Page 117 - JSOM Fall 2020

(which included all studies from the previous review), inves- a questionnaire and 12 cohort studies that directly collected
tigations were obtained from three retrieval databases: the orofacial injury data.
National Library of Medicine’s PubMed, Ovid Embase, and
the Cumulative Index to Nursing and Allied Health Literature. Table 1 shows the results of the meta-analysis. For studies in-
Studies were selected for the review if they (1) contained origi- volving both direct data collection (cohort studies) and ques-
nal quantitative data on orofacial injuries; (2) included groups tionnaires, the results were similar. Participants not wearing
involved in sports or exercise activities; (3) included MG us- MGs had more than twice the risk of an orofacial injury com-
ers and MG nonusers; (4) provided either risk ratios (RRs) pared with those wearing MGs.
and 95% CIs comparing injuries among MG users and MG
nonusers, or data that could be used to calculate these RRs TABLE 1 Results of Meta-Analysis Comparing MG User and MG
and 95% CIs; and (5) were written in English. Studies were Nonuser Groups
not included if (1) they involved activities other than sport or Summary Meta-Analysis Statistics
exercise; (2) they compared different types of MGs and did not Risk Ratio 95%
have a non-MG group; (3) they lacked original, quantitative No. of (MG nonusers/ Confidence
injury data; or (4) all or most (≥95%) of the athletes in the Type of Study Studies MG users) Interval
study wore MGs. Direct Data 12 2.33 1.59–3.44
Collection
An important consideration was the fourth inclusion criterion Questionnaire 11 2.32 1.04–5.13
in the aforementioned list. In the absence of RRs and 95% CIs MG, mouthguard.
comparing MG users and MG nonusers, the study had to pro-
vide at least four numbers: (1) injured MG users, (2) total MG The analysis suffered from several limitations. One limitation
users, (3) injured MG nonusers, and (4) total MG nonusers. was that there was only one randomized prospective study.
As shown in Figure 1, if these data were available in the arti- Randomized prospective studies are considered the gold stan-
cle, the orofacial injury risk in the MG user and MG nonuser dard in experimental design because they eliminate bias in as-
groups could be calculated. The RR was simply the orofacial signing participants to a group. For example, if individuals
20
injury risk in the nonuser group divided by that in the user voluntarily chose to use MGs during sports (i.e., not random-
group. If the RR was >1, the risk was higher in the MG nonus- ized), those individuals may differ in some important charac-
ers. Calculation of the 95% CIs was more involved but could teristic from individuals not wearing MGs. Perhaps voluntary
19
be accomplished with these four pieces of data. MG users may be more careful to avoid injury, whereas MG
nonusers may take more risks that expose them to a higher
FIGURE 1 Data necessary to calculate the risk of orofacial injuries injury risk. The likelihood of this problem (and others) is re-
in mouthguard users and nonusers and the risk ratio. duced by random allocation of participants to MG user and
nonuser groups. Other study designs included in the review
18
Injured Not Injured Total
were either nonrandomized prospective cohort studies (n = 6),
nonrandomized retrospective cohort studies (n = 3), one co-
Mouthguard a (n) b (n) a+b
hort ecological intervention studies (n = 3), or cross-sectional
surveys (n = 11). One-group ecological interventions compared
injuries in groups of athletes before and after the introduction
No a (n) d (n) c+d
Mouthguard of MGs. Another limitation was the inclusion of cross-sec-
tional surveys. Surveys suffer from problems of potential re-
call bias, social desirability bias, errors in self- observation,
Risk with MGs = a/a+b and errors in recall of events. 21,22 These survey studies were
Risk without MGs = c/c+d analyzed separately in the meta-analysis for this reason, but
Risk Ratio = Risk without MG/Risk with MG the results of these questionnaire studies tended to support the
MG, mouthguard; n, number of individuals in group. results of studies directly collecting injury data. Despite these
limitations, the currently available data strongly suggest that
Twenty-three articles met the inclusion criteria. Three studies MGs provide a high level of protection against orofacial inju-
involved a football team followed over several seasons and the ries during sport activities.
data were combined to obtain a single RR and 95% CI. An-
other study reported separately on orofacial injuries in foot- Conclusions
ball and ice hockey players, so the RR and 95% CIs for each
sport were treated separately. Thus, the sports examined in More studies are needed on the effectiveness of MGs during
these studies included football (n = 5 studies), rugby (n = 5 military training. The single study conducted in basic train-
studies), basketball (n = 4 studies), ice hockey (n = 1 study), ing suggests that the more training activities for which MGs
field hockey (n = 1 study), handball (n = 1 study), taekwondo are used, the greater reduction in orofacial injuries. Studies
(n = 1 study), and studies that involved a variety of sports on the effectiveness of MGs for prevention of sport-related
(n = 5). All types of MGs (i.e., stock, boil-and bite, custom) orofacial injuries have explored a wide variety of sports, used
were included in the review, with most studies examining MGs different injury case definitions, included different types of
of any type (n = 13), exclusively boil-and-bite (n = 3), exclu- MGs, and have methodological weaknesses stemming from
sively custom (n = 2), or both boil-and-bite and custom MGs weaker study designs. Nonetheless, the data included in the
(n = 2); two studies did not report the type of MG. Injury defi- systematic review suggest MGs can substantially reduce the
18
nitions also differed, with most studies appearing to include risk of orofacial injuries in sport activities and indicate MGs
only injuries to the teeth (n = 13), whereas others included any should be used where there is a significant risk of orofacial
orofacial injuries (n = 10). There were 11 studies that used injuries.

Mouthguards | 115