The present work consists of a systematic review of observational studies, including prospective, retrospective and cross-sectional investigations. For this purpose, the recommendations of the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) standards were followed.12 This approach attempted to provide a complete epidemiological profile of injuries in soccer referees, based on the analysis of a large number of subjects. The use of this strategy was justified considering that existing primary studies tend to have relatively small samples. Thus, conducting a review covering a larger number of subjects allowed to obtain a broader and more representative view of the situation. Furthermore, to date there is no knowledge of the existence of such a review, which underscores the relevance and originality of this study. This review, following its prior registration in PROSPERO, has been assigned the number CRD42024547881.

A systematic search of observational studies with a cut-off date of May 24, 2024, was conducted using the databases PubMed, Web of Science and ScienceDirect. For this purpose, the following search terms were used: Epidemiology, Epidemiologic, Epidemiological, Incidence, Prevalence, Injury, Injuries, Injured, Soccer and Soccer referees. These terms were combined in the following search equation: ((((((Epidemiology) OR (Epidemiologic)) OR (Epidemiological)) OR (Incidence)) OR (Prevalence)) AND (((Injury) OR (Injuries)) OR (Injured))) AND (("Soccer referees") OR (Soccer)). In the complementary material, the search history in each of the databases used, together with the number of records identified, is provided in detail.

The text selection process for this review was carried out in a rigorous and systematic manner. First, preliminary manuscripts obtained during the searches in each database were uploaded to the collaborative web application Rayyan.13 Then, two reviewers carried out an independent selection after analyzing the title and abstract of each text found. Subsequently, the reviewers carried out a second independent selection process after reading the full texts that had been selected in the first phase. During these two stages of selection, in the event of discrepancies between the reviewers, a third reviewer was involved to facilitate reaching a consensus and reaching an agreement.

Once the texts included in this review had been selected, the following data were extracted from each of them: title, authors, year of publication, journal of publication, type of study, follow-up period, injury definition, population, total number of injuries, and number of injuries by anatomic location or diagnosis, as applicable. These data were extracted by two reviewers independently and stored in an Excel spreadsheet. Subsequently, the extracted data were cross-checked to verify the accuracy of the extraction. In case of discrepancies, a third reviewer intervened to verify the information and reach a consensus on the final data.

The quality of the manuscripts included in this review was assessed using the STROBE checklist,14 which provides predefined criteria to classify the quality of manuscripts according to their presentation and reporting. It is important to note that STROBE does not constitute a quality scale of the evidence itself, but rather focuses on the quality of the study report. Therefore, to complement the assessment of study quality, we used the checklist for cohort studies provided by SIGN (Scottish Intercollegiate Guidelines Network) .15 This list contributed to the evaluation of the quality and methodological rigor of each of the selected studies. The application of these scales to the selected texts in this review was carried out by two investigators. In case of disagreements, a third reviewer intervened to facilitate consensus.

The results obtained in this study are presented quantitatively using descriptive statistics through tables and figures, which provide a visual representation of the data collected. These tables and figures show the prevalence and incidence of injuries in soccer referees, discriminated by anatomical location or specific diagnosis, as applicable. In addition, the results are presented qualitatively through a narrative that describes and analyzes the findings obtained. This qualitative narrative provides a detailed interpretation of the results, highlighting trends and patterns observed in the data.

To obtain the prevalence of injuries, retrospective and cross-sectional studies were used, while prospective studies were used exclusively to obtain the incidence of injuries. As previously mentioned, data were collected on the location and diagnosis of injuries, which were subsequently analyzed and operationalized according to the temporal period of analysis to obtain measures of prevalence and incidence. The ratio of injury incidence and prevalence data was calculated by dividing the total number of injuries per anatomic region or specific pathology by the total number of injuries reported in all anatomic locations or pathologies, respectively.

In the systematic search, 12,246 manuscripts were identified, of which 2460 were from PubMed, 2790 from Web of Science, and 6996 from ScienceDirect. This was reduced to 10,479 after eliminating duplicates. Once the titles and abstracts had been reviewed, 12 articles were selected for full review.9,16-26 Following this review, 6 articles presenting combined data for main and assistant referees were excluded,16,18-22 and one more, since it did not provide specific data for soccer referees,17 leaving a total of 5 articles for the final selection and synthesis.9,23-26Fig. 1 provides the flow chart summarizing the search and item selection process.

Fig. 1.

PRISMA flow chart suggested by Page et al.27

(0.48MB).

An analysis was carried out covering a total of 433 main referees and 467 assistant referees. Of these, 763 were examined to determine the prevalence of injuries, distributed in 382 main referees and 381 assistant referees, while 137 were studied to analyze the incidence of injuries, of which 51 were main referees and 86 were assistant referees. The age of the referees ranged from 25 to 45 years old.

In terms of the methodology used, three retrospective studies were identified that contributed to the prevalence analysis [9, 25 y 26], a prospective study used for incidence analysis,23 and another study that presented both prospective and retrospective characteristics, thus allowing the analysis of both the incidence and prevalence of injuries.24

A noteworthy aspect of this analysis was the consistency in the definition of injury used among the studies. Although some of them adopted literal definitions previously used in research on soccer players, others made slight adaptations, replacing terms such as "soccer players" or "footballers" with "referees". However, all studies agreed to consider an injury as any physical ailment resulting from soccer matches or training sessions related to refereeing, regardless of the need for medical attention or interruption of soccer activities. For a more detailed understanding of the characteristics of the included studies, a summary is presented in Table 1.

After performing an assessment of the degree of compliance of each study's report using the STROBE list, it was observed that the main weakness among the articles reviewed was related to several aspects. These included population eligibility criteria, sample size calculation, handling of missing data, dealing with loss to follow-up, sensitivity analysis, and notably, the limited use of flowcharts to facilitate understanding of the flow of subjects throughout the study, except for one study that employed this tool.

Regarding quality assessment, upon completion of the cohort study checklist provided by SIGN, one study was identified as being of low quality, while four of them were considered to be of acceptable quality. Among the aspects that presented major deficiencies were the assessment of the probability that some eligible subjects may present the outcome of interest at enrollment, the blinding or lack of recognition of blinding in those cases where it was impossible, and the identification and control of potential confounders. A summary of the reporting grade and quality of the studies included in this review is provided in Tables 2 and 3.

The most prevalent injury by anatomical location for main referees, according to the last match, was injuries in the Achilles tendon and ankle/foot, while for assistant referees was in the lower leg. During the physical tests, the most common injuries for the main referees were in the knee, hip and groin, and in the hip and groin for the assistant referees. Throughout their career, the most common injuries were in the hip and groin for the main referees, and in the knee for the assistant referees. In the last twelve months, the most common injury was in the thigh for main referees, and again in the knee for assistant referees. Table 4 provides a detailed summary of the prevalence of injuries, categorized by anatomical location, for the main referees. Similarly, Table 5 presents this information for the assistant referees.

On the other hand, in the last match mediated by the main referees, tendon strain was identified as the most prevalent diagnosis, while during the physical tests, "muscle contracture" was the most reported. Throughout their careers as referees, ankle sprains were the most prevalent diagnosis. Finally, during a 12-month follow-up, thigh muscle injuries and tendon strains again stood out as the most prevalent.

In the case of assistant referees, "muscle contracture" emerged as the most common diagnosis both in the last match refereed and during physical tests and 12-month follow-up. On the other hand, hamstring strains were established as the most prevalent throughout their career. An exhaustive analysis of the prevalence of injuries discriminated by diagnosis for main and assistant referees is detailed in Tables 6 and 7 respectively.

Regarding the incidence of injuries categorized by anatomical location, lower leg and Achilles tendon injuries were together reported as the most incident over a 12-month period for both main and assistant referees. Followed by these, the most frequent injuries were ankle/foot injuries in both types of referees. For a 30-day follow-up period, the most frequent injury was in the lower back for both groups. Table 8 presents the incidence data by anatomical location for both groups of referees.

Unfortunately, categorization and operationalization of incidence data by diagnosis was not feasible in this context. This limitation arises due to the size of the study population for this category of analysis, which has resulted in all diagnoses receiving an equal incidence percentage.