Perceived Costs of Care Influence Post-Acute Care Choices by Clinicians, Patients, and Caregivers

Abstract

Background/objective:

Older adults frequently receive post-acute care (PAC) after hospital discharge, but little is known about how perceived costs influence PAC choices. This research study sought to understand how clinicians, patients, and their caregivers evaluate the cost of Skilled Nursing Facility (SNF) care in their decisions about whether to utilize SNFs post hospital discharge.

Design:

Guided by principles of Social Constructivist theory, we conducted qualitative interpretative study using semi-structured interviews with clinicians, patients and caregivers.

Setting:

Three SNFs and three hospitals located in an urban area. Purposive sampling was used to maximize variability in SNFs, hospitals, units within hospitals, and staff.

Participants:

One hundred four participants: 25 hospital clinicians, 20 SNF clinicians, 20 hospital patients, 15 SNF patients, 14 hospital caregivers, and 10 SNF caregivers who were directly involved in patients’ transition from acute hospitalization to SNFs.

Measurements:

Central themes related to how perceived costs of care influence post-acute care choices

Results:

Clinicians, patients, and caregivers did not understand the nuances of SNF insurance coverage or out-of-pocket costs. They felt constrained by insurance coverage in their discharge disposition choices and faced delays in hospital discharge due to insurance authorization processes. Some clinicians reacted to these constraints by “documenting failure” — sending patients home to “fail” so they could justify SNF to insurers. Others changed their recommendations to provide patients “some” post-discharge care, even if inadequate, because of cost constraints. Clinicians discussed conserving resources to take maximal advantage of insurance-covered SNF days.

Conclusion:

Improving patient and caregiver understanding about costs and constraints of PAC would improve decision-making. There is a need for improved comprehension of cost and insurance coverage of SNF care for informed patient and provider decision making at the time of hospital discharge.

Introduction

Care transitions after hospitalization are common and represent a time of vulnerability during which adverse events are common and costs are substantial^1,2. Post-acute care (PAC) in skilled nursing facilities (SNFs) and home health (HH) is growing in prevalence ³ but represents the largest single source of variation in regional Medicare spending⁴, suggesting a lack of standardization in how these resources are used. While PAC costs have received increasing national attention^5,6, little is known about the role perceived costs play at a micro level – that is, at the level of decision-making about PAC around the time of hospital discharge.

Costs play a significant role in decision-making in other contexts ^7–9, and prior research suggests both providers and patients are uncertain about other critical aspects of PAC that inform decision-making.^10–15 Perceived costs may drive decision-making about PAC and could represent an important lever for improving decision-making and the value of PAC delivered. Indeed, many payment reforms involving PAC – from first-dollar cost sharing in Medicare Advantage plans¹⁶, to Bundled Payments for Care Improvement, to Medicare Spending per Beneficiary metrics – are intended to address costs through changes in decision-making and processes of care between the hospital and PAC.

We sought to understand the role perceived costs played in decision-making about PAC around the time of hospital discharge, looking at decisions made by interdisciplinary team members, patients, and caregivers in both the hospital and in SNF. We hypothesized that insurance constraints and qualifications for SNF would play a dominant role in clinician decision-making, but that patients and caregivers would be unaware of cost implications of the decision to go to a SNF.

Study Data and Methods

Interviews

This research was conducted as a secondary analysis of a larger study that sought to identify how decisions about PAC are made by patients, caregivers, and clinicians in the hospital and SNF.^10,11,17 Between February 2016 and September 2017, we conducted in-depth, semi-structured interviews with clinicians, patients, and caregivers purposively sampled from three hospitals and three SNFs located in an urban area.

To maximize variability in care settings and patient populations, our participants were recruited from a Veterans Affairs hospital, a university teaching hospital, and a safety-net community hospital serving indigent and immigrant populations. Our three SNFs included a VA-affiliated SNF and two community SNFs, one of which provided only short-term rehabilitative care. Hospital clinicians included physicians, nurses, social workers, case managers, physical therapists, and occupational therapists. SNF clinicians included physicians, nurse practitioners, administrators, directors of nursing, floor nurses, hospital liaisons, therapists, and social workers. Patients were identified by these participating clinicians if they were 65 years or older, were going to be discharged to SNF or recently were admitted to a SNF from the hospital and did not have significant cognitive impairment that would prevent them from making healthcare decisions. Caregivers were identified by asking the enrolled patients, “Besides your medical team, could you please tell us who is the one person who helps you most with your health?”

Qualitative analysts (EL, AL, RA) conducted one-on-one interviews with participants at hospitals, SNFs, or over the phone in a semi-private environment using a guide derived from the Ideal Transition of Care framework.¹⁸ We stopped interviews when we reached data saturation.¹⁹ All interviews were audio recorded, transcribed verbatim, validated and coded using ATLAS.ti version 7.5.11 (Scientific Software Development, Berlin, Germany). This study was approved by the Colorado Multiple Institutional Review Board.

Analysis

We employed a team-based approach to thematic analysis to explore emergent themes regarding insurance and cost when deciding on SNF care options. Guided by principles of Social Constructivist theory ²⁰, we used directed content analysis techniques ²¹ via a matrix analysis design ²² to understand and describe participants’ decision-making with insurance and cost as a primary factor. Matrix analysis is used to display categorized data in cells using rows and columns to observe what appears. We created columns of roles we interviewed and rows of what interview questions elicited discussions about cost of care. Our analysis was guided by the following questions: 1) When do cost and insurance-related discussions come up in the decision-making process? 2) How does insurance and cost influence participants’ SNF options? 3) What are the pressure points for clinicians in justifying patient referral to SNFs? Codes were developed inductively through group discussion until saturation was reached; coding consistency checks were made through out the coding phase and discrepancies were resolved by team discussion.^{14, 15}

Results

We conducted a total of 104 interviews with 25 hospital clinicians, 20 SNF clinicians, 20 patients and 14 caregivers in the hospital, and 15 patients and 10 caregivers in the SNF (Table 1). Overall, clinicians, patients, and their caregivers identified three major cost-related influences on SNF care decision making: 1) lack of patient, caregiver and clinician understanding of insurance coverage and out of pocket cost (Table 2); 2) constraints of insurance in treatment and discharge disposition choice (Table 3); and 3) consequences of cost of care constraints (Table 4).

Table 1:

Study participant characteristics

Characteristic	Clinician N= 45 (%)	Patient N=34 (%)	Caregiver N=23 (%)
Location of Participant
Hospital	25 (55)	19 (55)	13 ()
Skilled nursing facility	20 (44)	15 (44)	10 ()
Role of Participant
MD/Hospitalist	8 (17)	-	-
Nurses	10 (22)	-	-
Physical Therapy	8 (17)	-	-
Occupational Therapy	8 (17)	-	-
Case Manager	2 (4)	-	-
Social Worker	7 (15)	-	-
Administration	4 (8)	-	-
Demographics
Age, mean years (range)	-	75.2 (60–96)	57.95 (29–85)
Women	33 (73)	15 (44)	14 (60)
Veteran	3 (6)	19 (55)	2 (8)
Race/ethnicity
White/Caucasian	30 (66)	24 (70)	12 (52)
Black/African American	1 (2)	4 (11)	2 (8)
Mixed/biracial	4 (8)	3 (8)	2 (8)
Asian	2 (4)	1 (2)	2 (8)
Latino/Hispanic	2 (4)	0 (0)	1 (4)
Native American	1 (2)	1 (2)	1 (4)
No Data	2 (4)	0 (0)	1 (4)
Educational level Attained
Grade school	0 (0)	5 (14)	1 (4)
High school/GED	1 (2)	7 (20)	2 (8)
Some college	2 (4)	7 (20)	9 (39)
College graduate	8 (17)	11(32)	7 (30)
Post-graduate	30 (66)	4 (11)	3 (13)
No Data	0 (0)
Annual income ($)
<30,000	-	17 (50)	-
30,000–50,000	-	7 (20)	-
>50,000	-	7 (20)	-
Household size of patient
1	-	15 (44)	-
2–3	-	14 (41)	-
>3	-	4 (11)	-
Insurance coverage of patient***
Medicare	-	30 (88)	-
VHA	-	10 (29)	-
Private insurance	-	12 (35)	-
Medicaid	-	4 (11)	-
Other/no coverage	-	9 (26)	-
Caregiver relationship to patient
Child	-	-	12 (52)
Spouse	-	-	4 (17)
Sibling	-	-	3 (13)
Other	-	-	4 (17)

^***Respondents could have more than one payer for insurance coverage. Rounding may result in values >100%.

Table 2:

Participants description of their understanding of insurance coverage and out-of-pocket cost for patients being discharged to SNFs

Theme	Quotations	Role
1. Lack of Clinician, Patient and Caregiver Understanding of Insurance Coverage and Out-of-Pocket Cost	A. I would say 85% to 90% of the time, they really, whatever we say, they [physicians] are on board with, unless there is an issue with coverage, like, a Medicaid patient or someone with no SNF benefits either needs to go home or to acute rehab, and so if there is an issue with that, with the insurance payment, those sorts of things, then it becomes more of a well you need to change your recommendation because that’s not an available option. But I would say 85 to 90% of the time, the medical teams are very receptive to the recommendations that we make.	Hospital Physical Therapist
	B. If the insurance don’t back me, it’s [SNF is] out of the question. If the insurance don’t back me, I might as well try to stay here and if the insurance insists that they can’t back me on this, I’d go home. I know they’ve got a time limit on everything, but they would give me to more time if I was in an injury, during that car crash, I cracked my back, I broke my back or whatever. I kind of understand what goes on, you know?... I don’t want another thing if the insurance company doesn’t back me. Who’s gonna pay for [NAME] bill?	Hospital patient
	C. “I also had to appeal, not only for an extended stay, but for extended visitations because you have to also have, not respite care, but CNA care is not paid for by Medicare after a certain period of time, and so, it would have been nice to know that all that was gonna happen cause I ended up paying $300 a day. For my mother the first time. $300 a day, not a week, not a month. $300 a day and having to go through the appeal process and work a regular job and raise my daughter as a single parent and try to keep up with whether my mother was gonna live or die.”	Hospital Caregiver
	D. “ I think, a lot of patients look at their benefit and they say oh, Medicare is gonna pay for 100 days of skilled treatment for me, but what they, and the same with Kaiser, Kaiser runs the bus and the insurance companies run the bus on these patients, so basically and you know, you have to make progress for insurance to pay for your stay and to continue to pay for you to receive physical therapy, occupational therapy and 24 hour nursing care, you have to be making progress, and a lot of the patients don’t understand that cause a lot of them, if they just plateau, they say well I have 100 days by my Medicare, and a lot of them don’t understand that process, so that’s also a lot of education and hand-holding on this end…”	SNF nurse
	E. “Oh, I don’t know. If my insurance runs out tomorrow, I don’t know what I’m gonna do.”	SNF Patient

Table 3.

Participants description of constraints of insurance in transitioning patients to SNFs

Theme	Quotations	Role
2. Constraints of Insurance in Treatment and Discharge Disposition Choice	A. They told me that because of some rules in Medicare, that I would probably have to spend three days in the hospital before I can get into one of those places, and then I might not be able to get in because they may be full because they’re not that large of places…with that many empty beds.	University Patient
	B. Yeah, the only thing that throws us off is insurances, based on what they’re going to allow us to treat for.	SNF Physical Therapist
	C. If they are very limited with insurance options, and I know that those insurance option choices are not gonna give them a great experience, that’s hard… for me because then you’re operating on such a limited amount of choice.	University Social Worker
	D. “Well, if they’re not progressing once they get to that setting then it’s no longer going to be covered by insurance. They will cover them under their Medicare A days for X amount of time, then they go into their copay days, but ultimately, if they’re not demonstrating certain progress, then they’re not going to be able to stay in that setting.”	Hospital Occupational Therapist
	E. …. you can…get the patient on long-term Medicaid, so that means instead of…going to a rehab facility to get rehab, they will go to like a nursing facility to get just care… we know that they really need rehab, but their insurance doesn’t pay for rehab and yet they can’t go home because family can’t take care of them so then we send them to a facility for 30 days knowing that they’re not going to get a lot and that’s hard, so yes, they’re safe because they are not at home and somebody is caring for them “24 hours a day” but they’re deconditioning more because now they’re laying in a bed and so that’s one of those, too. It just doesn’t feel great, but it’s the best option? So it’s kind of like OK, this is the lesser of two evils so we’ve gotta go with it.	University Social Worker
	F. “I would be eligible for going to a civilian rehab…if I had Medicare I guess it is…when she [the social worker] found out I didn’t have it, then she said well then your only option really is the CLC [Community Living Center]”	VA Hospital Patient
	G. It’s finding a bed that works with that person’s insurance, it’s getting the insurance authorizations, so there’s constant barriers that we might be facing or you know, they have a bed available but after this whole process with getting insurance authorization, now that bed is no longer available.	Hospital Social Worker

Table 4.

Consequences of insurance constraints on patients, caregivers and clinicians

Theme	Quotations	Role
3. Consequences of constraints • Expressing frustration and anxiety Clinician workarounds • Documenting Failure • Changing Recommendations • Rationing Resources	A. There are a lot of …medically unnecessary delays in care. There are times when we decide …someone needs sub-acute rehab, but they don’t have their Medicare benefits or their Medicaid benefits, and so they sit here waiting for a week or two waiting … and at some point, somebody gets frustrated and they go home, like either the patient is like, I’m not waiting here anymore doing nothing. We need to go to rehab because I’m getting weaker, and I’d rather just go home and get home PT, even if that’s not our original recommendation.	Hospital physician
	B. We send people where we think they need more and we end up sending them home and it’s kind of like we have to almost get like documented failure, if that makes sense, like we have to send them home so that they can fail at home so that they can come back in and then we can try different tacks to get them the facility that they need.	Hospital Nurse
	C. They typically pay for 20 days and any days after that is going to be a copayment on the patient’s behalf, and I just kind of feel that’s in direct conflict with what the surgeon has said…	SNF Caregiver
	D. We make our discharge recommendation. That’s within our scope of practice. That’s what’s needed from us. That’s what we’re being asked to do, not only because of our expertise, but also because our payer sources need to know what do they need next and they’re asking us to tell them. Then the therapist feels like their recommendation isn’t being followed and the social worker is saying well, this is all that insurance will cover, and the therapist wants to fight harder to say, can we push them more to do the acute, like we’ll make sure the communication is clear in our notes, as we try all of the time, but if there’s an issue, can we call and talk to the medical director at the insurance or whatever. So, what will happen then sometimes is that the case manager, social worker will ask us to change our notes, which we don’t like to do just based on that recommendation, and we’re talking about this because insurance is all over the place, but it does really become an insur…, a payer source issue [...]	Hospital Physical therapist
	E. Hospitals are getting just as much pressure as we are. And to get people out the door sooner. And as a result of that, they are coming to me sicker, but my facilities do not have any more increased capacity to take care of sicker patients. I have a patient right now who had a massive bleed in his head. He has a trach and he’s not on a ventilator, but he’s on a collar to deliver continuous oxygen. If he develops a pneumonia, I’m gonna have to send him right back immediately because he doesn’t have, because he’s already so compromised.	SNF physician
	F. Right now, I think it’s really hard to do the right thing. It often takes heroic effort and even beyond that, impossible amounts of time to do this for every patient, so that’s why I said I was selective in who I do it for, not based on the person, but based on who I think needs it the most or would benefit from it the most or is at highest risk for harm if with it not being done, but I would love if it was very easy to have a conversation with the person who is going to be taking care of them on the other end, if it was very easy to ensure that the information that was critical got to that person or those people in a time-sensitive and time appropriate manner, if there was feedback for the provider about how things went and how things could be better for the future, cause we all want to improve, it’s not that people do the same crappy job over and over because they want to, if that happens, it’s because they don’t know or it’s challenging to do better.	Hospital physician

Lack of clinician, patient and caregiver, understanding of insurance coverage and out-of- pocket cost

Most hospital clinicians - other than social workers - said that they find costs confusing, and that they relied on social workers to navigate the insurance process. Some hospital clinicians discussed lack of awareness regarding covered services for patients and how surprised patients were after learning their insurance doesn’t cover certain services and/or how unaware clinicians were about covered services. For example, when a patient was cared for under observation status and SNF care was recommended at the end of the hospital stay, explaining a longer inpatient stay was required to access SNF benefits to the patient was difficult. While a “usual” inpatient stay followed by SNF might seem straightforward, clinicians commonly deal with “exceptions” to this rule such as patients with Medicaid (which does not cover post-acute care) or no payer for post-acute care, Medicare Advantage plans, or private pay plans that require them to defer to specialized staff such as social workers. On the other hand, some providers discussed adjusting their PAC recommendations based on what insurances will pay for rather than what they believe the patient needs (Table 2, Quote 1A).

Hospital patients and caregivers looked to their inpatient care team to help them understand PAC and what was covered. Hospital patients and caregivers did sense that insurance was a primary determining factor in what kind of PAC the clinician team might recommend, but did not know what will actually be covered, how much they may face in out-of-pocket costs, and how much cost of care uncertainity may affect patient’s recovery (Table 2, Quote 1B). Hospitalized patient’s caregivers who had prior PAC experience were concerned about paying out-of-pocket costs once patients reached the maximum number of insurance-covered SNF days (Table 2, Quote 1C).

In contrast, SNF clinicians did not mention confusion about insurance coverage. Some SNF clinicians noted that educating patients about cost of care upon admission is not ideal because insurance decision has already been made and patients are unaware of the insurance coverage requirements. For example, patients were unaware that insurance will cover their SNF stay only if they were making progress (Table 2, Quote 1D). They did, however, discuss insurance coverage as a primary means of determining patients’ eligibility for admission to their facility.

Once patients arrived at the SNF, their concern as well as their caregivers continued as to whether their care will be covered and whether they will be allowed to stay and continue to receive treatment if they need it but their insurance coverage runs out. Patients and caregivers’ expectations often did not match the reality of copays, care not covered, and limited number of SNF days (Table 2, Quote 1E).

Constraints of insurance in treatment and discharge disposition choice

Participants discussed various levels of constraint of cost on SNF care and its impact on decision making. These constraints included timing of care, whether costs/coverage permits SNF at all, limitations in coverage once at the SNF, quality of SNF patients could go to, and room availability at a desired SNF.

Participants frequently brought up discharge delays related to Medicare’s three-night rule when the patient was felt to be ready to go to SNF but had not completed enough hospital days (Table 3, Quote 2A) Additionally, participants identified instances where they assessed whether costs/coverage permit SNF placement at all. Some clinicians transparently said they recommend PAC based on what insurance allows them to treat (Table 3, Quote 2B & 2C) whereas others discussed number of days and intensity duration limits per patients based on their coverage (Table 3, Quote 2D). Hospital clinicians elaborated on the limitations of patients with only Medicaid coverage, requiring them to be admitted to respite care or long-term nursing home care to receive any PAC, which clinicians felt was inadequate but better than going home with no PAC. Similar to other healthcare decisions, insurance coverage was a significant constraint. For example, Veterans who only had VA benefits were limited to being placed at VA- affiliated SNFs regardless of whether this might be the best choice to meet their rehabilitation needs or preferences, while those with Medicaid coverage end up being placed in a long-term facility with respite instead of a rehabilitation facility. (Table 3, Quote 2E & 2F)

While the insurance coverage issue is being resolved, participants could lose a bed in their desired or approved SNF. Beds might have been available when the process started, but by the time patients are ready to be discharged, there was no guarantee that the approved SNF will be able to admit patients. (Table 3, Quote 2G)

Consequences of constraints

Participants reacted to cost of care dilemmas in two main ways: 1) Expressing frustration and anxiety; 2) finding work-arounds.

Expressing frustration and anxiety

Clinicians discussed frustrations associated with insurance related delays in transitions and patients leaving against medical advice, because they are not receiving the optimal care they deserve putting them at risk for complications. For example, some hospital clinicians discussed medically unnecessary delays in patient discharge from hospitals while trying to figure out a patient’s insurance status and coverage eligibility. These delays occurred after clinicians decided the patient was ready for SNF placement, but the insurance verification and eligibility process was time-intensive. This unnecessary hospital stay was frustrating enough that some patients left the hospital against medical advice or were discharged home without receiving the care they needed at a SNF (Table 4, Quote 3A & 3B).

Additionally, patients and caregivers said they had regular concerns and worries about whether insurance coverage would end before the patient recovered while at the SNF. They remained concerned about their SNF stay and the cost associated with treatments. This SNF cost anxiety made some patients re-assess their decision to go to a SNF. Additionally, patients and caregivers expressed frustrations about clear discordance between a clinician’s recommendations and insurance allowance in covered services (Table 4, Quote 3C).

Clinician workarounds

Beyond expressing frustration, clinicians responded to these pressures through workarounds where they “documented failure,” changed PAC recommendations to get patients “something” even though that care setting might not be best for that patient, and/or conserved resources by deciding the optimal time to discharge patients to maximize insurance- covered SNF days.

Documenting Failure

Hospital clinicians discussed using times when SNF was not covered to “document failure” at home, to make the case for SNF after what was an inevitable hospital readmission given the patient was discharged home from their index stay without proper support. These clinicians admitted that they know what is best for the patient, but by documenting failure, they are letting insurance know the urgent need for patient placement in a SNF (Table 4, Quote 3B).

Changing Recommendations

Hospital clinicians reported changing recommendations and treatment options based on what insurance allows and that those recommendations are not always what they believe is best for the patient. Several clinicians discussed changing recommendations based on what insurance allows to provide “something” for the patient even though it was not the best course of treatment. (Table 4, Quote 3D) For example, clinicians discussed changing their recommendation from a SNF to recommending patients be placed in a respite nursing home where there is no skilled nursing provided because it was safer for patients than going home. (Table 4, Quote 3E) A few SNF clinicians talked about being the catch-all place for patients who are discharged quicker and sicker from the hospitals but do not have insurance coverage for full rehabilitation. (Table 4, Quote 3E)

Conserving Resources

Hospital clinicians also discussed conserving resources by determining when best to discharge patients to maximize insurance-covered days. In these cases, discharge from the hospital was delayed to maximize the first 20 days of SNF, which are covered fully by Medicare, a decision primarily driven by insurance coverage. Although this seemed to benefit patients by allowing them to have more covered days at the SNF, the unnecessary stay at the hospital was costly to the healthcare system and potentially dangerous for patients (Table 4, Quote 3F).

Discussion

Participants discussed how cost of care and insurance coverage play a vital role in PAC decision-making. This ranged from patient and clinician lack of knowledge leading to uninformed recommendations and decisions, to constraints of current processes that must be worked around or negotiated. This was often done through clinicians documenting failure or changing recommendations to PAC that would be covered by insurance in some way, even if not the recommended level of care. These findings are important in the context of value-based payment reforms which alter the incentives behind these decisions.

Prior studies explored the rising cost of PAC and its impact on overall health care spending, quality of care and outcomes.²³ Few studies have explored what hospitalized patients would like to know to make informed decision about their PAC placement, including information on the upfront financial cost of services as well as insurance coverage.²⁴ Patient-physician communication on out-of-pocket costs may occur infrequently, leading some to recommend physician communication with patients about out-of-pocket costs as an important but neglected aspect of clinical practice.²⁵ To our knowledge this is the first research study that explored cost as it relates to clinician, patient and caregiver decision-making about SNFs.

At the macro level, issues with cost and quality in PAC are well-described, but evidence is mounting that the micro level is equally important in determining how PAC is used and how reforms might affect individual choices. Studies demonstrated the key role patients, caregivers, and clinicians play in PAC decision-making on a case-by-case basis. ^10,11,13 It seems plausible that reforms to control PAC costs at the macro level may have substantial effects on these individual case-by-case “micro” decisions, potentially creating maladaptive changes where patients are discharged quicker and sicker, resources are rationed, and patients, caregivers and mounting concerns about the right level of care. These patterns may perversely contribute to increased re-hospitalizations from PACs ²⁶ and a continued increase in healthcare spending.²⁶ Although participants explicitly described consequences of insurance constraints on patients, caregivers and clinicians, perhaps the most consequential impact of these constraints is in patient safety issues that arise from delayed care, altered medical records to satisfy insurance requirements and risks associated with allowing patients to “fail” to show the need for SNF. These extreme workarounds put patients in danger and clinicians in a professional dilemma due to moral distress.

Our results suggest both the macro and micro level may need to be targeted to meaningfully address costs without creating the potential for harm. Macro-level payment reforms increasingly hold both hospitals and SNFs accountable for costs and patient outcomes ^27–30, but it is unknown to what extent these translate to individual decision-making, and how critical this may be to whether these reforms meaningfully affect costs and outcomes. Our results suggest patients and providers perceive cost and payment constraints as at best nuisances and at worst barriers to be “worked around” in achieving what are perceived to be optimal PAC for patients. Obviously, “working around” macro-level reforms may limit their efficacy. Pairing these reforms with reforms at the micro-level that make patients and clinicians more knowledgeable about these constraints and allow more choice flexibility within them may help achieve their intended effect – reducing costs and improving outcomes.

For example, based on our findings, several micro-level reforms may address the concerns of our participants. First, patients, caregivers and clinicians need on-demand tools to inform cost and insurance coverage for post-acute care options at the time of hospital discharge. This would reduce/avoid confusion and lack of knowledge discussed as a barrier. This could be done using standardized shared decision-making aids that could help patients, caregivers, and clinicians make high quality PAC placement decisions. Second, policy level interventions should allow for more flexibility in creating a PAC “packages” to dissociate coverage from location, which will match individual patient needs better than all-size fits all approach practiced. Third, for patients who do not have SNF coverage, onetime coverage of SNF stay paid by Medicaid could potentially avoid long-term nursing home placement and increased cost to the patient and the system. Fourth, clinicians need avenues to document the need for PAC and instances where patients in need did not receive the care due to various reasons such as eligibility rules and insurance barriers. Tracking these barriers to optimal care and exploring its relationships to patient outcomes could inform macro-level reform.

Although rigorous qualitative methods were applied throughout the study, this study has limitations. Our study was conducted in an urban area; thus, findings might not be generalizable to rural or frontier settings. While we attempted to gather a diversity of perspectives, our interviews were limited to six sites (3 hospitals, 3 SNFs) and findings may vary in other markets. Additionally, interviewers had no way of verifying what was discussed by participants as covered versus what was actually covered by insurance, which could lead to some discrepancies. However, there was consensus among participants across roles and sites in what was discussed, providing evidence of consistency in reporting insurance coverage. A strength of the study was the diversity of perspectives within each site, including more than 100 participants from SNFs and hospitals, capturing a wide array of perspectives about how cost of care plays into patient SNF placement decision making. Additionally, we had an experienced, multi-disciplinary research team that conducted, analyzed, and interpreted these research results.

These results discuss pressures experienced by clinicians, patients, and caregivers, and suggest the need for targeted education about insurance policies, team-based decision making for patient placement, and incorporating bedside experiences in the changing policy environment. To avoid unwanted consequences of insurance constraints on patient safety and clinician integrity, the new emphasis on value-based purchasing for PAC offers both opportunity and caution about how to deliver the highest-value PAC.