Characteristics of and management strategies for 54 suicidal inpatients in a general hospital

This retrospective study was conducted in a large general hospital with 5,000 beds that provides treatment for more than 200,000 inpatients per year. The data were based on a patient safety adverse event network reporting system in which suicidal cases occurred in one hospital from November 2008 to January 2017. Information included patients’ demographic data (age, gender, marriage, date of admission, medical diagnosis, department, etc.), details of the event (time, methods and location of occurrence, etc.), reason or condition leading to the suicidal act, outcome and suggested methods to prevent recurrence.

Statistical analyses were performed with Microsoft Excel spreadsheets and SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Mean, standard deviation and percentage were used for descriptive analysis. The t-test was used to compare the different suicide methods by gender, and the chi-squared test was used to compare characteristics between patients who attempted suicide and patients who completed suicide. In statistical testing, P<0.05 was considered as significant.

A total of 54 cases were identified during the 9-year period. There were 29 women (53.70%) and 25 men (46.30%), with a female:male ratio of 1.16:1.0. The age of cases ranged from 20 to 80 years, and the mean age of cases was 54.87±12.38 years. The mean age was 54.60±11.63 years and 55.10±13.20 years, respectively, for men and women. Almost all cases (92.59%) were married. Females were more common among those who attempted suicide than among those who completed suicide; moreover, there was a significant difference for gender with regard to attempted and completed suicidal inpatients (χ²=4.562, P=0.033). Most cases were in the 45–59 years age group (n=25, 46.30%), followed by those above the age of 60 years (n=18, 33.33%), but no significant difference in age between attempted and completed suicidal inpatients was found (χ²=2.216, P=0.330) (Table 1), which may be related to the insufficient sample size of this study.

Table 1

Characteristics of suicidal inpatients, n (%).

Among the 54 cases, 22 (40.74%) patients had been treated in medical departments, 16 (29.63%) in surgical departments, 14 (25.93%) in the oncology center and two (3.70%) in the obstetrics and gynecology department. The most common medical diagnosis among suicidal inpatients was malignant neoplasms (n=35, 64.81%), followed by digestive diseases (n=5, 9.26%), cardiovascular diseases (n=5, 9.26%) and other diseases (n=9, 16.67%). Medical diagnoses among patients who had attempted or completed suicide were not significantly different (χ²=5.811, P=0.121) (Table 1). Among the malignant neoplasms, gastrointestinal cancer and leukemia or lymphoma accounted for 14.81% of cases, followed by lung cancer (n=7, 12.96%), cervical carcinoma (n=3, 5.56%), pancreatic cancer (n=3, 5.56%), liver cancer (n=3, 5.56%) and other cancers (n=3, 5.56%). Medical diagnoses of suicidal inpatients did not significantly differ by gender (t=0.03, P=0.722) (Table 2).

Table 2

Disease categories of suicidal inpatients.

Length of stay before suicidal behavior covered a wide range (1–42 days) with a mean of 11.31±9.60 days. Completed suicides occurred at a mean of 11.33±9.03 days after admission, and attempted suicides occurred at a mean of 11.33±10.18 days after admission. Male and female inpatients did not have a significantly longer length of stay from admission to suicidal behavior (t=0.474, P=0.801). Suicidal acts were most likely to occur in the first (n=25, 46.30%) and second (n=11, 20.37%) weeks of admission. Among the 24 suicide completers, 11 (45.83%) completed within the first week and four (16.67%) completed in the second week (Table 1).

Both attempted and completed suicides had peak hours, which occurred between 2 a.m. and 6 a.m. Among the 54 patients, 16.67% of suicides occurred between 6 a.m. and 10 a.m., 3.70% between 10 a.m. and 2 p.m., 18.52% between 2 p.m. and 6 p.m. and 16.67% between 6 p.m. and 10 p.m. and 7.41% between 10 p.m. and midnight. However, we did not find a significant effect of time on the occurrence of inpatient suicide (χ²=3.600, P=0.608) (Table 1).

Jumping from heights (n=25, 46.30%) and cutting (n=15, 27.77%) were the most common methods of suicide, followed by drug overdose (n=4, 7.41%), drowning (n=2, 3.70%) and hanging (n=4, 7.41%). A total of 68% (17/25) of inpatients died as a result of jumping from heights, which is the most lethal method of suicide. Nearly two-fifths of suicide attempters tried to harm themselves by using tools (scissors, fruit knives, etc.). One-third of suicide attempters used fatal methods (jumping from heights or hanging) that were similar to those used by suicide completers. There were 51 suicidal acts that occurred within the hospital, 25 (46.30%) occurred from the windowsill and five (9.26%) at the bedside and in the bathroom. Additionally, five suicides (13.3%) occurred outside the hospital. The majority of attempted suicides occurred in inpatients’ beds or inside the ward, and far more completed cases occurred from the windowsill. Evidences of significant differences were observed in the methods and locations of suicidal inpatients (χ²=16.848, P=0.005 and χ²=13.636, P=0.003, respectively) (Table 1).

As shown in Table 3, somatic disease accounted for 27.78% of the suicide acts. The second reason for suicidal behavior was economic pressure (24.07%). Hopelessness and poor social support were reported as 16.67% and 14.81%, respectively, of the causes of suicide. Depression and personality defects accounted for 9.26% and 7.41%, respectively, of the cases. Moreover, there was more than one relevant cause for each case.

Table 3

Relevant causes associated with suicide.

In the present study, most suicide cases were women, which is consistent with a previous study.7 Compared with Chinese men, Chinese women are more sensitive and fragile and more likely to commit suicide when they encounter adverse stimuli. It is noteworthy as reported by Wint and Akil and Bostwick and Rackley that more men commit suicide than women in general hospitals8,9 and more women than men are initially more willing to seek help after experiencing frustration in a Western country.

Patients with malignant tumors had the highest rate of suicidal acts, which is consistent with the related literature.10,11 Hung et al. showed that somatic disease, particularly malignant neoplasms, was the most common disease and increased the risk of suicide in hospitalized patients.12 These cancer patients all had suffered from despair, fear, severe pain, dysfunction or cognitive function, which directly affected their quality of life. Underlying these circumstances, depression and anxiety are expected, and suicidal ideation might appear among terminal patients.

It is important to note that 25 subjects committed suicidal acts within the first week and 11 in the second week of admission, which is similar to the findings of Hung et al.12 These results may suggest that patients are more likely to carry out suicidal acts within the first week after admission. The time of suicidal act is important for suicide prevention. In our study, we found that 37.04% of acts occurred between 2 a.m. and 6 a.m. The result is similar to other reports.11,13 This fact might be attributed to patients’ suicidal acts occurring in the early hours of the morning when supervision was sparser and families were absent; thus, there was a higher likelihood of success during this time period.

The most common and lethal means of suicide in general hospitals was jumping (25 cases). This result is similar to the findings of local Chinese articles.14,15,16 However, in Western populations, cutting is the most frequent method of suicide in general hospitals.17 The difference in suicidal methods appears to be culturally based. The choice of a suicide method is closely related to the availability of suicide tools. At present, the modern ward building of the general hospital is mostly high-rise; most of the hospital’s balconies and windows are not closed or are semi-closed like the psychiatric hospital. Undoubtedly, it is convenient for the patient to commit suicide. Moreover, in this study, cutting was another frequent method of suicide (15 of 54 subjects) because of its easy access for hurting oneself in the ward.

It is reported that hopelessness might make suicide risk increase at least three times; moreover, sustained hopelessness is more predictive of suicidal behavior than occasional hopelessness.18 Patients suffering from depression might be a warning sign before suicidal acts, but unfortunately, depression is often neglected by the medical staff.19 Therefore, a comprehensive evaluation of suicide risk is critical in these patients. According to their suicide risk factors, more attention should be paid to patients’ psychological state. Additionally, it is essential to develop and implement a targeted suicide prevention program for mental disorder patients.20,21

There is no solution to avoid the suicidal potential of inpatients. Modifying the process of patient assessment and care, strengthening staff training and communication and improving the physical environment are critically important. Therefore, to reduce inpatient suicidal behavior and ensure inpatient safety, this paper explores suicide preventive management strategies from the aspects of patients, medical staff and hospital environment, and recommendations are provided for each of these domains.