Background

Netball is a popular court-based team sport, played predominantly by females. The international governing body reports over 20 million participants across 117 nations spanning Africa, Americas, Asia, Europe and Oceania, with ongoing global growth [1]. However, netball’s intermittent, dynamic nature, involving repeated high-intensity sprints, jumps, landings, cuts and changes of direction [2,3,4,5], imposes considerable physical demands on players. These actions, combined with netball’s unique footwork rule, generate substantial forces [6,7,8] and player workloads [9,10,11,12,13]. Consequently, injury rates are high, ranging from 11.3–14 injuries/1000 player hours (h) at the community level [14,15,16,17], to elite rates from 54.8/1000 h at the 2019 Netball World Cup [18] up to 500.7/1000 h [19] in South African players. Hence, effective prevention strategies are crucial to support growing participation and minimise the negative impact of injuries at all levels.

Sports injury research, guided by van Mechelen et al.’s. ‘sequence of prevention’ [20] and the Translating Research into Injury Prevention Practice (TRIPP) [21] models, emphasises the importance of identifying the injury evidence base to inform prevention strategies. Hence, the initial crucial steps involve understanding the sport's injury problem through injury surveillance [22], followed by identifying the risk factors and mechanisms causing injuries [23, 24]. To ensure prevention strategies are effective, it is essential to collect accurate evidence using robust data collection methods. This requires the continual, systematic collection of high-quality data from injury surveillance systems across various settings [22], and a multifactorial approach to understand the complex interactions between multiple risk factors and injury mechanisms [23, 24].

Currently, there is limited review evidence describing the characteristics of methodologies used in netball injury research. Two recent netball reviews provide valuable synthesis of injury types, characteristics and risk factors, but only briefly address methodological considerations [25, 26]. Therefore, there is an urgent need for a comprehensive review of the methodologies used in netball injury research to establish the injury evidence base. Furthermore, while the recent consensus on netball video analysis framework [27] provides guidance for the assessment of injury mechanisms from match video, there is currently no consensus statement to inform injury surveillance methods in netball. Consequently, a scoping review of this area was considered appropriate to provide researchers with an overview of existing netball injury methodologies and to inform future research directions.

Therefore, the purpose of this scoping review is to evaluate the range and characteristics of methodologies used to describe 1) the incidence, severity and burden of injuries 2) the aetiology and mechanisms of injuries in netball. This information will be used to provide recommendations for future research to ensure the accuracy of the evidence base for targeted netball injury prevention.

Methods

Protocol

This review was conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-analyses extension for Scoping Reviews (PRISMA-ScR) and PRISMA 2020 updated statement [28, 29] (see Additional file 1 for PRISMA-ScR checklist).

Data sources and search strategy

A systematic, structured search strategy was developed with the assistance of a subject-specialist librarian. The electronic databases searched were SPORTDiscus, MEDLINE, CINAHL and Academic Search Complete (EBSCOhost), PubMed, Scopus and Web of Science from 1985 to 24th May 2024. The start date of 1985 was selected as Hopper (1986) [30] is recognised as the first peer-reviewed study on netball injuries [25]. The search terms used in all databases were “Netball*” AND “Injur*” AND (“incidence” OR “prevalence” OR “epidemiolog*” OR “risk*” OR “mechanism*” OR “cause*”). A secondary search of reference lists of included papers and Google Scholar was conducted to locate any additional studies eligible for inclusion.

Study selection

Following the removal of duplicates, the titles and abstracts were independently screened by two authors (SH, AFS) using the eligibility criteria. All articles that could not be excluded from this process were retrieved and underwent full-text screening. Where disagreements occurred, both authors met and discussed the studies until a consensus was gained. Hence, a third author was not required.

Eligibility Criteria

Eligible studies included those reporting data on netball injuries across all ages and levels of competition. These studies investigated the incidence, severity and burden, and/or the aetiology (risk factors) and mechanisms of netball injuries. Only studies published in English and peer-reviewed journals were included. Studies were excluded if they did not investigate netball, or they assessed the efficacy of prevention strategies, biomechanical factors in netball players un-related to injuries, or the physiological/movement demands of the game. Analytical studies that included netball athletes as part of a broader sports cohort but generalised findings across sports were also excluded e.g. Rigg et al. [31] and Almousa et al. [32]. Additionally, review articles, consensus statements, abstracts, and reports were excluded. All definitions of netball injuries were accepted. As outlined in the injury prevention literature [22, 24], aetiology is defined as the causes or risk factors that lead to injury. The injury mechanism is defined as the inciting event (playing situation and athlete behaviour) and biomechanical features resulting in injury [22].

Data extraction and analysis

Authors (SH, AFS) reviewed the included studies and discussed their categorisation, which was subsequently agreed by all authors. Studies were classified as descriptive epidemiological (describing the incidence and nature of netball injuries) or analytic epidemiological studies (identifying the association between specific risk factors and netball injuries or injury mechanisms), in a similar approach to Pluim et al. [33]. The descriptive epidemiological studies were further classified according to study design as studies using hospital/clinic records and insurance claim databases (hospital/clinic and insurance studies), or studies using injury data from netball competitions and/or historical injury data of match-play (netball competition studies). All studies were also classified by study design as prospective cohort, retrospective cohort or cross-sectional studies.

Data extraction from the included studies was conducted by the main author. Subsequently, the data from 14 studies (22%); descriptive epidemiology n = 10 (23%); analytic epidemiology n = 4 (19%), were verified by a second author (AFS). The data extracted included study details (author(s) and publication date), study design and data collection methods, data collection period, country of origin, population (including level, age and sample size), injury definitions and classifications, injury incidence and exposure, body regions, risk factors assessed and data analysis methods. Only those risk factors specifically related to netball injury data were included. The findings are summarised quantitatively with frequencies and percentages mapping the extent, nature, geographical distribution and range of methodologies in the studies.

Results

Study Selection

The database search yielded 655 studies, reduced to 199 following the removal of duplicates. After screening the titles and abstracts, 70 studies were identified for full-text screening. A further seven studies were identified through a secondary search of reference lists and 25 from Google Scholar, with 11 selected for full-text screening. Thus, a total of 81 studies received full-text screening. Subsequently, 65 studies were identified for inclusion in the review. A flowchart of the study selection process is shown in Fig. 1.

Fig.1
figure 1

Flowchart of scoping review selection process

Full size image

Review Findings

Tables 1, 2 and 3 provide a summary of the findings based on the study categories. Each table describes the study design and data collection methods, data collection period, country of origin, population, injury definitions, injury incidence and exposure methods and body region. The findings are also presented in graphical and tabular formats in Additional File 2.

Table 1 Methodological details and injury incidence of netball descriptive epidemiological studies using hospital records, clinic records and insurance claim databases
Full size table

Study Design

Of the 65 studies included in the review, 44 (68%) were descriptive epidemiological studies, while 21 (32%) were analytic epidemiological studies. The descriptive studies utilised injury data from hospital/clinic records and insurance databases in 25 studies (57%) (Table 1), while 19 studies (43%) collected data from netball competitions (Table 2). Most descriptive studies assessing hospital/clinic records and insurance data were retrospective in design (n = 20, 80%), while the netball competition studies more frequently utilised prospective study designs (n = 12, 63%) Similarly, most analytic studies (Table 3), were prospective in design (n = 11, 52%) with 7 (33%) using cross-sectional designs (Additional File, 2 Fig. 1). One analytic study reported both retrospective and prospective injury data [73], hence a total of 22 analytic epidemiology study designs are reported.

Table 2 Methodological details and injury incidence of netball descriptive epidemiological studies of competitions
Full size table
Table 3 Methodological details and results of netball analytic epidemiological studies
Full size table

Year of Publication

Eighteen descriptive epidemiology studies were conducted pre–2008 (41%), 14% of which reported data from pre–1998 [30, 34,35,36, 59, 60]. Post 2018, eight hospital/clinic record studies (18%) [51,52,53,54,55,56,57,58], and six (14%) netball competition studies [18, 68,69,70,71,72] have been conducted. The analytic research has increased considerably in the 15 years since 2008 (n = 15, 71%), with nearly half of these studies conducted since 2019 [17, 86,87,88,89,90,91]. Additional File 2, Table 1 presents the frequency of studies according to publication year. It is also important to note that all of the studies report injury data from a minimum of 1 year [34], up to a maximum of 16 years [36] prior to the publication date.

Country of Origin

Eight of the 77 netball countries affiliated to World Netball [1] have conducted injury research. Most studies were conducted in Australia (n = 32, 49%), New Zealand (n = 14, 22%) and South Africa (n = 11, 17%). Australian studies focused on descriptive studies of netball competitions [14, 15, 30, 59, 60, 62, 63, 67, 68, 71] and analytical studies [16, 17, 73,74,75,76,77, 81,82,83,84, 86]. In contrast, New Zealand largely utilised hospital/clinic and insurance data [35, 36, 38, 44, 46,47,48, 52, 54, 55]. South African studies assessed both descriptive studies of netball competitions [18, 19, 64, 66, 69, 70], and analytic studies [78, 80, 87, 90, 91]. Only four (6%) injury studies, comprising three hospital/clinic and insurance studies [39, 53, 57] and one analytic study [89], have been conducted in the UK, with no netball competition studies to date (Tables 1, 2 and 3, Additional File 2, Fig. 2).

Data Collection Period

A wide range of data collection periods were used across the netball studies with hospital/clinic or insurance data reporting the longest periods (Table 1). Most studies collected data for 4 years or more (n = 14, 56%) [36,37,38, 42, 44,45,46,47,48,49, 51, 52, 55, 58], or periods lasting 1 year (n = 9, 36%) [35, 36, 39,40,41, 43, 50, 54, 57], 2–3 years (12%) [35, 53, 56], or 10 months [34].

Descriptive netball competition studies collected data during netball seasons (n = 8, 42%), netball tournaments (n = 6, 32%) or over time periods (n = 5, 26%). The season data included studies assessing specific state or school leagues over one 14-week season [30], two seasons [60, 70], three 17-week seasons [71] and five 14-week seasons [59]. Other studies assessed injuries in players across one five-month season, two five-month seasons [14, 15] or one club during one 12-week season [67]. Studies assessing tournaments collected data for 3 days [62, 64], 4–6 days [19], 6 days [68] and 10 days [18], while those analysing time periods included 12 months [63, 66], 4 weeks of 1 season [61], one previous season [64], and 5 years [65].

The analytic studies similarly recorded injury data over seasons (n = 11, 52%), tournaments (n = 4, 19%) or time periods (n = 6, 29%). Season data assessed state leagues lasting one 14-week season [75, 84], injured players over one season [78, 79, 81, 83, 87, 90, 91] or two seasons [16] and one club over one season [17]. Other studies reported injury data from multi-day tournaments [73, 74, 76, 80], or time periods including the previous 12 months [85, 89], 4 years [86], 6 years 3 months [82] and 8.5 years [88]. One study collected data on all pervious injuries [77].

Study Populations

The populations investigated across the netball injury studies showed considerable variation. The hospital/clinic and insurance studies (Table 1) had the largest number of participants, ranging from 3 [40] to 11,757 [52], with 60% including > 100 participants [35,36,37,38, 43,44,45,46, 49,50,51,52,53,54, 58], and 40% > 1000 [37, 44,45,46, 49,50,51,52, 54, 58]. Most studies included a combination of children and adult age groups (60%) [35, 37, 38, 40, 41, 43,44,45,46,47, 49, 50, 53, 55, 58], with children typically categorised as under 15 years (y). A further seven (28%) studies analysed adults (15 y +) [34, 36, 42, 48, 51, 54, 56], while Hassan & Dorani [39] assessed children between 5–15 y.

The netball competition studies (Table 2) analysing season long competitions, included populations ranging from 37 [67] to 11,228 [59], with 56% (n = 5) < 300 participants [14, 15, 67, 70, 71]. The populations consisted of adults and children in four studies [14, 15, 30, 59], adults in three studies [60, 67, 71], while typical netball age categories; under 18, 19 and 21 were used by Sinclair et al. [70]. In studies assessing tournaments, populations ranged from 14 [72] to 1280 [19], with 50% < 200 [18, 62, 68, 69, 72]. Two studies analysed adults [18, 72], with four assessing a combination of adult and junior age categories (under15 to under 21 and senior level) [19, 62, 68, 69]. The four studies analysing time periods included populations ranging from 59 [65] to 1512 [61], with 50% > 1000 [61, 63]. Participants included junior [61], junior school [66], children and adult age groups [63] and under 16, 21 and senior age categories [65].

The analytic studies reported the smallest populations (Table 3). Those analysing seasons included cohorts ranging from 10 [91] to 368 [16] participants, of which 81% included populations of < 100 adult participants [16, 75, 78, 79, 81, 83, 84, 90, 91]. Tournament study populations ranged from 204 [74] to 1280 [80] participants, of which 75% had < 300, including under 16, under 21 and Open (adult) participants [73, 74, 76]. The six studies analysing time periods assessed populations of 16 [82] to 536 [86] athletes, typically < 200 (67%) [82, 85, 88, 89], two of which assessed the ACL injury mechanisms of elite athletes [82, 88].

Level of competition

The hospital/clinic and insurance studies (Table 1) mostly assessed the general population across all levels (68%) [37,38,39,40,41,42, 46,47,48,49,50, 52,53,54, 56,57,58] or netball populations across all levels (28%) [35, 36, 43,44,45, 51, 55], with one study investigating elite netball [34]. Specific competition levels or a combination of levels were more frequently analysed in netball competition and analytic studies (Tables 2 and 3). Studies analysing netball competitions assessed a combination of levels in six (32%) studies, reported as elite & sub-elite [19, 64, 65], elite & recreational [60] and recreational to competitive levels [30, 59]. Studies in this category also assessed players at the recreational/community (club) level [14, 15, 63, 67], junior and senior school level [69, 70], elite level [18, 71], sub-elite level [62], recreational junior level [61] and university level [72]. The analytic studies similarly assessed a combination of elite and sub-elite levels (29%) [73, 74, 76, 79, 84, 86], recreational/community (club) level [16, 17, 77, 89], university level [78, 90, 91] and elite level netballers [80, 82, 88]. Of the studies conducted at the elite level two analysed the Australia and New Zealand premiership (ANZ) [82, 88], one investigated the Netball World Cup [18], and one the Suncorp Super Netball competition [71].

Data Collection Methods

The methods of data collection in the hospital/clinic and insurance studies all involved diagnosis of injuries by medical professionals. In contrast, the netball competition and analytic studies used a wider range of data collection methods (Tables 2 and 3; Additional File 2, Fig. 4). Data was collected via player self-reporting of injuries in 47% of netball competition [14, 15, 62,63,64,65,66,67, 69] and 38% of the analytic studies [16, 74, 79, 81, 85, 87, 89, 91]. A combination of self-reporting and medical professional diagnosis also in combination with the coach/manager was used in 32% of netball competition [19, 30, 59,60,61, 68] and 43% of analytic studies [17, 73, 75, 77, 80, 83, 84, 86, 90]. Medical professionals, typically physiotherapists, diagnosed player injuries in 21% of netball competition studies [18, 70,71,72] and 19% of analytic studies [76, 78, 82, 88]. The data collection methods used in the netball competition and analytic studies were influenced by the level of competition, with medical professional diagnosis typically used at the elite level (80%) [18, 71, 82, 88] and self-report at the recreational/community level (75%) [14,15,16, 63, 67, 89].

Across the netball competition and analytic studies, only two netball injury studies captured longitudinal data of all injuries from ongoing and systematic injury surveillance systems. Toohey et al. [71] reports standardised injury data from a cohort of elite players in the Suncorp Netball Superleague, assessing 119 players from 8 teams across three seasons using the Australian Institute of Sport (AIS) customised Athlete Management System (AMS) database. Horgan et al. [86] also report 4 years of retrospective data from the same centralised database in a cohort of 536 elite and pre-elite athletes.

Body Regions

Most netball injury studies assessed injuries across all body regions (60%), shown in Tables 1, 2 and 3 and Additional File 2, Fig. 3. The most common specific body regions analysed were the knee and lower limb. Five (20%) hospital/clinic and insurance studies [42, 46, 49, 54, 56], and two analytic studies [82, 88] assessed the knee. Five (24%) analytic studies focused on the lower limb injuries [74, 79, 84, 91, 92], while 2 assessed lower limb and back injuries [73, 76]. Two further analytic studies assessed ankle injuries [81, 83]. The hospital/insurance data studies also assessed fractures across all body regions [39, 40, 57], dental injuries [38, 47], and Achilles Tendon injuries [48].

Injury Definitions

A wide range of injury definitions were used in the netball injury research (Tables 1, 2 and 3). Hospital/clinic or insurance studies used medical attention definitions in 44% of studies, referring to clinic or hospital attendance [34,35,36,37, 41, 44, 45, 50, 52, 53, 58], while 28% used medical attention definitions related to specific injuries; fractures [39, 40, 57]; ACL [42, 49, 56]; Achilles Tendon [48]. A further 28% included any complaint resulting in an insurance claim, in relation to all injuries [43, 51], dental injuries [38, 47] and ACL injuries [46, 54].

Netball competition studies used any or all complaints definitions in 58% of studies; five used any complaints that impaired performance [60, 61, 63, 64, 67], three any complaints leading to medical attention and time-loss [18, 70, 71] and three approved sports injury definitions [14, 15, 68]. Six studies (32%) used medical attention definitions [19, 30, 59, 62, 69, 72], two of which excluded minor injuries [30, 59] and time-loss from training or competition definitions was used in two studies [65, 66]. The analytic studies used all complaints definitions in six studies (38%) [17, 74, 78, 87, 90, 91] and medical attention and time-loss in two studies [16, 79]. Medical attention definitions were used in five (24%) studies [73, 75, 76, 80, 86], time-loss criteria in five studies (24%) [77, 83,84,85, 89], and definitions relating to specific injuries in three studies [81, 82, 88]. A small proportion of studies identified injuries as new or recurrent (n = 11, 17%). The term recurrent injury was mostly used and defined as the same injury as an index injury post recovery [16, 18, 19, 65, 80]. Subsequent injuries were defined by Toohey et al. [71] as any injury, following an initial injury in the time period.

Injury Severity and Burden

Injury severity definitions were reported in 40% of the hospital/clinic and insurance studies. Four studies used recognised injury severity scoring systems [35, 36, 39, 44], others reported the number or type of treatment [34, 57] and proxy measures based on the cost of injury [43, 52] or admission/length of stay in hospital [37, 41]. Fourteen (56%) of the netball competition studies reported injury severity, of which 50% used time-loss from participation definitions [18, 19, 60, 61, 68, 70, 71]. Other studies defined severity based on injury symptoms [59, 64], level of treatment [14, 15], treatment and time-loss combined [63] or pain ratings [66, 67]. Similarly, most analytic epidemiology studies reporting injury severity (38%) used time-loss definitions [78, 80, 85, 87] or specific injury scoring tools [81, 82] (Tables 1, 2 and 3). Severity ratings across the studies were typically based on grades or categories, either grades 1–3, or categories most commonly minor, moderate and severe. Only one study reported injury burden across the 65 included studies. Toohey et al. [71] defined burden as the product of mean severity and injury incidence.

Injury Classifications

Injuries were typically classified across the studies by body location or the location and type of injury, but recognised injury classification systems were only used in nine studies (14%). The International Classification of Diseases (ICD) [92] was used in six hospital/clinic or insurance studies [41, 44, 45, 49, 53, 58] and the Orchard Sports Injury Classification System (OSIICS) [93] was used in one hospital/clinic or insurance study [51] and two netball competition studies [68, 71]. Injuries were additionally classified by the mode of onset (traumatic or overuse) in two hospital hospital/clinic or insurance studies [34, 52], seven netball competition studies [18, 19, 62, 67, 68, 70, 71] and three analytic studies [73, 80, 87].

Injury Incidence rates

Tables 1, 2 and 3 show a small number of studies reported the total number of injuries only [52, 53, 72, 90], while others reported the proportion of injuries; 11 hospital/clinic and insurance studies [37, 39,40,41,42, 46,47,48, 54, 56, 57]; two netball competition studies [65, 66]; ten analytic studies [73–77, 79, 81, 82, 88,]. All other studies used a range of methods to report injury rates. The hospital/clinic and insurance studies typically used injury rates in relation to an actual or estimated population (n = 10, 40%); mostly including rates per 100,000 netball participants [35, 36, 44, 49, 50, 58] or 1000 participants [43, 45, 51, 55]. Netball competition studies mostly (47%) reported injury rates per 1000 player hours [14, 15, 18, 19, 61, 62, 68,69,70]. Other studies reported rates per 1000 [60, 62] or 10,000 players [63], per player per season [64], per 1000 players/match [30] or per 365 player days [71], while two used injury prevalence [18, 67]. Injury rates per 1000 player hours was the method reported in 29% of analytic studies [16, 17, 80, 83, 84, 87], while other methods included injuries per player [78], per 100 players per year [89] per player per year [83], daily probability [86] and injury prevalence [85].

Athlete Exposure

A variety of methods were used to calculate incidence rates based on athlete exposure hours (Tables 2 and 3). Studies mostly estimated match and/or training hours based on the average duration (hours) of playing and training in the time period [14, 15, 18, 62, 68, 69, 84, 87]. Only two studies calculated exposure based on individual match and training attendance records [17, 83]. Of the ten (53%) netball competition studies reporting athlete exposure, six used match exposure hours only [18, 19, 62, 68,69,70], with two combining match and training hours [14, 15]. Estimated individual exposure hours were determined in six studies [14, 15, 18, 62, 68, 70], while three calculated team hours [68, 69, 71]. The analytic studies (n = 6, 29%), utilised combined match and training hours in four studies [16, 17, 83, 84], and match hours in two [80, 87]. Individual exposure hours were used in three studies [16, 17, 80, 83] and team hours in two [84, 87]. Two further studies measured individual athlete exposure as the individual player match time in minutes before the injury occurred [19, 80].

Injury Mechanisms

The mechanism or event causing an injury was identified in seven (28%) hospital/clinic or insurance record studies, reporting injury events in categories including: overexertion, falls and collisions [35, 36, 39, 42, 44, 46, 48]. Eleven (58%) netball competition studies [18, 30, 59,60,61,62, 64,65,66, 68, 70] described injury mechanisms. The injury questionnaires used in these studies provided common injury cause options including: sharp twists/turns, falls, incorrect landing, collision with player, trip/slip, trodden on foot, sudden stopping, struct by player/ball, overexertion or other reasons. Hopper et al. [59] and Hume & Steele [62] provided further detail including the playing strategy (attack or defence), playing action e.g. intercepting, and movement e.g. shuffling, at the time of injury. Eight (38%) analytic studies reported mechanisms as part of their injury analysis [73, 75, 78, 80, 83, 86, 87, 90]. Three further studies had a specific injury mechanism focus, including Mullally et al. [89] who assessed injury situations in relation to previous injury. Two studies assessed injury mechanisms using systematic video analysis methods providing a comprehensive assessment of the events leading to ACL injury [82, 88]. These studies provided detailed descriptions of the game situation, player movement patterns, player behaviour and qualitative biomechanics of netball injuries to identify patterns in ACL injury causes.

Injury Risk Factors

The included studies have assessed a wide range of intrinsic and extrinsic risk factors and their association to injuries (Additional File 2, Table 2). The hospital/clinic or insurance studies assessed the smallest number of risk factors (median = 1, range 0–7 factors per study). The most common factors assessed were age (n = 12) [35, 39, 42,43,44,45, 48, 50,51,52,53, 55], gender (n = 8) [35, 40, 44, 45, 50, 52, 53, 58] and cost of injury (n = 3) [43, 50, 52]. Netball competition studies assessed a greater combination of risk factors (median = 4, range 0–11), with four studies analysing between 8 to 11 risk factors [30, 59, 65, 66]. The most frequent intrinsic factors assessed included age (n = 10) [14, 30, 62, 64,65,66,67,68,69,70], position (n = 8) [18, 19, 30, 64, 65, 69, 70, 72] and previous injury (n = 3) [15, 65, 68]. While the common extrinsic factors were weekly training (n = 8) [15, 30, 59, 60, 64, 68, 72], initial treatment required (n = 7) [15, 30, 59, 61, 64, 68, 72], training time (n = 6) [14, 30, 59, 65, 66, 68] and match quarter the injury occurred in (n = 6) [18, 19, 30, 59, 69, 70].

Commensurate with their purpose, the analytic studies assessed the widest range of risk factors (median = 6, range 3–15), with five studies assessing between 10 to 15 factors [17, 73, 75, 78, 87]. Table 3 and Additional File 2, Table 2 show the intrinsic factors most frequently analysed included age (n = 10) [17, 73, 75,76,77, 81, 83, 84, 87, 90], previous injury (n = 8) [16, 17, 73, 74, 77, 81, 83, 85, 87, 89], height (n = 8) and mass (n = 8) [17, 75, 78, 81, 83, 84, 87, 90], and playing position (n = 6) [76, 77, 82, 87,88,89]. A range of anatomical and biomechanical factors including limb dominance, postural stability, podiatric variables, ankle joint laxity and range of motion and lower body stiffness were assessed across 15 studies [17, 73,74,75, 77,78,79,80,81, 83,84,85, 87, 90, 91]. Physiological factors such as aerobic and anaerobic fitness, agility, strength, power, speed and flexibility, were additionally assessed across seven studies [75, 76, 78, 79, 83, 85, 87]. The extrinsic risk factors assessed included level of competition (n = 7) [17, 73, 75, 76, 81, 84, 85] and match quarter (n = 4) [73, 82, 88, 89] with a wide range of timing, training and treatment related factors also assessed across the 21 analytic studies.

Data Analysis Methods

The data analysis methods used across the studies included a range of descriptive and inferential statistics to describe the injury datasets (Fig. 2). Over 40% of the hospital/insurance records [34,35,36,37, 40, 41, 44, 46, 51, 58] and netball competition studies [19, 62,63,64,65, 69, 70, 72] reported descriptive statistics only. A small number of hospital/insurance record [38, 50, 55, 56] and netball competition studies [68, 71] reported odds ratios (injury probability), risk ratios (relative risk) or injury incidence rate ratios to describe differences between groups. Univariate inferential statistics were additionally used to assess the effect of various risk factors on injury in 60% of hospital/insurance record studies [38, 39, 42, 43, 45, 47,48,49,50, 52,53,54,55,56,57] and 53% of the netball competition studies [14, 15, 18, 30, 59,60,61, 66,67,68]. The chi-square test was the most frequent univariate test used in the descriptive studies (n = 19, 76%). Multivariate statistical tests were infrequent in these studies with only Fernando et al. [50] and Toohey et al. [71] using binary logistic regression models and generalised linear mixed models respectively.

Fig. 2
figure 2

Frequency of Netball injury studies by study design and data analysis methods

Full size image

Most analytic studies used inferential statistics to assess the effect of risk factors on injury (81%). Five studies used odds ratios [17, 77, 81, 84, 86], with risk ratios [84], absolute risk [86] and incidence rate ratios [16] also reported. Univariate statistics, including chi-square, t-tests, Mann–Whitney U tests, analysis of variance and univariate logistic regression, were used in 76% of studies [16, 17, 73,74,75,76,77,78,79, 81, 83,84,85, 87, 90, 91]. Five (24%) studies used multivariate tests, with all using multiple logistic regression models [17, 75, 77, 86, 87]. Adjustments for confounding variables was conducted in three studies [77, 86, 87]. The three studies with a focus on assessing injury mechanisms provided descriptive analysis only [82, 88, 89].

Discussion

This scoping review presents the first comprehensive overview of research methodologies used to determine injury incidence, aetiology, and mechanisms in netball. It complements the recent reviews of netball injury research by Downs et al. [25] and Whitehead et al. [26], highlighting methodological considerations aligned with the first two steps of the van Mechelen et al. [20] and TRIPP [21] injury prevention models. A total of 65 netball injury studies were included following screening, consisting of 44 descriptive epidemiological studies and 21 analytic epidemiological studies. The review highlights a scarcity of studies using systematic and ongoing injury surveillance, as well as limited methodological approaches to assess injury aetiology and mechanisms in netball. Without a specific consensus statement for netball to guide injury research, this review proposes potential future directions to enhance the quality of the netball injury evidence base.

The extent of the injury problem in netball is described in the 44 descriptive epidemiological studies and 19 (90%) of the analytic studies reporting injury data. However, 41% of descriptive studies and 29% of analytic studies were published between 1986 to 2008, with injury data collected an average of 3.6 years prior to publication. Furthermore, the majority of netball injury research has been conducted in Southern Hemisphere countries (88%), predominantly Australia (49%), and thus does not represent all netball-playing nations. Recent advancements in injury data collection methods [94], together with the growing professionalisation of netball with its increased physical demands [26, 95], and variations in playing styles across countries [1], emphasise the need for further research. This should encompass the diverse range of playing nations to fully understand the injury problem in line with the demands of the modern game.

The netball injury research has utilised various data sources, including hospital, clinic, and insurance databases (39%), as well as different competition formats, and specified time periods (descriptive epidemiology 29%; analytic epidemiology 32%). While hospital/clinic or insurance studies, utilise large populations and longitudinal data [51], they primarily capture severe injuries [22, 25], thereby underestimating injury incidence by neglecting milder cases. In contrast, data from netball competitions capture a broader range of injuries, providing a more accurate portrayal of the sport’s injury problems. Yet, studies vary considerably in observation periods, including short tournaments of 3–10 days (25%), league competitions over single or multiple seasons (50%), or specified time periods (25%). The lack of netball injury studies reporting longitudinal data from ongoing, systematic injury surveillance systems is a key finding of this study. Ekergen et al. [22] emphasised the need for such systems to provide high-quality data for effective injury prevention. However, only two (3%) netball studies report injury data from “true” injury surveillance systems [22]. Toohey et al. [71] collected injury data from a prospective cohort, in the elite Suncorp Netball Superleague over three consecutive seasons, using standardised methods [94]. While Horgan et al. [86] assessed retrospective data from the same centralised database (AMS), to assess the impact of risk factors on previously recorded injuries. The lack of comprehensive injury surveillance impacts the accuracy and reliability of the current netball injury research.

The current netball injury studies employed diverse methodologies to collect injury data, utilising prospective, retrospective and cross-sectional designs across the study categories. Study populations included a broad range of netball participants ranging from 3 [40] to 11,757 [52], with many including a combination of age-groups and participation levels, often lacking clear definitions. Indeed, Ferreira & Spamer [78] defined “elite” netballers as University first team players, while Janse van Rensburg [18] defined “elite” as those representing their country at the 2019 Netball World Cup. Injury diagnosis methods also differed, hospital/clinic or insurance studies using medical professionals, while competition studies used mostly medical staff at the elite level (80%) and self-report methods at the community/recreational level (75%).

Injury definitions varied across injury studies, with hospital/clinic or insurance studies mainly employing medical attention definitions (72%), while competition and analytic studies used a broader range, including all complaints (51%), medical attention (30%) and time-loss definitions (19%). Definitions of injury severity also varied, incorporating time-loss, treatment, symptom, hospital attendance and cost of injury criteria. To date, Toohey et al. [71] is the only study to report injury burden, a critical measure that combines injury frequency with its severity (typically measured in days lost) [94]. This metric allows for the identification of not only the most common injuries but also those that impose the greatest impact [96]. This understanding is vital for comprehensively assessing the repercussions of injuries within netball. Furthermore, only a small number of studies defined recurrent injuries, (14%) or used a recognized classification system for injuries (14%).

The variations in study design and data collection methods make it difficult to compare netball injury studies, and differentiate injury risks within defined populations. The methodological issues subsequently impact the reported incidence rates in the current netball injury research. Moreover, the different metrics for calculating injury incidence further confuse the extent of the injury problem. Although more recent competition studies [14, 15, 18, 19, 61, 62, 68,69,70] and analytic studies [16, 17, 80, 83, 84, 87] report injuries in relation to athlete exposure hours, differences in exposure calculation methods, including using match hours only, combining match and training hours, and using average team or individual hours have also impacted the reported incidence rates. This has led to incidence rates ranging from 11.3 to 89.4 injuries/1000 player hours (Table 1, 2 and 3). Additionally, two further studies [19, 80] calculated player exposure based on game time in minutes prior to injury rather than total exposure time over the study period. This different approach to calculating athlete exposure resulted in a very high injury incidence rate of 500.7 injuries per 1000 h.

To develop a clear understanding of the injury problem [20, 21], robust injury surveillance systems are crucial for netball to ensure accurate data informs the evidence base. The England Rugby Football Union (RFU), has effectively implemented such systems across elite men’s and women’s levels (PRISP and WRISP projects), community level (CRISP project) and university level (BUCS ISP project) [97] providing an effective model for netball. Currently, no netball injury research has assessed the UK Netball Superleague, or New Zealand ANZ Premiership, and only one study assesses the Australian Suncorp Super Netball. Therefore, future research should focus on the development of robust surveillance systems to provide consistent injury data to analyse all competitions at the elite level. Furthermore, there is a need to develop tailored surveillance systems for all levels of the game.

This study recommends adopting the standardised methods of data collection in the International Olympic Committee (IOC) consensus statement [94] to ensure consistent surveillance methods. This updates the recommendations of Downs et al. [25], who endorsed the rugby union consensus statement [98]. The guidelines include consistent use of either all complaints, medical attention or time-loss injury definitions, and time-loss severity definitions, depending on the study focus. They suggest using measures of injury burden that combine frequency and consequences, typically injury incidence multiplied by severity (time-loss days). Recommendations for classifying injuries are provided using consistent coding systems such as the Orchard Sports Injury Classification System (OSIICS) [93]. Furthermore, to standardise the reporting of injury rates, the IOC statement recommends recording individual player exposure hours and expressing injury incidence rates per 1000 athlete exposure hours for sudden-onset injuries. For gradual-onset conditions, it suggests reporting prevalence as the proportion of injured athletes [94].

In addition to the IOC guidelines, this study advises incorporating netball-specific demographic categories to define study populations. Age categories such as Senior/Adult, under 21, under 19, under 17, and Junior levels such as Under 16, Under 15, Under 14 are universally used across nations in international, national, and school-level competitions, providing a consistent framework. Inclusion of age mean and range will further describe the age distribution within each category. To describe level of play we recommend classifying netball populations according to Mckay et al.’s. [99] skill level and training status framework. Participants are categorised using the criteria of Tier 0–4: Sedentary, Recreationally Active; Trained/Developmental; Highly Trained/National Level, Elite/International Level. In this framework Elite/International netball competitions would include all International competitions and elite leagues including the UK Netball Superleague, Suncorp Super Netball in Australia and ANZ Premiership in New Zealand. The consistent reporting of injuries using these categories would provide greater clarity regarding the injury issues across age groups and playing levels. A summary of guidelines to identify the injury problem, adapted for netball, are provided in Fig. 3.

Fig. 3
figure 3

Netball injury research methodological recommendations

Full size image

The current research assessing injury aetiology and mechanisms in netball has notable limitations. Twenty-one analytical studies aimed to identify the factors causing injury, while a further 34 descriptive studies investigated isolated factors related to injury. Collectively, these studies have assessed a wide range of intrinsic and extrinsic risk factors, but typically only a small combination of factors within each study. Specifically, the analytic studies analysed a median of 6 risk factors across the 21 papers. Furthermore, most studies employed a reductionist approach, simplifying factors into units in a linear, unidirectional way. This approach is thought to restrict understanding of injury causes, particularly where interactions between multiple factors may determine injury potential [24, 100]. Only 11% of the netball studies used multivariate statistics to assess the impact of a range of risk factors on injury, and even these approaches are suggested to be insufficient to identify the complex interactions between multiple risk factors [100].

The mechanisms of injury, or inciting event leading to an injury, has been identified in a number of netball injury studies using a variety of methods. Some studies report the mode of onset as acute or overuse and/or classify the injury mechanism as contact/non-contact. A greater number of studies (45%) describe the injury inciting event, typically through athlete self-report or medical staff report, using pre-determined categories to guide the responses. This approach has provided some valuable information, but it provides only a simplistic description of the injury event and is often limited in accuracy, as it relies on biased retrospective recall [101]. Thus, the understanding of injury inciting events in netball requires further investigation. Thus far, only two studies have conducted a more comprehensive assessment of netball injuries using video-based methods to accurately describe the inciting event. Stuelcken et al. [82] and Belcher et al. [88] assessed the mechanisms of ACL injuries, providing a full description of the playing situation, movement patterns and player behaviour at the time of injury. However, no research to-date has developed video-based methods to assess a wider range of injuries and their causes in netball.

To better understand the aetiology and mechanisms of injury in the second step of the sequence of injury prevention [20, 21], aetiology research should employ a multifactorial approach. This should assess the complex interaction between multiple intrinsic and extrinsic factors, workload and the injury inciting event [24, 102]. Hence, studies need to make use of a dynamic model which describes the interaction between as many risk factors as possible, appropriate workload measures and the events leading to the injury. The multifactorial model additionally needs to account for the dynamic, recursive nature of sports injury. Such models include Windt & Gabbett’s [102] workload-injury aetiology model, developed from the original multifactorial models of Meeuwisse and colleagues [103, 104]. Accurate assessment of netball injury mechanisms, to inform the injury model, require a consistent approach. The development of video-based methods that fully describe the playing situation, player/opponent behaviour and accurately assess the biomechanics of injury are necessary to provide a complete assessment of the injury inciting event. Combining these video methods, where possible, with athlete and medical staff descriptions is recommended to provide a more comprehensive understanding of injury causality [23, 101]. To facilitate clear comparisons between studies, the definitions and terminology recommended in the recent consensus on netball video analysis framework [105] should also be adopted.

Finally, to analyse the non-linear interactions between these injury determinants a complex systems approach has been suggested by Bittencourt et al. [100] to be a more appropriate method of assessing sport injuries. The method identifies a risk profile from the interactions between the “web” of injury determinants. Appropriate statistical methods are necessary to identify injury predictions rather than relationships. These methods include recursive partitioning-based methods e.g. classification and regressions trees (CART) and random forests, or machine/statistical learning methods [100]. Figure 3 summarises the recommendations for netball injury aetiology and mechanism research methodologies. Future research should address these methodological concerns to provide an accurate netball injury evidence base which is critical to inform the development of targeted injury prevention strategies. This study provides a comprehensive summary of the research methodologies describing the extent of the injury problem and aetiology and mechanisms of injuries in netball. However, it is possible the search may not have identified all studies in the area.

Conclusion

This scoping review reveals a lack of systematic and ongoing injury surveillance systems in the netball injury research describing the injury problem. Studies exhibit considerable heterogeneity in methodologies, including study designs, injury definitions, data collection methods and injury reporting practices. Inconsistent methods of reporting injury rates and classification of study populations further limit the quality of evidence across different age groups and level of play. Research assessing injury aetiology often focuses on a limited number of risk factors, using reductionist approaches, while studies assessing injury mechanisms use simplistic descriptions, based on unreliable retrospective recall. Therefore, additional research is needed to comprehensively assess the netball injury problem, its causes, and mechanisms within the modern game, considering a broader spectrum of playing styles.

Accurately identifying key injury issues in netball, requires reliable and consistent injury surveillance systems across settings. The IOC consensus statement guidelines are recommended for the accurate collection of injury data, providing clear definitions, collection methods and reporting protocols. To understand the causes of netball injuries, a multifactorial approach is essential to assess the complex interaction between multiple intrinsic and extrinsic factors, player load and the injury inciting event. Detailed assessment of the inciting event should encompass the playing situation, player/opponent behaviour, and joint and whole-body biomechanics utilising video analysis and medical staff descriptions.