Professional Documents
Culture Documents
Author manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Author Manuscript
Abstract
Objectives—Pediatric behavioral health disorders and related emergency department visits are
increasing, but effects on emergency medical services (EMS) are unknown. This study’s objective
was to describe the statewide epidemiology of pediatric behavioral health-related EMS encounters
in Florida, including mental health and substance use.
Results—There were 22,254 pediatric behavioral health-related EMS encounters during the
Author Manuscript
study period, one-quarter of which were noted to have suspected or confirmed ingestion /
substance use. The median age was 16 and the majority of patients were female and white. A total
of 946 patients (4%) had an acute EMS behavioral / psychiatric intervention. EMS scene, ED
turnaround, and total EMS time were significantly longer for intervention patients. Of the 14
counties in the top quartile of percentages of intervention patients, 7 were rural, 10 did not have
any hospitals with child / adolescent psychiatric services, and 7 did not have any child
psychiatrists.
community paramedicine to alleviate EMS utilization and decrease the frequency of pediatric
behavioral health emergencies.
Address correspondence and reprint requests to: Dr. Jennifer N. Fishe, MD, Department of Emergency Medicine, University of
Florida - Jacksonville, 655 West 8th Street, Jacksonville, FL 32209. Tel: (904) 244-4046, jennifer.fishe@jax.ufl.edu.
Author Contributions: J.N.F. contributed to the study concept and design, acquisition of funding and the data, statistical analysis,
interpretation of the data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content. S.L.
contributed to the study concept and design, interpretation of the data, drafting of the manuscript, and critical revision of the
manuscript for important intellectual content.
Conflict of Interest: Neither J.N.F. nor S.L. have conflicts of interest or financial disclosures.
Fishe and Lynch Page 2
Keywords
Author Manuscript
Behavioral Health; Emergency Medical Services; Mental Health; Pediatrics; Substance Use
Introduction
Recent years have seen significant increases in the number of emergency department (ED)
visits for patients aged 15 years and older for behavioral health disorders (including mental
health and substance use).1 One study found that adult ED visits for mental or substance use
disorders increased from 28 per 1,000 ER visits in 2005 to 35 in 2011.2 For youth ages 18
years and younger, the prevalence of pediatric behavioral health disorders in the United
States is also rising, with attendant increases in ED utilization.3–6 For example, annual
pediatric behavioral health-related emergency department visits rose 26% between 2001–
Author Manuscript
Accordingly, EMS are increasingly providing psychosocial care related to behavioral health.
7–9 Some have construed use of EMS for behavioral health emergencies as a misuse of
emergency resources.8,10 Others have argued that further work should be done to optimize
EMS provider clinical decision-making related to behavioral health care.9,11 Since EMS
providers increasingly provide care for adult patients with behavioral health disorders,12
EMS agencies have pursued adult direct transport protocols to psychiatric hospitals and
community paramedicine initiatives to alleviate increases in EMS resource use such as
repeat calls, and in turn reduce ED utilization that might be avoidable.13–16 Additionally,
community paramedicine models are being developed that include behavioral health.10 Such
health care service delivery models expand paramedic and emergency medical technician
(EMT) roles to work with adults with serious mental illnesses.10 However, there is little
Author Manuscript
Methods
Study Design and Setting
Author Manuscript
This analysis was a retrospective observational study of pediatric patients ages 2 – 18 years
with a behavioral health-related EMS encounter from 2011–2016. We selected this age range
to capture the fullest picture of pediatric EMS utilization and because are infrequently
diagnosed in patients less than two years of age.17 We identified patients from Florida’s
EMS Tracking and Reporting System (EMSTARS) database, which contains EMS
encounters from over 100 EMS agencies and includes 74% of all statewide 911 EMS patient
calls during the study period.18 Most counties are served by one EMS agency.18 EMSTARS
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 3
data for the study period covered 64 of Florida’s 67 counties, of which 28 counties were
classified as rural by the Florida Department of Health (DOH).19 The 3 counties not
Author Manuscript
included were all classified as rural.19 In Florida, each EMS agency operates under its own
protocols, in contrast to other statewide systems. Accordingly, any official relationship
between EMS and mental health resources, and/or EMS protocols for behavioral health
encounters, vary widely. Individual agencies submit patient care reports (collected via
National EMS Information Systems (NEMSIS)-compliant and state-validated ePCR
software) to the Florida DOH Bureau of EMS which integrates the data into EMSTARS.
EMSTARS during the study period contained Florida-specific elements (both data variables
and their values) and national elements which conform to the NEMSIS-National Highway
Traffic Safety Administration (NEMSIS-NHTSA) Version 2.2.1 Data Dictionary elements.20
We obtained county-level data on child psychiatric resources (number of hospitals with
child / adolescent psychiatric services and child psychiatric providers in 2016) from the
Health Resources and Services Administration’s Area Health Resource Files (AHRF).21 The
Author Manuscript
University of Florida and Florida DOH Institutional Review Boards approved the study
(IRB201702645 and Protocol 180000U11, respectively).
Data Variables
We abstracted demographic, clinical, EMS, and geographic characteristics. EMSTARS
contains a unique patient identifier for each encounter, which allows for unique patient
encounter identification even if multiple EMS units respond to the same scene. Each vital
sign’s first recorded measurement was used (as many EMS decisions are made based on the
initial assessment).22 EMS times were calculated as minute intervals. We considered
negative time intervals as miscoded and those were excluded. Other variables were
abstracted directly from EMSTARS. After examining all the EMS medications and
procedures administered to the study sample patients, we characterized patients by those
who had one or more of certain a posteriori-defined acute behavioral / psychiatric
intervention(s) by EMS: administration of activated charcoal, diazepam, haloperidol,
Author Manuscript
Data Analyses
We report descriptive statistics and compared characteristics between patients who did and
did not receive an acute behavioral / psychiatric intervention (“intervention”). We performed
an analysis specifically examining the distribution of intervention and non-intervention
patients by age. Continuous variables were compared using unpaired t-test and Wilcoxon
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 4
Rank Sum tests, as appropriate. Categorical variables were compared using the Chi-Square
Author Manuscript
test, or Fisher’s Exact Test when there were less than 10 observations (a conservative
approach).23 The Kolmogorov-Smirnov test evaluated whether variables were normally
distributed (sample size greater than 2,000).24
We conducted geographic analyses classifying patients by the treating EMS agency’s home
county. We first classified counties as urban or rural (rural being 100 persons or less per
square mile).19 For each county, we calculated the percent of patients who received one or
more acute behavioral / psychiatric EMS interventions. Since percent by county was not
normally distributed, we divided counties into quartiles by the percent of patients who
received an acute intervention. Next, we described the behavioral health resources available
for each county and also by each quartile using the AHRF’s 2016 county-level data.
Missing data were excluded by individual variable (e.g., for a patient encounter missing
Author Manuscript
ethnicity data, that encounter was excluded from ethnicity analysis only, but the patient
encounter was still included for other analyses). We excluded missing data since we were
not able to determine whether EMSTARS’ data were missing at random due to its
aggregation of data from multiple heterogeneous agencies.18 Statistical analysis was
performed using SAS® version 9.4 (Cary, NC). Geospatial analysis was performed using
ArcGIS Desktop 10.4.1 (Redlands, CA).
Results
From 2011 to 2016, there were 3,491,241 patients of all ages in EMSTARS and 388,187
patients ages 2–18 transported by EMS, of which 22,254 were pediatric behavioral health-
related EMS encounters (Figure 1). Most patients (90%, n=20,023) had at least one provider
impression of behavioral / psychiatric disorder. One-quarter of patients (25%, n=5,670) had
Author Manuscript
at least one impression of poisoning / drug ingestion, intentional drug use, or alcohol-related
problems. The top three provider primary impressions were “behavioral / psychiatric
disorder” (n=13,739, 62%), “intentional drug use; related problems” (n=1,974, 9%), and
“alcohol related problems / delirium tremens” (n=1,020, 5%). There were 981 acute
behavioral / psychiatric interventions by EMS in 946 patients (4% of total patients, Table 2).
Of those 946 patients, 911 had only one intervention and 35 patients had 2 interventions.
Patient and EMS characteristics, by those who received one or more acute intervention(s)
versus those who received no acute behavioral / psychiatric interventions are displayed in
Tables 3a and 3b, respectively. For both groups, the median age was 16 and the majority
were female and white. Intervention patients had significantly higher heart and respiratory
rates, and decreased total Glasgow Coma Scores (GCS) and level of alertness (all p<0.0001).
Author Manuscript
Missing EMSTARS data is also documented in Tables 3a and 3b, with race, ethnicity,
respiratory effort, and level of alertness having the highest percentage of missing values.
Prehospital blood glucose levels were missing in 60% of patients, which precluded further
meaningful analysis between intervention and non-intervention patients. However, of the
3,847 patients with a GCS less than 15, 3,370 (88%) did not have a blood glucose level
recorded.
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 5
intervention patients were transported with lights and sirens (Table 3b, p<0.0001).
Intervention patients had significantly longer scene times (median 16 vs 13 minutes), ED
turnaround times (median 25 vs 20 minutes), and total EMS encounter times (median 66 vs
58 minutes) (all p<0.0001). The a priori-excluded negative time intervals comprised less
than 1% of all times. More patients with an intervention were noted to be improved
clinically by the EMS provider at the time of ED arrival (p<0.0001). Aside from those
medications which defined an intervention, both groups received few other medications
aside from oxygen (8% intervention, 13% non-intervention) and fluids (<1% intervention,
6% non-intervention). Intravenous (IV) access was low and did not differ between groups
(5% in both groups). When patients were divided into age categories reflective of different
developmental and school stages, the proportion of patients receiving an acute behavioral /
psychiatric intervention varied significantly amongst categories (p<0.0001) (Figure 2).
Author Manuscript
The EMS agencies’ county data were available for all but 71 patients (<1% of total). There
were no encounters in 3 of Florida’s 67 counties (all rural). As between urban and rural
counties, there were no significant difference in response time, scene time, and transport
time (all p>0.05). Notably, rural counties had shorter ED turnaround times and total EMS
times (both p<0.0001). The median county percentage of acute intervention patients was
2%, with an interquartile range of 0% – 7%, and not normally distributed (Shapiro-Wilk w
value 0.8, p-value <0.0001). Figure 3 maps county-level intervention quartiles and number
of hospitals with child / adolescent psychiatric services. Of the 14 counties in the highest
intervention percentage quartile, 7 were rural, and 7 were urban. Of the 7 rural counties, 6
did not have any hospitals with child / adolescent psychiatric services, and 5 did not have
any child psychiatrists in the AHRF database. Of the 7 urban counties, 4 did not have any
hospitals with child / adolescent psychiatric services, and 2 did not have any child
Author Manuscript
psychiatrists.
Discussion
This study describes the epidemiology of non-critically ill pediatric behavioral health-related
EMS encounters in the state with the fourth-largest pediatric population.25 The vast majority
of study and intervention patients were adolescents (median age 16 years), although the
thousands of younger patients who met inclusion criteria merit future study. Females had a
higher proportion of encounters than males. That finding is consistent with pediatric studies
showing a higher female prevalence of mental health disorders such as anxiety and
depression, and a higher proportion of psychiatric hospitalizations for suicidal ideation or
self-harm.3,4
Author Manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 6
The large proportion of substance use when paired with 62% of patients missing prehospital
Author Manuscript
blood glucose measurements (including 88% of patients with a GCS less than 15) prompts
concern for unrecognized hypoglycemia. Low rates of blood glucose testing may be due to
EMS provider safety concerns with potentially aggressive and unpredictable patients.
However, EMS education should encourage blood glucose testing whenever feasible, as
hypoglycemia may cause or exacerbate patient agitation.
EMS scene time, ED turnaround time, and total EMS time were all significantly longer for
pediatric patients with acute behavioral / psychiatric interventions. Longer EMS times
decrease available units for the next dispatch call. The 8-minute increase in median total
EMS times for intervention patients may or may not be operationally significant depending
on the EMS agency. Another explanation for interventions by rural EMS agencies may be
safety concerns for longer transports, however there was not a significant difference in EMS
transport time between intervention and non-intervention patients, as well as between urban
and rural counties. This still may reflect concerns EMS providers have regarding patient and
crew safety for transports in counties where there are few or no behavioral health services
that have potentially engaged or counseled the EMS agency on behavioral health training.
Alternatively, the percentage of interventions in rural counties may reflect poorly controlled
behavioral health disorders due to a lack of community behavioral health resources.
Author Manuscript
The high proportion of substance use/ingestions, the impact on EMS encounter times -
particularly for patients requiring interventions, and our analysis showing that some counties
with high acute intervention rates did not have access to many hospital-based resources
suggests that community paramedicine may provide a potential solution. Possible roles for
EMS providers under a community paramedicine initiative may include patients and families
counseling on the benefits and logistics of outpatient psychiatric care, assisting with care
coordination, and identifying available community behavioral health resources.10,28 Those
interventions may reduce EMS utilization through the decrease in avoidable and/or repeat
EMS pediatric behavioral health-related encounters. However, research involving
community paramedicine in this role is needed.
Author Manuscript
Whether urban or rural, this study’s findings stress the need to train EMS providers to
provide pediatric-specific behavioral health care during acute encounters or as part of
nascent community paramedicine programs. Local public health and EMS leadership in
areas with high or disproportionate levels of behavioral health-related EMS encounters
should examine the utility of direct transport protocols, community paramedicine, and care
coordination with the (decreasing) number of pediatric mental health providers.29 Future
research can assess how such programs reduce EMS resource utilization and avoidable
pediatric EMS behavioral health-related encounters.
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 7
Limitations
Author Manuscript
This study is limited by its retrospective design and its setting is limited to one state,
although Florida has the fourth-largest pediatric population in the United States.25 Missing
data varied greatly by variable and should be taken into consideration when interpreting
results. EMSTARS collects data from individual agencies that use different data capture
systems, and therefore does not have the same degree of data capture, quality, and
“cleaning” as does a registry.18 EMSTARS is a population-based dataset with voluntary
participation (albeit covering 64 of 67 counties), and is therefore subject to both selection
and information bias.18 EMSTARS contains both state-specific and NEMSIS-compliant
data, therefore not every variable and/or value (including EMS interventions such as
“psychological first aid”) may apply in other settings and EMS systems.20,18 Additionally,
our definitions of a behavioral health-related encounter and acute behavioral / psychiatric
intervention were made using clinical judgement. There may be relevant EMS encounters or
Author Manuscript
interventions not captured by those definitions. Further to this, we chose to exclude acutely
ill patients, which may not capture patients suffering critical medical illnesses or trauma
secondary to a behavioral/psychiatric problem. For example, our selection criteria would
exclude a patient with a cardiopulmonary arrest due to an opioid overdose. However, the
intention of this study was to characterize non-critically ill behavioral health EMS patients.
We were unable to ascertain the timing of administered medications or procedures with
respect to with each patient’s first documented vital signs, and therefore we are unable to
present whether or not any abnormal vital signs were before or after an EMS intervention.
ED outcomes were not available, limiting assessment of the results of EMS interventions
and each patient encounter. While we used the most recent AHRF data available to
characterize psychiatric resources, it is possible that new facilities were opened during or
after the study period that would not have been operative for the entire study period.21
Author Manuscript
Conclusion
Pediatric behavioral-health related EMS encounters have a significant proportion of
suspected ingestions / substance use, and disproportionate effects on rural agencies.
Increases in EMS resource utilization are occurring in settings with limited behavioral health
infrastructure. Those findings may represent an opportunity for direct transport protocols,
community paramedicine, and/or coordination with local or regional behavioral health
resources to decrease the frequency of EMS utilization and pediatric behavioral health
emergencies for patients and families.
Acknowledgements
Author Manuscript
The study investigators acknowledge Steve McCoy, Brenda Clotfelter, Karen Card, DrPH and Joshua Sturms from
the Florida Department of Health’s Bureau of Emergency Medical Oversight for their assistance and data
management. The study investigators acknowledge Erik Finlay from the University of Florida GeoPlan Center for
his assistance with this study.
Funding Source: Research reported in this publication was supported by the National Center for Advancing
Translational Sciences of the National Institutes of Health under University of Florida Clinical and Translational
Science Awards KL2TR001429 and UL1TR001427. The information, content, and conclusions are those of the
authors and should not be construed as the official position, policy, or endorsement by the National Institutes of
Health.
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 8
References
Author Manuscript
1. Weiss AJ, Barrett ML, Heslin KC, Stocks C. 2016 Trends in emergency department visits involving
mental and substance use disorders, 2006 – 2013. HCUP Statistical Brief #216. Rockville, MD:
Agency for Healthcare Research & Quality Available at: https://www.hcup-us.ahrq.gov/reports/
statbriefs/sb216-Mental-Substance-Use-Disorder-ED-Visit-Trends.jsp. Accessed November 26,
2018.
2. Ayangbayi T, Okunade A, Karakus M, Nianogo T. Characteristics of Hospital Emergency Room
Visits for Mental and Substance Use Disorders. Psychiatric Services 2017;68(4): 408–410.
[PubMed: 27974000]
3. Bitsko RH, Holbrook JR, Ghandour RM, Blumberg SJ, Visser SN, Perou R, Walkup JT.
Epidemiology and Impact of Health Care Provider-Diagnosed Anxiety and Depression Among U.S.
Children. J Dev Behav Pediatr. 2018;39(5):395–403. [PubMed: 29688990]
4. Teich JL, Mutter R, Gibbons B, Miller K, Lynch S, Ali MM. Use of Inpatient Behavioral Health
Services by Children and Adolescents With Private Insurance Coverage. Psychiatr Serv. 2018;69(9):
1036–1039. [PubMed: 29921189]
Author Manuscript
5. Pittsenbarger ZE, Mannix R. Trends in pediatric visits to the emergency department for psychiatric
illnesses. Acad Emerg Med. 2014;21(4):25–30. [PubMed: 24552521]
6. Lovegrove MC, Weidle NJ, Budnitz DS. Trends in Emergency Department Visits for Unsupervised
Pediatric Medication Exposures, 2004–2013. Pediatrics. 2015;136(4):e821–829. [PubMed:
26347435]
7. Horowitz L, Kassam-Adams N, Bergstein J. Mental health aspects of emergency medical services
for children: summary of a consensus conference. Acad Emerg Med. 2001;8(12):1187–1196.
[PubMed: 11733300]
8. Ford-Jones PC, Chaufan C. A Critical Analysis of Debates Around Mental Health Calls in the
Prehospital Setting. Inquiry. 2017;54: DOI: 10.1177/0046958017704608.
9. Shaban R Mental health and mental illness in paramedic practice: A warrant for research and
inquiry into accounts of paramedic clinical judgment and decision-making. Australasian Journal of
Paramedicine. 2015;2(3).
10. Prener C, Lincoln AK. Emergency medical services and “psych calls”: Examining the work of
urban EMS providers. Am J Orthopsychiatry. 2015;85(6):612–619. [PubMed: 26192707]
Author Manuscript
11. Parsons V, O’Brien L. Paramedic clinical decision-making in mental health care: a new theoretical
approach. Journal of Paramedic Practice. 2011;3(10):572–579.
12. Larkin GL, Claassen CA, Pelletier AJ, Camargo CA Jr. National study of ambulance transports to
United States emergency departments: importance of mental health problems. Prehosp Disaster
Med. 2006;21(2):82–90. [PubMed: 16770997]
13. Cheney P, Haddock T, Sanchez L, Ernst A, Weiss S. Safety and compliance with an emergency
medical service direct psychiatric center transport protocol. Am J Emerg Med. 2008;26(7):750–
756. [PubMed: 18774037]
14. Trivedi TK, Glenn M, Hern G, Schriger DL, Sporer KA. Emergency Medical Services Use Among
Patients Receiving Involuntary Psychiatric Holds and the Safety of an Out-of-Hospital Screening
Protocol to “Medically Clear” Psychiatric Emergencies in the Field, 2011 to 2016. Ann Emerg
Med. 2018 DOI: 10.1016/j.annemergmed.2018.08.422. [EPub ahead of print]
15. Creed JO, Cyr JM, Owino H, Box SE, Ives-Rublee M, Sheitman BB, Steiner BD, Williams JG,
Bachman MW, G Cabanas J, Myers JB, Glickman SW. Acute Crisis Care for Patients with Mental
Author Manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 9
18. Florida’s Prehospital EMS Tracking and Reporting System (EMSTARS) Data Dictionary Version
1.4. http://www.floridaemstars.com. Accessed December 15, 2018.
Author Manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 10
Author Manuscript
Author Manuscript
Author Manuscript
Figure 1:
Study Patient Selection
Author Manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 11
Author Manuscript
Author Manuscript
Figure 2:
Percent Study and Intervention Patients by Age Category
Author Manuscript
Author Manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 12
Author Manuscript
Author Manuscript
Author Manuscript
Figure 3.
Author Manuscript
Pediatric Mental Health Resources and Proportion of Pediatric EMS Behavioral Health-
Related Encounters that Received an Acute Behavioral / Psychiatric Intervention by County
from 2011–2016.
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 13
Table 1:
Inclusion Exclusion
Any primary or secondary impression of: Any primary or secondary impression of:
Behavioral / psychiatric disorder Cardiac arrest
Intentional drug use Airway obstruction
Alcohol related problems / Delirium tremens (DTs) Vaginal hemorrhage
Pregnancy / OB delivery
Hypovolemia / shock
Obvious Death
Respiratory arrest
Fever related problems / symptoms
Heat related illness
Author Manuscript
Sepsis
Sickle cell crisis
Stroke
Electrocution
*
All of the listed phrases in both inclusion and exclusion criteria are values for EMS provider primary and secondary impression in EMSTARS
EMS = emergency medical services, EMSTARS = EMS tracking and reporting system
Author Manuscript
Author Manuscript
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 14
Table 2:
Number of Acute EMS Behavioral / Psychiatric Interventions, Total N=981 for N=946 patients
Author Manuscript
Intervention N (%)
EMS Administered Medication
Diazepam 53 (6%)
Haloperidol 71 (8%)
Lorazepam 55 (6%)
EMS Procedure
Decontamination 1 (<1%)
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 15
Table 3a:
Patient characteristics of pediatric behavioral health-related EMS encounters in Florida between 2011–2016,
Author Manuscript
by those who did and did not receive an acute behavioral / psychiatric intervention
Demographics
Clinical Characteristics
SBP, [mean (SD)] 125.7 (22) 125.9 (19) 0.826 761 (3%)
Heart Rate (beats per minute), [mean (SD)] 108.2 (26) 100.9 (23) <0.0001 443 (2%)
Author Manuscript
Respiratory Rate (breaths per minute), [median (IQR)] 20 (16–24) 18 (16–22) <0.0001 696 (3%)
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 16
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 17
Table 3b:
EMS characteristics of pediatric behavioral health-related EMS encounters in Florida between 2011–2016, by
Author Manuscript
those who did and did not receive an acute behavioral / psychiatric intervention
EMS Characteristics
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.
Fishe and Lynch Page 18
Prehosp Emerg Care. Author manuscript; available in PMC 2020 September 01.