Table 1.
Main studies (in chronological order) examining the effectiveness of physical barriers on high-risk bridges for preventing suicide by jumping. By A-B study design we refer to a two-phase study comparing the post-intervention period B to the pre-intervention baseline A.
| Author(s) | Country | Type of Study | Main results |
|---|---|---|---|
| Lester (Lester, 1993)* | USA | Before-and-after single site study (A-B design) | Installing barriers significantly decreased the number of suicides by jumping without any increase at adjacent bridge locations. |
| O’Carroll et al. (O'Carroll and Silverman, 1994)* | USA | Before-and-after single site study (A-B design) | Installing 2.5 m tall barriers decreased the number of suicides by jumping to zero without any increase at other locations. |
| Beautrais et al. (Beautrais et al., 2009, Beautrais, 2001)* | New Zealand | Before-and-after single site study (A-B-A reversal design) | Removal of barriers was followed by a 5-fold increase in suicides. After reinstalling full-length safety barriers no more suicides by jumping were recorded. |
| Bennewith et al. (Bennewith et al., 2007)* | UK | Before-and-after single site study (A-B design) | Installing 2 m tall barriers along parts of the bridge halved the number of deaths by jumping during the following 4 years without any displacement to other locations. |
| Pelletier (Pelletier, 2007)* | USA | Before-and-after single site study (A-B design) | With a tall (3 m) safety fence, the number of suicides dropped from 14 (during 23 years pre-barrier) to zero (during 22 years post-barrier) without any increase at other locations. |
| Reisch et al. (Reisch et al., 2007) | Switzerland | Before-and-after multi-site study (A-B design) | Investigating the effects of barriers on method and site substitution, they found no substitution in women and some method substitution in men in those regions where no other suitable bridge was available. |
| Sinyor et al. (Sinyor et al., 2017, Sinyor and Levitt, 2010) | Canada | Before-and-after single site study (A-B design) | Suicide rates declined from 9.0 deaths/year pre-barrier to 0.1 post-barrier (5 m) with a short-lived increase at nearby sites and no associated increase in suicide by other means. |
| Atkins Whitmer et al. (Atkins Whitmer and Woods, 2013) | USA | Cost effectiveness single site study | Installing barriers would save lives and be highly cost effective. |
| Pirkis et al. (Pirkis et al., 2013, Pirkis et al., 2015)** | n/a | Systematic review and meta-analysis of data from 9 studies | Barriers resulted in an 86 % reduction in suicides by jumping with a 44 % increase at nearby sites, yielding a net decrease by 28 %. Barriers are much more effective compared to other interventions that encourage help seeking or third-party intervention. |
| Perron et al. (Perron et al., 2013) | Canada | Before-and-after single site study (A-B design) | Suicide rates by jumping decreased by over 75 % after installing a 2.5 m barrier with little to no displacement to other sites. |
| Law et al. (Law et al., 2014) | Australia | Before-and-after single site study (A-B design) | Tall (3.3 m) barriers reduced the number of suicides by 53.0 % (p = 0.041) immediately after installation and by 100 % in subsequent years, with no evidence of displacement to other bridge locations. |
| Hemmer et al. (Hemmer et al., 2017) | Switzerland | Before-and-after multi-site study (A-B design) | On average, barriers reduced suicides by about 69 %, with reductions of 100 % recorded for taller barriers (2.3 m minimum) that cover the entire length of the bridge. |
| Sæheim et al. (Sæheim et al., 2017) | Norway | Before-and-after multi-site study (A-B design) | Full-length barriers led to a 100 % reduction in suicides by jumping at 2 bridges, while suicides continued at a bridge that was only partially secured. |
| Berman et al. (Berman et al., 2022) | USA | Before-and-after multi-site study (A-B design) | Installation of a 2.4 m suicide fence at a high-risk inner-city bridge reduced suicides by 95 % without any lasting increase at other bridges in the same city. |
| Bandara et al. (Bandara et al., 2022) | Australia | Modeling study of cost effectiveness | Installing barriers at bridge and cliff sites would yield a 240 % return on investment over 10 years and produce significant monetary savings. |
*These studies are covered by the meta-analysis by Pirkis et al. (Pirkis et al., 2013) and have only included as separate items due to their historic relevance and significance as important landmark studies.
**The percentages reported in the Main result column are for Pirkis et al. (Pirkis et al., 2013). The 44 % increase at nearby sites is largely due to findings be Sinyor and Levitt (Sinyor and Levitt, 2010) who found a 63 % increase at nearby bridge locations. However, in a later study by Sinyor et al. (Sinyor et al., 2017), published 4 years after Pirkis et al., the authors found an overall decrease by 20 % at nearby locations and attributed the initial increase of 63 % to unfortunate media coverage immediately following the installation of the barrier.