In spite of almost Herculean efforts to improve hospital safety,
results seem at best to be underwhelming.1-4 Why should this be so? In
2001 Plsek and Greenhalgh pointed out that we deal with complex systems.5
It therefore seems reasonable to wonder whether a failure to employ the
Sciences of Complex Systems and Networks6-8 may have something to do with
this. Perhaps as we improve what we do it is matched by complexity and the
outcome is less than fully rewarding.
Transmission of multiple antibiotic resistant organisms (MROs) in
hospitals continues to be a major problem and Vancomycin-resistant
Entercoccus is in many places the current number one villain. It is
reasonable to wonder whether, until we understand more fully the
underlying network structure of transmission,6 we will really be able to
make progressive inroads into MROs.
It has been shown that waiting lists9 have a power law
distribution.10 It seems not unlikely that other events in hospitals such
as length of stay and medication errors may have power law distributions.
This may be important for analyzing and understanding these data. Uncommon
low probability high impact events are likely to occur more commonly in
our hospitals than conventional statistical approaches based on Gaussian
methods suggest.11
Although investigation of major adverse events often reveals a trail
of pre-existing malfunction, it is a characteristic of complex systems
that low probability high impact events may occur seemingly without prior
cause and therefore to be unpredictable. If an adverse event is
unpredictable, the only defense must be generic. We must make our systems
more sustainable and resilient.12 Resilience involves such things as
modularity, diversity and, above all, some redundancy. Hospitals have
often been reformed to make them more "efficient" and "productive". Highly
efficient complex systems are known to be vulnerable to failure. Nature
makes its complex systems more resilient by building in some redundancy.11
We desperately need to know the correct balance of efficiency and
redundancy. For example, if we run hospitals at near 100% capacity they
are probably going to spend a not inconsiderable part of their time fixing
problems that are related to their efficiency. There appears to be good
evidence that bed occupancy rates between 80% and 85% work better.
Pronovost and his colleagues1 have noted the importance of "bottom-
up" approaches. This is readily understandable in the work of Ostrom13 on
cooperative behaviour, and in complex system emergence and self-
organisation.14
References.
1. Pronovost P, Berenholtz S and Morlock L "Is quality of care
improving in the UK?" BMJ 2011;342:c6646.
2. Benning A, Dixon-Woods M, Nwulu U, Ghaleb M, et al "Multiple
component patient safety intervention in English hospitals: controlled
evaluation of second phase" BMJ 2011;342:d199.
3. Benning A, Ghaleb M, Suokas A, Dixon-Woods M, et al "Large scale
organizational intervention to improve patient safety in four UK
hospitals: mixed method evaluation" BMJ 2011;342:d195.
4. Landrigan C, Parry G, Bones C, Hackbarth A, et al "Temporal trends
in rates of patient harm resulting from medical care" NEJM 2010;363:2124-
34.
5. Plsek P and Greenhalgh T "The challenge of complexity in health
care" British Medical Journal 2001;323:625-8.
6. Watts D "Six Degrees" London, Vintage Books 2004, especially
chapter 6.
7. Mitchell M "Complexity" New York, Oxford University Press 2009.
8. Cioffi-Revilla C "Computational social science" WIREs
Computational Statistics 2010;2:259-71.
9. Papadopoulos M, Hadjitheodossiou M, Chrysostomou C Hardwidge C and
Bell A "Is the national health service at the edge of chaos?" Journal of
the Royal Society of Medicine 2001;94:613-6.
10. Brown C and Liebovitch L "Fractal Analysis" Thousand Oaks
California, Sage Publications, 2010.
11. Taleb N "The Black Swan" 2nd ed. Camberwell, Penguin Books 2010.
12. Orrell D "Economyths" London, Icon Books 2010, especially chapter
2.
13. Ostrom E "A general framework for analyzing sustainability of
socio-economical systems" SCIENCE 2009;325:419-422.
Rapid Response:
Hospital Safety and Complexity
Hospital Safety and Complexity
In spite of almost Herculean efforts to improve hospital safety,
results seem at best to be underwhelming.1-4 Why should this be so? In
2001 Plsek and Greenhalgh pointed out that we deal with complex systems.5
It therefore seems reasonable to wonder whether a failure to employ the
Sciences of Complex Systems and Networks6-8 may have something to do with
this. Perhaps as we improve what we do it is matched by complexity and the
outcome is less than fully rewarding.
Transmission of multiple antibiotic resistant organisms (MROs) in
hospitals continues to be a major problem and Vancomycin-resistant
Entercoccus is in many places the current number one villain. It is
reasonable to wonder whether, until we understand more fully the
underlying network structure of transmission,6 we will really be able to
make progressive inroads into MROs.
It has been shown that waiting lists9 have a power law
distribution.10 It seems not unlikely that other events in hospitals such
as length of stay and medication errors may have power law distributions.
This may be important for analyzing and understanding these data. Uncommon
low probability high impact events are likely to occur more commonly in
our hospitals than conventional statistical approaches based on Gaussian
methods suggest.11
Although investigation of major adverse events often reveals a trail
of pre-existing malfunction, it is a characteristic of complex systems
that low probability high impact events may occur seemingly without prior
cause and therefore to be unpredictable. If an adverse event is
unpredictable, the only defense must be generic. We must make our systems
more sustainable and resilient.12 Resilience involves such things as
modularity, diversity and, above all, some redundancy. Hospitals have
often been reformed to make them more "efficient" and "productive". Highly
efficient complex systems are known to be vulnerable to failure. Nature
makes its complex systems more resilient by building in some redundancy.11
We desperately need to know the correct balance of efficiency and
redundancy. For example, if we run hospitals at near 100% capacity they
are probably going to spend a not inconsiderable part of their time fixing
problems that are related to their efficiency. There appears to be good
evidence that bed occupancy rates between 80% and 85% work better.
Pronovost and his colleagues1 have noted the importance of "bottom-
up" approaches. This is readily understandable in the work of Ostrom13 on
cooperative behaviour, and in complex system emergence and self-
organisation.14
References.
1. Pronovost P, Berenholtz S and Morlock L "Is quality of care
improving in the UK?" BMJ 2011;342:c6646.
2. Benning A, Dixon-Woods M, Nwulu U, Ghaleb M, et al "Multiple
component patient safety intervention in English hospitals: controlled
evaluation of second phase" BMJ 2011;342:d199.
3. Benning A, Ghaleb M, Suokas A, Dixon-Woods M, et al "Large scale
organizational intervention to improve patient safety in four UK
hospitals: mixed method evaluation" BMJ 2011;342:d195.
4. Landrigan C, Parry G, Bones C, Hackbarth A, et al "Temporal trends
in rates of patient harm resulting from medical care" NEJM 2010;363:2124-
34.
5. Plsek P and Greenhalgh T "The challenge of complexity in health
care" British Medical Journal 2001;323:625-8.
6. Watts D "Six Degrees" London, Vintage Books 2004, especially
chapter 6.
7. Mitchell M "Complexity" New York, Oxford University Press 2009.
8. Cioffi-Revilla C "Computational social science" WIREs
Computational Statistics 2010;2:259-71.
9. Papadopoulos M, Hadjitheodossiou M, Chrysostomou C Hardwidge C and
Bell A "Is the national health service at the edge of chaos?" Journal of
the Royal Society of Medicine 2001;94:613-6.
10. Brown C and Liebovitch L "Fractal Analysis" Thousand Oaks
California, Sage Publications, 2010.
11. Taleb N "The Black Swan" 2nd ed. Camberwell, Penguin Books 2010.
12. Orrell D "Economyths" London, Icon Books 2010, especially chapter
2.
13. Ostrom E "A general framework for analyzing sustainability of
socio-economical systems" SCIENCE 2009;325:419-422.
14. www.science.org.au/nova/094/094key.htm
Competing interests: No competing interests