In-hospital cardiopulmonary resuscitation (CPR) requires a multi-professional team to assemble efficiently and provide coordinated care under time pressure with potentially high stakes outcomes.1 Effective leadership is critical.2,3 Ineffective leadership4,5 has been linked to CPR interruptions, delayed defibrillation, and delayed administration of epinephrine, any of which can decrease a patient’s likelihood of survival.6 The 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care recommend specific training in teamwork and leadership skills.7
In a review of the literature on leadership during CPR, Hunziker and colleagues noted that individual traits, like gender, affect the status of team members and might be important in code team dynamics so deserve further exploration.2 In most teaching hospitals, internal medicine residents in their second and third years (postgraduate year [PGY] 2 and 3) are expected to learn to lead CPR (“codes”). Women comprise approximately 44% of internal medicine residents8 and, despite a large body of research on gender and leadership,9–11 few studies have examined how gender influences code leadership experiences. In a thorough review of the literature on leadership in health care teams, Künzle and colleagues included no mention of gender.3 Wayne and colleagues assessed resident performance leading 64 actual CPR events and found no gender difference.12,13 Others studied only untrained medical students during simulated codes.14,15
We undertook the present study to explore internal medicine residents’ perceptions of and experience with leadership during codes. Specifically, we examined what behaviors internal medicine residents associate with effective code leadership and how gender influences their experiences.
ResultsOf the 30 residents invited to participate, 25 (83%) from 9 residency programs agreed (see Table 1). Of those who did not participate, four did not respond to the invitation emails and one declined due to time constraints.
We identified a total of 81 unique content areas and organized the data into five themes: (1) learning to be a code leader, (2) leadership defines code success, (3) ideal code leadership behaviors, (4) leadership and gender, and (5) integrating conflicting identities. In this report, we address the last four themes.
Leadership defines code successMost participants acknowledged that codes had “pretty dismal prospects.” In reflecting back, both male and female participants described the experience as successful if the code was controlled and had effective leadership. One participant said that he thought “less about the actual patient outcome” and more about his “control” of the room. Another described a code that ended when the “patient was getting better and was wheeled away” but, because the room was chaotic, said it “wasn’t a good code.” Poor leadership was consistently cited as the major contributing factor to a chaotic code: “if you’re quiet and wishy-washy, it just goes horrifically.”
Ideal code leadership behaviorsBoth male and female participants described the ideal code leader as a person who has an authoritative presence; speaks with a deep, loud voice; uses clear, direct communication; and has emotional control. The uniform advice to “assert yourself” was given by both male and female participants, who thought that “the most important thing” was to “establish yourself as a leader.” Participants overwhelmingly reported that it was important to be “authoritative”; to have a “strong presence” by being both seen and heard. One participant established her role as a leader by being “the focal point in the room…you know that I’m running the code because I’m the one who’s talking and I’m the one who’s standing at the foot of the bed and telling people what to do.” Others also mentioned the importance of being “tall.”
The code leader also must be “the main voice in the room” by “speaking in a very loud, calm, controlled voice.” Both male and female participants described raising the volume of their voice and being cognizant of their tone, choosing an “assertive version.” One participant said that he tried to have “authority” in his voice and would “spit out orders.”
Participants explained that the code leader should be the “chaos control factor” and project a composed demeanor to “help calm other people down.” They also described often being “panicked” but trying to appear as the “calmest person in the room.” One likened his behavior to a duck, saying: “If you’ve ever watched a duck on a pond, it looks as though it’s floating effortlessly across the lake. But if you’ve ever looked underneath at a duck’s feet, they’re paddling feverishly.”
Leadership and genderAll participants thought that men and women were equally effective leaders. One male PGY-3 participant remarked, “I can think of women that are just as assertive as any men when they’re running codes.” Both male and female participants described the same ideal leadership behaviors and described their struggle to achieve them at times. For example, the same male PGY-3 participant then said: “I’m not by nature extremely assertive…I don’t really like ordering people to do things, so it took me a little while to just get comfortable.” A few male participants commented on the need to act more assertively during codes, but largely female participants expressed this discomfort: “I just try my best to look authoritative…but it’s stressful” (female PGY-3 participant).
Although both genders perceived leader performance in a similar manner, several participants noted that adopting leadership behaviors was easier for men. One female PGY-3 participant observed that “tall men with a deep voice may naturally appear more authoritative” and that male residents have it “easier.” A male PGY-3 participant noted his advantage, saying:
Anyone who tells you that being a white male with a deep voice and who’s a little bit taller is not an advantage to being perceived positively, or perceived as in control, would be lying, in really any situation not just a code.
A female PGY-2 participant said:
I act differently during a code than my normal day-to-day behavior…the male residents do not have to alter their behavior quite as much…and [they] still command authority…In addition to remembering the ACLS algorithm and everything else you have to do during a code you’re also trying to assume this persona of being in charge and I think that’s probably a little more stressful [for women].
Both male and female participants described using directive language in leading codes, but women more than men felt uncomfortable with that communication style. One female PGY-3 participant said: “women are always afraid that they’re going to come across witchy with a ‘b’”. Almost half of the female participants described their apprehension in appearing “bossy” when leading codes, whereas no male participants expressed this concern. A female PGY-2 participant noted: “it can come off as bossing people around. Whereas I never felt that way when I see a male run a code; it just seems natural.” A few female participants talked about feeling “weird” or “awkward” about adopting such directive behavior. For example, a female PGY-3 participant said: “you just aren’t sure if people’s feelings are going to get hurt or if they’re going to be mad about it.”
Generally, the participants did not mention race and ethnicity, even when asked whether gender, race, or other factors might influence leadership. Of the five participants born outside the US, we found little difference in their ideal leadership behaviors. However, one female PGY-2 participant, born outside the US in a culture where gender norms are explicitly reinforced, was unable to abandon her normal communication style in codes. She explained: “I don’t give orders, I cannot. I just cannot. I find it very awful.”
Integrating conflicting identitiesFemale participants described strategies they individually discovered to help mitigate discomfort and provide confidence in assuming code leadership. For some, the physical indicators of institutional authority were important: “wearing your long coat, having a badge that says ‘resident’” (female PGY-2 participant). In addition, some noted that the shared knowledge that they were the institutionally designated CPR leader because they were carrying the “code pager” was affirming. Several female participants talked about the importance of mentally preparing and “assuming a code persona” when on call. One female PGY-3 participant said:
I do have to prepare mentally beforehand. I am a woman, my voice is not deep and I’m not tall or authoritative-looking…I have to mentally prepare to talk in an authoritative way so people respect and listen to me.
Physical signals helped some female participants look and feel more in control. One female PGY-3 participant prepared to lead by carrying a hair tie when she carried the code pager. At the beginning of a code, she would “put [her] hair back and just stand there and just order people around.” To her, putting her hair back meant: “I’m all business here and nothing is going to get in my way.” Several participants described adopting a powerful posture and would stand with their “chest pushed out” or arms crossed. One PGY-2 participant said that it was important that she assumed a wide “code stance” gripping the patient’s bed and leaning over it. A few female participants would ask to stand on a stool; one female PGY-2 participant said she will “stand on my toes to bring the focus to me.”
Both male and female participants talked about how the urgency of codes requires them to act in a directive, assertive manner. Most female participants felt able to suspend their usual “niceties” during codes, saying: “in that room it’s fine.” However, one female PGY-3 participant said that she is “super apologetic afterward”; she felt the need to tell one team member “I’m so sorry I was mean to you!” after leading a code. Knowing that she would apologize afterward allowed her to act uncharacteristically assertive during codes.
Female participants also acknowledged that their advice to other women leading codes would be different than their advice to men. They emphasized the intersection of gender and personality, and one female PGY-3 participant said: “the big thing is if you are a shy female, you’re probably going to have the most trouble.” Another said: “just be really encouraging [to] women to feel ok about dropping the ‘pleases’ and ‘thank yous.’” One suggested: “tell her to wear a ponytail holder.”
DiscussionThe internal medicine residents in this study confirmed that effective leadership is critical to a successful code.2,6,21,22 While a wealth of research has found little if any difference in the effectiveness of male and female leaders,9,23–25 gender has not been evaluated specifically in the task-oriented, time limited, urgent setting of a code.3 Our study therefore adds to the literature on code leadership by directly exploring gender. The male and female residents in our study expressed similar levels of confidence in their own abilities to lead, and many had observed effective code leadership enacted by both men and women, consistent with the findings of Wayne and colleagues.13
Cultural stereotypes about men and women are well known, even among those who do not endorse their content.26 The ubiquitous reinforcement of these stereotypes throughout life leads to implicit prescriptive rules about how men and women should and should not behave.10,11 Specifically, “agentic” behaviors, including being assertive, dominant, forceful, and acting as a leader, are socially more acceptable for men, while “communal” behaviors, including being cooperative, warm, soft-spoken, and polite, are socially more acceptable for women.26,27 Both men and women in our study described the ideal code leader in highly agentic terms--as a person who has an authoritative presence; speaks with a deep, loud voice; and uses clear, direct communication. In the literature, others also have described effective leadership in life-threatening situations in agentic terms. Bergs and colleagues,28 for example, found that the code leader needs to “take command and give direction,” and Yun and colleagues29 found that team members preferred directive (agentic) leadership in cases of major, life-threatening trauma. Cooper and Wakelam decided that the “consideration” (communal) aspect of leadership was relatively unimportant in a code situation compared to the more directive “initiating structure” leadership style and that poor leadership resulted from failing to “take command and give direction.”6
Figure 1 provides a conceptual presentation of our findings. Although a few male residents commented on how the need to become louder and more assertive during a code was contrary to their nature, more often women expressed the need to modify their “normal day-to-day” behavior to adopt the more directive and agentic behaviors required to enact code leadership. In adopting these agentic behaviors, female residents frequently expressed discomfort in violating prescriptive gender norms. In addition to receiving repeated admonishments to conform to the female gender norms of being “nice,” “polite,” and not too “bossy,” women learn from experience that they can suffer social penalties for engaging in behavior that is too agentic.10,30,31 Koch found that, even when viewed as more competent than men in leading a small task-oriented group, women elicited more non-verbal displays of negative affect from other team members.32
Rudman and colleagues named the social censure women receive from counter stereotypical behavior as “backlash” and noted that women express a fear of this backlash.33,34 The mental model of a leader is of a man35; thus, simply acting as a leader may be enough to trigger female residents’ fear of backlash, especially when forced to lead in a directive manner as they are during a code.36 In our previous study of internal medicine residents, we found that women frequently expressed concern about behaving in counter stereotypical ways by giving directive orders and that even men acknowledged that women had more constrained communication styles they could use for directing patient care.19 We were not surprised that female residents again expressed this concern given their need to behave in a highly agentic manner to lead a code. Many residents spoke about increased anxiety before codes, a numb feeling during codes, and anxiety afterwards that some described as “PTSD” like. However, only women voiced stress from needing to behave in the agentic, stereotypical male manner required for code leadership.
Our findings align with a large body of empirical research and are particularly important because we have uncovered strategies that female residents found useful to help integrate the conflicting identities of gender and code leader. Many of these strategies provided psychological reassurance that women could suspend female gender behavioral norms when leading a code. Some of these strategies find empirical support from research on embodied cognition, which argues that motor actions influence cognition and emotional states. In an experimental study manipulating body posture, Carney and colleagues37 found that participants who assumed a powerful versus a non-powerful posture felt more confident and were more likely to take action. Thus, the “code stance” that many residents described may be valuable in the same way, as a means of increasing confidence. In addition, Adam and Galinsky found that the symbolic meaning of wearing the white coat of a doctor improved cognitive performance.38 The benefit of what they referred to as “enclothed cognition” may be reflected in residents’ descriptions of the symbolic importance of holding the code pager, wearing a white coat, or tying back one’s hair in assuming code leadership.
Our study was exploratory and relied on the recollected lived experiences of the participants. Thus, the generalizability of these themes deserves further study. However, experimental research in social psychology supports that women would experience stress when engaging in counter stereotypical behaviors, such as leading a code, and that the strategies the resident’s identified would be helpful. Therefore, our study has implications for teaching CPR leadership to internal medicine residents. Training should include an “identity safety” statement reinforcing that both male and female residents are equally competent to lead codes.39 In other settings, providing information about how gender stereotypes might cause anxiety in male-gendered tasks has improved performance40 and appears to promote leadership self-efficacy in women at early career stages in science, engineering, and medicine.41
We recommend that future resident training in code leadership:
Provide clear affirmation that research finds no difference in the effectiveness of male and female code leaders,
Acknowledge that directive code leadership generally requires a greater departure from cultural gender behavioral norms for women than men, and
Describe some strategies (and the supporting research evidence) that have helped other women suspend gender expectations and adopt highly agentic behaviors while leading codes.
In summary, our findings support the importance of effective leadership in CPR. Male and female residents both described code leadership in highly agentic terms and acknowledged that enacting code leadership required female residents to make greater adjustments than male residents to their normal behaviors. Some female residents found such counter stereotypical behavior stressful. To mitigate this stress, female residents frequently adopted intentional rituals to give themselves permission to suspend gender role expectations while leading a code. Because of the importance of effective code leadership for team performance, our findings have implications for improving patient outcomes.