Public quality reporting and nursing home admission decisions

Brief Summary:

Nursing homes are facility-centric and strategic in their admission processes. Admission decisions are influenced by sensitivity to publically reported quality scores, and also by a potential resident’s staffing requirements.

It’s been twenty years since CMS launched Nursing Home Compare, now Nursing Home Care Compare (NHCC), a web-based consumer resource outlining care quality in nursing homes (NH). Quality metrics are summarized in scorecards grading each facility using a five-star rating. When the National Quality Forum developed the long-stay quality-measures, they purposefully incorporated only minimal adjustments for resident risk with few, well-defined exclusions.¹ A disadvantage of using unadjusted quality scores is the tendency to underestimate the care quality provided to less healthy patients. Furthermore, an incentive is created for facilities to artificially improve their scores by selecting healthier patients and avoiding frailer individuals.^2,3 Numerous metaphors describe this routine consequence of quality initiatives leveraging public reporting, including cherry-picking - selecting for easy, straightforward admissions likely to improve scores - and lemon-dropping - discharging difficult residents (usually to the hospital) who negatively impact quality-measures.^4,5 Some estimates attribute as much as half of NHCC quality-measure variation to unadjusted risk.^2,6 The unfair outcome of biased selection is worsening disparity as more desirable facilities recruit more desirable patients, thereby increasing the appearance of quality and thus their desirability.⁷

While the impact of public reporting on NH admission decisions remains hypothetical, disparities in NH quality and access have indisputably increased since NHCC’s debut, though explanations remain imperfectly understood.^7,8 As the current administration expands and enhances NHCC, increasing performance driven incentives, greater knowledge surrounding impacts of public reporting on admission processes and resultant outcomes is critical to avoid worsening NH disparities.⁹

To better understand factors influencing NH admission decisions, in 2020, we interviewed NH staff responsible for evaluating potential admissions regarding their resident selection processes. Before the COVID pandemic terminated recruitment, 14 participants from three urban and one rural facility in Virginia were enrolled. Participating facilities had enhanced staffing, were overwhelmingly white demographically, and in the highest quintile for both social determinants of health and antipsychotic use. Four zip-code matched facilities with low antipsychotic use declined participation.

Interviewees held multiple NH positions and included social workers, nurse managers, and administrators, the common thread being involvement in the NH’s decision to admit or retain residents. Interviews focused on factors/processes impacting admission decisions, including NHCC quality-measures, particularly antipsychotic use. Semi-structured, template-guided telephone interviews conducted by experienced NH clinicians were transcribed, maximizing anonymity of the NH and participant. Coding began with an inductive and deductive process to develop categories for the admission factor being addressed, yielding a-priori codes subsequently augmented to include emergent findings. Transcripts were coded independently by three investigators and consistent concepts and patterns began to emerge. Finally, through immersion and crystallization processes involving all team members, themes were delineated.

Three themes regarding NH admission and retention decisions emerged where respondents were in general agreement. Facilities are strategic in choosing the residents to admit. Though multiple interrelated factors contribute, admission decision-makers described purposeful processes for selecting the most ideal residents possible while still keeping the facility full. Goals were facility-centric. What was best for local populations or individual patients was not described. There was consensus that certain facilities have greater opportunity for selectivity while others are obliged to take residents no other NH wants. NHCC quality-measures influence admission decisions. While ambivalence regarding the efficacy of quality-measures to capture quality was universal, all participants endorsed scorecard awareness and described sincere facility-wide effort to optimize scores. No decision-maker reported difficulty forecasting the impact of individual admission decisions on quality-measure scores. All interviewees reported cherry-picking admissions. Most described lemon-dropping problematic residents. Staffing requirements are critical considerations. While acknowledging nuanced interplay between the competing elements influencing decisions, all participants emphasized the critical importance of an admission’s staffing needs. Residents with few staffing needs were desirable; residents requiring challenging, time consuming care were avoided.

Though limited in scope, these results achieved data saturation in analysis and begin to illuminate NH admission and retention processes, a subject rich in conjecture but lacking data. It is known that publically reported quality initiatives lead to cherry-picking, that cherry-picking increases disparities, and that disparities have increased since NHCC’s scorecard debuted.^4,5,7,8,10 Our results suggest these phenomenon are linked, that quality reporting has created new selection incentives resulting in increased sorting and greater disparity. Though NHCC scorecard considerations were not the sole or even primary catalysts in strategic admission processes, they did impact decisions. Staffing concerns affected all decision areas. This finding should raise concern given the pandemic has witnessed a simultaneous surge in resident care requirements and a contraction in staffing availability, a trend potentially incentivizing even greater admission sorting and intensifying disparities.

With quality-based compensation set to increase and few measures in place to curb facility-centric selection processes, it is likely that new or modified admission policies will be necessary to mitigate disparities.⁹ Some states modify compensation for risk, creating financial incentives to take sicker and more difficult patients. Alternatively, NHCC’s quality-measures could have greater adjustment for risk. There was wisdom in the Forum’s concern for potentially adjusting away true differences in quality.¹ That said, these results support the literature suggesting a meaningful fraction of quality-measure variation is attributable to resident characteristics, and quality improvement should be an ongoing, iterative process.⁶ NHCC is growing, adding quality-measures and increasing penalties for underperformers. Thus, reflection on both successes and improvement opportunities in existing programs may be timely.⁹ If quality reporting impacts admission decisions and admission decisions drive disparities, then facility-centric admission processes require continued research attention and perhaps enhanced oversight to avoid increasing disparities in NH quality and access.

Table 1.

Nursing home admission decision-makers’ quotes, by theme.

Theme	Quotes
1, 2, 3	“We’re fortunate, but there are facilities that have to take those residents who have a lot wrong with them, so there is never a real way for them to even the (NHCC) scores.
1, 2	“I understand why facilities will discharge residents to the hospital and not let them come back. Regulations focus on the process not the people.”
1, 2, 3	“Let’s just say it was somebody with Lewy Body. I know that can be especially difficult at some point for our staff. Well, it’s easy for me to sit back here in an office behind the wall with a computer, and I could admit people all day, but, it’s not my license I would have to worry about…. I’m gonna go to the (locked unit manager) or whoever’s taking care of this person and say, “Would you look at this referral?” You know her license is on the line; Nursing homes license is on the line…. Not every referral needs everybody’s opinion, but a lot of referrals need more than one opinion… as we say: need eyes.”
1, 2, 3	“I would not take a resident with aggressive, violent behaviors in this facility.”
1, 2 3	“Some off ‘em have physical or verbal aggression. They require a lot of one on one care, which we can’t really provide.”
1, 2, 3	“Was there any issues… has this patient been in another facility?… Were they kicked out?… was they any reason they became combative?… you really have to digest each and every referral and each and every little piece out. I look at the drugs.”
1, 2	“We have to be selective taking patients on antipsychotics.”
1, 2	“Some things like Seroquel and things, (um?)… Haldol, you know, things that are on the regulations that could get us in a lot of trouble…the unit manager of our dementia unit – locked unit, you know, she can look at something and say, “Um, I need to run this by Dr. [medical director], ‘cause, you know, they get here, and we’re just gonna have to start reducing the dose.” And the – and she very well may say, ‘Well, let me go talk to the nurse up at [hospital 1] and let them know that once they get here, the dose would have to be reduced to keep us in compliance, and we start that at [hospital 1] and see how they do while they’re there.’”
1, 2, 3	“We’re still getting referrals with a lot of folks with antipsychotics. But the degree of what they’re on, and how often they’re taking, depends on whether we can actually accept them… Like the Zyprexa… of course Haldol is a no-no under the regs - we cannot take anybody that has the scheduled Haldol.”
2	“We don’t use Haldol. Not for any client, ever. We use Seroquel or Risperdal.”
2	“I’m moving towards Depakote over Seroquel, Risperdal.”
3	“But nothing beats one on one (staffing). And unfortunately that just a luxury that I don’t foresee happening anytime soon.”
3	“Well staffing is always an issue… I think like any other facility, having enough people to do what we want to do is always a challenge…”
3	“For us five or six CNAs would be best. More often than not we have three or four.”
3	“… even in a great place like this, some ‘one-on-one,’ providing more staff, more people involved.”

^1.Facilities are strategic in choosing the residents to admit.

^2.Nursing Home Compare quality-measures influence admission decisions.

^3.Staffing requirements are critical considerations.