@article{ref1,
title="Cancer diagnosis and suicide outcomes: umbrella review and methodological considerations",
journal="Journal of affective disorders",
year="2021",
author="Calati, Raffaella and Filipponi, Chiara and Mansi, William and Casu, Diego and Peviani, Giulia and Gentile, Guendalina and Tambuzzi, Stefano and Zoja, Riccardo and Fornaro, Michele and Lopez-Castroman, Jorge and Madeddu, Fabio",
volume="295",
number="",
pages="1201-1214",
abstract="BACKGROUND: Suicide outcomes in cancer patients represent a major public health concern. We performed an umbrella review (UR) including all meta-analyses (MAs) and systematic reviews (SRs) published on the association between cancer and suicide outcomes. <br><br>METHODS: Eligible studies were searched in the main scientific databases up to January 23rd, 2021. Eligible MAs/SRs focused on all suicide phenotypes among cancer patients. Evidence of the association was extracted; the credibility and quality of the included studies were evaluated using ad-hoc tools, including &quot;A MeaSurement Tool to Assess systematic Reviews-2-Revised&quot; (AMSTAR-2-R). <br><br>RESULTS: Six MAs and 6 SRs were included. The standardized mortality ratio of suicide in cancer patients was 1.5 to 1.7-fold higher than in the general population. Risk factors for suicide outcomes among cancer patients were male sex and older age, a cancer diagnosis within the prior year, and some specific cancer sites. Among 107 associations, 90 (84.1%) were supported by high credibility of evidence (class II). However, all studies reported a large heterogeneity (I(2)> 50%) and the majority of them reported considerable heterogeneity (I(2)> 75%). All MAs used random-effects measures. All MAs but one assessed publication bias and only one disclosed it. The majority of MAs/SRs showed critically low quality based on AMSTAR-2-R. LIMITATIONS: We could not perform additional analyses due to the limited number of MAs. <br><br>CONCLUSIONS: This UR underlines the inflated risk for suicide among cancer patients. Upcoming, well-designed studies are needed to account for a broader set of variables. Several methodological issues likewise warrant attention.<p /> <p>Language: en</p>",
language="en",
issn="0165-0327",
doi="10.1016/j.jad.2021.08.131",
url="http://dx.doi.org/10.1016/j.jad.2021.08.131"
}