Patient assessment in emergency medical services: complexity and uncertainty in street-level patient processing.

Henderson, Alexander C.

INTRODUCTION

Though still a relatively nascent public function, emergency medical services (EMS) agencies have come to constitute a core local government service. As the initial line of emergency treatment in a larger health care system, EMS providers are tasked with the treatment of patients in urgent, complex, and uncertain situations, and providing transportation to advanced health care facilities. The potentially life-saving interventional and palliative treatments provided by EMS personnel are inarguably a vital social service that, for patients in medically precarious situations, "... may determine the outcome as much as the subsequent hospital-based care" (Institute of Medicine, 2007, p. 1).

The Institute of Medicine of the National Academy of Sciences estimates that approximately 16 million emergency department patients arrive via ambulance every year, an average of more than 30 patients transported every minute (2007, p. xiv). This staggering volume of service is handled by more than 840,000 EMS providers working for local government, nonprofit, or for-profit EMS agencies (American Ambulance Association, 2009). Despite the significant volume of calls for service answered, and the substantial number of individuals engaging in front-line service provision, there is a paucity of research examining the behavior of these individuals from the scholarly field of public administration.

Emergency medical services providers, in their capacity as front-line public servants, can be analytically examined as "street-level bureaucrats" (Lipsky, 1980). Previous empirical research into the behavior of street-level bureaucrats has focused on several distinct occupational areas, including law enforcement, nursing, front-line welfare workers, and teachers (Isett, Morrissey, & Topping, 2006; Lipsky, 1980; Maynard-Moody & Musheno, 2003; Riccucci, 2005; Vinzant & Crothers, 1998). Given the relationship between the fundamental tasks of certain professions and the behavior of individuals fulfilling those roles, it is important to examine EMS providers as separate and distinct from other street-level occupations (Hill & Hupe. 2003, p. 477),

As a profession, EMS has received substantial attention from the fields of clinical medicine, health care economics, operations research, community planning, and other related areas of interest, resulting in a body of empirical evidence on clinical care, provider education, and "systems-level" considerations (National Highway Traffic Safety Administration, 2001). There is, however, there is a lack of research into the behavior of emergency medical services providers at the front lines (Institute of Medicine, 2007). Specifically, there is a lack of understanding of the exercise of discretion and the internal and external sources of influence that act on EMS providers.

This exploratory research begins to narrow this gap in our understanding of EMS by examining the uncertainty and complexity of street-level patient processing. The act of determining the scope and extent of client need--perhaps the most difficult aspect of front-line occupations--is made increasingly complex by the nature of emergency medical service as based in the physiological and psychological needs of patients. Given the intricacy of many of these issues, and their inability to be easily observed, the information gathering process is often clouded or obscured by a variety of sources. And, though EMS provision is rooted in clinical medicine--an area of study not often subject to scrutiny in public administration literature--it serves to illustrate an important example of front-line public policy implementation. As noted previously, EMS has evolved into a core public service over the last several decades (IOM, 2007; NHTSA, 2001). There is an expectation that, in response to both common emergencies and crises and disasters, emergency medical health care will be available to the public regardless of the time of day or type of need. Though EMS agencies may be public, nonprofit, or private organizations, the responsibility for providing the service generally rests with local government organizations. This serves to reinforce the importance of the study of these organizations from a public policy implementation perspective.

As with other rule sets found in front line occupations, the clinical protocols employed in EMS care are numerous, complex, and sometimes conflicting. Though varying on a state-by-state basis, clinical protocols generally employ text and visual algorithms to describe the patient assessment process and treatment procedures. These protocols are generated either at the state-level, through standard notice-and-comment procedures used in the creation of administrative regulations, or at the local-level by regional EMS agencies in a similar fashion (NASEMSO, 2011). In all cases, these written clinical policies create foundation from which all EMS care is derived.

This study specifically focuses on the determination of patient need during incidents in which paramedics were presented with incomplete, vague, or misleading information by patients, family members, or through their own judgments of ascribed patient characteristics. Focusing on cases of difficult or failed information gathering efforts in front-line EMS is crucial in that these interactions have serious and potentially fatal results. To this end, this study investigates how several conventional sources of influence on the behavior of street-level bureaucrats may complicate the nature of patient processing in EMS organizations.

This article will first discuss past empirical research on both street-level bureaucrats and emergency medical services, followed by a discussion of the similarities and differences of emergency medical services as compared with other front-line functions. Research methods and case selection will then be discussed, outlining the use of semi-structured interviews with thirty (30) paramedics in the State of Pennsylvania. The results of these in-depth interviews--firsthand accounts of complex and difficult service interactions--will be presented and discussed in the context of street-level bureaucratic behavior. Three sources of influence found to be of primary importance in complicating the act of patient processing--paramedic communication skills, patient identity, and bystanders--will be presented and discussed. Limitations of this research, future avenues of investigation, and concluding remarks will follow.

LITERATURE REVIEW

Street-Level Bureaucrats in the Public Service

Street-level bureaucrats are responsible, at the most basic level, for policy implementation (Brodkin, 2011; Lipsky, 1980; Riccucci, 2005; Maynard-Moody & Musheno, 2003; Vinzant & Crothers, 1998). The complex tasks of front-line workers, often situated in human services agencies with voluminous rules and potentially ambiguous goals, are of vital importance to individuals receiving the services and in many cases represent core quality-of-life issues (Bovens & Zouridis, 2002; Hill, 1974; Keiser, 1999; Lipsky, 1980). The cumulative actions of bureaucrats at the front lines therefore constitute a critical area of interest not just for individual clients, but also for analysis and evaluation of public programs (Riccucci, 2005).

The defining characteristics of the concept of street-level bureaucracy, notably the ability to exercise discretion in an official capacity in face-to-face interactions with clients, also point directly to the most complex aspects of these positions (Vinzant & Crothers, 1998, p. 40; Davis, 1969; Handler, 1986). Front-line public servants exercise discretion in time-bound and uncertain situations--often engaging in rule-saturated tasks with potentially ambiguous goals--in response to the complex needs of the individuals they are serving (Keiser, 1999). Discretion, in the words of Dimock, ". is the very stuff of the daily duties in all branches of government" (1936/1967, p. 65).

Conceptually, discretion is dependent to a great extent on varying types of constraints. As Hupe and Hill (2007) noted, "[d]iscretion and rules are interrelated: as rules specify the duties and obligations of officials, discretion allows them freedom of action" (p. 280-281). Bureaucratic discretion, then, is defined and delimited by both the broad legislation drafted to create public programs as well as the administrative rules and regulations developed during the implementation process. In addition to rules, front-line bureaucratic behavior is also both regulated and constrained by the social and professional norms of human services agencies (Dworkin, 1977; Handler, 1986; Hupe & Hill, 2003; Lipsky, 1980; Scott, 1997). Thus, the difficulty of front-line work lies at the intersection of rule application or deviation, conformity to expectations of behavior, and situational factors, and represents ". a flexibility versus uniformity dilemma" (Loyens & Maesschalck, 2010, p. 67).

The nature of discretion and resolution of these difficult problems of street-level service can only be understood by examining sources of influence at the front lines (Handler, 1986; Riccucci, 2005; Vinzant & Crothers, 1998). Several sources of influence within organizations have been found to shape street-level bureaucratic behavior. Managerial influence on street-level bureaucrats has been studied empirically with varying results in the impact of managerial action across different occupations and organizational settings (May & Winter, 2007; Riccucci, 2005).

Cultural considerations have also been found to play a role in shaping street-level behavior. Kelly (1994), in a study of teachers and welfare workers, found that the organizational culture of an agency does indeed shape the ability of front-line workers to exercise discretion in line with their conceptions of justice (p. 138). Likewise, occupational culture--the norms, beliefs, and values related to how a specific job is accomplished--has been found to be important in studies of health care providers, and welfare workers (Isett, Morrissey, & Topping, 2006; Riccucci, 2005; Sandfort, 2000). Pertinent to the present discussion, Isett, Morrissey, and Topping (2006) found that front-line health care providers' perspectives are mainly shaped by the nature of their daily tasks. "... [W]hat matters to street-level bureaucrats is how they do what they do, not the larger questions of what they are doing or some abstract paradigmatic explanations of why they are doing it" (Isett, Morrissey, & Topping, 2006, p. 223).

External, extraorganizational sources have a similarly substantial bearing on front-line behavior. Notably, political principals have been found to both directly and indirectly shape front-line work behavior through the distribution of resources to the front-lines, and through the expression of preferences for certain types of behavior (Gilboy, 1992; May & Winter, 2007). Perhaps more central to this study, clients, and specifically the street-level bureaucrat's perception of the client's identity and worthiness, have been found to significantly impact front-line worker behavior (Maynard-Moody & Musheno, 2003). Scott (1997), in an experimental study of welfare eligibility workers, used multiple cases to depict clients with differing levels of need to evoke varying levels of compassion for the client's situation. On average, those clients who evoked a higher need for compassion received higher levels of assistance (Scott, 1997).

Client identity has been found to shape treatment in other sectors of the healthcare system (Cain & Kington, 2003; Fiscella, 2004). Todd et al. (2000), and Todd, Samaroo, and Hoffman (2004), found ethnicity to be a strong predictor of the administration of pain medications to patients presenting with similar types of injuries. In a study of 618 patients and their treating physicians, van Ryn and Burke (2000) found that patient race was associated with "... physician's assessment of patient intelligence, feelings of affiliation toward the patient, and beliefs about the patient's likelihood of risk behavior and adherence with medical advice; patient SES was associated with physician's perceptions of patient's personality, abilities, behavioral tendencies, and role demands" (p. 813).

Past Empirical Research in EMS

Though making substantive contributions to the understanding of EMS, the general body of empirical research on EMS is somewhat removed from the focus of this study. A review of this body of evidence does, however, allow for more concrete identification of gaps in knowledge. Previous research on EMS "systems" has addressed the creation of performance measurement regulations for EMS agencies (Narad & Driesbock, 1999), on theoretical discussion and empirical investigation of public versus private provision of EMS (David & Chiang, 2009), on racial disparities in systems-level service provision (David & Harrington, 2010), and on predicting EMS calls volume (Brown et al., 2007; Setzler, Saydam, & Park, 2009). Though these broader considerations may indirectly impact front-line behavior in tangible ways, they fail to account for the complexity and intricacy of service-level interactions of EMS providers.

Empirical research from other perspectives has ventured closer to the realm of the actual service-level interaction. Salerno, Wrenn, and Slovis (1991), in a retrospective review of 1,246 patient care charts, found that EMS providers deviated from written clinical protocols in 199 (16%) of the cases, with 55% of deviations considered "minor," 38% considered "serious," and 7% categorized as "very serious" (p. 1321). Though providing important findings regarding the nature of protocol deviations, this study does not consider the actual decision-making processes employed during services interactions. David and Brachet (2009) studied the relationship between the volume of EMS skills performed during services interactions by individual paramedics and improved patient outcomes, finding that increased exposure to prehospital skills is robustly related to reduced time-at-scene and transport times (p. 919). While also important to our knowledge of EMS, the dependent variable in this study--the amount of time spent at the scene of an incident and transporting a patient to the hospital--is a very crude measure of patient outcomes. Though certainly adding to knowledge, both studies do not examine the intricacies of actual service provision. This study serves, in essence, to bring to light the decision processes not found in either. The highly situational nature of EMS calls and complex relationships between rules, discretion, and influence requires a more specific and intense focus paramedic behavior responding to actual incidents.

Emergency Medical Services as a Street-Level Occupation

The essential tasks of front-line EMS providers share certain similarities and display important differences from those of police officers, welfare workers, teachers, and other street-level bureaucrats. As with other street-level occupations, EMS providers must follow a large body of rules and regulations, and are faced with substantial situational complexity and uncertainty as they evaluate patients presenting with complex physiological or psychological conditions, and provide services in potentially hazardous situations (Lightner, Brywczynski, McKinney, & Slovis, 2010, p. 57). And, as with other street-level occupations, EMS providers exercise discretion and may react to myriad types of influence in carrying out their duties.

However, both the nature of EMS work, based in the physiological or psychological needs of clients as opposed to social, economic, or legal needs, and the temporal immediacy of these incidents, creates some important differences from other street-level professions. First, the reason for client need, though potentially difficult to observe in other street-level positions, is in many cases unobservable in EMS. Second, the nature of EMS work is intrinsically urgent. Though clients of other public service agencies may require services of an emergent nature, response to those services can often be handled in days and hours as opposed to minutes and seconds. The intensity of an EMS call may also require urgency from beginning to end, a period of time that is substantially longer than other street-level professions. The present study seeks to investigate these difficult and complex service interactions in front-line patient care.

METHODS

Past research in the behavior of front-line workers has used both qualitative (Gilboy, 1992; Maynard-Moody & Musheno, 2003; Newman, Guy, & Mastracci, 2009; Sandfort, 2000; Skolnick, 1966; Vinzant & Crothers, 1998), mixed methods (Oberfield, 2010; Riccucci, 2005), and quantitative experimental studies (Scott, 1997). Given that the inherent subjectivity of both the concepts of influence and discretion, as well as the complexity of the decision-processes used in making patient care decisions, an interpretive methodological foundation is particularly appropriate for this study (Riccucci, 2010).

As this research is exploratory in nature, and includes several interrelated, complex, and unexamined concepts, a multiple-case-study design is appropriate (Yin, 1994). This design allows for what George and Bennett (2005) call "process-tracing," or the identification of the possible causal mechanisms at work for a given relationship between variables. Given the topic of interest, this research is also necessarily multi-level, allowing for contextual influence at the state, community, organizational, and individual levels (Yin, 1994). The unit of analysis is the individual street-level EMS worker providing service in a particular situation, allowing for the examination of different types of discretion, rule abidance and deviation, and influence as they are enacted by individuals. Full-time, career EMS providers working in organizations that provide emergency medical transportation services are the focus of this study.

Selection and Attributes of Participating Organizations

The three organizations examined in this study were selected from a single state in order to keep the political and regulatory context constant across organizations. Pennsylvania was chosen as the focus of this study as it has both a substantial number of emergency medical services providers--the third highest number of career emergency medical technicians and paramedics in the United States (Department of Labor, 2009)--and also displays a substantial call volume of approximately 1.8 million calls for service in 2008 (Pennsylvania Department of Health Bureau of Emergency Medical Services, 2009, p. 1).

Participating organizations were chosen for substantial call volume and for structured variation of organizational arrangements (e.g., fire department-based EMS ["FDEMS"]; police department-based EMS ["PDEMS"]; and hospital-based EMS ["HSEMS"]). Data on Pennsylvania municipalities from the 2000 U.S. Census were used as a sampling frame, and sorted by population size. Organizational characteristics for the official municipal emergency medical service was identified, and organizations were selected to represent each of the primary organizational arrangements while simultaneously remaining similar in demographic and geographic characteristics to control for this variation (West, 2001). Detailed descriptions of these organizations are provided in Table 1.

Participants for the semi-structured interviews were randomly chosen from a population of on-duty paramedics over a 2- to 3-day interview period at each location. A total of 10 front-line paramedics were interviewed from each participating organization, and all interviews were audio-recorded and transcribed. The interview process outlined here was pilot tested with paramedics from two agencies not involved in the research. Interviewee characteristics are provided in Table 2.

Semi-Structured Interviews

Semi-structured interviews were conducted to examine the nature of influence and the exercise of discretion in emergency medical services. Data gathered from interviews can move beyond simple description and can "... encompass the hows of people's lives (the constructive work involved in producing order in everyday life) as well as the traditional whats (the activities of everyday life)" (emphasis in original, Fontana & Frey 2005, p. 698). This study uses an interview method in a manner similar to Maynard-Moody and Musheno (2003) and Kelly (1994). Participants were asked to relate stories about their experiences working at the front-lines of emergency medical services, with a specific focus on service interactions that were particularly memorable, complex, challenging, or involved a nonroutine situation.

Using storytelling as a method of social science research has several potentially unique strengths for the study of front-line work (Bailey & Tilley, 2002; Connelly & Clandinin, 2006; Kelly, 1994; Maynard-Moody & Musheno, 2003). Researchers' working propositions are less obvious to the storyteller, and, as noted before, stories can depict complex interactions between multiple variables and localize these within certain situations. They can show "... what situations call for certain routines and how the specifics of a case do or do not fit standard practices" and can show the consequences of the actions of street-level emergency medical services providers (Maynard-Moody & Musheno, 2003, p. 29).

Data from the semi-structured interviews were analyzed using a grounded theory approach. Grounded theory is appropriate for this particular study in that it is allows for exploration of the data in a manner that brings out the processes that, combined, create the total package of patient treatment. Charmaz noted that this is a particular strength of grounded theory, and is brings to light the "... enacted processes, made real through actions performed again and again" (emphasis in original, 2005, p. 508). This method of analysis will focus on bringing to the fore the assessment and treatment processes that EMS providers use on a daily basis.

Data Reduction

The three stories selected for discussion here were chosen for several reasons. First, each incident represents some amount of difficulty at the intersection of rules, patient need, and context. In particular, these stories are representative of instances in which the assessment and treatment process was characterized by complexity, uncertainty, and vagueness. Second, they are comprehensive and detailed enough to allow for analysis and interpretation. The details provided by the paramedics in each were substantive enough to allow for identification of the nature of the situation and context of the care provided, the primary actors involved in the story, the impressions and reactions of the paramedics as they interacted with patients, and the primary source of influence that shaped paramedic behavior.

RESULTS AND DISCUSSION

Several narratives of front-line patient processing will be presented, each of which highlights how a primary type of influence may shape and complicate the process of street-level EMS. These influences range from concepts that are paramedic-centric--notably communication skills--to those that are more commonly found in front-line work, including the behavior of bystanders during the service interaction and patient identity. Importantly, each narrative represents street-level behavior that is policy-based, situationally contingent, and influenced by past experiences. Each incident will be introduced, outlining the general importance of that source of influence on the service interaction, followed by a description of the incident and discussion of key aspects of the narrative.

Paramedic Communication Skills and the Patient Assessment Process

A core function of any street-level role is that of determining client need, a process of variable complexity given the primary functions of the service in question or the particular citizen's needs. In those cases in which patient need is not relatively obvious to responding paramedics, the process of collecting data about the patient may be dependent on the paramedic's individual communication skills. The dyadic patient-paramedic communication that occurs during an emergency includes both the initial patient assessment and determination of need, as well as the continual reassessment of the patient after clinically appropriate medications and procedures have been performed. Both verbal communication skills, specifically the language used to convey meaning, and non and nonverbal communication skills (e.g., affect, body language) of responding paramedics may allow for more efficient or effective methods of determining patient need.

Discussing an incident in which the patient was not able to effectively communicate the nature of her chief compliant, one paramedic noted that he overcame this obstacle not through changing the topic of this diagnostic questioning but by altering the wording of the questions he asked. Though the standard set of questions he asked had been effective in previous patient assessment efforts, he found that he had to alter his approach to better communicate with the patient to determine need.

   ... [T]his was a call, it was an old lady. And it was one that made
   you think. And it was actually one comment that made me put my ear
   up, because she's complaining of all non-specific stuff, and this
   and that. And you're going through, and you're like, "What's going
   on? ... There's gotta be something more to this. And you go through
   and she's going "I feel washed out, and I feel this, and I feel
   that." And she's just complaining, "I just feel weak, I just ...
   I'm washed out. This is that." And it's just one that you're
   sitting there going, "There's got to be something going on." [The
   patient says] "... and this is what I felt like when I was in the
   hospital before." "When's before?" "Well before, it was like last
   week. They did one of those things on me, where they go in though
   your vein." She had an MI [myocardial infarction, a heart attack].
   She was in the middle of another one, again. But it ... just didn't
   show up in a 12 lead ..., but you call in and you say to the doc,
   "This is what she's telling me, and she's telling me this is how
   she felt last week, and this is what you wound up doing." And the
   doc met us at the back of the ambulance and said, "is this Miss so
   and so? Because that was bad last week." And it was one of those
   where you had to just keep digging. They just don't present
   themselves...

   Yeah, it was all the standard questions I was asking, but some of
   it was...some of it's in the way you have to reword your standard
   questions. I can ask anybody, "How don't you feel well today?" And
   I'll get an answer from "Well, I'm having pain in my chest" to "I
   just don't feel good" to "I don't know, something's not right."
   Depending on the answers they're giving you, you have to try to ...
   sometimes you have to phrase something in a different way. Like I
   probably asked seven or eight times about trying to figure out why
   she was weak. "What did they tell you the last time?" Until I
   finally said it in a way that it dawned on her that that's what she
   should tell me. But some guys would have just wrote it off as
   "Well, she doesn't feel good."

In this situation the paramedic was faced with a patient expressing her feelings of distress in very general, nonspecific terms. Despite this, some aspect of the patient's presentation cued the paramedic to make further inquiries about the nature of her complaints, which resulted in further information about her recent medical history. The medic noted that "... it was all the standard questions I was asking, but some of it was ... some of it's in the way you have to reword your standard questions." After making several attempts to gather pertinent information, the medic was able to ascertain that the patient had a recent history of a heart attack and interventional cardiac catheterization. This additional information, and the diagnostic testing that was spurred by the information, allowed the paramedic to conduct a more focused assessment and find that the patient was indeed having another heart attack.

Had the paramedic not recognized the cue to continue his line of questioning, the appropriate interventional therapies would not have been provided and the patient's condition would likely have rapidly deteriorated. The delay in receiving appropriate care, critical in emergencies such as this, could have been significant given the nature of the emergency department intake, assessment, and patient prioritization. The appropriate field diagnosis of this condition allowed for more rapid and focused treatment of this critical patient once she arrived at the emergency department.

When asked about what kind of paramedic is good at communicating with patients, one medic described it as being "... able to engage somebody enough to be able to get the accurate story out of them. Not just ask questions but ... hav[ing] follow-up questions to get the correct answer to the question you are asking. When they don't add up, explore enough so you know you have gotten the correct answer." Both patient assessment and communication skills, then, become important in determining patient need and related appropriate treatments. Another medic noted that "... the information gathering is the real art, which I think some people do well and some don't. I think that personally makes a good medic is the good listener, the person who knows what questions to ask and listens to the answers."

Those medics who are unable to probe for information relating to past medical history, the nature of the incident and progression of the patient's current condition, or other key variables such as prescribed medications or environmental factors may miss key information that would aid in determining appropriate treatment. After consciously working on his own perceived shortcomings in patient assessment, including communication skills, the same medic noted that he found himself in fewer situations like the story above. In reflecting on these past experiences, he noted that felt he often "... didn't do a good job getting their story." The medic who previously defined good communication skills also noted pointedly that few organizations "... will hold up somebody's command [authorization to practice] because they can't communicate with the patient." Organizations may not allow a paramedic with poorly developed intubation skills to advance to independent care and treat patients without supervision; however, paramedics with poor communication skills, which are inherently much harder to assess, are not as closely evaluated and may not be prevented from unsupervised practice.

Paramedic communication skills, then, become crucial in ensuring improved patient outcomes. Though the paramedic outlining this incident would have satisfied the requirements of the clinical policies governing patient assessment without expending additional effort, the discretionary decision to continue the assessment procedure was critical to the patient's outcome. Moving from paramedic-centered factors in the assessment process, other actors played a part in shaping how street-level EMS providers employed both rule-following and discretionary behavior.

The Effects of Bystanders on Patient Assessment and Treatment

Bystanders--family members, witnesses to an incident, or other key individuals--were also found to be important in patient assessment and treatment decisions. The behavior or reactions of these individuals, or the information they may provide during an incident, may be influential on paramedics as they attempt to determine patient need and therapies best matched to the patient's condition. Patient past medical history, accounts of the immediate situation and moments leading up to the request for EMS, and specialized knowledge or skills may aid paramedics as they assess patients or determine the appropriateness of treatments. An example of this type of influence was evident in a story about the rescue of a worker experiencing a medical emergency on the roof of a four-story building. In this case, family members provided information about the patient's past medical history of congestive heart failure (CHF) to responding paramedics, which then altered their perceptions about the possible causes of the patient's distress.

   I remember one that was really complicated ... We got called for
   respiratory distress ... We had to get the fire department there
   because this guy is up on the roof. He was a big, big guy and he
   was roofer and he was tarring this roof ... [H]is son is there with
   him working on the roof and ... [he's] is yelling down, "He can't
   breathe, he's got CHF." ... So a ladder truck gets there and sets
   up a ladder and we go up to the top of this four-story building.
   This guy is pale, diaphoretic. He is breathing very labored but at
   the same time, he is not following our commands. He is just kind of
   flailing and pushing us away and not cooperating at all. It is hot
   ... up there. It is like 90 degrees with the sun coming down, much
   less what is radiating off the roof that they were just tarring. It
   was so uncomfortable. You have a bunch of us up there. ... Everyone
   is screaming that he was just seen for CHF in the hospital last
   week and all this sort of stuff. Everything is just like, CHF, CHF,
   CHF, respiratory distress. That is all that is in my brain. I am
   trying to listen to this guy's lungs in this hectic atmosphere. He
   is a big guy. He is moving around. He is pushing away from me. He
   breathes out and I can hear some crackling in his lungs kind of but
   in retrospect it probably was not a very accurate. I probably did
   hear crackling but it was not as big of a deal as I made it out to
   be. I was like "he is acting this way because he is hypoxic and he
   can't breathe so that is why he is like flailing."

   His breathing is all labored so we slap on an oxygen mask and we
   try to take his blood pressure but he is uncooperative. It comes up
   kind of high, like 200s, something like that. Automatically I was
   thinking he is in respiratory distress and hypoxia and all this
   stuff. Let's put some nitro paste on him and we try to get IV
   access to give him captopril and ace inhibitors. His blood pressure
   is up, his respiratory rate is up and labored. He has some crackles
   in his bases. But, something just wasn't sitting right. We started
   down the CHF route and we started treating him for respiratory...
   trying to get the fluid off of his lungs and get his work of
   breathing down and all this sort of stuff.

From the outset, the narrator noted that this is a "complicated" call. Not only is the patient experiencing a serious medical condition--difficulty breathing according to the family members on-scene working with the patient--but the patient was also four stories above ground on a roof in the middle of a hot summer day and surrounded by excited bystanders. Upon arriving on-scene, and after reaching the roof, the patient's son made it clear that the patient has a medical history of CHF, and that he was recently released from the hospital after treatment for this condition.

A rapid assessment of the patient's respiratory function produces some evidence that this could indeed be the reason for his respiratory distress. The medic theorized that the respiratory distress is resulting in hypoxia, a systemic lack of oxygen, which could then be explaining the patient's somewhat agitated interactions with medics. With this potential diagnosis in mind, the medics began to treat the patient according to the protocol for CHF. However, something was not "sitting right" with the more experienced paramedic on the crew.

   He was like, "I don't think this is CHF." He has CHF, you could
   tell but that is not what is problem is right now. Something in [my
   partner's] head said "I think he's got a brain bleed." His
   breathing is labored and all that but he is ... pushing us away, he
   is exerting a lot of energy and that is why he is diaphoretic. He's
   confused. He is not responding to our instructions very well. He is
   answering our questions but only kind of.

   So, we switched .... We started treating as like a stroke or a
   brain bleed kind of thing instead ... By the time we got to the
   hospital, it is way more evident now that he is strapped down and
   he is off of the roof and he is in a controlled atmosphere in the
   back of the ambulance. He's acting goofy. He's acting confused.
   That is more of the issue than the breathing thing ...

   If it was really his breathing, he would not be like pushing me
   away when I tried to give him oxygen. That is not normal behavior
   for someone in respiratory distress. But that is something that is
   consistent with somebody who has a brain bleed or a stroke or
   something like that, some sort of altered mental status. So, we
   were able to change our course of action.

Recognizing some dissonance between the patient's exertion of energy to push the medics away during the assessment and the standard presentation of patients with congestive heart failure, the seasoned medic proposed an alternative diagnosis of a brain hemorrhage. Other signs, including poor response to questioning by paramedics and an inability to follow commands, leads the experienced medic to redirect the treatment toward this alternative diagnosis. Those therapies which were initially appropriate for a CHF patient but inappropriate for a potential brain bleed were discontinued, and other protocol-driven care for the new diagnosis were started.

The narrator, who noted that he "... been a medic at that time not even 12 months ... [and] was still relatively new," reflected on those things that were initially influential and his performance on the call.

   Everyone in the world is screaming at me that he just was seen in
   the hospital for his CHF and he has labored breathing. He looks
   [terrible] and you start thinking, okay, I think I hear some
   crackles but I was not sure but we just started going down that
   route and then you don't take into account the other things that
   you see. It is easy to get tunnel vision. You've gotta be able to
   step back sometimes.

He emphasized several times throughout his telling of the story that the patient's history of CHF thoroughly saturated the messages coming from family members and bystanders. This consistent exposure to this hypothesized reason for the patient's distress, as well as some objective verification in this assessment of the patient's lungs, was the primary reason that the medic did not consider other possible causes. Reflecting on this, he emphasized that sometimes you have to "step back" to get a better picture of the whole situation. Other signs that the first diagnosis was incorrect were initially ignored, minimized, or explained through other possible causal processes. Important in this case was the need to "... trust your clinical findings ..." and incorporate those findings with information provided by bystanders.

Bystanders were important in that they had either access to information about the patient or some other knowledge about the patient's condition that was helpful to paramedics to make informed decisions. These individuals were influential in either the information they brought to light or the nature of the means of communicating that information to paramedics. The paramedics in this case enacted a particular clinical protocol based on the initial information presented, and then changed course as they compared the clinical findings to the policies in place to guide clinical behavior.

In other instances paramedics derived information from other areas that influenced their decisions. Specifically, paramedic assessments of patient identity served to communicate information on probable patient condition and appropriateness of therapies. Although the accuracy of these more shallow judgments was potentially questionable, this type of guidance is important to discuss in considering the full range of influences on paramedic behavior.

The Impact of Patient Identity on Treatment

Patient identity, and the related concept of patient need, can serve to influence paramedic decision-making and behavior. Patient identity in this case refers to the ascribed characteristics perceived by a paramedic as important during a service interaction, including some combination of race, gender, socioeconomic status, and potentially other variables. This package of variables, and the resulting assessment of the patient in light of perceived patient identity, serves to communicate to the paramedic the appropriateness or inappropriateness of specific therapies based on their conceptions of probable patient need. In essence, the translation of an identity into a set of possible causes for clinical presentation may result in differences in treatment options and paramedic behavior.

Describing a situation in which identity was a factor in patient assessment, one paramedic noted that the patient's age was influential in causing him to rule out one specific cause of the patient's condition. Though the paramedic conducted an in-depth assessment of the patient and evaluated several potential causes for her unconscious state, the correct cause and appropriate treatment were ruled out due to the patient's age.

   [We had] a lady who was in her 60s who was unresponsive. I remember
   it was a holiday, because I remember the fact that the whole family
   was in the house, and everyone was really upset about ... grandma.
   And the family's frantic. ... And we ran the call for the
   unresponsive, we ... did all of the stuff that you would do for
   unresponsive. The family said she went to the bathroom and
   collapsed.

   And I remember getting to the hospital, and the nurse - who was
   very good - looks at her, and runs through it all and goes, "Did
   you try Narcan?" [The patient is] 65 years old ... two [milligrams]
   of Narcan, [and] she sits up in the bed. It was a heroin overdose.
   Never in my wildest imagination. Tunneled vision'd on the problem
   has to be from the diabetic, renal, something, that line, [maybe a]
   stroke. Forgot the basics ... open her eyes, and take a really
   decent look. You probably would have saw it. I was fairly young
   when that happened.

   [I was] just feeling [foolish] when that was all done. Stick to
   your basics, do what you should have done, which was to roll them
   all out. Had I really thought a little bit more about it, the fact
   that she had a liver problem from Hep C, which was probably from
   the habit. I just didn't draw all of the dots together. Probably
   same call, 30 year old, I would have dumped the Narcan second
   without thinking twice about it. Instead, I'm thinking old lady
   with a lot of problems. Could be an ammonia level, could be all
   kinds of stuff. None of the above, it was the easiest thing.

The treating paramedic's initial impression of the patient's condition was that it was potentially caused by a handful of clinical presentations, including low blood sugar, renal failure, or a stroke. A condition not considered, that of a drug overdose, was eventually identified as the cause upon arrival at the hospital. In relating the story, the paramedic was particularly disappointed with his diagnostic skills and lack of attention to all possible causes of the patient's condition. He noted that he "[f]orgot the basics ..." and should have simply checked the patient's pupillary response, and indicator of a possible overdose. The medic also notes, rather quickly, that he was a "younger medic," thus potentially indicating that his patient assessment abilities were not as refined and his experience treating unconscious patients. His treatment should have been, in his words, to "... roll them all out," meaning that he would have employed all possible skills and resources to treat the condition.

Perhaps the most important aspect of this story is the medic's retrospective statement about how he would treat a much younger patient with the same clinical presentation. He states confidently that on the "... same call, [with a] 30-year-old, I would have dumped the Narcan second without thinking twice about it." A 30-year-old unconscious patient would have received the necessary medication regardless of other indicators. Instead, because the patient was of an advanced age, his considerations for the probable causes of unconsciousness were different. Thus, the patient's age, and various assumptions about the patient's identity wrapped up in the paramedic's assessment of age, resulted in an inappropriate course of treatment.

IMPLICATIONS FOR THEORY AND PRACTICE

Several implications for theory emerge when considering these incidents together. First, this study suggests support for those theories of street-level bureaucracy that place occupational and professional culture at the fore. Professional norms that focus on comprehensive patient assessment become that much more important given the potential consequences of failing to adequately gather information. Paramedics working in professional cultures that encourage thorough patient interviewing techniques and verification of patient condition through multiple lines of questioning will be less likely to under- or misdiagnose a patient's condition. Adding to this, the professional norms and routines emerging from concrete experiences become that much more salient when the information gathering process is made difficult by situational characteristics. Prior experiences treating patients in clinically complex situations will improve abilities to recognize specific cues that indicate the appropriateness of certain therapies. The results presented here indicate that strong occupational and professional norms surrounding both communication with clients and attention to clinical cues can serve to mitigate potential hazards associated with service provision in complex, vague, and uncertain situations. And, importantly, this suggests the need for additional research to further unpack the nature of occupational culture in EMS.

Second, this study supports and enhances theories of street-level bureaucracy that emphasize the importance of patient identity in directing front-line worker behavior. Paramedic perceptions of patient identity were strongly tied to the imaginable realm of patient need. Ascribed identity, then, became an important means of ruling out differential diagnoses and directing clinical behavior. This also places more emphasis on client identity as an important influence for those occupations where client need is less observable. As noted, EMS is unique in that client needs are rooted in often-unobservable physiological and psychological conditions, making this finding particularly important. Though important for EMS, this is broadly applicable to many front-line public services occupations in which street-level bureaucrats are charged with sorting through complex and varied sources of information to gauge the nature and legitimacy of a client's specific needs in time-bound and uncertain situations.

Implications for practice become more evident when considering the incidents individually. In the first incident, paramedic communication skills emerged as key in discovering important aspects of a patient's past medical history, aspects of the current incident that give clues that may aid in sorting through differential diagnoses, and other contributing factors that may serve to improve the process of selecting the right medications and procedures for a particular patient. The paramedic's motivation to continue pressing the patient for more information about her specific symptoms until the crucial information emerged allowed both the most appropriate treatment to be provided to the patient and early hospital notification to occur. Interestingly, the aforementioned motivation is, in and of itself, discretionary behavior. The practical implications of this finding point to the importance of training paramedics to both look for cues which may signal a diagnosis other than that provided by the patient during the initial assessment, and to engage in alternative means of probing for important information.

In the second case, bystanders with intimate knowledge of the patient's medical history or other situational factors were influential in shaping diagnosis and treatment. The information provided directly by family members and bystanders served to push paramedics toward a specific diagnosis, one that was inappropriate for the patient's actual condition. Though an experienced paramedic eventually realized the misstep in diagnosis and treatment, there was nevertheless a delay in administering the appropriate clinical therapies to the critically ill patient. Considering this finding, this story highlights the practical importance of matching more objective clinical assessments, including those that are targeted at confirming or challenging the paramedic's initial impression of patient condition, with the information provided by bystanders or family members who may or may not have clinical expertise.

In the final case, a paramedic's assessment of the patient's identity, essentially his or her evaluation of need or worthiness for specific services based on a collection of ascribed characteristics, was influential. In the story of the unconscious elderly woman it was clear that the paramedic's assessment of the patient's identity was a driving force in the establishment of a clinical diagnosis. In particular, the patient's age provided a cue to the paramedic that it was appropriate to eliminate a specific diagnosis of a drug overdose. The paramedic ruled out this diagnosis, and some key assessment findings that would have pointed to the correct diagnosis, given past experiences with patients of the same age. The practical implications of this finding point to the need for paramedics to strive to incorporate both objective and subjective information on patient condition into the assessment process and corresponding treatment plans. This intense need to gather and process information on patient need will likely need to be an ongoing process, one which continually compares clinical diagnosis with data gathered during the service interaction.

LIMITATIONS AND FUTURE RESEARCH

Several limitations of this study are worthy of mention. First, this study reports only a handful of service interactions that illustrate concepts and relationships of interest. Though notable and certainly contributing to our knowledge of service interactions in front-line emergency medical care, these incidents do not represent a full body of interactions relating to the focus of this research. Future research should investigate additional incidents that offer support or challenge the relationships described here, with emphasis on gathering data that allows for generalization. However, given the exploratory nature of this study, this is an acceptable limitation.

Second, participating organizations and subjects were not chosen randomly, and do not represent the full universe of EMS providers. The substantial variation in clinical and operational rules, populations served, organizational arrangements, training, and other important variables found in the field of EMS does not easily allow for such generalizations. Examination of EMS in urban, suburban, and rural contexts, working in public, private, and nonprofit organizations, and with varying levels of training and experience should is an important next step.

Third, the use of narrative analysis has some inherent weaknesses. As Maynard-Moody and Musheno (2003) noted, "[s]tories are not facts or evidence waiting for interpretation; from the moment they are conceived through the many telling and retellings, they are the embodiment of the story-tellers interpretation" (p. 26). These stories do not, then, constitute objective data. Additionally, important data on embarrassing incidents, or those that are not legally, socially, or culturally acceptable, will generally not be revealed, thus "... perpetuat[ing] the conspiracy of silence" (Maynard-Moody & Musheno, 2003, p. 32). Future research should employ data collection methods such as observation or participant observation with follow-up interviews to probe for causal relationships from the perspectives of street-level EMS workers.

CONCLUSIONS

This exploratory research examining accounts of service interactions is illustrative of the potential complexity of the foundational act of patient processing. Analysis of narratives of action from front-line paramedics indicates that several types of influence may contribute to the complexity of these service interactions. Each of the incidents reviewed here, in which paramedics were presented with incomplete information by patients, family members, or through their own judgments of ascribed patient characteristics, is demonstrative of difficult or failed information gathering efforts in front-line EMS. In each of the stories related, some aspect of the service interaction either caused the paramedic difficulty in determining which treatments were appropriate or provided cues that pointed the paramedics in the wrong clinical direction.

The results of this study highlight the importance and impact of paramedic communication skills, influential bystanders with key information, and patient identity on the intrinsically difficult situations common to street-level patient processing. Though this does not constitute a fully developed and complete list of sources of influence on front-line EMS providers, these and other influential factors represent important considerations worthy of scholarly discussion and attention. Given the impact of EMS on both short-term patient care and long-term patient outcomes, additional research building on this study is necessary to more fully understand patient assessment and determination of clinical need during complex and uncertain service interactions.

REFERENCES

American Ambulance Association (2009). Ambulance facts. Retrieved November 10, 2009 from: http://www.theaaa.org/medi a/ ambul ancefacts .html.

Bailey, P.H., & Tilley, S. (2002). Storytelling and the interpretation of meaning in qualitative research. Journal of Advanced Nursing, 38(6), 574-583.

Brodkin. E.Z. (2011). Policy work: Street-level organizations under new managerialism. Journal of Public Administration Research and Theory, 21(Supplement 2), i253-i277.

Cain, V.S., & Kington, R.S. (2003). Investigating the role of racial/ethnic biases in health outcomes. American Journal of Public Health, 93(2), 191-192.

Charmaz, K.C. (2005). Grounded theory in the 21st century. In N.K. Denzin & Y.S. Lincoln (Eds.), Handbook of Qualitative Research Methods (3rd Ed.)(pp. 139-164). Thousand Oaks CA: Sage Publications, Inc.

Connelly, F.M., & Clandinin, D.J. (2006). On narrative method, personal philosophy, and narrative unities in the story of teaching. Journal of Research in Science Teaching, 23(4), 293-310.

Davis, K.C. (1969). Discretionary justice: A preliminary inquiry. Baton Route, LA: Louisiana State University Press.

Department of Labor (2009). Occupational employment and wages, May 2008. Retrieved December 19, 2009 from: http://www.bls.gov/oes/current/oes292041.htm

David, G., & Brachet, T. (2009). Retention, learning by doing, and performance in emergency medical services. Health Research and Educational Trust, 44(3), 902-925.

David, G. & Chiang, A.J. (2009). The determinants of public versus private provision of emergency medical services. International Journal of Industrial Organization, 27, 312-319.

David, G., & Harrington, S.E. (2010). Population density and racial differences in the performance of emergency medical services. Journal of Health Economics, 29, 603-615.

Dimock, M.E. (1936/1967). The role of discretion in modern administration. In J.M. Gaus, L.D. White, & M.E. Dimock. The frontiers of public administration. New York, NY: Russell & Russell.

Dworkin, R. (1977). Taking right seriously. Cambridge, MA: Harvard University Press.

Fiscella, K. (2004). Socioeconomic status disparities in healthcare outcomes: Selection biases or biased treatment? Medical Care, 42(10), 939-942.

Gilboy, J.A. (1992). Penetrability of administrative systems: Political "casework" and immigration inspections. Law & Society Review, 26(2), 273-314.

Goodsell, C.T. (1981). The public encounter and its study. In C.T. Goodsell (Ed.) The public encounter (pp. 3-20). Bloomington, Indiana: Indiana University Press.

Handler, J.F. (1986). The conditions of discretion: Autonomy, continuity, bureaucracy. New York, NY: Russell Sage Foundation.

Hill, D. (1974). The art of "muddling through": Policymaking and implementation in the field of welfare. Madison: Univ. of Wisconsin Press.

Hill, M., & Hupe, P. (2003). The multi-layer problem in implementation research. Public Management Review, 5(4), 471-490.

Hupe, P., & Hill, M. (2007). Street-level bureaucracy and public accountability. Public Administration, 85(2), 279-299.

Institute of Medicine (2007). Emergency medical services: At the crossroads. Washington, DC: The National Academies Press. Retrieved October 27, 2009 from: http://books.nap.edu/openbook.php?record_id=11629

Isett, K.R., Morrissey, J.P., & Topping, S. (2006). Systems ideologies and street-level bureaucrats: Policy change and perceptions of quality in a behavioral health care system. Public Administration Review, 66(2), 217-227.

Keiser, L.R. (1999). State bureaucratic discretion and the administration of social welfare programs: The case of social security disability. Journal of Public Administration Research and Theory, 9, 87-106.

Kelly, M. (1994). Theories of justice and street-level discretion. Journal of Public Administration Research and Theory, 4(2), 119-140.

Lightner, L., Brywczynski, J, McKinney, J., & Slovis, C.M. (2010). Shortness of breath: Prehospital treatment of respiratory distress. Journal of Emergency Medical Services, 35(4), 57-61.

Lipsky, M. (1980). Street-level bureaucracy: Dilemmas of the individual in public services. NY: Russell Sage Foundation.

Loyens, K., & Maesschalck, J. (2010). Toward a theoretical framework for ethical decision making of street-level bureaucracy: Existing models reconsidered. Administration & Society, 42, 66-97.

May, P.J., & Winter, S.C. (2007). Politicians, managers, and street-level bureaucrats: Influences on policy implementation. Journal of Public Administration Research and Theory, 19, 453-476.

Maynard-Moody, S. & Musheno, M. (2003). Cops, teachers, counselors: Stories from the front-lines of public service. Ann Arbor, MI: The University of Michigan Press.

Narad, R.A., & Driesbock, K.R. (1999). Regulation of ambulance response times in California. Prehospital Emergency Care 3(2), 131-135.

National Association of State Emergency Medical Services Officials (2011). 2011 National EMS Assessment. Retrieved February 10, 2012 from: http://ems.gov/pdf/2011/National_EMS_Assessment _Final_Draft_12202011.pdf

National Highway Traffic Safety Administration (2001, December). National EMS research agenda. Retrieved October 28, 2009 from: http://www.nhtsa.gov/people/injury/ems/ems agenda/EMSResearchAgenda.pdf

Newman, M.A., Guy, M.E., & Mastracci, S.H. (2009). Beyond cognition: Affective leadership and emotional labor. Public Administration Review, 69(1), 7-19.

Oberfield, Z.A. (2010). Rule following and discretion at government's frontlines: Continuity and change during organizational socialization. Journal of Public Administration Research and Theory, 20(4), 735-755.

Pennsylvania Department of Health Bureau of Emergency Medical Services (2009). 2008 annual report. Retrieved December 19, 2009 from: http://www.portal.state.pa.us/portal/server.pt?open= 18&objID=445347&mode=2

Prottas, J.M. (1978). The power of street-level bureaucrats in public service bureaucracies. Urban Affairs Review, 13, 285-312.

Riccucci, N.M. (2010). Public administration: Traditions of inquiry and philosophies of knowledge. Washington, DC: Georgetown University Press.

Riccucci, N.M. (2005). How management matters: Street-level bureaucrats and welfare reform. Washington, DC: Georgetown University Press.

Salerno, S.M., Wrenn, K.D., & Slovis, C.M. (1991). Monitoring EMS protocol deviations: A useful quality assurance tool. Annals of Emergency Medicine, 20(12), 1319-1324.

Sandfort, J.R. (2000). Moving beyond discretion and outcomes: Examining public management from the front lines of the welfare system. Journal of Public Administration Research and Theory, 10(4), 729-756..

Scott, P.G. (1997). Assessing determinants of bureaucratic discretion: An experiment in street-level decision-making. Journal of Public Administration Research and Theory, 7(1), 35-57.

Setzler, H., Saydam, C. & Park, S. (2009). EMS call volume predictions: A comparative study. Computers & Operations Research, 36, 1983-1851.

Todd, K.H., Deaton, C., D'Adamo, A.P., & Goe, L. (2000). Ethnicity and analgesic practice. Annals of Emergency Medicine, 35(1), 11-16.

Todd, K.H., Samaroo, N., & Hoffman, J.R. (1993). Ethnicity as a risk factor for inadequate emergency department analgesia. Journal of the American Medical Association, 269(12), 1537-1539.

van Ryn, M., & Burke, J. (2000). The effect of patient race and socio-economic status on physicians' perceptions of patients. Social Science & Medicine, 50, 813-828.

Vinzant, J.C., & Crothers, L. (1998). Street-level leadership: Discretion and legitimacy in the frontline public service. Washington, DC: Georgetown University Press.

Weatherly, R. (1979). Reforming special education: Policy implementation from state level to street level. Cambridge, MA: MIT Press.

West, E. (2001). Management matters: The link between hospital organisation and quality of patient care. Quality in Health Care, 10, 40-48.

Yin, R.K. (1994). Case study research: Design and methods (2nd Ed.). Thousand Oaks, CA: Sage Publications.

Table 1
Characteristics of Selected Organizations

       Calls /    Total    Total    Area    Pop. Per   Median
Org.     Yr.       Emp.    Pop.     (sq.     Sq. Mi.   HHI ($)
       (2009)    (FT/PT)             mi.)

FDEM   10,000    20 / 30   75,000    20      4,999     45,000
S
HSEM   11,000    20 / 14   95,000    10     10,000     50,000
S
PDEM   15,000    29 / 20   115,00    20      7,000     35,000
S                               0

Source: U.S. Census, 2010; figures
summed for multiple municipalities
and rounded to de-identify
organizations; FT denotes full
time employees, PT denotes part
time employees; HU denotes housing
units; HHI denotes household income

Table 2
Characteristics of Interviewees

                        Median    Range

Age                      35.5     25   57
Years as a Paramedic       10    0.5   31
Years in Organization       6    0.5   30

Gender                   Female-4; Male-26