Many (un)happy returns: Challenges and opportunities for potentially avoidable ED visits after discharge to a skilled nursing facility

Return patient visits after hospital discharge to a skilled nursing facility (SNF) are costly, frequent, and potentially preventable. Among hospitalized Medicare patients discharged to SNF, 23.3% are either readmitted or die within 30 days.¹ In their study, Tian et al. find that among SNF patients sent to the emergency department (ED), 75% are admitted, while 25% are ED same-day or next-day discharges (SDD). They explore the epidemiology of and resource utilization associated with ED SDD, based on the notion that this subset of ED visits may be avoidable.² Though the cost of alternatives is not presented and may not be known, in aggregate, these visits cost Medicare an estimated $24.6 million annually.

In their study, referrals from an SNF to the ED were less likely among female patients and those over 85. Reasons for ED visits were not captured. However, head injuries, urinary tract infection (UTI), and chest pain were the most frequent diagnoses for SDD ED visits, consistent with other studies.² In a multivariable analysis, the authors identified chronic conditions (dementia, congestive heart failure, chronic obstructive pulmonary disease), and facility characteristics (rural SNF setting and those with >70% Medicaid patients) as being associated with an SDD ED visit when compared to patients with no ED visit. Evaluation and monitoring required in certain conditions, such as head injury, may make these SDDs of high value. However, avoiding preventable ED visits for SNF patients (including the burdens of transportation, waiting times, and disruption of postacute care) would be beneficial to this vulnerable population and lessen the burden on already overcrowded EDs.

Preventing avoidable SDDs implies the ability to identify the 25% of such ED visits a priori. Some proposed alternatives might obviate ED visits among post-discharge SNF patients: implementing evidence-based strategies to manage common clinical problems on site with remote decision support; increasing clinical evaluations by mobile providers (physicians or advanced practice providers) or through virtual visits; and increasing real-time, on-site testing and interpretation capabilities (imaging, urinalysis, electocardiogram [EKG]). In addition, certified medical directors (CMDs) could be leveraged to implement solutions to reduce referral.

Each of these has advantages and challenges. Implementing an urgent care model within SNF has shown promise in reducing ED referral but the data are limited.³ Telehealth urgent care within SNF or SNF-based physician specialists could mitigate ED referral, but these have not been studied. Also, literature regarding telehealth overall suggests it leads to higher ED utilization, but in-person evaluation may result in directed diagnostic testing.⁴ Unfortunately, the rates of bundled payments to SNFs do not include on-site testing and urgent physician evaluation services, likely limiting their uptake. Supplemental payments to cover these services could reduce ED referrals, but future research is needed to ascertain the effect of well-defined pilot programs. Assertive implementation of strategies to manage common clinical problems holds promise, such as the Agency for Healthcare Research and Quality falls management program. Though this program was intended to reduce the risk of falls by targeting individual-patient-specific risk factors to directly address, translation into clinical practice has been hampered by high implementation costs, poor compliance, and staffing challenges. In the case of UTIs, conventional practices leading to over-testing⁵ and the absence of clinical expertise at SNFs may contribute to ED referral. While “UTI” may be the ultimate diagnosis, referral to the ED may also be prompted by concerning stroke-like symptoms or altered mental status. That notwithstanding, implementing stewardship measures and improving local testing protocols could obviate the need for ED clinical and laboratory evaluation.⁶ For patients with chest pain, augmented evaluation strategies (e.g., modified HEART score with EKG testing) could be implemented to risk stratify patients for referral to the ED.

In rural communities, the ED has an expanded safety net role and is increasingly used due to provider shortages.^7,8 In response, the Biden–Harris Administration has announced $129 million in United States Department of Agriculture Emergency Rural Health Care Grants to increase access to care in rural communities.

Other potential solutions could focus on quality initiatives. Performance measures emphasizing on-site care (e.g., percentage of patients treated for UTI in-house as opposed to ED) could be reported to the public and payors. Additionally, federal law mandates reporting hours spent on CMD quality work. These measures could supplement traditional SNF quality performance measures (e.g., percentage of patients with delirium), given that traditional measures have not been consistently associated with risk of readmission,¹ which exhibits patterns similar to ED referral. Such measures would require careful crafting to avoid perverse incentives that might discourage appropriate ED referrals. Further study is necessary to evaluate proposed measures in reducing discretionary and preventable ED visitation.

Taken together, the significant volume of ED referrals and SDDs among postdischarge SNF patients warrants consideration of strategies to enhance approaches to on-site evaluation and acute care, recognizing that many ED referrals are wholly necessary. Several approaches hold promise and are worthy of further evaluation to minimize disruption of postacute care.