All the work of the hand is rooted in thinking. -David Sudnow
Work presented at the Making Healthcare Safer Conference, St. Andrew’s (UK)
Safety in design of input technologies of modern surgery
Medical workplace studies reveal significant difficulties are encountered when teams use computers in sterile environments. These studies highlight the safety concerns that arise out of socio-technical practices to maintain aseptic barriers and the alteration of such practices to accomplish the clinical goals of the procedure. Responding to these concerns human-computer interaction (HCI) researchers have prototyped a variety of input technologies for surgical environments, including foot pedals, speech recognition systems for voice control, and gesture recognition systems for body-based control.
Although the design of these input technologies are motivated by a concern for making modern surgery safer for patients and medical staff, these technical interventions can have their own unintended consequences. This raises the issue of whether design processes for developing new input technologies, based in the culture of HCI research, are appropriate when HCI researchers develop technologies in a medical context. This poster is meant to generate debate and discussion of design processes for the development of input technologies to address the needs for safe, subtle and efficient control of computers in modern surgery.
The goals of this document are to: 1) raise awareness and sensitivity to what Dourish calls the paradox of technical intervention and 2) to illicit feedback from conference attendees. Specifically to discuss what are safe design practices to explore the potential of new input technologies in medical environments.
To ground the discussion the author will present her own on-going study of gesture and body movements of catheter work of cardiac electrophysiology. The study is in response to recent prototype of touchless interaction systems for image navigation and manipulation images in neurosurgery and interventional radiology. This fieldwork will lead to a design process of synthetic gesture vocabularies development. The intention of the gesture vocabulary is to compliment but not override the meanings of naturally occurring body movements of the medical workplace and to do so safely.
There are a number of material arrangements of equipment. The arrangements are determined in a large part by what equipment the lab holds. For instance a stereotaxis machine is roughly the size of a compact car. The types of equipment present in a lab are related to what types of procedures the lab preforms or has performed in the past. This makes analyzing the layout of an EP lab a rather idiosyncratic task. The description will depend to a large degree on what sort of lab one has access to.
However, I’ve learned through my informal engagements with EP lab workers that these materials arrangements actually fall into two categories related to two ways EP medical teams are formed and how and which team members fill the role of performing catheter work.
Several earlier studies of human computer interaction in sterile environments have noted significant challenges in the proxied control of imaging technologies. Proxied interaction with imaging technologies is one of the core components of catheter work in image guided medical procedures. This study extends that work narrowly, by documenting and analyzing the hands and body movements comprising catheter work to answer two questions. One, what is the hand and body movements involved in catheter work? Two, what are the circumstances of interruptions in catheter work emerging out of technical breakdowns?
To address these questions the following study is proposed. The researcher would make an ethnomethodoloical observation and documentation of catheter work in the cardiac electrophysiology procedures conducted at a out-of-state hospital. The researcher would place two video cameras in one of the procedure rooms. She would be positioned in the corresponding control room (see figure 1).
Figure 1 caption: The EP lab has two facilities. Each facility is comprised of a control and procedure room. The researcher would be positioned in the control room making observations of the flow of directives of passed through the intercom from individuals in the control room to the person doing catheter work in the procedure room.
One camera will record standard digital video. The other will record motion in 3-D. The cameras will be placed underneath a boom of monitors (see figure 2).
The cameras will be directed at the hands and upper body of the nurse or cardiologist performing catheter work. It is important to note that cameras will be stationary and the frame in view will not change. Thus it is expected that a person could and would walk out of the camera’s view. By directing the cameras on the hands and upper torso of the person doing catheter work, several elements of the environment will not be in view, such as:
o the contents of the monitors displaying patient information
o the patient’s body and face
o the face and bodies of other persons in the procedure room
o the people and information displays in the control room
The person doing catheter work will be partially in view. The focus on the camera is directed at hands and upper body, but partial view of the person’s masked face be recorded at times. For instance, if a person needed to bend into the fixed shot, the camera would record that person’s head. Several elements of surgical garb worn when performing catheter work will further ensure their anonymity (see figure 3).
Figure 3 caption: Surgical garb is a required part of catheter work in the sterile environment of the procedure room. The person recorded will have safety glasses, a sterile cap, a face mask and a neck covering on their head. Though not pictured here the person doing catheter work will wear a surgical gown, scrubs, and a lead apron. Together these elements help further preserve the anonymity the person being filmed.
An audio recorder will also be present in the procedure room. The audio recorder will capture sound coming from the intercom system. The intercom system mediates the dialog between the control and procedure rooms. Members of the medical team, whose work is distributed across the two rooms, use the intercom to coordinate their work. Several studies document the nature of such communication. The dialog is primarily composed of directives given by a doctor in the control room to the person doing catheter work in the procedure room. Much of the dialog is a clipped exchange of questions, directions and confirmations about the placement and manipulation of catheters. Here is an typifying example. Doc 1, from the control room: “I want you to draw a line from there (pointing with the mouse cursor) to there (drawing a path with the mouse cursor). Doc 2, from the procedure room, manipulates the catheter in the way described, then asks, “like this?” Doc1 replies, “yes, good.”
The audio recording is intended to contextualize the researcher’s video analysis of the hand and body movements. It’s primary utility is to help answer the second research question. As such, the audio analysis will be focused on identifying the circumstances in technical breakdowns and understand the flow of directives that lead up to and follow interruptions in catheter work.
Two final and supplementary forms of data will be collected. To understand the context of the dialog directing the progress of the procedure underway, field notes will be made by the researcher. The researcher will be stationed in the control room for these observations. She will note the processual character of catheter work as directed from the control room, answering questions like the following:
- What happens first?
- What happens next?
- What directives come in what order?
- Is there a pattern in what the person doing catheter work has been instructed to do?
The last form of data collected is interview data. Since the goal is to understand the hand and body movements involved in catheter work, on occasion the researcher may review a piece of video with the person recorded in order to understand what those movements represent or how the movement functions in their work. The interview might also play back portions of audio to order orient the interviewee to the particular section of the procedure seeking clarification about his or her catheter work. A limited number of other circumstances might occasion an interview. For instance, when analysis of video data is inconclusive the researcher, may interview the person who was performing catheter work to explain interruptions in that work. The goal here will be to determine whether a technical breakdown occurred and what that its circumstances were. An example of the interview questions follows.
- When you left the frame here (playing back video and audio) what was going on?
- When you got this directive (play back audio) how did you interpret the message?
- When you went into the control room were you seeking clarification about information displayed in the monitors? Was this related to some instructions you received from the control room?
Instead of audio recording interviews the interviewer will make notes.
In all forms of data will be anonymous from the point of recording. Thus the data will not be de-identified because no identifying information will be recorded. To accomplish this generic pseudonyms will be used. The recordings of video and audio will be used for analysis only and won’t be distributed online.
Drawing on socio-technical understandings of workplace [2, 5], my research contributes an in-depth look at the movement of bodies in a surgical environment. This is motivated by a discomfort with how the HCI community has defined and evaluated “natural user interfaces” and “embodied interactions”. Both descriptors make weighty claims about the configuration of human and machine. The SST community is especially sensitive to the complexities in the organization of action in systems  and so it is with this lens I have looked into a field where new computing inputs will shortly be introduced . We in the gesture research community might consider video of such prototype systems a call to action. The SST literature has demonstrated it is possible to take a more nuanced view of interaction through close observation, including studies of science labs and hospitals [5, 4, 7]. So what would my research contribute?
Medical workplaces are a compelling setting for interaction design. They are compelling because they are important. The work done in medical workplaces has noble overtones (Sethuraman, 2006). Thus the design of technologies supporting valuable work also taken on this air of social good. However there are many other motivations for designing for medical workplaces.
A catheter is “a tubular medical device for insertion into canals, vessels, passageways, or body cavities for diagnostic or therapeutic purposes (as to permit injection or withdrawal of fluids or to keep a passage open)” (catheter, n.d.).
The catheter is a fundamental part of work in cardiac electrophysiology (EP), so much so that often an EP lab is referred to as the “Cath Lab.” The purpose of this document is to define catheter work and why it is important to understand the socio-technical configuration of work in an EP lab.