The effects of the Patient Driven Payment Model (PDPM) and COVID-19 on nursing and therapy staffing levels among skilled nursing facilities
Highlight box
Key findings
• The study found significant declines in nursing staff intensity—registered nurse, licensed practical nurse, and certified nursing assistant—and therapist staffing intensity-occupational therapy, occupational therapy assistants, physical therapy, physical therapy assistants, and speech therapy, from the pre-Patient Driven Payment Model (PDPM) period.
• For-profit and chain-affiliated skilled nursing facilities (SNFs) had larger reductions in nursing staffing levels and higher therapy staffing levels than not-for-profit and independent facilities.
• As SNF occupancy rates increased towards the end of the pandemic, nursing and therapy staffing levels continued to decline.
• The Coronavirus Aid, Relief, and Economic Security (CARES) funding was not significantly associated with staffing levels. Nurse and therapy staffing reductions persisted despite the influx of CARES funding.
What is known and what is new?
• PDPM is the first SNF payment model that links reimbursement to the complexity of care and resident needs.
• SNF management may have reduced therapy services due to the change with therapy reimbursement within PDPM, and decreasing occupancy, during the pandemic.
• Occupancy rate increased towards the end of the pandemic but nurse and therapy staffing intensity continued to decline.
• Nursing staff intensity declined, potentially influenced by COVID-19 related challenges and workforce shortages.
What is the implication, and what should change now?
• SNF management faces the challenge of adapting to a nursing-centric payment model, necessitating policy adjustments amid, new staffing regulations, evolving payor mix, and demographic shifts.
Introduction
The U.S. healthcare system is shifting from fee-for-service to outcomes-based reimbursement models. This shift extends to skilled nursing facilities (SNFs). In 2019, the Patient Driven Payment Model (PDPM) was introduced in the SNFs to maintain budget neutrality (1) and address deficiencies within the previous prospective payment system (PPS) (2). For decades, SNFs operated under the PPS, which based reimbursements on the volume of physical, occupational, and speech therapy minutes provided to residents, as categorized within predetermined Resource Utilization Groups (RUGs) (3). However, concerns about this model’s focus on therapy volume led the Office of Inspector General at the U.S. Department of Health and Human Services (DHHS) to recommend a shift towards a payment model that would prioritize resident characteristics and care needs (4). In response, the Centers for Medicare and Medicaid Services (CMS) introduced the PDPM. This new model significantly altered the financial dynamics within SNFs by removing the incentive to maximize therapy minutes (5).
The PDPM established a more holistic reimbursement approach by balancing therapy with non-therapy services and incorporating greater emphasis on nursing care components, such as heart disease, cancer, Alzheimer’s disease, diabetes, and depression (3). Under the PPS reimbursement model, financial incentives were tied to therapy services volume (2), whereas the PDPM shifts the focus to more comprehensive evaluation of patient needs, emphasizing clinical complexity and the intensity of nursing care required. As such, PDPM may encourage SNFs to increase their census of clinically complex patients, as higher comorbidity scores now lead to higher reimbursements (5).
Prior research exploring staffing changes since the implementation of PDPM in 2019 falls into three categories: (I) research exploring the early impact of PDPM on staffing prior to coronavirus disease 2019 (COVID-19; January 2018 to March 2020); (II) research examining the impact of COVID-19 on staffing (January 2020 to September 2022); and (III) research examining changes in staffing patterns before and during COVID (January 2019 to March 2022).
Findings from the early implementation of PDPM indicated significant shifts in staffing patterns within SNFs. A study by McGarry and colleagues (6) showed a noticeable reduction in therapy staffing levels during the initial phase of PDPM (September through December 2019), with no significant increases in nursing staff. Similarly, Rahman and colleagues (7) found a significant overall decrease in therapy minutes post-PDPM compared to pre-PDPM levels.
Studies examining staffing patterns during COVID-19 have shown mixed results. A mixed methods study by Brazier and colleagues (8) found an overall reduction in nursing staff levels among 40 nursing homes during the COVID-19 pandemic (January 2020 through September 2022). Shen and colleagues (9) also observed declines in nursing staff levels among nursing homes experiencing severe COVID-19 outbreaks (June 2020 through January 2021). In contrast, a national study by Werner and colleagues (10) found no significant declines in nursing staff levels after adjusting for resident census (January 2020 through September 2020).
Finally, Prusynski and colleagues (11) explored changes in therapist staffing prior and during the pandemic (January 2019 to March 2022). Their study showed that PDPM resulted in a decrease in total therapist staffing, and this pattern persisted two years into the pandemic, despite fluctuations during the pandemic.
However, to date there have been no studies that have examined the effects of PDPM on both nursing and therapist staffing intensity, before, during, and after COVID-19. More specifically, our study using national data to provide a comprehensive analysis of changes in nurse and therapy staffing from 21 months pre-PDPM (January 2018 to September 2019) to 51 months post-PDPM (October 2019 to December 2023). We segmented the staffing data into five periods to capture the evolving impact of both PDPM and COVID-19: pre-PDPM (January 2018 to September 2019), post-PDPM/pre-COVID (October 2019 to February 2020), COVID first wave/peak (March 2020 to December 2020), COVID vaccine introduction (January 2021 to December 2021), and COVID endemic management (January 2022 to December 2023) (11).
Under PDPM, SNFs that admit clinically complex patients with multiple comorbidities can receive higher reimbursement. As such, SNFs may position themselves to increase revenues from clinically complex residents by increasing registered nurse (RN) staffing intensity. As such, we hypothesized that SNFs would increase nursing staff intensity in response to the PDPM policy changes. In addition, we hypothesized that SNFs would decrease therapy staffing intensity, due to the refocus on nursing services, in response to the PDPM policy changes.
Methods
Our study was a retrospective study examining the relationship between PDPM and changes in nurse and therapy staffing intensity from January 1, 2018 to December 31, 2023. The study was deemed non-human subjects by the University of Alabama at Birmingham Institutional Review Board.
Data sources
Data for our study from 2018 through 2023 was obtained from multiple sources to provide a comprehensive view of SNF staffing, organizational, and market characteristics. The CMS Payroll-Based Journal (PBJ) provided data on staffing hours for RNs, licensed practical nurses (LPNs), certified nursing assistants (CNAs), therapists, and therapist’s assistants. The CMS Provider of Service (POS) included data on SNF characteristics, such as for-profit status, total beds, and occupancy rate (12). Brown University’s Long-Term Care Focus (LTCFocus) provided case mix data. CMS Medicare Cost Reports were used to obtain payer mix data—Medicare, Medicaid, and private pay (13). The Area Health Resource Files (AHRF) provided data on market factors at the county level, such as Medicare Advantage (MA) penetration, per capita income, and the Rural-Urban Commuting Area (RUCA) codes. The U. S. DHHS Provider Relief Fund identified COVID-19 payments under the Coronavirus Aid, Relief, and Economic Security (CARES) Act. The Centers for Disease Control and Prevention (CDC) COVID-19 Public File collected data on COVID-19 SNF cases (14), and the CDC COVID-19 Data Tracker included data on county-level COVID-19 new cases.
Study sample
The study sample consisted of all non-hospital-based Medicare and Medicaid-certified SNFs (15) that treated at least one short-stay resident per day for the study period 2018–2023. We used year-month data from January 1, 2018, to December 31, 2023. Our study’s baseline data was January 2018 to September 2019: pre-PDPM. The study sample comprised 931,865 skilled nursing home (SNFs) year-months, or 80,721 SNF-year observations (14,015 unique SNFs).
Measures
The variables operationalization is shown on Table 1. The study’s dependent variables comprised eight nursing/therapist staffing measures: registered nurse hours per 100 resident days (RN PRD), LPN hours per 100 resident days (LPN PRD), certified nurse aide hours per 100 resident days (CNA PRD), physical therapist hours per 100 resident days (PT PRD), PT assistant hours per 100 resident days (PTA PRD), occupational therapist hours per 100 resident days (OT PRD), OT assistant hours per 100 resident days (OTA PRD), and speech therapist hours per 100 resident days (ST PRD).
Table 1
| Variable name | Description | Data source |
|---|---|---|
| Nursing staffing hours | Nurse hours per 100 residents for RN, LPN, and CNA | PBJ |
| Therapy staffing hours | Therapist hours per 100 residents for OT, OT assistant, PT, PT assistant, speech therapist | PBJ |
| For-profit status | Ownership status | POS |
| Total beds | Total number of certified beds | POS |
| Occupancy rate | Resident count/certified beds | POS |
| Chain affiliation | Chain affiliation | POS |
| Payer mix | Percentages of inpatient days whose primary payor is Medicare, Medicaid, vs. other pay | Cost report |
| Case-mix (acuity index) | Residential Classification System Resource Utilization Groups (RUG) resident-level assessment information | LTCFocus |
| Herfindahl index | Measures market concentration to determine market competitiveness (range, 0–100). The sum of the squared market shares of nursing homes in a county | POS |
| Per capita income | Total national income divided by the number of people in the nation | AHRF |
| Medicare advantage penetration | Percentage of Medicare beneficiaries enrolled in a Medicare Advantage plan in the county | AHRF |
| Geographic location | Rural-Urban Commuting Area (RUCA) codes 1–3 are identified as urban areas, while codes 4–10 are rural | AHRF |
| CARES funding | CARES funding per nursing home bed for August and October 2020 | DHHS Provider Relief Fund |
| COVID-19 SNF cases | Average weekly COVID-19 confirmed cases per 1,000 SNF residents | CDC |
| COVID-19 community cases | COVID-19 community cases per 10,000 population | CDC |
RN, registered nurse; LPN, licensed practical nurse; CNA, certified nursing assistant; PBJ, Payroll-based Journal; OT, occupational therapist; PT, physical therapist; POS, provider of service; LTCFocus, Long-Term Care Focus; AHRF, Area Health Resource File; CARES, Coronavirus Aid, Relief, and Economic Security; DHHS, Department of Health and Human Services; COVID-19, coronavirus disease 2019; SNF, skilled nursing facility; CDC, Centers for Disease Control and Prevention.
The study’s independent variable was a categorical variable comprising the post-PDPM/COVID periods with the pre-PDPM as the reference category: pre-PDPM (January 2018 to September 2019), post-PDPM/pre-COVID (October 2019 to February 2020), COVID first wave/peak (March 2020 to December 2020), COVID vaccine introduction (January 2021 to December 2021), and COVID endemic management (January 2022 to December 2023). The yearly coefficients for time showed the trend in staffing patterns over the study’s period.
The control variables included organizational and environmental/market variables that may be associated with staffing levels. Organizational variables included: for-profit status, total beds, occupancy rate, chain affiliation, percentage of Medicare residents, percentage of Medicaid residents, and case-mix (acuity index). For-profit status is a dichotomous variable indicating whether a facility is for-profit (1= yes; 0= no). For-profit SNFs have been associated with lower nursing staff levels. For-profit SNFs may reduce nursing staff levels to minimize costs and increase profitability (16). Size was measured by the number of beds. Given the larger staff of larger facilities, they may be better equipped to restructure internally to meet new staffing requirements imposed by the external environment (17). Occupancy rate consists of the percentage of beds that are occupied. Higher occupancy rate has been associated with higher staffing levels (10,18). Chain affiliation is a dichotomous variable indicating whether a facility is part of a multi-facility organization (1= yes; 0= no). Chain affiliated facilities may be better resourced to meet staffing requirements (19). Payer mix consists of the percentage of Medicare and Medicaid residents. SNFs with a higher Medicare payor mix may have higher staffing ratios to address the needs of the post-acute resident population (20). Case-mix index was used to control for patient acuity across SNFs. Higher resident acuity requires increased nursing staff levels (18).
Environmental/market-level variables included: the Herfindahl index, per capita income, MA penetration, geographic location, CARES funding per bed, COVID-19 SNF cases, and COVID-19 community cases. The Herfindahl-Hirschman Indexes (HHI) is a measure of market competition (i.e., market concentration). HHI ranged from 0–100, where values closer to 0 indicate perfect competition, and 100 indicates a monopoly. Per capita income was used to control for differences in economic conditions across markets that may impact SNFs’ staffing levels (21). MA/managed care penetration represents the proportion of Medicare beneficiaries enrolled in an MA plan in the county. Studies suggest that MA patients are more likely to be discharged home rather than to a SNF after a hospital visit, contributing to lower SNF occupancy rates, which in turn impacts SNF staffing levels (22). Geographic location identifies urban versus rural counties. RUCA codes 1–3 were classified as urban areas, while codes 4–10 were classified as rural areas. The CARES funding variable identifies payments made to SNFs during the COVID-19 pandemic. Approximately $1.9 billion of the CARES funding were allocated toward increased SNF staffing levels and education (August 2020; October 2020) (16,23). COVID-19 SNF cases and community cases were included as to account for the effect of the pandemic on staffing levels. SNFs with higher staffing levels had a lower probability of COVID-19 outbreaks and deaths (24,25).
Analysis
Descriptive statistics were used to summarize the characteristics of the different variables of our study sample. Categorical data was described using frequencies, while continuous data was described using means and standard deviation. We modeled the data using random effects linear regression to explore the relationships between nurse and therapy staffing levels and PDPM, before, during, and after the pandemic periods. Random effects models account for the repeated facility observations over time. Data were analyzed across five distinct periods: pre-PDPM (January 2018 to September 2019), post-PDPM/pre-COVID (October 2019 to February 2020), COVID first wave/peak (March 2020 to December 2020), COVID vaccine introduction (January 2021 to December 2021), and COVID endemic management (January 2022 to December 2023). State fixed effects controlled for interstate differences, while year fixed effects accounted for time trends. STATA/BE 17.0 was used to perform all analyses. Findings with a P value <0.05 were considered statistically significant.
Results
The descriptive statistics of the study sample of SNFs year-months are reported in Table 2. Staffing intensity for RN PRD, LPN PRD, and CNA PRD increased from the pre-PDPM period to the peak of the first COVID-19 wave. Following the COVID peak, there was a slight decline in staffing intensity (Figure 1). As of 2023, staffing intensity for RNs and LPNs had returned to the levels observed pre-PDPM; however, CNA staffing intensity remained lower than the pre-PDPM levels. Staffing intensity for PT PRD, PTA PRD, OT PRD, OTA PRD, and ST PRD decreased from the pre-PDPM across the other study period (Figure 2). A majority of the SNFs were for-profit and part of a chain. Regarding the occupancy rate, there was a slight increase in the period of post-PDPM/pre-COVID compared to pre-PDPM, but a significant decrease in the other periods. Residents that were covered by Medicare and Medicaid showed slight variations but remained relatively stable. The acuity index remained relatively stable. Per capita income and MA penetration increased from the pre-PDPM period across the other periods. CARES funding was distributed mostly in the first wave/peak of COVID-19. Finally, COVID-19 cases in SNFs and the community spiked during the pandemic; however, COVID-19 community cases experienced a decline in 2023.
Table 2
| Variables | Pre-PDPM (Jan 2018 to Sept 2019) | Post-PDPM/Pre-COVID (Oct 2019 to Feb 2020) | COVID first-wave/peak (Mar 2020 to Dec 2020) | COVID vaccine introduction (Jan 2021 to Dec 2021) | COVID endemic management (Jan 2022 to Sept 2023) | P value |
|---|---|---|---|---|---|---|
| Dependent variables | ||||||
| Registered nurse PRD | 41.11 (29.09) | 43.06 (31.03) | 47.36 (34.38) | 44.78 (32.52) | 41.06 (29.58) | <0.001 |
| Licensed practical nurse PRD | 79.46 (30.71) | 79.36 (30.86) | 84.10 (34.59) | 81.77 (33.47) | 79.18 (38.33) | <0.001 |
| Certified nursing assistant PRD | 214.75 (52.85) | 213.78 (52.44) | 215.94 (60.86) | 203.27 (60.67) | 201.29 (131.49) | <0.001 |
| Occupational therapist PRD | 7.65 (8.31) | 7.50 (8.53) | 6.96 (7.32) | 7.14 (7.41) | 6.82 (8.74) | <0.001 |
| Occupational therapist assistant PRD | 8.92 (8.22) | 8.88 (8.61) | 7.70 (7.35) | 7.94 (7.30) | 7.43 (7.97) | <0.001 |
| Physical therapist PRD | 7.90 (8.62) | 7.78 (8.76) | 7.36 (8.04) | 7.45 (7.98) | 6.90 (8.47) | <0.001 |
| Physical therapist assistant PRD | 10.59 (9.53) | 10.57 (9.88) | 9.27 (8.51) | 9.53 (8.44) | 8.88 (8.88) | <0.001 |
| Speech language therapist PRD | 4.97 (4.63) | 4.96 (4.57) | 4.79 (4.63) | 4.76 (4.45) | 4.27 (4.75) | <0.001 |
| Control variables | ||||||
| For-profit status | ||||||
| No | 3,577 (26.1) | 3,613 (26.5) | 3,588 (26.1) | 3,498 (25.9) | 3,421 (25.7) | 0.59 |
| Yes | 10,127 (72.9) | 10,009 (73.5) | 10,170 (73.9) | 9,991 (74.1) | 9,916 (74.3) | – |
| Total beds | 112.92 (58.29) | 111.85 (58.12) | 111.92 (58.63) | 112.17 (59.04) | 112.11 (58.84) | <0.001 |
| Occupancy rate | 81.20 (14.33) | 81.19 (14.55) | 72.17 (15.30) | 70.94 (16.04) | 75.44 (16.61) | <0.001 |
| Chain affiliation | ||||||
| No | 5,312 (38.8) | 5,373 (39.4) | 5,452 (39.6) | 5,421 (40.2) | 5,367 (40.2) | 0.08 |
| Yes | 8,392 (61.2) | 8,249 (60.6) | 8,306 (60.4) | 8,068 (59.8) | 7,970 (59.8) | – |
| Percentage Medicare residents | 13.03 (12.30) | 13.36 (12.72) | 14.00 (12.40) | 13.30 (11.76) | 13.22 (11.93) | <0.001 |
| Percentage Medicaid residents | 55.89 (26.20) | 55.73 (26.31) | 55.92 (25.92) | 56.24 (25.93) | 55.93 (25.16) | <0.001 |
| Percentage other pay residents | 31.01 (23.24) | 30.85 (23.26) | 30.01 (23.22) | 30.24 (23.06) | 30.85 (22.50) | <0.001 |
| Acuity index | 12.21 (1.39) | 12.17 (1.60) | 12.18 (1.70) | 12.15 (1.96) | 12.15 (1.95) | <0.001 |
| Herfindahl index | 3.27 (3.90) | 3.28 (3.92) | 3.26 (3.91) | 3.25 (3.90) | 3.27 (3.92) | 0.61 |
| Per capita income | 52,071 (15,685) | 54,155 (16,113) | 56,200 (15,993) | 60,520 (17,041) | 60,552 (17,110) | <0.001 |
| Medicare Advantage penetration | 34,460 (14,088) | 36.496 (13.800) | 38.277 (13.735) | 41.87 (13.36) | 44.85 (13.04) | <0.001 |
| Geographic location | ||||||
| Urban | 9,603 (70.1) | 9,566 (70.2) | 9,654 (70.2) | 9,460 (70.1) | 9,356 (70.2) | >0.99 |
| Rural | 4,101 (29.9) | 4,056 (29.8) | 4,104 (29.8) | 4,029 (29.9) | 3,981 (29.8) | – |
| CARES funding per bed | 0.00 (0.00) | 0.00 (0.00) | 278.71 (661.85) | 0.00 (0.00) | 0.00 (0.00) | <0.001 |
| COVID-19 SNF cases | 0.0 (0.00) | 0.00 (0.00) | 14.21 (48.91) | 4.03 (16.25) | 9.55 (22.46) | <0.001 |
| COVID-19 community cases | 0.0 (0.00) | 0.00 (0.00) | 64.92 (81.05) | 87.21 (80.06) | 55.16 (141.89) | <0.001 |
Data are presented as mean (standard deviation) for continuous variables and frequency (%) for categorical variables. SNF, skilled nursing facility; PDPM, Patient Driven Payment Model; COVID-19, coronavirus disease 2019; PRD, per resident day; CARES, Coronavirus Aid, Relief, and Economic Security.
The multivariate results are organized and presented in four separate sections: changes in nursing staff patterns over time, control variables related to nursing staff patterns, changes in therapy staffing patterns over time, and control variables related to therapy staffing patterns.
Changes in nursing staff patterns over time
Table 3 presents the random-effects regression results for RN, LPN, and CNA staffing intensity. Compared to RN staffing intensity in the pre-PDPM period, RN staffing intensity increased by 0.6 and 1.5 hours per 100 residents days in the post-PDPM/pre-COVID and COVID first wave/peak periods, respectively; but then decreased by 0.8 and 3.2 hours per 100 residents days in the COVID vaccine introduction and COVID endemic management periods, respectively. Compared to LPN staffing intensity in the pre-PDPM period, LPN staffing intensity decreased by 2.7, and 3.0 hours per 100 residents, in the COVID vaccine introduction and the COVID endemic management periods, respectively. Compared to CNA staffing intensity in the pre-PDPM period, CNA staffing intensity decreased by 3.2, 21.5, and 17.4 hours per 100 residents, in the COVID first wave, COVID vaccine introduction, and COVID endemic management periods, respectively. Given the general pattern of nursing staff reductions during the post-PDPM and COVID periods, our did not support Hypothesis 1.
Table 3
| Variables | Registered nurse PRD | Licensed practical nurse PRD | Certified nurse assistant PRD | |||||
|---|---|---|---|---|---|---|---|---|
| β-coef. | P value | β-coef. | P value | β-coef. | P value | |||
| PDPM/COVID periods | ||||||||
| Pre-PDPM | Reference | Reference | Reference | |||||
| Post-PDPM/pre-COVID | 0.643 | <0.001 | −0.197 | 0.12 | −0.268 | 0.54 | ||
| First COVID wave/peak | 1.538 | <0.001 | 0.173 | 0.40 | −3.132 | <0.001 | ||
| COVID vaccine introduction | −0.845 | <0.001 | −2.692 | <0.001 | −21.481 | <0.001 | ||
| COVID endemic management | −3.218 | <0.001 | −2.991 | <0.001 | −17.303 | <0.001 | ||
| Control variables | ||||||||
| For-profit status | ||||||||
| No | Reference | Reference | Reference | |||||
| Yes | −1.160 | <0.001 | −0.802 | <0.001 | −14.524 | <0.001 | ||
| Total beds | −0.103 | <0.001 | −0.046 | <0.001 | −0.111 | <0.001 | ||
| Occupancy rate | −0.342 | <0.001 | −0.480 | <0.001 | −0.801 | <0.001 | ||
| Chain affiliation | ||||||||
| No | Reference | Reference | Reference | |||||
| Yes | −0.005 | 0.94 | 0.126 | 0.19 | −4.511 | <0.001 | ||
| Percentage Medicare residents | 0.111 | <0.001 | 0.119 | <0.001 | 0.097 | <0.001 | ||
| Percentage Medicaid residents | −0.016 | <0.001 | −0.027 | <0.001 | −0.145 | <0.001 | ||
| Acuity index | −0.081 | <0.001 | 0.049 | 0.04 | 0.701 | <0.001 | ||
| Herfindahl index | −0.416 | <0.001 | −0.790 | <0.001 | −0.722 | <0.001 | ||
| Per capita income | −0.001 | <0.001 | 0.001 | 0.23 | 0.001 | <0.001 | ||
| Medicare Advantage penetration | 0.060 | <0.001 | −0.169 | <0.001 | −0.133 | <0.001 | ||
| Geographic location | ||||||||
| Urban | Reference | Reference | Reference | |||||
| Rural | −6.888 | <0.001 | −5.928 | <0.001 | −6.677 | <0.001 | ||
| CARES funding per bed | −0.001 | 0.045 | −0.001 | <0.001 | −0.002 | <0.001 | ||
| COVID-19 SNF cases | 0.001 | <0.001 | 0.001 | <0.001 | 0.001 | <0.001 | ||
| COVID-19 community cases | 0.001 | 0.07 | 0.001 | <0.001 | −0.009 | <0.001 | ||
PDPM, Patient Driven Payment Model; COVID-19, coronavirus disease 2019; PRD, per resident day; CARES, Coronavirus Aid, Relief, and Economic Security; SNF, skilled nursing facility.
Control variables related to nursing staff patterns
For-profit SNFs were associated with a decreased nursing staff intensity hours per 100 residents across all three categories: RN (−1.2), LPN, (−0.8), and CNA (−14.5) (Table 3). SNFs that were chain affiliated showed a decreased nursing staff intensity per 100 residents across one category: CNA (−4.5). RN and LPN staffing intensities were not statistically significant with chain affiliation. Higher occupancy rate was associated with a decrease of nursing staff hours per 100 residents (−0.3 RN, −0.5 LPN, and −0.8 CNA). An increase in the proportion of Medicare residents was associated with increased RN, LPN, and CNA staffing intensity, while an increase in the proportion of Medicaid residents was negatively associated with RN, LPN, and CNA staffing intensity. Decreased market competition was associated with lower RN (−0.4), LPN (−0.8), and CNA (−0.7) staffing intensity. Increased MA penetration was associated with increased RN staffing intensity (0.1), but lower LPN (−0.2) and CNA (−0.1) staffing intensity. Every $1,000 increase in per capita income was associated with a decrease of 1 RN hour but an increase of 1 CNA hour per 100 resident days. Regarding geographic location, nursing staff intensity for rural SNFs was associated with a reduction of 6.9 RN, 5.9 LPN, and 6.7 CNA hours per 100 residents compared to urban SNFs. Finally, CARES funding showed a small negative effect on staffing hours. A $100 increase in CARES funding was associated with a decrease of 0.1 RN, 0.1 LPN, and 0.2 CNA nursing staff hours per 100 residents.
Changes in therapy staffing patterns over time
Table 4 presents the random-effects regression results of OT, OTA, PT, PTA, and ST staffing intensity. Compared to OT staffing intensity in the pre-PDPM period, hours per 100 residents days decreased by 0.1, 0.5, 0.9, and 1.0 in the post-PDPM/pre-COVID, COVID first wave/peak, COVID vaccine introduction, and COVID endemic management periods, respectively. Compared to OTA in the pre-PDPM period, hours per 100 residents days decreased by 0.7, 1.2, and 1.6 in the COVID first wave/peak, COVID vaccine introduction, and COVID endemic management periods, respectively. OTA staffing intensity did not change significantly during the post-PDPM/pre-COVID period. Compared to PT staffing intensity in the pre-PDPM period, hours per 100 residents days decreased by 0.1, 0.5, 1.1, and 1.4 in the post-PDPM/pre-COVID, COVID first wave/peak, COVID vaccine introduction, and COVID endemic management periods, respectively. Compared to PTA staffing intensity in the pre-PDPM period, hours per 100 residents days decreased by 0.8, 1.1, and 1.5, in the COVID first wave/peak, COVID vaccine introduction, and COVID endemic management periods, respectively. PTA staffing intensity did not change significantly in the post-PDPM/pre-COVID period. Compared to ST staffing intensity in the pre-PDPM period, hours per 100 residents decreased by 0.2, 0.3, and 0.6 in the COVID first wave/peak, COVID vaccine introduction, and COVID endemic management periods, respectively. ST staffing intensity did not change significantly in the post-PDPM/pre-COVID period. Given the increasing reduction in therapist staffing post-PDPM, particularly since COVID, hypothesis 2 was supported.
Table 4
| Variables | Occupational therapist PRD | Occupational therapist assistant PRD | Physical therapist PRD | Physical therapist assistant PRD | Speech language therapist PRD | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β-coef. | P value | β-coef. | P value | β-coef. | P value | β-coef. | P value | β-coef. | P value | |||||
| PDPM/COVID periods | ||||||||||||||
| Pre-PDPM | Reference | Reference | Reference | Reference | Reference | |||||||||
| Post-PDPM/pre-COVID | −0.087 | 0.004 | 0.024 | 0.42 | −0.131 | <0.001 | 0.035 | 0.28 | −0.036 | 0.043 | ||||
| First COVID wave/peak | −0.462 | <0.001 | −0.689 | <0.001 | −0.483 | <0.001 | −0.755 | <0.001 | −0.231 | <0.001 | ||||
| COVID vaccine introduction | −0.946 | <0.001 | −1.213 | <0.001 | −1.128 | <0.001 | −1.092 | <0.001 | −0.344 | <0.001 | ||||
| COVID endemic management | −0.997 | <0.001 | −1.593 | <0.001 | −1.414 | <0.001 | −1.525 | <0.001 | −0.631 | <0.001 | ||||
| Control variables | ||||||||||||||
| For-profit status | ||||||||||||||
| No | Reference | Reference | Reference | Reference | Reference | |||||||||
| Yes | 0.275 | <0.001 | 0.050 | 0.18 | 0.159 | <0.001 | 0.173 | <0.001 | −0.003 | 0.88 | ||||
| Total beds | −0.019 | <0.001 | −0.010 | <0.001 | −0.018 | <0.001 | −0.014 | <0.001 | −0.009 | <0.001 | ||||
| Occupancy rate | −0.051 | <0.001 | −0.034 | <0.001 | −0.058 | <0.001 | −0.043 | <0.001 | −0.028 | <0.001 | ||||
| Chain affiliation | ||||||||||||||
| No | Reference | Reference | Reference | Reference | Reference | |||||||||
| Yes | 0.084 | <0.001 | 0.104 | <0.001 | −0.058 | 0.007 | 0.018 | 0.47 | 0.099 | <0.001 | ||||
| Percentage Medicare residents | 0.109 | <0.001 | 0.138 | <0.001 | 0.107 | <0.001 | 0.170 | <0.001 | 0.056 | <0.001 | ||||
| Percentage Medicaid residents | −0.006 | <0.001 | −0.001 | 0.054 | −0.005 | <0.001 | −0.001 | 0.13 | −0.001 | 0.01 | ||||
| Acuity index | −0.032 | <0.001 | −0.013 | 0.03 | −0.018 | 0.001 | −0.035 | <0.001 | 0.043 | <0.001 | ||||
| Herfindahl index | −0.160 | <0.001 | −0.158 | <0.001 | −0.167 | <0.001 | −0.168 | <0.001 | −0.127 | <0.001 | ||||
| Per capita income | 0.001 | 0.006 | 0.001 | <0.001 | 0.001 | 0.18 | 0.001 | <0.001 | 0.000 | <0.001 | ||||
| Medicare Advantage penetration | 0.003 | 0.22 | 0.019 | <0.001 | 0.012 | <0.001 | −0.010 | <0.001 | −0.021 | <0.001 | ||||
| Geographic location | ||||||||||||||
| Urban | Reference | Reference | Reference | Reference | Reference | |||||||||
| Rural | −2.097 | <0.001 | −0.682 | <0.001 | −2.393 | <0.001 | −1.126 | <0.001 | −1.303 | <0.001 | ||||
| CARES funding per bed | 0.001 | 0.003 | 0.001 | <0.001 | 0.001 | <0.001 | 0.001 | <0.001 | 0.001 | 0.01 | ||||
| COVID-19 SNF cases | −0.003 | <0.001 | −0.005 | <0.001 | −0.003 | <0.001 | −0.005 | <0.001 | −0.003 | <0.001 | ||||
| COVID-19 community cases | 0.001 | <0.001 | −0.001 | <0.001 | 0.001 | <0.001 | −0.001 | <0.001 | 0.001 | <0.001 | ||||
PDPM, Patient Driven Payment Model; COVID, coronavirus disease; PRD, per resident day; CARES, Coronavirus Aid, Relief, and Economic Security; SNF, skilled nursing facility.
Control variables related to therapy staffing patterns
For-profit SNFs were associated with increased therapy staffing intensity hours per 100 residents days across three categories: OT (0.3), PT (0.2), and PTA (0.2). OT assistants and ST staffing intensities were not statistically significant. Chain affiliated facilities showed increased therapy staffing intensity per 100 residents days across three categories: OT (0.1), OTA (0.1), and ST (0.1), and decreased in one category: PT (−0.1). A 10% increase in occupancy rate was associated with a decrease in therapy staffing hours per 100 residents days (−0.5 OT, −0.3 OTA, −0.6 PT, −0.4 PTA, and −0.3 ST). An increase in the proportion of Medicare residents was associated with increased therapist (OT, OTA, PT, PTA, ST) staffing intensity, while an increase in the proportion of Medicaid residents was negatively associated with therapist staffing intensity, except for PTA which was not significant. Decreased market competition and increased per capita income were associated with lower therapist (OT, OTA, PT, PTA, ST) staffing intensity. Increased MA penetration was associated with increased OTA and PT staffing intensity, but lower PTA and ST staffing intensity. Every $1,000 increase in per capita income was associated with an increase of 1 therapy staffing hour across all types of therapists and therapist assistants, except for PT which was not significant. Regarding geographic location, therapy staffing intensity for SNFs located within a rural area was associated with a reduction of 2.1 OT, 0.7 OTA, 2.4 PT, 1.1 PTA, and 1.3 ST hours per 100 residents when compared to SNFs located in urban areas. Finally, a $100 increase in CARES funding was associated with an increase of 0.1 therapy staffing hours per 100 residents, across all types of therapists and therapists assistants.
Discussion
This study examined the association between PDPM policy changes and nursing and therapy staffing intensity capturing data across five distinct periods to assess the evolving impact of both PDPM and COVID-19 on nurse and therapist staffing intensity. Regarding nursing staff, RN staffing increased during the post-PDPM/pre-COVID and COVID first wave/peak periods; however, there was a decrease in RN staffing intensity after the COVID vaccine introduction period. LPN and CNA staffing intensity showed a decline, particularly since the COVID vaccine introduction period. In terms of therapist staffing, we found decreased OT, PT, and ST staffing intensity starting from the post-PDPM/pre-COVID period, while OTA and PTA staffing intensity decreased beginning with the COVID first wave/peak period.
The initial increase in RN staffing during the post-PDPM and COVID first wave/peak periods may be attributed to SNFs adjusting to PDPM’s nursing-centric payment model (3) and responding to COVID’s initial impact on nursing homes. However, this proved to be temporary. There were overall reductions in nursing staff particularly since the COVID vaccine introduction, which may be attributed to several factors. First, nursing staff shortages and the increased cost associated with managing COVID-19 (8) likely played a significant role. Additionally, COVID-19 transmission may have caused workforce burnout or fear of infection (9), further reducing the pool of available RNs, LPNs, and CNAs. Furthermore, SNFs have faced challenges in attracting and retaining workers due to the low wages in a highly competitive environment (26).
The overall therapy staffing reductions may have been influenced by a combination of factors, including the effects of PDPM and the COVID-19 pandemic. For instance, the decrease in therapy staffing levels might be attributed to SNF management reducing skilled therapy due to the limited financial incentives under the new PDPM payment model. Our findings align with the current literature. Prusynski et al. (11) findings regarding therapy staffing levels, found a decrease in total therapy staffing between January 2019 and March 2022. McGarry and colleagues (6) findings regarding therapy staffing levels, found a decrease of therapy staffing between September 2019 and December 2019. Similarly, Rahman’s study (7), which measured therapy minutes rather than therapy staffing levels between January 2018 and March 2020, found a total reduction of therapy minutes following the implementation of PDPM.
In 2022, CMS reduced the fee schedule for therapy assistants to 85% for skilled and Part B therapy services (27). The decrease in therapy assistant staffing levels may reflect SNFs response to this fee reduction by increasing therapy ratios and decreasing therapy assistant ratios.
The CARES Act funding had minimal impact on nurse and therapy staffing intensity. SNFs may have directed CARES funds towards other immediate operational needs, such as improved infection control protocols and increased personal protective equipment (PPE) (28).
On average, for-profit and chain-affiliated SNFs had lower nursing staffing intensity than not-for-profit and independent facilities. This may be attributed to cost containment efforts, such as reducing nursing staff levels, which research suggests are more common in for-profit and chain-affiliated SNFs to improve profitability (29). On the other hand, for-profit and chain-affiliated SNFs had higher therapy staffing intensity than not-for-profit and independent homes. This may be potentially due the reliance of for-profit and chain-affiliated nursing facilities on therapy services to generate revenue.
Policy implications
Our study findings have significant implications for future nursing home regulations and reimbursement models. The observed decline in both nursing and therapy staffing levels, particularly in the context of the PDPM and the COVID-19 pandemic, underscores the need for policymakers to closely monitor staffing level and financial incentives.
The ongoing reduction in nursing staff intensity, despite the nursing-centric approach of the PDPM, suggests that the model’s incentives may not be fully aligned with ensuring adequate staffing levels. Given concerns with current SNF nursing staff levels, the Biden administration has recently implemented new staffing regulations. These regulations (CMS Final Rule for Minimum Staffing Standards for Long-Term Care Facilities), which took effect on May 2024, establish minimum staffing standards for long-term care facilities, including a total nursing staff standard of 3.48 hours per resident day (HPRD), which must include at least 0.55 RN HPRD and 2.45 CNA HPRD (30). These new guidelines, will be part of the annual facility assessments and enable surveyors to cite nursing homes for non-compliance if they find staffing levels to be inadequate (31).
The PDPM’s focus on higher reimbursement for residents with more complex health conditions may inadvertently incentivize nursing homes to prioritize revenue generation over appropriate staffing levels. Policymakers should explore more nuanced reimbursement structures that not only reward the admission of clinically complex patients but also ensure that these facilities have the necessary staffing to provide adequate care.
The continued decline in nursing home staffing intensity, especially following the COVID-19 pandemic, signals the need for workforce development policies. Recruitment and retention strategies, including competitive wages and improved working conditions, should be prioritized, particularly in for-profit and chain-affiliated nursing homes, which have been shown to maintain lower staffing levels.
Research indicates that COVID-19 transmission led to reduced SNF occupancy rates, which limited therapy access to the SNF population. As occupancy declined, therapy staffing levels may have been reduced, impacting the SNFs’ capacity to deliver outpatient Part B therapy services. Further research is needed to explore the potential correlation between therapy staffing intensity and the provision of outpatient Part B therapy. Additionally, the reduction in therapy assistant staffing levels may reflect SNFs adjustments in anticipation of CMS’s reduced therapy assistant fee schedule. Future research should examine the impact of this fee reduction on therapy staffing intensity.
Limitations
There are several limitations to this study. First, it is challenging to fully isolate the effect of the PDPM from that of the pandemic. However, our study analyses longitudinal data that includes pre-PDPM data (January 2018 to September 2019) as well as COVID endemic management data (2022 to 2023). In addition, we controlled for variations in COVID-related variables, such as COVID infection rates at the SNF and community levels, as well as the distribution of CARES funding. Second, our study did not explore how variations in the Medicare payer mix may have affected nursing homes’ response to PDPM/COVID. Third, we focused on individual changes in nursing and therapist staffing levels. However, future research could explore the complex interplay and substitutions between different types of staffing within SNFs. This would involve examining how changes in one type of staffing might influence or compensate for changes in others. Fourth, our study focused on the effect of PDPM on staffing levels during the pandemic. Future research is needed to determine if patient outcomes have been affected within the PDPM model. Finally, further studies should continue to assess the long-term impacts of PDPM on staffing and patient outcomes, particularly as the nursing home industry recovers from the effects of the COVID-19 pandemic.
Conclusions
Since 1999, SNF management has largely focused on a therapy-centric model. However, the shift to the PDPM in 2019 brought a new focus on nursing-centric care, tailored to the acuity and clinical complexity of residents. Despite this change, we found a decline in SNF nurse and therapy staffing levels, potentially due to financial constraints and workforce shortages. This highlights a critical need for SNF leadership to reevaluate and update their operational strategies to ensure they are aligned with the new reimbursement environment. These changes are vital to navigating the evolving payer mix and the increasing complexity of resident care, ensuring that both quality of care and financial viability are maintained. While PDPM represents a critical shift in nursing home reimbursement policy, further adjustments to financial incentives may be necessary to ensure they promote adequate staffing levels and high-quality care across both nursing and therapy services. Future research should examine the long-term effects of PDPM on staffing and resident outcomes, investigate the relationship between therapy staffing and outpatient Part B therapy services, and analyze the impact of CMS’ therapy assistant fee schedule reduction on staffing levels.
Acknowledgments
None.
Footnote
Provenance and Peer Review: This article was commissioned by the editorial office, Journal of Hospital Management and Health Policy for the series “Healthcare Finance: Drivers and Strategies to Improve Performance”. The article has undergone external peer review.
Peer Review File: Available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-36/prf
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-36/coif). The series “Healthcare Finance: Drivers and Strategies to Improve Performance” was commissioned by the editorial office without any funding or sponsorship. N.B. and R.W.M. serves as the unpaid Guest Editors of the series. R.W.M. serves as an unpaid editorial board member of Journal of Hospital Management and Health Policy from March 2023 to February 2025. The authors have no other conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. IRB approval and informed consent was not required as there is no human subject involved.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
References
- CMS. Fiscal Year (FY) 2022 Skilled Nursing Facility (SNF) Prospective Payment System (PPS) Final Rule (CMS-1746-F) 2021 [cited 2021 12/09/2021]. Available online: https://www.cms.gov/newsroom/fact-sheets/fiscal-year-fy-2022-skilled-nursing-facility-snf-prospective-payment-system-pps-final-rule-cms-1746
- Department of Health and Human Services, Office of Inspector General, Services DoHaH. Questionable billing by skilled nursing facilities. Washington, DC: Department of Health and Human Services, Office of Inspector General, Services DoHaH; 2010 Dec. Contract No.: OEI-02-09-00202.
- 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;83:39162-290. [PubMed]
- CMS. Medicare Program; Prospective Payment System and Consolidated Billing for Skilled Nursing Facilities: Revisions to Case-Mix Methodology. 2017. Available online: https://www.federalregister.gov/documents/2017/06/14/2017-12324/medicare-program-prospective-payment-system-and-consolidated-billing-for-skilled-nursing-facilities
- CMS. SNF PPS: Patient Driven Payment Model. U.S. Department of Health and Human Services (HHS); 2018 [cited 2021 11/15/2021]. Available online: https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/Downloads/MLN_CalL_PDPM_Presentation_508.pdf
- McGarry BE, White EM, Resnik LJ, et al. Medicare's New Patient Driven Payment Model Resulted In Reductions In Therapy Staffing In Skilled Nursing Facilities. Health Aff (Millwood) 2021;40:392-9. [Crossref] [PubMed]
- Rahman M, White EM, McGarry BE, et al. Association Between the Patient Driven Payment Model and Therapy Utilization and Patient Outcomes in US Skilled Nursing Facilities. JAMA Health Forum 2022;3:e214366. [Crossref] [PubMed]
- Brazier JF, Geng F, Meehan A, et al. Examination of Staffing Shortages at US Nursing Homes During the COVID-19 Pandemic. JAMA Netw Open 2023;6:e2325993. [Crossref] [PubMed]
- Shen K, McGarry BE, Grabowski DC, et al. Staffing Patterns in US Nursing Homes During COVID-19 Outbreaks. JAMA Health Forum 2022;3:e222151. [Crossref] [PubMed]
- Werner RM, Coe NB. Nursing Home Staffing Levels Did Not Change Significantly During COVID-19. Health Aff (Millwood) 2021;40:795-801. [Crossref] [PubMed]
- Prusynski RA, Humbert A, Leland NE, et al. Dual impacts of Medicare payment reform and the COVID-19 pandemic on therapy staffing in skilled nursing facilities. J Am Geriatr Soc 2023;71:609-19. [Crossref] [PubMed]
- CMS. Nursing Homes including Rehab Services 2022 [cited 2022 04/05/2022]. Nursing Home Compare]. Available online: https://data.cms.gov/provider-data/search?theme=Nursing%20homes%20including%20rehab%20services
- CMS. Cost Reports: U.S. Centers for Medicare & Medicaid Services; 2022 [updated 12/31/2021; cited 2022 04/05/2022]. cost report data on SNFs updated on 12/31/2021]. Available online: https://www.cms.gov/research-statistics-data-and-systems/downloadable-public-use-files/cost-reports
- CDC. Nursing Home COVID-19 Data Dashboard 2022 [updated 01/19/2021. Available online: https://www.cdc.gov/nhsn/covid19/ltc-report-overview.html
- HRSA. SNF COVID-19 Relief Payments by State 2020 [cited 2021 12/5/2021]. Available online: https://www.hrsa.gov/provider-relief/data/targeted-distribution
- Pradhan R, Weech-Maldonado R, Harman JS, et al. Private equity ownership and nursing home financial performance. Health Care Manage Rev 2013;38:224-33. [Crossref] [PubMed]
- Zinn JS, Weech RJ, Brannon D. Resource dependence and institutional elements in nursing home TQM adoption. Health Serv Res 1998;33:261-73. [PubMed]
- Bowens CS. Relationship Between Skilled Nursing Facility Nurse Staffing Levels and Resident Rehospitalizations [D.H.A.]. Ann Arbor: Walden University; 2019.
- Hirth RA, Zheng Q, Grabowski DC, et al. The Effects of Chains on the Measurement of Competition in the Nursing Home Industry. Med Care Res Rev 2019;76:315-36. [Crossref] [PubMed]
- Mukamel DB, Kang T, Collier E, et al. The relationship of California's Medicaid reimbursement system to nurse staffing levels. Med Care 2012;50:836-42. [Crossref] [PubMed]
- Shin DY. Measuring the Relationship between Contextual Factors, Nurse Staffing Patterns, and Hospital Performance. All ETDs from UAB. 2958. 2014. Available online: https://digitalcommons.library.uab.edu/etd-collection/2958
- Teigland C, Pulungan Z, Shah T, et al. As It Grows, Medicare Advantage Is Enrolling More Low-Income and Medically Complex Beneficiaries 2020. Available online: https://www.commonwealthfund.org/sites/default/files/2020-05/Teigland_Medicare_Advantage_beneficiary_trends_ib.pdf
- HHS. HHS Announces Allocations of CARES Act Provider Relief Fund for Nursing Homes 2020. Available online: https://www.hpnonline.com/regulatory/article/21149513/hhs-announces-allocations-of-cares-act-provider-relief-fund-for-nursing-homes
- Joshi S. Staffing Shortages, Staffing Hours, and Resident Deaths in US Nursing Homes During the COVID-19 Pandemic. J Am Med Dir Assoc 2023;24:1114-9. [Crossref] [PubMed]
- Gorges RJ, Konetzka RT. Staffing Levels and COVID-19 Cases and Outbreaks in U.S. Nursing Homes. J Am Geriatr Soc 2020;68:2462-6. [Crossref] [PubMed]
- Newswire P. Nursing shortage in South Florida has healthcare facilities scrambling to counter the talent drain: LifeWings Peak Performance is hosting a quickly arranged webinar in response to a critical situation aggravated by the COVID-19 pandemic [Internet]. New York: Newswire P; 2021 [updated 2021 Apr 28]. Available online:
https://www.floridatrend.com/article/31159/nursing-shortage-in-south-florida-has-healthcare-facilities-scrambling-to-counter-the-talent-drain - CMS. Medicare Program; CY 2022 Payment Policies Under the Physician Fee Schedule and Other Changes to Part B Payment Policies; Medicare Shared Savings Program Requirements; Provider Enrollment Regulation Updates; and Provider and Supplier Prepayment and Post-Payment Medical Review Requirements. 2021.
- Sciacca A. Hayward nursing home, parent company must pay almost $20 million for neglecting patients, jury says. SiliconValley.com: Bay Area News Group; 2021 [updated 10/15/2021; cited 2021 11/18/2021]. Available online: https://www.siliconvalley.com/2021/10/14/hayward-nursing-home-parent-company-must-pay-almost-20-million-for-neglecting-patients-jury-says/
- Harrington C, Hauser C, Olney B, et al. Ownership, financing, and management strategies of the ten largest for-profit nursing home chains in the United States. Int J Health Serv 2011;41:725-46. [Crossref] [PubMed]
- CMS. Medicare and Medicaid Programs; Minimum Staffing Standards for Long-Term Care Facilities and Medicaid Institutional Payment Transparency Reporting. 2024.
- CMS. Revised Guidance for Long-Term Care Facility Assessment Requirements 2024 [updated June 18, 2024]. Available online: https://www.cms.gov/files/document/qso-24-13-nh.pdf
Cite this article as: Shapley R Jr, Weech-Maldonado R, Davlyatov G, Orewa GN, Patterson J, Borkowski N. The effects of the Patient Driven Payment Model (PDPM) and COVID-19 on nursing and therapy staffing levels among skilled nursing facilities. J Hosp Manag Health Policy 2025;9:3.
