Table 1 Study characteristics and findings

Non-intervention studies

1

van Houwelingen, C., et al. (2013) [27]

Railway systems in the Netherlands and Germany

Ecological study, comparing Dutch and German railway suicides in terms of railway and population parameters.

1475 Dutch and 6105 German railway suicides in 2000 – 2007

Railway suicides

•  Railway length

Positive association:

•  Train frequency

•  Train frequency (RR = 0.74)

•  Number of train passengers

•  Population density (RR = 0.59)

No association:

•  Population density

•  Railway length (RR = 1.49)

•  Number of passengers (RR = 1.20)

2

Hegerl, U., et al. (2013) [28] & Ladwig, K.–H., et al. (2012) [29]

Railway system, Germany

Quasi experimental, ecological study, comparing railway suicidal behaviours before and after the railway suicide incident.

310 railway suicidal behaviours in index period of 2006–2008, 188 in index period of 2009 (Index period = 11th November to 31st December of the year).

Railway suicides and suicide attempts

Railway suicide of the famous German football goal-keeper.

•  Railway suicidal behaviours in the index period increased by 1.81 (IR), after adjusted for daily temperature.

•  Railway suicidal behaviours increased by 2.2 (IR) after 28 days of the incidence, after adjusted for daily temperature.

•  Railway suicidal behaviours increased by 18.8% in the following two years.

3

Niederkrotenthaler, T., et al. (2012) [30]

Subway system, Vienna, Austria

Ecological study, measuring the associations between station characteristics and subway suicidal behaviours.

107 subway suicide attempts and 185 subway suicides in 1979 – 2009

Railway suicides and suicide attempts

•  Presence of a surveillance unit

Positive association:

•  Train types

•  Presence of a surveillance unit (RRs = 1.65, RRa = 1.93)

•  Station construction on the street level versus other construction

•  Distance between stations

•  Faster train (associated with railway suicides after full adjustment RRsf = 3.53, but not associated with railway suicide attempts)

•  Distance between a station to historical sites

•  Number of stations operated in extensive media reporting period (RRs =1.82, RRa = 1.71)

•  Number of stations operated in the years of extensive media reporting

•  Drug scene (RRs = 1.49, RRa = 2.56)

•  Number of train passengers

•  Number of train passengers (RRs = 1.02, RRa = 1.03)

•  Distance between a station to historical sites (only for railway suicide attempts, RRa = 1.78)

•  Station used as local drug scene

No association:

•  Socioeconomic status of the population in the station’s neighbourhood

•  Station construction

•  Distance between stations

•  Socio-economic status of station’s neighbourhood

4

Kunrath, S., Baumert, J., & Ladwig, K-H. (2011) [31]

Railway system, Germany

Quasi experimental, ecological study, comparing railway suicidal behaviours before and after the presumed railway suicide incident.

747 railway suicidal acts from December to March in 2004–2007 Index period: first 2 months after the presumed railway suicide incident (January and February 2007).

Railway suicides and suicide attempts

Media coverage of a presumed railway suicide on the main railway trunk line following by 3 investigators killed by an express train.

•  Daily railway suicidal behaviours increased by 44% following extensive media coverage of the presumed railway suicide incident in the index period, after controlling for unemployment rates and temperature.

5

Van Houwelingen, C. A. J., Kerkhof, A. J. F. M., & Beersma, D. G. M. (2010) [32]

Railway system, the Netherlands

Ecological study, measuring the associations between railway and population parameters and railway suicide rates.

5178 railway suicides and 517 railway suicide attempts in 1980–2007

Railway suicides and suicide attempts

•  Railway length

Positive association:

•  Train mileage

•  General suicides (similar trend)

•  Passenger kilometre

No association:

•  Railway length

•  Free transport for students

•  Train mileage

•  Population density

•  Passenger kilometre

•  General suicides

•  Free transport for students

•  Presence of high risk populations near railways

•  Population density

6

Baumert, J., Erazo, N., & Ladwig, K. H. (2006) [33]

Railway system, Germany

Ecological study, measuring the associations between availability of railway and train and railway suicide trend.

8653 railway suicides and 857 railway suicide attempts in 1991-2000

Railway suicides and suicide attempts

•  Railway length

For subjects aged ≤ 65 years old,

•  Train mileage

Positive association:

•  Passenger kilometre

•  Railway length (AAPC = 3.2)

Negative association:

•  Passenger kilometre (AAPC = −2.7)

No association:

•  Train mileage

For subjects > 65 years old,

Negative association:

•  Train mileage (AAPC = −8.8)

•  Passenger kilometre (AAPC = −10.4)

No association:

•  Railway length

7

Clarke, M. (1994) [34]

Railway system, England and Wales, the United Kingdom

Ecological study, measuring the associations between availability of railway and train and railway suicide

4171 railway suicides in 1852-1947

Railway suicides

•  Growth of railway length

Positive association:

•  Number of train passengers

•  Growth of railway length

•  Number of train passengers

8

Schmidtke, A., & Häfner, H. (1988) [35]

Railway system, Germany

Quasi-experimental, ABABA design (A = baseline phase, B = broadcasting phase)

Years 1976–1984. First broadcasting in 1981, second broadcasting in 1982.

Railway suicides

Twice-broadcast six-episode weekly serial showing the railway suicide of a 19-year-old male student.

•  Railway suicides increased by 175% in the period during and just after the two broadcasts, for the group whose age and sex were most similar to those of the film model.

Intervention studies

9

Matsubayashi, T., Sawada, Y., & Ueda, M. (2012) [36]

71 metro train stations, Japan

Quasi experimental, ecological study, comparing railway suicide number pre- and post-installation of blue light.

The average number of suicides per station-year observations was 0.164 in 2000-2010

Railway suicides

Installation of blue light in 11 stations with blue light and 60 stations without blue light

•  Railway suicides decreased by 84% after installation of blue lights, after controlling for the number of suicides in the previous years, use of faster train, proximity to a psychiatry hospital.

10

Law, C. K., et al. (2009) [37]

Underground system, Hong Kong

Quasi-experimental, ecological study measuring railway suicide before and after the installation of platform screen door.

76 railway suicides in 1997-2007

Underground railway suicides

Installation of platform screen door in 30 stations in year 2002

•  Railway suicides decreased by 59% since the installation of platform screen door, adjusted for age and gender.

•  No sign of suicide substitution to unsealed platforms.

•  Railway suicides of deceased with psychiatric profile decreased by 84%.

11

Niederkrotenthaler, T., & Sonneck, G. (2007) [38]

Subway system, Vienna, Austria

Quasi experimental, ecological study, measuring subway suicide before and after the introduction of media guidelines.

Year 1982/83 to 2004/5

Subway suicides

Introduction of media guidelines in 1987/88.

•  Subway suicides decreased by approximately 62 cases following the introduction of media guidelines, after controlling the passenger number.