Qualitative data can provide key insights into the decision-making processes of subjects [6], rendering it particularly useful for investigating facilitating factors and barriers in the adoption of novel medical interventions. Toward this aim, the authors conducted four semistructured focus group interviews in June 2018.

Resident focus groups were advertised during the intervention facility’s regularly held, facility-wide resident town halls and were further advertised through flyers posted throughout the facility. Two resident focus groups were conducted by the authors and included 19 residents with and without personal experience of telemedicine calls (see interview guides in Multimedia Appendix 2). Participating residents were asked to fill out anonymous demographic forms, but very few did so, rendering the demographic information collected insufficiently representative of the sample to report. Staff were recruited through the facility’s resident director, who asked staff both on and off shift to consider participating and compensated staff for their time spent in a focus group. The two staff focus groups included 4 safety staff and 3 nurses’ aides from the on-site clinic. Informed consent was obtained from all participants prior to commencing the focus group interviews and recording. The study was approved by the Institutional Review Board of the Longwood Medical Area in Boston, Massachusetts, United States.

Interviews were recorded and professionally transcribed. The first author led qualitative data analysis, using a thematic analysis approach and NVivo 12 software (QSR International) to organize resulting themes. Employing a thematic analysis approach to analyzing the data allowed for the identification of themes, both within each unit of data analysis (ie, each transcribed focus group) and across the data set [7]. Thematic analysis is a six-step qualitative method of finding and interpreting patterns in the data: familiarization with the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, and producing a report [7]. The authors utilized a theory-driven approach to thematic analysis, which involves focusing on a previously identified research objective [7]; in the case of this study, the objective was to identify factors that facilitated or acted as barriers to resident participation in the intervention.

The first step of thematic analysis, familiarization with the data, began during data collection and preparation, with the authors discussing and taking notes between focus groups on initial ideas and potential patterns. This process continued through the first author’s verification of the transcripts for accuracy, which involved multiple close readings of each transcript to confirm alignment with the audio recordings. The second step of thematic analysis, initial coding, involves beginning to identify and group data into relevant categories [7]. The first author looked for manifest and latent data, coding for content explicitly stated in the data, as well as attending to statements that highlighted a participant’s attitude or assumptions [7].

After completing initial coding, the first author organized initial codes into related groups and began to search for preliminary themes within each of these lists of codes organized by topic [7]. At this juncture, the first author presented organized initial codes and resulting preliminary themes to the senior author for feedback, with the two authors discussing these preliminary results until a consensus on preliminary themes was reached. Next, qualitative transcripts were loaded into NVivo 12, and the first author utilized NVivo 12 to organize and review the initial themes identified through this consensus process. This review involved reconsidering both the relevance and importance of each initial theme as it related to the research objective and the congruence of identified themes across the data set as a whole, resulting in a further refined set of themes that were then named and defined [7]. The analysis process concluded with the first author presenting identified themes to all four authors, engaging in additional discussion and refinement of themes until consensus was reached. As all four authors participated in the data collection via co-leading focus groups, engaging in this final step supported the trustworthiness of findings by creating an opportunity to re-examine the extent to which the presented themes sufficiently represented the data and to build consensus on the salience of finalized themes.

EMT-trained safety staff did not perceive telemedicine as a valuable tool, viewing it as undermining their autonomy in decision making and increasing their workload. Reducing resident transports to the ED was not a goal embraced by safety staff, who described approaching all calls with the assumption that transport would likely be necessary.

Safety staff approached calls with the perspective that resident transport to the ED was likely: “We normally respond to all emergencies like probably they are going to consist of transports.” Another safety staff participant reiterated, “...the majority of the calls that I get, they usually warrant transport anyway, so I just go, ‘There’s no point [in a telemedicine consult].’ That’s kind of where I’m at with it.” If the medical need for transport appeared uncertain, staff preferred to call an ambulance: “If there is a borderline [case]...maybe you err on the side of just calling EMS [emergency medical services] out to handle that assessment.” Operating from the perspective that transport was likely necessary and/or best in the case of uncertainty, staff viewed telemedicine as an unnecessary obstacle, delaying the resident’s needed emergency transport: “...we have to wait for the telemedicine conference, it still ends up in transport anyway, so it’s just delaying the inevitable.” Thus, safety staff also perceived telemedicine as undermining their professional judgment:

This safety staff participant independently determined that transport was necessary, and regarded this decision as undermined by the telemedicine physician. When transport did end up being necessary in this case, the safety staff participant’s view of telemedicine consult as an obstacle rather than an aide was cemented.

Staff expressed that telemedicine was of limited use for the community’s residents: “...the majority of stuff we deal with here, it doesn’t fit the bill. Old people don’t bounce. They fall and break.” Another participant described common conditions for which residents call safety staff: “...shortness of breath, and fevers...but most of the time, they’re 102, they’re septic, they’re coughing up phlegm, their lungs are nasty. You’re like, ‘They need to go.’” One participant indicated that falls with pain are frequent: “I always tell them, ‘I can’t see through you, so you’re going to have to go get an x-ray or an MRI [magnetic resonance imaging] or whatever it needs to be.’” Staff highlighted that they were accustomed to making independent triage decisions and that the telemedicine physician “asking the same questions” was not helpful.

Safety staff felt that the program’s goal to deter ED visits was the wrong goal and that telemedicine would better address minor medical concerns at the on-site clinic. Safety staff indicated that “the resistance you have from our department is we’re emergency response,” whose “primary goal is security, medical response.” The nursing staff at the on-site clinic “deal with the nonemergency, ‘Hey, I don’t feel well,’ and they walk into their office looking for the Tylenol.” Safety staff indicated more willingness to utilize the telemedicine consult if the safety staff called to the case had already determined for themselves that transport was not necessary. A participant indicated that telemedicine could be beneficial if and when “I don’t think it warrants going to the hospital. It may be an alternative to going to their own doctor or to an urgent care. If I don’t see emergency treatment as really necessary.” The safety staff’s self-description of their role as emergency responders and the emphasis on telemedicine as potentially more suitable for nonemergency cases highlights their perception of telemedicine intervention as inappropriate for emergency use.

Safety staff indicated that telemedicine increased their workload, describing the software as difficult to navigate: “Just the sheer amount of time that it takes...typically there’s three [safety staff on duty]. We have well over 900 residents we’re checking on. Sometimes the telemedicine can take a little longer than I think is warranted.” This was echoed by another staff member who described telemedicine as increasing the difficulty of juggling multiple demands: “It can be time-consuming. We have a lot of things going on, more than one medical going on...it can be very time-consuming.” Safety staff participants also described a number of technical issues that acted as a barrier to use, such as inconsistent wireless internet connectivity throughout the facility: “...that’s come up in certain parts of the campus...We just try and make do and move around if possible.” In combination with the safety staff’s perception of telemedicine as inappropriate for emergency situations, the staff’s experience of telemedicine as additional work led to the perception of telemedicine as adding burden without adding value.

Safety staff also reported that residents were reluctant to use telemedicine. One safety staff member said, “Honestly, they don’t directly request it, and I think a lot of that has to do with they don’t want to change from the old-school days of actually seeing a doctor. A lot of them, they go, ‘No, they can’t do anything for me over the phone or over the tablet. I actually need to go in.’” Another staff member agreed, saying, “...when I do offer it to residents, they only want to see their doctor and then they want to be in person...to talk to someone who knows them personally.” One participant described this perceived reluctance as a lack of comfort with receiving medical care through an unfamiliar medium:

In contrast to staff perceptions, resident focus group participants described multiple benefits of using telemedicine. Residents emphasized that avoiding an ambulance transport and having their concerns successfully resolved by the telemedicine physician were the most significant benefits of a telemedicine consult.

Residents expressed a strong interest in avoiding trips to the ED whenever possible, mentioning the long wait times, financial costs, and potential health risks of ED visits. Residents interviewed identified avoiding an ED visit as a primary benefit of using the telemedicine intervention. A resident described satisfaction with the quality of the telemedicine consults she received, emphasizing that an unnecessary transfer to the ED was prevented in both of her experiences with the telemedicine intervention:

Another resident emphasized the value of telemedicine in preventing emergency transport:

Other residents who had not yet tried telemedicine indicated a willingness to do so if the opportunity arose. A resident expressed the following:

Residents also described an unanticipated benefit of telemedicine: encouraging hesitant residents to go to the emergency room when necessary. A resident described his experience as follows:

The participant described how he agreed to the transport and was diagnosed with a “fatal heart condition” in the ED. He concluded, “Everybody I’ve talked to about telemedicine, I just tell them they saved my life...convincing me that it was necessary to go. I wouldn’t have gone.”

Resident focus group data indicated that residents were satisfied with the results of telemedicine use and/or expressed a willingness to try it. A few residents did indicate concern that telemedicine would delay treatment:

However, the majority of residents expressed confidence in their ability to determine when telemedicine might be useful.