Magnet hospital staff nurses describe clinical autonomy☆☆☆★

Background on autonomy 

In an early definition, autonomy was described as the power to determine what needs to be done in providing patient care, to act on assessments, and to accept accountability for decisions.16 In 1982, directors of nursing defined autonomy as the “freedom to practice,” “independence in nursing and decision making,” and the “ability to self-govern.”17 More recently, autonomy has been variously defined as “control over work,”5, 18 the “amount of work-related independence, initiative, and freedom either permitted or required in daily work activities;”6 “freedom to act on what you know;”15, 19 “freedom from bureaucratic restraints;”7 and building on a body of expert knowledge, reflecting competence, and allowing for accountability and authority in decision making.1

The classic definition of professional autonomy, traditionally considered the ultimate criterion of status as full professionals, is “socially granted and legally defined freedom to make practice decisions without technical evaluation from sources outside the profession.”20 Although Porter-O'Grady8 suggests that no discipline may be autonomous in practice any longer, 2 kinds of autonomy—clinical (ie, rooted in the clinical act) and professional (ie, autonomy focusing on the discipline of nursing)—continue to be noted and desired.15, 21 Beth Israel nurses equate professional and organizational autonomy and define it as the opportunity to work in an environment free from rules and regulations that have little to do with patient care. They define clinical autonomy as the scope of clinical practice for which a nurse is accountable.21

Many investigators continue to study and analyze autonomy as though it were a singular concept and measure multiple concepts with the same tools.3, 6, 22, 23 Kennerly15 suggests that the resulting confusion might explain the unexpected results in her study of the effects of shared governance on autonomy. Analysis of 4 common tools used to measure autonomy led the authors to conclude the following: (a) scale items such as “I rarely ask a patient a personal question,”3 lack face validity; (b) items such as, “On my service, my supervisors make all the decisions; I have little control,”6 do not differentiate between clinical autonomy, professional autonomy, or control over nursing practice23, 24; (c) many scale items such as, “I refused to administer a contraindicated drug,”22 are now standard nursing practice and therefore do not particularly measure autonomy; and (d) quantity of autonomy is measured by the total number of scale items answered affirmatively and assumes that all scale items are of equal weight and importance.3, 6, 23

The study 

The sample consisted of approximately 20 staff nurses from each of 14 magnet hospitals selected from the 3 groups of magnet hospitals known to exist in August 2000: 5 were “original” magnet hospitals so designated by McClure et al25 and studied extensively for 17 years by the authors.26 Two were among 14 hospitals studied by the authors from 1989 to 1993 and judged to have met the same criteria as the original magnet hospitals (ie, high staff nurse job satisfaction, attraction, retention, and productivity).19 These 7 hospitals are referred to as “original—staff nurse” magnet hospitals because the selection criteria focus on staff nurses' attraction, retention, satisfaction, and productivity and look at structural and organizational criteria as they affect nursing practice, not as criteria determining magnetism. The remaining 7 hospitals, called “ANCC-designated” magnet hospitals, focus on the chief nurse executive and structural and organizational criteria. These are hospitals in which the chief nurse executive and the nursing department met the credentialing criteria for nursing service administrators and who had received Magnet Nursing Service Recognitionawards as of August 2000. In each hospital, 20 staff nurses volunteered for the study on the basis of meeting the following 3 criteria: reasonably satisfied with their current job, feel that they can give good nursing care, and were “successful employees” as defined by average or better performance evaluations.

Clinical autonomy ranked-category scale 

Category 1: autonomous patient care action 

This type of autonomy occurred in great frequency, was organizationally sanctioned, and was patient-care focused (ie, at least 1 discipline other than nursing was involved). If action was not successful or appropriate, feedback was constructive, not punitive.

Examples 

1.“We were doing a cath on this diabetic patient who had to have peritoneal dialysis, so I said to the doc, ‘We need to make arrangements for peritoneal dialysis after this procedure.’ The doc says, ‘that should have been taken care of on the east side.’ But I didn't let it go. I said, ‘but they didn't take care of it; we need to do it.' So I did. I talked to the patient and the wife and made all the arrangements. The physician said to me afterwards, ‘I'm really glad you took care of that.' To me, that's autonomy. I see what needs to happen for a patient. I have the knowledge to do it, and I follow through.”

2.“A very elderly patient with advanced Alzheimer's with decreased mental status was ordered to have an MRI to rule out brain metastasis? Stroke? There was no reason for it. It wasn't going to change his treatment one bit. It would have been very hard on the patient because, at best, he would have had to have conscious sedation, at worst, a general anesthetic. I talked to the resident about it yesterday, and he said he wanted it done, wrote the order, and underlined it twice. Then today, he says to me, ‘Make sure that MRI is done.’ I still didn't think it would be in the best interests of the patient, so I took it to the team, and we decided not to do it. I talked to the brother and he said the family wanted it done because family all over the US wanted to know. But, when I explained that it would not make any difference in care, he called them and they understood.”

3.“I had a patient go into respiratory distress. I assessed the patient—rapid respirations, anxiety, color, etc—gave him more oxygen, ordered a chest x-ray because I thought it was fluid overload. Then I called the doctor and gave him a complete report. It was fluid overload, and we got him out of that fast. I felt good because I knew I had made a correct assessment. The other nurses and nurse manager said ‘good call,’ and that made me feel good, too. But if I'd been wrong, there would not have been a big deal made of it. I would have been approached in a professional manner—not someone shaking a finger at me, or scolding, ‘This is how you could have done this better. Or, this is what you missed.”

Category 2: autonomous nursing care action 

This kind of autonomy occurred in great frequency, was organizationally sanctioned, and was nursing-care focused. It provided freedom to act in the best interests of the patient within established nursing parameters (eg, teaching, discharge planning, skin care) and nursing care protocols. There was no evidence of negative organizational sanction, and frequently organized study was involved.

Examples 

1.“From previous experience and knowing this particular patient, I decide whether alternating pain meds is better than giving a single type of pain med all the time deciding when to get patient out of bed—the patient may become tachycardic versus possibly getting a deep vein thrombosis.”

2.“Our CNS [clinical nurse specialist] did research on angio patients. This has led to changes and a wider scope of decision-making regarding these patients. It has broadened the protocols.”

Category 3: limited autonomy 

This type of autonomy occurred in moderate frequency, sanction by some authority was required, and action was patient-focused. The freedom to act was or was believed to be restricted in some way. Frequency was characterized as “to some extent, to a limited extent.” Essence of this classification is that the nurse knew what to do and wanted to do it but did not feel free to act until the sanction of some authority was obtained.

Examples 

1.“Standing orders tell us what we can do and when—I had a patient who really needed a respiratory consult, but he had to wait until I could get a written order first. The patient lost a whole day of treatment within these constraints (doctor's orders, policy and procedures, routines, standardized care plans), I have freedom to do what's best for the patient.”

2.“I'd ask my nurse manager, charge, or the other nurses before acting outside of standing orders. I had an 89-year-old woman with a massive CVA [cerebrovascular accident]. She had previously made herself a DNR [do not resuscitate]. She's nonverbal now she had an N/G [nasogastric ] tube in for feeding and she pulled it out. The intern ordered a new feeding tube to be reinserted. I was very uncomfortable with this so I went to my nurse manager and she worked it out with the intern.”

Category 4: unsanctioned autonomy 

This type of autonomy occurred in low frequency and as taking place “some, sometimes, on rare occasions” and “if I really feel strongly and am willing to take a chance.” Examples were usually patient-care focused, and lack of organizational sanction was either “known” to be or perceived by the nurse to be negative. This lack of support was often identified as a feeling of risk, being “out on a limb,” or taking a chance with the nurse's professional license. The examples in this category were often a series of actions in which the nurse executed an autonomous action and received some kind of negative feedback. Subsequent actions in the same category were then executed covertly with only “the other nurses I work with knowing.” The first 2 examples in this category are grouped because often interviewees indicated that they did not want “others” to know of their covert or clandestine actions on behalf of the patient.

Examples 

1.A number of examples involved bypassing a physician who “would not listen to our observations and assessments or what we had to say about the patient” and either covertly doing what we know “should be done” or going to a different physician whom we knew would listen and give us the order that the patient needed.

2.Another group of examples involved the nurse being “blocked” in functioning in the best interests of the patient because of physician's research or because patient satisfaction superceded the health/medical needs of the patient. The nurse advocated for the patient by first going directly to the physician and then, if blocked, attempted to get organizational sanction for the desired action by going to the nurse manager or supervisor. Oftentimes, the latter was not forthcoming even though the supervisor would acknowledge that the nurse was “right” in what he or she wanted to do. These examples are classified as “unsanctioned” rather than “no autonomy” because “you can function autonomously here, providing you don't interfere with physician's research or with getting good ratings on patient satisfaction scales.”

3.“Nurses should be able to do patient teaching like diabetic teaching we really do know how to do that, you know, what to teach and the best order of things. I've tried, but we can't do that here. You have to call the diabetic educator, and she determines what will be taught and on what day and how it will be taught. I feel so restricted. To give you a concrete example, the best time to teach a patient insulin injection is when it's time for him to have his insulin—but no, we're supposed to wait for the diabetic educator to come. But she's got lots of patients. So we do it anyway. We just go ahead with the teaching and everything else like we used to.”

Category 5: No autonomous practice 

Frequency was characterized by “no, never, not here.” Scope was undeterminable because there was no action. Nonsupport for autonomy was stated or implied. The description often indicated “acceptance” on the part of the nurse.

Examples 

1.“No. We can't do anything here without a written doctor's order. You're asking for trouble if you do don't even think about ‘using your head’ and getting the patient what he needs—even if it's an emergency.”

2.“No, never. If you don't have an order, call the supervisor, and she'll get it, even if it's for simple things like getting a patient up to use the bathroom.”

3.“Too much red tape—policies, procedures, routines; ‘the way we usually do things around here’ gets in the way of the nurse acting for and in the best interests of the patient.”

In a few instance (N = 3), nurses' responses to the autonomy question could not be classified. This usually involved new graduates—a classic response being: “Autonomy!! Surely you jest. No way am I ready for that yet!” Three additional nurses were unable to come up with any description or example. The final number of usable, classifiable responses was 270 out of 279. See Table 1

Table 1. Distribution of the 270 examples by ranked category

	Rank/Category	Number of nurses	Percent of nurses
	1. Autonomous patient care action	117	43.33%
	2. Autonomous nursing care action	47	17.42%
	3. Limited autonomy	35	12.96%
	4. Unsanctioned autonomy	27	10.00%
	5. No autonomous practice	44	16.30%
		270	100.00%

for the distribution of the 270 examples by ranked category.

The distribution of examples into the highest category ranged from a low of 10% in 3 hospitals to 80% or more in 5 hospitals. The distribution of examples into the last 2 categories ranged from 10% or less in 6 hospitals to 55% in 1 hospital and 75% in 1 hospital.The remaining 6 hospitals ranged from 25% to 30% in these categories. These ranges covered both the “original-staff nurse” and the “ANCC-designated” magnet hospitals.

No attempt was made to ascertain whether nurses' perceptions of degree of autonomy or sanction were correct or whether their descriptions of the clinical situation were accurate. Classification into categories was made on the basis of the nurses' perceptions, as illustrated by the examples and descriptions provided.

Relationship between autonomy category rankings and quality of care and job satisfaction rankings 

Staff nurses in this study ranked autonomy third out of 37 factors abstracted from the original magnet hospital study25 as being very important to giving quality care.26 Autonomy also has consistently been associated with high nurse job satisfaction.11 The following 2 questions guided this analysis:

At the beginning of each interview, nurses were asked to rate, on a scale of 1 to 10 (10 being high), the quality of care given on their units and their own job satisfaction. Autonomy rankings and rankings for quality of care and job satisfaction were then compared for each nurse. Results of Spearman Rho correlations for ranked data by hospital, corrected for ties, showed significant positive correlations between autonomy rankings and quality care rankings in half of the hospitals (Table 2).

Table 2. Significance of relationships between ranked autonomy categories and nurse job satisfaction and productivity of quality care: Spearman RHO corrected for tied ranks

	Hospital # satisfaction	Productivity of quality care	Job
	1	.35	.40*
	2	.21	.60†
	3	.22	.52*
	4	.34	.44*
	5	.46*	.47*
	6	.67@008c	.42*
	7	.55*	.57*
	8	.41	.42
	9	.35	.65†
	10	.06	.54*
	11	.51*	.58†
	12	.44*	.54†
	13	.39*	.39*
	14	.86†	.49*

*Significant for sample size at .05 level; †Significant for sample size at .01 level.

Discussion 

Although autonomy has long been considered essential to magnetism, 26% of the staff nurses in this study described examples reflecting unsanctioned or no autonomy. To the extent that the ranked categories define quantifiably different levels of autonomy (and examples were provided for each category so that readers could judge the validity for themselves), this is a very surprising result. Anecdotally, the hospitals in which half of the 20 nurses interviewed described “no autonomy” situations were those hospitals in which several nurses at different levels in the organization voiced concern that “they were not magnet hospitals” or that autonomy was negatively affected by merger or advent of a large cadre of new physicians “who haven't learned how to work with us yet.” These also were the hospitals in which nurses reported poor physician relationships32 and little or no control over practice.24

In 5 of the 14 hospitals, almost all of the nurses interviewed described examples of autonomous patient and/or nursing care action (the few nurses who didn't were usually new graduates or new hires). This was expected and is supported by the magnet hospital literature. The ranked-category scale developed in this study requires more study, development, and establishment of psychometrics. If it proves to be a valid indicator of degree of autonomy, then it might well be used as 1 indicator of magnet hospital status, particularly since it focuses on the staff nurse and hospital as a whole rather than the chief nurse executive, management, and the nursing department.

This research is significant for the following reasons: (1) it elicited from staff nurses examples of what they understood to be autonomy, rather than suggesting examples of autonomous behavior as do most of the usual measures of autonomy; (2) it clearly demonstrated that when staff nurses used the word autonomy, they meant “clinical autonomy” and action beyond the usual standard of practice; and (3) it confirmed the very important relationship between autonomy and nurse job satisfaction.

The results of this research are particularly important and useful to nurse managers. The support and sanction of nurse managers (and through them, of upper administration) are essential to nurses functioning autonomously in 3 ways. Competence is a prerequisite to autonomy. You can't act on what you know, unless you know. Nurse managers are crucial in providing opportunities for nurses to keep up their knowledge base. Another essential component of autonomy is freedom. How does one feel freedom? The best indicator of freedom is trust. “We could not function autonomously without the support of our nurse manager she's super and the reason why we have an ‘untiller’ not an ‘iffer’ culture. Our nurse manager will back you 100% until we give her reason not to trust us far better that than the philosophy that ‘if you do thus and so and show me over and over again that you will do it that way, then I'll trust you.’”

Nurse managers must provide opportunities for nurses to develop and maintain exquisite competence; must communicate trust; and most of all, must empower nurses to function autonomously. To empower means to “share power with,” to enable (ie, motivate through enhancing personal efficacy a “can-do” attitude). “Confidence in the employee” and “fostering autonomy” rate as the most empowering leader behaviors.7 To empower, nurse managers must support the “risk taking” essential to a culture of excellence.33

Lastly, the practice of clinical autonomy must be rewarded. If nurses are to practice autonomously, then behavior reflecting autonomous practice ought to appear on peer review as well as other evaluation tools. It is surprising how little this occurs. For example, a review of published peer review instruments34 disclosed only 1 of 30 items related to clinical autonomy, for example, “Takes initiative to advocate for patient/family.”

The lack of progress toward increasing nurses' perceptions of the opportunity to function autonomously has serious implications for the nursing shortage.2 As a major component of job satisfaction, promotion of autonomy within nursing could do much as recruitment and retention strategies. Suggestions for fostering development of autonomy—decentralization, empowerment, increasing educational level of clinical nurse, role modeling, and role taking—are as appropriate now as they were in 198835 and in 1993.19 Autonomy has also been noted to be 1 of the 2 factors essential to job contentment and is positively related to commitment to the organization and intent to stay.36

The findings of this research have implications for further research. Scales to measure autonomy need to differentiate between clinical autonomy and professional autonomy and “control over practice.”24 As Ballou4 noted, “Construct validity is essential in the development and use of research instruments. Without operational definitions and defining attributes, conceptual measurement is futile.” It would be extremely useful to correlate scores on self-reported clinical autonomy scales with the scale developed here from nurses' descriptions of actual practice. The scale developed in this research also would be useful in measuring the effectiveness of educational efforts to increase clinical autonomy of students and nurses. Additional interviews with nurses functioning at a high level of autonomy would help to clarify exactly what nurse managers need to do to convey organizational sanction and support. If autonomy is a highly desirable behavior related to nurse effectiveness, job satisfaction, and retention—3 of the 4 cornerstones of magnetism—then we need to be clear what we mean by autonomy, develop tools to measure it, and conduct research to identify the factors leading to and supporting it.