Potential US Health Care Savings Based on Clinician Views of Feasible Site-of-Care Shifts

This survey study assesses clinician views on the feasibility of site-of-care shifts from the hospital to alternative care settings and estimates potential net savings that would be associated with these shifts.

Care activities were categorized into four types: facility-based care such as management of acute pneumonia or heart failure; procedures such as cardiac bypass surgery or a breast biopsy; evaluation and management (EM) visits such as a preoperative visit or a new patient visit with physical exam; and ancillary services, which included labs such as STD testing on blood, urine, and genital specimens or genetic testing for cancer for non-hospitalized, non-emergency patients and imaging such as a CT of the spine for non-admitted patients or a MRI of the brain for non-hospitalized, nonemergency patients.
Individual CPT and DRG descriptors were grouped in a way that respondents would easily understand.For example, "percutaneous coronary intervention," "other coronary angiography," "percutaneous coronary intervention (PCI) and stent placement," and "PCI intervention, other or unspecified" were all tied to the care activity: "Cardiac catheterization including coronary angiogram and stent placement (PCI)."Spending and volume of labs and images performed within the context of an admission (i.e., DRG) were included as part of the admission, and not under "ancillary."Observation encounters and emergency department visits were asked about separately by specialty, and all were grouped under "facility-based care."Similarly, anesthesia for specific procedures was grouped with the procedure, and for each procedure relevant respondents were asked to determine whether that procedure could safely take place in the context of an inpatient admission, in a hospital-based outpatient department (which would still be categorized as facility-based in the aggregate responses), an ambulatory surgery center, a physician office, or the patient's home.Creating distributions: For each care activity, an original site of care distribution was created for the survey.This was developed directly from the claims data for commercial and Medicare.The same panel of 33 physicians reviewed each of these distributions to ensure they represented expected reality to avoid any data irregularity issues.

Enablers and barriers:
A list of enablers and barriers was developed based on presurvey interviews with clinicians and a literature review.Nine enablers were included in the survey: availability of alternative sites; physician and patient awareness of alternative sites and their capabilities; convenience of giving or receiving care at alternative sites; the integration of alternative sites with electronic health records; patient out-of-pocket costs incentivizing the use of alternative sites; perceived equivalency of quality of care delivered at alternative sites; favorable reimbursement at alternative sites; technological advancement; and the integration of alternative sites with existing provider workflows and clinical protocols.Four barriers were included: certificate of need laws restricting the development of alternative sites; perceived loss of continuity of care at alternative sites; limited clinician privileges or affiliations with alternative sites; and economic incentives (such as payment arrangements and ownership models).

Survey administration:
A total of 4,608 respondents completed the screening questions for the survey, and 1,069 respondents (23.2%) met the criteria to participate.A total of 1,783 respondents (38.7%) were screened out because the desired number of respondents with their specialties had already been attained; 949 respondents (20.6%) were removed from the pool because they did not pass one of the other screening © 2024 Sahni NR et al.JAMA Network Open.
questions (for example, not currently practicing or practicing fewer than 20 hours per week).A total of 807 respondents (17.5%) were excluded from the results for other factors (for example, failing "sense check" questions, being inactive after initiating the survey).
All respondents practiced more than 20 clinical hours per week.To ensure substantial input from non-physician clinicians, the number of clinicians in certain specialties was increased (for example, for cardiology, 26.8% of the 41 respondents were non-physician clinicians).Physicians who had completed residency before 1980 or after 2019 and non-physician clinicians under the age of 25 or over the age of 80 were excluded.

Survey questions:
Any respondent completing the full survey answered a total of 6 to 21 questions related to care activities, depending on the respondent's specialty.For two specialties with more than 20 care activities (general surgery, which had 41, and radiology, with 27), additional survey respondents were recruited, and respondents were randomized to receive only half of their specialty's care-activity questions.
Each of the 312 survey questions were asked in the format below.