Discussion
This study investigated the association between health insurance status and disposition in patients who sustained lower extremity fractures, with a specific focus on identifying care deviations from recommended discharge pathways. Insurance type was found to be independently predictive of the likelihood of discharge to facility-based rehabilitation. Traditional Medicare had high rates of discharge to AR or SAR, with a low rate of care deviation from discharge recommendations. While rates of discharge to any facility remained high in patients with managed Medicare, discharge to AR markedly decreased, with a resultant increase in care deviation. The demographic and clinical similarities of these Medicare insurance types offer an opportunity to consider the potential influence of commercial health insurance carriers in the administration of government-sponsored healthcare.
Rehabilitation after traumatic injury
PM&R physicians consider injury type, preinjury and current functional status, comorbidities, ability to participate in PT and admission guidelines for rehabilitation facilities to determine what level of care a patient is likely to require and to ultimately offer recommendations for postacute care needs.4 5 Early evaluation by PT and PM&R has been shown to decrease HLOS and complication rates, including falls and readmissions.5 6 17 While PT directly assesses the patient’s current functional capabilities, PM&R physicians take a global view of the patient, including in their assessment comorbid conditions and medical/neurological impairments, rehabilitation trajectory, insurance payer type and insurance-specific requirements to qualify for inpatient rehabilitation. Even though these services work closely in conjunction with one another, variations in practice perspectives and training may lead to occasional discrepancies in their recommendations. These discrepancies are usually reconciled after subsequent evaluations and interdisciplinary discussion. If a discrepancy persists, PM&R recommendations carry more authority as the physician order to treat.18
Traditional Medicare versus managed Medicare
This study demonstrated high rates of discharge for all Medicare patients to rehabilitation facilities and low rates of discharge home after hospitalization. The characteristics of both traditional and managed Medicare patient groups were consistent with the general Medicare population: older with higher comorbidity burden, largely white and non-Hispanic and more often female. They presented most often after low-level falls, sustained femur and pelvis fractures most commonly and typically received less restrictive activity recommendations. These factors likely contribute to an increased expected benefit from a higher level of rehabilitation after discharge.3 19 Despite parallel demographics between managed and traditional Medicare, managed Medicare patients were discharged much less frequently to AR, reflective of a deviation in care from PT and PM&R recommendations. Moreover, on multivariate analysis, patients with managed Medicare were 54% less likely than traditional Medicare patients to be discharged to AR.
SAR utilization was unique, with low rates of care deviation across all payor types. We suspect that patients who ultimately require SAR have significant functional deficits and a limited ability to participate in the prolonged PT sessions required for AR, may be deemed unsafe for home discharge, and are left with no option other than SAR. These explanations align with known complexities in discharge planning and challenges in care transitions for Medicare patients after hospitalization.20–22 One study found that 90% of patients reported at least one barrier to discharge readiness inhibiting an effective transition home, which may include limited resources to provide appropriate levels of care at home, and likely relates to social determinants of health that are difficult to articulate with this study design.22 23
Medicare patients overall had a commendably low HLOS when discharged to a facility, compared with other payor types, though managed Medicare patients had a HLOS on average 2 days longer than those covered by traditional Medicare. This difference in HLOS for traditional Medicare and managed Medicare patients is likely due to two factors. The first involves the widely recognized benefit of early operative intervention (within 48 hours) for geriatric patients with hip fractures, the leading fracture type in Medicare patients.24 25 The second is the lack of prior authorization required by traditional Medicare. If criteria for postdischarge services are met, the patient can be discharged to the recommended discharge destination based on inpatient provider recommendations alone.26 When compared with traditional Medicare patients, those with managed Medicare may have prolonged HLOS due to non-clinical factors, which we suspect arises from the administrative burdens imposed by the insurance carrier administering the Medicare program; the lack of this administrative barrier in traditional Medicare may streamline the team’s ability to discharge patients with the necessary postacute services. Increased HLOS and care deviation from discharge recommendations, particularly for AR, portrays managed Medicare and its restrictive prior authorization as a barrier to access clinically important postacute services. In 2021, 2.1 million managed Medicare prior authorization requests were denied, and only 11% of those denials were appealed.27 Artificial intelligence algorithms are reportedly being used by Medicare Advantage plans to determine when coverage can be tenably denied to enrollees, who are likely unaware of the decreased prospects of receiving fundamentally important postacute services.28 29 Limitations on care access to the detriment of short-term and long-term recovery, heightened administrative burdens, with a prolonged HLOS carries significant financial and clinical implications.
Coverage for postacute services for working age adults
The other insurance types also demonstrated notable demographic and clinical similarities among themselves, differing from the Medicare groups. Medicaid, managed Medicaid, PI and self-pay patients were more often male, had lower comorbidity burden, similar ISS distributions and mechanisms of presentation, and increased rates of tibia-fibula, knee and ankle fractures than the Medicare groups. While approximately 90% of all Medicare patients were discharged to AR or SAR, this was significantly less common for the remaining insurance groups, who were largely discharged home. Home discharge is not necessarily inappropriate and may even be optimal for a younger and healthier patient population. However, consideration of discharge recommendations for AR demonstrates high rates of care deviation between recommended and actual disposition for PI, Medicaid, managed Medicaid and self-pay patients, which did not occur when patients in these insurance groups were recommended for home discharge. The high rates of care deviation may be a demonstration of the barrier prior authorization represents when seeking access to postacute services. The ability for third-party commercial insurance entities to position themselves as ‘gatekeepers’ in the care continuum is an intentional feature. From the perspective of commercial insurance payors, there is a clear financial incentive to limiting utilization of high-cost facility-based AR. Though medical record review in this study was unable to capture denials, we suspect the persistence of this finding and the directionally after a deviation in care suggests the commercial entity administering the health insurance plan is exerting its invisible hand of influence on patient care. This stark disparity in the face of objective medical recommendations yields concern that financial costs associated with facility-based rehabilitation may be influencing a patient’s opportunity to access this health resource, given that the cost of an inpatient rehabilitation stay has been estimated at $10 000–$24 200 U.S. dollars, with an average daily cost of $716 U.S. dollars.7 30 These high costs have the potential for financial devastation. Increased access to commercial health insurance coverage and increased Medicaid enrollment after Medicaid expansion resulting from the Affordable Care Act has shown improved access to rehabilitation services and enhanced protection from catastrophic health expenditures.31–33 However, it is also increasingly recognized that having health insurance does not necessarily protect against financial vulnerability, and catastrophic health expenditures can be high even in the privately ensured.34–36 In our study, while <20% of PI patients were ultimately discharged to AR, Medicaid, managed Medicaid and self-pay patients were even less likely to access this resource, representing only a fraction of the patients for whom AR was formally recommended.
Medicaid does not require prior authorization and functions similarly to traditional Medicare, providing coverage for a variety of injury-related care.37 Medicaid eligibility varies widely, and in our state, emergency Medicaid can be obtained to cover inpatient hospital expenses after emergent conditions but does not cover postdischarge services. We were unable to delineate how many patients in our group received emergency Medicaid coverage but suspect a substantial proportion given low rates of discharge to facilities, high rates of care deviation and prolonged HLOS.
Readmission rates in our study population were highest in Medicare, and when separated by disposition, were highest in patients discharged to AR. Medicare patients are at higher risk for readmission given their age and comorbidity status. Given the presence of an AR affiliated with our facility, the threshold for readmission may be lower due to proximity, which may also lend itself to better capture of readmissions.
Previous literature has described disparities in access to postacute care for minority populations after injury,38 39 though insurance status has been shown to mitigate this disparity.40 While race and ethnicity characteristics were significantly different across insurance categories, on multivariate analysis, they were not found to be statistically significant factors in likelihood of disposition to AR or SAR, with one important exception. Black patients had a higher likelihood of discharge to SAR, without significantly different rates of care deviation from discharge recommendations. This finding may suggest a bias on the clinician’s part, where black patients may be less likely to be recommended for AR.
Limitations
This study has several limitations that constrain the ability to draw further conclusions. Beyond its retrospective nature, there were evident differences among the insurance groupings that limit comparative analysis. Age is an inherent confounder given criteria for Medicare eligibility. We sought to mitigate this through our multivariate analysis model, and by subgroup analysis of patients over the age of 65 (though this attempt was likely underpowered). The Medicaid insurance category also likely contained a significant proportion of patients whose hospitalization was covered under Emergency Medicaid.41 Furthermore, for Emergency Medicaid or even self-pay patients with demonstrated need for postacute services, many may have received facility placement under governmental provisions for ‘charity care’, which we were unable to quantify.
Next steps
Despite these limitations, this study identified a key disparity through the influence health insurance status has on access to postacute care. The frequent care deviations from recommendations provided by licensed clinicians for patients enrolled in commercially administered health insurance suggest a departure from clinician-driven care and likely arise as a consequence of these entities acting as gatekeepers biased towards limiting care utilization. State and federal policy that establishes standards for coverage of post-AR services can decrease the administrative burdens imposed by prior authorization requirements and mitigate unregulated intrusion into clinician-directed care seeking to optimize patient recovery after injury. Instead, the utilization of PM&R physicians, specialized and trained in patient rehabilitation, to guide decisions on postacute care would facilitate a return to clinician-guided care, with the goal of ensuring patients receive appropriate post-AR, with improved functional outcomes and decreased long-term morbidity.
Future research can be structured to establish causality of adverse third-party influence at the population level and to identify potential impacts on long-term mobility and functional independence after acute traumatic injury.
