Introduction
Mass gatherings such as soccer games, pageants and concerts occur frequently throughout the world. Commonly, mass gatherings impact health services such as onsite health services, ambulance and prehospital emergency medical services, hospital emergency departments, and acute medical care services including operating theatres. Furthermore, mass gatherings are important sites to research health behaviors because they help researchers to understand how to manage large numbers of people in temporary environments.
Throughout the literature, the term “mass gathering” is defined in many ways. Commonly, the defining factor of a mass gathering is related to the number of attendees at an event, such as an event with >25,000 attendees. However, on closer examination, a mass gathering seems more complex than this. An alternative and perhaps more appropriate definition of a mass gathering is: “a situation (event) during which crowds gather and where there is the potential for a delayed response to emergencies because of limited access to patients or other features of the environment and location. This potential delay requires planning and preparation to limit (or mitigate) the hazards inherent in a mass gathering and ensure timely access to appropriate health care is available.”Reference Arbon1 Throughout this discussion, the term attendee will be used to describe a spectator or participant of an event.
According to Arbon,Reference Arbon2 it has been suggested that there are three distinct domains which influence the health service presentation of patients at mass gatherings: environmental, psychosocial and biomedical. The environmental domain includes factors such as the nature of the event, availability of drugs or alcohol, venue characteristics and meteorological factors. The psychosocial domain includes the crowd mood and behavior, crowd culture, and reason for attendance. The biomedical domain includes factors such as demographics and health status of spectators and participants.
This paper aims to initiate international discussion of the need for consistency in the reporting of data from mass gatherings, while acknowledging that meaningful data collection and reporting across societies and mass gatherings needs to be flexible. The current situation of data collection and reporting is presented, along with a possible minimum data set, and future opportunities are outlined.
Current Situation
When examining the biomedical domain of the mass-gathering literature, the focus is on categorizing presenting injury or illness, reporting patient presentation rates (PPRs) or medical usage rates (MURs), and exploring other factors, such as transportation to hospital rates (TTHRs).
Injury/Illness Categorization
In the earlier mass-gathering literature, authors commonly listed a breakdown of specific types of injuries and illness of patients who presented to health services at mass gathering sites.Reference Milsten, Maguire, Bissell and Seaman3, Reference Zeitz, Zeitz, Arbon, Cheney, Johnston and Hennekam4 For example, Rose et alReference Rose, Laird, Prescott and Kuhns5 reviewed data from six and a half years of patient presentations at college football games in the United States. However, the authors did not commonly make reference to the origin of these lists of types of injuries and illness; therefore, this patient presentation method is author dependent, and cannot be generalized to other mass-gathering events.
Another concern when presenting specific levels of data is that some categories may have larger counts than others. For example, in the comparison of injuries and illnesses from US football, baseball, and rock concerts, a large amount of presentations (69%) are termed as “medical related” with no further explanation.Reference Milsten, Maguire, Bissell and Seaman3 By categorizing presentations as “medical related,” the types of presentations are not well defined.
In addition, the 2002 FIFA World Cup data illustrated that “other” and “unrecorded” accounted for 24.9% and 21.7% of the total presentations respectively.Reference Morimura, Katsumi and Koido6 This unspecified data highlights how having large counts in categories such as “other” limits the insight gained at an event. While reporting at a specific category level, some reduce the number of counts in an “other” category to represent less than one percent.Reference Zeitz, Zeitz, Arbon, Cheney, Johnston and Hennekam4 This strategy is more useful to determine the true types of presentations.
To describe the severity of injury and illness, some authors report patient presentations in a broader, nonspecific manner. This may include categories such as minor, intermediate, and major.Reference Olapade-Olaopa, Alonge and Amanor-Boadu7 Alternative categories have included “basic-level,” “advanced-level,” and “life-threatening level.”Reference Yazawa, Kamijo, Sakai, Ohashi and Owa8 When these broad categories are used, the authors either provide descriptors,Reference Olapade-Olaopa, Alonge and Amanor-Boadu7 or examplesReference Yazawa, Kamijo, Sakai, Ohashi and Owa8 of the types of injuries and illnesses included in each category. On occasion, it may be worthwhile to have a specific breakdown of injuries and illness which cannot be articulated from nonspecific levels of categorization. Broad illness and injury categories can be determined from reports of specific levels of categorization.
Patient Presentation Rates
Within the literature, terms such as “patient presentation rate” (PPR) and “medical usage rate” (MUR) are used interchangeably. These are crude ratesReference Pagano and Gauvreau9 that refer to the number of attendees who present to onsite health services, in comparison to the overall number of attendees.
$$\bi PPR = \frac{\bi {Attendees \ who \ present \ to \ the \ onsite \ health \ service}}{{Total \ number \ of \ attendees \ at \ the \ event}}\eqno\rm$$The PPR provides insight into the onsite health service usage. However, PPR does not always reflect the acuity of individual patients, which may influence the onsite health service requirements. Additionally, event duration may be an important factor not explicitly considered in PPR, as PPR may vary over hours, days or weeks. In the literature, PPRs are presented as either raw numbers, or as presentations per 100, 1,000 or 10,000 attendees, with no consideration of the length of the event.
Raw numbers are used on occasion to highlight the number of patient presentations.Reference Feldman, Lukins, Verbeek, MacDonald, Burgess and Schwartz10, Reference Zeitz, Zeitz and Arbon11 During the early development of mass-gathering research and evaluation, authors reported PPR per 100 attendees.Reference Hnatow and Gordon12 Following the 2002 FIFA World Cup, PPR was reported per 1,000 attendees.Reference Morimura, Katsumi and Koido6 This trend is similar to others who report per 1,000 attendees.Reference Zeitz, Zeitz, Arbon, Cheney, Johnston and Hennekam4 In contrast, some have reported PPR as presentation per 10,000 attendees. This paper encourages the standardizing of PPR as presentations to onsite health services per 1,000 attendees for generalizability across all events.
$$\bi PPR \ = \ \frac{{\displaystyle{Attendees \ who \ present \ to} \atop \displaystyle{the \ onsite \ health \ service}}}{{Total \ number \ of \ attendees \ at \ the \ event}} \ \times \bf 1,\!000\eqno\rm$$Other Rates
Transport to hospital rate (TTHR) provides insight into the prehospital ambulance or emergency medical service usage. In the literature, TTHR has been reported as a percentage,Reference Hiltunen, Kuisma and Maataa13 as presentations per 1,000 attendees,Reference Zeitz, Zeitz, Arbon, Cheney, Johnston and Hennekam4 or as presentations per 10,000 attendees.Reference Hiltunen, Kuisma and Maataa13 As variability exists in reporting TTHR, this paper encourages the standardized reporting of TTHR as presentations to onsite health services per 1,000 attendees.
$$\bi TTHR \ = \ \frac{{\displaystyle{Attendees \ who \ are \ transported} \atop \displaystyle{to \ hospital \ by \ ambulance}}}{{Total \ number \ of \ attendees \ at \ the \ event}} \ \times \bf 1,\!000\eqno\rm$$A rate that has not been widely reported in the literature is the referral to hospital rate (RTHR). This rate includes patients who are transported to hospital (TTHR). Additionally, it includes patients who are referred to hospital and do not travel by ambulance. This rate gives some insight into the usage rate of hospital emergency departments in the vicinity of the mass gathering and the value of onsite care in regards to hospital avoidance.
$$\bi RTHR \ = \ \frac{{\displaystyle{Attendees \ who \ are \ referred} \atop \displaystyle{to \ hospital \ by \ all \ means}}}{{Total \ number \ of \ attendees \ at \ the \ event}} \ \times \bf 1,\!000\eqno\rm$$Other Data Collection
In addition to categorizing injury and illness and highlighting various rates, some authors report on patient demographics.Reference Milsten, Maguire, Bissell and Seaman3 This data provides additional insight into the “type” of patients at mass gatherings. Some authors report on the level of care, making comparisons of onsite health resources, such as number of medical officers, nurses, paramedics, volunteers, and ambulances compared to the number and type of patients treated.Reference Nguyen, Milsten and Cushman14 Additionally, some authors include patient disposition, such as return to the event or transported to hospital.Reference Milsten, Maguire, Bissell and Seaman3
Minimum Data Set
In collecting biomedical data from mass gatherings, there may be an agreed-upon minimum data set.15 A minimum data set is a tool that can be used to collect de-identified patient-level information for the purpose of making comparisons across societies and individual mass gatherings. Introduction of a minimum biomedical data set for mass-gathering evaluation and research is proposed. The proposed minimum data set (Table 1) was developed based on injury and illness categorizations of: (1) published authors in the mass-gathering literature; (2) the “injury surveillance national minimum data set” from the Australian Institute for Health and Welfare;15 (3) the “event and emergency first aid minimum data set” from St John Ambulance Australia;16 and (4) the authors’ experience of undertaking research and evaluation and as practicing clinicians at mass gatherings.
Table 1 Patient Data Set and Entry Codes
In addition to presenting a minimum data set, a data dictionary with associated descriptors relating to data entry codes (Table 2) is provided to assist in differentiating among the various categories and to assist in providing consistency in reporting.
Table 2 Data Dictionary to Supplement Minimum Data Set and Entry Codes
An example of a data collection tool and data entry using Microsoft Excel 2010 (Microsoft Corp., Redmond, Washington USA) is shown in Figure 1. This data includes the minimum data set in Table 1 and categories from the data dictionary in Table 2.
Figure 1 Example of Data Collection Tool Abbreviations: Enviro, environmental; Inj Loc, injury location; MH, mental health; Out, outcome/disposition; Rx Dur, duration of treatment
Future Opportunities
Currently, mass-gathering data is collected and held by individual persons or organizations undertaking research and evaluation. To enhance understanding of the complexities of mass gatherings, there is a need for consistent collection of data by individuals and organizations. Having consistent data will provide the possibility for meta-analysis of events, and comparison of similar events within different societies or the comparison of a single event over time. In addition a consistent data set would better inform health services about their possible involvement and requirements at mass gatherings and inform event managers about health risks and implications of their event.
Retrospective review of mass gathering data has been proposed as an accurate predictor of PPR and TTHR.Reference Zeitz, Zeitz and Arbon11, Reference Zeitz, Schneider, Jarrett and Zeitz17 However, as retrospective information about a future mass gathering is not available, being able to compare similar mass gatherings in different societies, or different mass gatherings, may be sufficient to gain some insight into the likely PPR, TTHR and RTHR. Statistical analysis using odds ratios or chi-squareReference Pagano and Gauvreau9 may be a first step in gaining a better understanding of some of the variances among societies and mass gatherings.
Modeling to predict health service requirements at Australian mass gatherings has been published.Reference Arbon, Bridgewater and Smith18 However, these models are limited, as they were generated from Australian populations with >25,000 attendees per mass gathering, and were developed more than a decade ago. Predicting and modeling health service resources is important for health workforce strategies at mass gatherings. In predicting health resources at a mass gathering, the PPR, TTHR, and RTHR would be considered the outcome (dependent) variables. Explanatory (independent) variables from the biomedical, psychosocial and environmental domains should be included in any modeling. With a consistent data set, it can be argued that predictive modeling would more closely forecast the realities of a mass gathering.Reference Pagano and Gauvreau9
A minimum biomedical data set and agreed method of reporting rates and outcomes associated with mass gatherings will allow for the retrospective comparison of events and prospective predictive modeling of events. The information derived from retrospective comparison and predictive modeling can aid in mass-gathering medical services planning. Highlighted above are some possible approaches to data analysis. The specific details of a possible data analysis plan and possible data analyses are not the focus of this paper; however, they should be taken into consideration in any overarching conversation about consistency in a minimum data set.
Conclusion
This paper has highlighted the research and evaluation of the biomedical domain of mass gatherings as being varied, haphazard and author dependent, particularly in terms of data collection and reporting. This is illustrated in terms of the various data collection and reporting of patient categories, rates and other biomedical-related information. It is proposed that a minimum data set and data dictionary be developed to begin discussion of the need for consistency in collecting and reporting data. Moving to a more expert consensus approach, and beyond a haphazard, author-dependent approach, will allow development of mass-gathering health service research and theory.
Abbreviations
- MUR:
medical usage rate
- PPR:
patient presentation rate
- TTHR:
transportation to hospital rate
