WILDERNESS & ENVIRONMENTAL MEDICINE, 24, 2 7 (2013) ORIGINAL RESEARCH Ten Years of Search and Rescue in Yosemite National Park: Examining the Past for Future Prevention Stacy M. Boore, MD; Dov Bock, MS From the University of California at San Diego, CA (Dr Boore); and Yosemite National Park, CA (Ms Bock). Objective. To identify trends in demographics of search and rescue incidents in Yosemite National Park and to ascertain patient opinions regarding incident prevention. Methods. Retrospective data on backcountry incidents from Yosemite Search and Rescue Patient Care Reports from 2000 to 2009 was examined, and a statistical description of data was completed to determine patient demographics. Patients assisted in the backcountry from 2007 to 2009 were surveyed, and a statistical description of data elicited factors that contributed to the incident. Results. Yosemite Search and Rescue received 1088 backcountry medical calls in 10 years. Of these calls, 55% of patients were male, 46% of incidents occurred while hiking on a trail, and extremity pain was the chief complaint in 53% of cases. The gender of patients was found to be associated with the activity causing the incident. Patient surveys reported that 79% of incidents occurred on clear days, 13% of patients felt that better gear would have helped prevent the incident, and 14.4% of patients felt they were underprepared. Fifty-four percent of patients reported that they were experts at the activity they were participating in. Appropriate footwear and sufficient water were suggested by most patients as items that would have prevented or minimized the incident. Conclusions. Many patients reported that they were experienced, well-prepared, and physically fit at the time of the incident. An educational effort regarding appropriate gear, trail conditions, and patient demographics may help decrease injury and illness in the Yosemite backcountry in the future. Key words: National Park Service, Yosemite National Park, search and rescue, wilderness medicine, prevention Introduction Search and rescue (SAR) operations have gained increasing recognition in the last 2 decades, and research on regional and national SAR statistics has been conducted in locations throughout the United States. 1 13 Studies have defined the epidemiology 1 13 and the financial cost 1 3,6,9,10 of SAR incidents, but to our knowledge no published studies have surveyed patients to assess causation and methods of prevention of these incidents. Yosemite National Park in California is the third busiest national park in the country, with 3,951,393 visitors in 2010. 14 The park is about the size of Rhode Island, and Support: Friends of Yosemite Search and Rescue and The Ohio State University Alumni Fund. Previous publication: Preliminary findings presented as a poster at the George Wright Society s Biennial Conference, New Orleans, LA, March 2011. Corresponding author: Stacy M. Boore, MD, 200 West Arbor Drive, San Diego, CA 92103 (e-mail: sboore@ucsd.edu). includes 1101 square miles of designated wilderness area. 15 The Search and Rescue program in Yosemite National Park (YOSAR) responds to an average of 240 backcountry incidents each year. 15 YOSAR consists of rangers, community volunteers, SAR dogs, and 2 teams of 10 members, one located in Tuolumne Meadows, the high-elevation region of Yosemite National Park, and one located in Yosemite Valley, the busier and lowerelevation region of the park. The teams are trained in high-angle rigging, backcountry rescue, and medical care. Every medical incident responded to by Emergency Medical Services (EMS) in Yosemite National Park has an associated Patient Care Report (PCR). Providers certification levels include EMT-B, parkmedic, paramedic, and RN. A parkmedic is a similar certification to a paramedic, but the curriculum and scope of practice are specific to providing care in national parks. 16 A Preventative Search and Rescue Program (PSAR) was created in 2007, according to an internal briefing by Joy
Search and Rescue in Yosemite National Park 3 Marschall of Yosemite National Park, with the goal of reducing the number of preventable SAR incidents occurring in the park. Yosemite is an ideal location to study the epidemiology of backcountry medical incidents, to assess the possibility of prevention of these incidents, and to assist the PSAR program in reducing SAR incidents in the future. A previous study of incidents in Yosemite National Park was conducted using data from 1990 to 1999, 9 but because of inconsistent reporting of incidents before 2000, a follow-up study from 2000 to 2009 was needed. In addition, to our knowledge no studies have surveyed previous patients to assess information not recorded on patient care reports (PCRs), including trail conditions, weather conditions, and final diagnosis after transport to a clinic or hospital. Finally, no studies have attempted to determine whether incidents could have been prevented. The goal of this study was first to determine the epidemiology of incidents from 2000 to 2009 in Yosemite National Park, and then to survey patients to determine details about the incidents, contributing factors, and whether anything could have prevented or minimized the severity of the incident. Methods All PCRs from 2000 to 2009 were examined, and the following information was extracted and compiled: date of incident, case number, SAR number if the incident was in the backcountry, patient name, age, gender, provider name, provider certification level, location of incident, chief complaint of the patient, and mechanism of injury. Backcountry, for the purposes of a SAR classification in Yosemite National Park, is defined as not being accessible by ambulance. Only the subset of backcountry incidents for those assigned a SAR number were used in this study, and both traumas and illnesses were included. All individual activities and chief complaints listed on the PCRs were grouped into the categories listed in Table 1. When the incident was primarily medical and unrelated to an activity, the activity category used was medical. When more than 1 chief complaint was listed, the most severe injury or illness was used. In cases in which all complaints were of equal severity, the first complaint listed was used. For example, when ankle pain and deformity were listed as well as fatigue, ankle pain and deformity were considered to be the primary complaint and grouped under pain, extremity. These assignments were made by the primary author and were not analyzed for interobserver variance. A statistical description of PCRs was completed, and a Pearson 2 test was used to determine association between variables, with significance for probability values Table 1. Demographic information of patients from PCRs compared with information as reported by survey respondents Variable PCR data 2000 2009 (n 1088) Survey data (n 131) Gender Male 55% 53% Female 45% 47% Age (years) 18 13% 7% 18 25 18% 11% 26 44 28% 20% 45 65 32% 47% 65 9% 15% Subdistrict Yosemite Valley 21% 28% Tuolumne Meadows 65% 59% Other subdistricts in 14% 13% YNP Activity Medical 34% Climbing 9% 3% Hiking on trail 46% 49% Hiking scrambling 11% 8% Chief complaint Pain, extremity 53% 51% Gastrointestinal 11% 12% problem Altered mental 10% 12% status Head pain 9% 10% Chest pain 4% 7% Difficulty breathing 4% 4% Fatigue 6% 3% Severe or fatal injury 3% 1% less than.05. Because repeated testing was done on different comparisons using the same data, a Bonferroni correction was used to provide a conservative estimate of significance. Seven comparisons were made, giving a raw probability value of less than.007. The second part of this study surveyed all patients who became injured or ill in the backcountry (assigned a SAR number) from 2007 to 2009 via a mailed questionnaire. Addresses were obtained from the PCR reports. Inclusion criteria for surveyed patients included the following: patients requiring assistance from YOSAR between 2007 and 2009, regardless of age (surveys sent to pediatric patients could be completed by the parent or guardian). Three years was chosen as a cutoff to allow for a large enough number of patients to perform a statistical analysis, while attempting to minimize recall bias. Exclusion criteria included the following: patients without a domes-
4 Boore and Bock tic mailing address recorded on the PCR. The survey asked questions regarding environmental and trail conditions, preparation, and possible preventable factors of the incident. A statistical description of survey data was completed and a Pearson 2 test was used to determine an association between variables. Microsoft Excel (2007; Microsoft Corp, Redmond, WA) was used for data compilation and statistical descriptions, and SPSS version 19.0 (SPSS Inc, Chicago, IL) was used for analysis of association of variables. The National Park Service Research and Collecting Permit Office reviewed and approved the study. Results During the study period, there were a total of 3218 medical calls in Yosemite National Park. Of these, 1088 were considered backcountry and therefore received a SAR number in addition to a case number. On average, 108.8 backcountry medical incidents occurred each year in Yosemite National Park, ranging from 70 to 150 incidents each year. Seventy-eight percent of backcountry incidents occurred between June and September, with July being the busiest month (27%). Demographic information, activities, and chief complaints for all incidents are provided in Table 1. Cross-tabulations of PCR data were completed and analyzed using a Pearson 2 test to determine whether there was an association between variables. Comparisons included activity vs gender, chief complaint vs gender, activity vs age, chief complaint vs age, activity vs subdistrict, chief complaint vs subdistrict, and activity vs chief complaint. All associations were found to be significant using the corrected probability value of less than.007. One of these associations is pictured, gender vs activity (Figure 1). Three hundred thirty-six patients met inclusion criteria for the survey, and 28 were eliminated because of the lack of a US postal address. Three hundred eight patients qualified for inclusion and were mailed questionnaires. Seventeen questionnaires were returned by the Postal Service because of a change of address. One hundred thirty-one completed questionnaires were returned either by mail or electronically, for a response rate of 45%. A statistical description of survey data was completed, and demographic information, activities, and chief complaints are listed in Table 1. A Pearson 2 test using a probability value of less than.05 did not show an association between variables in Table 1. Additional information from surveys included the timing of the incident: 29% of incidents occurred in first half of the trip or activity, 17% occurred near the halfway point, and 54% occurred in the second half of the trip or activity. The majority of patients reported that their incident occurred on clear days (79%), while only 15% reported the weather as hot, 2% reported snow, and less than 1% reported rain. The time of day was reported as afternoon in 71% of cases, and night in only 9% of cases. Patients were also asked to assess what factors may have helped prevent the incident, as shown in Figure 2. Only 13% of respondents agreed that better equipment would have helped to prevent or minimize their incident, while 75% disagreed that better equipment would have helped. A comparison of what equipment or resources patients had with them at the time of the incident to what equipment or resources they would have liked to have had with them is shown in Table 2. Also shown is what equipment the group of patients who had an incident 100% 90% 80% 70% 50% 30% 20% 10% 0% 29% 38.8% Medical (no associated activity) 3.6% 13.1% 58.8% 35.1% 8.6% 13.1% Climbing Hiking on Trail Hiking off Trail, Scrambling Female Male Figure 1. Association of gender and activity. Percentage of all female incidents in each activity category, and percentage of all male incidents in each activity category. Association between variables is significant (P.007).
Search and Rescue in Yosemite National Park 5 100% 90% 80% 96% 89% 75.3% 83.3% 70% 50% 30% 20% 10% 4% 11% 24.8% 16.7% 0% Knowing the weather Knowing effects of altitude Knowing the terrain Better equipment Agree Disagree Figure 2. Importance of factors in incident prevention. Percentage of survey respondents who agreed that these factors would have prevented or minimized the incident. Table 2. Percentage of respondents who had or desired these resources to prevent or minimize their incident Resource Possessed (n 131) Desired (n 131) Desired by hikers on trails a (n 64) Appropriate 91% 1% 2% clothing Appropriate 79% 11% 14% footwear Sufficient water 84% 10% 20% Sufficient food 71% 5% 0% First aid supplies 48% 2% 3% Headlamp/ 47% 2% 0% flashlight Batteries 0% 0% Map 47% 0% 0% Compass 41% 1% 0% Global 6% 0% 0% positioning system (GPS) Cell phone 29% 6% 5% Whistle 32% 2% 0% Companions 75% 2% 0% Emergency plan 18% 1% 0% Turnaround time 17% 1% 0% Weather forecast 33% 2% 0% None 2% 49% 59% Other 0% 16% 7% b a Desired by the subset of patients injured or ill while hiking on a trail. b This 7% wrote in trekking poles as the desired resource. while hiking on a trail would have wanted with them to prevent or minimize the incident. A comparison of the experience level of the patient to what supplies the patient had with them at the time of the incident is shown in Figure 3. Forty-six percent of patients reported themselves as beginners (participated in the activity fewer than 4 times), while 54% reported themselves as experts (participated in the activity more than 10 times). A comparison of activity by experience level using survey data demonstrated that patients were experts in 100% of climbing incidents, 89% of hiking-scrambling incidents, and 45% of hiking on trail incidents. Patients were beginners in no climbing incidents, 11% of hiking-scrambling incidents, and 18% of hiking on trail incidents. When asked whether they thought their preparation was adequate, 85.6% of respondents thought that it was adequate, while 14.4% thought it was not. Finally, educational efforts by the National Park Service were assessed on the survey. Fifty-eight percent of respondents indicated that they had not received education from the National Park Service before their trip on safety and injury prevention, while 19% of respondents indicated that they had received education. Of the 19% who reported that they received education before the trip, 80% agreed that the education was useful, 20% neither agreed nor disagreed, and no one thought it was not useful. Discussion The epidemiology of medical incidents in the Yosemite backcountry does not differ greatly from the demographics of patients in other national parks and wilderness areas. 1 13 There were significant associations between variables in the data from 2000 to 2009, indicating that certain people are more likely to be injured doing certain activ-
6 Boore and Bock 100% 90% 80% 70% 50% 30% 20% 10% 0% 61% 39% Communication Device 65% 64% 35% 36% First Aid Supplies Headlamp Map/Compass/GPS Beginner Expert Figure 3. Experience level and gear choices. A comparison of supplies held by beginners to supplies held by experts as reported on surveys. ities. As shown in Figure 1, gender is associated with activity, which may only reflect different rates of participation in a given activity by different genders. Alternatively, it could actually reflect different rates of injury doing a certain activity for different genders. Without demographic data on the overall participants in each category, and not just the subset of injured participants, this information cannot be used to assess risk for injury when participating in an activity. Survey responses elicited areas in which preventive efforts could be focused. Educating visitors on how to travel over dangerous terrain, and possibly posting signs regarding dangerous areas, is one potentially effective intervention, as shown in Figure 2. Additionally, educating park visitors on appropriate equipment for various activities could help prevent or minimize medical incidents in the backcountry. As shown in Table 2, although many park visitors are not carrying the Ten Essentials commonly recommended, 17 these are not necessarily the items they thought would have helped prevent or minimize their injury or illness. Appropriate footwear and sufficient water were the 2 items that would have prevented or minimized incidents among those hurt while hiking on trails, the most common activity resulting in injuries. Education efforts should begin to move away from the traditional recommendations and target those items that were actually implicated in injury occurrence. Items that could be recommended based on this study include appropriate footwear, sufficient water, sufficient food, and trekking poles. Although cell phones were also suggested as being useful by survey respondents, a cell phone cannot help prevent incidents or help patients to self-rescue, and are cautiously recommended. A study by Pope and Martin 18 of visitors to the Lost Coast Trail found that visitors perceive inexperience to be the most important factor contributing to the need for rescue in the wilderness, whereas bad luck was perceived to be the least important factor. In contrast, patients in our study did not believe that better knowledge or preparation would have prevented or minimized their incident, and many wrote that they thought their injury was caused by bad luck. The statistical data regarding the lack of association between injury and experience level could be used as a part of an educational effort to inform visitors that many injury victims are experienced. Of all patients who had cell phones with them at the time of their injury, a majority were self-reported beginners at the activity, whereas of all those who had headlamps, first aid supplies, or a map, compass, or GPS, a majority were self-reported experts at the activity. Cell coverage in Yosemite National Park is inconsistent, and this finding may not be useful in preventing SAR incidents in Yosemite. Studies could further elicit the association between experience level and the perception of the importance of technology in rescue, as well as the prevalence and actual importance of technology in SAR operations. Survey results indicated that although education is not currently reaching a majority of patients, those that receive this education find it useful. A new forum for education aimed at reaching a larger percentage of park visitors is one of the next steps in preventing medical incidents in the Yosemite backcountry. LIMITATIONS Surveyed patients could have been affected by recall bias, especially those completing the survey years after the incident. Nonparticipation is another potential bias. We did not capture the entire population in our survey responses. Possible reasons for the response rate of 45% include the
Search and Rescue in Yosemite National Park 7 following: the person had moved from the provided address, or the address was recorded incorrectly; the person declined to respond; or the person had passed away since the incident. Assignment of activities and chief complaints into categories could have introduced bias as assignments were not made by both authors to assess for interobserver variance. The protocol for these assignments attempted to eliminate bias, but future work should make an effort to have multiple authors independently assign categories. Additionally, future PCRs could document a primary chief complaint followed by secondary complaints. Finally, those patients who self-rescue and initially present to a clinic or hospital are not captured in this study. Conclusions Patient demographics in the Yosemite National Park backcountry are similar to other national parks, and the most common medical incidents in the backcountry of Yosemite National Park were musculoskeletal injuries to extremities that occurred while patients were hiking on trails. Preventive efforts could include education targeted at appropriate gear choices and trail conditions. Specific gear recommendations should be made based on actual incidents, rather than a standard list, and will likely vary in different parks and with different activities. Informing visitors that a large percentage of previous patients considered themselves well-prepared experts may dispel visitors beliefs that only beginners are involved in medical incidents and encourage safe behavior. Acknowledgments Support from the Search and Rescue Program in Yosemite National Park allowed completion of this project, with particular help from John Dill. Financial support was provided by Friends of Yosemite Search and Rescue and The Ohio State University College of Medicine Alumni Fund. We would also like to thank Jeni Squiric of the Statistics Department at The Ohio State University for her help. References 1. Heggie TW, Heggie TM. Search and rescue trends and the emergency medicine service workload in Utah s National Parks. Wilderness Environ Med. 2008;19:164 171. 2. Heggie TW. Search and rescue in Alaska s National Parks. Travel Med Infect Dis. 2008;6:355 361. 3. Heggie TW, Heggie TM. Search and rescue trends associated with recreational travel in US National Parks. J Travel Med. 2009;16:23 27. 4. McIntosh SE, Brillhart A, Dow J, Grissom CK. Search and rescue activity on Denali, 1990 to 2008. Wilderness Environ Med. 2010;21:103 108. 5. Schmidt TA, Federiuk CS, Zechnich A, Forsythe M, Christie M, Andrews C. Advanced life support in the wilderness: 5-year experience of the Reach and Treat team. Wilderness Environ Med. 1996;3:208 215. 6. Wild FJ. Epidemiology of mountain search and rescue operations in Banff, Yoho, and Kootenay National Parks, 2003 06. Wilderness Environ Med. 2008;19:245 251. 7. Ela GK. Epidemiology of wilderness search and rescue in New Hampshire, 1999 2001. Wilderness Environ Med. 2004;15:11 17. 8. Johnson J, Maertins M, Shalit M, Bierbaum TJ, Goldman DE, Lowe RA. Wilderness emergency medical services: the experiences at Sequoia and Kings Canyon National Parks. Am J Emerg Med. 1991;9:211 216. 9. Hung EK, Townes DA. Search and rescue in Yosemite National Park: a 10-year review. Wilderness Environ Med. 2007;18:111 116. 10. Heggie TW, Amundson ME. Dead men walking: search and rescue in US National Parks. Wilderness Environ Med. 2009;20:244 249. 11. Forrester JD, Holstege CP. Injury and illness encountered in Shenandoah National Park. Wilderness Environ Med. 2009;20:318 326. 12. Johnson RM, Huettl B, Kocsis V, Chan SB, Kordick MF. Injuries sustained at Yellowstone National Park requiring emergency medical system activation. Wilderness Environ Med. 2007;18:186 189. 13. Montalvo R, Winegard DL, Bracker M, Davidson TM. Morbidity and mortality in the wilderness. West J Med. 1998;168:248 254. 14. United States National Park Service Public Use Statistics Office. Available at: http://www.nature.nps.gov/stats/. Accessed April 10, 2012. 15. United States National Park Service, Yosemite National Park Statistics. Available at: http://www.nps.gov/yose/ naturescience/park-statistics.htm. Accessed April 10, 2012. 16. Kaufman TI, Knopp R, Webster T. The Parkmedic Program: prehospital care in the National Parks. Ann Emerg Med. 1981;10:156 160. 17. Cox SM, Fulsaas K, eds. Mountaineering: The Freedom of the Hills, 7th ed. Seattle, WA: The Mountaineers Books; 2003. 18. Pope K, Martin SR. Visitor perceptions of technology, risk, and rescue in wilderness. Int J Wilderness. 2011; 17:19 26, 48.