Trends and characteristics of terrorist attacks using edged weapons over five decades: implications for mass casualty...

Terrorist attacks using edged weapons have increased globally over the past decade. Their epidemiology over a sustained period remains poorly characterized, limiting the evidence available to inform mass casualty preparedness. This study aimed to describe the global distribution of these attacks over five decades, quantify their kill-to-wound ratio (KWR) relative to firearm and explosive attacks, and characterize attack, target, and perpetrator patterns.

A retrospective cross-sectional study was conducted using the Global Terrorism Database, covering 1970 to 2020. Attacks were identified using a precoded primary weapon classification. Events flagged as doubtful terrorism were excluded. The KWR was calculated using complete-case casualty data and a Poisson-based method with 95% confidence intervals (CI) and compared with firearm and explosive comparator datasets extracted using the same filters.

A total of 2,616 eligible attacks were identified across 12 regions and 120 countries. Two temporal surges were identified: 1989 to 1997, peaking at 177 attacks, and 2014 to 2020, accounting for 43.7% of all attacks and peaking at 235 attacks in 2015. The complete-case aggregate KWR was 2.45 (95% CI 2.35-2.55), exceeding that of firearms (1.76, 95% CI 1.74-1.78) and explosives (0.38, 95% CI 0.37-0.38), with non-overlapping confidence intervals across all three weapon categories. The edged weapon KWR was 1.39 times that of firearms and 6.45 times that of explosives. Of 109 identified lone-actor attacks, 85.3% occurred between 2014 and 2020, concentrated in Western Europe (47.7%) and North America (27.5%); however, lone actors accounted for 8.1% of all attacks during that period (93 of 1,144). Lone-actor attacks produced a KWR of 0.34 compared with 2.67 for group-actor attacks. The median casualty count was one victim per attack.

Edged weapon terrorism produced an aggregate fatality-dominant casualty profile that differed from the wound-dominant pattern of explosive attacks. Preparedness frameworks informed primarily by blast and ballistic injury may not fully reflect this profile. Context-specific planning should account for both the aggregate fatality-dominant pattern and the potential wound-dominant profile seen in coordinated multi-perpetrator events.