Strength of evidence for primary prevention of sports-related concussion
| Study | Overview and results | Risk of bias | Quality | Importance | PICO questions |
| Benson et al 38 | Prospective cohort. Compared half vs. full face shield hockey helmets. No difference.RR=0.97 (0.61–1.54). | Unlikely | Low ⊕⊕ | Low | 4 |
| Benson et al 39 | Prospective cohort. Evaluated prior data set evaluating sessions of play lost due to concussion relative to half vs. full face shield hockey helmets. More sessions lost for half vs. full.RR=4.07 (3.48–4.74). | Unlikely | Low ⊕⊕ | Low | 4 |
| Black et al 52 | Prospective cohort. Significantly increased risk of concussion when body-checking is allowed in youth hockey. IRR=2.83 (1.09–7.31). | Likely | Moderate⊕⊕⊕ | Low | 5 |
| Black et al 53 | Retrospective cohort. Significant reduction in concussion in youth hockey league comparing year before and after implementation of rule prohibiting body-checking. IRR=0.34 (0.21–0.56). | Likely | Very low⊕ | Very low | 5 |
| Collins et al 36 | Prospective cohort. Compared new vs. standard-design football helmets. Significant decrease with new design.RR=0.69 (p<0.027). | Likely | Moderate⊕⊕⊕ | High | 4 |
| Collins et al 67 | Prospective observational. Evaluated correlation of preseason neck strength measurements and incidence of concussion in high school athletes. Significant correlation with mean neck strength but no comparison of exercise vs. no exercise. | Unlikely | Very low⊕ | Low | 7 |
| Collins et al 43 | Retrospective. No difference in concussion rate between new and recertified helmets. Data suggest difference between helmet brands. | Unlikely | Low ⊕⊕ | Low | 4 |
IRR, incidence rate ratio; RR, relative risk; PICO, population, intervention, comparator, and outcomes.