Introduction
In recent decades the need for more consistent emergency medical services at mass gatherings has been identified. The presence of a medical team—able to provide rapid assessment of patients, including triage, treatment, stabilization and transportation of those seriously injured to off-site medical facilities—is increasingly discussed in the literature.Reference De Lorenzo1-Reference Hnatow and Gordon3 The Mass Gathering Medicine (MGM) Interest Group (http://www.ubcmgm.ca) in the Department of Emergency Medicine at the University of British Columbia, Canada, consists of interested clinicians who have provided care at community events for more than two decades. In 2008, an academic team committed to studying MGM in a rigorous manner was assembled.
Despite the fact that mass gatherings are frequent occurrences with some common nomenclature (Table 1), record keeping and data collection with regard to the presenting injuries and illnesses that occur during such events remain challenging. A variety of patient encounter forms (PEF) are in use, and currently there is no standardization with regard to a minimum data set, although such efforts are underway in the international context. Furthermore, no systematic approach to data collection and analysis for determining the relative influence of event and patient-related factors on illness and injury rates exists in the published literature. Standardized data collection tools for use during mass gatherings, encompassing both an event and patient focus, would be invaluable for researchers as this would allow for comparison between events and over time, building both the theory and evidence base for mass-gathering medicine.
Table 1. Commonly Used Abbreviations
aAll rates are per 1,000 attendees/staff/volunteers at event.
bMTR is preferred by the team over TTHR (transfer to hospital rate), also used in the MGM literature, since not all transfers are to a hospital.
The present report describes the development and pilot implementation of a Web-based event and patient registry for the purpose of capturing event and patient encounter data for a cross-section of mass-gathering events during a three-year period in British Columbia, Canada. The Registry will provide a comprehensive sample database of events and patient encounters, allowing for descriptive observations regarding patient presentations at different types of events over several seasons. Study funding permits the enrollment of a limited number of events on an annual basis for three years. Ethics approval was obtained from the University of British Columbia's Behavioural Research Ethics Board.
Background and Literature Synthesis
In reviewing the available mass-gathering medicine literature, the team focused on three main areas: (1) Current methods of data collection with regard to mass gatherings; (2) How results generally are reported; and (3) What a systematic approach to data collection during mass gatherings (ie, an event and patient registry) might provide for researchers, event planners, members of the medical team, and local governments or permit-granting agencies?
Mass-Gathering Medicine, Data Collection, Data Analysis, and Reporting
In relation to data collection and patient characteristics, many data fields were in common use (eg, name, date, time, chief complaint, treatment provided). Unique fields specific to certain types of events (eg, bib number, category of event participant) also were identified. Of note, fields postulated to be important were found to be “missing” both within the literature and on many of the reviewed PEFs. For example, in contrast with other fields of medicine, there are few reports with regard to patient encounters for health-promoting services or illness prevention despite the fact that individuals may present in significant numbers for such services. Gutman, Lund, and TurrisReference Gutman, Lund and Turris2 reported that 27.7% of individuals presenting to the medical team during the 2009 World Police and Fire Games received services focused on prevention, including massage therapy, taping, and other prevention-focused interventions. Overall, there was a wide heterogeneity in ways that researchers, medical professionals, and event organizers have captured data. The majority of event reports reviewed had no centralized record-keeping plan.
The reporting of patient encounter types and acuity was similarly heterogeneous (Table 2). In order to capture the range of patient presentations at events, a variety of strategies was used, including grouping similar presenting complaints and reporting percentages (eg, “lacerations/abrasions” or “musculoskeletal injuries”), reporting according to body system affected (eg, “dermatology” or “cardiac”), and/or according to final diagnosis (eg, “diabetes” or “asthma”). Most common by far was to employ a mix of these three approaches, making comparisons between and among populations and events quite challenging. In addition, measures of acuity and acuity determination varied widely. For example, some teams used “minor” and “major” designationsReference Olapade-Olaopa, Alonge, Amanor-Boadu, Alese, Adeyinka and Ogunbunmi4 and others used additional categories.5 Given the well-recognized need for rigorous approaches to medical support planning, it was anticipated that this will be an area of interest for MGM researchers in the coming decade.
Table 2. (continued) Examples of Categories Used to Classify Patient Encounters
In terms of understanding event characteristics, there are two landmark papers in the MGM literature. Milsten et alReference Milsten, Maguire, Bissell and Seaman21 synthesized the existing mass-gathering literature and summarized the many variables including weather (temperature and humidity), event duration, location (indoor vs. outdoor), geography, event boundaries (fenced, contained vs. unbound), event type, alcohol and drug availability, crowd seating (mobile vs. seated), density, mood, and average age that influence patient presentation rate (PPR). ArbonReference Arbon22 presented conceptual work with regard to the psychosocial, environmental, and biomedical domains of mass gatherings and their influence on PPR. As well, there has been ongoing work both retrospectivelyReference Ranse and Zeitz23-Reference Hartman, Williamson and Sojka26 and prospectivelyReference Shah27 to verify the utility of specific variables in predicting the PPR. For example, ShahReference Shah27 accurately predicted presentation rates drawing on 11 years of data from the Galway Races, based solely on the previous PPRs.
Arguably the main purpose of collecting event and patient data is to be able to plan medical responses in advance of a future event. Lack of consistent data collection and analysis of the injuries and illnesses that occur during particular types of events makes such planning difficult.Reference Woodall, Watt and Walker24, Reference Hartman, Williamson and Sojka26 ArbonReference Arbon28 noted the lack of a consistent approach to the collection of data to be one of the “impediments” to the development of the theory in mass-gathering research. Several researchers have worked in the area of modelling for the prediction of PPR. Arbon and colleagues provided a model for risk assessment based on a number of predictors (eg, weather, crowd mobility, event boundaries) for events with no previous medical support.Reference Arbon, Bridgewater and Smith29 Although this approach has shown some promise, it lacks utility for inter-day variability. Zeitz and colleagues published a model based on a retrospective review of events. This model has been shown to be helpful for events for which there is prior medical experience in terms of patient presentation.Reference Zeitz, Zeitz and Arbon30 However, both models have their limitations, and neither provides data regarding staffing requirements.
There has been early work on how the composition of the medical team impacts medical transfer rate (MTR) and ambulance transfer rate (ATR).Reference Grange31, Reference Martin-Gill32 More recently, Hartman and colleagues devised a scoring system for predicting resource use based on predictive factors as identified by Milsten.Reference Hartman, Williamson and Sojka26 Their results showed a good prediction of medical resource needs for “minor” and “intermediate” patient presentations. However, it seems their model has not yet been validated prospectively.
Methods
The study team reviewed the literature to identify event and population factors that have been shown to affect PPR, MTR and ATR as described above. Furthermore, an attempt was made to assemble an inclusive list of event factors that reasonably could have an influence on PPR, either in reducing it (protective/preventative factors), or increasing it (risk factors). Consultation with local, experienced providers and medical response organizers also was utilized to identify factors that were believed to influence the PPR at various types of events, as well as descriptions of the teams and their capabilities that were believed to influence the MTR. Finally, the research team sought out the reporting fields that event organizers would be interested in generating from the Registry and built automated reporting capacity into the Registry design. These data sources and field characteristics were collated into blueprint documents for event, patient, and report data/calculation fields and submitted to the database designer.
Architecture of the Registry
The Mass Gathering Medicine Registry was designed to collect and report on medical incidents at mass-gathering events. The Registry home page (http://www.ubcmgm.ca/registry) is accessible to the public. Research team members access the Registry through their unique, encrypted login ID and passwords.
The developer created the most recent version of the Registry based on an outline of the data elements provided by the team in April 2010. The design of the Registry (Table 3) was based on a number of factors:
Table 3. Mass Gathering Medicine Online Registry Project Development Package: Software and “Back-End” Description (in use August 17, 2012)
• The need for a flexible solution that could grow with the project;
• Preference for well-supported, open-source software to minimize reliance on a single developer or company for future maintenance or expansion;
• A solution that was cost-effective yet robust;
• An easy-to-use interface that was adaptable to various types of users; and
• A system that was maintainable by a non-technical audience.
Security of Registry Data
The data set is built with three tiers of dependent data, allowing for flexibility in data clustering. The top level of the data is “Event Name” with associated descriptor fields. Below that, “Event Dates” can be added to the “Event Name” in a dependent fashion. Patient encounters then are added to the database under the specific “Event Date.”
In planning the Registry, security was a major consideration due to the sensitive nature of the data being collected (Table 4). In order to control access to the registry, users were categorized into one of four categories (administrators, primary investigation team, co-investigators, data entry) for access and authentication purposes.
Table 4. Security Features
Results
The data fields for the MGM Registry were derived through an iterative process of team writing through three draft documents (Event Fields, Patient Fields, Reporting Fields), until a common set of data fields and field properties were agreed upon. Thereafter, the forms were circulated to event producers and medical service providers familiar with the MGM environment for feedback and modification. International input was sought from other research teams studying mass-gathering medicine (Dr. Paul Arbon, personal communication, November 11, 2009) regarding the preliminary work, confirming that similar patient fields had been incorporated into the “minimum data set” that another team also had hypothesized to be relevant. In formulating the MGM Registry, the approach of creating a “maximum reasonable data set” was pursued in order to capture as much information as possible to allow hypothesis generation. Therefore, the “minimum data set” is defined by the mandatory fields (Appendices 1, 2, and 3).
Drawing on common variables cited in the mass-gathering literature, the MGM Registry's patient fields were constructed with mandatory fields and optional fields (where variability was anticipated). From this work, a generic, customizable, single-page Patient Encounter Form (PEF) was derived. A multi-patient minor treatment log was created, based on the identified need to improve charting compliance for the more numerous, minor patient presentations to MGM teams. Sample forms are viewable at http://www.ubcmgm.ca/registry.
The research team primarily captures event and patient data prospectively and, when practical, enters these data on the day of the event. The data fields for characteristics of each event are in Appendices 1 and 2; data fields for patient encounters are in Appendix 3. For economy of data entry, when a brief encounter is recorded, it can be entered into the Registry through an abbreviated interface that does not include the fields for more detailed encounters (ie. vital signs, detailed past medical history). There also is a “clone” feature for the most minor presentations (eg, acetaminophen request, band-aid, simple hydration) or other requests referred to as “dispensary,” the most minor of minor interactions with the medical team, but the most numerous.
Finally, in order to “give something back” to event planners and medical directors who consent to have data from their events entered into the Registry, there are value added reporting features. For each event, a report is generated and provided for the event planner (Appendix 1). A second report, generated for the medical director, outputs all of the patient records (prior to de-identification) in an electronic, portable document format (PDF) for legible record keeping. The event and patient data that have been entered into the MGM Registry as of July 19, 2011 are summarized in Table 5.
Table 5. Events Enrolled in the University of British Columbia, Department of Emergency Medicine, Mass Gathering Medicine Registry, through July 19, 2011
Abbreviation: PPR, patient presentation rate.
aLack of standardized charting for data prior to 2010 makes PPR less accurate (underestimates).
Discussion
Despite numerous descriptive reports of patient numbers, case mixes, presentation rates, and specific encounters in the literature, currently no uniform guidelines for data collection at mass-gathering and mass-participation events exist in the Canadian context. The MGM Registry is one possible method of supporting data collection and data analysis with regard to mass gatherings that, in the future, might provide information and knowledge to underpin MGM policy development.
The research team for this project was able to locate a single report in the literature with regard to the use of an electronic registry for mass gatherings. Zeitz and colleagues reported on a pilot study in which an electronic patient surveillance system was used by South Australia St John Ambulance.Reference Zeitz33 A limitation of their study was the strict exclusion of data from all medical presentations, as well as blister and uncomplicated splinter injuries. Patient encounters that occurred as a result of traumatic injury were the exclusive focus of data collection, and they had the ability to map the location of injuries when they occurred. Zeitz and colleagues were interested in mapping potential hazards at specific events in real time so that further injuries could be prevented. Accordingly, their team had the capacity for real-time surveillance and communication with other agencies and groups on-site to have hazards addressed in a timely fashion. Currently, the MGM Event and Patient Registry does not capture information regarding geographical location of the injury on-site. Nevertheless, injury trends have been identified. For instance, a higher than expected rate of “kitchen” injuries at musical events with volunteer staff was observed, permitting in-time training to be implemented. Future ability for live reporting and communication with other agencies can empower the research team to transfer this knowledge to appropriate authorities in a more timely fashion, based on Zeitz's and our own work.
Two related research questions arose during the development of the “Patient Encounters” section of the Registry related to: (1) acuity determination, and (2) categorizing illness and injury. At this time, no validated triage or acuity scoring system for use in the somewhat unique setting of mass gatherings exists.5 In categorizing the type of the patient encounter, there is a high degree of heterogeneity. Nguyen et al assigned level of care based on time spent with patients (<5 minutes minor care, basic 5-15 minutes, advanced >15 minutes).Reference Nguyen, Milsten and Cushman34 They grouped patients requesting medication or a bandage as a “minor category.” Johnsson and colleagues devised a three-tier category system based on the urgency of the medical needs of the presenting patients where Category I represented immediate need, and Category III could await a “delayed” response.Reference Johnsson, Ortenwall, Kivi and Ledelin18 In creating the Registry, there needed to be a way to “measure” acuity for a wide variety of purposes including (but not limited to) equipment and human resource planning. Drawing on the traditions of STARTReference Navin, Sacco and McCord35 and CTAS,Reference Bullard, Unger, Spence and Grafstein36, Reference Murray, Bullard and Grafstein37 a five-category triage system was created and is described elsewhere.5 One of the unique contributions of this approach is the ability to capture patient encounters in which acuity is a non-issue (eg, the treatment of a blister, sunscreen request, water request) under the category “dispensary.” Testing of this triage system currently is underway (see http://www.ubcmgm.ca - Research section).
Arbon noted that the use of common categories for patient presentation (illness and injury groups) would facilitate sharing of data and comparison across events.Reference Arbon22 For the MGM Registry, the developers adopted a three-pronged approach to injury and illness categorization so that Registry data might be compatible with that of other researchers, regardless of the reporting categories used:
(1) Incorporation of ICD-10 codes, which allow the data entry team to categorize illnesses and injuries according to systems (eg, diseases of the respiratory system);
(2) The Registry captures information about presenting complaint (eg, “shortness of breath); and
(3) The discharge diagnosis when provided by the care provider (eg, asthma exacerbation).
Limitations
The nature of MGM events present some challenges when it comes to coordinated data entry, as data entry is not always possible on site. The specific event profile has been a factor. At events such as music festivals, the number of patients presenting each hour is more or less flat, permitting research team members to “keep up” with data entry. In contrast, at sporting events such as the Vancouver Sun Run, a 10 km run with >55,000 attendees, or the Vancouver International Marathon, simultaneous data entry was not possible due to the massive peak of patient presentations over a short period of time. Mobile events such as the Ride to Conquer Cancer, a two-day cycling charity ride involving dispersed providers, a moving medical response team, and roaming Internet access issues, made data entry in real time too difficult to accomplish with existing resources. Similarly, real-time data entry is challenging when multiple sites of care are utilized at an event.
Completeness of documentation remains a challenge, and when data entry is not accomplished on-site, there is no opportunity for the research team to approach providers to “back fill” the data. At times the legibility of the completed PEFs also has been difficult to interpret. This is a problem common to health care charting in many environments.
The full burden of illness and injury at MGM events is underestimated. At present there is no access to patient data for those who leave the event for care at community clinics or emergency departments without presenting to the on-site medical team.
Presently, funding is available for entering the data at accessible events for one final year of the three-year study period. Outstanding issues including data entry costs, expenses related to maintenance and portable Internet access, funding for ongoing research projects, and expansion of the use of the Registry provincially, nationally, and internationally will dictate its continued utilization.
Future Directions
Consistency of data collection will allow researchers and medical providers in diverse jurisdictions to systematically describe both the population of Events as well as the population of Patients. Ultimately, this will allow for observations and comparisons within and between unique events and ultimately, the pooled population of patients in similar events provincially, nationally, and perhaps in the future, internationally. Researchers would be in a position to analyze risk factors, safety features, as well as the size and nature of medical teams and their influence on the ATR and MTR. Currently, there is a dearth of models in the literature that predict PPR prospectively. There is an identified need for uniform guidelines for the provision of health services at mass gatherings that has not yet been addressed due to lack of consistent research frameworks and an absence of theory development.Reference De Lorenzo1, Reference Arbon22, Reference Arbon28 This lack of evidence, to some extent, has hindered the development of policy with regard to mass gatherings and medical support, at least in the Canadian context.
Ultimately, the adoption of a centralized Registry could allow MGM research teams to more accurately:
(1) Describe the patient populations at various unique MGM events;
(2) Document PPR and case mixes at similar and different MGM events;
(3) Hypothesize regarding event variables that affect the PPR, ATR, and MTR;
(4) Analyze the effects of the size and composition of the medical team on the ATR and MTR;
(5) Describe the equipment, medications, and protocol variability and hypothesize regarding their influence on transfer rates; and
(6) Document the frequency of rare events (eg, cardiac arrest, major trauma, mass-casualty incidents) through the collation of data from many events over multiple years and participating jurisdictions.
In the future, the incorporation of local emergency department visit data is planned. In addition, the design of the Registry could permit real-time data entry by care providers using portable devices such as iPads.
Conclusions
The Mass Gathering Medicine Event and Patient Registry provides an effective tool for recording both event and patient-related variables and automating relevant reports for event organizers in the context of mass-gathering and mass-participation events. Standardizing data collection will serve researchers and policy makers well. The systematic collection of data supports hypothesis generation and the advancement of theoretical foundations in MGM.
Acknowledgments
The authors acknowledge the work of Trivent Consulting in database design, Dr. Samuel Gutman and RockDoc Consulting, the British Columbia Ambulance Service Special Operations Division, SportMedBC, St John Ambulance Service, and the Vancouver Folk Music Festival Society.
Abbreviations
- ATR:
ambulance transfer rate
- MGM:
mass-gathering medicine
- MTR:
medical transfer rate
- PEF:
patient encounter form
- PPR:
patient presentation rate
