Complexity Theory: Developing a Complexity Lens – Part 1.

David Aron, MD, MS  December 28, 2015.

Complexity is popular, made so by novels and films such as Jurassic Park and mass market non-fiction like Chaos: Making a New Science, and Complexity: The Emerging Science at the Edge of Order and Chaos, and many others. The field has been a rich source of metaphors, some more apt than others, and the terminology has been used to put a scientific veneer on evidence light (if not evidence free) opinions and unsophisticated analyses.  Nevertheless, there have been many efforts to advance the field, usually within traditional disciplines, but sometimes in a more transdisciplinary fashion.

The first problem comes with the definition. There are many definitions and descriptions of complex systems.  Melanie Mitchell’s definition has the virtue of brevity and clarity (at least superficially). She defines complex systems as “collections of elements, interacting nonlinearly, which produce emergent behavior.” The only problem with this definition is the use of the words: collections, elements, interacting, nonlinear, produce, and emergent behavior. How these terms are defined and characterized makes a great deal of difference. Nevertheless, although complexity theory/science is still relatively immature, there is still much of value to be learned. It has been argued by David Byrne that the “complexity frame of reference” was a more apt term. This approach was also endorsed by Boulton, Allen, and Bowman. Different scholars and practitioners have parsed complexity in different ways. Several have listed components, while others have divided complexity into broader categories that included several elements. Morin, a philosopher, divided the field into restricted complexity (a scientific and methodological approach) and generalized complexity (a philosophical and epistemological approach).  Manson, a geographer, divided the field into algorithmic, deterministic, and aggregate complexity. Geyer and Rahani, policy scientists divided complex systems into physical, biotic, and conscious.  Another approach would be to divide the field into academic versus experienced complexity (after Beautement and Broenner).

One of the issues is the separation between what might be termed academic complexity and what we experience as complexity.   In much of the academic literature complexity is approached primarily with mathematical models. This is a consequence of the hegemony of “science” and the particular view that rules in the groves of Academe.  These mathematical models are both interesting and informative. For example, the phenomenon of birds flocking or fish schooling can be described in three relatively simple mathematical equations. (see http://www.red3d.com/cwr/boids/ ), demonstrating that simple rules can result in complex behavior. (We err when we go on to conclude that complex behavior always results from simple rules or that we can impose simple rules on a complex system and have completely predictable results.) More recently, qualitative approached have been used, particularly in the context of the increasing popularity of “mixed methods.” A part of the stimulus to the use of qualitative methods is the desire to gain a richer picture of the experience of complexity.  We experience complexity in the world and often depend on intuition and sensemaking. For my course in the Doctorate in Management Program at the Weatherhead School of Management, on managing in complex systems I have felt the need to bridge academic and experiential complexity. Scholarly complexity may constitute that bridge.  I defined scholarly complexity as the conscious and systematic application of principles (regardless of their origin) that underlie complex systems to problems of practice in the real world – praxis. This frame of reference is a different way of looking at things and thus it involves a “complexity lens.”  In Part 2, I will describe the development of a complexity lens.

References:

  • Mitchell M (2009) Complexity: A Guided Tour. Oxford University Press, Oxford UK.
  • Byrne D, Callaghan G. (2014) Complexity Theory and the Social Sciences: The state of the art. Routledge, New York.
  • Boulton JG, Allen PM, Bowman (2015) Embracing Complexity: Strategic Perspectives for an Age of Turbulence. Oxford University Press, Oxford UK.
  • Beautement P. and Broenner C. (2011) Complexity Demystified, Triarchy Press. Axminster, UK.
  • Boisot, M. H., & McKelvey, B. (2010). Integrating modernist and postmodernist perspectives on organizations: A complexity science bridge. Academy of Management Review, 35(3), 415-433.
  • Geyer R., Cairney P. (20016) Handbook on Complexity and Public Policy. Edward Elgar Publishing, Cheltenham, UK.
  • Manson S. (2001) Simplifying complexity: a review of complexity theory. Geoforum 32:405-414.
  • Morin, E. (2007) Restricted Complexity, General Complexity, in: C. Gershenson, D. Aerts & B. Edmonds (eds) Worldviews, Science and Us, Philosophy and Complexity (London, World Scientific) pp. 5–29.
  • Richardson KA. (2008) Managing complex organizations: complexity thinking and the science and art of management. E:CO 10:13-26.

Nothing new under the sun

David C. Aron, MD, MS, OC (optimistic cynic)

Nothing new under the sun.

What has been is what will be,
and what has been done is what will be done,
and there is nothing new under the sun.

Is there a thing of which it is said,
“See, this is new”?
It has been already
in the ages before us.

 

There is no remembrance of former things,
nor will there be any remembrance
of later things yet to be among those who come after.

Ecclesiastes  1:9-11 (Jewish Publication Society 1917)

The April 2010 issue of AHRQ Research Activities Newsletter led with the astonishing headline: “Patients admitted to the hospital on weekends wait for major procedures.” (http://archive.ahrq.gov/news/newsletters/research-activities/apr10/0410RA.pdf )  This finding was based on a report Characteristics of Weekday and Weekend Hospital Admissions that was based on data from 2007. (http://www.ncbi.nlm.nih.gov/books/NBK53602/pdf/Bookshelf_NBK53602.pdf)

 

There have been many critiques of health services research. Sometimes it has been called ‘academic,’ meaning not of practical relevance or of only theoretical interest. This particular article brings up another criticism that I have heard (and made) that health services researchers prove the obvious.  The authors of the report have used rigorous methods to support their conclusion and I suppose that should make us feel more comfortable in making the statement with scientific certainty.  However, one has the sense that this is something that could have been predicted more than 2000 years ago.

Genesis 2.1-3.  And the heaven and the earth were finished, and all the host of them. And on the seventh day God finished His work which He had made; and He rested on the seventh day from all His work which He had made. And God blessed the seventh day, and hallowed it; because that in it He rested from all His work which God in creating had made.

Exodus 20:7-10. Remember the sabbath day, to keep it holy. Six days shalt thou labour, and do all thy work; but the seventh day is a sabbath unto the LORD thy God, in it thou shalt not do any manner of work, thou, nor thy son, nor thy daughter, nor thy man-servant, nor thy maid-servant, nor thy cattle, nor thy stranger that is within thy gates; for in six days the LORD made heaven and earth, the sea, and all that in them is, and rested on the seventh day; wherefore the LORD blessed the sabbath day, and hallowed it.

Deuteronomy 5:11-14. Observe the sabbath day, to keep it holy, as the LORD thy God commanded thee. Six days shalt thou labour, and do all thy work; but the seventh day is a sabbath unto the LORD thy God, in it thou shalt not do any manner of work, thou, nor thy son, nor thy daughter, nor thy man-servant, nor thy maid-servant, nor thine ox, nor thine ass, nor any of thy cattle, nor thy stranger that is within thy gates; that thy man-servant and thy maid-servant may rest as well as thou. And thou shalt remember that thou wast a servant in the land of Egypt, and the LORD thy God brought thee out thence by a mighty hand and by an outstretched arm; therefore the LORD thy God commanded thee to keep the sabbath day.

The idea of a day of rest for the working class was an innovation and by no means the usual practice. The Jewish Sabbath was criticized by 1st Century Roman Seneca the Younger who was quoted by Augustine in City of God:  ‘Seneca, among the other superstitions of civil theology, also found fault with the sacred things of the Jews, and especially the Sabbaths, affirming that they act uselessly in keeping those seventh days, whereby they lose through idleness about the seventh part of their life, and also many things which demand immediate attention are damaged.( De Civitate Dei. 6:11) Nevertheless the concept caught on.

Since on the seventh day even God rested and no procedures were performed, is it surprising that delays have been introduced in the system?  I am not suggesting that the fact that the Sabbath appears in the Bible is scientific proof that this is the cause of waiting for procedures.  If anything, it is only correlation rather than proof of causality. Rather, it raises the issue of how much rigorous research is required.  Should we do an RCT to be sure of the causal relationship? Further, if waiting for procedures for patients admitted on weekends was not demonstrated, would you believe it? Is there a plausible counterfactual that would explain those results?  That is also not to say that there is nothing important to be learned here. Which procedures are delayed and whether that makes a difference in outcomes are legitimate research questions that have implications for hospital staffing.  Let us just not let ourselves be carried away with our brilliance. (For the record I am not aiming this at the authors of the study which included many more findings of considerable interest; rather, the idea that this was a headline struck me as funny.) This led me to ask the question, how much of the findings of health services and related management research, particularly in my area of implementation of evidence into practice, were at least foreshadowed in the Hebrew Scriptures.

There are several articles that provide so called lessons learned or prove the importance of one factor or another and these have been published in “high impact journals” (which means that other health services researchers cite them and not that they have high impact on actual practice – I am as guilty as the next in trying to publish in such journals). For example, some of the factors associated with implementation success include:

Leadership –  One need only read the Book of Exodus; and the ever popular importance placed on the leader’s vision.  See Proverbs 29:18

Spatial sufficiency (which probably means enough space)– which facilitates co-location of primary care and mental health clinics  Putting aside the laws of physics which don’t allow two things to occupy the same space simultaneously, one might recall that the Israelites are reminded that God led them out of Egypt. The Hebrew word for Egypt (מִצְרַיִם ) literally means the narrow places. Even then, spatial sufficiency was important.

I have no doubt that there are other examples. Health services researchers would be more useful if they looked at how to implement something in the face of poor leadership or how to accomplish things when space is a constraint. It is time for a change.

Bible translations and numbering from the JPS 1917 Edition

 

http://www.mechon-mamre.org/p/pt/pt0.htm

11/30/15

The Veterans Health Administration as a Wicked Problem

David C. Aron, MD, MS

The Veterans Health Administration as a Wicked Problem — A commentary on Reforming the Veterans Health Administration — Beyond Palliation of Symptoms by Brett P. Giroir, M.D., and Gail R. Wilensky, Ph.D., N Engl J Med 2015; 373:1693-1695.

A recent article in the New England Journal of Medicine by the co-chairs of a blue ribbon panel summarizes the panel’s assessment of the Veterans Health Care System in light of recent scandals related to access to care, quality of care, and falsification of data. (http://www.nejm.org/doi/full/10.1056/NEJMp1511438) The title alone: “Reforming the Veterans Health Administration — Beyond Palliation of Symptoms” promises that recommendations will get to the “root causes” of VHA’s problems and identify real solutions. Unfortunately, the identification of problems is too shallow and doesn’t really get at their complexities and interdependencies. The report doesn’t adequately address the conflicting interests and political power of stakeholders nor does it deal with the issues of how and why the VA got to where it is now.  Thus, it is not at all surprising that the report notes that it includes numerous operational recommendations for the near term, few of which are unexpected. They are not only expected, they have been known for years. For example, the need for more exam rooms and increased staff-to-patient ratios to enhance physician productivity was noted in the Advanced Clinic Access initiative, a prior national VA effort to improve access beginning more than 15 years ago. (http://www.researchgate.net/profile/Mark_Meterko/publication/228478818_The_implementation_and_effectiveness_of_advanced_clinic_access/links/00b7d526555e853f00000000.pdf)

The key question is: if the problems were known, why have they not been addressed successfully?  First, as noted in the report, VHA is one of the largest health care systems in the US and size alone (9.1 million enrollees, 20,000 physicians, 1600 facilities, 288,000 employees) brings with it the challenges of minimizing undesired heterogeneity. In addition the various conflicting interests play out in a dynamic environment where how you got to where you are matters. This is path-dependence writ large.  All of these and other factors contribute to making the situation intractable and resistant to solution. In fact, the situation meets the criteria for constituting a “wicked problem.”

Rittell and Weber, two urban planners, described a class of problems that they called wicked. These wicked problems had a number of characteristics:

  1. There is no definitive formulation of a wicked problem
  2. Wicked problems have no stopping rule
  3. Solutions to wicked problems are not true-or-false, but good-or-bad
  4. There is no immediate and no ultimate test of a solution to a wicked problem
  5. Every solution to a wicked problem is a “one-shot operation”; because there is no opportunity to learn by trial-and-error, every attempt counts significantly
  6. Wicked problems do not have an enumerable (or an exhaustively describable) set of potential solutions, nor is there a well-described set of permissible operations that may be incorporated into the plan
  7. Every wicked problem is essentially unique
  8. Every wicked problem can be considered to be a symptom of another problem
  9. The existence of a discrepancy representing a wicked problem can be explained in numerous ways. The choice of explanation determines the nature of the problem’s resolution
  10. The planner has no right to be wrong

http://www.ask-force.org/web/Discourse/Rittel-Dilemmas-General-Theory-Planning-1973.pdf

These characteristics describe VHA in a nutshell and make clear why the panel’s recommendation: “First, the urgent need for strategic vision and dynamic decision making argues for a new VHA governance board that is representative, expert, empowered, and relatively insulated from direct political interactions” is not only laughable in the current hyperpolarized political environment, but likely to be nearly if not totally impossible.

How is the VA a wicked problem? For brevity’s sake, I will touch on only a few of the characteristics. First, VA has numerous stakeholders who perspectives differ and objectives conflict.  These include not only veterans and their families, but also their representatives in veteran service organizations, members of Congress, employees both unionized and non-unionized, professional and non-professional, the media, auditors, medical schools and other professional schools, and taxpayers, just to name a few.  For some members of Congress, VHA is a source of jobs for constituents; any attempt to realign facilities with need is resisted.  Similarly, unions are resistant to efforts that might result in staffing reductions. Physicians are recruited to VHA often come because they want to do research and teaching as part of their job. The current focus of attention leading to the blue ribbon panel is related to access to care.  However, although care for veterans is the primary mission, it is not the only one (the others being research, education, and backup for the Dept. of Defense). Time spent by physicians doing research (a major recruitment incentive) is time not seeing patients. Similarly, time spent lecturing in medical school or teaching in other venues (another recruitment incentive) is time not seeing patients. Moreover, physician trainees are less efficient than staff physicians, especially in ambulatory care, the focus of access. Physician trainees also have schedules that conflict with continuity of care. Meanwhile, medical schools want to preserve the inpatient setting for training students and residents while preserving high tech care, supporting research and teaching, and maintaining business relationships where VA facilities send and pay for patients to be seen in the affiliated hospitals. Auditors such as the Inspector General and GAO have an important role, but they also have to justify their role by finding problems. (See Michael Power. The Audit Society, 1997 and also Organized Uncertainty, 2007)  Media thrive on scandal and are even more dependent on it in the 24-hr news cycle.  The multiple stakeholders have the power (to varying degrees) to assert their values and interests. When these conflict, change is difficult. The problems are dynamic. Today it is access; tomorrow it may be something else and today’s solution may be tomorrow’s problem.  For example, as primary care in VHA expanded, the specialties required to support primary care were to a large extent ignored, perhaps because of the costs involved. Decreased access to specialists was predictable. This is not to say that there is no access problem. Rather, it means different things to different people. The patient’s personal physician? Any physician? A nurse? It remained rather murky.  Because success is generally determined in terms of objectives, when there is ambiguity in purpose, there will be a lack of clarity about criteria for success.Although various standards have been set, there is little if any evidence on which those standards are based. When the answer to question such as how quickly does someone need to be seen is “it depends,” then characterizing what the problem is can be difficult.  Moreover, “common sense” solutions may not be so simple in practice. Improving access such as after hours clinics requires changes in staffing way beyond the clinic. In addition to physicians, nurses, and clerks, there has to be additional capacity in supporting services such as radiology, laboratory, and pharmacy.  At the Phoenix VA Medical Center, ground zero for the access scandal, there was a shortage of primary care physicians and an inability to recruit them, a common problem nationwide. Common sense solutions may have unintended consequences.  The use of clinical reminders in the electronic health record to improve quality has not only led to check-box medicine, but has also been described as “death by a thousand clicks.”  This analysis just touches the surface, but the work necessary to have a better grasp of the issues should have been conducted in the Independent Review, but alas was not. In addition, some of the biases of the panel clearly show through. The consideration of the electronic health record is a case in point.

The panel made one statement: “electronic health record (EHR) has been stagnant for a decade, and clinicians are frustrated with the lack of integration and mobility and the feature deficits as compared with commercial systems.”  This is simply untrue. For example, in the  Medscape 2014 satisfaction her satisfaction survey (http://www.medscape.com/features/slideshow/public/ehr2014#7), the VA’s EHR called CPRS was the highest rated. The EHR with the highest market penetration is in the middle of the pack. That has been my experience talking to colleagues. Anyone who has had experience with two different systems prefers CPRS.  This is not to say that CPRS can’t be improved. It can and it should. Rather, the idea that the private sector must be better does not conform to the evidence. The panel also came down hard on VA because of the lack of interoperability of CPRS with the EHR used in the Dept. of Defense. This is fair comment. However, it is important to take note of a few things.  First, VA and DoD spent around $1 billion unsuccessfully trying to develop a joint EHR. DoD then chose a commercial package that promises interoperability – we shall see. ( Second, where is interoperability in the private sector. At an affiliated institution, the different modules in its EHR do not communicate with each other, much less any other institution’s system.   Parenthetically, in my own VA facility, we have access to the EHR in two non-VA health care facilities, albeit not our primary affiliate.  The problems with EHRs in the private sector have been well documented. http://hcrenewal.blogspot.com/

There seems to be an assumption that the private sector does things better. There are two health care systems in the US that are of comparable size – Community Health Systems (Brentwood, Tenn.) and HCA (Nashville, Tenn.).  (http://www.beckershospitalreview.com/lists/15-largest-for-profit-hospital-operators-2014.html) Do they have the degree of uniformity of quality demanded of the VA? Do they routinely share “best practices” and have similar operations?  Maybe they do, but we don’t know.  What we do know is that each of these systems has paid out millions of dollars in settlements over issues such as Medicare fraud, Medicaid fraud, Ethics in Patient Referrals Act (also known as the Stark law), the False Claims Act, and other federal and state laws and regulations. One could go on and on. It is not that I think that many of the recommendations are wrong. Quite the contrary, although because the problems are so intertwined, unintended and unanticipated consequences are likely, but though they call for systems thinking, a critically important issue for such a situation, the panel did not evidence much of it.

Additional references.

  1. Independent Assessment of the Health Care Delivery Systems and Management Processes of the Department of Veterans Affairs. http://www.va.gov/opa/choiceact/documents/assessments/Integrated_Report.pdf
  2. Rittel, W. J. and Webber, M.M. 1973. Dilemmas in a General Theory of Planning, Policy Sciences. 4 (2): 155-169.
  3. Australian Public Service 2007. Tackling Wicked Problems: A Public Policy Perspective. Canberra: APSC.
  4. Camillus, C. 2008. Strategy as a Wicked Problem. Harvard Business Review 86 (5): 99- 106.
  5. Conklin, 2006. Wicked Problems and Social Complexity, in Dialogue Mapping: Building Understanding of Wicked Problems. Chichester: John Wiley.
  6. Head, W. 2008. Wicked Problems in Public Policy. Public Policy, 3 (2): 101-118.