Linda Aiken: Improving health care outcomes through research (Part One)
Published 5/10/2014, Vol. 38, No. 1
When Linda H. Aiken, PhD, RN, FAAN, FRCN, speaks, health care listens. Renowned for her leadership in using performance measures to demonstrate relationships between nursing care and patient outcomes and an authority on causes, consequences and solutions for nurse shortages around the world, Aiken is Claire M. Fagin Leadership Professor in Nursing at the University of Pennsylvania (UPenn) School of Nursing, director of the Center for Health Outcomes and Policy Research at the school and also professor of sociology at UPenn. Prior to joining the University of Pennsylvania, she was vice president of the Robert Wood Johnson Foundation. In 2010, she was among the first cohort of inductees into the Sigma Theta Tau International Nurse Researcher Hall of Fame.
Recently, Karen H. Morin, DSN, RN, ANEF, FAAN, past president of the Honor Society of Nursing, Sigma Theta Tau International (STTI), spoke with Aiken about Aiken’s research career—what motivated her to follow the path she took, her most recent findings and their implications for the future of nursing. Here's part one of Morin's interview.
Morin: Dr. Aiken, thanks so much for taking time to discuss the evolution of your research, and to highlight significant implications. I would like to begin by having you set the stage for this fabulous program of research that you’ve undertaken over time. What piqued your curiosity, initially, to go down this road?
Aiken: I graduated from the University of Florida’s baccalaureate nursing program in 1964. That was an important time relative to the American Nurses Association (ANA) debate over entry into practice. As students, we were very much looking at the big debate about whether the baccalaureate should be set as the entry-level qualification for nurses. In 1965, the ANA did establish that as a goal, to be achieved by 1985. All of my career, I’ve been watching us push that goal back further and further.
Also, throughout my career, I’ve been interested in how differently nursing education seems to be thought of, as compared to education in other fields. In other fields, especially in health and medicine, there is much more emphasis on a longer education and more education, and there’s always been emphasis in the research literature about the importance of higher qualifications and better patient outcomes. The best example, with which we are all very familiar, is the very long educational pathway that we’ve adopted for physicians in the United States. It’s the longest in the world, the most intense. As a nation, we obviously place high value on physicians being very well educated. And for decades, there’s been good research that shows that additional qualifications such as board certification of physicians are associated with better patient outcomes.
Coming back to our debate about nursing education and qualifications, it seemed to be less about education, less about qualifications and more about how fast can we educate nurses to meet the day-to-day demands. With doctors and nurses working so closely together in patient care but being so different with regard to qualifications, it seemed to me that nursing needed to update its thinking, in light of the demands of medicine and health care. So, I’ve always been interested in taking a look at whether qualifications and education matter in nursing as much as they have been shown to matter in medicine. And, of course, they do.
Every 10 percent increase in the proportion of nurses with a BSN degree in hospitals is associated with a 5 percent decline in mortality.
Morin: Yes, of course, they do! It’s nice to know that something many of us knew intuitively now has such good solid data to support it, and we have you to thank for that. I chuckled when you said you were right in the throes of the ANA discussions, because I am just a little behind you. My education was in Canada, but I listened to my sister, which I rarely did, who was older and was a nurse educator in Canada, who said to me when I was thinking about being a nurse, “Well, the wave of the future is a baccalaureate education. So, if you are going to be a nurse, you better go to university to get your degree.” So thanks for providing that context, because I don’t know if everybody knows that history.
Since then, you’ve done a lot of work that has been incredible in terms of informing us and providing evidence supporting more education. Would you summarize some of the major research you’ve done, both in the United States and in other countries, bringing us up, perhaps, to a couple of years ago?
Aiken: Ever since spending my first night in charge as a new BSN graduate, my main area of inquiry has been the organizational context in which nursing practice takes place. From that first night on, I was aware that the organizational context for nursing practice was problematic in so many ways. Even though, as a young, well-educated nurse with good beginning expertise, I could see it would take everything I had and more to really provide the care that patients deserved. And that was because the context of care created so many obstacles to professional nursing practice. So I became interested in that, why it was the case and if it could be changed.
So really, my whole career, from that first night as a nurse until now, I’ve been studying and trying to scientifically document in an empirical way the main dimensions of the organizational context in which nurses practice, and asking the question: Does the organizational context of practice have an impact on nurse outcomes? In other words, how does it relate to nurse burnout, overall career satisfaction of nurses, on whether or not they will even stay in clinical care. And, most importantly, how does the organizational context of care affect patient outcomes?
Now, studying the organizational context of care is a very complicated and difficult task, because you can’t go to large data sets and immediately find any measures of this. So I spent a lot of my early career trying to figure out how to measure organizational context or culture, what many people in health care consider to be sort of a soft thing that is probably not that important. When we really figured out how to measure the nurse practice environment, it opened the door for advances in nursing outcomes research. To measure the nurse work environment, we took a page from organizational sociology. I have my PhD in sociology, so I tend to approach things from a sociological perspective, in terms of conceptual frameworks, methods, and so forth.
Organizational sociology proposes that a really good way to understand organizations is not so much to ask the people at the top what goes on—the elites in the organization—but to ask the people who really work in the main part of the organization, the people who come in contact with everyone, and use them as informants about the organizational context. We thought, what better informants are there—particularly in hospitals but also in health care in general—than nurses, because they are in close proximity with patients and families, and with doctors.
I’ve always been interested in taking a look at whether qualifications and education matter in nursing as much as they have been shown to matter in medicine.
Nurses are employed in a line relationship with managers, so they get the full brunt of communication with managers, for better or for worse. And they are responsible for making all the resources come together around patients, so they have contact with materials management and, really, everything in the hospital. Once we decided we could use nurses as informants, we realized we could do large-scale surveys of them and ask them where they work. So this approach has been the basic breakthrough for our scientific method.
We concluded that, if you surveyed 100,000 nurses selected from state registry lists, you would have a whole population of nurses. You could ask them where they worked and, through these surveys, completed at their homes, they would provide empirical measures of what their workplaces were like, wherever they worked. They would also give you the names of their employers, so then we could aggregate nurses together and have, without ever really studying a hospital per se, 300 nurses who worked in a particular hospital really characterize the context of their work environment.
So some of our very first publications related to the question, what is the impact of variation in nurse staffing on patient mortality? And, of course, the answer was, it is quite significant. But once we figured out how to actually survey nurses, then we could ask them about the work environment and the composition of the nurse workforce at the health care delivery level, about which no information was then available. So even though we knew nationally what proportion of nurses have a baccalaureate degree, we never knew what proportion in Hospital X had a baccalaureate degree or whether there was any variation in the percentage of BSN nurses from hospital to hospital. After developing our survey technique, we found that some hospitals had no nurses with baccalaureate degrees, while others had more than 60 percent.
We were fascinated by that because, usually, when there is that large a variation in an input factor in health care, there are consequences for patients, and that’s exactly what we found in our initial study. So we (Aiken, et al.) published a paper in JAMA in 2003, which was really the watershed paper, I think, because, not only did it show that nurse education has an impact on mortality, but it also demonstrated a new perspective on studying the impact of nurse education.
We said, while it’s true that education is a characteristic of an individual, it can also be a characteristic of a hospital. Hospitals can have a higher proportion of nurses with BSNs or a lower proportion. Looked at that way, it becomes a hospital characteristic that can be set by hospital management. Briefly, what we found in that study was, every 10 percent increase in the proportion of nurses with a BSN degree in hospitals is associated with a 5 percent decline in mortality.
Morin: I was in Michigan when you presented that at the Michigan Nurses Association meeting, which was just fabulous. So now, you’ve expanded this. You’ve replicated this work in other institutions and looked at it from an international perspective. If my memory serves me correctly, the same findings are present when we look at it in different health care systems. Am I correct?
Aiken: Yes. One of the things that has motivated our international work is, around the world people look at the United States and they say, “Well, they do a lot of research there, but that’s the United States, and it’s not like any other place.” We found that our big paper in JAMA (Aiken et al., 2003) had limited influence in Europe, for example, where they said, “Oh, well, that doesn’t apply here.” So we decided that we should test out some of these ideas internationally, to figure out if these are generic concepts that you could find in every health care system, no matter the country, what kind of resources they had, or how it was organized and financed.
Both better staffing and better education have a much bigger effect on mortality in hospitals that have a good work environment.
We (Aiken et al., 2001) launched the international hospital outcome study in five countries: the United States, Canada, England, Germany and New Zealand. In Canada, Carole Estabrooks and her colleagues found almost the same finding as in the United States: A 10 percent increase in BSN nurses was associated with about 5 percent decline in mortality (Estabrooks et al., 2005). We have since replicated that finding in Belgium and published it. And now we are doing a very large-scale study of 12 countries in Europe, China and South Africa.
In the meantime, to permit people to move easily between countries in the European Union, the EU established a standard for nursing education in Europe, and they set that standard at the baccalaureate degree. Now, of course, in Europe many of the countries are like the United States, in terms of their resources, but we’ve also looked at this in terms of countries with more transitional economies, and we find the same association. A paper we’re just getting ready to publish shows that hospitals in China with a higher proportion of BSN nurses provide better outcomes for patients, so it seems to be universal.
Morin: This is very exciting because now you are beginning to see the answers, globally, to questions you asked a long time ago. Now, since then, you’ve expanded your research to look at a few things besides education that influence patient outcomes. How did you arrive at those other variables? I know it still has to do with the context of the health care system and the actual institution, but what got you there?
Aiken: All of us who work in health care observe that it’s a complicated phenomenon, and we don’t want to oversimplify it in terms of cause and effect. So we are always looking for evidence of a causal link, a link between better educated nurses and better outcomes, for example. That’s somewhat difficult to determine just from the cross-sectional data we have, so we’ve been on the track to create more of an understanding of a causal link, so that we could say to hospital managers, or even countries, “There is a high probability that if you increase the proportion of nurses at your hospital (or in your country), you would see a related improvement in patient outcomes.”
To do that, we’ve delved in more. One of the first things we did was take the idea of nurse specialty certification, because that relates to an idea I spoke about earlier with regard to physicians, the longstanding research literature which shows that the larger the proportion of board-certified physicians, the better patient outcomes. So we thought, OK. We should look at that, too, because we have an expanding nurse specialty certification program and, actually, many hospitals are trying to encourage all of their nurses to get specialty certification.
We wanted to find, what do you get more value from (or is there an interaction between them)? Specialty certification, which does not necessarily require a specific educational credential, or a nursing education credential? Or is it a mix of both? So we did a big study that looked at the proportion of nurses in a hospital that had specialty certification—we’re talking about national specialty certification from a specialty organization or the ANA—and we found that there is a relationship between hospitals that have more nurses with specialty certifications but only for those hospitals that have a high proportion of BSN nurses. In other words, we found that, as in medicine, specialty certification in nursing is associated with better outcomes but only for nurses who have at least a baccalaureate education. These results were published in Sigma Theta Tau International’s Journal of Nursing Scholarship in early 2011 (Kendall-Gallagher, Aiken, Sloane, & Cimiotti).
Aiken: And, of course, that’s what we have in medicine, as well, because we have these physicians who have many long years of education and then get specialty certification. So, it’s not a surprise that we’re finding this, but it is a bit shocking to some of the hospitals, because we’re able to say to them, in terms of getting value for your investment, yes, support specialty education for your BSN nurses, but for those nurses who don’t yet have a BSN, you would get more, in terms of benefits to your hospital and to patients in your hospital, if you helped those nurses without a BSN get their BSN and wait until later for specialty certification.
Morin: Has that message been well received?
Aiken: It’s a little hard to tell. I think people are very perplexed. One of the things our research does is come up on a problem from a different angle, and it takes time for people to digest the information, because it challenges the conventional way of thinking about things.
We also have another new paper out that takes a similar stance. We go back to the two papers we published in JAMA, where we showed that assigning fewer patients to each nurse is associated with lower mortality (Aiken et al., 2002) and hospitals having more nurses with baccalaureate degrees is associated with lower mortality (Aiken et al., 2003). But now we know how to measure the work environment, so we wanted to know, what does a work environment have to do, if anything, with the impact of better staffing and better education on mortality?
The answer is, and we (Aiken et al.) published this paper in Medical Care in 2011, is that both better staffing and better education have a much bigger effect on mortality in hospitals that have a good work environment. So we have found across-the-board positive impact of nursing education. Nursing education is always associated with better patient outcomes, whether the work environment is good or bad, but it has a bigger effect in hospitals with a good work environment. Likewise, staffing has a very large effect on reducing mortality in hospitals with a good work environment but no effect in hospitals with a poor work environment.
This suggests that, if you have limited resources, you want to decide, what do I do first? Staffing costs a lot of money. If nurses are practicing in a very poor environment, you could throw away a lot of money on improving staffing. So, for those hospitals, the message is, improve your work environment first and then you’ll get more value out of your staffing.
Morin: Very interesting! That helps explicate the interaction between everything, that it’s not a simple A plus B event, nor should it be, in light of the complexity of health care today.
Aiken: Just another note on that. We now have data at two points in time on large panels of hospitals, so, of the hospitals that really improved the proportion of BSNs over time, we can now ask, “Can we point to an improvement in patient outcomes that seems to be associated with that?” And, the answer is, yes, we can. So it’s starting to build this causal link that says to a manager, if I do X, will it yield Y? We are beginning to build a case that says, yes, it will.
Morin: That’s very exciting. Now, let’s put this in context. This has taken place over a period of how many years?
Aiken: It’s really two decades of serious research.
Morin: I think this is important to mention. As I think about young investigators and conversations we have with them about what their contributions are going to be, they need to appreciate that, although we like things to accelerate, it does take time to provide the evidence, the solid evidence, that will then help them move the science forward. I think, sometimes, they think it should be done the next day. RNL
Aiken, L.H., Clarke, S.P., Cheung, R.B., Sloane, D.M., & Silber, J.H. (2003). Educational levels of hospital nurses and surgical patient mortality. JAMA, 290(12), 1617-1623. Retrieved from http://jama.ama-assn.org/content/
Aiken, L.H., Clarke, S.P., Sloane, D.M., Sochalski, J.A., Busse, R., Clarke, H., Giovannetti, P., … Shamian, J. (2001). Nurses’ reports of hospital quality of care and working conditions in five countries. Health Affairs, 20, 43-53. Retrieved from http://content.healthaffairs.org/content/20/3/43.full
Aiken, L.H., Clarke, S.P., Sloane, D.M., Sochalski, J., & Silber, J.H. (2002). Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA, 288(16), 1987-1993. Retrieved from
American Nurses Association. (1965). A position paper. New York: Author.
Estabrooks, C.A., Midodzi, W.K., Cummings, G.G., Ricker, K.L., &
Giovannetti, P. (2005). The impact of hospital nursing characteristics
on 30-day mortality. Nursing Research, 54(2), 74-84.
Kendall-Gallagher, D., Aiken, L. H., Sloane, D. M., & Cimiotti, J. P. (2011). Nurse specialty certification, inpatient mortality, and failure to rescue. Journal of Nursing Scholarship, 43(2), 188-194. doi: 10.1111/j.1547-5069.2011.01391.x
To comment on this article please sign in.