A Literature Review of Suicide in Cancer Patients

Linda Anguiano, MSN, RN; Deborah K. Mayer, PhD, RN, AOCN, FAAN; Mary Lynn Piven, PhD, PMHCNS-BC; Donald Rosenstein, MD

Cancer Nurs. 2012;35(4):E14-e26.

Abstract and Introduction

Abstract

Background: Cancer survivors have a higher suicide rate than the general population. Oncology nurses need to have knowledge and skills in assessing risk for suicide in this population.
Objective: This study aimed to conduct a literature review on risk factors for and incidence of suicide in patients with cancer and to identify potential screening tools.
Methods: PubMed, CINAHL, and PsycINFO databases were searched to identify research articles in peer-reviewed journals from 1999 to 2009. The variables under study included suicide rate, cancer type, demographic characteristics, and signs and symptoms associated with suicide. In addition, articles focused on suicide risk assessment tools were also included.
Results: Twenty-four articles met the inclusion criteria. As in the general population, suicide risk was higher among men with cancer as compared with women with cancer. Patients aged 65 years or older with cancer have a higher rate of suicide compared with those younger than 65 years, with rates highest among men 80 years or older. Specific diagnoses associated with higher suicide rates include prostate, lung, pancreatic, and head and neck cancers. The first year after diagnosis carries a higher risk for completed suicide. Multiple risk assessment tools have been developed and are effective in identifying patients with depression or hopelessness, factors associated with higher risk for suicide. However, no tools exist that sensitively and specifically predict suicide.
Conclusion: The incidence of suicide in someone with a cancer diagnosis is approximately double the incidence of suicide in the general population. Early detection of depression in special cancer populations, such as older male patients, may help identify those at greatest suicide risk.
Implications for Practice: Oncology nurses should be aware of cancer patients considered at higher risk for suicide. Systematic screening for suicidal ideation and behavior may identify cancer patients at high risk and facilitate appropriate mental health evaluation and treatment.

Introduction

Suicide was ranked as the 11th leading cause of death in the United States, responsible for 33,289 deaths^[1] in 2006. Risk factors associated with suicide in the general population include chronic diseases, pain, depression, both older age and youth, living alone, and unemployment. Gender also influences suicide rates with males at 4 times higher risk of suicide than females. Suicide was the seventh leading cause of death for males and the 16th for women^[2] in 2006. In addition, Americans 65 years or older have a higher incidence of suicide. According to the National Institute of Mental Health, 14.2 of 100,000 older Americans died as a result of suicide in 2006 as compared to 10.6 of 100,000 people in the general population. More than 90% of people who commit suicide in the general population have depression, mental illness, and/or substance abuse.^[2]

The suicide rate for persons with cancer is estimated to be twice the rate in the general US population.^[3] Yet oncology nurses are often unable to identify factors that may put the cancer patient at higher risk.^[4] Failure to recognize suicide risk in patients with cancer may lead to further morbidity and mortality. Therefore, the purpose of this review of the literature was to identify (1) risk factors associated with suicide in cancer patients and (2) evidence-based assessment methods.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table. Suicide in Cancer Patients: Literature Review
Citation	Sample	Method	Findings	Weaknesses
Abbey et al⁵	200 Hospice patients with a life expectancy of less than 6 mo.	Cross-sectional study. The BHS, as well as other distress measures, was administered by trained staff.	All scales were valid and reliable measures of hopelessness. The 7-and 13-item subscales out preformed the original 20-item BHS in the prediction of suicidal ideation and desire for hastened death. The data suggest that the 20-item BHS may be improved when applied to a terminally ill patient by elimination of problematic items.	Small sample Lack of ''gold standard'' to measure hopelessness. State of the art facility. Mostly white and Catholic participants.
Akechi et al. Symptom indicator of severity of depression in cancer patients: a comparison of DSM-IV criteria with alternative diagnostic criteria. Gen Hosp Psychiatry Jpn. 2008.	5431 Cancer patients were referred during the study period.	Retrospective study using a computerized database from 1996 to 2003. Of the 5431 cancer patients, 728 (12.8%) had been diagnosed with depression using an inclusion approach.	DSM-IV, diagnostic criteria have a low ability for discriminating the severity of depression. As several other studies suggest the somatic symptoms among the DSM-IV criteria may not be useful markers for the severity of depression in cancer patients. Not participating in medical care and social withdrawal seem to be good markers of moderately severe depression in cancer patients.	No measurement of depression used. Would have been useful to compare with the Beck Inventory Scale or HRSD. Generalized to only Referred patients with major depression. Cultural differences cannot be randomized.
Akechi et al. Clinical factors associated with suicidality in cancer patients. Jpn J Clin Oncol. 2002.	1713 Psychiatric consultations referred to the National Cancer Centre Hospital East, Japan, from 1996 to 1999.	Retrospective study using database of 1713 consults. Diagnostic and trained staff administered DSM-IV.	62 (3.6%) were related to suicidality, including 44 cases of suicidal ideation, 10 attempts, and 8 requests for euthanasia. Impaired physical functioning and major depression were significant associated factors. Management of major depression and improving physical functioning may help to prevent suicide.	Physician bias History of depression and suicide not addressed Social support factors, family life not addressed
Akechi et al. Suicidality in terminally ill Japanese patients with cancer. Am Cancer Soc. 2003.	140 Terminally ill patients with cancer, whose survival time was estimated to be less than 6 mo.	Follow-up study of consecutive outpatients who registered with the palliative care service. Structured clinical interviews for DSM-III-R to assess suicidal ideation.	At baseline, 8.6% had ideation and 5.0% had interest. Self-reported anxiety and depression were associated with ideation (P = .003). Changes in ideation and interest occurred in 38% and 15% of patients. Suicidality can change in terminally ill patients.	Institutional biasV1 institution Small samples No identification of factors that predict changes in suicidal thoughts. No validated measures
Akechi et al⁶	1721 Cancer patients referred to a psychiatric hospital.	Retrospective study of database using 1721 patients with cancer. Modified DSM-IV diagnostic inclusion approach, which is based on depressive symptomatology regardless of presumed etiology.	220 (12.8%) were diagnosed with major depression. 113 of these 220 had suicidal ideation. The results of this study found that more than half (51.4%) of the referred cancer patients with depression had suicide ideation. The most common cancer site was lung (n = 54, 24.5%). Significant risk factors identified include the following: Poor physical status (OR = 1.29; 95% CI, 1.03–1.63; P = .03) Severe depression (OR = 1.80; 95% CI, 1.89–2.37; P = .0001 Poor physical functioning and major depression are important indicators of suicidal ideation.	Physician bias Physical symptoms other than pain not included. History of depression or anxiety not addressed
Akechi et al⁷	89 Newly diagnosed unresectable non–small cell lung cancer patients. Consecutive patients who were diagnosed in the National Cancer Center Hospital East, Japan.	Baseline and 6-mo follow-up study using the HRSD to assess. Suicidal ideation by a psychiatrist using item 3 of the HRSD. Major depression and adjustment disorders were assessed using the DSM-III-R.	13 of 89 (15%) patients had suicidal ideation. Multivariate analysis indicated pain (OR = 3.72; 95% CI = 1.12–14.69; P = .04). And in depressive disorder, OR = 27.97; 95% CI, 5.18–214.14; P = .0003 were significant predictive factors. Earlier pain management and appropriate psychiatric intervention is needed.	Sample small Sampling bias Type of pain not assessed No family history Lung cancer only.
Breitbart et al⁸	92 Terminally ill cancer patients who passed a cognitive screening test.	Prospective survey conducted in a 200-bed palliative care hospital in New York, New York. Self-report measure on SAHD, DSM-IV, as well as other distress measures were administered and read to patients by trained staff.	15 (16%) had high desire for death using SAHD scale. 16 (17%) DSM-IV classified as major depression. Depressed patients were 4 times more likely to have a high desire for hastened death compared with patients without depression (47% vs 12%). Based on SCID interviews, 15 (17%) of the 89 patients met DSM-IV criteria for major depression. 16 (17%) of the 92 patients had a high desire for hastened death.	Small sample GeneralizabilityV1 institution
Ciaramella and Poli⁹	100 Consecutive cancer patients from the pain therapy and palliative care unit, Santa Chaira Hospital, Italy.	Depression was assessed using a structured clinical interview for DSM-III-R (SCID), Endicott criteria, and HRSD.	49% of patients were depressed using the SCID compared with 29% using Endicott criteria. 28% were depressed using both. When modified for psychological symptoms in place of somatic symptoms, the prevalence rate dropped from 49% to 29%. These 2 structured interviews were equally valid. Suicide was not predicted by lifetime depression but by the severity of present depression, pain, and metastasis.	One institution—bias Cultural influence Small sample
Filiberti et al. Suicide and suicidal thoughts in cancer patients. Tumori. 2001.	Suicidal cancer patients	Review of literature about cancer in suicide patients was done using MEDLINE, PSYLIT, and peer contacts.	Increased communication between physician and patient and palliative individualized care may help reduce suicide in cancer patients.	Older review of literature
Hem et al¹⁰	490,245 Cancer patients registered in the Cancer Registry of Norway from 1960 to 1997.	Retrospective cohort study using the Cancer Registry of Norway, which includes site, diagnosis, grade, type, stage, and time of cancer diagnosis from 1960 to 1997. Suicide was defined by the ICD diagnosis of suicide.	(N = 490,245 patients with 520,823 cancer diagnosis) Death certificates defined suicide. 589 Cancer patients (407 males and 182 females) committed suicide. Risk was highest in the first months after diagnosis, male sex, and respiratory organs. Risk elevated for males and females, SMR of 1.55 (95% CI, 1.41–1.71) and 1.35 (95% CI, 1.17–1.56). Risk was highest in the first months after diagnosis. The risk was significantly increased after with the diagnosis with cancer of the respiratory organs (SMR, 4.08; 95% CI, 2.96–5.4). With all cancer types, the SMRs varied from 0.76 to 3.67. Bronchus, trachea, and lung cancers carried the highest suicide risk in men. Buccal cavity and pharynx carried the highest among women.	Norway study—generalizability concern
Kendal¹¹	1.3 million cancer patients from SEER from 1973 to 2001.	Sex comparative study of persons with invasive cancers. Data were reviewed from individuals with invasive cancer from all SEER data including anatomic sites, gender, and age at diagnosis, race, and marital status.	The frequency of completed suicide for females was 0.02% about one-fifth of males. For males, the strongest association with suicide was with distant metastasis. 4.8 times excess in total suicides in male over female Divorce carried a higher incidence of suicide. Suicide hazard for African Americans was decreased. The high-risk patient was male, with head and neck cancer or myeloma, advanced disease, little social support, and limited treatment options.	Suicide reasons not studied
Llorente et al¹²	Men 65 y and older in South Florida during 1983–1993.	Population-based retrospective cohort review.	667 Completed suicides, 20 were prostate cancer related (3% of the total male suicide sample). The average annual incidence of suicide for men was 55.32 per 100,000 persons, but for men with prostate cancer, the rate was 274.7 per 100,000. The risk of suicide in men with prostate cancer was 4.24 times that of an age and gender-specific cohort.	Information obtained from different sources. ME recall bias One geographical area
Lloyd-Williams et al¹³	79 Hospice inpatients (Lefevre study) 100 Inpatients with metastatic cancer (Lloyd study) 197 Palliative care patients (Chochinov study)	Literature review: Librarian: Cochrane review. Three studies were identified that compared 7 screening tools. HADS. 12-Item general health questionnaire (GHQ-12, Lefevre study). EPDS (Lloyd study). Beck Depression Inventory (Chochinov study). Are you depressed?	The question ''Are you depressed'' has a sensitivity of 1 and a specificity of 1 and a positive predictive value of 1. The 10-item EPDS has a sensitivity of 0.81 and a specificity of 0.79 and a positive predictive value of 0.53. The 14-item HADS has a sensitivity of 0.77 and a specificity of 0.89 and a positive predictive value of 0.48. The GHQ-12 did not give sensitivities and specificities because of its evaluation of somatic symptoms. Beck Depression Inventory, 13 items; sensitivity of 0.79 and a specificity of 0.71 and a positive predictive value of 0.27.	2003 Review Librarian bias
Miller et al¹⁴	1,408 New Jersey residents 65 y and older enrolled in Medicare and in a pharmaceutical insurance program.	Case control study of suicide risk associated with medical illness in older Americans. Period of study, 1994–2002.	The only medical condition that was associated with suicide was cancer (OR, 2.3; 95% CI, 1.1–4.8). 128 Suicides in patients aged 65 y and older. Prostate cancer was the predominant cancer among cases and controls. Of the 19 cases of suicide among patients with cancer, 8 (42%) had prostate cancer. Of the 64 who did not commit suicide, 27 or 43% had prostate cancer.	Study only examined patients on Medicare and a drug prescription program for low-income elderly. Unrecognized depression common in elderly.
Misono et al³	3,594,750 Patients in the SEER program of the National Cancer Institute who were diagnosed with cancer from 1973 to 2002.	Retrospective cohort study of suicide in persons with cancer. Data from the SEER included age at diagnosis, gender, race marital status, and year at diagnosis. Patients were identified as committing suicide if the cause of death was coded suicide and self-inflicted injury. Contingency tables of suicide rates were used to help comparison with the general population as well as different anatomic sites.	Rate of suicide among SEER; 31.4 of 100,000 person years compared to the general US population of 16.7 of 100,000 person years. Patients with cancer have nearly twice the incidence of suicide than the general population. Suicide rates were highest in patients with cancers of the lung and bronchus (81.7 of 100,000; SMR = 5.74; CI, 5.30–6.22) followed by stomach cancers (71.7 of 100,000 person years; SMR = 4.68; 95% CI, 3.81–5.70), and cancers of the oral cavity and pharynx (53.1 of 100,000 person years; SMR = 3.66; 95% CI, 3.16–4.22). Suicide risk was highest immediately after diagnosis; however, it remained increased for more than 15 y after diagnosis. Male sex, white race, and older age had a higher risk.	Cause of death is often misclassified, sometimes as unexplained. Death may be misclassified as accidental. Tobacco and ETOH may increase risk.
Passik and Breitbart¹⁵	Review of depressive disorders in patients with cancer of the pancreas.	Clinical and research review of data on the connection between depression and cancer of the pancreas.	One study found that 76% of patients diagnosed with cancer of the pancreas had psychiatric symptoms compared with only 17% in patients with colon cancers. Another study found 50% of patients with cancer of the pancreas met criteria for depression. Another study reviewed found that 38% of 131 patients with cancer of the pancreas had significant symptoms of depression using the Beck Depression Inventory. Patients with cancer of the pancreas often have depression and anxiety that occur more often than other types of cancer. Identification and treatment of depression can enhance quality of life in pancreatic cancer patients. Depression is best managed with therapy, cognitive behavioral interventions, and antidepressant medications.	Older article Limited to pancreas
Recklitis et al. Suicidal ideation and attempts in adult survivors of childhood cancer. J Clin Oncol. 2006.	226 Adult survivors of cancer seen in clinic.	Evaluated with depression suicidal ideation tools. 29 or 12.8% reported suicidality.	12.8% reported suicidal symptoms. Suicidal thoughts in adult survivors of childhood cancer are related to cancer treatments and physical and physiological well-being.	Convenience sample Generalizability Participants in clinic setting
Nissim et al¹⁶	27 Ambulatory patients aged 45–82 y with advanced lung or gastrointestinal cancer.	Qualitative study. Patients were recruited through theoretical sampling from outpatient clinics at a large cancer center in Toronto, Canada. 54 Audiotape interviews on ''What is life like for you these days'' and questions were imbedded into interview regarding desire for hastened death. Later reviewed by trained staff.	Based on SADH cutoff scores, 10 (37%) were classified as having high SADH scores, 7 had moderate (26%), and 10 had low (37%). The experience of hastened death in advanced cancer was found to be multidimensional and has 3 distinct categories. 1. Desire for hastened death as a hypothetical exit plan 2. Desire for hastened death as an expression of despair 3. Desire for hastened death as a manifestation of letting go	Small sample Limited to large urban cancer center.
Robinson et al¹⁷	206,129 Men and 211,443 women diagnosed with cancer in Southeast England between 1996 and 2005. Database of Thames cancer registry.	Population-based retrospective study of database. Suicide was identified as either from ICD or text in the death certificate.	166 Suicides (117 in men and 49 in women). Mean age at diagnosis of those committing suicide, 67.9 in men and 63.4 in women. Increased suicide in men, SMR = 1.45; 95% CI, 120–173. In women, the SMR was lower and did not reach statistical significance (SMR = 1.19; 95% CI, 0.88–1.57). Both sexes had a downward trend in relative risk of suicide with increased time since diagnosis.	Generalizability
Rosenfeld et al. The schedule of attitudes towards hastened death; measuring desire for death in terminally ill cancer patients. Cancer. 2000.	92 Terminally ill cancer patients, all with a life expectancy of less than 6 mo (after admission to a palliative care hospital).	Interview survey: Patients were administered the SAHD, a self-reported measure. A clinician rated measure of desire for death.	15 (16.3%) Patients endorsed >10 items indicating a high desire for hastened death. The SADH seems to be a reliable and valid measure of desire for death in terminally ill cancer patients.	Sample terminally ill from same institution. State of the art institution, good palliative services. Voluntary, those unwilling to discuss not used.
Walker et al¹⁸	A survey of 2924 cancer patients from an outpatient clinic at Regional Cancer Center in Edinburgh, United Kingdom.	Cross-sectional survey Consecutive patients used touch screen computers before their oncology visits. Each patient completed the 9-item patient health questionnaire, which included question 9 that asks if they are better off dead or of hurting themselves in the previous 2 wk and the HADS 14-item self-report scale and also the 30-item European Organization for Research and Treatment of Cancer Quality of Life Questionnaire.	Data obtained from 2924 patients, 7.8% or 229 were positive responders (95% CI, 6.9–8.9). 8% of outpatients who attended the oncology clinic felt they would be better off dead or thought of hurting themselves for at least several days in the previous 2 weeks. Emotional distress; significant pain; and, to a lesser extent, older age were all associated with positive responders.	Regional bias—generalizability. Self-reported questionnaire versus interview. Cross-sectional—not all data collected.
Yousaf et al¹⁹	564,508 Cancer patients in the Danish Cancer Registry.	Population-based cohort study. Subjects had a cancer diagnosis between 1971 and 1999 were followed up by the Danish Causes of Death registrar for completed suicide, excluding non–melanoma skin cancers.	A total of 1241 suicides (740 men, 501 women) were observed. The overall incidence of suicide was 1.7% for men and 1.4 for women. Suicide rates where higher for men in the first 3 mo and from 3 to 12 mo for women. Increased suicide risk after diagnosis. Suicide risk increased with increased stage of cancer diagnosis. Cancer patients between the ages of 50 and 79 y committed most suicides.	Generalizability

Abbreviations: BHS, Beck Hopelessness Scale; CI, confidence interval; DSM-III-R, Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; EPDS, Edinburgh Postnatal Depression Scale; HADS, Hospital Anxiety and Depression Scale; HRSD, Hamilton Rating Scale for Depression; ICD, International Classification of Diseases; OR, odds ratio; SADH, Schedule of Attitudes toward Hastened Death; SEER, Surveillance, Epidemiology, and End Results; SMR, standardized mortality ratio.