Objective To synthesise the current evidence regarding the risk factors, the injury prevention strategies and the profile of injured skiers and snowboarders in terrain parks (TPs) and half-pipes (HPs).
Design Systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines.
Data sources Literature searches from six electronic databases and manual searches were performed.
Eligibility criteria for selecting studies Inclusion criteria were: (1) publication based on original data; (2) injuries sustained in TPs or HPs; (3) recreational skiing or snowboarding injuries; (4) observational or experimental study design with a comparison group.
Results No study explored the risk factors in HPs or the prevention strategies in TPs or HPs. From the literature retrieved, there is strong evidence that skiing or snowboarding in a TP is a risk factor for head, neck, back and severe injuries. Two papers assessed the risk factors for injuries in TPs, mainly demonstrating that features promoting aerial manoeuvres or a large drop to the ground were associated with higher feature-specific injury rates. The profile of injured skiers and snowboarders in TPs described in the literature suggested some evidence of associations between factors including activity, sex, skill level, helmet use, age and TP injuries.
Summary/conclusions This systematic review demonstrates the need for studies identifying the risk factors for injuries to skiers and snowboarders and on interventions to reduce the risk of injury in TPs and HPs. Studies addressing the issue of TP design should be considered.
PROSPERO registration number CRD42016045206.
- injury prevention
- intervention effectiveness
- risk factor
Statistics from Altmetric.com
Since the demonstration of moguls and aerials competitions at the 1988 Winter Olympic games, other disciplines like ski and snowboard cross, half-pipe and slopestyle appeared in Olympic competitions and Winter X games.1 2 The high levels of competition promote the execution of spectacular acrobatic manoeuvres.3 4 Ski area terrain parks (TPs) have become much more popular since the early 2000s.5–9 In 2011, the National Ski Areas Association have reported the results of a survey showing that 94% of the ski areas under their surveillance had at least one TP.10
TPs include features like jumps, half-pipes (HPs), quarter-pipes, mushrooms, boxes, rails, kickers and so on.9 Their design promotes acrobatic manoeuvres and heights that can increase the risk of injury.9 There is evidence that the risk of severe injuries is higher in TPs than on regular slopes, and that these injuries are more likely to occur in self-reported ‘experienced’ skiers and snowboarders.6 8 9 11–20 TP injuries are more likely to affect the head, neck and spinal cord.3 8 21–23 A similar aetiology to TP injury is also observed in high-level competitions in skiers and snowboarders performing on similar features.3 4 24–30
Ski areas with TPs have to deal with participants who want to be challenged but must balance this against ensuring a secure environment. As concluded in the systematic review of Hume et al,19 using Haddon’s prevention concepts, the design of TPs should consider prevention of injuries. Research related to playground and skatepark injuries has demonstrated an association between the quality of the physical environment and the risk of injury,31–34 and this may also be true for TPs. According to McNeil et al,35 most TPs have not been designed by a qualified engineer.
Research on TP injuries is a relatively new field of interest. Considering the Translating Research into Injury Prevention Practice framework developed by Finch,36 a few risk factors are reported in the literature and intervention effectiveness and the implementation context are still insufficiently described. In addition, population targets should be defined to create more efficient prevention countermeasures.36–38 In the perspective of injury prevention in TPs, interventions aiming to create a safer physical environment are important, as they do not require any overt action on the part of the user to provide his/her safety.6 9 35 39–42 Nevertheless, this passive strategy should be paired with active strategies involving human adaptation.43 This systematic review has a primary aim to synthesise and highlight the current evidence regarding the risk factors and the injury prevention strategies for skiers and snowboarders in TPs and HPs. The secondary aim is to define the profile of injured skiers and snowboarders in TPs and HPs.
For the purpose of this review, TPs were defined by any delimitated physical location that included features such as a HP, quarter-pipe, jump, box, rail, mushroom or other man-made features installed by operators of ski areas. Only studies including clear details on the physical location of TPs were included in this systematic review. In addition, all studies with or without detail on the physical location of HPs were included because they are not always part of TPs but are always designed and maintained by ski area operators.
Data sources and search strategy
To ensure the quality of the review, the methods used respect the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analysis).44 The protocol of this research has been registered in PROSPERO and can be retrieved online (https://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42016045206).
Literature searches were done in the following six electronic databases: PubMed; SPORTDiscus (1985 to present); CINAHL Plus (1937 to present); Web of Science Core collection (1900 to present); Physical Education Index (1970 to present); ProQuest Dissertations and Theses Global Database. Two independent searches (OA, TPM) were conducted before 1 August 2016 over a 2-month time period. The initial search strategy was conducted a second time in the original databases (8 May 2018) to update our results. The strategy used (figure 1) involved a combination of subject headings, keywords, related synonyms and Boolean operators. We focused on five main concepts: (1) Terrain park AND (2) Sport AND (3) Risk factor OR (4) Injury OR (5) Intervention. An example using the PubMed database is available in online supplementary files 1 and 2 . Databases were limited to humans and English-language publications but were not limited to a publication date. Only peer-reviewed literature (manuscripts and published abstracts) and dissertations were included. A manual search in ASTM volumes (Skiing and Trauma Safety, volumes 8–20) and in reference lists of selected studies was also performed. Duplicate studies were removed at this step.
Selection of studies
Independent screening of all publications in each database by title and abstract was done by two authors independently (OA and TPM or OA and MDL). Potential literature kept from at least one author was retrieved as full text. In addition, full articles were retrieved when an abstract was missing and the title was relevant. The same authors have independently reviewed the full text of all articles retrieved to identify final study selection. Eligibility criteria were assessed by two reviewers (OA and CG). Disagreements were resolved through consultation with a third reviewer (AM). To be included in this systematic review, studies had to meet the following criteria: (1) publication based on original data; (2) injuries sustained in TPs or HPs; (3) recreational skiing and/or snowboarding injuries; (4) observational or experimental study design with a comparison group including randomised controlled trial, quasi-experimental, cohort, case-control (prospective or historical) and cross-sectional. Review articles and case studies were excluded.
Data extraction and risk of bias assessment
Three reviewers (OA, TPM, CAE) extracted data from the included studies into a structured table. Information was included on references (first author and year of publication), characteristics of the studies, characteristics of the participants, outcomes, exposures, control condition description, reported incidence rates-proportions and/or other useful statistics, measures of association and quality assessment of studies. When relevant data were missing in the publication, authors were contacted.
The Newcastle-Ottawa Scale (NOS),45 designed especially for cohort and case–control studies, was independently applied to each included article by two reviewers (OA and BEH or OA and CG) to assess quality. On a total score of eight points, this tool aimed to evaluate each article on selection of study subjects (four points), comparability of comparison groups (one point) and exposure ascertainment (three points). A modified version of the NOS was used to assess case–control and cross-sectional studies where one less point was awarded for selection criteria. In the case of a disagreement on the NOS score, a third author (AM) was consulted.
A priori, we planned to conduct meta-analyses on our primary outcome (risk factors and prevention strategies) to provide estimated measures of effect. However, the limited number of studies and some heterogeneity precluded meta-analysis. Consequently, we decided to report our findings using narrative syntheses. Our secondary outcome (the profile of injured skiers and snowboarders in TPs) was also reported with this approach.
The search strategy initially identified 268 records and 193 were duplicates. Reviewers screened 75 studies by title and abstract and identified 38 as potentially relevant. In addition, 25 articles were retrieved through reference lists of chosen-so-far articles and 3 articles were identified in updated search for further examination. A total of 66 full-text articles, abstracts and dissertations were assessed for eligibility. Forty-nine records did not meet eligibility criteria. Finally, 12 studies obtained through databases and one through our manual search were included (n=13). Manuscripts on the same study were grouped at this stage. Study designs were case–control (n=10) or cross-sectional (n=3). To summarise the literature search, the PRISMA flow diagram is presented in figure 2.
Data extraction and risk of bias assessment
Because of the study designs (case–control and cross-sectional), the quality of studies was assessed using only the modified NOS. For all papers, points awarded by reviewers ranged from four to six (scale range 0–7). Subsequently, the mean (5.2) and the median (5.0) scores for the methodological quality were considered as similar. Percentage of agreement between assessors, calculated for all items (84/91), was 92%. The Cohen’s Kappa coefficient (κ) was 0.84, corresponding to an excellent level of agreement46 between assessors. Briefly, the quality assessment revealed that (1) data on exposure were frequently missing; (2) number of injuries may be underestimated because usually they are not all reported (ski patrols bypassed); (3) the non-response rate is rarely reported or poorly described; (4) not all studies addressed the issue of confounding through design or analysis. The results of the risk of bias assessment and the relevant details of studies are presented in online supplementary file 3.
Supplementary file 3
Profile of the injured skiers-snowboarders in TPs
According to our systematic review and eligibility criteria, before 2018, thirteen studies related to TP injuries were conducted in six different countries by eight different groups of researchers (one from Austria, one from Canada, one from France, one from Norway, one from Spain and three from the USA). Most articles had a major focus on injuries that occurred on regular slopes and provided little information about TP injuries. This section highlights the most common characteristics of injured persons in TPs reported in the literature. Although this is our second outcome of interest, we present it at first to facilitate the comprehension of the following sections on risk factors for injuries and injury prevention strategies (our primary outcome of interest).
The profile of skiers-snowboarders injured in TPs was compared with the profile of skiers-snowboarders injured on regular slopes. Note that each characteristic should be interpreted separately and not as a set.
Among the literature retrieved, seven studies6–8 22 47–49 reported the type of activity practised when participants were injured in a TP or on a regular slope. The collected data suggest that the odds of being a snowboarder (rather than a skier) are four times higher in TPs than in regular slopes. Moreover, in TPs, snowboarders were reported to have a higher risk of traumatic brain injury whereas skiers were reported to have a higher risk of other types of head injury.49
Three papers7 8 22 reported the sex of skiers-snowboarders injured in a TP or on a regular slope. The univariate analyses suggested that the odds of being a male (rather than a female) are four to eight times higher in TPs than on regular slopes. A similar result was obtained by one study using a multivariable logistic regression analysis (OR 3.46; 95% CI 2.10 to 5.72).22
The skill level of the skiers-snowboarders injured in a TP or on a regular slope has been explored by two studies.7 8 Compared with non-expert, the odds of being an expert are two to three times higher in TPs than on regular slopes.
Three studies7 8 22 provided data on the proportion of helmet use for injured participants in TPs and on regular slopes. The literature demonstrates that the odds of observing an injured skier or snowboarder using a helmet in TPs were greater than on regular slopes.
The univariate analyses provided in studies7 8 22 indicate that skiers and snowboarders injured in TPs were significantly younger than those injured on regular slopes. This is also supported by a multivariable logistic regression analysis done by Ruedl et al 22 (OR 1.06, 95% CI 1.04 to 1.08).
Other potential characteristics of injured skiers-snowboarders in TPs
A relatively large number of characteristics of injured skiers-snowboarders in TPs have been sporadically reported by several authors. More precisely, snow conditions, state resident, season pass holder, number of days at resort, owned equipment, ski lessons, wind, clear visibility and having jumped were other variables assessed in studies. These characteristics require additional research.
Synthesis of the risk factors for injuries
In TPs, the overall injury rates reported are 0.9/1000 runs to skiers and 0.75/1000 runs to snowboarders.9 50 There is evidence that skiers and snowboarders in TPs suffered from injuries that are more likely to be severe than those sustained on regular slopes. As shown in seven studies,6–8 22 23 49 51 the odds of having a head, neck, back or severe injury (vs other types of injury or minor injuries) are greater in TPs. In 2007, Goulet et al 6 suggested that, when an injury occurs in a TP, skiers (adjusted OR 1.36; 95% CI 1.21 to 1.53) and snowboarders (adjusted OR 1.14; 95% CI 1.05 to 1.23) were more likely to be evacuated by ambulance than skiers-snowboarders injured on regular slopes. The most common mechanism of severe injury in TPs reported by authors was a fall after the execution of a jump and/or an aerial manoeuvre.6 8 22 48
Two papers included in our systematic review of the literature have specifically assessed risk factors for injuries in TPs. According to the study of Russell et al 9 52 53 conducted on a population of snowboarders, self-rated expert, listening to music (no), previous snowboard injury (no), temperature between −10°C and 0°C, and snowboarding over night are factors that significantly increase the odds of having an injury in a TP. Furthermore, the authors concluded that features promoting aerial manoeuvres or a large drop to the ground have higher odds and rates of injury. The results suggest that, compared with rails, the odds of injuries are higher for HPs (OR 9.63; 95% CI 4.80 to 19.32), jumps (OR 4.29; 95% CI 2.72 to 6.76), mushrooms (OR 2.30; 95% CI 1.20 to 4.41), kickers (OR 1.99; 95% CI 1.27 to 3.12) and boxes (OR 1.37; 95% CI 0.85 to 2.20). The odds of injuries were lower for quarter-pipes (OR 0.51; 95% CI 0.23 to 1.13). An association of similar magnitude has been observed for odds of severe injuries with feature type for HPs (OR 13.28; 95% CI 5.84 to 30.19), jumps (OR 5.97; 95% CI 3.38 to 10.55) and kickers (OR 2.32; 95% CI 1.31 to 4.12). Using a similar methodology, Carús and Escorihuela50 54 have studied the risk factors for injuries to skiers and the feature-specific ski injuries in TPs. Their findings demonstrated that novice skill level and medium wind conditions are risk factors for severe ski injuries in TPs (compared with minor injuries). The analysed data on feature-specific ski injuries showed that, compared with boxes (which have the lowest odds of severe injuries), C-rails (OR 9.1; 95% CI 0.6 to 13.18), HPs (OR 4.5; 95% CI 0.3 to 6.27) and big jumps (OR 3.0; 95% CI 0.3 to 3.53) have increased odds of severe injuries versus minor injuries. Authors concluded that features requiring advanced techniques or promoting aerial manoeuvres or/and large drops to the ground have higher odds and rates of injuries. In parallel, other evidence shows that when jumps are removed from TPs, the odds of severe injuries decline (OR adjusted 0.72; 95% CI 0.54 to 0.97).55
Synthesis of the prevention strategies
There is no study examining the prevention strategies in TPs or HPs.
TP literature on the profile of the injured skiers-snowboarders, the risk factors for injuries and the prevention strategies
The primary aim of this study was to synthesise and highlight the current evidence regarding risk factors for injuries and injury prevention strategies in TPs and HPs. In summary, this systematic review identified no study examining prevention strategies in TPs or HPs and two articles on risk factors for injuries to skiers or snowboarders in TPs. As a secondary aim, we report the profile of injured skiers and snowboarders in TPs. This synthesis provides new information about the characteristics of injured participants in TPs. One of the original contributions of this manuscript is that it demonstrates that the results of the published studies strongly support the fact that skiing or snowboarding in a TP, instead of on regular slopes, is a risk factor for head, neck, back and severe injuries. This supports the need for research on injury prevention strategies in TPs and HPs.
Regarding the level of evidence,56 among all the current scientific literature in this field, only studies that can be more susceptible to bias have been published (case–control and cross-sectional). Despite this, the quality assessment using the modified NOS showed that these studies have a relatively good quality regarding their methods. In addition, the evaluation reveals four recurring weaknesses in the TP injury literature. Data on exposure were often missing. The number of injuries is believed to be underestimated as most studies use ski patrol reported injuries, but do not capture those injuries reported to the emergency department or medical clinics. Nevertheless, this current method is still one of the most valid and reliable to record injuries in ski areas.57 Moreover, non-response rates were rarely provided or poorly described for several studies. If the response rates are different for the comparison groups (cases/controls), it should be reported, as well as the potential reasons for the difference. The real reason may lead to a selection bias between groups owning to different intrinsic characteristics. Lastly, not all studies addressed the issue of confounding through design or analysis. Univariate analyses are common in the current literature leaving uncontrolled confounding a potential limitation.
Profile of the injured skiers-snowboarders in TPs
The most reported type of activity of participants injured in a TP or on a regular slope was snowboarding.6–8 22 47–49 However, the proportion of snowboarders uninjured in TPs is not reported in these studies. Generally, the results show that, among those injured, the odds of being a snowboarder was four times greater in TPs than on regular slopes. The risk of sustaining a traumatic brain injury is also greater to snowboarders.49 The intrinsic psychosocial characteristics of the participants involved in a specific activity can also influence the odds of injury. For instance, risk-taking behaviours of participants may be different between skiers and snowboarders.14 58 59 Moreover, admittedly, injury locations for these activities are different. Skiers are more affected by lower extremity injuries and snowboarders by upper extremity injuries.60
Among skiers-snowboarders injured, the odds of being a male was approximatively four to eight times greater in TPs than on regular slopes.7 8 22 Once again, no information was given on uninjured in TPs, but other papers in the literature agreed that males frequently used TPs.9 50 In ski slopes, the work of Paquette et al 58 61 shows that male skiers and snowboarders exhibit riskier behaviours. In fact, in many activities, males like to challenge themselves through risky behaviours more than females do.62–64 TPs can be seen as a social environment where small groups of friends are observed. Men have been observed to take more risks in such conditions.63
The studies exploring the skill level (expert, non-expert) among injured skiers-snowboarders provide evidence that the odds of being an expert (rather than a non-expert) are two to three times higher in TPs than in regular slopes.7 8 Expert skiers-snowboarders may expose themselves to features with higher injury rates that require a very precise technique (eg, big jumps, HPs, C-rails).9 50
Skiers-snowboarders injured in TPs were more likely to be wearing a helmet than skiers-snowboarders injured on regular slopes.7 8 22 This result is not really surprising because helmets are now mandatory in most TPs. It is also important to consider that helmet users have other characteristics (eg, man, young, snowboarder, expert) that could account for this finding.8 65 The TPs literature shows that skiers-snowboarders injured in TPs are often helmet users, but do not provide sufficient information about the protective effect of helmets. Strong evidence exists in the ski-snowboard literature indicating that helmets reduce the risk of head injuries.65
As reported in the literature, skiers-snowboarders injured in TPs were significantly younger than those injured on other slopes.7 8 22 Adolescents, who have a risk-taking reputation, are frequently using TPs.8 In this population, ‘the heightened risk taking during adolescence is likely to be normative and biologically driven’ by the slow maturation of the cognitive-control system.66 In regular slopes, younger participants also appear to express risk-taking behaviours and recklessness.58 64
Risk factors for injuries
As the results show, skiers and snowboarders are more likely to suffer from a head, neck or back injury when they use TPs compared with regular slopes. Furthermore, as shown by Goulet et al,6 Ruedl et al,51 and Stenroos and Handolin67 (excluded from the systematic review), injuries sustained in TPs of ski areas or urban environments are more likely to be severe. Newton’s laws of motion established more than 300 years ago can be applied to the study of TP injuries.68 In TPs, the height that users often achieve results in a significant transfer of force to the skier-snowboarder. These forces may be beyond the threshold for tissue damage (ie, injury). As stated by the International Ski Federation (FIS), future research aiming to study recreational ski-snowboard injury mechanisms might benefit from use of a video surveillance system and perform video analyses to understand more accurately the situation in TPs and HPs.69 70
According to our systematic review and eligibility criteria, studies assessing risk factors for injuries in TPs are rare. Two studies (grouping five papers)9 50 52–54 showed that self-rated expert skill level, listening to music (no), previous snowboard injury (no), temperature between −10°C and 0°C, snowboarding over night and features promoting aerial manoeuvres and/or large drops to the ground are risk factors for snowboard injuries in TPs and that novice skill level, medium wind conditions and features promoting aerial manoeuvres and/or large drops to the ground are risk factors for severe ski injuries in TPs. In parallel, Goulet et al 55 concluded that removing jumps from TPs at two ski areas reduced the proportion of severe injuries over a time period where no change was observed in the proportion of severe injuries at ski areas where jumps were not removed from TPs. As proposed by many authors, these findings support the need for defining standards for optimised design and maintenance of TPs and features to reduce injury risk.6 9 35 39 41 42 50 71 These kinds of engineering approaches show some promises as a potential prevention strategy to decrease the injury risk in TPs.35 39 41 42 71 72
This systematic review retrieved seven studies concluding that skiing or snowboarding in a TP is a risk factor for head, neck, back and severe injuries, and none refuted this. This finding is an original contribution to the field of skiing and snowboarding safety research. In parallel, we also demonstrated in this manuscript that no prevention strategy aiming to reduce injuries in TPs or HPs has been scientifically examined. There is an important gap.
An engineering approach considering TP or HP design and management might help prevent injuries, but there is no actual evidence on the effectiveness of any injury prevention strategies. Future research should focus on how to design and maintain a safer environment.
This systematic review is restricted by the insufficient literature and the limited study designs retrieved. Generally, the literature assessed with the NOS was made up of well-designed case–control and cross-sectional studies, though these designs can be more susceptible to bias.56 Furthermore, we observed weaknesses in methods used, where confounding factors were not addressed.
Nevertheless, this is a major finding raising an important problem for injury prevention in ski areas. In this manuscript, we clearly demonstrate the worrying situation of head/neck, back and severe injuries in TPs. The key message implied here is that there is a dearth of scientific literature on the injury prevention process36 73 and aetiology of injuries in TPs. We need to explore scientifically the association between TP or HP activities and the risk of injury to develop standards to guide future prevention strategies.
To our knowledge, this is the first systematic review focusing on injuries in TPs and HPs. As mentioned in the objectives of this manuscript, our first aim was to synthesise and highlight the current evidence regarding the risk factors and the injury prevention strategies for skiers and snowboarders in TPs and HPs. We found that only two studies assessed risk factors for injuries in TPs. Their results showed that self-rated expert, listening to music (no), previous snowboard injury (no), temperature between −10°C and 0°C, snowboarding over night and features promoting aerial manoeuvres or/and large drops to the ground are risk factors for snowboard injuries and that novice skill level, medium wind conditions and features promoting aerial manoeuvres or/and large drops to the ground are risk factors for severe ski injuries. In addition, we observed that the literature strongly agrees that skiing or snowboarding in a TP, instead of on regular slopes, is a risk factor for head, neck, back and severe injuries. Paradoxically, we found that no study examining prevention strategies in TPs or HPs has been published and that the risk factors for injuries in TPs and HPs are understudied. As a secondary aim, this study attempted to paint a picture of the profile of injured skiers-snowboarders in TPs and HPs. The results suggested some evidence of associations between factors including activity, sex, skill level, helmet use, age and TP injuries. The original contribution of this systematic review provides relevant information for stakeholders including skiers-snowboarders, patients, healthcare professionals, healthcare administrators and ski area administrators who want to be involved in the injury prevention process. An engineering approach aiming to secure the physical environment by developing standards to optimise design and maintenance of TPs and features might be an effective strategy to prevent injuries in ski areas. Future research should focus on how to design and maintain a safer environment.
What is already known about this subject?
Compared with injuries sustained on regular slopes, terrain park (TP) injuries are more likely to be severe.
Injured participants in TPs represent a homogeneous population of younger male snowboarders.
Interventions to reduce injuries in TPs and half-pipes (HPs) and risk factors for injuries in TPs and HPs are understudied.
Improvement of TP and HP design and management may prevent injuries.
What are the new findings?
Skiing or snowboarding in a TP, instead of on a regular slope, is a risk factor for head, neck, back and severe injuries.
No study examining prevention strategies to reduce injuries in TPs or HPs was found.
In TPs, the risk factors for injuries (all injuries vs no injury) to snowboarders are the type of feature, self-rated expert, listening to music (no), previous snowboard injury (no), temperature between −10°C and 0°C, and snowboarding over night and the risk factors for severe injuries (severe injuries vs minor injuries) to skiers are the type of feature, novice skill level and medium wind conditions.
There is some evidence of associations between factors including activity, sex, skill level, helmet use, age and TP injuries.
Supplementary file 1
Supplementary file 2
OA and TPM were funded through studentships provided by an Alberta Innovates Collaborative Research and Innovation Award held by CAE and BEH.
Contributors OA and CAE conducted the conception and design of the systematic review and the systematic review plan and registration (PROSPERO). OA and MDL built the search strategy used in databases. OA and TPM searched articles in databases and did the selection process. OA, CG and AM assessed the selection criteria. OA, BEH and CG contributed to the risk of bias assessment. OA, BEH, CG and AN-A were responsible for the interpretation of results. All authors contributed to the final systematic review plan (PROSPERO), reviewed the manuscript critically and gave final approval of the manuscript. OA was responsible for drafting the article.
Funding Funding for this systematic review of the literature comes from the ministère de l’Éducation et de l’Enseignement supérieur of Québec, Canada, and the Sports Injury Prevention Research Centre (SIPRC) of the University of Calgary, Alberta, Canada.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.