Description
Read the article titled: COVID-19 Infections, PandemicRelated Social and Economic Impacts,
and Changes to Mental and Self-Rated
Health Among Latinx Immigrant
Housecleaners in New York City:
The Safe and Just Cleaners Study. The article is attached. The rubric to follow i attached. The questions to fill out is attached,
Unformatted Attachment Preview
RESEARCH & ANALYSIS
COVID-19 Infections, PandemicRelated Social and Economic Impacts,
and Changes to Mental and Self-Rated
Health Among Latinx Immigrant
Housecleaners in New York City:
The Safe and Just Cleaners Study
Sherry Baron, MD, MPH, Isabel Cuervo, PhD, Dhwanil Shah, MPH, Ana Gonzalez, BA, Homero Harari, ScD, and Deysi Flores, BA
Objectives. To estimate impacts of COVID-19 infections and social and economic sequelae on mental
AJPH
and self-rated health among Latinx immigrant housecleaners in New York City.
housecleaners initially surveyed before the pandemic between August 2019 and February 2020. We
measured rates of self-reported COVID-19 infections, COVID-19 antibodies, and pandemic-related social
and economic sequelae and examined predictors of mental and self-rated health changes using logistic
regression models.
Results. Fifty-three percent reported COVID-19 infections, consistent with the rate demonstrating
August 2023, Vol 113, No. 8
Methods. From March to June 2021, we conducted a follow-up study with 74% retention of 402
COVID-19 antibodies. During shutdown of nonessential services, from March 22 to June 8, 2020,
29% worked as housecleaners, although this was not associated with higher COVID-19 infection rates.
COVID-19–related stigma at work, lost earnings owing to COVID-19 infections, housing insecurity, food
insecurity, and unsafe homes, including experiencing intimate partner verbal abuse, were statistically
associated with changes in mental or self-rated health compared with prepandemic measures.
Conclusions. The disproportionate impact and virtually nonexistent safety net housecleaners
experienced during the first year of the pandemic highlight the importance of inclusive stopgap
measures to mitigate economic insecurity and its sequelae. (Am J Public Health. 2023;113(8):893–903.
https://doi.org/10.2105/AJPH.2023.307324)
I
nequities in the distribution of
families, especially immigrants and
rigorously and effectively monitors and
COVID-19 infection–related morbidity
others who were excluded from unem-
intervenes in upstream social and eco-
and mortality by race, ethnicity, and
ployment compensation, stimulus
nomic factors linked to health
economic status are well-documented.1
payments, and other health and social
inequities.3
Also important to understand is the dis-
assistance programs.2 Understanding
parate effects of the pandemic’s social
these factors can contribute to better
distribution of pandemic-related eco-
and economic sequelae on health. Loss
public health planning consistent with
nomic and social impacts on health,
of earnings were potentially significant
public health leaders’ calls for a new
the National Institutes of Health funded
for many low-income workers and their
Public Health 3.0 model that more
the COVID-19 Social, Behavioral, and
To better characterize the extent and
Research
Peer Reviewed
Baron et al.
893
RESEARCH & ANALYSIS
Economic (SBE) Impacts initiative, which
and street outreach in 4 of 5 boroughs
employment characteristics before the
provided supplemental funding to
of NYC and 2 suburban communities.
pandemic as those who participated in
ongoing health studies, prioritizing inves-
The prepandemic survey was interview-
the follow-up study except nonpartici-
tigators actively engaged with dispropor-
er administered in person in Spanish,
pants were more likely to report they
4
tionately affected populations. We
and 70% of those who had expressed
were the family primary wage earner
report on findings from a COVID-19 SBE
interest in the study completed the sur-
before the pandemic (59% of nonparti-
study that targeted Latinx immigrant
vey between July 2019 and February
cipants vs 39% of participants; x2
housecleaners in New York City (NYC)
2020. The prepandemic survey partici-
P < .01).
enrolled in the Safe and Just Cleaners
pants were similar to the estimated
Study,5 an ongoing community-based
343 527 housecleaners in the United
participatory research study exploring
States in 2019. Participants were 99%
the role of working conditions and expo-
female, average age was 44 years, all
We measured participants’ COVID-19
sures to household cleaning products
were foreign-born and had lived in the
status using self-report because it was
on the health of housecleaners, one of
United States on average 15 years, and
difficult to access medical care or test-
the most common occupations for
only 14% reported feeling comfortable
ing in the early months of the pandemic
documented and undocumented Latinx
with spoken English. Most were self-
in NYC,9 especially because 49% of the
women.6
employed, worked an average of
cohort reported not having health in-
22 hours per week for an average of
surance on their prepandemic survey.5
survey before the pandemic, between
3 clients, and earned less than $18 000
We measured whether respondents
July 2019 and February 2020, to 402
per year; 44% were the primary wage
ever had COVID-19 with the survey
Spanish-speaking housecleaners in the
earners for their families.5
question “Do you think that you had
AJPH
August 2023, Vol 113, No. 8
We had recruited and administered a
For this follow-up COVID-19 study, we
COVID-19?” or whether they reported
that survey, exploring pathways
attempted to recontact the 402 house-
ever having a positive COVID-19 nasal
through which the housecleaners’ em-
cleaners who participated in the pre-
swab test. For those reporting they had
ployment and working conditions af-
pandemic survey at least 3 times by
had COVID-19, we asked whether they
fected their mental and self-rated
telephone and text messaging in Span-
had been hospitalized. We also asked
health, were previously reported. In
ish to invite them to participate. For
the number of household members
this follow-up study, conducted be-
those who agreed to participate, a tele-
who ever had COVID-19, whether family
tween March and June 2021, we resur-
phone survey lasting approximately
members or close friends had died of
veyed our participants to (1) document
30 minutes was administered in Span-
COVID-19, and whether the respondent
housecleaners’ experience with
ish between March 18 and June 11,
had received or planned to receive a
COVID-19 infections and SBE effects
2021, by bilingual or native Spanish
COVID-19 vaccination.
during the pandemic, (2) measure
speakers trained in survey administra-
changes in the housecleaners’ mental
tion techniques. We collected and man-
whether participants worked as a
and self-rated health compared with
aged data using the REDCap tool
housecleaner during the citywide shut-
findings from our prepandemic survey,
hosted at the Icahn School of Medicine
down for nonessential work (March
and (3) explore how housecleaners’
at Mount Sinai,8 and participants were
22–June 8, 2020), their pay during that
experiences with COVID-19 infections
offered a $30 incentive.
period, and whether they returned to
NYC metropolitan area. Results from
5
Employment-related items captured
Of the 402 housecleaners who partic-
housecleaning work after the shut-
sured changes in the participants’ men-
ipated in the prepandemic survey, 296
down. We drew other SBE measures
tal and self-rated health.
(74%) participated in the COVID-19
from the COVID-19 resources in the
follow-up survey, 27 (7%) were not in-
PhenX Toolkit,10 including COVID-19–
terested, 24 (6%) had disconnected
related stigma at work, defined as
and SBE effects might predict mea-
METHODS
894
Survey Measures
7
telephones, and 55 (14%) did not re-
“feeling afraid or embarrassed to tell an
In 2019 we recruited the initial cohort of
spond. The housecleaners who did not
employer if they were to have COVID-19”
402 housecleaners through partnerships
participate in the COVID-19 follow-up
and a 2-item domestic insecurity mea-
with community-based organizations
survey had similar demographic and
sure asking whether, since the beginning
Research
Peer Reviewed
Baron et al.
RESEARCH & ANALYSIS
of the pandemic, the participant felt safe
Platelia test for IgA, IgM, and IgG, which
variables significant at the a 5 0.05
at home and whether they experienced
targets the SARS-CoV-2 (severe acute
level. We estimated adjusted odds
intimate partner verbal abuse. For hous-
respiratory syndrome coronavirus 2)
ratios (AORs) along with 95% confi-
ing insecurity, we asked about the stabili-
nucleocapsid protein (manufacturer
dence intervals (CIs). We dichotomized
ty of their current housing situation and
sensitivity 5 98.0%; specificity 5 99.3%).
predictor variables unless otherwise
the amount of back rent they owed. To
The study laboratory also validated this
specified. To differentiate the magni-
assess food insecurity over the previous
assay for use with DBS, which found
tude of food insecurity, we created an
12 months, we used the 2-item food in-
100% sensitivity and 100% specificity.18
ordinal scale (0–4) by summing the 2
security screening tool, which has good
Participants received letters from a
responses, with “sometimes” experienc-
sensitivity, specificity, and accuracy in
study physician explaining their results.
ing food insecurity contributing 1 point
and “often” contributing 2 points for
adults, including Spanish speakers and
low-income respondents.11
Data Analysis
each of the questions. We excluded
To measure changes in mental and
participants with missing information
self-rated health, we repeated the mea-
After conducting descriptive analyses, we
on a variable only from analyses in-
sures from our prepandemic survey:
compared demographic, employment-
volving that variable. We conducted
the Center for Epidemiologic Studies
related, and social characteristics of those
all analyses using SAS version 9.2 (SAS
Depression 10-item scale (CES-D-10),12
who reported having had COVID-19 to
Institute, Cary, NC).
Cohen’s Perceived Stress 10-item scale
these characteristics of those who did
(PSS),13 and the single-item self-rated
not by using the x2 test for categorical
overall health scale,14 using the Spanish
variables or the t test for continuous
validated versions of each measure.15
variables. We next examined predictors
AJPH
RESULTS
At the time of the survey, 153 partici-
stress (PSS), and self-rated health using
pants (51.7%) reported that they had
separate logistic regression models for
ever had COVID-19, and of those, 9
each outcome measure. To dichoto-
(5.9%) reported being hospitalized. The
We assessed the validity of COVID-19
mize these outcome measures, we
seasonal distribution of cases (spring
self-reports using an at-home self-
used the same cutpoints as in our pre-
2020: 55%; summer and fall 2020: 16%;
collected dried blood spot (DBS) kit,
pandemic study.5 For the CES-D-10, we
and winter 2021: 29%) mirrored the
replicating the procedures used in an-
used the recommended cutpoint
overall pattern in NYC.21 Among those
other national study.16 DBS kits were
(≥ 10).19 For the PSS, we used a cutpoint
sent from and returned to the study
20
suggested for health screenings (≥ 14)
whether they had COVID-19, 5 partici-
laboratory (Molecular Testing Laborato-
and consistent with the mean score
pants (1.7%) reported a positive nasal
ries, Vancouver, WA) via the US Postal
found in the Hispanic Community
swab test. Of the 296 participants, 218
COVID-19
Serology Measures
who did not think or did not know
Service in self-addressed, stamped
Health Study.
For self-rated health, we
(74%) consented to receive a DBS kit,
envelopes containing a biohazard bag.
compared those reporting poor or fair
and 116 (53% of those receiving a kit
A Spanish-language video, developed
health to those reporting good, very
and 39% of survey participants)
and validated by another research
good, or excellent health.
returned their kit. The rate of self-
17
team
and customized with an intro-
15
Our modeling used a stepwise
reported COVID-19 was similar be-
duction by our study’s outreach worker,
elimination method in multiple stages
tween those who returned the DBS kit
demonstrated procedures to complete
beginning with demographic and
and the others (Table 1). Of the 116
the home test. We reminded partici-
COVID-19–related variables, then
returned samples, only 5 (4%) were
pants at least 3 times to return their
employment-related variables, then
indeterminant and 66 (57%) were posi-
kits and offered an additional $20 in-
other SBE variables. At each step, we
tive for antibodies. Among those with
centive when kits were returned.
retained variables significant at a 5 0.2.
positive antibodies, 18% reported not
In the final step, we added the baseline
having had COVID-19, and 27% of
cimens for total antibodies using the
prepandemic value of the health out-
those reporting having had COVID-19
Bio-Rad Laboratories (Hercules, CA)
come being modeled and retained all
tested negative for antibodies (Table 1).
The study laboratory tested DBS spe-
August 2023, Vol 113, No. 8
of depression (CES-D-10), perceived
Research
Peer Reviewed
Baron et al.
895
RESEARCH & ANALYSIS
TABLE 1— Self-Reports of COVID-19 Infections and Dry Blood Spot (DBS) SARS-CoV-2 Antibody Home
Testing Results: Safe and Just Cleaners Study, New York City, March–June 2021
Self-Report of
Ever Having Had
COVID-19
Antibody Negative
DBS Results (n = 45),
No. (%)
Antibody
Indeterminant DBS
Results (n = 5),
No. (%)
All Participants
(n = 296), No. (%)
Returned DBS Kit
(n = 116), No. (%)
Antibody Positive
DBS Results (n = 66),
No. (%)
158a (53)
63b (54)
49b (74)
12 (27)
2 (40)
Don’t Know
21 (7)
8 (7)
5 (8)
3 (7)
0 (0)
No
117 (40)
45 (39)
12 (18)
30 (67)
3 (60)
Yes
Note. SARS-Cov-2 5 severe acute respiratory syndrome coronavirus 2.
a
Includes 5 participants who reported positive COVID-19 nasal swab tests and “no” (n 5 4) or “prefer not to answer” (n 5 1) when asked whether they had
had COVID-19.
b
Includes 1 participant who reported a positive COVID-19 nasal swab test but answered “prefer not to answer” when asked whether they had had
COVID-19.
AJPH
August 2023, Vol 113, No. 8
Impacts
896
their client (Table 2). Many houseclea-
At the time of the survey, March to
ners who reported losing pay during
June 2021, most respondents (89%;
Table 2 provides descriptive statistics
the shutdown because they were con-
n 5 265) were employed, yet 207 (70%)
for COVID-19 infection–related and SBE
cerned they might infect their client
reported earning less than $1000 per
effects. COVID-19 deaths among family
likely lost pay because they themselves
month from all jobs, and 133 (45%)
members and close friends were
had COVID-19. Among those who lost
reported being their family’s primary
reported by 152 (51%) of participants.
paid workdays during the spring 2020
wage earner. Since March 2020, only
Participant acceptance of the COVID-19
shutdown period, 23 (53%) reported
8% (n 5 23) of all participants had re-
vaccine appeared high. The vaccine be-
having had COVID-19 during spring
ceived unemployment compensation,
came available to all adults in NYC at
2020 compared with 61 (24%) of all
the beginning of April 2021. During the
other study participants (P < .01).
32% (n 5 94) received Supplemental
first half of our survey administration
At the end of the shutdown, 231 parti-
period, March 18 to April 30, 2021, only
cipants (78%) worked as housecleaners,
27% (n 5 44) had received at least 1
17 (6%) worked in a job other than
COVID-19 vaccine shot. For those sur-
housecleaning, and 48 (16%) were not
veyed between May 1 and June 11, 66%
employed. For the working houseclea-
(n 5 88) reported receiving at least 1
ners, 128 (55%) reported they were ex-
shot. Most (76%) of those not vaccinat-
tremely or very concerned about having
ed intended to get vaccinated.
enough clients to meet their financial
Nutrition Assistance, and 11% (n 5 34)
had received any other form of governmental financial benefits.
Given such low earnings and benefits,
it is unsurprising that reports of food insecurity were common, with 254 (86%)
indicating they “sometimes” or “often”
experience food insecurity on at least 1
of the 2 food insecurity questions.
Most participants (93%; n 5 275)
needs. We also found that only 84 (36%)
reported they were still working in
felt they could take sick leave, paid or
housecleaning in March 2020 just be-
unpaid, during the pandemic without
fore the citywide COVID-19 shutdown
retaliation. Given widespread concern
not have a fixed place to live, 118 (40%)
period (March 22–June 8, 2020), and of
about infection control, 142 (61%)
worried about having a secure living sit-
those, 85 (31%) continued to work as
reported using more disinfectant clean-
uation in the future, and 71 (24%)
housecleaners during the shutdown,
ing products at work. Almost half of all
reported being behind on rent by at
144 (52%) reported not receiving any
respondents (47%; n 5 138) reported
least $1000. Regarding domestic safety,
housecleaning income during the shut-
that if diagnosed with COVID-19, they
excluding 7 participants who did not
down, and 43 (16%) reported losing
would be likely or somewhat likely to
feel comfortable answering these ques-
paid housecleaning workdays because
feel afraid or embarrassed to disclose
tions, 59 (20%) reported feeling only a
they were concerned they might infect
this information to their employer.
little safe or unsafe inside their own
Research
Peer Reviewed
Baron et al.
Housing insecurity was also common.
Although only 16 participants (5%) did
RESEARCH & ANALYSIS
TABLE 2— COVID-19–Related Impacts; Employment, Social, and Economic Impacts; and Mental and SelfRated Health: Safe and Just Cleaners Study, New York City, March–June 2021
No. (%)
COVID-19–related impacts
Infection and exposure
158a (53)
Self-reported ever had COVID-19 or a positive test
Hospitalized with COVID-19
9 (3)
Not hospitalized but had COVID-19
149 (50)
Other members of household ever had COVID-19
182 (61)
Family/friends in US or other country died of COVID-19
152 (51)
Vaccination status
Vaccinated with at least 1 COVID-19 shot
132 (45)
Among 163 surveys completed March 18–April 30
44 (27)
Among 133 surveys completed May 1–June 11
88 (66)
Not vaccinated but probably will get vaccinated
123 (42)
Employment-related impacts
Still worked as a housecleaner in February 2020
275 (93)
85 (31)
Lost paid work days during shutdown because they might infect their client
43 (16)
Did not work and received no compensation from clients during shutdown
AJPH
Worked as housecleaner during March–June 2020 shutdown
144 (52)
August 2023, Vol 113, No. 8
Worked as housecleaner after the end of the shutdown in June 2020
231 (78)
Could take sick leave (paid or unpaid) during pandemic without retaliation
84 (36)
Extremely/very concerned about having enough clients to meet needs
128 (55)
Used more disinfectant cleaning products at work after the shutdown
142 (61)
If diagnosed with COVID-19, would be likely/somewhat likely to feel afraid or embarrassed to disclose this information to your employer
138 (47)
Among those who reported having COVID-19
75 (49)
Among those who reported not having COVID-19
63 (44)
Earning per month from all jobs at time of the survey, $
Not working
31 (11)
≤ 500
120 (41)
500–1000
87 (29)
≥ 1000
45 (15)
Received governmental financial benefits since March 2020
Unemployment compensation
23 (8)
Supplemental nutrition assistance benefits
94 (32)
Any other government assistance
34 (11)
Food insecurity, in the past 12 mo
Q1. Worried your food would run out before you had money to buy more
Sometimes
138 (47)
Often
102 (34)
Q2. The food you bought did not last and you didn’t have money to buy more
Sometimes
152 (51)
Often
61 (21)
Responding “sometimes” or “often” to Q1 or Q2
254 (86)
Other social and economic impacts
Housing insecurity
Does not have a fixed place to live
16 (5)
Continued
Research
Peer Reviewed
Baron et al.
897
RESEARCH & ANALYSIS
TABLE 2— Continued
No. (%)
Currently has a place to stay but worries about the future
118 (40)
Owes more than $1000 in back rent
71 (24)
Domestic insecurity since the pandemic beganb
Q1. Level of feeling safe inside their own home (only a little safe/unsafe)
59 (20)
Q2. Someone with whom they had romantic relationships yelled at them or said things that made them feel scared or bad
about themselves?
35 (12)
Either Q1 or Q2 (but not both)
86 (30)
Mental and self-rated health outcomes
Depression (CES-D 10 ≥ 10)
Prepandemic survey
65 (22)
During pandemic survey
100 (34)
New onset since prepandemic survey
59 (20)
Moderate/high perceived stress (Cohen’s PSS ≥ 14)
Prepandemic survey
130 (44)
During pandemic survey
160 (54)
New onset since prepandemic survey
63 (21)
AJPH
August 2023, Vol 113, No. 8
Fair/poor self-rated health
Prepandemic survey
86 (29)
During pandemic survey
129 (44)
New onset since prepandemic survey
68 (23)
Note. PSS 5 Perceived Stress 10-item scale. Study sample size was n 5 296.
a
Includes 5 participants who reported positive COVID-19 nasal swab tests and “no” (n 5 4) or “prefer not to answer” (n 5 1) when asked whether they had
COVID-19.
b
This excludes 7 participants who did not feel comfortable responding to questions about the safety of their home.
home, 35 (12%) reported verbal abuse
from a romantic partner, and 86 partici-
Socioeconomic Conditions
and COVID-19
Mental and Self-Rated
Health Measures
Compared with prepandemic responses, more participants exceeded
the threshold cutpoint for the mental
We examined whether employment sta-
COVID-19 and SBE-related factors
tus, living conditions, or demographic
(Table 4). In the fully adjusted models,
characteristics were related to a respon-
we found increased odds for depres-
dent’s having had COVID-19. When com-
sion for those reporting COVID-19
paring housecleaners reporting COVID-
(OR 5 1.95; 95% CI 5 1.10, 3.45 for
19 or a positive test to others in the
nonhospitalized participants;
study, the only difference found was in
OR 5 5.09; 95% CI 5 0.91, 28.36 for
the likelihood of living with household
hospitalized participants). We also
members who also had had COVID-19:
found increased odds for perceived
and self-rated health measures. We
89% of those reporting COVID-19 versus
stress among those not yet vaccinated
found that 20%, 21%, and 23% of parti-
30% of others (Table 3).
cipants formerly below the cutpoint
exceeded the cutpoint on the follow-up
(OR 5 2.39; 95% CI 5 1.28, 4.45), those
with household members who had
Predictors
survey for depression, perceived stress,
898
adjusting for prepandemic measures,
found increased odds associated with
pants (30%) reported at least 1 of these
aspects of domestic insecurity.
and fair or poor self-rated health,
COVID-19 (OR 5 1.93; 95% CI 5 1.04,
3.60), and those with family members
and fair or poor self-rated health, re-
Multivariable logistic regression mod-
or close friends who died of COVID-19
spectively (Table 2).
els for depression, perceived stress,
(OR 5 1.84; 95% CI 5 1.01, 3.35).
Research
Peer Reviewed
Baron et al.
RESEARCH & ANALYSIS
TABLE 3— Self-Reported COVID-19 Status by Demographic, Social, and Employment Conditions: Safe
and Just Cleaners Study, New York City, March–June 2021
Ever Had COVID-19a (n = 158),
Mean 6 SD or No. (%)
Did Not Have COVID-19
(n = 138), Mean 6 SD or No. (%)
Pb
Age, y
44.9 6 9.7
44.7 6 9.9
.88
Years living in United States
17.4 6 8.5
16.1 6 9.5
.21
Primary school
50 (31.7)
35 (25.4)
High school
74 (46.8)
61 (44.2)
Education level
.5
General equivalency diploma
5 (3.2)
6 (4.4)
At least some college
29 (18.4)
36 (26.1)
English comfort
.38
Uncomfortable
60 (38)
49 (36.3)
More or less
78 (49.4)
61 (45.2)
Comfortable
20 (12.7)
25 (18.5)
Yes
61 (38.6)
54 (39.4)
No
97 (61.4)
83 (60.6)
Primary family wage earner
.89
.73
72 (45.6)
69 (50)
No
84 (53.2)
67 (48.6)
Don’t know
2 (1.3)
2 (1.5)
0
18 (11.4)
96 (69.6)
1
33 (20.9)
21 (15.2)
2
35 (22.2)
5 (3.6)
3
32 (20.3)
9 (6.5)
≥4
40 (25.2)
7 (5.1)
No. of people in household with COVID-19
.001
Asked to work during COVID-19 shutdown
(March–June 2020)c
.29
Yes
41 (27.5)
44 (34.9)
No
107 (71.8)
82 (65.1)
Don’t know
August 2023, Vol 113, No. 8
Yes
AJPH
Health insurance
1 (0.7)
0
Employment in housecleaning after end of the
shutdown in June 2020
.2
Worked continuously
22 (13.9)
30 (21.7)
Worked on and off
101 (63.9)
78 (56.5)
Stopped working in housecleaning
35 (22.2)
30 (21.7)
a
We defined reported COVID-19 as self-reporting ever having had COVID-19 or a positive nasal swab test for COVID-19 at the time of the survey:
March–June 2021.
b
P value for continuous variables from the t test and for categorical variables from the x2 test.
c
This variable includes only those still working as housecleaners in March 2020, right before the shutdown (n 5 275).
Regarding employment-related fac-
they had COVID-19 (OR 5 3.09; 95%
to housecleaners’ concern that they
tors, we found increased odds for per-
CI 5 1.71, 5.58). We found increased
might infect their clients (OR 5 2.21;
ceived stress among those likely or
odds for fair or poor self-rated health
95% CI 5 1.05, 4.69).
somewhat likely to feel afraid or embar-
for those reporting having lost wages
Regarding other SBE effects, we
rassed to tell their employer whether
during the pandemic shutdown owing
found increased odds for perceived
Research
Peer Reviewed
Baron et al.
899
RESEARCH & ANALYSIS
TABLE 4— COVID-19 Infection–Related and Social and Economic Impacts Predictors of Perceived Stress,
Depression, and Self-Rated Health: Safe and Just Cleaners Study, New York City, March–June 2021
Moderate/Severe
Perceived Stress
(PSS ≥ 14), OR (95% CI)
Depression
(CES-D 10 ≥ 10),
OR (95% CI)
Poor/Fair Self-Rated
Health,a
OR (95% CI)
NA
NA
NA
NA
NA
NA
5.09 (0.91, 28.36)
NA
COVID-19 infection–related impacts
Household member COVID-19 status
≥ 1 member had COVID-19
1.93 (1.04, 3.60)
No member had COVID-19
1 (Ref)
COVID-19 deaths of family/close friends
Experienced deaths
1.84 (1.01, 3.35)
Did not experience deaths
1 (Ref)
COVID-19 vaccination status
Not yet vaccinated
2.39 (1.28, 4.45)
≥ 1 vaccine shot
1 (Ref)
Self-reported COVID-19 disease
Hospitalized with COVID-19
NA
Had COVID-19 but not hospitalized a
1.95 (1.10, 3.45)
Didn’t have/don’t know
1 (Ref)
August 2023, Vol 113, No. 8
Employment-related impacts
Would feel afraid or embarrassed to disclose to employer whether they had COVID-19
Likely/somewhat likely
3.09 (1.71, 5.58)
Somewhat unlikely/unlikely
NA
NA
NA
2.21 (1.05, 4.69)
1 (Ref)
Pay during shutdown
Lost housecleaning pay during shutdown because
respondent might infect client with COVID-19a
NA
AJPH
Did not work during shutdown or if worked, did not lose
pay because respondent might infect client
1 (Ref)
Other social and economic impacts
Food insecurityb
Yes
1.35 (1.05, 1.73)
1.46 (1.15, 1.84)
1.57 (1.23, 2.00)
No
1 (Ref)
1 (Ref)
1 (Ref)
2.04 (1.06, 3.91)
2.12 (1.17, 3.83)
2.11 (1.14, 3.89)
1 (Ref)
1 (Ref)
1 (Ref)
NA
NA
Domestic insecurity
Unsafe/a little unsafe and/or experienced IP verbal abuse
Felt safe and no IP verbal abuse
Housing insecurity
No fixed home/worried about future
Had a stable place to live
1.83 (1.01, 3.33)
1 (Ref)
Prepandemic mental and self-rated healthc
Perceived stress (PSS ≥ 14)
4.40 (2.41, 8.02)
NA
NA
Depressed (CES-D 10 ≥ 10)
NA
3.42 (1.78, 6.58)
NA
Poor/fair self-rated health
NA
NA
5.09 (2.70, 9.59)
Note. CES-D 10 5 Center for Epidemiologic Studies Depression 10-item scale; CI 5 confidence interval; IP 5 intimate partner; NA 5 variable not included in
the final model because it was eliminated in the stepwise regression analysis; OR 5 odds ratio; PSS 5 Cohen’s Perceived Stress scale. We adjusted all
models for age, educational attainment, and whether the participants indicated they were the primary family wage earner in their prepandemic survey.
We included the prepandemic mental and self-rated health variables in the models only for that same outcome measure.
a
Models containing this variable included only the 275 housecleaners who were still working in housecleaning in March 2020 because those no longer
working were not asked this question.
b
4-point food insecurity scale: 1 point each for answering “sometimes” on the 2-item food insecurity measure, 2 points each for answering “often,”
reference is answering no to both questions.
c
We collected all prepandemic measures between August 2019 and February 2020.
900
Research
Peer Reviewed
Baron et al.
RESEARCH & ANALYSIS
stress for those reporting no fixed
immigrant Latinx populations, especially
was associated with higher psychologi-
place to live or worry about future
in the first year of the pandemic.1
cal distress and lower self-rated health
housing (OR 5 1.83; 95% CI 5 1.01,
Almost one third reported working as
during the COVID-19 pandemic.27
3.33). We found increased odds for
housecleaners during the shutdown of
Fewer studies have examined
those reporting food insecurity
nonessential work, from March 22 to
COVID-19–related risks associated with
(OR 5 1.35; 95% CI 5 1.05, 1.73 for per-
June 8, 2020, although we did not find
domestic safety and intimate partner
that this was associated with higher
abuse because of challenges in collect-
rates of reported COVID-19 infections.
ing such data,28 although public health
95% CI 5 1.23, 2.00 for self-rated
We also documented a myriad of other
experts have highlighted the potential
health). Similarly, those experiencing
SBE effects, including reduced employ-
risks.29 Finally, the rates of employ-
domestic insecurity had increased
ment and low earnings, perceived
ment, housing, and food insecurity
odds for all 3 outcomes (OR 5 2.04;
COVID-19–related stigma at work, food
documented in our study are consis-
95% CI 5 1.06, 3.91 for perceived
and housing insecurity, and feeling
tent with findings from a national sur-
stress; OR 5 2.12; 95% CI 5 1.17, 3.83
unsafe at home or experiencing inti-
vey of Spanish-speaking domestic
for depression; and OR 5 2.11; 95%
mate partner verbal abuse. As has
workers conducted by the National Do-
CI 5 1.14, 3.89 for self-rated health).
been widely reported in other stud-
mestic Workers Alliance during the first
ies,23 we found deterioration in house-
6 months of the pandemic.30 Our study
cleaners’ mental and self-rated health
adds to these previous studies by cap-
compared with prepandemic levels,
turing how multifaceted SBE stressors
with COVID-19 infection, employment,
together with the direct effects of
and other SBE impacts contributing.
COVID-19 infections contributed to
Lastly, although studies have found
changes in mental and self-rated health
mixed results related to vaccine accep-
in our sample of Latinx immigrant
DISCUSSION
Our study provides insight into the
Latinx immigrant housecleaners in NYC
at the beginning of the second year of
tance among Latinx populations,
the COVID-19 pandemic (March–June
findings suggest a high level of
2021), and we were able to compare
acceptance.
24
our
housecleaners in NYC.
Consistent with the communitybased participatory research values of
mental and self-rated health measures
Our findings are consistent with oth-
to measures collected in the 6 months
er COVID-19 studies that included larg-
and following the model of Public
before the pandemic-related shutdown
er and more diverse populations. For
Health 3.0, we were committed as a
in NYC. Our findings contribute to the
example, a national survey found that
community–academic partnership to
literature by providing a unique and
all racial and ethnic minorities— but es-
using our data to influence policy to im-
comprehensive picture of the depth of
pecially those having limited English
prove workers’ social and economic
the pandemic’s impact on a population
proficiency, less education, and lower
conditions. Our community partner,
of workers who faced a multitude of
incomes and those living in large
Make the Road New York, is a major im-
disproportionate risks.
cities—were more likely to report that
migrant advocacy organization in New
We found that 53% of our participants
people acted afraid of them because of
York, and it responded to these find-
reported ever having had COVID-19 be-
suspected COVID-19 infection.25 A na-
ings by providing essential emergency
fore being surveyed between March 18
tional food insecurity study using the
support through food pantries and fi-
and June 11, 2021, and this high rate of
same 2-item measure found an in-
nancial assistance. Beyond these stop-
infection was confirmed by COVID-19
creased risk of food insecurity among
gap measures, it prioritized promoting
antibody tests. The housecleaners’ rate
Hispanic and low-income populations
policies to improve financial equity.
of infection was almost double the rate
and also found associations between
Make the Road New York, in coalition
documented over a similar period in
becoming food insecure and anxiety
with other community organizations
26
our overall Safe and Just Cleaners Study
a national COVID-19 seroprevalence
and depres
