−0.11	−0.01	0.77	0.887	0.26	0.73	0.713-2.356	0.394
Axis II disorders	0.91	0.08	0.83	0.275	−0.29	0.77	0.392-1.429	0.380
Grading of suicidality
No suicidality	REF	REF	REF	REF	REF	REF	REF	REF
Low	1.69	0.15	0.93	0.070	−0.12	0.09	0.414-1.907	0.762
Moderate	1.60	0.13	1.00	0.111	1.15	7.37	1.375-7.195	0.007
High	3.64	0.20	1.42	0.011	1.74	7.99	1.713-19.488	0.005

DDPRQ indicates Difficult Doctor-Patient Relationship Questionnaire^23,24; DSI, Difficulty Single-item, a single item measuring the difficulty of the patient as perceived by the professional; ICPT, Interpersonal Community Psychiatric Treatment Study^15,16; MATCH, MATCH-cohort study¹⁴; MINI, Mini Neuropsychiatric Interview; REF, reference.

TABLE 3 - Multivariate Linear Regression Model With DDPRQ and DSI as Dependent Variables (P≤0.05)

	Perceived Difficulty (DDPRQ) ICPT and MATCH (N=176)				Perceived Difficulty (DSI) ICPT and MATCH (N=176)
Variables	B	SE	β	P	OR	Wald	95% CI	P
Sociodemographic variables
Age	0.09	0.03	0.22	0.003
Sex (male vs. female)	2.33	0.83	0.20	0.006
Marital status (married vs. unmarried)	1.92	0.85	0.17	0.025
Grading of suicidality (high vs. not high)
No suicidality	REF	REF	REF	REF	REF	REF	REF	REF
Low	1.44	0.89	0.12	0.109	0.89	0.09	0.414-1.907	0.762
Moderate	1.10	0.97	0.08	0.262	3.04	7.37	1.355-6.854	0.007
High	3.96	1.44	0.21	0.006	7.11	7.99	1.843-27.435	0.005

Bold indicates statistically significant values (P≤0.05).

DDPRQ indicates Difficult Doctor-Patient Relationship Questionnaire^23,24; DSI, Difficulty Single-item, a single item measuring the difficulty of the patient as perceived by the professional; ICPT, Interpersonal Community Psychiatric Treatment Study^15,16; MATCH, MATCH-cohort study¹⁴; OR, odds ratio; REF, reference.

DISCUSSION

In this study, we assessed the association between perceived difficulty of patients by mental health professionals and the grading of suicidality. We hypothesized that a higher grading of suicidality would be associated with more perceived difficulty. Perceived difficulty indeed was significantly associated with moderate and higher grades of suicidality, meaning that perceived difficulty increased with increased grading of suicidality.

Not much empirical research has been done on perceived difficulty, but the existing literature reports that contributing factors involve a combination of professional factors (eg, poor communication skills, stress management), patient factors (eg, a personality disorder, self-destructive behavior), and organizational factors (eg, conflicts within a team).^9,26 A cross-sectional survey found that perceived difficulty cannot be explained by individual patient characteristics but rather by treatment characteristics perceived by the mental health professional, such as ‘’feeling powerless.”²⁷ Our findings are in part consistent with these conclusions. However, we did find some patient characteristics that appeared to contribute to perceived difficulty. As far as we know, no other research has suggested or found that marital status, age, and sex contribute to perceived difficulty.

Previous studies have stressed the complexity of working with suicidal patients.^28–30 More than half of the patients in our sample had some degree of suicidality (57.4%). A diligent clinical assessment of suicide risk by mental health professionals is important when a patient has suicidal thoughts or plans.³¹ In addition to assessment, decisions also need to be made about necessary care, with potentially important consequences for the patient.³² The results of our study showed that patients with a moderate or high level of suicide risk were likely to be perceived as difficult in the interpersonal relationship by their mental health professional. Several other studies have found similar results,^28,30,33 but they did not differentiate among levels of suicidality. Difficulties in clinical work with suicidal patients may occur due to a lack of knowledge about suicidality or the use of ineffective interventions (eg, nonsuicide contracts),³⁴ but they may also result from negative prejudices and attitudes toward suicidal patients.¹¹ The feeling of not being taken seriously by their mental health professional is harming to patients’ feelings and may strengthen their feelings of incompetence or hopelessness. Systematic research has shown an association between a strong therapeutic alliance and fewer suicidal thoughts; this finding was reported in longitudinal studies in which the alliance with a mental health professional was evaluated.³⁵ The therapeutic alliance, therefore, requires special attention when treating patients with some level of suicidality.³⁶

Strengths and Limitations

To our knowledge, this study is the first attempt to describe the direct association between perceived difficulty and suicidality. Our analysis has a few limitations. First, the cross-sectional nature of our data precludes predictive, causal conclusions. Future research could further clarify the relationships among perceived difficulty and level of suicide risk by, for example, utilizing longitudinal and experimental methodologies.

Second, we realize that the suicide grading measure used in the current study was relatively limited in scope. The items used in this research were taken from the MINI Plus,¹⁷ as part of the assessment of Axis I diagnoses, which is not a structured assessment of suicide risk. Future research should carefully examine the replicability of the current results using more sophisticated suicide risk measures.³⁷

Relevance for Clinical Practice

In this study, moderate and high gradings of suicidality were significantly associated with perceived difficulty by community mental health nurses. Training to improve practices in dealing with suicidal patients is recommended; however, resources for such training are scarce. In The Netherlands, the Dutch Multidisciplinary Guideline for the Assessment and Treatment of Suicidal Behavior³⁸ recommends the use of the Chronological Assessment of Suicide Events (CASE) approach. This approach, which was originally developed by Shea,³⁹ includes (1) gathering information related to risk factors, protective factors, and warning signs of suicide; (2) collecting information related to the patient’s suicidal ideation, planning, behaviors, desire, and intent; and (3) making a clinical formulation of risk based on all of the information. A recent Dutch study by Jongkind et al⁴⁰ found that the CASE approach is used in a simplified form in The Netherlands, and the authors recommended that practitioners be educated to utilize the CASE approach more thoroughly. The Collaborative Assessment and Management of Suicidality (CAMS) approach, which was developed by Jobes,⁴¹ is a treatment framework in which a patient and a mental health professional work together to keep the patient stable, ideally in outpatient therapy. Its specific goals are the evaluation, treatment, and management of chronic suicidal behavior.

A recent meta-analysis comparing the effectiveness of CAMS with other approaches showed promising results in terms of significantly lower suicidal ideation and general distress, significantly higher treatment acceptability, and significantly lower hopelessness.⁴² The CAMS approach has been fully implemented in Danish and Norwegian mental health care, but not yet in other countries as far as we know. Another well-known model used in the United States is the Assess, Intervene, and Monitor for Suicide Prevention (AIM-SP) model, which has been proposed as a framework for implementing zero suicides in clinical care. “Assess” refers to the use of screening and risk assessment to identify patients at risk. “Intervene” consists of conducting suicide-specific brief and psychosocial interventions. “Monitor” provides strategies for ongoing monitoring and increased contact during known high-risk periods. AIM-SP provides guidelines for clinical training and best practice in suicide prevention that can be applied in a wide range of care settings.⁴³ The framework can be used in long-term outpatient care.⁴⁴

Besides evidence-based programs or frameworks, clinical supervision and support and feedback from colleagues remain an embedded resource for practice quality in community mental health institutions to increase competence and decrease stress, and they have been associated with decreased depressive symptoms in mental health care professionals.⁴⁵ Since patient “difficulty” as perceived by mental health professionals often results in ineffective treatment, higher levels of care use, and persistence of symptoms,^45–47 further research is needed concerning the factors underlying the perceived difficulty to enhance outcomes for patients and improve mental health professionals’ understanding of these issues.

CONCLUSION

We observed that perceived difficulty is significantly associated with moderate and high gradings of suicidality, increasing age, female sex, and unmarried status.