Study design

This study used interviewer-administered semi-structured interviews with traditional grounded theory approach and iterative thematic analysis to observe the perceived barriers and facilitators to accessing HCS for older adults with serious life limiting illness [17]. The traditional grounded theory approach emphasizes the discovery of emerging patterns from the data and it differs from the constructivist grounded theory approach, in which the researcher co-constructs the theory along with the participants [18, 19]. We engaged in an iterative process between data collection, analysis, and concepts generation and identified emerging patterns from the data until the concepts were bounded and the body of data was deemed saturated.

Study participants and recruitment

We used expert sampling, a type of purposive sampling method, to identify key informants from the PRIM-ER site leadership who could provide information based on their knowledge, experience, and expertise [20, 21]. Participants were physicians and nurses from a variety of EDs across the United States (US) who were actively practicing medicine at their respective EDs during the time of their interviews.

Study procedures

Interviews were conducted via ZOOM teleconferencing services and via phone. The interviews were conducted by the lead author, JDH, using a pre-scripted interview guide (see S1 File). The interviewer has previous training and experience conducting interviews, as well as experience with qualitative analysis. The interview guide was developed by authors JDH, AC, and CRG prior to initiation of the interviews and pilot tested with an emergency medicine physician to ensure clarity and relevance. Interviewees were sent the interview questions in advance of their scheduled interview. All interviewees were educated on the topic and goals of the study and were introduced to the interviewer prior to study initiation.

The interview questions were developed to better understand the provider’s perspectives to facilitators and barriers to HCS available at their respective ED. For this study, we defined HCS as services such as hospice, nursing homes, home health, home care, rehabilitation, physical therapy, occupational therapy, skilled nursing care, etc. The interviewees were reminded prior to starting the interview, and throughout the interview, that we were interested in discussing services that are available outside of their hospital system, such as services provided in a patient’s home or private/community-based organizations for care of older adults with serious life-limiting illness. The interviews were audio-recorded, and notes were documented throughout each interview. Debriefing sessions occurred among the research team after every 3^rd interview to review interviewee responses, adjust the interview guide as needed for clarity and relevance, and assess for data saturation. All interviews occurred between December 2019 –March 2020. All interviews took 45 minutes or less. After 18 interviews, we achieved data saturation, defined as the point when no novel information emerged from the data.

Analysis

We analyzed the notes of the interviews to assess for emerging themes. We transcribed verbatim the recorded audio files of each interview using Landmark Associates Inc. transcription services. The intent of transcribing the recorded audio was to use the participants quotations to develop the themes. Two members of the research team developed the initial codebook, and the codebook was revised at every debriefing session. The research team compared and reviewed codes to achieve consensus on the themes that emerged from the data.

We performed a thematic analysis using an inductive approach [22]. First, we pre-coded the notes of the interviews, generating initial codes. Thereafter, we analyzed the transcripts for additional inductive codes that were missed in the interview notes. All transcripts were independently double coded by members of the research team. We compared codes across all transcripts during the final analysis phase and condensed codes into meaningful themes. Themes were generated with the aid of Dedoose qualitative analysis software version 8.3.43 [23]. Disagreements on coding among independent coders were reconciled via further review and discussion of the transcript until consensus was achieved. We completed cross-investigator and cross-interview analyses. We presented the results as quotations of the study participants, while removing the interviewer’s filler words for clarity and brevity.

Methodological rigor

To enhance the credibility of this study, we reported how the participants were guided by specific questions in the interview guide. We created a codebook, and we iteratively reviewed the codebook as new information emerged from the data. All members of the research team agreed on the codes and the themes in the final codebook. Furthermore, we used the participants’ quotes to create the themes. In ensuring dependability, we have provided the questions in the interview guide and the procedure of the coding process. To demonstrate the transferability of this study, we provided contextual descriptions of the study participants and the analytical process. The lead author maintains the audit trail of all the interview notes, debriefing sessions, and analytical decisions for the purpose of auditability.

Provider familiarity with local HCS

Providers were familiar with local HCS and able to provide a general list of services as well as names of specialty services, such as programs or organizations who provide care specific for patients with dementia.

“I feel like we have the usual home health services that are available. There’s regular home health, hospice, nursing homes, and some adult daycare programs. There are like 50 hospices, quite a huge amount, given that it’s a very small city. Other things that we have down here is an extended acute care hospital system. These are long-term acute care hospitals. I think they have the most of them down here, I think has the highest concentration of them.”—Participant 16, Physician

The most frequently mentioned HCS include hospice, home aides, and nursing homes. Additional services mentioned include home hospice, physical therapy, caregiver respite, skilled nursing care, hospital at home, occupational therapy, home-based palliative care, transportation services, adult day care, counseling, general education, insurance counseling, home medical equipment, and sub-acute rehabilitation. Additionally, interviewees provided details on specialty programs for specific populations their health systems serve such as geriatric specific mixed-service organizations, religious groups (ex. Jewish community services) and/or specific disease cohorts (ex. patients with dementia).

“We have a very large Jewish community here, and they have their own Jewish living facilities as well as home care agencies.”- Participant 6, Nurse

“There’s home hospice resources, and then there’s a program called PACE, Program of All-Inclusive Care for the Elderly. It’s a government grant-funded program for low-income senior citizens. It has the potential for really comprehensive care. They offer all kind of geriatric care, including their own hospice program within it”- Participant 2, Physician

Provider opinions on helpful and desired HCS

Providers found hospice, physical therapy, occupational therapy, and visiting nursing services to be the most helpful HCS to support their patients. The most desired services that providers wish they had access to, or improved access to, include home-based palliative care, full time (24/7) social work/case management for improved referral workflow, greater caregiver support, more general education on HCS, and greater long-term follow-up with patients. All providers reported they feel it is important for emergency providers to be aware of HCS and have systems in place to expedite referrals to these services.

“If we know the community in which our patients are coming from and the resources that are available and the support that can be accessible for those patients, I think overall the care of the patient can be strengthened, and it can be set up for greater success for whatever the outcome we anticipate being for the specific patient. I think that’s a responsibility of all professions to know what resources are available.”- Participant 12, Physician

Benefits HCS provide to patients

Interviewees believed HCS play an important role in helping patients attain care goals. For example, one provider said;

“I feel like it’s [HCS] important to not only take care of their medical needs, but to take care of their holistic needs too. Help them do what they want to do at the end of life and not be contained in a hospital. Help them see who they want to see and do what they want to do and be comfortable, not only medically, but also spiritually, and have them tie up all their loose ends. I think that home health aides and nurses can really help with that that aspect too.”—Participant 14, Nurse

Numerous benefits of HCS for older adults with serious life-limiting illness were identified by our interviewees. The most frequently mentioned benefits include avoiding the ED and inpatient admission, symptom management, and caregiver/family support. Additional benefits identified include supporting activities of daily living/home support, personalized care focusing on cultural and/or religious needs, improved medication management, a more holistic approach to care including spiritual, functional, and emotional support, as well as improved support and education on the process of dying.

“I think end-of-life needs are very complex. Medical nursing care, cultural community and religious aspects can be better addressed in the home or community setting. Those components to care are more at the forefront and can be introduced in a more respectable honor fashion. Community access, information and care—the more heavily that is involved in that community setting, then hopefully that would impact patients having to come into the hospital.

There’s so many challenges, obstacles, and higher risk situations that happen for these patients who are potentially at end of life when they do come into the hospital, and it can make it a less personalized experience for that person in that time. If home and community services were more readily accessible and a little bit more attention or emphasis was placed in that environment, I think it would add to the holistic care of the patient.”- Participant 18, Nurse

Barriers and facilitators to accessing HCS

The most frequently mentioned barriers to accessing HCS include services that are not accessible full-time (24/7) to receive patient referrals, insurance/payment barriers, and the busy nature of the ED. Additional barriers mentioned include inadequate staffing to support HCS referrals, patients refusing services, the location/distance of patients when trying to establish HCS referrals, provider reluctance to refer to services, the requirement to obtain permission from other managing providers (ex. oncologist) prior to a referral, provider preferences to admit patients to the hospital, and having a smaller ED with limited resources.

“When we have so many inpatients that we’re holding—we’re the biggest inpatient unit in the hospital, and then we have ER patients coming in on top of that, I feel like sometimes we don’t always have the time to address those needs.”- Participant 14, Nurse

Interviewees almost unanimously cited case management and social workers as being the greatest facilitators for linking patients to community services.

Staff and procedures for referring patients to HCS

Most interviewees mentioned contacting case management and social workers as their primary means of facilitating a patient’s referral to HCS. For instance, one participant mentioned,

“We do have social work in our ER 24/7. They take the main responsibility for offering these services to the families, to the patients—how to get in touch with resources, things like that.” -Participant 15, Nurse

Additional processes for referral that were frequently mentioned are an assessment of the patient’s physical condition, a referral placed through the electronic medical record, having a discussion with the patient/family, and having the patient meet with a HCS representative. When discussing the key staff members in the referral process, ED-based case management and social workers were most frequently mentioned, followed by general hospital-based case management and social workers.

Limitations

While the geographic spread of our participants was varied, there was no provider representation from the western half of the US due to the timing of the interviews and the COVID-19 pandemic. In addition, most of our interviewees were white females, which limits obtaining perspectives from male providers or providers from minority groups. Interviewees were also providers who served as site leaders for the PRIM-ER research study, which is a national palliative-care study, and thus are highly engaged in this topic area and may not represent the general providers knowledge on this topic. Their advanced knowledge and interest in palliative care due to their involvement in the PRIM-ER study may bias their answers towards a greater familiarity of local HCS compared to other emergency providers.