The study took place from July 2018 to February 2019 in San Francisco, California. Approval for this study, including a waiver of informed consent for use of the retrospective dataset of pediatric encounters, was obtained from the Institutional Review Board of the University of California, San Francisco. Electronic informed consent was obtained from all study participants. Data were obtained from three sources in a single, urban EMS system: a retrospective review of local pediatric EMS encounters over one year; survey data of EMS practitioners’ comfort with pediatric skills using a 7-point Likert scale; and qualitative data from focus groups with EMS practitioners assessing their experiences with pediatric patients and their preferred training modalities.
Local EMS system
The San Francisco Emergency Medical Services System serves the City and County of San Francisco, an area of 47 mile2 with a daytime population of approximately 1.3 million and a night time population of 800,000. The 9–1-1 Emergency Medical Services call volume (demand for ambulance service) is approximately 120,000 calls annually, of which approximately 4% are for patients under the age of 18 years. All 9–1-1 calls in San Francisco are answered and dispatched by a single entity, The Emergency Communications Center at the Department of Emergency Management. Three EMS organizations respond to 9–1-1 calls: San Francisco Fire Department (SFFD, 75% of call volume), King American Ambulance Company (King, 15% of call volume), and American Medical Response (AMR, 10% of call volume). In addition, there are 4 additional private ambulance companies that provide interfacility transport for adult and pediatric patients that are not tracked by the San Francisco EMS Agency. All 9–1-1 patients are transported to one of 13 receiving hospitals. Among these, there is one Level 1 Trauma Center that serves both adults and children, as well as two pediatric critical care centers.
Review of local Pediatric EMS encounters
All medical 9–1-1 calls managed by SFFD are logged using an electronic database (ESO, Austin, TX, USA). This database does not include data from calls managed by AMR or King. The SFFD database is compiled from the electronic EMS charts completed by the prehospital care team. We retrospectively analyzed the patient characteristics of all SFFD prehospital encounters for patients 18 years and younger from January 1, 2018 to December 31, 2018. The following characteristics were obtained and summarized as percentage of total numbers: time and date of encounter, patient’s age, gender, race/ethnicity, chief complaint, the primary impression (primary complaint as specified by the prehospital provider), acuity of transport to receiving hospital (low, emergent, or critical), and patients’ level of distress (mild, moderate, severe). Race/ethnicity data were self-reported by patient or caregiver and documented by EMS provider into predefined categories within electronic EMS chart (ESO, Austin, TX, USA).
The most frequent primary complaints were calculated for all encounters and for the following age groups: 0 to 1 years of age, 2 to 5 years of age, 6 to 11 years of age, 12 to 15 years of age, and 16 to 18 years of age. Some primary impressions were independently assigned to broader categories by two investigators (KP, NG) to better reflect the most common type of encounter. Any discordance was discussed and agreed upon by all investigators. For example, “trauma” included all primary impressions involving injury, burn, or hematoma. “Neurology” included primary impressions of headache and altered mental status, but excluded seizures given the high prevalence of seizures among pediatric EMS encounters.
Survey design and collection
An electronic survey was developed to obtain information on EMS practitioners’ comfort with caring for critically ill pediatric patients. Respondents were asked to use a seven-point Likert scale (ranking from “extremely uncomfortable” to “extremely comfortable”) to rate 34 aspects of pediatric care divided into six clinical domains based on EMS education standards [28]: respiratory, shock, cardiac arrest, care of the newborn, trauma, and other (“other” included skills related to seizure management, using length-based weight estimation, toxidromes, and managing concerned parents).
The survey was initially piloted on a group of 6 EMS practitioners, including leadership representatives from each of the three EMS providers in San Francisco, and revised to reflect common feedback. The survey was distributed to EMS practitioners electronically using a secure email link through a local EMS listserv comprising approximately 400 members, as well as through a San Francisco EMS social media group with approximately 458 members. An estimated 858 possible EMS providers were eligible to participate in survey, though an unknown number of providers belonged to both email listserv and social media group. Participants were eligible to complete the survey if they were an active EMS practitioner working in San Francisco. There is total of approximately 2400 registered EMS providers in San Francisco. The total active workforce is unknown; local EMS providers are not required to have an email address or belong to unified directory. The survey was live from August 1, 2018 to October 30, 2018.
Qualitative focus groups design and analysis
We conducted focus groups with a convenience sample of prehospital providers at each of the three major EMS agencies in San Francisco: SFFD, King American, and AMR. Participants were approached in-person by investigators after change of shift. In order to reflect the mixed nature of EMS teams, each group included a mix both Emergency Medicine Technicians (EMTs) and Paramedics. Given participants were approached after change of shift, each group contained participants from a single agency. No other persons were present during the focus groups aside from the researchers and participants. The focus groups were planned to comprise a minimum of three and maximum of twelve participants and were facilitated by one or more authors trained in qualitative interview techniques (KP, JB, and NG). Each group lasted between 30 to 60 min. They were continued until no new themes emerged from interviews, thereby precluding a predetermined number of participants consistent with qualitative methods [29]. Sampling from each agency reflected the volume of calls at each agency (for example, approximately three-quarters of sessions took place with SFFD providers).
The focus group guide was developed by the authors with support from other medical educators with experience in qualitative methods, and in accordance to prior qualitative research guidelines [30]. The guide was piloted on six EMS practitioners and edited to reflect common feedback (Appendix 1). All focus groups started by asking participants to recall their last encounter of a critically ill pediatric patient. Participants were prompted to discuss the case in detail and elicit challenges in caring for pediatric patients. The second half of the interview focused on training and learning, including types of learning formats that are beneficial, and prior experience with simulation-based exercises.
All focus group sessions were recorded and transcribed (RevRecorder, San Francisco, CA, USA). A “constant comparison” analysis was conducted by three authors (KP, NG, JB), whereby transcripts were repeatedly read and an initial framework of key codes was developed and refined with successive readings. Mutually agreed-upon definitions and examples for each code were developed; codes were reviewed and revised with disagreements in coding resolved by team consensus. Final themes were determined using the principles of grounded theory that are common for qualitative analysis [29, 31]. After a final analysis, all themes were discussed with the present authors (KP, JB, CC, NG, AZ), as well as a larger interdisciplinary team, including EMS leadership, local EMS providers, emergency medicine providers, and medical educators with experience in qualitative analysis.