Open Surgery.

Open surgery consists of cutting skin and tissues to expose organs and other structures. Open surgery provides direct access to the operative site with depth perception and dexterity for one or more sets of hands. Open surgery is the only option for many procedures and for certain kinds of patients (e.g., obese patients, patients with prior surgeries or multiple adhesions). However, open surgery usually requires a long incision, which leaves a visible scar. The trauma caused in gaining access to the organs can result in a more painful recovery, a longer healing process, prolonged hospital stays, a higher risk of infection [5], [6], and sometimes even disability and morbidity [7].

Laparoscopic Surgery.

Minimally invasive procedures have advantages for certain kind of operations [6], [8]: shorter hospital stays, reduced postoperative pain, fewer infections, and better cosmetic outcomes [3]; however, they also have disadvantages for the surgeon: limited dexterity, loss of depth perception, camera instability, awkward movement of instruments and scopes (e.g., fulcrum effect), poor ergonomics, fatigue, and miscommunication caused by the reversed image on the monitor [5], [6].

Robotic-Assisted Surgery.

A surgical robot is a self-powered, computer-controlled manipulator that can be programmed to aid in positioning and manipulating surgical instruments [9]. The robotic manipulator acts as a remote arm extension governed by the surgeon's movements [3], [10]. Robotic-assisted surgical techniques can enable surgeons to carry out more complex tasks than standard laparoscopic surgery and achieve better performance for specific operations [11]. Other advantages include greater dexterity and accuracy, scalable motions, camera stability, improved surgeon ergonomics, elimination of tremor, depth perception, and better patient outcomes [5], [6], [10], [12], [13]. Surgical robots also help hospitals market themselves as cutting edge [11]. However, a robotic system lacks tactile and force feedback [2], [9], affords the surgeon less control over patient safety [14], has the risk of malfunction or failure [14], has risks associated with port placement [15], is bulky, suffers incompatibilities with conventional laparoscopic instruments, has less availability of parts [13], and sometimes requires surgeon troubleshooting [14]. A further disadvantage is that a robotic procedure can take longer than a standard laparoscopic procedure because of increased setup time [16]. Robotic-assisted surgery also costs more than other techniques because of the fixed cost of the robotic system (on average $1.5 million [17], higher maintenance and support costs [12], and the cost of expensive equipment upgrades. Nevertheless, robotic-assisted procedures receive the same reimbursement in the United States as laparoscopic procedures from commercial health insurance and federally administered Medicare [18].

This study focuses on the da Vinci surgical system because of its extensive use in diverse subspecialties [13], [19], [20], [21], [22], [23]. The da Vinci is manufactured by Intuitive Surgical, Inc. Currently, the da Vinci is the only actively marketed surgical system to have received approval from the US Food and Drug Administration and the only system to have received approval for cutting and suturing. The da Vinci consists of a surgeon's console, four robotic arms, and a video tower [6].

The use of robotic systems has been gaining popularity in several surgical subspecialties, such as cardiac, thoracic, urological, gynecological, pediatric, and general surgery [24]. However, despite this growth, some surgeons in these specialties are not inclined to adopt robotic-assisted surgery. Healthcare professionals, and especially physicians, can be slow in adopting new technologies [25]. Common factors in slow adoption have been incorporated in the main constructs and subconstructs of the UTAUT model [26]. Some of these factors have been identified as important to the acceptance of healthcare technology. However, the UTAUT model has not been specifically applied to robotic-assisted surgery. In doing so, the following factors proposed by previous research are relevant:

Self-Governance.

Physicians' adoption of a technology is highly self-governed because of the specialized services they provide [27]. Thus, physicians' perception of a technology is the most important factor in its voluntary use.

Long Training Requirements.

Lack of acceptance may be attributed to the challenges of coping with new devices, adapting new kinds of instrumentation, and learning new operative maneuvers [28]. As a result, surgeons have to interrupt their practice to attend educational programs, seminars, and training sessions. For some devices like surgical robots, uniform training standards among hospitals are lacking.

Lack of Clear Benefits.

Some surgeons believe traditional practices are sufficient to treat their patients. For example, previous research has found that in general surgery the application of robotic technology has not translated into improved patient outcomes [16].

Users.

In relation to the proposed modified UTAUT model, Perceived Usefulness, Facilitating Conditions, and Attitude toward Using Technology were the key facilitators of users' adoption of robotic-assisted surgery (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Proportion of comments categorized according to the modified UTAUT subconstructs.

https://doi.org/10.1371/journal.pone.0016395.t004

Within the Performance Expectancy main construct, the most influential subconstruct (with 221 pro-adoption comments) was Perceived Usefulness. Users were attracted to the many enhanced functions of the robot: better visualization, increased precision, better dexterity, elimination of hand tremor, better suturing, better instrumentation, better angle of placement, easier access, and better ergonomics. Another persuasive element in this construct is that robotic technology improves on standard laparoscopy, allowing the performance of closed surgeries as if they were open. In addition, the users trusted this technology because of the reliability of the robot. In the Extrinsic Motivation subconstruct, improved patient outcomes influenced the surgeons' decision to adopt surgical robots. These positive outcomes include reductions in bleeding, pain, blood clots, infection rate, complication rate, and post-operative adhesions. Personal Outcome Expectations indicated that robotic-assisted surgeons had experienced an increase in patient referrals, which in turn increased job satisfaction.

Within the Social Influence main construct, the most influential subconstruct with 40 pro-adoption comments was Subjective Norm. Surgeons viewed robotic technology as market driven. Because of public advertisement, media coverage, and online information, patients are encouraged to seek out the latest technology. These patients are thus attracted to hospitals and surgeons where robotic-assisted surgery is available and marketed. In this study, the users claimed not to care about the effects of performing robotic-assisted surgery on their personal image; users stated that they performed robotic-assisted surgery, because they wanted to improve patient outcomes.

Within the Facilitating Conditions main construct, the most influential subconstruct (with 208 pro-adoption comments) was Facilitating Conditions. Surgeons expressed confidence in using this technology because of the support provided by Intuitive Surgical, Inc. A technical support representative is onsite most of the time or easily reached by phone.

Attitude Toward Using Technology is another main construct that had a definite impact on the sample's adoption of robotic technology. The construct contains 125 pro-adoption comments made by the robot users. Most users indicated that they have fun performing robotic-assisted surgery. They held high expectations for the future development of this technology, such as the ability to perform surgery remotely and the improved portability of the robot. Development of additional practice-specific instruments will allow the adoption of this technology by other specialties, such as general surgery, cardiac surgery, and plastic surgery. Surgeons also indicated that as they became more experienced using the technology, their technique and comfort level improved. Overall, these surgeons are early adopters [1]; their positive attitude toward using new technology has a great impact on their behavior.

The Leadership subconstruct was added to contextualize the UTAUT model for this study. Most robot users are involved in training and proctoring other surgeons and otherwise helping them to adopt this technology. Some of these surgeons also help the manufacturers make instruments more effective and durable. These improvements allow the specialties to accommodate more procedures.

Nonusers.

In relation to the modified UTAUT model, similar to the users, the results among the nonusers showed that Perceived Usefulness and Facilitating Conditions are the subconstructs that increase surgeons' expected success in adopting robotic-assisted surgery (Table 4).

Within the Performance Expectancy main construct, the most influential subconstruct with 99 pro-adoption comments was Perceived Usefulness. Persuasive elements in this construct were similar to those identified by robot users. The Social Influence main construct has a slightly positive influence on robotic-assisted surgery adoption with 10 pro-adoption comments. Robot nonusers expressed sentiments similar to those of users: that the technology is market-driven. Within the Facilitating Conditions main construct, the Facilitating Conditions subconstruct has 67 pro-adoption comments. Even robot nonusers aware of the good onsite technical support. Training provided by the technical support team was also identified as a positive influence on surgeons' likelihood of adopting robotic technology in the future.

Users.

In relation to the modified UTAUT model, the results among the users showed that Perceived Usefulness, Perceived Ease of Use and Complexity, and Perceived Behavioral Control were the major barriers to the adoption of robotic-assisted surgery (Table 4).

Within the Effort Expectancy main construct, Perceived Ease of Use and Complexity, the only subconstruct, received 86 contra-adoption comments. Among the contra-adoption comments, a steep learning curve was a major barrier. Surgeons were accustomed to feeling the organs, including the amount of pressure being applied to an organ, during an operation. However, that tactile feedback is lost with robotic technology. Surgeons stated they needed to perform at least 25 operations to learn to ‘feel’ with their eyes. Another element that slowed the decision to adopt robotic technology is the setup of the robot, which is cumbersome and time-consuming.

Nonusers.

The results among the nonusers showed that Perceived Usefulness, Personal Outcome Expectation, Perceived Ease of Use and Complexity, and Perceived Behavioral Control were the four major barriers to the adoption of robotic-assisted surgery (Table 4).

Like the users, within the Effort Expectancy main construct, Perceived Ease of Use and Complexity was the best represented subconstruct with 62 contra-adoption comments. Among the contra-adoption comments, a steep learning curve was again a major barrier. A lack of tactile feedback is the main reason. Unlike robot users, nonusers did not want to take the time to learn to ‘feel’ with their eyes, and they did not want to learn on their patients. Another barrier was increased operative time because of the time required to set up the robot.

Performance Expectancy is another main construct that has a negative impact on the adoption of robotic technology among nonusers. Among its subconstructs, Personal Outcome Expectation contained 33 contra-adoption comments, and Extrinsic Motivation contained 24 contra-adoption comments. Some nonusers stated that robotic-assisted surgery is merely a marketing tool. In OB/GYN and urology, some robot nonusers said they were losing patients to surgeons who perform robotic-assisted surgery. In cardiothoracic surgery, some surgeons could not justify using a robot because of limited patient volume for the procedures in which a robot can be used.

Despite the advantage of improved post-operative patient outcomes (Table 1), some nonusers noted an increased chance of bowel injury using robotic technology. They argued that the claim that robotic-assisted surgery provides both better margin data and better cancer treatment is false.

Within the Facilitating Conditions main construct, Perceived Behavioral Control had 42 contra-adoption comments. Nonusers noted that hospitals had not encouraged them to perform surgery with the robot. The perceived return on investment was low: Although the robot is an expensive piece of equipment, reimbursement from private insurance companies and Medicare is the same as for regular laparoscopic procedures. Therefore, because the potential for economic gain seemed marginal, surgeons did not want to invest time in learning to use the robot.

Model Streamlining.

Each main construct of the UTAUT model includes two or more subconstructs. However, some of the subconstructs have similar definitions in the context of robotic-assisted surgery. Because the same comment could often be categorized under multiple subconstructs, streamlining these subconstructs was necessary after the interviews. Hence, similar subconstructs were combined, such as the Perceived Ease of Use, Complexity, and Ease of Use subconstructs of Effort Expectancy. Another issue with UTAUT is that some of the subconstructs constitute the positive and negative anchors of the same dimension, while other subconstructs have no opposing anchors. This deficiency had to be remedied so that we could place pro-adoption and contra-adoption comments under each subconstruct.

Main Construct Addition.

The UTAUT constructs are significant in determining users' behavioral intention and actual use of a technology. In this study, almost half the participants had already adopted robotic technology; some of them were even involved in promoting the technology to their colleagues and assisting robot developers in improving it. When the comments were categorized, some comments did not fit into existing UTAUT constructs. Because many of these comments reflected the surgeons' technology adoption behavior, two new constructs were added, Attitude toward Using Technology and Leadership, thus expanding the UTAUT model to six constructs. In the validation of the model, Venkatesh et al. [26] found the observed relation between Attitude Toward Using Technology and Behavioral Intention to be spurious because both constructs are strongly affected by Performance Expectancy and Effort Expectancy. However, in this qualitative study, the comments on Attitude Toward Using Technology were valuable regardless of their incremental validity. Figure 1 depicts the expanded UTAUT model developed by the aforementioned contextualization methods.

The following methods were used to assess and enhance reliability and validity:

Reliability.

When using human coding, intercoder reliability must be established. Intercoder reliability is the number of times two coders agreed on a unit divided by the total number of units coded [38]. Two coders working separately divided the transcriptions of the interviews into concepts, categorized these comments, and compared the results. The initial intercoder reliability was 74%. Disagreements in coding were discussed until a consensus was achieved.

Validity.

Two validation procedures were employed in this research study: (1) participant data triangulation—facts and opinions provided by one participant were checked by other participants to build a coherent justification for the choice of constructs; and (2) member-checking—each participant evaluated the accuracy of his or her categorized comments. Five participants proposed minor corrections to their comments, which were then incorporated into the results.