Content uploaded by Huiwen Xu
Author content
All content in this area was uploaded by Huiwen Xu on Aug 24, 2020
Content may be subject to copyright.
Journal Pre-proof
Shortages of Staff in Nursing Homes During COVID-19 Pandemic: What Are the
Driving Factors?
Huiwen Xu, PhD, Orna Intrator, PhD, John R. Bowblis, PhD
PII: S1525-8610(20)30691-5
DOI: https://doi.org/10.1016/j.jamda.2020.08.002
Reference: JMDA 3592
To appear in: Journal of the American Medical Directors Association
Received Date: 26 June 2020
Accepted Date: 6 August 2020
Please cite this article as: Xu H, Intrator O, Bowblis JR, Shortages of Staff in Nursing Homes During
COVID-19 Pandemic: What Are the Driving Factors?, Journal of the American Medical Directors
Association (2020), doi: https://doi.org/10.1016/j.jamda.2020.08.002.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Inc. on behalf of AMDA -- The Society for Post-Acute and Long-Term
Care Medicine.
Title: Shortages of Staff in Nursing Homes During COVID-19 Pandemic: What Are the Driving
Factors?
Huiwen Xu, PhD
1*
, Orna Intrator, PhD
2,3
, John R. Bowblis, PhD
4,5
(1) Department of Surgery, Cancer Control, University of Rochester School of Medicine and Dentistry,
Rochester, NY, USA
(2) Department of Public Health Sciences, University of Rochester School of Medicine and Dentistry,
Rochester, NY;
(3) Geriatrics & Extended Care Data Analysis Center (GECDAC), Canandaigua VA Medical Center,
Canandaigua, NY
(4) Department of Economics, Farmer School of Business, Miami University, Oxford, OH;
(5) Scripps Gerontology Center, Miami University, Oxford, OH;
*Corresponding author: Huiwen Xu, PhD, Department of Surgery, Cancer Control, University of
Rochester School of Medicine and Dentistry. 265 Crittenden Blvd., BOX 420658, Rochester, NY 14642,
USA. Phone: +1 585.275.2090; Fax: 585.461.4532; Email: Huiwen_Xu@URMC.Rochester.edu; Twitter:
@Dr_HuiwenXu.
Running title: Shortages of Staff in Nursing Homes during COVID-19
Key words: Staff Shortages, Personal Protection Equipment, COVID-19, Nursing Homes.
Funding sources: This research did not receive any funding from any agencies in the government or
private sectors.
Word, reference and graphics count:
• Word count: 2,999
• Number of text pages: 11
• References: 42
• Tables: 2
• Figure: 2
Brief Summary: Staff shortages in nursing homes are mainly driven by COVID-19 factors, such as
resident and staff with COVID-19, as well as PPE supply. Most nursing home and market factors, and
Medicaid reimbursement rate were insignificant.
Acknowledgements: No.
Conflict of Interests: John Bowblis owns Bowblis Economic Consulting, which provides consulting
services to long-term care providers. None of the material discussed in this paper are directly related to
these services.
Journal Pre-proof
1
Shortages of Staff in Nursing Homes During COVID-19 Pandemic: What Are the Driving Factors?
1
2
Abstract (300 words)
3
Objectives: During the COVID-19 pandemic, U.S. nursing homes (NHs) have been under
4
pressure to maintain staff levels with limited access to personal protection equipment (PPE).
5
This study examines the prevalence and factors associated with shortages of NH staff during
6
COVID-19 pandemic.
7
Design: We obtained self-reported information on staff shortages, resident and staff exposure
8
to COVID-19, and PPE availability from a survey conducted by the Centers for Medicare &
9
Medicaid Services in May 2020. Multivariate logistic regressions of staff shortages with state
10
fixed-effects were conducted to examine the effect of COVID-19 factors in NHs.
11
Setting and participants: 11,920 free-standing NHs.
12
Measures: The dependent variables were self-reported shortages of licensed nurse staff, nurse
13
aides, clinical staff, and other ancillary staff. We controlled for NH characteristics from the
14
most recent Nursing Home Compare and Certification And Survey Provider Enhanced Reporting,
15
market characteristics from Area Health Resources File, and state Medicaid reimbursement
16
calculated from Truven data.
17
Results: Of the 11,920 NHs, 15.9%, 18.4%, 2.5%, and 9.8% reported shortages of licensed nurse
18
staff, nurse aides, clinical staff, and other staff, respectively. Georgia and Minnesota reported
19
the highest rates of shortages in licensed nurse and nurse aides (both > 25%). Multivariate
20
regressions suggest that shortages in licensed nurses and nurse aides were more likely in NHs
21
having any resident with COVID-19 (adjusted odds ratio (AOR) = 1.44, 1.60, respectively) and
22
any staff with COVID-19 (AOR = 1.37, 1.34, respectively). Having one-week supply of PPE was
23
Journal Pre-proof
2
associated with lower probability of staff shortages. NHs with a higher proportion of Medicare
24
residents were less likely to experience shortages.
25
Conclusions/Implications: Abundant staff shortages were reported by NHs and were mainly
26
driven by COVID-19 factors. In the absence of appropriate staff, NHs may be unable to fulfill
27
the requirement of infection control even under the risk of increased monetary penalties.
28
Keywords: Staff Shortages, Personal Protection Equipment, COVID-19, Nursing Homes.
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Journal Pre-proof
3
Main Text (2,999 words)
43
Introduction
44
The epicenter of the COVID-19 pandemic in the United States (U.S.) has been in long-term care
45
facilities, particular nursing homes (NHs).
1
The first COVID-19 case in a NH was confirmed in a
46
Kirkland, Washington facility on February 28, 2020.
2
Since then, the Center for Medicare and
47
Medicaid Services (CMS) reported 107,389 confirmed cases and 71,278 suspected cases of
48
COVID-19 among residents based on self-reported data by NHs released in June 20, 2020.
3
NH
49
residents are extremely vulnerable to COVID-19 because they are older, functionally impaired,
50
and have multiple comorbidities
1, 4
This frail population thus bore over 27.5% of all confirmed
51
cases resulting in death.
3
In fact, a New York Times analysis claims that NH residents and
52
workers accounted for one third of COVID-19 death in the U.S.
5
53
A critical aspect of NH care is staff.
6-10
Prior to the pandemic, NH staff was the single
54
largest cost to operating a NH.
11
NHs must staff positions to provide direct care to residents
55
but also ancillary services, such as housekeeping and food service. Examples of staffing
56
categories include: licensed nurses [i.e. registered nurses (RNs) and licensed practical nurses
57
(LPNs)], nurse aides that assist licensed nurses and provide direct care to residents [certified
58
nurse aides (CNAs)], clinical staff (i.e. physicians and other advanced practice providers), and
59
other ancillary staff (e.g. recreation and food services).
12
Research suggests that NHs with
60
higher staffing levels tend to provide better quality of care,
4, 6-10, 12-16
but low wages, less-
61
desirable work environments compared to alternatives have made it difficult for NHs to hire
62
and retain staff.
10, 17-19
Reliance on government payment models, such as Medicaid which
63
Journal Pre-proof
4
reimburses at or below actual costs,
20
further limits NHs’ ability to increase wages or offer other
64
benefits to hire and retain staff.
65
These structural challenges have only become worse for NHs during the coronavirus
66
pandemic.
21, 22
NH workforce does not have the luxury of being able to social distance, as their
67
job requires close contact with the residents. At the early stage of the pandemic, NHs lacked
68
the life-saving personal protection equipment (PPE) to prevent the transmission within the
69
facility.
23-25
The shortage put staff at increased risk of contracting the virus, with staff
70
suspected of having contracted COVID-19 required to quarantine for at least 14-days. The net
71
result was existing NH staff were often sidelined. Yet, other factors also created pressure. NHs
72
needed to implement infection control protocols, including isolating residents who were
73
suspected of having the virus.
26
The ban on visitors to NHs also reduced the availability of some
74
informal care provided to residents by visiting relatives. This created a situation in which time
75
and effort needed from NHs staff increased, yet structural factors made it more difficult to
76
address,
8, 22, 25, 27
creating the potential for a staff shortage.
22
CMS acknowledged this shortage
77
by temporarily suspending the competency requirement for providing direct care to residents,
28
78
but the additional $600 per week federal unemployment benefit hurt the ability of NHs to
79
recruit needed staff.
22
80
The coronavirus pandemic has led to an urgent shortage of staff faced by NHs,
22, 25
yet
81
which facilities and what factors drove these shortages are not well understood. To mitigate
82
this knowledge gap, we analyzed the first-ever national COVID-19 nursing home staff data from
83
CMS to examine the staffing shortages in NHs. Understanding the potential predictors of
84
Journal Pre-proof
5
staffing shortages can help policy makers and NH administrators implement effective
85
interventions to combat staff shortages.
86
Methods
87
Data sources
88
We consolidated several publicly available datasets to create our analytic file. We
89
download the Nursing Home COVID-19 Public File (COVID-19 File) from the Nursing Home
90
Compare (NHCompare) website for COVID-19 related information, including detailed self-
91
reported data on the number of resident and staff COVID-19 cases, supply of PPE, and
92
shortages of staff. As of June 15, 2020, CMS published data of the weeks ending on May 24 and
93
31, 2020 for each certified NH and conducted the quality check of the data.
3
Of the 15,451 NHs
94
with data on May 31, 2020, 80.1% (12,375) passed the quality check. NHs that did not pass the
95
quality check tended to be smaller, for-profit, and with lower Five-star Ratings.
96
The COVID-19 File was merged with other data to obtain NH characteristics, particularly
97
the April 2020 monthly NHCompare archive database and the Certification And Survey Provider
98
Enhanced Reporting (CASPER). The NHCompare archive contains summary information about
99
each NH, including select measures of facility structure, nursing staff levels, and star ratings.
100
This information is updated regularly by CMS and contains the most recent publicly available
101
information regarding facilities. CASPER captures a snapshot of each facility’s payer-mix and
102
resident case-mix prior to the pandemic. CASPER includes data collected as part of initial and
103
annual recertification inspections of all Medicare and Medicaid certified nursing homes, with
104
these inspections occurring every 9 to 15 months. Because CASPER is available with a lag, we
105
utilized the most recent inspection for each facility that occurred from August 2018 through
106
Journal Pre-proof
6
October 2019 (with a median date of March 28, 2019). The 2010 Rural-Urban Commuting
107
Areas Codes (RUCAs) that incorporate information on both population size and commuting time
108
were downloaded to define the rurality of NHs.
29
County market factors were obtained from
109
the 2018-2019 Area Health Resources File.
30
Truven Health Analytics’ 2016 report on Medicaid
110
expenditures for NHs was used to estimate state Medicaid reimbursement rates.
15, 31
111
Study cohort
112
The primary analysis included all free-standing NHs with COVID-19 information in the
113
week of May 31, 2020, that could be merged with NHCompare and CASPER data, resulting in
114
11,920 unique NHs.
115
Dependent Variables
116
The dependent variables included whether the NH self-reported a shortage in staff
117
(yes/no) for the following type of staff: licensed nurse staff, nurse aides, clinical staff, and other
118
staff. Licensed nurse staff included RNs and LPNs. Nurse aides included the CNAs, nurse aides,
119
and medication aides/technicians. Clinical staff referred to physician, physician assistant,
120
advanced practice nurse. Finally, other staff included all staff not mentioned in the categories
121
above (e.g. ancillary services such as housekeeping).
122
Covariates
123
Covariates associated with potential shortages included COVID-19 factors, NH and
124
market characteristics, and state policy relating to NHs.
4, 9, 15, 32, 33
COVID-19 factors included
125
the cumulative number of residents and staff diagnosed with COVID-19 per 100 beds. We
126
scaled the number of cases to 100 beds to account for differences in facility size. PPE has been
127
shown to be very critical in preventing the transmission of COVID-19. We included three binary
128
Journal Pre-proof
7
variables indicating whether a NH had one-week supply of N95 masks, eye protection, and
129
gowns.
130
We used CASPER data to extract NH characteristics that might be associated with
131
staffing shortages: staffing levels (RNs, LPNs, and CNAs measured in hours per resident day
132
(HPRD)), NH structure (ownership, chain status, total beds, occupancy rate, and dementia
133
special care unit), resident case-mix and payer-mix (case-mix acuity index, % Medicaid residents,
134
and % Medicare residents), rurality, and NHCompare Overall Five-star Rating.
4, 15, 32, 34-36
Rurality
135
of NHs was determined from zip codes merged with 2010 RUCAs.
37
NHs were grouped into
136
urban, large rural city/town (micropolitan), and small rural town/ isolated small rural town
137
(rural).
38
138
We included the following factors that described the NH market identified as the county
139
in which the NH was located:
4
primary care physician per 1,000 population, concentration of
140
total NH beds measured by the Herfindahl–Hirschman Index (HHI),
39
Medicare Advantage
141
penetration rate (% Medicare Advantage of all Medicare beneficiaries in the county), median
142
household income ($), and % older population (≥ 65).
4, 15, 32-34
Medicaid reimbursements were
143
approximated by the ratio of a state’s total Medicaid expenditure on NHs derived from Truven
144
reports
31
divided by the total number of Medicaid bed days estimated from the number of NH
145
residents with Medicaid payer reported in CASPER data.
15
Finally, we included state effects to
146
control for unobserved fixed differences across states.
147
Statistical analysis
148
Descriptive analyses were conducted to show staff shortages, COVID-19 factors, NH
149
characteristics, market factors, and state policy. We also compared these factors by whether
150
Journal Pre-proof
8
the NHs had any staff with COVID-19 and tested the statistical significance of differences
151
between NHs with and without any staff with COVID-19 using t-tests for continuous variables
152
and Chi-square tests for binary variables. We then conducted four separate multivariate
153
logistic regressions to examine factors associated with shortages of staff with standard errors
154
clustered at county level, as many COVID-19 policies including reporting are county-based. We
155
dichotomized the residents and staff with COVID-19 at 1 to indicate whether the facility had
156
any confirmed COVID-19 for easy interpretation and to avoid potential bias in reported cases.
157
Overall five-star rating was categorized as 4 or 5 stars vs 1-3 stars as an indicator of high rating.
158
Continuous variables (except staffing and Medicaid rates) were standardized at overall means
159
and standard deviations to reduce variance and simplify the comparison of parameter
160
estimates.
33
The data of COVID-19 File in the week of May 24, 2020 were analyzed as
161
sensitivity analysis and shows similar results (not reported).
162
All statistical analyses were performed in SAS 9.4 (SAS Institute Inc., Cary, NC) and Stata
163
16.0 (StataCorp LLC, College Station, TX).
164
Results
165
Descriptive results by whether NHs had any staff with COVID-19 are presented in Table
166
1. Of the 11,920 NH sample, 15.9%, 18.4%, 2.5%, and 9.8% reported shortages of licensed
167
nursing staff, nurse aides, clinical staff, and other staff, respectively. On average, 5.7 (standard
168
deviation (SD) = 30.2) residents and 3.8 (SD = 18.9) staff per 100 beds were confirmed with
169
COVID-19; 82.2%, 89.8%, and 79.6% NHs had one-week supply of N95 masks, eye protection,
170
and gowns, respectively. The majority of NHs were for-profit (71.8%), chain-affiliated (60.6%),
171
with most residents paid by Medicaid (59.4%), and located in urban areas (66.3%). About one
172
Journal Pre-proof
9
half (46.3%) NHs had overall five-star rating ≥4. The average % Medicare Advantage
173
penetration was 31.6% and state on average reimbursed NHs $179.7 per resident day. Table 1
174
also suggests that almost all predictors were significantly different in NHs having any staff with
175
vs without COVID-19, except for one-week supply of gowns, CNA staffing level, and overall five-
176
star rating (all P ≤ 0.01). NHs having any staff with COVID-19 were more likely to experience
177
shortages of licensed nurse, nurse aides, clinical staff, and other staff.
178
Figure 1 presents the geographic variation of staff shortages in licensed nurse and nurse
179
aides. NHs in east and middle west states had a greater percentage of reported shortages, with
180
the following states reporting the highest rate of shortages in licensed nurse and nurse aides
181
(both >25%): District of Columbia, Georgia, Minnesota, and Rhode Island. Figure 2 suggests
182
that number of residents and staff with COVID-19 were highly correlated and also varied by
183
states. Connecticut, District of Columbia, Massachusetts, and New Jersey reported over 20
184
residents and 10 staff per 100 NH beds. Together, Figure 1 and 2 imply that states with higher
185
number of residents and staff with COVID-19 were more likely to report shortages in licensed
186
nurse and nurse aides.
187
Multivariate logistic regression results are shown in Table 2. NHs having any resident
188
with COVID-19 were more likely to experience shortages of nursing staff, nursing aides, clinical
189
staff, and other staff (adjusted odds ratio (AOR) =1.60, 1.44, 2.10, and 1.71, respectively; all P
190
<0.01). Similarly, NHs with any staff with COVID-19 were more likely to report all shortages of
191
all types of staff (AOR ranges 1.34-1.43; all P <0.01). Having one-week supply of eye protection
192
and gowns were associated with lower probability of staffing shortages.
193
Journal Pre-proof
10
Previous staffing levels were not associated with staffing shortages during the pandemic
194
(at the 5% level), except that NHs with higher RN staffing level were less likely to report
195
shortages in licensed nurse staff. Most NH structure factors were not significantly associated
196
with staff shortages, except for occupancy rates. NHs with higher occupancy rates were less
197
likely to have shortage in licensed nurse staff, nurse aides, and other staff. NHs with more
198
Medicare residents were less likely to have shortages in licensed nurse staff, nurse aides, and
199
other staff. No differences in staff shortages were found among NHs located in urban,
200
micropolitan, or rural areas. NHs with ≥4 overall ratings were less likely to report shortages in
201
licensed nurse staff, nurse aides, and other staff (AOR=0.79, 0.83, 0.85, respectively; all P <
202
0.01). Most market factors in the model were not associated with staffing shortages, except
203
market competition for shortage in other staff and % Medicare Advantage penetration for
204
shortage in licensed nurse staff. Finally, Medicaid reimbursement rates were not associated
205
with any shortage in staff.
206
Discussion
207
Using publicly available staffing data, we found that 16%-18% of NHs reported shortages
208
in licensed nurse staff and nurse aides during the coronavirus pandemic. These reported
209
shortages were not evenly distributed across states, with one out of four facilities in states like
210
Georgia and Minnesota reporting shortages in licensed nurses or aides. Those numbers are
211
concerning as licensed nurse and nurse aides are the essential workers who provide most of the
212
direct care to residents. Adequate staffing levels are required to provide high quality care to
213
residents.
7, 12
A recent paper reported that higher RN staffing levels prior to the pandemic was
214
associated with fewer COVID-19 cases in a sample of Connecticut NHs.
35
215
Journal Pre-proof
11
A major finding of our study is that staff shortages were associated with COVID-19
216
related factors. NHs having any resident or staff with COVID-19 were significantly more likely to
217
experience shortages of all types of staff, with resident cases of the virus having a stronger
218
effect on licensed nurse staff and clinical staff than nurse aides. This might be due to the fact
219
that COVID-19 residents require clinical care usually at the level of RNs or physician. Even in
220
May 2020, 20% of NHs did not have one-week supply of gowns, calling for help from federal
221
and state governments.
24, 25
NHs with one week supply of eye protection and gowns were less
222
likely to report staff shortages, reinforcing the importance of PPE on staff security. Importantly,
223
available supply of N95 masks, which are required only for closer procedure care, were not
224
related to shortages.
225
Findings support the expectation that NHs with higher staffing levels prior to the
226
pandemic might be less susceptible to shortages during the pandemic: better RN staffing was
227
related to less shortages in licensed nurses and nurse aides (marginally), but not to clinical staff
228
or other staff.
35
Higher CNAs prior to pandemic was marginally related only to less shortages in
229
nurse aides. Unexpectedly, higher CNAs was related to higher clinical staff shortages, possibly
230
indicating that those NHs were more likely to rely on post-acute care. Shortage of clinical staff is
231
different from shortage of front line staff. Clinical staff are often only present some days of the
232
week and many respond through phone or video calls.
233
NHs that took care of more post-acute, Medicare paid residents, were less likely to have
234
shortages in clinical staff, as did non-profit NHs. A reason for this may be states putting
235
temporary bans on elective surgeries, which led to reduced Medicare post-acute care stays.
236
While previous work suggests that COVID-19 cases were higher in urban areas,
7
we found no
237
Journal Pre-proof
12
difference in reported staff shortages in rural versus urban NHs. The lack of available workforce
238
in rural market prior to pandemic makes rural NHs more vulnerable to COVID-19, even they
239
were less likely to have staff with COVID-19 compared to urban NHs (21.8% vs 5.5%).
40
NHs
240
with overall star rating ≥4 were less likely to report staff shortages, suggesting they might be
241
more resilient to the pandemic.
35
242
Our results suggest that self-reported shortages in NH staffing are primarily associated
243
with COVID-19 related factors. However, NHs are still faced with multiple other challenges.
244
Media attention has put pressure on regulators to punish NHs given the large number of deaths
245
seen nationally. However, the structure of NH care, the fact that NH residents are frail and
246
more susceptible to the virus, and early miss-steps such as not providing NHs PPE when
247
supplies were scarce and sending coronavirus patients to NHs may have led to this situation.
248
CMS has recently increased the civil monetary penalties up to $20,000 per instance for non-
249
compliance with infection control.
41
This places great financial challenges on NHs, especially
250
considering that most NH care is reimbursed by Medicaid at lower than operating cost. Even
251
prior to the pandemic, NHs were reliant on higher margin Medicare residents to provide
252
financial cushion to invest in staff and quality.
42
Indeed, NHs with higher Medicare prevalence
253
were less likely to suffer staff shortages. Securing the financial health of NHs that allow them to
254
address these staff shortages needs to be a priority which might help NHs assure that fewer
255
residents are exposed to Covid-19.
256
While our study highlights staffing shortages in the NHs, we acknowledge several
257
limitations. Our findings may not be generalizable to all NHs, as 20% NHs did not pass CMS’
258
data quality check. Information regarding the COVID-19 factors and whether the NH had a staff
259
Journal Pre-proof
13
shortage are self-reported and may be inaccurate. Finally, the most up-to-date information
260
regarding facility characteristics are unavailable requiring us to rely on resident and payer-mix
261
characteristics from 2018-19, and Medicaid reimbursement rates from 2016.
262
Conclusions and Implications
263
About 1 out of 6 NHs self-reported having a shortage in licensed nurse and nurse aide
264
staffing during the COVID-19 pandemic. These shortages are not evenly distributed across
265
states. Staff shortages are mainly driven by COVID-19 factors, such as resident and staff with
266
COVID-19, as well as PPE supply. Policymakers should further support NHs to prevent the
267
transmission of COVID-19 among their vulnerable residents and valuable workforce, and help
268
them acquire sufficient PPE. Current policy efforts that focus on preventing the spread of the
269
infection within and across NHs include (dis)incentives such as large fines might be counter-
270
productive. Monetary penalties might motivate NHs to avoid violation of infection control, but
271
without funds to hire and retain staff, NHs lack the capacity to fulfill the requirement. The
272
availability of high quality direct care workers becomes even more critical as many states are
273
reopening the economy and lifting bans on visitors to NHs.
274
275
References
276
1. Kamp J, AW, M. As U.S. Nursing-Home Deaths Reach 50,000, States Ease Lockdowns.
277
The Wall Street Journal. 2020.
278
2. McMichael, TM, Currie, DW, Clark, S, et al. Epidemiology of Covid-19 in a long-term care
279
facility in King County, Washington. N Engl J Med 2020;382(21):2005-2011.
280
Journal Pre-proof
14
3. Center for Medicare and Medicaid Services. Nursing Home COVID-19 Public File 2020.
281
https://data.cms.gov/stories/s/COVID-19-Nursing-Home-Data/bkwz-xpvg. Accessed
282
6/10 2020.
283
4. Xu, H, Bowblis, JR, Li, Y, et al. Construction and Validation of Risk-adjusted Rates of
284
Emergency Department Visits for Long-stay Nursing Home Residents. Med Care
285
2020;58(2):174-182.
286
5. Yourish, K, Lai, KR, Ivory, D, et al. One-third of all US coronavirus deaths are nursing
287
home residents or workers. New York Times 2020.
288
6. Harrington, C, Zimmerman, D, Karon, SL, et al. Nursing Home Staffing and Its
289
Relationship to Deficiencies. The Journals of Gerontology: Series B 2000;55(5):S278-
290
S287.
291
7. Min, A, Hong, HC. Effect of nurse staffing on rehospitalizations and emergency
292
department visits among short-stay nursing home residents: A Cross-sectional study
293
using the US Nursing Home Compare database. Geriatric Nursing 2018.
294
8. Charlene Harrington, Helen Carrillo, Garfield, aR. Nursing Facilities, Staffing, Residents
295
and Facility Deficiencies, 2009 Through 2014. Kaiser Family Foundation; 2015.
296
9. Backhaus, R, Verbeek, H, van Rossum, E, et al. Nurse Staffing Impact on Quality of Care
297
in Nursing Homes: A Systematic Review of Longitudinal Studies. J Am Med Dir Assoc
298
2014;15(6):383-393.
299
10. Harrington, C, Swan, JH. Nursing Home Staffing, Turnover, and Case Mix. Med Care Res
300
Rev 2003;60(3):366-392.
301
11. SV, M. Analysis of Nursing Home Costs. 1998.
302
Journal Pre-proof
15
12. Bowblis, JR, Roberts, AR. Cost-Effective Adjustments to Nursing Home Staffing to
303
Improve Quality. Med Care Res Rev 2018;77(3):274-284.
304
13. Saliba, D, Weimer, DL, Shi, Y, et al. Examination of the New Short-Stay Nursing Home
305
Quality Measures: Rehospitalizations, Emergency Department Visits, and Successful
306
Returns to the Community. Inquiry 2018;55:46958018786816.
307
14. Kim, H, Harrington, C, Greene, WH. Registered Nurse Staffing Mix and Quality of Care in
308
Nursing Homes: A Longitudinal Analysis. The Gerontologist 2009;49(1):81-90.
309
15. Xu, H, Intrator, O. Medicaid Long-term Care Policies and Rates of Nursing Home
310
Successful Discharge to Community. J Am Med Dir Assoc 2020;21(2):248-253 e241.
311
16. Antwi, YA, Bowblis, JR. The Impact of Nurse Turnover on Quality of Care and Mortality in
312
Nursing Homes: Evidence from the Great Recession. American Journal of Health
313
Economics 2017;0(ja):1-50.
314
17. Lapane, KL, Hughes, CM. Considering the employee point of view: perceptions of job
315
satisfaction and stress among nursing staff in nursing homes. J Am Med Dir Assoc
316
2007;8(1):8-13.
317
18. Bourbonniere, M, Feng, Z, Intrator, O, et al. The use of contract licensed nursing staff in
318
US nursing homes. Med Care Res Rev 2006;63(1):88-109.
319
19. Harrington, C, Edelman, TS. Failure to Meet Nurse Staffing Standards: A Litigation Case
320
Study of a Large US Nursing Home Chain. Inquiry 2018;55:46958018788686.
321
20. Eljay, L. A report on shortfalls in Medicaid funding for nursing center care. Report for the
322
American Health Care Association, December 2016.
323
Journal Pre-proof
16
21. Chris Kirkham, Lesser, B. Pandemic exposes systemic staffing problems at U.S. nursing
324
homes. Reuters. 2020.
325
22. Stateline, S. Staffing nursing homes was hard before the pandemic. Now it's even
326
tougher. Fierce Healthcare. 2020.
327
23. Livingston, E, Desai, A, Berkwits, M. Sourcing Personal Protective Equipment During the
328
COVID-19 Pandemic. JAMA 2020;323(19):1912-1914.
329
24. Jewett, C. Nursing Homes Run Short Of COVID-19 Protective Gear As Federal Response
330
Falters. Kaiser Health News. 2020.
331
25. Abbasi, J. “Abandoned” Nursing Homes Continue to Face Critical Supply and Staff
332
Shortages as COVID-19 Toll Has Mounted. JAMA 2020.
333
26. Center for Medicare and Medicaid Services. Long-Term Care Nursing Homes Telehealth
334
and Telemedicine Tool Kit 2020.
335
27. Spanko, A. Nursing Home Staffing Strain Could Hit ‘High Point’ This Week as COVID-19
336
Crisis Rages. Skilled Nursing News. 2020.
337
28. Center for Medicare and Medicaid Services. Nursing Home Five Star Quality Rating
338
System updates, Nursing Home Staff Counts, and Frequently Asked Questions. 2020.
339
29. Hart, G. Rural-Urban Commuting Areas Codes (RUCAs) 2010; 2014.
340
https://ruralhealth.und.edu/ruca. Accessed August 1 2018.
341
30. US Department of Health and Human Services. Area health resource file. . 2019: .
342
31. Steve Eiken, Kate Sredl, Brian Burwell, et al. Medicaid Expenditures for Long-Term
343
Services and Supports (LTSS) in FY 2016. 2018.
344
Journal Pre-proof
17
32. Grabowski, DC, Stewart, KA, Broderick, SM, et al. Predictors of nursing home
345
hospitalization: a review of the literature. Med Care Res Rev 2008;65(1):3-39.
346
33. Intrator, O, Grabowski, DC, Zinn, J, et al. Hospitalization of nursing home residents: the
347
effects of states' Medicaid payment and bed-hold policies. Health Serv Res
348
2007;42(4):1651-1671.
349
34. Freeman, S, Bishop, K, Spirgiene, L, et al. Factors affecting residents transition from long
350
term care facilities to the community: a scoping review. BMC Health Serv Res
351
2017;17(1):689.
352
35. Li, Y, Temkin-Greener, H, Gao, S, et al. COVID-19 infections and deaths among
353
Connecticut nursing home residents: facility correlates. J Am Geriatr Soc 2020;n/a(n/a).
354
36. Abrams, HR, Loomer, L, Gandhi, A, et al. Characteristics of U.S. Nursing Homes with
355
COVID-19 Cases. J Am Geriatr Soc 2020;n/a(n/a).
356
37. Hart, LG, Larson, EH, Lishner, DM. Rural definitions for health policy and research. Am J
357
Public Health 2005;95(7):1149-1155.
358
38. WWAMI Rural Health Research Center. Using RUCA Data; 2017.
359
http://depts.washington.edu/uwruca/ruca-uses.php. Accessed 11/26 2017.
360
39. Bowblis, JR, North, P. Geographic Market Definition: The Case of Medicare-Reimbursed
361
Skilled Nursing Facility Care. INQUIRY: The Journal of Health Care Organization,
362
Provision, and Financing 2011;48(2):138-154.
363
40. Roberts, AR, Bowblis, JR. How Does Rurality Influence the Staffing of Social Service
364
Departments in Nursing Homes? Gerontologist 2018;58(3):540-545.
365
Journal Pre-proof
18
41. Centers for Medicare & Medicaid Services. COVID-19 Survey Activities, CARES Act
366
Funding, Enhanced Enforcement for Infection Control deficiencies, and Quality
367
Improvement Activities in Nursing Homes. Baltimore, Maryland; 2020.
368
42. Weech-Maldonado, R, Pradhan, R, Dayama, N, et al. Nursing Home Quality and Financial
369
Performance: Is There a Business Case for Quality? Inquiry 2019;56:46958018825191.
370
371
372
373
374
375
376
377
378
379
380
381
382
383
Journal Pre-proof
19
List of Tables and Figures
384
Table 1. Nursing Home Characteristics, Market Factors, and State Policy
385
Table 2. Multivariate Logistic Regressions Models Examining Factors Associated with Reported
386
Shortages of Licensed Nurse, Nurse Aide, Clinical and Other Staff in Nursing Homes on May 31,
387
2020
388
Figure 1. Percentage of Nursing Homes Reporting Staff Shortages by State
389
Figure 2. Prevalence of Nursing Home Residents and Staff with COVID-19
390
Journal Pre-proof
Table 1. Nursing Home Characteristics, Market Factors, and State Policy
Variables
All Nursing Homes
(N=11,920)
Staff with COVID
-
19 (N=4,466)
Staff without
COVID-19 (N=7,454)
P
value
M
ean
(SD)
or N(%)
M
ean
(SD)
or N(%)
M
ean
(SD)
or N(%)
Outcome Measures
Shortage of
Licensed
Nurs
e
Staff (RN+LPN)
1
,
897 (15.9%)
877 (19.6%)
1
,
020 (13.7%)
<0.01
Shortage of
Nurse Aides
ǂ
2
,
189 (18.4%)
996 (22.3%)
1
,
193 (16.0%)
<0.01
Shortage of
Clinical Staff (MD+NP+PA)
301 (2.5%)
161 (3.6%)
140 (1.9%)
<0.01
Shortage of
Other Staff
1
,
170 (9.8%)
567 (12.7%)
603 (8.1%)
<0.01
COVID
-
19
F
actors
Total
Residents with
COVID
-
19
per 100 beds
5.68 (30.24)
13.86 (35.58)
0.79 (25.31)
<0.01
Total
Staff with
COVID
-
19
per 100 beds
3.79
(18.86)
10.13 (29.77)
0.00 (0.00)
<0.01
H
as
One
-
Week Supply of N95 Masks
9
,
796 (82.2%)
3
,
727 (83.5%)
6
,
069 (81.4%)
<
0.0
1
H
as
One
-
Week Supply of Eye Protection
1
,
0705 (89.8%)
4
,
088 (91.5%)
6
,
617 (88.8%)
<0.01
H
as
One
-
Week Supply of Gowns
9
,
486
(79.6%)
3
,
559 (79.7%)
5
,
927 (79.5%)
0.82
Nursing home characteristics
Staffing
RN Staffing
Level (
HPRD
)
0.66 (0.42)
0.68 (0.42)
0.65 (0.42)
<0.01
LPN Staffing
Level (
HPRD
)
0.86 (0.33)
0.88 (0.32)
0.85 (0.34)
<0.01
CNA
Staffing Level (
HPRD
)
2.29 (0.53)
2.28 (0.54)
2.29 (0.52)
0.33
Structure
Ownership
For
-
profit
8
,
561 (71.8%)
3
,
190 (71.4%)
5
,
371 (72.1%)
<0.01
Government
647 (5.4%)
192 (4.3%)
455 (6.1%)
No
t
-
for
-
profit
2
,
712
(22.8%)
1
,
084 (24.3%)
1
,
628 (21.8%)
Chain Affiliated
7
,
205 (60.6%)
2
,
541 (57.0%)
4
,
664 (62.7%)
<0.01
Number of Beds
108.47 (58.65)
129.75 (72.16)
95.73 (44.12)
<0.01
Occupancy Rate (0
-
100)
78.91 (16.55)
81.61 (15.00)
77.30 (17.22)
<0.01
Dementia Special Care Unit
1
,
688 (14.2%)
694 (15.6%)
994 (13.4%)
<0.01
Resident and
P
ay
er M
ix
Case
-
mix Acuity
I
ndex
10.42 (1.35)
10.56 (1.32)
10.34 (1.36)
<0.01
% Medicaid Paid (0
-
100)
59.44 (23.45)
58.75 (24.49)
59.84 (22.80)
0.01
% Medicare Paid (0
-
100)
13.11 (12.88)
13.85 (12.99)
12.66 (12.79)
<0.01
Rurality
Urban
7
,
871 (66.3%)
3
,
698 (83.1%)
4
,
173 (56.2%)
<0.01
Micropolitan
1
,
703 (14.3%)
384 (8.6%)
1
,
319 (17.8%)
Journal Pre-proof
Rural
2
,
300 (19.4%)
368 (8.3%)
1
,
932 (26.0%)
Overall Five
-
star Rating
≥4
5,519 (46.3%)
2,105 (47.1%)
3,414 (45.8%)
0.16
Market Factors
at County Level
Primary
Care Physician
per 1000
P
opulation
0.60 (1.21)
0.93 (1.49)
0.40 (0.96)
<0.01
Competitive
M
arket (HHI < 0.15)
8
,
872 (74.4%)
3
,
803 (85.2%)
5
,
069 (68.0%)
<0.01
% Medicare Advantage Penetration
(0
-
100)
31.63 (13.75)
32.67 (13.13)
31.00 (14.08)
<0.01
Median Household Income ($)
58
,
206.04
(15,797.69)
63
,
264.92
(17,467.63)
55
,
139.06
(13,820.35) <0.01
% Older population
(
≥ 65) (0
-
100)
16.90 (4.02)
15.95 (3.43)
17.47 (4.24)
<0.01
State Policy
Medicaid
Reimbursement Rates
179.74 (52.43)
185.89 (57.38)
176.06 (48.87)
<0.01
Note: RN= Registered Nurse, LPN= Licensed practical nurse, HPRD= hour per resident day, MD=physician, NP= Nurse Practitioner,
PA=Physician Assistant, CNA =certified nursing assistant, Micropolitan= Large Rural City/Town, Rural= Small Rural Town/ Isolated
Small Rural Town, HHI= Herfindahl-Hirschman Index;
+
P values measures whether nursing homes of staff with vs without COVID-19 had the same characteristics using t-tests for
continuous variables, and Chi-square tests for binary variables;
ǂ
Nurse aides included the certified nursing assistant, nurse aide, medication aide, and medication technician.
Data sources included the COVID-19 Nursing Home Dataset for COVID-19 related information, Nursing Home Compare Data (April
2020) for facility characteristics, Certification and Survey Provider Enhanced Reporting (2018-2019) for NH characteristics, and Area
Health Resources File (2018-2019) for market factors.
Journal Pre-proof
Table 2. Multivariate Logistic Regressions Models Examining Factors Associated with Reported Shortages of Licensed Nurse, Nurse
Aide, Clinical and Other Staff in Nursing Homes on May 31, 2020
Variables Licensed Nurse Staff Nurse Aides Clinical Staff Other Staff
OR (95 CI)
OR (95 CI)
OR (95 CI)
OR (95 CI)
COVID-19 Factors
Any resident with COVID-19 1.60*** (1.38 - 1.85) 1.44*** (1.25 - 1.67) 2.10*** (1.54 - 2.87) 1.71*** (1.43 - 2.05)
Any staff with COVID-19 1.37*** (1.19 - 1.58) 1.34*** (1.17 - 1.53) 1.43** (1.05 - 1.94) 1.38*** (1.17 - 1.64)
Has One-Week Supply of N95 Masks 1.14 (0.90 - 1.43) 1.02 (0.84 - 1.25) 1.25 (0.87 - 1.79) 0.98 (0.78 - 1.24)
Has
One
-
Week Supply of Eye
Protection 0.70*** (0.55 - 0.89) 0.64*** (0.52 - 0.79) 0.46*** (0.31 - 0.69) 0.78* (0.61 - 1.01)
Has One-Week Supply of Gowns 0.53*** (0.44 - 0.64) 0.55*** (0.47 - 0.65) 0.77 (0.54 - 1.10) 0.57*** (0.47 - 0.70)
Nursing Home Staffing
RN Staffing Level (HPRD) 0.66*** (0.49 - 0.89) 0.80* (0.62 - 1.04) 1.18 (0.70 - 1.99) 0.81 (0.58 - 1.13)
LPN Staffing Level (HPRD) 0.91 (0.72 - 1.15) 0.97 (0.78 - 1.22) 0.97 (0.61 - 1.55) 0.94 (0.69 - 1.28)
CNA Staffing Level (HPRD) 1.00 (0.87 - 1.14) 0.88* (0.77 - 1.00) 1.34* (0.98 - 1.84) 1.00 (0.85 - 1.19)
Nursing Home Structure
Ownership (Ref: For-profit)
Government 1.31* (0.99 - 1.72) 1.20 (0.92 - 1.55) 1.07 (0.58 - 1.97) 1.35* (0.99 - 1.85)
Not-for-profit 1.00 (0.86 - 1.18) 1.08 (0.93 - 1.26) 0.67** (0.45 - 0.99) 0.99 (0.82 - 1.19)
Chain Affiliated 0.96 (0.85 - 1.08) 0.90* (0.80 - 1.00) 1.25 (0.96 - 1.62) 0.91 (0.78 - 1.06)
Number of beds
+
0.98 (0.91 - 1.05) 1.01 (0.94 - 1.08) 0.92 (0.78 - 1.08) 0.91** (0.83 - 1.00)
Occupancy Rate
+
0.86*** (0.80 - 0.92) 0.91*** (0.85 - 0.98) 0.98 (0.82 - 1.17) 0.90** (0.82 - 0.99)
Dementia Special Care Unit 1.08 (0.92 - 1.25) 1.07 (0.92 - 1.24) 1.29 (0.91 - 1.85) 1.22* (1.00 - 1.49)
Nursing Home Resident and Payer
Mix
% Medicaid Paid
+
1.00 (0.92 - 1.10) 1.06 (0.98 - 1.15) 0.97 (0.81 - 1.16) 1.09* (0.98 - 1.21)
% Medicare Paid
+
0.79*** (0.70 - 0.90) 0.82*** (0.74 - 0.91) 0.81* (0.65 - 1.01) 0.80*** (0.70 - 0.92)
Case-mix Acuity Index
+
1.01 (0.93 - 1.10) 0.99 (0.92 - 1.07) 1.02 (0.87 - 1.19) 1.02 (0.94 - 1.12)
Rurality (Ref: Urban)
Micropolitan 0.96 (0.79 - 1.16) 0.83* (0.70 - 1.00) 0.95 (0.62 - 1.48) 0.90 (0.71 - 1.15)
Rural 1.11 (0.90 - 1.37) 0.96 (0.78 - 1.17) 0.98 (0.60 - 1.58) 0.97 (0.75 - 1.26)
Overall Five-star Rating ≥4 0.79*** (0.70 - 0.89) 0.83*** (0.74 - 0.93) 0.89 (0.68 - 1.16) 0.85** (0.73 - 0.98)
Market Factors
Primary Care Physician
+
0.92* (0.85 - 1.01) 0.98 (0.90 - 1.06) 1.04 (0.91 - 1.19) 1.01 (0.90 - 1.14)
Competitive Market (HHI < 0.15) 1.01 (0.85 - 1.19) 0.93 (0.79 - 1.09) 0.86 (0.58 - 1.29) 0.75*** (0.62 - 0.92)
Journal Pre-proof
% Medicare Advantage Penetration
+
1.11** (1.00 - 1.22) 1.07 (0.97 - 1.17) 1.02 (0.84 - 1.24) 1.02 (0.90 - 1.17)
Median Household Income ($)
+
0.96 (0.88 - 1.03) 0.94 (0.87 - 1.02) 0.99 (0.84 - 1.15) 1.05 (0.95 - 1.15)
% Older population (≥ 65)
+
1.04 (0.96 - 1.13) 1.04 (0.97 - 1.12) 1.07 (0.93 - 1.24) 1.08 (0.98 - 1.18)
Medicaid Reimbursement Rates 0.97 (0.90 - 1.03) 0.97 (0.91 - 1.04) 1.04 (0.86 - 1.27) 1.00 (0.92 - 1.09)
Observations 10,870 10,928 10,666 10,859
Note: OR= odds ratio, CI= confidence interval, RN= Registered Nurse, LPN= Licensed practical nurse, CNA =Certified nursing assistant,
HPRD= hour per resident day, Micropolitan= Large Rural City/Town, Rural= Small Rural Town/ Isolated Small Rural Town, HHI=
Herfindahl-Hirschman Index;
Standard errors were clustered at county level;
State fixed effects were not presented;
*** p<0.01, ** p<0.05, * p<0.1;
+
Continuous variables were standardized with a mean of zero and a standard deviation of one.
Journal Pre-proof
Figure 1. Percentage of Nursing
Homes Reporting Staff Shortages by
State
Notes:
+
Licensed nurse staff included the registered nurse, licensed practical nurse, and vocational nurse
as reported by the provider;
ǂ
Nurse aides included the certified nursing assistant, nurse aide, medication aide, and medication
technician as reported by the provider.
Panel B.
Reported
S
hortage of
Nurse Aides
ǂ
Panel A.
Reported
Shortage of
Licensed
Nurs
e
Staff
+
Journal Pre-proof
Journal Pre-proof
Figure 2. Prevalence of Nursing Home Residents and Staff with COVID-19
Note:
Number of cases were scaled to 100 beds to account for differences in nursing home size.
Panel A. Total Number of Residents with COVID-19 per 100 Beds
Panel
B
. Total
N
umber of
S
taff with
COVID
-
19
per 100
B
eds
Journal Pre-proof
Journal Pre-proof
Journal Pre-proof
Journal Pre-proof
Journal Pre-proof
