Choosing the Best and Scrambling for the Rest: Hospital–Nursing Home Relationships and Admissions to Post-Acute Care

Abstract

Objective:

We explored post–Affordable Care Act hospital and skilled nursing facility (SNF) perspectives in discharge and admission practices.

Method:

Interviews were conducted with 138 administrative personnel in 16 hospitals and 25 SNFs in eight U.S. markets and qualitatively analyzed.

Results:

Hospitals may use prior referral rates and patients’ geographic proximity to SNFs to guide discharges. SNFs with higher hospital referral rates often use licensed nurses to screen patients to admit more preferred patients. While SNFs with lower hospital referral rates use marketing strategies to increase admissions, these patients are often less preferred due to lower reimbursement or complex care needs.

Conclusion:

An unintended consequence of increased hospital-SNF integration may be greater disparity. SNFs with high hospital referral rates may admit well-reimbursed or less medically complex patients than SNFs with lower referral rates. Without policy remediation, SNFs with lower referral rates may thus care for more medically complex long-term care patients.

Introduction

Since the Affordable Care Act (ACA) went into effect in 2010, hospitals and skilled nursing facilities (SNFs) have attempted to enhance care transitions between hospitals and post-acute care (PAC) facilities and reduce the rate of 30-day hospital readmissions. Bundled payments, the existence of the Nursing Home Compare report card, the growth of accountable care organizations (ACOs), and Medicare-managed care offer potential paths to better PAC placement and reduced re-hospitalizations (Mor & Besdine, 2011; Mukamel, Ladd, Weimer, Spector, & Zinn, 2009). In addition, more patients receive care at home following a hospitalization (Payne, DiGiuseppe, & Tilahun, 2002).

Many of the new capitated and value-based payment models have led to increased integration of hospitals and PAC. The type and extent of integration between and among hospitals and associated SNFs vary throughout the country (Rahman, Foster, Grabowski, Zinn, & Mor, 2013). Research has shown that tighter integration of hospitals and PAC helps to improve care for older patients through the articulation of shared care goals and better communication and coordination among care providers (King et al., 2013; Lage, Rusinak, Carr, Grabowski, & Ackerly, 2015; Rahman, Foster, et al., 2013; Maly et al., 2012). For example, efforts to enhance communication and coordination among facilities help reduce errors in care transitions, such as through the INTERACT Tool (Ouslander & Berenson, 2011).

With the advent of the ACA legislation and its concomitant developments in the organization of health care, the impact of these changes on care quality is only beginning to be explored, and how patients and their families are affected also remains unclear (Pesis-Katz, Phelps, Temkin-Greener, Spector, Veazie, & Mukamel, 2013). While data accumulate to help track population trends and outcomes associated with these developments, it is important to know how individuals in administrative positions within both hospitals and SNFs are responding locally to these changes. With greater integration overall, how are hospitals deciding where to refer patients for PAC? How are PAC facilities positioning themselves to optimize patient mix, reimbursement, and referral rates from hospitals? Given evidence that nursing homes operate in a tiered system along racial and socioeconomic lines (Li, Cai, & Cram, 2011; Mor, Zinn, Angelelli, Teno, & Miller, 2004; Rahman, Grabowski, Intrator, Cai, & Mor, 2013), what are the future implications of increased organizational integration for patients, especially for those who are poorer and have mental health and other complex care needs?

In this study, we visited health care markets throughout the country and interviewed individuals in administrative positions in hospitals and SNFs to ascertain their views on the evolving market dynamics and determine their responses to the multiple new pressures of the post-ACA world that impact them and affect their decisions related to SNF admissions. We know of no other current studies that attempt to gauge how individual hospital and SNF administrators strategize to manage the transfer of older patients from hospitals to SNF. Using a qualitative interview strategy in varied markets throughout the United States, we explored how administrators on both sides of the hospital–SNF transition process position themselves on behalf of their institutions to deliver the best care they can to older Americans given the rapidly changing policy and legislative environment in which they work.

Method

We selected eight cities throughout the United States for site visits. This multiple case study method was used to highlight differences and similarities by recognizing the uniqueness of each site and identifying trends across sites (Padgett, 2012; Yin, 2012, 2014). Variation was sought among the sites. Criteria for selection related to the prevalence of Medicare Advantage programs, geographic region of the country, hospital-PAC referral patterns and re-hospitalization rates. We targeted institutions in the Northeast, South, West, and Midwest regions of the country. Within each market, we selected one hospital with a low re-hospitalization rate and one with a higher re-hospitalization rate. We used Medicare fee-for-service claims data from 2013 to calculate these rates. We also used the 2013 Medicare claims data to determine the SNFs in each market to which each hospital was sending a high proportion of patients. We applied the Herfindahl-Hirschman index (HHI)index to the volume of hospital referrals to SNFs to indicate the strength of their relationships (i.e., a high HHI represented stronger relationships between the SNF and its referring hospitals). Within each of the hospitals and SNFs, we spoke to administrators with extensive knowledge of their facilities’ practices concerning discharge planning, institutional strategy, and daily operations. We interviewed discharge planners, business development executives, chief medical officers and hospitalists at hospitals, and directors of nurses, administrators, and admissions coordinators at for-profit and not-for-profit SNFs. A total of 138 interviews were conducted in these sites. Approximately 46% or 64 interviews were of hospital personnel, including 21 physicians (33%), 23 vice presidents/administrators (36%), and 20 discharge planners (31%). SNF interviews comprised 54% of the total interviews (74 individuals), including 24 administrators (32.4%), 24 directors of nurses (32.4%), and 26 admissions coordinators (35.1%). Occasionally, an individual in a particular role was not present for an in-person interview during the site visit and a telephone interview subsequently took place.

We developed separate interview guides for our hospital respondents depending on their role and asked questions in the following areas: (a) referral sites and discharge processes, (b) processes for sharing data and improving transitions, (c) efforts to reduce re-hospitalizations, and (d) hospital-PAC network development. Our interview questions for SNF personnel focused on (a) which hospitals referred to them, (b) types of patients whom they accepted/specialized in, (c) patterns of interaction with hospitals, (d) strategies for generating hospital referrals, and (e) programs for improved transitions and reduced re-hospitalizations. We piloted and tested our interview questions with individuals in the corresponding roles in a major hospital system and two nursing facilities in a large city in the Northeast and refined our protocols using feedback from these interviews. A sample of interview questions is listed in Table 1. We obtained participants’ permission prior to recording each interview. IRB permission was granted by our university.

Table 1.

Sample Interview Questions.

For hospital discharge planner
What is your rationale for referring to certain SNFs or other providers?
How have these preferred providers changed over time?
What choices do you have in terms of post-acute care settings?
How might bed availability influence your decision?
Do you have access to post-acute provider quality data? If so, what type of data? Does this include data on readmission rates? Who uses these data? How?
Please describe any programs to improve transitions to PAC.
How do you exchange information with SNFs or other providers? Have you modified the information provided on your documentation to facilitate discharge to SNFs?
Do you know any SNF admissions coordinators? If so, how often do you speak with them? Do they see potential patients here in the hospital?
Does the SNF contact the hospital prior to a potential readmission?
For SNF admissions coordinator
We are interested in the type of patients admitted to your facility, and by type we mean diagnoses. Is there a particular type of case that you specialize in?
Which hospitals refer to your facility? What kinds of patients does the hospital send you?
Do you have any contracts with managed care companies for post-acute care services?
Please describe the typical admission process.
Who communicates with whom? What kind of information do you receive?
Do you see patients in the hospital?
Does your facility provide any information to referring facilities or managed care organizations to help patients in deciding to use your facility?
How important would you say the location of the provider usually is to patients and families?
What else is usually important to patients or their families?
When you are seeking additional referrals or “preferred provider” status from a hospital, do you show them data on quality metrics and/or readmission rates?

Note. SNF = skilled nursing facility; PAC = post-acute care.

Interviews averaged 40 minutes in length and were transcribed for analysis. Our team embarked on a rigorous analysis process to code the transcripts and identify emerging patterns or themes across transcripts (Crabtree & Miller, 1999; Miles, Huberman, & Saldana, 2014; Padgett, 2012; Weston et al., 2001). Codes were applied to segments of the transcripts to group similar kinds of text together. Codes in the preliminary coding scheme were based on the questions we asked in our interview guides. For example, as we asked about processes used in admitting patients from hospitals, codes such as “patient choice/options” and “patient/family preferences” were among those we applied to text when those subjects arose. We then modified and refined the scheme in an iterative fashion to add codes and refine code definitions. Additional codes resulted when material emerged from interviews that were unanticipated or different from the specific questions that we asked. The resulting coding scheme reflected both the a priori codes and areas of interest from the interview questions as well as the unanticipated findings. See Table 2 for a sample of codes related to hospital discharge and SNF admission.

Table 2.

Partial Coding Scheme: Codes Pertinent to Hospital Discharge and SNF Admission.

Codes for discharge planning/SNF admission/transitions
A.	Patient choice/options (including any difference in this by diagnosis or other)
B.	Patient/family preferences
C.	Patient/family education/information (about what to expect at SNF, what insurance covers, etc.)
D.	Patient clinical characteristics/specialty care needs (including assessment of)
E.	Role/function of PAC network
F.	Influence of PAC providers (e.g., have power to pick and choose)
G.	Influence of insurance company
H.	Influence of MD/PCP
I.	Influence of discharge planner/social worker/care manager
J.	Communication between hospital and SNF (how/what information is transferred)
K.	Volume of admissions
L.	Change in discharges/admissions over time (types of patients or how accomplished)
M.	Type of patients admitted or avoided
N.	Liaison with potential patients
O.	Quid pro quo arrangements
P.	Marketing
Q.	Programs to improve transitions
R.	Effect of programs to improve transitions
S.	Other

Note. SNF = skilled nursing facility; PAC = post-acute care; MD = medical doctor; PCP = primary care provider.

The entire research team consisted of five members. Five read all interviews from the first two sites and four individually coded each transcript. In team meetings that followed, the four coders discussed and then refined the coding scheme and code definitions according to how well they continued to fit the data in the transcripts; reviewed preliminary patterns we perceived in the data; and reconciled our interpretations of the first coded transcripts that each team member had prepared. Once we completed analysis of the interviews from the initial two sites (n = 48), we streamlined the process by dividing into two subteams of two members each. Each team member in the subteams coded the transcripts independently, then met to reconcile final coding. Each subteam also discussed potential themes after coding of each site’s transcripts was completed. Members rotated in the subteams to enhance validity and ensure that criteria for team decisions remained consistent among members. The full team of five convened regularly to discuss emerging themes from each site, track the prevalence of proposed themes across transcripts, and search for disconfirming evidence to explain discrepant information and better understand the range of responses. Throughout analysis, we kept a comprehensive audit trail that recorded ongoing team decisions, including selection and definitions of codes and the discussion of emerging themes (Curry & Nunez-Smith, 2015; Holloway & Wheeler, 1996; Lincoln & Guba, 1985; Miles et al., 2014; Ritchie & Lewis, 2012).

Results

We interviewed 138 respondents at 41 facilities throughout the country. Basic characteristics of the facilities are shown in Table 3. Of the 16 total hospitals, one was a for-profit hospital, 13 were not-for-profit, and two were owned by the state or local government. They ranged in size from 320 to 990 beds. Of the 25 SNFs in the sample, 17 were for-profit, with a bed size that ranged from 20 to 200. All but one accepted both Medicare and Medicaid patients. There was a full range of 1–5 Centers for Medicare & Medicaid Services (CMS) star ratings across the 25 sites with over half of the participating SNFs having a rating of 4 or 5 stars. Interview data indicate that four hospitals had established preferred provider networks with local SNFs based on where they had historically discharged patients in the past (McHugh et al., 2017).

Table 3.

Participating Hospital and SNF Characteristics.

Site no., hospital or SNF	Hospital and SNF characteristics
	No. of bedsa	Profit/ownership	SNF overall quality star rating
Site 1, Hospital 1	500	Non-profit	NA
Site 1, Hospital 2	320	State/local government	NA
Site 2, Hospital 1	420	For-profit	NA
Site 2, Hospital 2	480	Non-profit	NA
Site 3, Hospital 1	410	Non-profit	NA
Site 3, Hospital 2	340	Non-profit	NA
Site 4, Hospital 1	630	Non-profit	NA
Site 4, Hospital 2	680	Non-profit	NA
Site 5, Hospital 1	430	Non-profit	NA
Site 5, Hospital 2	420	Non-profit	NA
Site 6, Hospital 1	990	Non-profit	NA
Site 6, Hospital 2	350	Non-profit	NA
Site 7, Hospital 1	450	Non-profit	NA
Site 7, Hospital 2	460	State/local government	NA
Site 8, Hospital 1	360	Non-profit	NA
Site 8, Hospital 2	940	Non-profit	NA
Site 1, SNF 1	90	For-profit-corporation	*****
Site 1, SNF 2	310	For-profit-corporation	***
Site 1, SNF 3	70	For-profit-corporation	***
Site 2, SNF 1	240	Non-profit-corporation	****
Site 2, SNF 2	120	For-profit-corporation	**
Site 2, SNF 3	120	For-profit-corporation	****
Site 3, SNF 1	210	Non-profit-corporation	****
Site 3, SNF 2	130	Non-profit-corporation	***
Site 3, SNF 3	110	Non-profit-corporation	****
Site 4, SNF 1	190	For-profit-corporation	*****
Site 4, SNF 2	20	Non-profit-church	*****
Site 4, SNF 3	120	For-profit-partnership	*****
Site 5, SNF 1	100	For-profit-corporation	***
Site 5, SNF 2	50	Non-profit-corporation	****
Site 5, SNF 3	180	For-profit-corporation	***
Site 5, SNF 4	50	For-profit-corporation	****
Site 6, SNF 1	80	Non-profit-corporation	***
Site 6, SNF 2	230	For-profit-corporation	*****
Site 6, SNF 3	160	Non-profit-church	****
Site 7, SNF 1	150	For-profit-corporation	*
Site 7, SNF 2	130	For-profit-corporation	***
Site 8, SNF 1	120	For-profit-partnership	***
Site 8, SNF 2	30	For-profit-corporation	****
Site 8, SNF 3	200	For-profit-corporation	*****
Site 8, SNF 4	140	For-profit-corporation	***

Note. SNF = skilled nursing facility; NA = not applicable.

^aRounded to further protect anonymity.

^*= CMS Star Rating (1=5 stars)

The three themes that are presented below arose from the analysis of our qualitative data. Together they help illuminate patterns about how hospitals and SNFs are changing their interactions and approaches to discharging and admitting older patients from hospitals to SNFs.

Theme 1: Hospitals face pressures to reduce lengths of stay and readmissions for older patients discharged to SNF, and they attempt to create networks with preferred SNFs.

Pressures to decrease hospital length of stay (LOS) result in a tighter time frame for discharging older patients and in some cases have led to hospitals developing a method to fine-tune referral preferences. As urgency to discharge increases, hospital discharge planners told us it is helpful to have certain reliable PAC facilities as discharge sites. In choosing desirable SNFs, for example, hospitals seem to consider their past referral relationships with area SNFs as care routines, prior performance and reputation, tendency to re-hospitalize, and facility personnel are known. Another consideration hospital personnel noted is the physical proximity of the SNF to the patient’s or family member’s home, especially because this factor is deemed important to patients and their families (Gadbois, Tyler, & Mor, 2017; Tyler et al., 2017; Tyler, McHugh, Shield, Winblad, & Gadbois, 2016). While hospitals recognize the priority of patient choice in facility selection (Baier et al., 2015), their historic relationships and prior referral rates with SNFs can help in guiding discharge discussions to those facilities considered feasible. One hospital administrator described her hospital’s method to establish SNF partners:

There’s overcapacity in this market and so we … said, “Well, we don’t need all of you” [and] in a preferred partner process we created a competitive process; every quarter those post-acute providers push out data to us on our key metrics that we identify.

(Site 4, Hospital 2, Administrator)

With attention on preventing re-hospitalizations, hospital physicians’ experiences with SNFs may also influence how hospitals guide patients. One hospital administrator noted,

[Physicians] are very tuned into when their patients are readmitted from SNFs … and they can say, “No … [these SNFs] just readmit patients for everything.” So we don’t want to send patients there.

(Site 7, Hospital 1, Vice President of Strategy)

The pressure to discharge patients quickly creates challenges in finding appropriate and willing SNFs while also ensuring their ongoing relationships with SNFs remain viable. Hospitals keep their options open for future placements, as one hospital discharge planner said:

If I have a … high risk [patient], and I have no takers, none … I’ve got to work with … some of our not-so-desirable [SNFs], because they’re going to take some of our highest risk patients, and we have to help them.

(Site 2, Hospital 2, Discharge Planner)

As hospitals balance SNFs’ capacities against the necessity of discharging the patient as soon as possible, they may use a quid pro quo strategy with SNFs. A SNF admissions coordinator related,

Occasionally we get a social worker begging us to take a complex patient because they are having a hard time [placing the person]. And we will do our best to try to help them because it’s a two-way street, you know?

(Site 3, SNF 1, Admissions Coordinator)

Medically complex patients pose particular difficulties for hospital discharge and SNF admission. While such patients “may be more complex than [SNFs] want” (Site 3, Hospital 1, Director of Care Management) and can become long-term care patients with a lower rate of Medicaid reimbursement, such admissions still maintain the SNFs’ volume of referrals. A SNF administrator said,

I am fairly blunt with the [hospital’s] head social worker that when we get one that’s a mess, a train wreck, we’re going to take it because they’re our referral source.

(Site 4, SNF 1, Administrator)

Another SNF administrator said,

They’re like, “Well, please just take this patient” … even though you know it’s going to be, as we call them, a train wreck … [and will] end up long-term care … you’re hoping that case manager is going to hold to their word and give you some good … ortho patients … that you want.

(Site 2, SNF 1, Director of Nurses)

SNFs may refuse the prospective admission. “The more complex patient, it can be harder to find a bed for them,” said one hospital respondent (Site 3, Hospital 1, Manager of Care Integration). Another hospital discharge planner acknowledged the desire by SNFs to admit patients they consider desirable, such as those who need short-term and Medicare-reimbursed orthopedic rehabilitation and reject those considered less preferred:

A 78-year-old hip: [SNFs] are waiting in line. I will have no problem … She likely has Medicare, she’ll pay well and they are all … saying, “Me, me, me! I want her!” Unless there’s something like, you know, a really mean family member or something like that.

(Site 1, Hospital 2, Discharge Planner)

A documented behavioral health history can result in “facilities that will turn them away. Some facilities take the best of the best … and then there are those facilities that take the worst of the worst … to fill their beds” (Site 2, Hospital 1, Inpatient Case Manager).

Theme 2: SNFs noted how they are stressed by hurried hospital discharges and other pressures. SNF respondents described an underlying anxiety about the faster pace of admissions, and they related their efforts to accommodate to the changes.

In the increasingly rapid discharge process, a SNF administrator of a 3-star rated facility noted, “We’re on this fast moving train and we don’t know where it’s going … there’s a lot of uncertainty” (Site 8, SNF 1, Administrator). This respondent emphasized the increasing rate of change:

We’re all trying to adapt … the environment it’s changing … if I were to leave this industry today and come back in a year, I’d be lost … It’s really crazy … nobody really knows what what’s going to happen … It’s consolidating more and more and … SNFs are being cut out … the hospitals are going after the SNFs that have the programs that are going to benefit them and ultimately benefit the patients.

(Site 8, SNF 1, Administrator)

A hospital administrator also commented on rapid change in the system:

Been a lot [of changes] … There’s been, yeah, a lot of buying and selling of nursing homes … I think it’s difficult to, it’s very difficult … to make … to be profitable in a nursing home … So it’s really tough.

(Site 6, Hospital 1, Vice President of Strategy)

While hospitals work to place their discharged patients in facilities that they hope will not result in a 30-day re-hospitalization, SNFs told us they often suffer from insufficient information about the prospective admitted patient who may be sicker, more complex, more costly to care for, and more likely to stay for long-term care. Though they desire patients with straightforward and short-term post-operative orthopedic rehabilitation needs, they still benefit from continued hospital referrals and agreement to accept a mix of patients considered both preferable and well-reimbursed versus those not as well-reimbursed.

In this process, it is necessary but sometimes difficult to obtain information about potential patients who are considered complex or challenging. A SNF director of nurses explained,

[Hospitals] don’t want to tell us because they’re having trouble placing them … We had one yesterday that … had someone sitting with them 24 hours a day. We can’t provide that [amount of care and oversight] here. But verbally the referral source didn’t tell us that.

(Site 3, SNF 3, Director of Nurses)

A SNF administrator noted that his facility was “not equipped” for a psychiatric patient, but “last week we had one and the referral had been a little scrubbed clean” (Site 4, SNF 1, Administrator). The speed required for making an admissions decision is weighed against the possibility of a good referral. A SNF director of nurses said,

Sometimes my social worker’ll call me, say, “I have a referral. It looks really good. It’s a hip. No behaviors,” and I go, “Just accept ‘em.” So we’ve gotten a lot more relaxed about just accepting people without really reviewing everything as long as we can kind of tell that there’s really … no behaviors or real complicated things. We want to get back to the discharge planner right away.

(Site 5, SNF 2, Director of Nurses)

Higher levels of managed care and decreased LOS have helped to create intense competition among SNFs as they vie for those patients who are highly reimbursed. This pattern was noted by both hospitals and SNFs governed by the tight timetable for accepting or rejecting potential admissions. A SNF administrator related,

If they’re ready to be discharged that day, they send a referral out. Literally … you’ve got five minutes to get back to us. And if you’re longer than five minutes, two or three people have already said yes.

(Site 1, SNF 3, Admissions Coordinator)

Another said discharges happen, “Probably within an hour … depends on how complicated it is. Some places do it in 30 minutes, so it’s very competitive” (Site 3, SNF 3, Administrator). SNFs indicate an understanding of how future referrals may be jeopardized when they reject referrals of patients seen as less desirable. They want to maintain the relationship and referrals. As one director of nurses stated,

They used to never tell you about problems because they knew that maybe you wouldn’t take ‘em, but because it’s so competitive, you rethink that process often. Like, there’s a lot of patients that I wouldn’t have thought about taking that I’ll take now without any problem, because we just have to stay competitive.

(Site 5, SNF 2, Director of Nurses)

Complaining about how saturated the market was getting in his area, a SNF administrator said, “You know, everybody’s fighting for the same patients … It’s like a business. At the end of the day it’s bloody” (Site 4, SNF 2, Administrator), while another said, “It’s become more competitive because you got to keep your beds filled, you know?” (Site 3, SNF 3, Director of Nurses).

Theme 3: Focuses on how SNF admission strategies appear to be changing

Theme 3a: SNFs that receive a high rate of referrals from hospitals use strategies to maximize admissions they consider preferable.

SNF admission strategies may depend on the strength of their hospital relationships and referral patterns. Staff members at SNFs and hospitals have historical relationships that develop over time and color how they perceive each other’s willingness to accept referrals and each other’s capacities to care for complex patients. Depending on the strength of their relationships, such as the volume of referrals SNFs have with area hospitals, SNFs use different strategies both for screening out potential admissions considered less preferable and casting a wide net to attract as many admissions as possible for financial viability.

As measured by the HHI index, SNFs in our sample that had a higher volume of referrals from a few select hospitals showed some difference in their admissions behavior than SNFs that receive a lower volume of referrals from a higher number of hospitals (the average HHI in such “high” SNFs was 0.47 compared with 0.33 in “low” SNFs). Stronger relationships seemed to be associated with the ability to screen and specialize in patients who had primarily predictable, short-term stays, and this seemed to be the case regardless of SNF ownership status as profit or not-for-profit organizations. These facilities’ admissions coordinators often had a nursing background. They said they telephone the hospital for more information, or send an individual or team to the hospital to do an assessment of a potential admission to make sure the admission is appropriate. Said one,

Our nurse liaison can go visit the patient at the hospital … We like to see everybody in person just to say “Hi,” and also to see if there’re any behaviors ever, if they have a dementia diagnosis.

(Site 4, SNF 1, Admissions Coordinator)

A SNF’s director of nurses explained the importance of obtaining complete information on the patient:

As long as the screening process goes well, you know, then we usually know up front what’s going to happen. I mean, there are times when you get patients that the record says one thing and when they get here, they’re a little bit different than what you were expecting.

(Site 6, SNF 1, Director of Nurses)

Another added,

If I see that they’re receiving some behavior medication, I’ll ask social services to call and find out—are they having behaviors? Have they been using the medication? And I want to find out about the behaviors.

(Site 5, SNF 2, Director of Nurses)

In addition to screening out more complex patients, some SNFs appear to adapt their abilities to care for patients they consider preferable. For example, a 5-star rated facility discussed numerous preferred provider relationships with area organizations and attributed its success to

Our ability to do a quick turnaround with rehab, hips in three weeks, knees in two weeks … we’re also, a preferred provider for the bundled payment with the [hospital] system … what it means for us is we have an increasing flow of orthopedic patients … we have a guaranteed time frame of 14 days, for standard replacements … we’re getting, younger orthopedic patients rather than just the 89 year olds. So we’re getting younger, healthier people with a fast turnaround.

(Site 8, SNF 3, Admissions Coordinator)

Theme 3b: Some lower rated SNFs appear to specialize and create additional services to be able to care for more complex patients.

SNFs that appeared to be less preferred by hospitals for their referrals received patients from a broad array of hospitals and cared for an extremely varied patient mix. Respondents from these less selective SNFs admitted numerous kinds of patients and described marketing strategies to recruit patients from a number of hospitals. One said,

We tend to take everyone … we don’t really pick and choose … if we have an open bed and we can take somebody, we usually will.

(Site 1, SNF 2, Assistant Administrator)

In contrast to SNFs with stronger relationships with hospitals, the admissions coordinators in these facilities frequently had backgrounds in marketing. One described going to hospitals and doctors’ offices to highlight the SNF’s services and programs, noting,

We’re in each hospital every single day meeting with the case managers, letting them know what availability we have on beds, asking them for referrals.

(Site 2, SNF 2, Admissions Coordinator)

A 2-star rated facility that “take[s] everyone” had two marketers who “talk to the hospitals, doctors, everybody and get, like, tips, like where we can go find somebody that needs something as far as … referring” (Site 1, SNF 2, Administrator). Bearing in mind the principle of guaranteeing patient choice, another described how the marketing person attempts to provide information about the SNF and is,

… the community liaison, so his job is to go to the hospitals and see the social workers, provide them with the material and then the social workers know that [place] is a viable option [for placing the patient].

(Site 4, SNF 1, Admissions Coordinator)

Some hospital administrators related how strongly some SNFs advocate for admissions to their facilities. One said,

Some of [the SNFs] have been very aggressive about wanting their name listed. You know, like the discharge planner has like a little cheat shee-, you know … little talking points and … their nursing home would be listed … in those talking points … So they’ve just gotten, some of them very aggressive … you have to say the patient has a choice but that, you know, there are sort of gray areas there, I think.

(Site 6, Hospital 1, Vice President of Strategy)

Another said SNFs frankly “solicit.” “I mean the facilities come up [here]. They bring you cake and food and things like that. They’re trying to build a relationship” (Site 2, Hospital 1, Director of Case Management). However, SNFs are aware of changes in hospital receptivity and how hospitals may push back. One SNF director of nurses said,

We don’t actually go out to the hospitals … they don’t want us … going out and marketing … that used be a big thing that we used to do … [But hospitals] are … just too busy … you know, for us to go marketing … So we market via phone, like when our census gets low … [we call them and] say, “Hey, we have beds available” … And then we get referrals.

(Site 5, SNF 2, Director of Nurses)

Some lower rated SNFs also appear to create additional and specialized services to care for more complex patients. Perhaps in response to higher rated SNFs admitting more patients with orthopedic and less complex needs, other SNFs may be adapting by accepting patients who require specialized services for behavioral and other complicated needs. While hospitals take steps to find SNFs with good reputations and low re-hospitalization rates, less highly rated SNFs admit numerous kinds of patients from a broad range of hospitals. For example, a 2-star rated facility provided numerous services and drew patients from all area hospitals:

The only service that we do not provide is ventilators … we do trachs … feeding tubes … in-house dialysis, we have a respiratory program … And we also do piped in oxygen, which a lot of facilities don’t have.

(Site 2, SNF 2, Director of Nurses)

A 1-star rated facility noted admitting,

A lot of patients that most long-term care facilities wouldn’t take because of the high acuity … TPN [total parenteral nutrition] … wound VACs [Vacuum Assisted Closure] … reconstructive skin flaps for wounds … patients [with a cancer diagnosis] … receiving chemotherapy and radiation treatments off-site …

(Site 7, SNF 1, Administrator)

As SNFs that are preferred by hospitals specialize to focus on bundles and produce short LOS, SNFs that have lower referral rates from hospitals appear to accept a broad array of patients from a large geographic area as a response to get by in increasingly competitive times.

Discussion

How hospitals and SNFs interact has undergone numerous changes in recent years. Our interviews indicated the greater stress that seems to be experienced on the part of both hospitals and SNFs due to decreased hospital lengths of stay and the greater need to place discharged patients in SNFs as rapidly as possible. This trend has meant that hospitals often rely on trusted historic SNF relationships that may provide some assurance that their patients will be well cared for and with a reduced likelihood of being re-hospitalized. As hospitals are now penalized for 30-day readmission for key diagnoses, the incentive to place patients in such trusted facilities is greater than in previous years.

Our interviews indicate that SNFs with a plethora of referrals of more highly reimbursed Medicare patients with few if no complicating conditions may screen out less desirable, more complex patients. Meanwhile, SNFs with lower quality ratings or those with weaker ties to referring hospitals have seen their rate of referrals go down, and seem to use marketers to elicit referrals and keep their beds filled for continued financial viability. Hospitals, for their part, may continue to refer to less highly rated SNFs when they fill a geographic or complex care need, or when other more highly rated SNFs will not accept patients with dementia or behavioral disturbance or those receiving Medicaid reimbursement. As these trends continue, more facilities may become selective about their admitted patients such that patients with lower reimbursement, mental illness, or other complex care requirements may have to choose among SNFs with lower quality ratings.

The perspectives presented here seem to paint a picture that adds to a dismaying trend in SNFs of unequal access to care and increased disparities (Li et al., 2011; Mor et al., 2004; Rahman et al., 2010). Research has demonstrated, for over a decade, how less advantaged patients who rely on Medicaid reimbursement tend to be housed in SNFs located in socioeconomically poorer regions of the country with decreased quality and greater rates of deficiencies (Mor et al., 2004).

The experiences of hospital discharge planners and SNF admissions coordinators whom we interviewed show how the pace of transfers resulting from decreased lengths of hospital stay and increased competition among SNFs may intensify the pressures experienced by both hospitals and SNFs. In the rush to discharge patients to SNFs that have a lower rate of 30-day re-hospitalization, the patients who are less likely to become a long-term care Medicaid patient and qualify for Medicare reimbursement seem to be at a premium and have an advantage. These patients generally have shorter, more predictable and less complicated paths to discharge and are consequently desired by SNFs. Patients without adequate financial resources or with complicating co-morbidities, chronic illnesses, or behavioral disorders are harder to place and may have to choose among SNFs with lower quality ratings that will accept them because they need to fill beds (Li et al., 2011). While not covered in this study, an additional trend that contributes to the competition among SNFs is that greater numbers of hospitalized older patients with orthopedic procedures are discharged home with home health services.

Also apparent in how hospital discharge planners and SNF admissions coordinators report this hectic pace of transfer to SNF is the ever-increasing difficulty in obtaining accurate information about the patients who need SNF care (Brennan et al., 2002). We suggest that policy initiatives recognize the potential of increased disparities in care from these pressures and work to remedy this situation; the troubling trends noted here are likely to worsen with the increasing numbers of older adults in our nation who will need care.

Limitations

While qualitative studies such as this one cannot be considered representative or generalizable given the limits of their relatively small size, our interviews with 138 administrative staff in hospitals and SNFs is considered robust by qualitative standards. However, facilities that agreed to participate in interviews may be different in important ways from those that did not participate. Our selection of SNFs may not be representative because we chose them based on hospital referral patterns and not based on measures of quality. Our selection of sites aimed to sample varied facilities from the Northeast, South, West, and Midwest of the United States and display a range of structural features such as, ownership, size, location, and referral patterns, provide a breadth to the sample though SNFs most often had star ratings of 3 or more. In addition, our approach to interview analysis and reconciliation was rigorous from the coding process to the identification of potential themes and the consideration of competing interpretations (Crabtree & Miller, 1999; Curry & Nunez-Smith, 2015; Weston et al., 2001). We also limited our interview questions to specifically focus on transitions from hospitals to SNFs; understanding how patients move from hospitals to the home setting was beyond the scope of this research.

Conclusion

The pressure to discharge hospital patients quickly creates pressure on both hospitals and SNFs. As hospitals seek to place patients in reliable SNFs that have a history of lower re-hospitalization rates, they also refer patients with more medical complexity who will yield a lower reimbursement rate to SNFs with lower quality ratings so the patient is placed somewhere. Meanwhile, SNFs balance their preference for patients with relatively simple and highly reimbursed rehabilitation needs with keeping referring hospitals happy and amenable to future discharge placements.

While medically complex patients receive Medicare reimbursement in the post-acute phase, they still require a high level of nursing care resources in the SNF. More SNFs are shifting a greater proportion of their beds from long-stay to short-stay. Patients with dementia, eventual long-term care needs, and multiple and complex co-morbidities may be increasingly perceived as less desirable as their Medicare reimbursement is limited, they are likely to transition to Medicaid, and they are consequently more likely to prevent a short-term rehabilitation patient from occupying the same bed. As competition intensifies to maximize selection of short-stay patients with Medicare, admit those with private pay insurance, and avoid those considered less desirable because their reimbursement rates are lower, patients’ equal access to excellent SNF care is increasingly jeopardized. Demographic trends guaranteeing greater care complexity with increased longevity only seem to worsen the outlook for these patients. Policy experts should take the words of our respondents and their sobering assessments into account to improve access to excellent care for all patients. As our population grows rapidly older, we must address the increasingly troubling consequences of these market forces and create practical and equitable solutions to them.