Objective Current data suggest that a history of traumatic open skin wounds may be a risk factor for infectious endocarditis, with limited evidence. We tested the hypothesis that traumatic skin wound is a risk factor for infectious endocarditis.

Methods Using the Japan Medical Data Center (JMDC) database (4 650 927 people aged 20–64 years, 2012–2018) and the Kumamoto database (493 414 people aged ≥65 years, 2012–2017), we conducted nested case–control and self-controlled case series (SCCS) analyses.

Results In the JMDC database, 544 cases hospitalised for infective endocarditis (IE) were matched with 2091 controls; 2.8% of cases and 0.5% of controls were exposed to traumatic skin wounds in the previous 1–4 weeks, with an adjusted OR of 4.31 (95% CI 1.74 to 10.7). In the Kumamoto database, 4.0% (27/670) of cases and 1.1% (29/2581) of controls were exposed to traumatic skin wounds in the previous 1–4 weeks, with an adjusted OR of 4.15 (95% CI 2.04 to 8.46). In the SCCS, the incidence rate ratios for IE were 2.61 (95% CI 1.67 to 4.09), 1.73 (95% CI 1.01 to 2.94), 1.19 (95% CI 0.63 to 2.27) and 1.52 (95% CI 0.82 to 2.74) for the Kumamoto database and 3.78 (95% CI 2.07 to 6.92), 1.58 (95% CI 0.64 to 3.89), 1.60 (95% CI 0.65 to 3.94) and 1.29 (95% CI 0.47 to 3.53) for the JMDC database at 1–4, 5–8, 9–12 and 13–16 weeks after traumatic skin wound, respectively, compared with the baseline period.

Conclusions This study suggests that traumatic skin wound is a risk factor for IE 1–4 weeks after the wound.