Abstract
The use of robotics is becoming widespread in healthcare. However, little is known about how robotics can affect the relationship with patients during an emergency or how it impacts clinicians in their organization and work. As a hospital responding to the consequences of the COVID-19 pandemic “ASST dei Sette Laghi” (A7L) in Varese, Italy, had to react quickly to protect its staff from infection while coping with high budgetary pressure as prices of Personal Protection Equipment (PPE) increased rapidly. In response, it introduced six semi-autonomous robots to mediate interactions between staff and patients. Thanks to the cooperation of multiple departments, A7L implemented the solution in less than 10 weeks. It reduced risks to staff and outlay for PPE. However, the characteristics of the robots affected staff's perceptions. This case study reviews critical issues faced by A7L in introducing these devices and recommendations for the path forward.
Introduction
On February 21, 2020, the first Italian case of COVID-19, caused by the SARS-CoV-2 virus, was diagnosed in Codogno. 1 Italy soon became one of the world’s hardest-hit countries in numbers of infections and deaths. On June 1, 2020—the day lockdown restrictions were officially loosened, Italy’s death toll topped 33,000. The emergency resulted in the overcrowding of Italian hospitals and extraordinary demand for Personal Protection Equipment (PPE). The consequential rise in PPE prices posed challenges to hospital budgets.
The number of cases combined with shortages in PPE supply significantly raised the risk of infection for health workers. According to the Italian National Medical Association, during the first wave of the pandemic in Italy (February 21-June 1, 2020), almost 30,000 health workers contracted the SARS-CoV-2 virus while treating infected patients and 172 died as a result.2,3 A key issue in this context has thus been how to interact and communicate with COVID-19 patients, who are isolated from both healthcare staff and family. This isolation negatively affects the quality of care, organizational dynamics, and patient experience.
To address the issue, the “ASST dei Sette Laghi” (A7L) a healthcare facility located in the city of Varese, Italy, decided to follow recent trends and introduce advanced robots as intermediaries between staff and patients.4,5 Recent studies show that individual-care robots may play an important role in healthcare services delivery and post-medical treatment.6,7 However, little pre-pandemic research could be found evaluating the use of robots in responding to infectious disease outbreaks and as a means for interacting with isolated patients. 8
The objective of this case study is to present and discuss an implementation of advanced robots in a healthcare facility as a response to the COVID-19 pandemic. We present the innovative solution in use during the first wave of the pandemic, together with recommendations and lessons learned from the process.
Advanced robots taking care of humans in healthcare organizations
Robotics has witnessed increasing use in healthcare operations in recent years. 9 The domain of healthcare is diversified and advanced robotics finds application in a wide range of activities, from prevention and diagnostics to surgery, therapy, and care support. 10 Previous literature has categorized healthcare robots by the type of task assigned to them. For example, Okamura et al. identified six fields of application for healthcare robotics 11 : (1) surgical and interventional robotics, (2) robotic replacement of diminished/lost function, (3) robot-assisted recovery and rehabilitation, (4) behavioural therapy, (5) personalized care for special-needs populations, and (6) wellness/health promotion. A more recent review by Joseph and colleagues focused only on humanoid robots and distinguished three categories of these devices used in healthcare 12 : (1) teleoperated service robots, (2) humanoid robots for pain relief, and (3) humanoid robots for ageing population care. These categorizations highlight the broad range of possible robotic functions in healthcare.
There are many ways that robots can potentially add value to healthcare servces. 10 In general, replacing or enhancing specific human activities, especially complex and repetitive ones, by robots reduces labour costs and can increase the quality of service. Yet, while there is strong demand for robotics to tackle growing problems in healthcare, there are numerous barrier issues such as regulatory framework, technological maturity, and societal acceptance. 10
However, most studies investigate the value of robots in performing complex or routine-based activities. By contrast, there is still much to be learned regarding their potential roles in cognitive and emotional tasks, where humans are still expected to be superior to robots. This area of study gains importance with the growing attention to patient-centric practices. 13
A critical factor in the successful application of advanced robots in healthcare is acceptance by human workers.10,14 Technology acceptance is commonly understood as one’s perception of the technology after using it or observing it in use and refers to one’s willingness to adopt the technology. Identification of applications and operations that influence this acceptance supports managerial decision-making towards successful implementation strategies. However, to date, very few studies have investigated the acceptance of healthcare robots in the context of infectious diseases and specifically in managing isolated patients. A notable exception is Kraft and Smart, who investigated the acceptance of teleoperated robots that allow healthcare workers to perform some of their duties at a safe distance from infected patients. 8 However, more case studies and first-hand experience are needed to improve decision-making in this challenging context.
Method
Context
Daily use of PPE at A7L hospital during the first phase of the pandemic.
Abbreviations: PPE, personal protection equipment; A7L, ASST dei Sette Laghi; FFP2, Filtering facepiece 2 standard.
They decided to implement new, innovative forms of healthcare worker-patient communication. Given the unprecedented emergency, the goal was to quickly (1) reduce infection risks to health workers, (2) reduce budget pressure in the hospitals treating COVID-19 patients, and additionally, given the severity and unpredictability of COVID-19 impact at the beginning of the emergency, and (3) choose a solution that was easy to implement and learn so as not to put any additional pressure on the medical staff.
Data sources
We collected data used in this study through (1) semi-structured interviews, (2) observations, (3) collection of textual materials from secondary sources, and (4) a quantitative survey. Some of the co-authors of this study are employed at the A7L, thus facilitating access to the documents and the details of the project and the devices introduced in both departments. We also collected first-hand insights from the employees involved in the introduction of the solution. Furthermore, to evaluate the adoption of the solution, we used a survey to collect opinions about the robots from 76 A7L healthcare workers (employees of the two departments). The survey was reviewed for research ethics and approved by the hospital board. All the respondents gave their consent to collect their responses in compliance with applicable legislation and research ethics. We asked the employees to fill out a short questionnaire about technology acceptance and adoption, using items from a well-established technology acceptance model. 15 The participants had to evaluate the robots in terms of trust towards the robot, ease of use, robot performance, intention to use in the future, and general relationship with them on a 1-7 Likert scale. The response rates for SICU and IDU were 20% and 37%, respectively.
Case presentation
SICU and IDU overview in the first phase of the pandemic (March 13, 2020 - May 3, 2020).
Abbreviations: SICU, semi-intensive care unit; IDU, infectious disease unit; MD, Medical doctor.
Two models of robot were used. Three units of an IVO robot produced by ELMEC Informatica were introduced in the IDU. This robot takes the form of a rack mounted on gyroscopic wheels and carrying a touchscreen display and synchronized sensors. Three units of an anthropomorphic, child-size robot nicknamed “Tommy,” produced by Omitech (Figure 1), were introduced to the SICU. Comparison of both robots - IDU on the left, SICU on the right. 
Both models were equipped with HD cameras, a touchscreen display, and audio receivers and transmitters. They differed mainly by shape/appearance. Both robots had the same aims: To help monitor patient vital signs by reading equipment present in the room via camera; To enable robot-mediated communication between the healthcare staff and patients by using audio and touchscreen functions.
Healthcare staff could control the robots using dedicated software installed on a PC or smartphone. Messages could be typed remotely on a keyboard and transmitted to the patients via the robots. On their part, the patients could use the touchscreen to provide feedback to the healthcare staff. Moreover, “Tommy” was able to measure patients' vital signs (heart rate/oxygen saturation - SpO2), alerting staff in the event of unusual situations (eg, a person on the floor), and to perform pre-triage remotely. Thus, the robots became an intermediary in communication with infected patients, which decreased the amount of contact time between patients and healthcare staff and thus the risk of the infection to the latter. The IT department monitored the process remotely and could intervene in case of need; IT staff was ready to support the healthcare staff and maintain the robots (see Figure 2 for the organizational scheme). Organizational scheme.
Evaluation of the solution
From the economic point of view, the introduction of both robot models to provide support in interactions with COVID-19 patients was a success. After the hospital started using the robots to mediate interaction with COVID-19-positive patients, the estimated reduction in the daily use of PPE (Filtering facepiece 2 standard, gloves, and biohazard suits) was 18% (Table 2), somewhat relieving pressure on the monthly budget. We report that the robots were donated to A7L; however, the impact of acquisition options (purchase, rent, and lease) is not within the scope of this study.
In terms of acceptance by staff, the introduction of the robots was a partial success. While they fulfilled their tasks, not all employees were enthusiastic about this solution (Figure 3 and Table 3 show sample results per unit). Responses by SICU personnel trended positive on all questions, while those by their colleagues in the IDU were significantly more negative (on average, 2 points less on a 1-7 Likert scale). Given that both robots offered the same functionality, we may hypothesize that the robots’ form does matter when using them as a communication tool in the healthcare context. The SICU staff who interacted with the anthropomorphic robot, which presented social cues to the users, evaluated the robot far more positively than IDU staff on all questions. Among other issues, IDU workers reported incompatibility with other technologies and difficulty learning how to use it and in applying it in daily operations. Not surprisingly, while SICU workers expressed a high intention to use “Tommy” in the future, IDU employees were less enthusiastic (average score of 6.15 vs 3.14 out of 7, respectively). Evaluation survey results. Sample survey answers per unit. Note: Higher percentage between the two units is marked with bold. Abbreviations: SICU, semi-intensive care unit; IDU, infectious disease unit. 
Discussion
Robots offer both advantages and disadvantages in managing emergencies like the one caused by COVID-19. As the case study shows, implementing a robot in doctor/nurse-patient communication can reduce budget pressure driven by increased PPE use. A second key benefit was that by not directly contacting infected patients, healthcare staff felt less exposed to risk and thus safer. The implementation of such a solution may lead to relevant transformation within healthcare organizations, in terms of both processes and culture. 13
However, our findings suggest that the decision to introduce such a robot cannot be dictated by budget needs alone. Change management issues and impact on ways of working must be taken into account. The case showed that degree of organizational adaptation conditions ability to exploit the full potential of the robot. To overcome this issue, the introduction of robots in healthcare facilities should be characterized by high levels of organizational cooperation across departments, disciplines, and competencies (including non-medical departments). The case of A7L shows that the more proactively departments are involved in the process, the easier its implementation across the hospital. Such a move should be carefully studied and planned in consultation with healthcare staff to work towards the redesign of clinical processes embedding patient experience dimensions, such as communication processes. This is a crucial step in gaining acceptance of the type of technological innovation we discuss herein.
Furthermore, humanoid robots appear to arouse far fewer negative reactions from the staff than less anthropomorphic forms, a fact that should be considered in choosing device type. Previous research has demonstrated that people accept robots better if they are similar to humans, provided they do not trigger negative reactions as a result (eg the “uncanny valley” effect—an unsettling feeling people get when interacting with robots that resemble humans almost but imperfectly). 16 In any case, while interface and human-like robots can potentially help, results are likely to be sub-optimum if attention is not dedicated to (a) the evolution of organizational culture and (b) changes in ways of working.
Based on the experience at A7L, we can provide the following recommendations for implementation of robots in healthcare facilities: An interdisciplinary team should lead the organizational redesign and implementation phase. ASST dei Sette Laghi assembled a team across all involved departments and studied the solution’s feasibility before making any binding decisions. Managers who implement robotic solutions should study the robot’s compatibility with the existing equipment and have a clear understanding of the personnel needs that a robot could meet. Provide proper training and highlight organizational implications and benefits to the employees who will interact with the robots. Many IT-based projects fail because of employee reluctance to “get on board.”
17
The employees in the IDU in our study complained of difficulty in using their robot. Thus, close consultation and interaction with staff is crucial to achieving smooth, efficient, and thus successful implementation. When possible, use anthropomorphic robots as they create more positive reactions among individuals. Based on the studied case and the literature on robotics, the chances of success increase the more the employees “like” the robot.16,18
Limitations and future research
Our case study is based only on one facility. The results may be informative for administrative staff of similar facilities but are not generalizable. Further studies are needed to explore how other institutions respond to analogous problems. There is also a cultural issue to be investigated, since only Italy-based employees participated in this study.
Furthermore, our case study presents a particular situation in which the advanced robot was introduced as an emergency response under epidemic pressure. Thus, all managerial decisions were conditioned by a highly unusual and unpredictable situation: the epidemic at that time. Future research could investigate acceptance of new robotic solutions in situations where emergency pressure on staff is not a predominant feature.
Conclusions
Advanced robots in healthcare facilities can reduce budget pressure in infectious disease emergencies by reducing need for PPE. In this study, we presented a case study of an Italian healthcare facility (A7L) that took the opportunity to respond to the COVID-19 emergency by implementing advanced robots in its operations. Overall, A7L administrators considered the undertaking to be a success. New advanced robots have been installed since November 2020 (the second wave of the virus in Italy) to guarantee even greater support and further explore the impact and potentials of using robots in clinical practice. The experience of A7L presented in this case study contributes to healthcare management literature by providing first-hand experience from a hospital under pressure. It thus stands as a contribution to the discussion on the use of these solutions in clinical facilities.
