Early methods of anesthesia delivery provided a total dose of anesthesia agent. More sophisticated inhalers used mechanical indicators to preset flow, providing a limited measure of safety and control. Mask and inhaler were diverse in design. They ranged in acceptance and evolved in a parallel timeframe; a commonality was the requirement to be portable, allowing early anesthetists to take their practice to the patient. Primarily these devices were designed by anesthetists around their own anthropometric capability, requiring the practitioner to observe the patient whilst controlling the mask or inhaler. Symmetry in use and close proximately to the patient is identified as a predominate design criterion that allowed for personal comfort and control (Figure 2).

Administration required two handed use; being a symmetrical system, either hand could be used. This freedom in administrative positioning, patient viewing and handedness required skill and control, a dexterity only gained from experience. Observation of archived equipment reveals that these early devices could be used either left or right handed, however with a need for a precise and controlled seal over the patients face, instructions (Boyle 1907), publications (Thomas 1975) and an early ICI training film (Thomson 1944) show the right hand was to dominate the mask. The patient’s physiology was representative of today’s monitoring interface, all in the visual field of the anesthetist. While cloth, mask and bottle, and equally the inhaler, are no longer used, their influence on design is marked by their duration of use into the 1950s.

The second stage of evolution occurred in replacing the hand held craft of anesthesia with scientific apparatus and mechanical interaction. Similar to inhalers, these were designed by anesthetists. These were portable devices more suitable for hospitals and the operating theatre; comprising standalone rather than hand held technology. By transferring anesthesia administration from the drop bottle and inhaler, to calibrated flowmeters and vaporizers, anesthetists were able to control more complex technology. While the design of these was flawed in mis-connections, poor calibration, and materials that perished, wore and failed, the ergonomics and positional activity was an anesthetist derived carryover from mask and bottle behavior. The diversity and experimental nature of the majority of apparatus conceived between 1900 and the 1930s continued the provision for established methods of use in line with the practitioner as designer. The apparatus remained biased to the tradition of close proximity to patient, symmetry, retention of the mask and predominantly left handed use, i.e. right hand on mask and left hand available for equipment.

The early apparatus designs were the results of innovation and modification amongst many practitioners, Wilkinson (2002) describes the origins of the popular Boyle apparatus, as an adaption of the Marshall apparatus that in turn had been influenced by Gwathmey’s designs. These early machines were mounted on a stand to be at a comfortable working height, and demonstrated a direct correlation between technology innovation and the transfer and acceptance of earlier use methods. This is ascertained firstly from the placement of apparatus components in a comprehensible left handed flow pattern from left to right to the outlet. Secondly, the mask was a continuation of earlier designs that still required dexterity and control to place appropriately. This design stage retained an established use method originating in right hand in control of the rag, mask or inhaler and typically left handed operation (Figure 2). The success of the Boyle apparatus may partially reside in anesthetists’ acceptance of their own iterative design and an evolutional approach to use methods based on normal use and recognition of ergonomic consistency.

The third stage occurs when the apparatus technology became a table based system, a consolidation that offered the anesthetist a work surface, with drawers; a workplace. To accommodate the table, the apparatus equipment was mounted at the rear, leading to a design where established technology becomes standardized through design integration. This created a workspace in the operating room and promoted a change in equipment positioning from the left of the patient to the right. This changed established methods of use, handedness and reach. The table based design dictated a behavioral change that opposed the flexibile use of the mask, inhaler and apparatus. In the case of the anglo-commonwealth Boyle machine, this was compounded by retaining a traditional left handed component arrangement to a machine that was situated and used on the right. Component arrangement was later revised to suit right hand use in the 1950s and illustrates how standardization and reliance on slow iterative change became a hallmark of the anesthesia machine. The Boyle machine focused on providing a design solution to the earlier diversity of individual practitioner arrangements in a multiple user hospital setting. The machine format biased normal use to the chassis or furniture rather than the practitioner’s anthropometrics. The right handed position that opposes the use methods of earlier devices is a direct result of compromise in integrating technology with furniture. The table defines an area for auxiliary equipment and a workspace dominated by the prevalent right handed. This contradicted the established method of mask application, leading to anesthetists creating, conforming and adopting a new awkward use method.

During this period, the table based format was adopted in the USA (Thomas 1975). One example exists of practitioner design that questions industry consensus and demonstrates a personal wish to mitigate behavioral change. Illustrated in A.G.McKenzie’s (2008) historical note, the Gillies machine of 1951 shows the interface and controls centered and brought to the front of the table.

The workstation differs little in physical form from its predecessor. While continuing to retain the table, its increased size, the addition of mechanical ventilation systems with long breathing circuits, and workplace congestion has forced placement further away from the patient (Figure 2). Its inception is marked by a 30 years design hiatus that established an awkward yet familiar convention in use methods aligned to the table based machine design. Weinger (1999) examined studies that led to awareness of patient safety during the 1980s, including anesthetists’ contributions to patient monitoring technologies, pulse oximetry and capnography. These additions are described by Cooper (1978) and Westhorpe (1992) as a haphazard application, in turning the simple machine format into a “Christmas tree”.

Recognizing the resulting ergonomic inadequacies of stacking new technology, investigations by Duri (1973) Cooper (1978), and Boquet (1980) promoted the argument that the current machine was in need of a complete redesign, resembling an “accident waiting to happen” (Cooper 1978 and Westhorpe 1992). The premise was that in stepping out of the 1950s convention, a new workstation design that both looked and worked differently to resolve ergonomic issues would be accepted as better. Complete redesign rather than integration was applied as a method to resolve the anesthetist’s workplace. Novel equipment resulted; the Gambro Engstrom 2000, Engstrom Elsa/EAS, Physio BV Physioflex, Dräger Cicero, and Siemens Kion machines reflected Duri, Cooper and Boquet’s research. While advancing capability, technology and patient safety in anesthesia, these designs were not universally accepted by the anesthesia community. In time the technology was iteratively adapted into machines retaining the traditional table based format and awkward use method.

Since the 1990s manufactures have sought to resolve the collection of components by integrating them into one device and mitigating ergonomic difficulties with adjustable screens and left hand versions. Recognition of dexterity in operating knobs, dials and touch screens as opposed to the patient and mask is prevalent in these designs. This places expectations on the anesthetist’s physical capability, both in the distance of reach and observation of patient and machine, as machine position in the operating room is not standardized i.e. the patient is in front and the workstation (monitoring and control of anesthesia) is behind (Dalley 2004).

Figure 2

Fig 2 Illustrations of equipment in use and corresponding diagrams of developing workplace awkwardness. source: A Geoffrey Kaye Museum of Anaesthetic History, Australian and New Zealand College of Anaesthetists, Melbourne, Australia. B (Boyle 1907). C (Thomas 1975)

The critical incident technique was introduced to anesthesia by Cooper, Newbower, Long, & Mc-Peek (1978) identifying the relationship between human error and equipment. Reducing human error has since remained a significant aim in anesthesia, however little consideration is given to pre-machine use methods and the long term impact of inherited habits. Evidence of these is incorporated in drawings, photographs and historical descriptions as legacies of pioneering anesthetists (Thomas 1975 and Maltby 2002).

Challenges to a convention that united workplace analysis with design, were initiated by the predesign investigation by Duri (1973), who promoted the role of the anesthesiologist in design and used human factors methods to examine the environment of the anesthetist. In an attempt to provide useful information for the redesign of the complete man-machine system, Duri quantified awkwardness in a series of link analysis diagrams of machine placement, and the time the anesthesiologist’s attention was distracted from the patient. A similar study by Seagull (2004), termed awkwardness as “dispersion of attention” and proposed that these measures are conspicuously absent in research but useful to designers. Duri’s investigation used memomotion filming, ₁ timings and link analysis including interviews, whereas Seagull used eye tracking video to analyze task related events. These two investigations span 30 years, have similar hypotheses, and while allowing for technical advances they also share methods and comparable results.

Similar hypotheses exist in Boquet’s 1980 analysis of the anesthesia workplace, however Boquet uses human factors analysis, advanced techniques in eye tracking, and industrial design methods to create design prototypes. Comparing the illustration of Boquet’s 1980 prototype to Wilkinson’s 1930s description of joining the table to apparatus, it would seem Boquet has separated the machine back into table and apparatus and reversed 50 years in search of normal behaviors to resolve his design. Boquet’s task and link analysis and the recording of interactions is a reactive design method. Therefore, any design conclusions are based purely on the individual tasks, sequence and frequency constrained by the studied machine format. Consequently Boquet’s conclusion of an apparatus type layout as better, implies that anesthetist’s use methods were only new habits to overcome design deficiencies in the machine.

The technology inspired design of Cooper’s (1978) Boston Anesthesia System (BAS), differs from Boquet’s industrial design bias and the notion of understanding compositional interaction through task analysis. Cooper’s hypothesis applied research ‘by design’, starting with an indictment of the mechanical anesthesia machine as unsafe where as digital electronics were potentially safer. A comparison can be made to the Siemens Kion workstation (released in 2001) a radical ergonomic layout consisting of three independently rotating parts: column, user interface, and trolley. Effectively a product designer’s intuitive hypothesis that does not communicate a practitioner centered use (Liu 2003). Neither of these concepts acknowledged the prior or established behavior of anesthetists as important precedents, but instead presented a compromise in physical format, endeavoring to coerce a new use method that was yet be established or understood.

Recently David Liu (2010) evaluated head-mounted display technology as a means to resolve awkwardness by providing the anesthetist with continuous visual information while monitoring the patient. Liu’s study has in effect returned the anesthetist’s positioning back more than 100 years, to the mask and bottle pose and normal behavior aligned to the activity rather than to the machine.

Studies of the anesthesia workplace, and the questioning of established convention through new designs, exhibit the perceived importance of human factors study. On a micro level much has been achieved through iterative design, new technologies and attention to detail, conversely these endeavors have inspired little change in the man-machine relationship. The scales remain tipped in favor of patient safety and information management, with the anesthetist progressing from normal use positioning where patient and device are one, to covering and monitoring more than 180 degrees. The anesthesia machine, as a table based system, bred an expertise and acceptance for a format that disregards awkwardness by favoring convention and safe change. This can be substantiated by the demise of novel machines discussed earlier that invested in human factors and challenged tradition yet suffered limited acceptance.