Table 4 Characteristics of the 2 randomised controlled trails on prevention strategies according to the PICO-framework

Andersson et al. (2017) [2]

660 female (49%) and male adult elite handball players from 45 teams participating in two highest leagues in Norway, ~ 22 years old on average, participating irrespective of shoulder injury status at baseline

Teams were block randomised into intervention (n = 22 teams, 331 players) and control group (n = 23 teams, 329 players); intervention group performed the Oslo Sports Trauma Research Centre Shoulder Injury Prevention Program (10 min exercises to improve internal rotation ROM, shoulder external rotation/scapular strength, kinetic chain, and thoracic mobility) 3 times per week during warm-up over 7 months (2014/15), control group performed normal warm-up; baseline questionnaire followed by monthly online monitoring of shoulder injuries

Group differences between prevalence of shoulder problems and substantial shoulder problems (moderate/severe reductions in training or inability to participate therein) in dominant arm

Prevalence of shoulder problems/substantial shoulder problems was 17%/5% in intervention and 23%/8% in control group during observation period, intervention group had 28% lower risk to sustain shoulder problems (odds ratio: 0.72, p = 0.04) than control group, no differences between groups for substantial shoulder problems (odds ratio: 0.78, p = 0.23)

Sakata et al. (2019) [39]

219 female (< 1%) and male youth baseball players from 16 teams participating in regional league in Japan, 9–11 years old, participating irrespective of shoulder injury status at baseline

Teams were block randomised into intervention (n = 8 teams, 117 players) and control group (n = 8 teams, 120 players); intervention group performed the modified Yokohama Baseball-9 Throwing Injury Prevention Program (10 min stretching exercises to improve shoulder/elbow/hip ROM, dynamic mobility exercises to improve scapular/thoracic function, and lower extremity exercises to improve balance) at least once per week during warm-up over 12 months (2015/16), control group performed normal warm-up; baseline questionnaire followed by clinical/ultrasonographic shoulder assessment every 4 months and ball throwing speed pre/post intervention

Group differences between incidence of shoulder and/or elbow injuries; ball throwing speed as performance measure; and differences in defined risk factors as shoulder/elbow/hip ROM, thoracic kyphosis angle, and modified Star Excursion Balance Test performance

Incidence of pooled shoulder and/or elbow injuries was lower (hazard ratio: 1.94, p = 0.010) in intervention (1.7/1000 athlete exposures) than control group (3.1/1000), no differences for isolated shoulder (hazard ratio: 2.08, p = 0.076) and elbow injuries (hazard ratio: 1.79, p = 0.052); ball throwing speed increased more on average (p = 0.010) in intervention (+ 6.4 km/h) than control group (+ 4.1 km/h); intervention group showed also improved shoulder horizontal adduction ROM deficit in dominant side, hip internal rotational ROM in non-dominant side, and thoracic kyphosis angle (p < 0.03)