Pandemic Media: Designing Safety

by Nick Travaglini

 

In my presentation, I'll look at how the healthcare system was able to work through this period of acute need. In so doing I will explore the idea of 'designing safety' in 2 senses. The first portion will lay out a theoretical framework based on the concept of 'resilience' as utilized by a discipline called "resilience engineering." This field utilizes decades of research in human factors and ergonomics to study how organizations and systems succeed in spite of pressures and failures, and how those learnings can be applied in a forward-looking manner.

 

I'll look at research outlining a dynamic model of safety developed by Richard Cook and Jens Rasumssen (2005) and tie it together with an examination of the necessity of "inefficiencies'' for humanitarian work by Sidney Dekker and Nalini Suparamaniam (2006). These articles provide a framework for understanding the workings of complex adaptive systems and how one might think with David Woods (2018) about developing "adaptive capacity" within a system.

 

In the aftermath of crises like the PPE shortage, resilience engineering's approach can teach us how a system operates normally, thereby helping us to learn from incidents like the PPE shortage and to prepare for the future. The upshot is that we can use that model of normal work to adjust and build infrastructures that provide the necessary adaptive capacity for systems to handle future incidents.

 

After that, I'll examine 3 cases of people designing PPE for HCWs. These designs were conducted and distributed online, and sometimes directly involved HCWs. I will discuss how these groups of designers self-organized, the designs they actually produced, and how they distributed those designs and got them into the hands of HCWs. I propose that if we bring resilience engineering ideas to bear on these case studies we can glean useful insights and then consider a few proactive approaches to designing safer systems at the level of local hospitals and at even greater scales.

 

To begin, I'll introduce a model of dynamic safety originally developed by Jens Rasmussen and further explored in Cook and Rasumssen's “Going solid”: a model of system dynamics and consequences for patient safety. In this paper, the authors examine a hospital that has seen operational connections within it get tightened over time as the system has striven for various efficiencies; one example is pressure to move patients more quickly from ward to ward in order to increase throughput. They propose a 3-sided model composed of economic, human workload, and acceptable (safety) performance boundaries; a point within this envelope represents the current position of operations. They say that as the gradients of the economic and workload boundaries become steeper (people are asked to do more with less), the operating point is pushed through its safety margins (or buffers) and approaches the marginal safety boundary.

 

They say that as the gradients of the economic and workload boundaries become steeper (people are asked to do more with less), the operating point is pushed through its safety margins (or buffers) and approaches the marginal safety boundary. Over time, people "flirt with the margin" and learn that they can seem to get away with crossing the marginal boundary (and are sometimes rebuffed when small incidents occur). But as they set precedents over time, operations inch closer and closer to the actual performance boundary. The continual resetting of precedents and tightening of operational couplings can make a portion of the system "go solid," which doesn't mean it can't function. Instead, the couplings set up non-linear emergent effects ("accidents") that can more easily push the system past its marginal and actual performance boundaries. By "non-linear," I mean that some small operational change, if tightly coupled to other parts of the system, can massively affect those other parts in ways that can be extremely difficult to predict or for people to intercede to stop the runaway process.

 

With this in mind, we can understand the argument that Dekker and Suparamaniam (2006) make in Of hierarchy and hoarding: How “inefficiencies” actually make disaster relief “work”. They argue that local "inefficiencies" on the part of workers and teams in the field are necessary for successful humanitarian aid efforts. They state: Tight coupling is normally introduced in organizations to gain efficiency, to maximize resource utilization, but it leaves few buffers for absorbing unanticipated surges in workload or resource demand in relief missions. Team leaders actually attempt to reduce coupling, by hoarding a share of resources and hold it back for possible use in more pressurized times. This is consistent with findings from other fields. For example, in healthcare contexts, where nurses hoard “beds” (Cook & Rasmussen, 2003 [sic]), and aircraft maintenance, where local crews hoard tools, hide their real “procedures” of how to get certain jobs done, or overestimate task times to buffer against unanticipated fluctuations in work pace and difficulty (McDonald, Corrigan, & Ward, 2002). Thus what makes a system “inefficient” also makes it work: without hamstering, relief team leaders would be unable to respond to contingencies, and unable to satisfy important organizational goals. 

 

In fact, team leaders maintain that hoarding is what makes them “efficient”:

 

“If we are to be efficient, we have to get as much as possible when we can from the mother organization or any other agency that is ready to provide for us. Just like the refugees or war victims affected in the mess out there, we are also hoarders. We keep, hide. Of course, many times we share what we can, but there is prestige which also means we have to show that we are successful. It means saving up for a rainy day like in the bank”. (Suparamaniam, 2003, p. 124)

 

The point here is that part of the dynamic safety model is people's local efforts to "hoard" "inefficiencies," or to actively undermine attempts to maximize efficiency by the molar organization. And this effort is necessary for the molar organization to function at all.

 

Finally, we can introduce Woods and his conception of resilience as a verb. In this paper, Woods proposes that resilience in complex adaptive systems should be understood as the potential to adapt to handle adverse events. He says: "Resilience concerns the capabilities a system needs to respond to inevitable surprises. Adaptive capacity is the potential for adjusting patterns of activities to handle future changes in the kinds of events, opportunities and disruptions experienced" (1). Those adaptive capacities include the "inefficiencies" previously discussed, which make it possible for people to "fluently" handle "SNAFUs."

 

The concern is that as a system stretches and builds upon its past successes (doing ever more with ever less), it sets itself up for bigger failures by eating into its adaptive capacity. As Woods puts it, "How can organizations flourish despite complexity penalties?" His approach is "to build and sustain the ability to continuously adapt." This shakes out as 4 attributes:

 

1) Encourage initiative by people on the ground within the system (don't overburden them with rules and procedures);

2) Facilitate mutual trust and reciprocity so that people will make short-term sacrifices for the greater good (and see the favor returned later);

3) Embrace surprise and how it challenges dominant understandings; and

4) Proactively learn about how people successfully handled incidents in the past to create a model of how work is normally done and then provide the support they actually need (4).

 

With all this in mind, we can use case studies of how people handled the PPE shortage—of how people took initiative and helped one another (if imperfectly) get through—and then consider a few changes to improve resilience.

 

Nick Travaglini studies resilience in complex systems. He is a student at the New School, in the Liberal Studies department at NSSR.

To view the full presentation please go to: Pandemic Media: Designing Safety

 

 

Response to Presentation

by Professor Rory Solomon

 

Hi, everybody. Thanks for being here. It's a pleasure to be joining you this afternoon. And big thanks to Nick as well as Rachel, Isabel, and Guillermina for sharing all this wonderful work here. And also thanks to my faculty correspondents Josh and Deborah and I echo both of your sentiments about how great it's been to see all this work. And thanks to Sumita, of course, for inviting me to participate. I just want to mention very briefly, I feel compelled to say that I'm actually an alum of the Media Studies program here, and so this event is my first opportunity to present in this capacity as a faculty member. And so it's just really lovely to feel like it's this full circle, and it's a privilege for the consort of participating in stitching together the lineages of this community, which is really nice.

 

So, in this great paper, Nick has given us a survey of several cases when 3D printing practices have intervened in supply-chain shortages of personal protective equipment caused by this pandemic. So, in this response here, I want to suss out what I read as some of the major themes of this research and also offer a few concepts and some other scholars that we might consider as additional ways to consider these themes. I just want to be clear before proceeding that I'm offering these remarks and these concepts, not in any way to present a mission in Nick's paper, but rather, just as opportunities for further discussion, and perhaps some avenues that might possibly work for continuing this line of research.

 

So some of the major themes here that I was drawn to that you were thinking about, like the DIY aspects of this to DIY ethos, and in particular, DIY versus wider, systemic, or structural processes and circulations. DIY as a mode of practice, and what other ideologies, practices and discourses with which it's compatible, and which one other thing does it challenge. And this also makes me think about ad hoc modes or like kind of market oriented modes versus what we might call planning oriented modes, you know, like planned economies and planned modes of provisioning sort of social goods, and of course, resiliency, resiliency and crisis and resiliency and safety, which is the dominant theme here. So, from one vantage point, at least, we can say that this PP crisis was kind of a crisis of market failure, right? Our society is largely organized around this kind of market system for the provision of the manufacturer and delivery of goods. And, in this particular case, the market system had kind of failed on its promise to have this capacity to deliver in this ad hoc fashion, based on supply and demand, there was a speed, the spike in demand and supply was not able to keep up. And so an alternative to society and politics organized around a market economy is often described as a planned economy when we don't allow supply and demand processes to run unregulated. And so I wonder if this is just an open question, if we believe that a planned economy could have possibly functioned better here? And, could this have anticipated the exigencies of this pandemic? And, if so, how would that have worked?

 

Because I think when we talk about resiliency, we're often talking about the existence of surplus, which I think Nick alludes to quite a bit throughout the paper, for take a somewhat different example, in communications networks, when we talk about resiliency, it's often as it relates to redundancy. So, think about this, the internet pioneer Paul Baran, with his mesh network diagram and stuff, and he talks about how a redundancy of routes for information circulations creates a resilient communication network because it means that if some links are destroyed, then the overall system will still function.

 

So that type of resiliency means that in non crisis moments, we have an overabundance or surplus, some fields laying fallow or some manufacturing processes that are not operating at full tilt to their capacity. And that means that some production, you know, some aspects of the system of production is not operating at its limit. And so I think part of the question here might be: is it possible to ensure that we can have this resiliency excess sitting around and not not being fully unified utilized? How might we defend the surplus?