Prevention of COVID19 Infection in Nursing Homes by Chemoprophylaxis With Hydroxychloroquine (PREVICHARM)

The objective of this study is to evaluate the effectiveness of hydroxychloroquine
chemoprophylaxis in healthcare personnel and residents:

1. the incidence of secondary cases of SARS-CoV-2 infection in residents

2. the incidence of SARS-CoV-2 infection in the staff. To assess the effect of measures to
promote infection control. Design: Controlled, randomized, triple-blind cluster
multicentre trial. Subjects: Institutionalized older people and nursing homes staff in
four Autonomous Communities.

Sample: 880 professionals, 1050 residents in 84 residences. They will be given 800mg of
hydorxychloroquine the first day and 400mg in the subsequent four days. Nasopharyngeal
exudate will be taken for baseline RT-PCR, and at six and 14 days after the adminsitration of
hydroxychloroquine.

Mortality, compliance, safety, symptoms, and hospital admission will be evaluated
secondarily.

The highest risk group for SARS-CoV-2 infection is that of people over 60 years of age, with
cardiovascular disease, diabetes, COPD and immunosuppression, reaching up to 32% of those
infected. But, in addition, the impact is especially insidious among institutionalized older
people. The Spanish MInistry of Health has developed some regulations with measures to curb
the impact in this context. WHO places institutionalized people as a very vulnerable group
due to their permanent close contact with other people in the residence, the abundance of
group activities, the presence of people with cognitive impairment and the need for help for
their activities of daily living, which involves close physical contact with healthcare
personnel. Even the interventions for preventing and controlling the infection could
seriously affect the mental health and well-being of this vulnerable people, even if they do
not contract the infection. In the province of Malaga (Spain), a checklist has been developed
consisting of 13 items related to knowledge and compliance with the rules of prevention,
self-protection, isolation and circuits, carried out by case management nurses. This
instrument contains many of the WHO recommendations and may be an element that could make a
difference in this context.

A significant proportion of those infected are known to have had close contact with probable
or confirmed SARS-CoV-2 cases. In institutionalized people, the main source of contact (once
family visits have been eliminated) is caused by cross-contamination with the care staff who
provide direct care or other infected residents. In nursing homes, staff ratios are much more
limited than in acute care hospitals (aggravated by the demand for professionals due to the
epidemic). Furthermore, the possibility of isolation of infected patients is very limited: On
average, residences only have 15% of their rooms prepared for isolation.

Among the various therapeutic strategies being tested, the use of hydroxychloroquine (HCQ)
emerges as a viable option, albeit with limitations in the studies. This use has already been
proposed for a long time due to its potential antiviral action, since it alkalinizes the
cellular pH necessary for binding of the virus to the host cell and interferes with
glycosylation of ECA2 receptors and blocks endosomatic transport of the virus. HCQ
administered in a safe dose range is distributed throughout the body, including the lung
parenchyma after oral administration, with few adverse effects. Compared to the antivirals
under study for SARS-CoV-2 infection, it is much more feasible in real clinical practice
conditions to access HCQ, a drug on which, in addition, there is enormous experience, its
safety is well known, It is funded by the National Health System (NHS), it is cheap and
accessible on the NHS. There are currently several clinical trials underway and a consensus
has been developed that recommends their use in patients with SARS-CoV-2 pneumonia. Regarding
the prophylactic use of HCQ, another trial in Spain has just started to use it in a "test and
treat" strategy in patients among the general population, but it does not directly affect the
institutionalized population. Our study aims to explore whether the administration of
chemoprophylaxis in nursing homes is a protective strategy for both institutionalized people
and the professionals themselves.

This is an controlled, randomized, triple-blind cluster study.The primary target population
of the study consists of professionals who provide direct care in nursing homes (nursing
assistants and nurses) and institutionalized older people.

The sampling frame will be integrated by the Communities of Andalusia, Madrid, Aragon and the
Autonomous Community of Navarra. A conservative sample calculation has been designed, due to
the current uncertainty regarding incidence. The incidence of Andalusia has been taken as the
lower limit and the incidence provided in the study by Kimball et al. as an upper limit, and
bearing in mind that many nursing homes are taking measures to isolate and relocate
residents. To decrease the minimum infection rate in residents from 25% to 10% (delta =
-0.15), with an alpha of 0.05, a power of 80%, assuming a minimum of 13 residents per cluster
with a CCI of 0.1, with a replacement rate of losses of 25%, a total of 1050 residents would
be necessary, with 7 clusters in each step, distributed in a total of 84 nursing homes.
Regarding workers, taking the references from Andalusia and the study by McMichael et al. to
decrease the minimum infection rate in professionals from 18% to 5% (delta = -0.13), with the
same power parameters, assuming a minimum of 10 professionals per cluster with a CCI 0.1, and
a replacement rate of 5% 880 professionals with the same number of clusters and nursing homes
are required.

The PREVICHARM study was originally conceived at the end of March 2020 as a randomized
clinical trial by stepped wedge clusters, since it was a methodological path that at that
time allowed all participating subjects to receive the drug as chemoprophylaxis, in a
staggered manner, without losing the comparison with a control group. After the evolution of
the pandemic and the research around HCQ, the research team has considered the need to
introduce some modifications in the design, to eliminate the staggering in the trial
sequence, maintaining the design controlled and randomized by clusters, but, introducing an
assignment of two parallel arms by triple blind to, on the one hand, have a totally blinded
design that improves the robustness of the conclusions and, on the other, simplify the
operationalization of the trial and thus further strengthen its feasibility. A criterion on
the period of time necessary for inclusion has also been updated, reducing it to 8 days, in
light of current knowledge on the predictive value of PCR in the course of infection in
exposed patients.

The modifications do not affect the objectives, or hypotheses, or the outcomes, or the
evaluation procedures thereof, or the follow-ups or the evaluation of the safety of the
trial.

Variables: Sociodemographic, adjustment and control of confounders (Cluster, follow-up time,
hours worked by each professional during the study, gradient of application of
non-pharmacological measures); Primary endpoints: Number of secondary cases of SARS-CoV2
infection among residents, and SARS-CoV-2 infection in caregivers who provide direct care;
Secondary endpoints: mortality, compliance with treatment, symptoms of SARS-CoV-2 infection,
hospital admission of personnel and adverse events.

All the nursing homes in the province will be invited to participate and after randomization,
the phased inclusion of residences will begin, requesting the consent of professionals to be
included and proceeding to collect basal exudate and study variables. In the randomized
nursing homes in each step as an intervention, the HCQ regimen will be provided and the
nasopharyngeal exudate follow-up will be performed for RT-PCR at 6 and 14 days. In those in
the control group step, an exudate will be collected on the same days than intervention
group. Once enrolled, subjects will receive a double naso-oro-pharyngeal swab and will be
provided blinded medication (HCQ or placebo, as appropriate) for the entire chemoprophylaxis
cycle. Subsequently, smears will be taken again at 6 and 14 days. The sample collection
procedure will be standardized through the protocol developed by the Ministry of Health and
technicians will be trained in advance in its performance.

The technicians will go to each residence according to a cadence pre-established by the
research team.

Adverse events will be monitored at each weekly visit using a standardized procedure
(gastrointestinal disorders, appetite disturbances, hypoglycemia, weight loss, decreased
visual acuity, photophobia, impaired visual accommodation with symptoms of blurred vision,
deafness, tinnitus, and vertigo, muscle weakness, symptoms of liver and kidney disorders,
neurological disorders). Serious and non-serious adverse events will be distinguished and
will be notified by means of a standard procedure to the Research Ethics Committee, and to
the Spanish Agency for Medicines and Health Products. An insurance policy has been contracted
for risk coverage.

The study will comply with all the precepts of the Declaration of Helsinki and its subsequent
revisions, as well as the Standards of Good Practice. Participants will be informed of the
characteristics and purpose of the study and will be asked for informed consent in accordance
with the regulations in force in Spain using the form and the corresponding information
sheet. The study has been approved by the Malaga Provincial Research Ethics Committee and the
Spanish Agency for Medicines and Health Products, as well as the participating institutions.

All the trial monitoring process and pharmacovigilance will be carried out by the Clinical
Trial Units of the Spanish Clinical Research Network of the participating Centers.

Exploratory data analysis, bivariate analysis (chi square, Student t, ANOVA or their
non-parametric equivalents in case of non-normality) will be performed and the main
hypothesis will be tested using generalized linear binomial and Poisson models, also taking
the cluster, as random effects factor, and time as a fixed factor. Intrasubject analyzes will
be carried out using mixed generalized linear models. All the analyses will be performed by
intention to treat, using multiple imputation models.

The total estimated duration is a maximum of 12 months.