Suicide After Breast Cancer: an International Population-Based Study of 723 810 Women

University of Helsinki, Helsinki, Uusimaa, Finland
CancerSpectrum Knowledge Environment (Impact Factor: 15.16). 11/2006; 98(19):1416-9. DOI: 10.1093/jnci/djj377
Source: PubMed

ABSTRACT Few studies have examined long-term suicide risk among breast cancer survivors, and there are no data for women in the United States. We quantified suicide risk through 2002 among 723,810 1-year breast cancer survivors diagnosed between January 1, 1953, and December 31, 2001, and reported to 16 population-based cancer registries in the United States and Scandinavia. Among breast cancer survivors, we calculated standardized mortality ratios (SMRs) and excess absolute risks (EARs) compared with the general population, and the probability of suicide. We used Poisson regression likelihood ratio tests to assess heterogeneity in SMRs; all statistical tests were two-sided, with a .05 cutoff for statistical significance. In total 836 breast cancer patients committed suicide (SMR = 1.37, 95% confidence interval [CI] = 1.28 to 1.47; EAR = 4.1 per 100,000 person-years). Although SMRs ranged from 1.25 to 1.53 among registries, with 245 deaths among the sample of US women (SMR = 1.49, 95% CI = 1.32 to 1.70), differences among registries were not statistically significant (P for heterogeneity = .19). Risk was elevated throughout follow-up, including for 25 or more years after diagnosis (SMR = 1.35, 95% CI = 0.82 to 2.12), and was highest among black women (SMR = 2.88, 95% CI = 1.44 to 5.17) (P for heterogeneity = .06). Risk increased with increasing stage of breast cancer (P for heterogeneity = .08) and remained elevated among women diagnosed between 1990 and 2001 (SMR = 1.36, 95% CI = 1.18 to 1.57). The cumulative probability of suicide was 0.20% 30 years after breast cancer diagnosis.

  • [Show abstract] [Hide abstract]
    ABSTRACT: To further validate an objective measure of physical illness burden, the Cumulative Illness Rating Scale (CIRS). Survey with correlation of CIRS ratings made from physician interviews and review of medical records with post-mortem ratings made independently at tissue autopsy. Victims of completed suicide investigated by both psychological and tissue autopsy (n = 72). CIRS ratings made by examination of tissue at autopsy were highly predictive of analogous ratings based on historical data, accounting for 75% of the variance in CIRS scores. Taking autopsy findings as the gold standard of objective health assessment, historical ratings tend to underestimate physical illness at high levels of tissue pathology and to overestimate burden at lower levels. The CIRS score, when derived from all available sources of medical information, is a valid objective measure of physical illness burden and has broad applicability to research in geriatrics.
    Journal of the American Geriatrics Society 02/1993; 41(1):38-41. DOI:10.1111/j.1532-5415.1993.tb05945.x · 4.22 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Case presentation This is an unusual case where a 49-year old female patient with known schizophrenia, paranoid type and a history of early-stage breast cancer, which was treated more than 6 years earlier, attempted suicide. Computed tomography and magnetic resonance imaging after this incident revealed the presence of multiple brain metastases as the first symptomatic site of recurrent cancer. Further staging lead to the diagnosis of lung, hilar and mediastinal lymph node metastases and histology confirmed estrogen receptor-positive metastatic cancer. Treatment consisted of whole-brain radiotherapy and letrozole. Twenty-one months later, the patient is in continued partial remission.
    Cases Journal 02/2008; 1(1):195. DOI:10.1186/1757-1626-1-195
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: An efficient screening instrument which can be used in diverse settings to predict suicide in different populations is vital. The aim of this study was to use the five-item Brief Symptom Rating Scale (BSRS-5) as a screening instrument for the prediction of suicide ideation in psychiatric, community and general medical settings. Five hundred and one psychiatric, 1,040 community and 969 general medical participants were recruited. The community participants completed a structured telephone interview, and the other two groups completed the self-report BSRS-5 questionnaire. The logistic regression analysis showed that the predictors of suicide ideation for the psychiatric group were depression, hostility and inferiority (p < 0.001, p = 0.016, p = 0.011), for the community group, inferiority, hostility and insomnia (p < 0.001, p < 0.001, p = 0.003), and for the general medical group, inferiority, hostility, depression and insomnia (p < 0.001, p = 0.001, p = 0.020, p = 0.008). The structural equation model showed the same symptom domains that predicted suicide ideation for all three groups. The receiver operating characteristic curve using the significant symptom domains from logistic regression showed that for the psychiatric group, the optimal cut-off point was 4/5 for the total of the significant dimensions (positive predictive value [PPV] = 78.01%, negative predictive value [NPV] = 79.05%), for the community group, 7/8 (PPV = 68.75%, NPV = 96.09%), and for the general medical group, 12/13 (PPV = 92.86%, NPV = 88.48%). The BSRS-5 is an efficient tool for the screening of suicide ideation-prone psychiatric inpatients, general medical patients, and community residents. Understanding the discriminative symptom domains for different groups and the relationship between them can help health care professionals in their preventative programs and clinical treatment.
    BMC Psychiatry 07/2008; 8:53. DOI:10.1186/1471-244X-8-53 · 2.24 Impact Factor