Participants across different medical specialties expected that a novel sensor would enhance operative outcomes by offering real-time feedback on the surgeons’ exerted force. All participants identified scenarios within their disciplines where the application of inappropriate force causes harm to patients or prevents task completion. Therefore, all participants predicted that real-time feedback about their exerted force could increase patient safety within their operating environment: “I think if you know exactly how much force you’re applying, that may make you rethink what manoeuvres you’re doing to deliver that baby” (OG2).

Participants agreed that providing real-time feedback on exerted force using a sensor could positively affect patients’ health. Beyond patient safety, several participants indicated that the provision of force feedback could enhance their operation speed. During specific procedures, practitioners must carefully approach the correct magnitude of force; suitable force feedback would enable them to skip time-expensive approaches to reach adequate force levels. This would improve efficiency and effectiveness and could also create financial gains:

While the feedback could increase surgeons’ effectiveness and efficiency, it could also create a psychological benefit, which enhances their operations’ outcomes. For example, many surgeons reported that receiving information about how much force they are applying would boost their comfort and provide assurance about their performance: “It is a problem in today’s increasingly justiciable times that we still rely on feelings” (AN1). Furthermore, participants reported that their increased confidence could impact their operative performance positively. Finally, the indirect force feedback could also benefit operative outcomes in terms of patients’ psychological well-being during surgery. That is, feedback on the surgeons’ exerted force was proposed to comfort patients who are not under anesthesia: “It can be a tool for trust building in scenarios where women have been previously sexually assaulted or had very traumatic internal examinations” (OG3).

In summary, the interviews demonstrated that surgeons could profit from using a novel force sensor during practice to apply appropriate magnitudes of force, which could enhance operative effectiveness and efficiency.

Offering real-time feedback on surgeons’ exerted force through the use of a novel sensor was identified to benefit the training of junior doctors, as existing medical training methods are inappropriate and unstructured. Participants reported that, in most cases, junior doctors learn to perform surgery through “learning by doing” (GS1). Unsurprisingly, the current shortcomings of surgical training for junior doctors were judged problematic by participants who feared the outcomes of untrained surgeons’ mistakes:

Furthermore, the interviewees criticized that, where surgical trainees are learning in practice, the present teaching style is dominated by the apprentice model, whereby senior surgeons instruct junior surgeons through qualitative feedback, which is exceptionally challenging for trainees due to individual differences between trainers and trainees. Most participants agreed that there is a demand for objective feedback about their exerted force. Thus, the prospect of a force sensor that generates real-time feedback on forces elicited enthusiasm for its implementation in training inside operating theaters:

While many participants acknowledged the benefit of the sensor in practical training, others proposed that the sensor could be used in simulation training. According to the interviewees, existing methods and models for simulation training are unavailable, unsuitable, or unaffordable. Two-thirds of all participants expressed that a novel force sensor could demonstrate high usability in simulation training outside the operating theater. The main benefit would be that junior surgeons could develop their tactile senses and learn how much force is appropriate in specific situations before applying their skills to actual patients: “I think it would be an absolutely incredible tool because that way you could practice as much as you want before you apply it to a human being and potentially sever that facial nerve” (PD2).

Altogether, the interviews demonstrated that junior practitioners could profit most from a novel force sensor as a source of objective feedback on how much force they are and should be applying. Thus, they could use it to refine their skills outside and to guide their forces inside the operating theater.

Across all medical specialties and experience levels, participants could imagine using a force sensor to assist them in various maneuvers in their work and identified maneuver-specific characteristics. Almost all participants expressed that they would appreciate an additional source of feedback in procedures where there is no external indication of how much force they are applying; thus, they must rely on their experience to determine appropriate magnitudes of pressure: “From my experience, it’s just based on what you’ve experienced previously because it’s a blind procedure, you can’t see what you’re doing” (OG1).

The interviews uncovered different reasons for surgeons’ limited awareness of their exerted force. For example, visual feedback from the surgical site can be obstructed or masked where sensitive tissue is hidden through other tissue or blood flow at the operating site: “They’re bleeding, and you can’t see whether you’re in the axial direction of the root” (DS4).

In some procedures, surgeons may be reliant on auditory feedback from patients to determine whether the force applied is appropriate. However, the use of anesthesia may result in application of inappropriate force not being apparent until after the anesthesia has worn off: “I can only say that if I don’t have any complications postoperatively, then obviously the procedure was fine, there’s not much more feedback there really” (DS3).

Thus, there is a significant opportunity to apply a force sensor during maneuvers where surgeons are unaware of their exerted force due to the absence of alternative sensory input such as direct visual or auditory patient feedback.

Procedures that require force sensing to avoid harm to the patient are usually defined by the practitioners’ interaction with particular bodily tissue. Participants frequently referenced the opportunity to apply a novel force sensor when dealing with tissue sensitive to excessive intraoperative force: “With nerves and vessels, you obviously have to be extremely careful because it depends on the size, but they’re usually quite delicate” (PD3).

Although most participants named fine, delicate tissues among those most sensitive to excessive pressure, one participant added: “When you’re required to apply quite larger forces for certain things, then it becomes quite important to know exactly how much” (PD3). The most force-sensitive tissue types are not necessarily the most delicate ones. Instead, participants labeled tissue as sensitive to pressure based on its importance for bodily functioning. The tissue types most frequently classified as critical for bodily functioning by participants across all disciplines were blood vessels and nerves:

Additionally, participants pronounced that tissue types are sensitive to the amount of force applied when they have poor healing abilities. In that context, bones were frequently mentioned by general surgeons:

Meanwhile, in dental surgery, connective tissue, which supplies other structures with blood, is deemed poorly healing: “If unnecessary pressure is applied to an instrument, this can lead to tissue tears, which are then difficult to manage” (DS1). Thus, the sensor was also considered to be useful when practitioners handle coarse tissue.

In addition to force-sensitive tissues, those procedures that require force sensing to avoid causing harm to the patient are usually defined by the surgeons’ use of specific instruments: those that are susceptible to transmitting excessive force. Across medical specialties, participants reported using a variety of instruments that, when applied inappropriately, can cause significant harm to their patients and thus could benefit from real-time feedback: “If you clamp an artery too much, it’s broken” (GS2). However, one participant remarked that the potential to cause harm to the patient is not evident for all tools:

The interviews have shown that one thing instruments that can apply excessive force to human body tissue have in common is that the force needed to reach the operative goal is specified, but the mechanics of the instrument do not define practitioners’ force application: “With most of the devices that we use for suturing and such things, the pressure is already predefined by the mechanics” (PD1).

Thus, many surgical instruments are designed in a way that prevents the application of excessive forces. Examples of such instruments were electrocautery scissors, forceps, and saws:

Even with acknowledging the potential benefit of real-time feedback on exerted forces, many participants doubted the successful definition of specific safe force ranges to provide them with informative real-time feedback on the force they exert on their patients. Some participants argued that, despite their extensive experience, they do not know what the right or wrong amount of force to be applied is during specific procedures: “Although I have been doing this for many years, I wouldn’t know at what pressure a problem for the bone occurs” (DS4). Furthermore, participants struggled to imagine how such force ranges would be determined while considering all variables that determine appropriate forces: “I would probably find it difficult because there are so many factors that are variable due to the different instruments and different tissues that I can’t standardise” (PD1).

Participants doubted the establishment of safe force ranges for pressure measured by a novel sensor because they reported exerting a complex composite of forces on their instruments. For instance, some participants remarked that they apply pressure with multiple areas on the hand during instrumental work. Others mentioned also applying pulls and torques during procedures with significant physical components:

While the componential exertion of force makes it difficult to determine how much force is adequate when measured at the sensor, most participants also raised concerns about the vast inter- and intraindividual differences between patients’ tissues. Most participants feared that these differences in tissue quality would prevent the definition of force ranges calibrated to patient needs:

Nonetheless, some interviewees mentioned that individual differences could be predicted using advanced imaging methods or accounted for by creating force ranges for standardized patient groups: “If you have a bell curve for different types of people you could sort of establish a range of forces that could potentially show the doctors where it gets critical” (PD3). However, this approach cannot account for intraindividual tissue differences such as the effects of scarring from earlier operations. Consequently, the tissue properties’ unpredictability could limit the feedback’s informativeness. Overall, participants identified several issues that could complicate establishing safe force ranges adapted to the specific procedure and patient.

Many participants were doubtful about the need for a sensor and the feedback it generates and posited different reasons surgeons might not need a sensor implemented in their work. First, force feedback was voted redundant whenever surgeons could rely on other sources of information to determine how much pressure they should apply. For some participants, their years of clinical experience were sufficient information: “I think once you’ve done a couple of thousand, you don’t necessarily need a quantitative measure because your experience and your sense of feel is used to it” (AN2).

In addition to relying on their experience, several participants explained that, in most procedures, visual or tactile feedback is available at the operating site, making the additional force feedback redundant:

While sensory information might make force feedback from a novel sensor redundant, participants found scenarios in which any information about their exerted force is unnecessary. Whenever the potential to cause harm is very slight, participants claimed no need for feedback on their exerted forces. This was often the case in coarse procedures where the required force is unconstrained:

Although some participants judged the force feedback unnecessary for their work, others questioned the costs relative to the benefits: “I’m not sure how useful this would be only because you’ve got to think whether it’s worth paying for something like that when actually removing something on the abdomen or the leg is a pretty easy procedure in the first place” (PD2).

While these comments referred to preoperative costs, other participants expected adverse operative outcomes. Some participants were worried about the negative impact on their tactile senses from wearing the sensor: “Surgeons work a lot with the tactility of the fingertip and if there’s something in front of it, then that can also be a disadvantage” (GS2). Additionally, some participants were cautious about the impact the interpretation of the real-time feedback would have on their performance: “I would like to be looking at the lesion not at something else to look at pressure” (PD2). Furthermore, experienced surgeons were worried about junior surgeons’ neglect of organic information: “The disadvantage of technology is sometimes that young anaesthetists no longer look at people” (AN1). Further, some medical practitioners who regularly interact with their patients during surgery explained that real-time feedback could stress the patients. Therefore, the interviews revealed a series of procedure- and practitioner-specific factors that limit the value of real-time feedback on surgeons’ intraoperative forces.

While existing force-sensing systems applied in medicine are predominantly applied to surgical tools, a novel force sensor was developed to be applied to surgical gloves. Many surgeons appreciated the prospective benefits of wearing a sensor on their surgical gloves during instrumental surgery. DS1 proposed implementation of a force sensor on the glove when operators expect to use many instruments in quick succession: “In order to be able to keep up with the speed, I would suggest to rather have it on your finger.” Other proposed advantages of wearing the sensor on the glove were the relative ease of installation and use:

Looking ahead, an advanced, wireless force sensor would allow for effortless installation and use at both the tool-tissue interface and the surgeon-tool interface. However, in addition to its implementation during instrument use, many participants suggested the application of the sensor to the glove for manual examinations, particularly during diagnostic maneuvers:

During instrumental and manual manipulation, surgeons expressed various preferences regarding a sensor’s local application on a glove. Participants’ preference for where a sensor should be applied depended on the instruments used during the respective maneuver. However, different operators also described having personal preferences regarding using certain instruments: “This is a bit specific, because different people will use the syringe in different ways” (AN1). While in most cases, a sensor’s application on a glove was voted preferable by the participants, some interviewees voiced their preference for force measurement at the instrument to know the magnitude of the force exerted at the tool-tissue interface rather than the surgeon-tool interface:

The interviews demonstrated that force measurement conducted at the tool-tissue interface was generally more requested when the forces exerted by the surgeons were complex composites of different hand areas:

Many participants expressed their opinion regarding the feedback modality when considering the reception of real-time feedback during interventions, which were mostly based on a prototype used in the opening video (Multimedia Appendix 1): “But I think if you were to devise an instrument and it had a red bar on it and my force was going towards the red bar or the higher end of the red bar, then I would start paying attention to that” (OG2).

However, most participants who discussed the impact of the visual feedback criticized that the visual feedback could distract surgeons from the operating site: “But a display means that you have to look at it again, and of course you don’t look away when you’re operating” (GS2). Instead, those participants were pleading for visual feedback “where you don’t have to widen your field of vision” (GS3).

Auditory feedback was initially considered to be inappropriate by the developers of the sensor due to the existing audible alarms in the operating theater. This view was confirmed by those participants who reported working in a noisy environment and supported by those working in particularly stressful conditions, where audible feedback was thought to stress the surgeons and the patients:

Nonetheless, many interviewees suggested auditory feedback as a viable alternative to visual feedback. The primary benefit of auditory feedback is that it can quickly catch the surgeons’ attention: “The one that we use more is the sound because as soon as it starts to go off the radar with the beeping, you kind of know that that’s the area that you’re interested in” (PD3).

Other participants, who criticized auditory and visual feedback, suggested administering tactile feedback: “I suppose something a bit like the electric toothbrushes, if you’re pressing too hard it will vibrate” (AN2). However, other participants contradicted this again since: “if your fingertips vibrate all of a sudden, it might be strange” (GS3). Altogether, the interviews revealed that surgeons must choose whatever feedback works best for them, depending on which senses are occupied by environmental stimuli and which are available.

Using the sensor could benefit surgeons in two ways: by enhancing the operative outcomes of interventions and improving the training of junior surgeons. We found that participants expected that wearing the sensor during surgery could enhance operative outcomes, reinforcing the notion that real-time feedback on exerted forces can improve the quality of surgery [1,18,35]. Nonetheless, like Ebrahimi et al [35], most existing research has only assessed the impact of feedback on the exertion of intraoperative force measured at the tool-tissue interface. Our findings extend this knowledge by suggesting that feedback on exerted forces could also enhance operative outcomes when measured outside the patient’s body at the surgeon-tool interface.

Most participants agreed that the sensor could help junior surgeons develop a feeling for adequate intraoperative forces, overcoming the difficulties of learning from qualitative feedback described in the scientific literature [2,22]. We found that a novel force sensor could be used for training purposes both in and out of the operating theater. This supports Rhienmora et al [36], who suggested that the feedback serves to develop a feel for adequate forces and can be interpreted for assessment purposes. Thus, the sensor could accelerate surgical skill acquisition through objective feedback in real time and serve as a novel quantifiable metric.

First, our results revealed that surgeons should use a force sensor when performing maneuvers where other sources of information on intraoperative forces are masked. These results corroborate existing studies that have attributed poor surgical outcomes to errors of perception rather than poor motor skills [37,38] and initially inspired researchers to test the provision of feedback on exerted forces during surgery [39]. Because organic force feedback is often masked in standard open surgeries, we propose that feedback on exerted forces from a novel sensor should be provided in those scenarios.

Second, our results demonstrated that a force sensor should be applied when tissues are susceptible to inappropriate force due to their relative importance for bodily functioning. This finding supports previous research that has highlighted the importance of force sensing when surgeons deal with delicate structures in retinal or nervous tissue, as they require minimal force [11,12,25]. Our findings extend this view, as they propose that a particular tissue’s importance for bodily functioning does not just depend on its delicacy but also on that tissue’s healing qualities. The current results demonstrate that a sensor could also have high applicability when dealing with more coarse tissue with poor healing abilities, such as bone tissue.

The application of a novel sensor when dealing with instruments prone to cause harm to the patient was also identified as an opportunity, as previous research has indicated that the design of interactive medical devices can provoke user error [40]. Our study supports these findings since it revealed that where the exertion of intraoperative force is not predefined by the instrument’s mechanics, the chance to harm patient tissue is high (in contrast to instruments where the mechanics define the exertion of force). Although our interviews revealed some examples of the latter (eg, electrocautery scissors, forceps, and saws), this has not been reported in prior literature.

The implementation of accurate force ranges is challenged by vast individual differences between patients’ tissues. This finding is supported by previous studies that identified significant differences in tissue qualities between individuals [26] and within individuals [27]. Building on the findings by Jacquet et al [41], which demonstrate that significant variability in skin mechanical properties persists even after controlling for sex, age, and BMI, our findings suggest that fixed force ranges for real-time feedback may be insufficiently informative for patients whose tissue qualities deviate from the norm. Thus, the sensor should be applied in cases where inter- and intraindividual differences are minor or predictable.

Additionally, implementing force ranges is challenged by the multimodality of the forces that surgeons use to manipulate their patients. Apart from the linear manual pressure applied on the fingertip, which the sensor can achieve, intraoperative forces can be complemented by torques or pulls, which are activated by different parts of the surgeon’s body, such as the wrist or the biceps [42]. Therefore, this sensor should only be applied in scenarios where the exertion of intraoperative force is less complex.

Finally, surgeons’ ability to rely on their experience and other information sources challenges the need for real-time feedback on exerted forces to be displayed. Our findings support existing conceptions of direct visual and tactile feedback’s impact on the operator’s judgment [43]. During surgical manipulation of human tissue, visuotactile neurons in the human brain allow surgeons to combine visual and haptic information into one judgment [44]. Our study supports the existing conception that surgeons judge their exerted force by relying on their experience and that surgical clinical judgment necessitates decisions blending explicit knowledge derived from evidence with tacit knowledge drawn from experience [45]. Thus, the sensor might be redundant when surgeons have enough explicit and tacit knowledge to base their judgment on. However, in high-risk situations or low-visibility scenarios, the sensor may still be of use to experienced surgeons.

Although participants did not interact with a physical prototype, their feedback highlights key design considerations for future iterations and provides an initial framework for usability testing in subsequent studies. Importantly, these design considerations need to be adapted for different surgical specialties. For example, dentists may require sensor placement optimized for fine instrument handling, whereas obstetricians may need configurations suited for manual examinations. Opie et al [46] explored the use of a sensorized glove in a usability study to assist with vaginal examinations during childbirth. It consisted of 18 participants either specializing in childbirth or with the intent to work in women’s health. That study involved creating a realistic environment where the usability of the device could be assessed in a safe and effective manner. Participant feedback from that study had a direct influence on the design of the sensorized glove for future iterations.

We identified that a novel force sensor and the associated real-time feedback should be implemented in multiple areas of the hand depending on the instrument. Additionally, by measuring intraoperative force at the surgeon-tool interface, a novel force sensor could also be used for manual treatments and diagnostic procedures to improve surgeons’ application of adequate forces during manual procedures as previously suggested in physiotherapy where they showed a temporary performance improvement when provided with force feedback [47,48].

When the force exerted onto tissue is a complex composite of forces applied by the hand, the sensor should be at the instrument tip. Aggravi et al [29] measured the contact forces exerted on three surgical tools and discovered a significant difference in forces applied by the thumb and index finger. Although these findings imply that a glove could have multiple force sensors, interpreting the different feedback strands could be challenging. This was supported by our participants, who suggested that measuring force at the tooltip to receive information on the sum of the force applied by different hand parts was more practical.

Finally, the indirect force feedback modality should depend on the accessibility of the visual, auditory, and tactile senses. Previous research discussed the benefits and drawbacks of providing visual, auditory, and tactile feedback in clinical settings [49-52], leading to the conclusion that the choice of feedback modality cannot be generalized. These findings were confirmed by a recent survey among health care professionals who stressed the importance of predetermining the best procedure-specific feedback modality based on the load on different senses in the operating environment [53]. Ultimately, it might come down to creating custom configurations of more than one feedback modality, as introduced by Tajadura-Jiménez et al [54], who introduced multisensory integration processes used to alter the perceived materiality of a touched surface to shape people’s touch behavior.