This cohort study reports on long-term trends in nursing home closures and how facility characteristics are associated with different closure patterns.
Key Points
Question
What are the long-term trends in US nursing home closures and ownership transitions and how are facility characteristics associated with different types of operational change?
Findings
In this cohort study of 33 110 US nursing home locations from 1985 to 2025, nearly half of the facilities followed simple entry-exit patterns, while approximately one-fifth remained operational despite ownership changes. Organizational restructuring, including change of ownership and within-facility capacity reconfiguration, became more common over time and was more prevalent than permanent facility closure or market exit.
Meaning
These findings suggest that nursing home closures often reflect broader organizational restructuring rather than true market exit, highlighting the importance of incorporating ownership transitions and capacity changes into nursing home policy and oversight.
Abstract
Importance
US nursing home closures, whether planned or unexpected, can disrupt continuity of care, place additional burdens on remaining facilities, and disproportionately affect vulnerable populations. It is important to understand the trends and factors associated with these closures, as such knowledge can inform policy decisions, guide resource allocation, and help develop strategies to mitigate negative effects on both residents and the broader health care system.
Objective
To examine nursing home operational dynamics in the US and to identify how facilities transition through openings, closures, and changes of ownership (CHOW).
Design, Setting, and Participants
A longitudinal cohort study used 40 years of data from the Centers for Medicare & Medicaid Services Provider of Services File, covering the period from 1985 through June 2025. Facility trajectories were examined using physical location identifiers rather than facility names or provider identifiers.
Main Outcomes and Measures
Facility lifecycle status (active, closed, or reopened), CHOW events, net changes in certified bed capacity, facility size, availability of specialized beds for patients with dementia, Medicare and Medicaid participation, and staffing intensity were assessed. Descriptive statistics and multivariate logistic regression were used to analyze trends and differences between closed and active facilities.
Results
A total of 39 449 records were analyzed, representing 33 110 unique physical locations and 80 257 event records. Nearly half of all entities followed an entry-exit trajectory without any CHOW, accounting for 46.29% (n = 15 328). A substantial proportion of entities remained active while experiencing at least 1 CHOW (20.48% [n = 6780]). CHOW events accounted for 13 053 occurrences (16.26%) of lifecycle activity, and capacity changes were increasingly concentrated in recent years in select states. Overall, organizational restructuring and within-facility capacity adjustment became more prominent over time than entry or exit.
Conclusions and Relevance
This study found that nursing home markets were characterized by continuous organizational change, with ownership transitions and capacity reconfiguration increasingly substituting for entry and exit. Examining the location-based life cycle of the facility shows that closure often reflects restructuring within evolving ownership structures rather than a single terminal event. These findings suggest that nursing home policy and oversight should attend to ownership transitions and capacity reconfiguration, alongside measures of facility entry and exit.
Introduction
Nursing homes (NHs) are a critical component of the US health care system, providing long-term skilled and custodial care to older adults and individuals with chronic illnesses or disabilities.1,2,3 They are especially important for populations covered by Medicare and Medicaid, which finance most long-term care services. NHs function as dual-purpose facilities, delivering both postacute skilled nursing care (primarily Medicare-financed) and long-term custodial care (primarily Medicaid-financed), with substantial differences in length of stay, clinical needs, and reimbursement.4 As the US population ages and demand for long-term care grows, maintaining stable access to NH services is a key policy priority.5 Despite their importance, NHs face persistent financial, workforce, regulatory, and reimbursement pressures that have contributed to closures over time.6,7,8,9,10,11,12 These closures disrupt continuity of care, reduce service availability, and place additional strain on remaining facilities, often worsening geographic and socioeconomic disparities in access.4,7,10,11 Compared with hospital closures, NH closures disproportionately affect a highly vulnerable population requiring long-term care and are associated with adverse outcomes, including functional decline, increased hospitalizations, and caregiver burden.4,7,9
NH closures are commonly classified by the Centers for Medicare & Medicaid Services (CMS) as voluntary or involuntary.12,13 Voluntary closures are often driven by strategic decisions such as mergers, acquisitions, restructuring, or market exit, influenced by financial pressures and reimbursement constraints.14 In contrast, involuntary closures are more often linked to regulatory enforcement, quality concerns, or sustained operational and workforce challenges, with implications for resident safety and continuity of care.11,12,15,16
Evidence also shows that closures are not evenly distributed and disproportionately affect facilities serving Medicaid-dependent, low-income, and racial and ethnic minority populations, raising concerns about inequitable access to care.3,7,17,18,19 These challenges may intensify as the population aged 75 years and older continues to increase, even as alternative care settings expand.11,12
At the same time, the evolving regulatory and payment landscape has further influenced NH viability and closure patterns. Since the 1980s, shifts from retrospective cost-based Medicare reimbursements to prospective payment systems, including the Resource Utilization Groups and, more recently, the Patient-Driven Payment Model, have reshaped financial incentives and operational practices within NHs.20,21,22,23 While these reforms aimed to enhance efficiency and better align payments with patient needs, they have also introduced new challenges, particularly for smaller or rural facilities that may find adaptation difficult, thereby impacting their sustainability and contributing to closure risk.18,24,25
Prior research suggests that regulatory and market changes have mixed and context-dependent effects on NH behavior and system capacity.3,8,22,24,25,26 However, most studies treat closures as discrete events, with limited attention to long-term facility trajectories, including ownership transitions, temporary closures, and reopenings.3,8 To address this gap, this study uses CMS Provider of Services (POS) data from 1985 to 2025 to examine NH operational dynamics over time, capturing how facilities transition through openings, closures, and changes of ownership (CHOW) within evolving structural and policy environments. Specifically, we aimed to describe NH closure trends in the US and evaluate how facility characteristics are associated with different closure patterns using longitudinal site-level facility trajectories that distinguish true permanent closures from operational transitions such as CHOW, temporary closures, and reopenings.
Methods
This longitudinal cohort study was conducted to examine NH closures and organizational transitions in the US from 1985 through June 2025, corresponding to the second quarter of 2025. The study period was June 2025 (using quarter two 2025 data). Facility trajectories were analyzed at the physical location level to distinguish true closures from CHOW and reopenings. Publicly available facility-level data from the CMS POS files were used. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline and according to the policies of the University of Massachusetts institutional review board was exempt from institutional review board approval and informed consent because it used publicly available nonidentifiable data.
Data Sources
All data were retrieved from CMS, POS File—Hospital & Non-Hospital Facilities, released for quarter 2 of 2025. This comprehensive administrative dataset includes detailed structural, operational, and ownership characteristics of all Medicare-certified facilities, including program participation (Medicare, Medicaid, or both), CHOW status, bed capacity, staffing levels by role, and reasons for facility termination.
Sampling and Study Participants
A total of 114 882 data records were retrieved from the CMS POS—Hospital & Non-Hospital Facilities. From this dataset, 39 403 records were specifically selected for skilled nursing facility (SNF) and nursing facility (NF) (Dually Certified & Distinct part), SNF, and NF. A total of 15 499 records were excluded because they represented facility types not relevant to the analysis, including general or short-term hospitals, home health agencies, hospice facilities, critical access or rural hospitals, and psychiatric or children’s facilities. The unit of analysis was defined as the physical location of each nursing home rather than individual event records, facility names, or facility identification numbers (IDs). Each record was linked to the facility ID; however, facility IDs were not used as the primary unit because they may change following ownership transfers. A location-based identifier was constructed using standardized address and state information to track continuity of operations at the same facility site over time. This allowed the analysis to distinguish true facility closures from temporary interruptions or ownership transitions, ensuring that events were attributed to the physical facility rather than administrative changes in facility identifiers.
Variables
The variables studied include facility life cycle status (active, closed, or reopened), CHOW events, net changes in certified bed capacity, facility size, availability of specialized beds for patients with dementia, Medicare and Medicaid participation, and staffing intensity. NHs that remained continuously enrolled in Medicare or Medicaid from 1985 to 2025 were considered active as of 2025.
NH closures were identified using CMS administrative data on termination of Medicare or Medicaid enrollment, as defined by the program termination code. Importantly, under CMS definitions, termination encompasses a range of circumstances and does not necessarily indicate physical closure of a facility, as some categories reflect CHOW, mergers, or administrative transitions rather than cessation of operations. Because these categories may capture fundamentally different underlying processes, the CMS classification was not used directly in the analysis. Instead, termination events were reclassified into analytically meaningful groups based on the underlying mechanism of exit or transition: (1) regulatory and compliance terminations included codes 05, 06, and 33, representing failures to meet health, safety, participation, or accreditation requirements; (2) financially driven terminations included code 08 (nonpayment of fees); (3) ownership and organizational transitions included code 07, reflecting changes in facility status or structure; and (4) voluntary exits included codes 01 to 04, capturing mergers, closures, reimbursement dissatisfaction, and other voluntary withdrawals.
NH bed capacity refers to the number of beds a facility is licensed to operate. A longitudinal, physical, location-level dataset of certified bed counts used net changes to capture variation in capacity over time, reflecting increases or decreases in the number of certified beds. Both absolute and proportional changes in bed capacity were calculated. A meaningful capacity change was defined as an event in which the absolute change in certified beds was at least 10 beds or the proportional change was at least 10% relative to the prior capacity. This combined threshold was used to distinguish substantive changes in facility capacity from routine fluctuations and minor operational adjustments.
An Alzheimer disease and related dementias (ADRD) unit was defined as having at least 1 bed specifically designated for patients with ADRD. Medicare or Medicaid or both indicates if the facility participates in Medicare only, Medicaid only, or both programs. Number of CHOWs refers to the total number of CHOWs an NH has undergone throughout its operation (reported numerically as CHOW, CHOW ≤ 1, CHOW = 1, and CHOW ≥ 2), spanning from 1985 to the second quarter of 2025.
The unit of analysis was defined at the facility location rather than the facility ID, as facility IDs may change following ownership transfers. A location-based identifier was constructed using standardized address and state information to track continuity of operations at the same physical site over time. This approach permitted the identification of facilities that continue to operate under new ownership despite changes in facility numbers.
To distinguish true facility closures from temporary interruptions or ownership transitions, termination events were linked to subsequent CHOW records. Facilities were classified as permanently closed only if no reopening or ownership transition occurred at the same location within 12 months. Facilities that resumed operations at the same location—under the same or a different facility ID—were classified as temporary closures or ownership transitions rather than permanent exits. A sensitivity analysis assessed robustness to alternative reentry windows (30, 180, 365, and 730 days). Facilities were grouped at the location level into 3 mutually exclusive states: active, permanently closed, and reopened. Reopening was defined as a new entry following termination within each time window. This approach allowed evaluation of whether life cycle classifications were sensitive to alternative assumptions regarding the timing of reactivation.
Statistical Analysis
All analyses were conducted using SAS version 9.4 (SAS Institute Inc). Descriptive and inferential analyses examined NH characteristics associated with closures and organizational transitions in the US from 1985 to 2025, including structural, staffing, operational, and geographic patterns across life cycle groups (active, closed, reopened, and ownership transitions). Descriptive statistics summarized facility characteristics across groups, with comparisons of bed size assessed using the χ2 test and Cramér V. Staffing variables were compared using the Kruskal-Wallis test due to nonnormal distributions and unequal group sizes. Geographic variation in closures was assessed using state-level counts and visualized through maps and trend analyses. Multivariable logistic regression models were used to examine key outcomes, with odds ratios (ORs), 95% CIs, and model discrimination (C-statistic) reported. Sensitivity analyses tested robustness using alternative closure definitions, varying reentry windows (6, 12, and 24 months) and excluding facilities with incomplete location data.
Results
A total of 39 449 records were analyzed, representing 33 110 unique physical locations and 80 257 event records. The 33 110 unique physical locations (location) that were identified were classified based on their longitudinal operational status and ownership history. From 1985 through the end of the study period, entry events were the most frequent, accounting for 39 449 observations (49.15%), followed by termination events (24 697 [30.77%]). CHOW events accounted for 13 053 occurrences (16.26%), while capacity change events were the least frequent (3058 [3.81%]). Figure 1 shows substantial changes in facility life cycle events from 1985 to 2025. In 1985, approximately 7400 events were observed, with entry (approximately 4900) and termination (approximately 1000) comprising the largest components, followed by capacity change (approximately 1200) and CHOW (approximately 300).
Figure 1. Bar Graphs Showing Temporal Trends in Facility Life Cycle Events, 1985-2025.
Facility events over time by type are shown. Note: Quarter two 2025 includes only partial-year data (January-June 2025) and does not represent a full calendar year. CHOW indicates change of ownership (number of CHOWs reported numerically as CHOW≤1, CHOW = 1, and CHOW≥2).
After 1985, activity declined sharply: totals decreased to approximately 1900 in 1990, 1600 in 1995, and approximately 1100 in 2000, with entry steadily decreasing while termination and CHOW became more comparable in magnitude. From 2005 to 2020, total events stabilized at lower levels (approximately 600 to 1000 annually) with a more balanced mix of event types. CHOW became relatively more prominent during this period (eg, approximately 345 in 2015 and approximately 250 in 2020), while entry and termination remained moderate and capacity change remained minimal. A clear structural shift has occurred in the most recent years. In 2024, total events increased again to approximately 1500, driven predominantly by capacity change (approximately 896), which became the dominant event type. In 2025 (partial year), total events remained elevated (approximately 600), again led by capacity change (approximately 571), while CHOW (approximately 12), entry (approximately 4), and termination (approximately 10) were minimal.
With regard to facility types among event records, nearly half were dual-certified facilities (19 678 [49.88%]), followed by nursing facilities (NF; 17 656 [44.76%]) and SNFs (2115 [5.36%]). Substantial temporal variation was observed across facility types. NF events dominated in the late 1980s, peaking at approximately 6500 to 6700 events per year, with more than 85% attributed to terminations. After this peak, NF events declined to approximately 1000 to 1200 in the mid-1990s, and decreased further to fewer than 200 events annually by 2015 to 2025. Dual-certified facilities exhibited relatively stable activity over time, ranging between 300 and 900 events per year, with no clear long-term trend but with peaks in the early 1990s and early 2010s. In contrast, SNF events remained consistently low throughout the study period, generally less than 200 events per year, declining gradually to fewer than 50 events annually in recent years (eFigure 1 in Supplement 1).
From 1985 to 2025, the distribution of life cycle patterns showed that nearly half of all entities followed an entry-and-exit trajectory without any CHOW, accounting for 46.29% (n = 15 328) (Figure 2). A substantial proportion of entities remained active while experiencing at least 1 CHOW, with 20.48% (n = 6780) classified as active with CHOW = 1. Entities with persistent operational stability and no ownership transitions comprised 14.84% (n = 4912). In contrast, a smaller share (5103 [15.41%]) exited permanently after undergoing 1 or more CHOWs. Few entities exhibited dynamic life cycle behavior involving reopening after closure, with 2.62% (n = 866) reopening once and 0.36% (n = 119) reopening multiple times. Entities experiencing multiple CHOWs while remaining active were rare (2 [0.01%]).
Figure 2. Flow Diagram Showing Longitudinal Classification of Nursing Homes by Activity Status, Closure–Reopening Patterns, and CHOW History.
CHOW indicates change of ownership; ID, identification number; NHs, nursing homes.
From 1985 through 2025, substantial geographic variation was observed in net changes in bed capacity across states. Texas had the largest positive net bed change (n = 17 971), far exceeding all other states. This was followed by Massachusetts (n = 8881), Illinois (n = 5740), New York (n = 4970), and Pennsylvania (n = 4522), with additional increases in Missouri, New Jersey, and Connecticut. In contrast, the largest net bed decreases were observed in Wisconsin (n = −5033), Iowa (n = −3159), Minnesota (n = −1865), and Ohio (n = −1824), with smaller decreases in Colorado and Kansas (Figure 3).
Figure 3. Map Showing State-Level Net Changes in Facility Bed Capacity Across the US (1985-2025).
Quarter two 2025 includes only partial-year data (January-June 2025) and does not represent a full calendar year. This figure was created using SAS software version 9.4 (SAS Institute Inc).
Across 1985 to 2025, facility terminations were largely driven by ownership or organizational transitions, which consistently accounted for most exits each year (Figure 4). The highest observed level occurred in 1985 (n = 5997), followed by sustained high counts in the late 1980s to early 1990s (eg, 967 in 1991, 875 in 1989), before gradually declining over time. Regulatory and compliance terminations remained comparatively low throughout the period but persisted steadily, ranging from approximately 60 to 92 per year in the mid-1980s, then stabilizing at approximately 10 to 40 per year in recent years (eg, n = 35 in 2023; n = 39 in 2024). Voluntary and exits for other reasons were also consistently small, generally below 200 annually in early years and typically fewer than 10 to 20 per year in the 2010s to 2020s. Financial distress–related terminations were rare throughout, appearing only sporadically (eg, n = 10 in 1985; typically 0-3 in most subsequent years).
Figure 4. Bar Graphs Showing Temporal Trends in Facility Termination.
Temporal trends in facility terminations by exit type are shown. Quarter two 2025 includes only partial-year data (January-June 2025) and does not represent a full calendar year.
Facility distribution varied substantially by life cycle group. Program participation was dominated by dual Medicare–Medicaid facilities (34 764 [43.3%]), followed by Medicaid-only (28 632 [35.7%]) and Medicare-only (3861 [4.8%]) (Table). Differences in program participation across life cycle groups were statistically significant (χ212 = 2236.6; P < .001; Cramér V = 0.118; 95% CI, 0.115-0.121). Most facilities were small (<100 beds; 42 131 [52.2%]), followed by 100 to 300 beds (37 094 [46.2%]), with few large facilities (>300 beds; 1032 [1.3%]). Bed size distribution also differed significantly across life cycle groups (χ212 = 2350.6; P < .001; Cramér V = 0.121; 95% CI, 0.118-0.124) (Table). Staffing intensity varied widely across life cycle groups. Registered nurse (RN) staffing was highest in entry-and-closed facilities (76.83 per 1000 beds) and lowest in active CHOW ≥ 2 facilities (9.17 per 1000 beds), while aide staffing was highest in entry-and-closed facilities (205.82 per 1000 beds). Administrative staffing remained low across all groups (mean [SD] 1.87 [26.42] per 1000 beds overall). Differences in staffing intensity across life cycle groups were statistically significant for all staffing measures. Presence of beds for patients with ADRD was observed in 7023 facilities (8.8%), with higher proportions among closed–reopened facilities (CHOW ≤ 1, 736 of 3871 [19.0%]; CHOW ≥ 2, 149 of 797 [18.7%]) compared with active facilities (active no CHOW, 673 of 4918 [13.7%]; entry-and-closed, 2265 of 36 230 [6.2%]). Compared with active facilities without CHOW, facilities with a single CHOW had significantly lower odds of having beds for patients with ADRD (OR, 0.75; 95% CI, 0.68-0.83). Reduced odds were also observed among facilities with a history of closure without reopening (OR, 0.58; 95% CI, 0.53-0.64) and entry-and-closed facilities without CHOW (OR, 0.42; 95% CI, 0.38-0.46). In contrast, closed–reopened facilities showed higher odds of presence of beds for patients with ADRD, including CHOW ≤ 1 transitions (OR, 1.48; 95% CI, 1.32-1.66) and CHOW ≥ 2 transitions (OR, 1.45; 95% CI, 1.19-1.76) (eFigure 2 in Supplement 1).
Table. Program Type, Life Cycle Status, Bed Size Distribution, and Staffing.
| Life cycle group | CHOW status, No. (%) | P value | Cramér V (95% CI) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Active (no CHOW) | Active + CHOW = 1 | Active + CHOW ≥ 2 | Closed + CHOW history (not reopened) | Closed, reopened (CHOW ≤ 1) | Closed, reopened (CHOW ≥ 2) | Entry + closed (no CHOW) | Total | |||
| Bed size | ||||||||||
| <100 | 2670 (6.3) | 5826 (13.8) | 0 | 9745 (23.1) | 1660 (3.9) | 248 (0.6) | 21 982 (52.2) | 42 131 (100) | <.001 | 0.121 (0.118-0.124) |
| 100-300 | 2145 (5.8) | 7654 (20.6) | 10 | 10 936 (29.5) | 2163 (5.8) | 541 (1.5) | 13 645 (36.8) | 37 094 (100) | ||
| >300 | 103 (10.0) | 90 (8.7) | 0 | 180 (17.4) | 48 (4.7) | 8 (0.8) | 603 (58.5) | 1032 (100) | ||
| Total | 4918 (6.1) | 13 570 (16.9) | 10 | 20 861 (26.0) | 3871 (4.8) | 797 (1.0) | 36 230 (45.2) | 80 257 (100) | ||
| Having ADRD beds | ||||||||||
| No ADRD beds | 4245 (86.3) | 12 128 (89.4) | 10 (100.0) | 19 103 (91.6) | 3135 (81.0) | 648 (81.3) | 33 965 (93.8) | 73 234 (91.2) | <.001 | 0.117 (0.114-0.120) |
| With ADRD beds | 673 (13.7) | 1442 (10.6) | 0 | 1758 (8.4) | 736 (19.0) | 149 (18.7) | 2265 (6.2) | 7023 (8.8) | ||
| Total | 4918 (100) | 13 570 (100) | 10 (100) | 20 861 (100) | 3871 (100) | 797 (100) | 36 230 (100) | 80 257 (100) | ||
| Program participation | ||||||||||
| Medicare only | 418 (12.3) | 232 (6.8) | 0 | 472 (13.9) | 35 (1.0) | 0 | 2704 (79.9) | 3861 (100) | <.001 | 0.118 (0.115-0.121) |
| Medicaid only | 170 (0.6) | 106 (0.4) | 0 | 7429 (26.5) | 36 (0.1) | 0 | 20 891 (72.4) | 28 632 (100) | ||
| Dual participation | 4330 (12.5) | 13 232 (38.1) | 10 | 12 960 (37.3) | 3800 (10.9) | 797 (2.3) | 12 635 (36.4) | 34 764 (100) | ||
| Staffing intensity, mean (SD) | ||||||||||
| RNs per 1000 beds | 11.35 (72.22) | 21.07 (686.25) | 9.17 (32.71) | 47.44 (674.48) | 35.15 (186.12) | 28.65 (114.80) | 76.83 (547.66) | 57.19 (788.83) | <.001 | NA |
| Aides per 1000 beds, mean | 46.90 (442.57) | 41.40 (204.54) | 57.40 (181.45) | 175.68 (1247.20) | 146.66 (425.27) | 109.52 (215.26) | 205.82 (1042.08) | 144.63 (791.10) | ||
| Administrators per 1000 beds | 0.75 (7.12) | 0.60 (16.62) | 0.51 (3.94) | 2.40 (63.56) | 1.26 (9.08) | 3.99 (61.68) | 3.62 (53.23) | 1.87 (26.42) | ||
Abbreviations: ADRD, Alzheimer disease and related dementias; CHOW, change of ownership (number of CHOWs reported numerically as CHOW ≤ 1, CHOW = 1, and CHOW ≥ 2); NA, not applicable; RN, registered nurse.
In sensitivity analysis, across all specifications, the number of continuously active facilities remained constant at 8997 (33.51%), while permanently closed facilities ranged from 17 008 (63.34%) under the 30-day window to 16 186 (60.28%) under the 730-day window. The proportion of reopened facilities increased from 847 (3.15%) under the 30-day specification to 1669 (6.22%) under the 730-day specification.
Discussion
This study used longitudinal CMS POS data to characterize NH life cycle trajectories in the US from 1985 to 2025, with a focus on operational transitions, CHOWs, and facility closures at the physical location level. NH operations exhibited substantial life cycle dynamism, suggesting that facilities should be viewed as evolving organizational entities rather than static institutions. Nearly half of locations (15 328 [46.29%]) followed a simple entry-exit pattern without CHOW, while 20.48% (6780) remained active despite ownership transitions, indicating that restructuring is a key mechanism of adaptation. Overall, NH markets reflect recurrent organizational reconfiguration rather than stable institutional persistence.27
Prior literature has shown that NH closure is influenced by facility financial performance, payer mix, staffing intensity, ownership structure, and local market competition, with Medicaid dependence, small size, and low occupancy increasing closure risk, while higher Medicare share, greater acuity, and prior CHOWs are associated with reduced closure likelihood.3,7,8,17,24,25,28 Our findings extend this literature by showing that “closure” in administrative data often reflects heterogeneous life cycle processes. Rather than a strict terminal event, closure may represent restructuring, consolidation, or reentry. Nearly half of facilities remained continuously active, while others exited permanently or reopened, often alongside repeated CHOWs, suggesting cyclical rather than linear lifespans. CHOW events accounted for 16.26% of events (n = 13 053) and increased over time, consistent with consolidation trends. Contrary to prior findings linking CHOW to higher closure risk,11 facilities without CHOW were more likely to close, suggesting that weaker or newer facilities may exit without acquisition. CHOW events appeared to be primarily market-driven rather than regulatory, although they may still disrupt staffing and care continuity.6,11,12,27
Capacity changes showed a similar evolution, with increasing concentration in recent years, indicating a shift toward within-facility restructuring rather than entry or exit. This trend was particularly evident in Texas, which experienced the largest net increase in certified bed capacity (n = 17 971), highlighting substantial geographic heterogeneity driven by policy and market conditions. Overall, system change increasingly occurs through expansion, contraction, and reconfiguration of existing facilities rather than turnover.
Staffing and service specialization further reflect structural differences. RN staffing ranged from 9.17 per 1000 beds in CHOW-intensive facilities to 76.83 in entry-exit facilities, although differences may partly reflect bed size effects. Lower staffing in CHOW facilities suggests instability and cost pressures, while higher values in entry-exit facilities may suggest early-stage operations or denominator effects. ADRD services were present in 8.8% of facilities, with a higher prevalence in CHOW-reopened groups (18.7%-19.0%) and lower prevalence in entry-exit facilities (6.2%), suggesting that specialization is more common in stable or complex organizations.
Finally, termination patterns indicate that exits reflect a mixture of organizational restructuring, ownership transitions, and true market exit rather than a single underlying process. The prominence of CHOW-related transitions suggests that consolidation and reclassification are more common than regulatory enforcement or abrupt failure. From a policy perspective, this may overstate regulatory exits while underrepresenting restructuring. These dynamics occur within broader regulatory and payment shifts, including Medicare and Medicaid reforms and the Nursing Home Requirements of Participation, which impose detailed compliance standards that may trigger administrative exits or restructuring rather than true market failure.16,17,28,29
Historically, these patterns reflect the sector’s transformation under evolving payment systems and consolidation pressures.22,26 The shift toward dual Medicare–Medicaid reliance and skilled post–acute care has been accompanied by declining entry and exit and increasing ownership transitions.29,30 This is consistent with broader financialization trends, including multifacility chains and private equity involvement, where ownership transfers replace facility closure as the primary adjustment mechanism.31
Limitations
This study has limitations. It relies on CMS POS administrative data, which lack detailed financial, clinical, and organizational information, limiting the assessment of underlying drivers of life cycle transitions. CHOWs and closures may reflect heterogeneous processes (eg, mergers, rebranding, partial restructuring) that cannot be fully distinguished. Although the study spans 4 decades, its observational design limits causal inference; results should be interpreted descriptively. Staffing and service measures are also limited proxies for care quality. Future research should link life cycle trajectories to resident outcomes and incorporate ownership structure (eg, private equity) and local market conditions to better explain restructuring dynamics.
Conclusions
This study provides a comprehensive longitudinal characterization of NH life cycle dynamics in the US over 4 decades using CMS POS data. Across 33 110 physical locations, we found that NH markets are defined less by stable institutional persistence and more by continuous organizational change, including entry, exit, ownership transitions, and capacity reconfiguration. Nearly half of facilities followed simple entry-exit patterns, while a substantial share remained active despite CHOWs, indicating restructuring as a key mechanism of adaptation. Notably, what is traditionally labeled as closure often reflects heterogeneous processes embedded within broader life cycle pathways, including consolidation and reclassification rather than solely regulatory or financial failure. Over time, system dynamics shifted from entry-exit dominated activity toward increasing reliance on ownership transitions and capacity adjustments, with pronounced geographic variation, including substantial bed expansion in states such as Texas. These patterns are consistent with a maturing and increasingly consolidated sector, potentially influenced by changing reimbursement structures and broader financialization of long-term care. Overall, these findings highlight that NH policy evaluation should move beyond static facility counts to incorporate dynamic measures of organizational change, ownership instability, and capacity reconfiguration to better reflect the evolving structure of the long-term care system.
eFigure 1. Temporal Trends in Provider Lifecycle Events by facility types, 1985–2025
eFigure 2. Lifecycle Group by Bed Size (left) and ADRD Status: Odds Ratios (right)
Data Sharing Statement
References
- 1.Tyler DA, Fennell ML. Rebalance without the balance: a research note on the availability of community-based services in areas where nursing homes have closed. Res Aging. 2017;39(5):597-611. doi: 10.1177/0164027515622244 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ruiz Aragoneses R, Navarro Chamizo M, Calvache González E, Steegmman L, Almaraz D, Bermejo Higuera JC. [Wish to hasten death in older adults living in nursing homes in the Spanish population: A descriptive study]. Rev Esp Geriatr Gerontol. 2025;60(6):101685. doi: 10.1016/j.regg.2025.101685 [DOI] [PubMed] [Google Scholar]
- 3.Xu L, Sharma H, Wehby GL. Effects of nursing home closures on occupancy and finances of nearby nursing homes. Med Care Res Rev. 2025;82(2):153-164. doi: 10.1177/10775587241296182 [DOI] [PubMed] [Google Scholar]
- 4.Law AC, Bosch NA, Song Y, et al. Patient outcomes after long-term acute care hospital closures. JAMA Netw Open. 2023;6(11):e2344377. doi: 10.1001/jamanetworkopen.2023.44377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.National Academies of Sciences, Engineering, and Medicine (US) . The National Imperative to Improve Nursing Home Quality: Honoring Our Commitment to Residents, Families, and Staff. A Consensus Study Report of the National Academies of Sciences, Engineering, Medicine. The National Academies Press; 2022. [PubMed] [Google Scholar]
- 6.Bowblis JR. Ownership conversion and closure in the nursing home industry. Health Econ. 2011;20(6):631-644. doi: 10.1002/hec.1618 [DOI] [PubMed] [Google Scholar]
- 7.Sharma H, Bin Abdul Baten R, Ullrich F, MacKinney AC, Mueller KJ. Nursing home closures and access to post-acute care and long-term care services in rural areas. J Rural Health. 2024;40(3):557-564. doi: 10.1111/jrh.12822 [DOI] [PubMed] [Google Scholar]
- 8.Werner RM, Chen X, Coe NB, Olenski AR. State nursing home minimum staffing mandates: increased staff levels, minimal impact on finances and closures, 2010-23. Health Aff (Millwood). 2026;45(3):269-277. doi: 10.1377/hlthaff.2025.01223 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bach-Mortensen AM, Goodair B, Degli Esposti M. Does outsourcing enable the survival of good care homes? A longitudinal analysis of all care homes in England, 2011-2023. BMJ Public Health. 2024;2(2):e001227. doi: 10.1136/bmjph-2024-001227 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Xu L, Sharma H, Wehby GL. The effect of nursing home closure on local employment in the United States. Health Serv Res. 2023;58(3):744-752. doi: 10.1111/1475-6773.14125 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hughes K, Feng Z, Li Q, Segelman M, Oliveira I, Dey JG. Rates of nursing home closures were relatively stable over the past decade, but warrant continuous monitoring. Health Aff Sch. 2023;1(2):qxad025. doi: 10.1093/haschl/qxad025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Weaver RH, Roberto KA, Brossoie N. A scoping review: characteristics and outcomes of residents who experience involuntary relocation. Gerontologist. 2020;60(1):e20-e37. doi: 10.1093/geront/gnz035 [DOI] [PubMed] [Google Scholar]
- 13.Liao JM, Ibrahim SA, Huang Q, et al. The proportion of marginalized individuals in US communities and hospital participation in bundled payments. Popul Health Manag. 2022;25(4):501-508. doi: 10.1089/pop.2021.0334 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lin SC, Zheng J, Epstein A, et al. The extended impact of the COVID-19 pandemic on long-term care residents in Medicare with frailty or dual Medicaid enrollment. J Am Geriatr Soc. 2024;72(11):3457-3466. doi: 10.1111/jgs.19131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lepore M, Scales K, Anderson RA, et al. Person-directed care planning in nursing homes: a scoping review. Int J Older People Nurs. 2018;13(4):e12212. doi: 10.1111/opn.12212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Centers for Medicare & Medicaid Services (CMS), Department of Health and Human Services (HHS) . Medicare program; prospective payment system and consolidated billing for skilled nursing facilities (SNF) final rule for FY 2019, SNF Value-Based Purchasing Program, and SNF Quality Reporting Program. Final rule. Fed Regist. 2018;83(153):39162-39290. [PubMed] [Google Scholar]
- 17.Centers for Medicare and Medicaid Services (CMS), Department of Health and Human Services (HHS) . Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long-term care hospital prospective payment system and policy changes and fiscal year 2017 rates; quality reporting requirements for specific providers; Graduate Medical Education; hospital notification procedures applicable to beneficiaries receiving observation services; technical changes relating to costs to organizations and Medicare Cost Reports; finalization of interim final rules with comment period on LTCH PPS payments for severe wounds, modifications of limitations on redesignation by the Medicare Geographic Classification Review Board, and extensions of payments to MDHs and low-volume hospitals. Final rule. Fed Regist. 2016;81(162):56761-57345. [PubMed] [Google Scholar]
- 18.Siconolfi D, Shih RA, Friedman EM, et al. Rural-urban disparities in access to home- and community-based services and supports: stakeholder perspectives from 14 states. J Am Med Dir Assoc. 2019;20(4):503-508.e1. doi: 10.1016/j.jamda.2019.01.120 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Li Y, Harrington C, Spector WD, Mukamel DB. State regulatory enforcement and nursing home termination from the Medicare and Medicaid programs. Health Serv Res. 2010;45(6 Pt 1):1796-1814. doi: 10.1111/j.1475-6773.2010.01164.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Spector WD, Limcangco MR, Ladd H, Mukamel D. Incremental cost of postacute care in nursing homes. Health Serv Res. 2011;46(1 Pt 1):105-119. doi: 10.1111/j.1475-6773.2010.01189.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Leiken A, Sexton TR, Silkman RH. A model to assess the quality-cost tradeoff in nursing homes. Health Serv Res. 1986;21(2 Pt 1):145-160. [PMC free article] [PubMed] [Google Scholar]
- 22.Prusynski RA, Brown C, Johnson JK, Edelstein J. Skilled nursing and home health policy: a primer for the hospital clinician. Arch Phys Med Rehabil. 2025;106(2):311-320. doi: 10.1016/j.apmr.2024.08.017 [DOI] [PubMed] [Google Scholar]
- 23.Liu Z, Meehan A, Brazier JF, Shield R, Gadbois EA. Implementing the patient driven payment model-perspectives from skilled nursing facility administrators. J Appl Gerontol. 2024;43(6):688-699. doi: 10.1177/07334648231223296 [DOI] [PubMed] [Google Scholar]
- 24.Hoffman GJ, Ha J, Fan Z, Li J. Associations between rural hospital closures and acute and post-acute care access and outcomes. Health Serv Res. 2025;60(3):e14426. doi: 10.1111/1475-6773.14426 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Keesee E, Malone T, Gurzenda S, Pink G. Rural hospital closures and nursing home outcomes. J Rural Health. 2025;41(2):e70026. doi: 10.1111/jrh.70026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Miller KEM, Gupta R, Polsky D. Growth of the Program of All-Inclusive Care for the Elderly and the role of for-profit programs. Health Aff Sch. 2025;3(1):qxae174. doi: 10.1093/haschl/qxae174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dalton K, Howard HA. Market entry and exit in long-term care: 1985-2000. Health Care Financ Rev. 2002;24(2):17-32. [PMC free article] [PubMed] [Google Scholar]
- 28.Centers for Medicare & Medicaid Services (CMS, Department of Health and Human Services (HHS) . Medicare and Medicaid programs; requirements for long-term care (LTC) facilities; notice of facility closure. Final rule. Fed Regist. 2013;78(53):16795-16806. [PubMed] [Google Scholar]
- 29.Centers for Medicare & Medicaid Services (CMS), Department of Health and Human Services (HHS) . Medicare Program; Hospital Inpatient Prospective Payment Systems for Acute Care Hospitals and the Long-Term Care Hospital Prospective Payment System and policy changes and fiscal year 2018 rates; quality reporting requirements for specific providers; Medicare and Medicaid electronic health record (EHR) incentive program requirements for eligible hospitals, critical access hospitals, and eligible professionals; provider-based status of Indian Health Service and Tribal Facilities and Organizations; costs reporting and provider requirements; agreement termination notices. Final rule. Fed Regist. 2017;82(155):37990-38589. [PubMed] [Google Scholar]
- 30.Sharma H, Perraillon MC, Werner RM, Grabowski DC, Konetzka RT. Medicaid and nursing home choice: why do duals end up in low-quality facilities? J Appl Gerontol. 2020;39(9):981-990. doi: 10.1177/0733464819838447 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Jin X, Uda K, Ishimaru M, et al. The effect of business operating systems on nursing home termination. Int J Public Health. 2023;68:1605439. doi: 10.3389/ijph.2023.1605439 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure 1. Temporal Trends in Provider Lifecycle Events by facility types, 1985–2025
eFigure 2. Lifecycle Group by Bed Size (left) and ADRD Status: Odds Ratios (right)
Data Sharing Statement
