Considerable interest and concern regarding stress experienced by nurses has been generated over the past decade. Stressors associated with the conflictual, complex, and demanding world of clinical nursing have taken their toll on the physical and psychological well-being of its practitioners (National Institutes for Occupational Safety and Health, 1978). Stress not only has a negative effect on the well-being of nurses, but the quality of their nursing care capabilities are adversely affected. The influence of extreme stress/anxiety on ineffective performance (Hebb, 1955; Malmo, 1958; Schlosberg, 1954) and mental inefficiency (Berkun, Bialek, Kern, & Yagi, 1962; Osier, 1954) has been well documented. In a recent study, Jamal (1984) found a negative linear relationship between work stress and job performance, motivation, and patient care in a large sample of nurses.
Of particular concern to researchers in the areas of stress and nursing has been the nature of Stressors encountered by intensive care nurses. In a selected review, Jacobson (1983) pointed out there was agreement in the findings of the major studies related to sources of stress for intensive care nurses. Interpersonal conflicts, the nature of direct patient care, inadequate knowledge and skill, and the physical work environment emerged as major categories of Stressors in three of the most extensive studies (Claus & Baily, 1980; Huckabay & Jagla, 1979; Jacobson, 1978).
Jacobson (1983) suggests that since the Stressors affecting intensive care unit nurses are known, attention now needs to focus on modifying changeable Stressors and facilitating the coping skills of nurses. The latter will serve to reduce the negative impact of inevitable Stressors (e.g., interpersonal conflicts, negative aspects of the work environment). Although stress management self-help articles and stress reduction/coping skills groups have proliferated recently, research efforts testing the efficacy of such programs with nursing samples are sparse. There is also little reported research in the area of prevention-oriented programs, i.e., programs aimed at teaching nursing students methods of coping with work-related stress. It seems critical to investigate intervention strategies that focus on helping nursing students prepare themselves for the often unavoidable Stressors of their future work.
We have an interest in evaluating the efficacy of several coping skills training programs to be used with baccalaureate nursing students. Our initial methodological concern was to find a context in which coping skills could be applied and evaluated. Two common settings for the evaluation of training methods are clinical sites and laboratory situations. Advantages and disadvantages to evaluating the outcome of coping skills training in a clinical setting versus a laboratory setting were considered. Although clinical studies maximize realism and enhance the ability to generalize results, such gains are made at the cost of measurement precision and internal validity. Laboratory experiments, on the other hand, maximize precision and variable control at a cost of realism and external validity (McGrath, 1970).
As a compromise, we decided to use an experimental simulation strategy because it appears to combine the positive features of both field (clinical) studies and laboratory experiments. According to McGrath (1970), this strategy differs from laboratory experiments in that it is a deliberate attempt to recreate or simulate "real-life" situations in a concrete, realistic manner whereas laboratory experiments represent an abstraction of a real-life situation. Experimental simulations differ from field studies in that the setting is more controlled, resulting in more effective control of intervening variables and greater precision of measurements. Simulations have the additional advantage of serving as a training methodology as well as in research applications.
Development of the Simulations
A cardiac arrest situation was selected for simulation because it is potentially one of the most stressful circumstances confronting nurses. The life and death emergency nature of the crisis in and of itself may produce stress. Urgent time demands, aversive environmental conditions, and interpersonal confrontations with other health care providers may also create stress. Accurate perceptual skills, effective and efficient problem-solving ability, and the performance of complex psychomotor tasks are required of nurses in this stressful situation.
Prior to developing the actual simulation, a sample (N = 40) of senior baccalaureate nursing students responded to a questionnaire wherein they were asked to project their level of anxiety in response to 35 potential cardiac arrest ("Code Blue") Stressors. The Stressors identified by the students as most potentially anxiety-evoking were similar to those identified in previous studies by nurses practicing in intensive care unit settings ( Jacobson, 1983 1. An analysis of the student responses indicated that the most uncomfortable Stressors could be clustered under three categories: (a) health condition of the patient, e.g., the patient's condition worsens; <b) interpersonal Stressors, e.g., doctors arguing, no one clearly in charge, family member arrives and is anxious and demanding; and (c) miscellaneous environmental Stressors, e.g., faulty equipment, eversive sounds, aversive odors. These clusters of Stressors became the basis upon which the stressful simulations were structured. Both simulations included a simulated patient (mannequin) whose condition was reported to be rapidly deteriorating. The environmental and interpersonal variations of the cardiac arrest situation were developed to include each of the Stressors identified by the students as most likely to be anxiety-evoking. Both simulations were developed to include time-urgent demands for the performance of a variety of intensive care tasks.
Purpose of the Study
One of the inherent problems of using a simulation strategy is the question of whether or not participants will respond in the simulation as they would in the "real" situation (McGrath, 1970). The question of subject motivation is obviously quite important when examining a phenomenon such as stress in a less-than-actual situation. The student in a cardiac arrest simulation, for example, knows that if she fails to perform a procedure correctly, no one will actually be endangered. Although the simulations were developed to be as realistic as possible, the question that needed to be answered was whether or not they had the capacity to elicit stressful reactions in the participants. The purpose of this study was to assess the stress-eliciting capacity of two variations of a cardiac arrest simulation on a sample of nursing students.
Subjects: Twenty-seven nursing students in their senior year of a baccalaureate program at the University of Missouri participated in the study. Subjects were recruited from a practicum course. All participants previously had formal instructions in cardiac arrest procedures and had passed performance tests related to this content. Each subject had satisfactorily met the CPR performance criteria of the American Heart Association (AHA) and mastery tests related to preparing intravenous medication.
Procedure: Subjects met with a research assistant who reminded them that the study was an evaluation of the effects of stressful situations nurses might be confronted with in an intensive care setting. Questions were answered and verbal consent was obtained from each subject. Pulse and blood pressure measurements were taken; both State and Trait scales of the STAI were then administered. The Trait scale was administered as a pretest to evaluate the equivalency of the two groups on anxiety proneness. Immediately following the pretesting, subjects participated in one of the two cardiac arrest simulations. Subjects were videotaped during the exercise in order to obtain an objective measure of performance. Immediately following the simulation exercise, pulse and blood pressure measurements again were taken; the State portion of the STAI was administered. Subjects were then debriefed. Questions and discussion related to the simulation exercise were encouraged.
Simulation Setting, Instructions, and Tasks: All subjects participated in ene of two variations of a cardiac arrest situation that required them to perform tasks similar to those undertaken by nurses in intensive care unit settings. Simulations were conducted in a videotape studio equipped as an intensive care unit. In the room were a mannequin on a hospital bed, a fully equipped crash cart, an IV apparatus, a defibrillator, an EKG monitor, and a bedside stand. In both simulations, subjects were instructed to perform identical tasks. All instructions for performing the tasks during the simulation were on audiotape; instructions were piped into the simulation room via speakers.
Immediately upon entering the simulation room, subjects were given the following instructions via a taped recording; "Imagine that you are a staff nurse on a critical care unit. Moments ago you admitted a 35-year-old woman with complaints of chest pains. During her admission process, she grasps her chest and reports that the pain is worsening. Her color is pale, her lips are blue. She stops speaking. This is a "Code-Blue" situation. A code has been called. Please listen carefully and follow the instructions given to you as quickly and as accurately as possible. The instructions will be preceded by "Nurse STAT." Whenever you hear "Nurse STAT," follow the instructions as quickly and as accurately as possible."
During the simulation, subjects were asked to perform the following tasks: (a) administer CPR, (b) place the patient on a cardiac monitor, (c) heparinize the syringe, (d) prepare an epinephrine intravenous drip, (e) prepare one ampule sodium bicarbonate, and (f) recall the name and dosages of three medications the subject was ordered to administer (10 seconds following the order).
All tasks were intended to simulate stress reactions. Two of the tasks served a second function; the CPR administration and the memory task also provided performance measures.
Cardiac Arrest Simulation With Environmental Stressors: This simulation included several Stressors common to the environment of an intensive care unit, such as odors, sounds, equipment problems, and patient variables.
1. Odors: Containers of cotton swabs soaked in acetone or alcohol were located around the room to simulate typical ICU odors.
2. Sounds: A tape of actual ICU sounds were piped into the room. Examples of sounds included loud moaning, crying, and delirious verbalizations from patients, suctioning noises, ventilator alarms, phones ringing, and monitor sounds.
3. Equipment malfunction: The syringe used for bicarbonate was prepared so it would not inject properly; the ECG cables did not attach smoothly into the electrode pads.
4. Direct patient care Stressors: There was simulated vomitus ( pea soup ) on the mannequin's lips and Vaseline on the mannequin's chest area to simulate diaphoresis.
Cardiac Arrest Simulation With Interpersonal Stressors: In this simulation variant, four potentially stressful interpersonal situations were portrayed via a tape recording.
1. Physician argument: Two physicians were portrayed as perceiving themselves with equal authority in controlling the emergency team members and disagreeing strenuously on the most appropriate emergency intervention.
2. Angry, impatient physician: A doctor-nurse confrontation depicted the doctor as angry and critical of another nurse for handing him the wrong size tube. A loud crashing noise suggestive of the physician throwing the tube on the floor completes the episode.
3. Ambiguous orders: Within the context of the physician argument scenario, two conflicting orders are given to the subject.
4. Spouse of patient anxious and shouting at nurse: while the subject is attempting to prepare an emergency medication, the husband of the patient shouts at the nurse to hurry, questions her competence, and asserts she will be blamed if his wife dies.
During both simulations, the subject was urged to hurry because the patient's condition was deteriorating. Thus, a greater sense of urgency and increased time pressure was conveyed.
An individual's response to work-related stress has been conceptualized to include three aspects of known functioning: (a) affective, (b) physiological, and (c) behavioral (Beehr & Newman, 1978). To assess the effectiveness of the simulation exercises in creating stress it was necessary, therefore, to use multiple measures that included affective, physiological, and behavioral reactions to stress.
The affective measure used was the ?-State scale of the State Trait Anxiety Inventory (Spielberger, Gorsuch, & Lushene, 1970). State anxiety has been conceptualized by Spielberger et al. (1970) as a transitory emotional state characterized by subjective but consciously perceived feelings of tension and apprehension and heightened autonomie arousal. The ?-State scale consists of twenty statements that ask subjects to indicate how they feel at a particular moment in time.
The physiological measures were pulse rate and blood pressure. Perceived anxiety pulse rate and blood pressure were measured before and after participation in the simulation; pre- post-change scores were the dependent variables.
Performance on CPR administration and a medication memory task were the behavioral measures. CPR performance was evaluated using a standard "check-off' list of 14 discrete motor operations associated with the administration of CPR. Each of the items conformed to the American Heart Association criteria for satisfactory performance of CPR, which all subjects had fulfilled within the previous year. Raters observed the videotaped performance of each subject and evaluated their performance against the rating scale. Post-test only data were collected on these task performance behavioral variables. Pre-tests were not conducted to avoid practice effects. Although all subjects had a common baseline of performance on CPR tasks, we did not consider this as a pre-test score because it was not possible to control for the length of time between when a subject had met the AHA criteria and participation in this study. Instead, the number of items on the list successfully completed was the dependent measure used. In the medication memory task, subjects were asked to recall the correct orders for the administration of three emergency drugs, approximately 10 seconds after the orders were given. The number of correct items, either medications or dosages, was the memory task measure.
An analysis of variance comparison indicated no significant differences in any of the measures of stress between the interpersonal and environmental variants of the cardiac arrest simulation. Since both groups were equivalent on all of the dependent measures, data from both groups were analyzed together.
Pre- and post-means and standard deviations of the affective and physiological measures are shown in the Table. Students t (one-tailed) tests conducted on each of the affective and physiological measures showed several significant changes from pre to post testing. Significant increases in pulse t(26) = 7.62, p < .0001, systolic blood pressure t(26) = 1.96, p. < .03, and state anxiety t(26) = 6.15, g < .0001 were found. No significant change in diastolic blood pressure was found.
An inspection of the ratings on CPR performance showed that 14 out of the 27 subjects performed less than 50% of the required CPR tasks; 18 of the subjects performed less than 66% of the skills adequately. On the medication memory tasks, with a possible range of scores of 0-7, the actual range was 2-6 and the median score was 3. None of the subjects successfully recalled all three medications and their dosages.
The results of this study demonstrated that both variants of the cardiac arrest simulation elicited physiological and affective manifestations of stress in the student nurse subjects. Significant physiological changes included increase in pulse rate and systolic blood pressure. Although it might be questioned if these increases could be attributed to the physical exertion required in the CPR performance task, this is not likely since there was at least a 7-minute time lapse between the CPR activity and post testing. The finding of no significant increase in diastolic blood pressure is consistent with previous stress research. Manuk, Craft, and Gold (1978) found that Type A (coronary prone) subjects exposed to experimental Stressors showed a pronounced systolic response while experiencing decreased diastolic pressure.
A significant increase in self-reported state anxiety was also noted. Even though it was obvious to the subjects that this was not a "real" situation, the participants felt more tension, apprehension, and heightened autonomie arousal than before the simulation experience. The felt anxiety may have in part been due to their awareness that their actions were being videotaped for later observation by the research team, which included their instructor in critical care nursing. In a laboratory experiment this might be considered a confounding variable. In a simulation such as this one, however, this does not present a problem but contributes to the realness of the simulation; in actual cardiac arrest situations there are observers of a nurses performance (e.g., head nurses, supervisors). It should be noted that although there was no attempt to systematically observe and classify non-verbal behaviors during the simulations, most of the subjects demonstrated behaviors commonly associated with anxiety or apprehension, such as rolling of the eyes, grimacing, sighing, nervous laughter, licking of lips, and walking around in circles.
While there were clear increases on the physiological and affective measures, conclusions that might be drawn on performance measures are somewhat limited. At some point during the year preceding participation in the study, each subject satisfactorily met the American Heart Association performance criteria for the administration of CPR, During the simulation, however, only one out of 27 subjects met each of the 14 criteria for satisfactory performance. Whether the dramatic decline in performance was due to the elapse of time and forgetting or due to the simulation Stressors themselves is not clear. There is some evidence that suggests CPR performance may fall below mastery level in only eight weeks (Friesen & Stotts, 1984).
One of the expectations of nurses in emergency situations is that they be able to recall and respond to verbal instructions given to them. In this study, none of the subjects wrote the medication orders as they were given although a pad and pencil were conveniently available. In the medication memory task, none of the subjects were able to accurately recall all of the drugs and dosages. On the average, subjects were able to identify only two of the medications and one of their dosages correctly. While 100% accuracy of medication order recall is expected of nurses, these same expectations are beyond what would be normally expected of student nurses. In part, the failure of the subjects to perform this task satisfactorily could be due to their inexperience. On the other hand, these students were within weeks of completing their baccalaureate degrees in nursing; these cardiac arrest simulation exercises indicate clearly how stressful situations elicit performance that is well below the standards by which these students would soon be evaluated. This further underscores the importance of using empirically tested stress-producing simulations in the education of nursing students.
PRE AND POST TEST MEANS AND STANDARD DEVIATIONS OF STATE ANXIETY AND PHYSIOLOGICAL MEASURES
Correlations between increases in physiological and affective measures and performance on the CPR and medication memory tasks were low and not significant. This indicates that stress effects are truly multifaceted, affecting behavior, physiological arousal, and emotional state. This lends further evidence for the importance of multivariate measures of stress.
Having demonstrated that the simulations have the capacity to elicit stress, we plan to experimentally evaluate the stress-reducing efficacy of a variety of coping skills in the context of the two simulations. Conducting the outcome studies in this context will provide a standardized set of Stressors that could obviously not be arranged in an actual clinical situation.
In addition to research applications, such a simulation could also serve as an excellent training methodology for the learning of motor, cognitive, and coping skills necessary in demanding stressful situations. The simulations offer many more elements of realism than might normally be found in a non-interactive learning laboratory. In a recent study, Benjamin & co-workers found that students who were exposed to stress producing, distracting conversation at the time of mastery testing performed at a higher level in the clinical setting than did students in a control group (Benjamin, Benson, Grim, Edgar, Kautsch, Langland, Flynt, & Rauthel, 1984). These researchers pointed out the need for nursing students to learn motor skills in the context of a situation that more closely approximates the stressful clinical environment. The cardiac arrest simulation could be available for students who have previously demonstrated mastery in intensive care unit procedures. In this way it could function as a transitional learning step between the controlled and relatively non-stressful learning lab to the more stressful clinical setting.
- Beehr, T.A., & Newman, J.E. (1978). Job stress, employee health, and organizational effectiveness: A facet analysis, model, and literature review. Personnel Psychology, 31, 665-669.
- Benjamin, G., Benson, R., Crim, B., Edgar, C., Kautsch, C., Langland, R.M., Flynt, T.D., & Raithel, J.A. (1984). Use of extraneous conversation in the teaching of nursing skills. Journal of Nursing Education, 23, 109-112.
- Berkun, M., Bialek, H., Kern, R., & Yagi, K. (1962). Experimental studies of psychological stress in man. Psychological Monographs, 76, 1-39.
- Claus, K. & Bailey, J. (1980). Living with stress and promoting well-being. St. Louis, MO: The C.V. Mosby Co.
- Friesen, L. & Stotts, N.A. (1984). Retention of basic cardiac life support content: The effect of two teaching methods. Journal of Nursing Education, 23, 184-191.
- Hebb, D.O. (1955). Drives and the CNS (conceptual nervous system). Psychological Review, 62, 243-254.
- Huckabay, L. & Jagla, B. (1979). Nurses' stress factors in the intensive care unit. Journal of Nursing Administration, 9(2), 21-25.
- Jacobson, S. (1978). Stressful situation for neonatal intensive care unit nurses. American Journal of Maternal-Child Nursing, 3, 144-150.
- Jacobson, S.F. (1983). Nurse stress in intensive and nonintensive care units. In S.F. Jacobson & H.M. McGrath (Eds.), Nurses under stress (pp. 61-83). New York: John Wiley & Sons.
- Jamal, M. (1984). Job stress and job performance controversy: An empirical assessment. Organizational Behavior and Human Performance, 33, 1-21.
- Malmo, R.B. (1958). Measurement of drive: An unsolved problem in psychology. In M.R. Jones (Ed.), Nebraska Symposium on Motivation, Vol. 6 (pp. 229-265). Lincoln, NE: University of Nebraska Press.
- Manuck, S.B., Craft, S., & Gold, K.J. (1978). Coronary prone behavior pattern and cardiovascular response. Psychophysiology, 15, 403-411.
- McGrath, J.E. (1970). Major methodological issues. In J.E. McGrath (Ed.), Social and psychological factors in stress (pp. 41-57). New York: Holt, Rinehart, & Winston, Inc.
- National Institute for Occupational Safety and Health. (1978). Proceedings of the conference on occupational stress. (DHEWNIOSH Publication No. 78-156). Washington, DC: U.S. Government Printing Office.
- Osler, S.F. (1954). Intellectual performance as a function of two types of psychological stress. Journal of Experimental Psychology, 47, 115-121.
- Schlosberg, H. (1954). Three dimensions of emotion. Psychological Review, 61, 81-88.
- Spielberger, C.D., Gorsuch, R.L., & Lushene, R.E. (1970). The State- Thait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press.
PRE AND POST TEST MEANS AND STANDARD DEVIATIONS OF STATE ANXIETY AND PHYSIOLOGICAL MEASURES