The Journal of Continuing Education in Nursing

Original Article 

Effectiveness of Nursing Education to Prevent Inpatient Suicide

Nancy N. Manister, DNS, APRN, FNP-BC; Stephanie Murray, BSN, RN-BC; John Martin Burke, RNC; Madeline Finegan, DNP, MS, RN, NEA-BC; Mary E. McKiernan, MSN, RN-BC, NEA-BC

Abstract

Background:

Inpatient suicide is recognized by The Joint Commission as a preventable sentinel event associated with inadequate patient and environmental assessments. Strategies are needed to meet this Joint Commission requirement.

Method:

Community hospital nurses were provided with classes to increase knowledge of inpatient suicide, patient assessments, and appropriate care. Independent pre- and postclass assessments were performed to measure nurses' confidence when talking to patients about suicidal thoughts and to assess nursing knowledge of actions to take when an at-risk patient is identified.

Results:

Education significantly increased nurses' confidence talking to patients about three of four confidence measures and increased their knowledge of actions to take when suicidal thoughts are identified. Confidence was lower in nurses with more experience; therefore, experience does not appear to increase confidence, and education is needed.

Conclusion:

This study demonstrated that a focused in-service education program can increase nurses' confidence to dialogue with patients about suicidal thoughts and nurses' knowledge of actions to prevent inpatient suicide.

J Contin Educ Nurs. 2017;48(9):413–419.

Abstract

Background:

Inpatient suicide is recognized by The Joint Commission as a preventable sentinel event associated with inadequate patient and environmental assessments. Strategies are needed to meet this Joint Commission requirement.

Method:

Community hospital nurses were provided with classes to increase knowledge of inpatient suicide, patient assessments, and appropriate care. Independent pre- and postclass assessments were performed to measure nurses' confidence when talking to patients about suicidal thoughts and to assess nursing knowledge of actions to take when an at-risk patient is identified.

Results:

Education significantly increased nurses' confidence talking to patients about three of four confidence measures and increased their knowledge of actions to take when suicidal thoughts are identified. Confidence was lower in nurses with more experience; therefore, experience does not appear to increase confidence, and education is needed.

Conclusion:

This study demonstrated that a focused in-service education program can increase nurses' confidence to dialogue with patients about suicidal thoughts and nurses' knowledge of actions to prevent inpatient suicide.

J Contin Educ Nurs. 2017;48(9):413–419.

Suicide rates are increasing, and it is now the 10th leading cause of death in the United States (Centers for Disease Control and Prevention, 2016). In the hospital setting, risk for inpatient suicide is highest among psychiatric patients in psychiatric units; however, risk does not stop at the doors of a general hospital unit. Therefore, risk for inpatient suicide is an important patient assessment for all providers to perform.

Inpatient suicide was first addressed by The Joint Commission as a National Patient Safety Goal (NPSG) in 2004 and is the fourth most commonly reported sentinel event (Joint Commission on Accreditation of Healthcare Organizations [JCAHO], 2016b). As part of the Joint Commission's (2016b) root cause analysis, it was found that inadequate patient and environmental assessment were the most frequently identified factors associated with this devastating event (JCAHO, 2016a). In response to the root cause analysis, the Joint Commission outlined requirements and recommendations for NPSG 15.01.01: Conduct a risk assessment that identifies specific patient characteristics and environmental features that may increase or decrease the risk for suicide (JCAHO, 2016a).

To meet the Joint Commission requirement for NPSG 15.01.01, one community hospital met with an independent behavioral health care risk management and quality consultant to do an institutional evaluation of practice. Recommendations were made to provide nursing staff with education, and a robust mandatory training program was developed by the Nursing Professional Development Department. This program included education on suicide risk factors, warning signs that may require imminent action, how to be alert to changes in behaviors or routines, and review of the hospital policy on environmental precautions to take when a patient at risk is identified.

Current literature was reviewed so that best practices and current evidence on inpatient suicide prevention and nurses' knowledge and confidence to assess patients for suicidal thoughts and behaviors would inform program development. It was considered that nurses' unfavorable attitudes toward suicide and their lack of confidence to assess and work with suicidal patients may negatively affect care (Jacobson, Osteen, Jones, & Berman, 2012; Nader, Tran, Baranyai, & Voracek, 2012; Neville & Roan, 2013). In addition, clinical experiences and religious beliefs have been identified as important factors influencing the assessment process (Neville & Roan, 2013). Discussing sensitive topics related to risk assessment and ongoing evaluation of suicidal patients may be anxiety provoking for some nurses and may even conflict with personal values. However, without nurses' accurate assessment and identification of suicidal thoughts, confidence to discuss suicide and death with patients, and appropriate environmental preparation patients are at risk. Therefore, these important points were addressed in the education program described in this article.

Theoretical Rationale

The Neuman Systems Model (Neuman, 2011) was used to frame the current study. This model envisions the individual as protected by a series of concentric circles, and at the center of these circles is the core of the individual or basic structure. The two outermost circles are the flexible line of defense, which protects against stressor invasion, and inside of this is the normal line of defense—the usual level of patient wellness. These lines surround the lines of resistance, which are coping behaviors, and at the center is the core. When the individual is faced with overwhelming stressors, and these stressors invade the flexible line of defense and normal line of defense, the lines of resistance are activated to prevent a core response. If the lines of resistance are not effective, a core response may occur, demonstrated by a negative change in health (Neuman, 2011). The authors of the current study envisioned suicidal thoughts and increased risk for suicide as a core response.

The authors envisioned the role of the nurse as preventing suicide by helping the patient to strengthen the lines of resistance by assessing and identifying patients who are at risk for suicide, having confidence to talk with patients about suicide and death, and knowing how to activate the appropriate resources to protect and strengthen the patient. Therefore, the education program for nurses was viewed as an intervention of prevention.

Purpose

The purpose of this study was to assess the effectiveness of a staff education program on the In-Patient Suicide Prevention class and program to increase nurses' confidence to talk with patients about suicidal thoughts and their knowledge about the actions to take when a patient with suicidal thoughts is identified. The program was designed to address Joint Commission recommendations to “educate all staff in patient care settings about how to identify and respond to patients with suicide ideation” (JCAHO, 2016a).

Method

Study Design and Participants

This study used a pretest–posttest design to survey clinical nurses at a community hospital in Connecticut. Clinical nurses were invited to participate in this voluntary study in conjunction with attendance of a mandated 1-hour class on In-Patient Suicide Prevention. This class was repeated 40 times, and 577 nurses attended the classes. Approval for this study was granted by the hospital's institutional review board.

Procedures

Nurses signed up for this class electronically and were provided with an additional electronic link to complete the anonymous and voluntary pretest survey. Participation was considered tacit consent, and the consent form described that the participant would be contacted 4 weeks after class to complete an anonymous and voluntary posttest. Four weeks after each class, nurses who attended the class received an invitation with an electronic link to the survey; a reminder was sent at 5 weeks. It was not known whether the nurses who attended the class had completed the pretest survey; however, participants were asked on the anonymous posttest survey whether they had completed the pretest and also whether they had attended the class. Pre- and posttest surveys were distributed via a URL link to SurveyMonkey™, and the blind cc feature was used for survey links sent via e-mail. Participation in these surveys was encouraged by offering a pizza party to the two nursing units with the highest percentage of nurses participating.

Instruments

Two instruments were developed for this study by an expert work group of five nurses, with overlapping specializations in behavioral health, research, administration, and education.

Self-Confidence. As with other studies that measure self-confidence in nurses (Cook et al., 2013; Taylor-Fishwick, Okafor, & Fletcher, 2015), questions were designed to assess clinical nurses' self-confidence. Included were four questions that started with the stem “I am confident in my ability to ask a patient about,” followed by strong emotions, such as depression or hopelessness; experiencing behavioral symptoms, such as irritability, agitations, anger, or restlessness; feelings or thoughts of self-harm; and feelings or thoughts of death or dying. Participants were asked to rate their self-confidence on a 6-point Likert scale (1 = never; 6 = always). Cronbach's alpha for the pretest (.97) and posttest (.96) scales were excellent.

Patient Scenario. A scenario was developed to depict an at-risk patient. This scenario described a 45-year-old patient who was just given a fatal diagnosis and was distraught. He was new to the area with no local family or friends, and he stated that his kids “would be better off” without him. He described his mother's slow death, stating that “it is better for a patient to take control of the situation and end the suffering early on.” Participants were asked to free write about the following: List five precautions that you would implement for this patient as part of the patient's plan of care.

Demographics. Demographic characteristics of the preand posttest samples were comparable (Table 1). Questions included years as an RN; years worked at hospital; education level; whether in school; behavioral health experience; and personal knowledge of a suicide (not a patient). The pretest asked about previous knowledge of the topic of suicide, and answers for nurses without behavioral health experience included member of National Alliance on Mental Illness (.6%); recently attended a lecture or conference (3%); wrote a paper (3%); read a journal article (14%); discussed with colleagues (32%); reviewed hospital's self-study module (46%); and experience with patients (39%). Nurses worked on 18 different units, and this was captured solely to award the incentive prizes (missing 12%).

Demographic Characteristics of Pre- and Postclass Participants Without Behavioral Health Experience

Table 1:

Demographic Characteristics of Pre- and Postclass Participants Without Behavioral Health Experience

Data Analysis

Data analysis was performed using SPSS® version 23. During the study design phase, it was expected that nurses would answer more genuinely if the survey was truly anonymous, and so the decision was made to offer the pre- and posttest surveys as independent surveys. The pre-test survey was completed by 461 participants. Of these, 45 nurses had experience in behavioral health as a nurse (missing = 10), 20 as a behavioral health technician or aide (missing = 22), and 10 nurses as both a nurse and technician or aide. Nurses who had behavioral health experience (n = 75), as well as those who missed this question (n = 31), had higher scores on confidence questions, and these two groups were not included in the main pretest data analysis.

The posttest was completed by 200 participants. Inclusion criteria for the posttest main analysis included completion of the pretest, attendance of the class on In-Patient Suicide Prevention, and not having behavioral health experience. Forty nurses did not complete the pretest survey (one was missing) and one did not attend the class, leaving 159 cases. These two criteria were important, as they comprised the study intervention. Excluded were nurses with behavioral health nursing experience (n = 19), nurses with behavioral health experience as an aide or technician (n = 8), those with both experiences (n = 1), and an additional 14 nurses who were missing this information. Those with behavioral health experience and those who missed this question had higher mean scores on confidence questions. A total of 117 cases met the inclusion criteria for the main posttest analysis.

Missing data on the four confidence questions was minimal (<1%), and these missing values were replaced with mean values. However, many cases were missing data on the pre- (n = 51) and posttest (n = 25) question asking to list patient care precautions for a suicidal patient. To determine whether the participants who did not provide patient care precautions were different from those who did provide patient care precautions, mean scores on the four self-confidence questions were compared for both the pre- and posttests, and no significant difference was found between scores.

It was believed that providing anonymity to participants would increase participation and provide more authentic answers. Scores on the self-confidence questions for the pretest (n = 355) and posttest (n = 117) groups (bootstrapped at 10,000 samples) (Table 2) were analyzed independently and not as matched pairs. These mean scores were then compared using the summary independent samples t test. This provided a conservative estimate of confidence.

Summary t Test: Comparison of Pretest (N = 355) and Posttest (N = 117) Group Means on Self-Confidence Questions

Table 2:

Summary t Test: Comparison of Pretest (N = 355) and Posttest (N = 117) Group Means on Self-Confidence Questions

A significant difference was found between pre- and posttest scores on three of the four self-confidence questions. For these three questions, the effect size was calculated (Wilson, 2015), and the Cohen's d effect size ranged from .24 to .29, signifying a small effect size (Table 2). No significant increase was found in nurses' confidence to ask patients about strong feelings such as depression or hopelessness. However, confidence to ask about experiencing behavioral symptoms such as irritability, agitation, anger, or restlessness; feelings or thoughts of self-harm; and feelings or thoughts of death or dying did have a small but significant increase.

Participants were asked to free write responses to the patient scenario, and these were coded with numerical codes for analysis. Acceptable precautions were based on the hospital policy “Suicide Precaution Guidelines for the Primary RN,” which describes actions to take when a patient is identified as being suicidal. Some examples of actions to take are assignment of a 1:1 sitter; notify dietary to provide only plastic utensils and dishes on meal trays; move the patient closer to the nurse's station; call the physician or nurse practitioner; and remove hazards from the environment. Content from this policy was discussed in the 1-hour education class. A scoring sheet to code answers was created, and interrater reliability was established.

The nurses from the pretest group who completed this section of the questionnaire (n = 304) most commonly chose the following actions to take: notify the physician or nurse practitioner (including psychiatrist) (71%); assign a 1:1 sitter (49%); perform patient assessments (48%); remove hazardous objects from the room (39%); and nonspecific suicidal precautions (24%) (Table 3). Only two cases were missing precautions on this free-text question.

Most Common Precautions Suggested for Suicidal Patient Scenario

Table 3:

Most Common Precautions Suggested for Suicidal Patient Scenario

In the posttest group who completed this section of the questionnaire (n = 92), similar actions were most commonly described: notify the physician or nurse practitioner (including psychiatrist) (86%); perform patient assessments (64%); assign a 1:1 sitter (52%); remove hazardous objects from the room (49%); and nonspecific suicidal precautions (29%). There were more missing free-text answers for the posttest, and some stopped completing the question after the fourth answer (n = 20). One third of both groups included precautions that were appropriate for this patient's diagnosis and social situation; however, they were not priority actions to take for a suicidal patient.

Comparison of demographic data from pre- and postclass groups was similar, especially on influential demographics such as nonpatient suicide experience, and education. However, postclass responses were higher in nurses with more years of nursing experience, and it was considered that this may have influenced increases in confidence and knowledge. Yet, this was not the case. For example, mean values for confidence in the post-test group were higher among nurses with less than 15 years of experience, compared with those with more than 15 years of experience. In addition, for postclass knowledge, nurses with less experience had higher mean numbers for “notify the physician or nurse practitioner” (42 versus 37) and “perform patient assessments” (24 versus 22).

Discussion

This study examined the effectiveness of an education program on In-Patient Suicide Prevention to increase nurses' confidence to talk with patients about suicidal thoughts and their knowledge about actions to take when a patient with suicidal thoughts is identified. The education program was effective in significantly increasing nurses' confidence to ask patients about experiencing behavioral symptoms such as irritability, agitation, anger, or restlessness; feelings or thoughts of self-harm; and feelings or thoughts of death or dying with a conservative small effect size. Although there was a small increase in nurses' confidence to talk with patients about hopelessness and depression, it was not significant. These findings are consistent with other studies of the positive effect of education on confidence to work with patients who are at risk for suicide (Jacobson et al., 2012), and other health problems (Cook et al., 2013; Taylor-Fishwick et al., 2015; Terry & Cutter, 2013). Missing data was minimal on these questions on both the preand postclass assessments. It was considered that higher levels of confidence in nurses with more than 15 years of experience may reflect changes in nursing education to include end-of-life care. It is important to note that increased years of nursing experience do not necessarily increase confidence to discuss suicidal thoughts with patients.

To measure appropriate actions to take when a patient is identified as suicidal, a patient scenario was developed to depict a patient at risk for suicide. The patient was described as socially isolated, distraught, had a fatal diagnosis, and verbalized a perspective of using personal control to end suffering. Nurses were asked to free write interventions based on this scenario. This strategy was used so nurses would have to identify actions, rather than choose from a prepopulated list. The authors considered appropriate actions to take in response to this scenario from the evidence-based hospital policy, which was reviewed with nurses in the education program.

The most common actions listed were similar, and these responses were described for pre- and posteducation nursing actions (Table 3). Important increases in described nursing actions included a 16% increase in performance of patient assessments; a 15% increase in notify the physician or nurse practitioner (or psychiatrist), a 10% increase in the removal of hazards from the room; a 3% increase in assigning a 1:1 sitter; and a 5% increase in nonspecific suicide precautions. Of these, the reported increase in patient assessment was most important, as this is a critical principle of nursing care and the basis for nursing actions. This evidenced the effectiveness of the education program on appropriate nursing actions to take when suicidal thoughts are identified. It is also possible that nurses more easily identified this patient as suicidal after class discussion of suicidal thoughts and behaviors. More answers were missing on the posteducation responses to this scenario, and this may be attributed to survey fatigue rather than lack of knowledge.

During the program, educators led discussion of the policy to prevent inpatient suicide, resulting in a rich dialogue on suicide, patient assessments, and preventive measures. Although not measured, educators reported two distinct themes: that nurses were not clear on various aspects of the hospital policy, in particular about the process of making a patient room safe; and verbalization of discomfort in talking with patients about suicidal thoughts and death.

Limitations

Study findings may be limited by convenience sampling, not having a comparison group, having a single northeastern U.S. location, and patient populations cared for at this community hospital. Study design considered that anonymity would be important, which appears to be true, as demonstrated by patterns of missing data on specific demographic questions that might be used to identify a nurse, such as one with behavioral health experience or who works in nursing unit. Matching pre- and postclass surveys was sacrificed to increase anonymity and greater authenticity. However, the large difference between preand postclass participation may limit generalizability of findings. Analyzing data as independent samples provided conservative results, and therefore the reported study effect size may actually be higher.

Conclusion

Preventing inpatient suicide is an important nursing action that benefits patients and their families, and also protects health care providers from the devastating effects of losing a patient to suicide. This study demonstrated that a focused in-service education program can increase nurses' confidence to dialogue with patients about suicidal thoughts and their knowledge of actions to prevent inpatient suicide, which may lead to improved patient care. Discussion of the hospital policy with educators provided an opportunity to clarify nursing assessments and actions and to express feelings about (inpatient) suicide. This study may benefit other institutions that are developing programs to meet Joint Commission standards.

This study also tested relations of the Neuman Systems Model (Neuman, 2011) that increasing nurses' confidence to talk with patients about suicidal thoughts and knowing what actions to take when an at-risk patient is identified, can lead to appropriate care of patients and strengthening patients' lines of resistance, thereby decreasing risk for suicide. The class provided to nurses was an intervention designed to prevent inpatient suicide. It is suggested that these relations be tested further by measuring the number of patients who are assessed as needing “suicide precautions” before and after an education program.

Next steps for sustaining this initiative are to develop online training for nurses who are new to this institution, and continued discussion of inpatient suicide risk and end-of-life patient care in journal club presentations. It is also recommended that comparisons be made between online training and in-person training.

References

  • Centers for Disease Control and Prevention, National Center for Health Statistics. (2016). Leading causes of death. Retrieved from http://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm
  • Cook, R.S., Gillespie, G.L., Kronk, R., Daugherty, M.C., Moody, S.M., Allen, L.J. & Falcome, R.A. Jr.. (2013). Effect of an educational intervention on nursing staff knowledge, confidence, and practice in the care of children with mild traumatic brain injury. Journal of Neuroscience Nursing, 45, 108–118. doi:10.1097/JNN.0b013e318282906e [CrossRef]
  • Jacobson, J. M., Osteen, P., Jones, A. & Berman, A. (2012). Evaluation of the recognizing and responding to suicide risk training. Suicide and Life-Threatening Behavior, 42, 471–485. doi:10.1111/j.1943-278X.2012.00105.x [CrossRef]
  • The Joint Commission. (2016a). Sentinel event alert 56: Detecting and treating suicide ideation in all settings. Retrieved from https://www.jointcommission.org/assets/1/18/SEA_56_Suicide.pdf
  • The Joint Commission. (2016b). Summary data of sentinel events reviewed by The Joint Commission. Retrieved from http://www.jointcommission.org/assets/1/18/2004-2015_SE_Stats_Summary.pdf
  • Nader, I.W., Tran, U.S., Baranyai, P. & Voracek, M. (2012). Investigating dimensionality of Eskin's attitudes toward suicide scale with Mokken scaling and confirmatory factor analysis. Archives of Suicide Research, 16, 226–237. doi:10.1080/13811118.2012.695271 [CrossRef]
  • Neuman, B.M. (2011). The Neuman Systems Model. In Neuman, B.M. & Fawcett, J. (Eds.), The Neuman Systems Model (5th ed., pp. 3–33). Upper Saddle River, NJ: Pearson.
  • Neville, K. & Roan, N.M. (2013). Suicide in hospitalized medical-surgical patients: Exploring nurses' attitudes. Journal of Psychosocial Nursing and Mental Health Services, 51(1), 35–43. doi:10.3928/02793695-20121204-01 [CrossRef]
  • Taylor-Fishwick, J.C., Okafor, M. & Fletcher, M. (2015). Effectiveness of asthma principles and practice course in increasing nurse practitioner knowledge and confidence in the use of asthma clinical guidelines. Journal of the American Association of Nurse Practitioners, 27, 197–204. doi:10.1002/2327-6924.12147 [CrossRef]
  • Terry, J. & Cutter, J. (2013). Does education improve mental health practitioners' confidence in meeting the physical health needs of mental health service users? A mixed methods pilot study. Mental Health Nursing, 34, 249–255. doi:10.3109/01612840.2012.740768 [CrossRef]
  • Wilson, D.B. (2015). Practical meta-analysis effect size calculator. Retrieved from http://www.campbellcollaboration.org/escalc/html/EffectSizeCalculator-SMD1.php

Demographic Characteristics of Pre- and Postclass Participants Without Behavioral Health Experience

DemographicPreclass % (n = 355)Postclass % (n = 117)
Years as an RN
  1 to 2239
  2 to 51720
  6 to 102517
  11 to 15125
  <152146
  Missing data23
Years at a hospital
  1 to 21921
  2 to 51918
  6 to 102821
  11 to 151412
  <152025
  Missing data03
Education
  Nursing diploma38
  Associate degree2414
  Bachelor's degree5154
  Bachelor's degree in nursing or other degree1516
  Master's degree nursing or advanced practice nurse76
  Missing data02
Currently in school
  Yes2314
  No7783
  Missing data03
Personal knowledge of person who attempted or completed suicide (not a patient)
  Yes4442
  No5658
  Missing data00

Summary t Test: Comparison of Pretest (N = 355) and Posttest (N = 117) Group Means on Self-Confidence Questions

Self-Confidence QuestionGroup MeanSDtp95% CICohen's d
I am confident in my ability to ask a patient about strong emotions such as depression or hopelessness.−1.236.217[−0.394, 0.089]
  Pretest4.63381.21388
  Posttest4.78631.22354
I am confident in my ability to ask a patient about experiencing behavioral symptoms, such as irritability, agitations, anger, or restlessness.−2.588.010[−0.562, −0.078].29
  Pretest4.62821.20064
  Posttest4.94781.08950
I am confident in my ability to ask a patient about feelings or thoughts of self-harm.−2.267.024[−0.554, −0.040].24
  Pretest4.59601.34799
  Posttest4.89331.13244
I am confident in my ability to ask a patient about feelings or thoughts of death or dying.−2.643.008[−0.627, −0.093].29
  Pretest4.38141.41239
  Posttest4.74141.19713

Most Common Precautions Suggested for Suicidal Patient Scenario

PrecautionPretest % (n = 304)Posttest % (n = 92)
Notify physician or nurse practitioner7186
Assign a 1:1 sitter4952
Patient assessment4864
Remove hazardous objects from the room3948
Nonspecific suicide precautions2429
Authors

Dr. Manister is Associate Professor and Director, CEIN/BS Program, University of Connecticut, Storrs; Ms. Murray is Professional Development Specialist, Behavioral Health and Infusion Centers, Mr. Burke is Clinical Nurse, Behavioral Health, Dr. Finegan is Director, Professional Practice and Magnet Program, and Ms. McKiernan is Director, Professional Development, Stamford Hospital, Stamford, Connecticut.

The authors have disclosed no potential conflicts of interest, financial or otherwise.

Address correspondence to Nancy N. Manister, DNS, APRN, FNP-BC, Associate Professor and Director, CEIN/BS Program, University of Connecticut, Augustus Storrs Hall, U-4026, 231 Glenbrook Road. Storrs, CT 06269; e-mail: nancy.manister@uconn.edu.

Received: December 30, 2016
Accepted: May 30, 2017

10.3928/00220124-20170816-07

Sign up to receive

Journal E-contents