Open access Protocol BCG vaccination to reduce the impact of COVID-­19 in healthcare workers: Protocol for a randomised controlled trial (BRACE trial) Laure F Pittet ‍ ‍, Nicole L Messina ‍ ‍, Kaya Gardiner ‍ ‍, Francesca Orsini, Veronica Abruzzo, Samantha Bannister ‍ ‍, Marc Bonten, John L Campbell ‍ ‍, Julio Croda ‍ ‍, Margareth Dalcolmo ‍ ‍, Sonja Elia ‍ ‍, Susie Germano ‍ ‍, Casey Goodall, Amanda Gwee ‍ ‍, Tenaya Jamieson, Bruno Jardim ‍ ‍, Tobias R Kollmann ‍ ‍, Marcus Vinícius Guimarães Lacerda ‍ ‍, Katherine J Lee ‍ ‍, Donna Legge, Michaela Lucas ‍ ‍, David J Lynn ‍ ‍, Ellie McDonald ‍ ‍, Laurens Manning ‍ ‍, Craig F Munns ‍ ‍, Kirsten P Perrett ‍ ‍, Cristina Prat Aymerich ‍ ‍, Peter Richmond ‍ ‍, Frank Shann ‍ ‍, Eva Sudbury, Paola Villanueva ‍ ‍, Nicholas J Wood, Katherine Lieschke, Kanta Subbarao ‍ ‍, Andrew Davidson, Nigel Curtis ‍ ‍, the BRACE trial Consortium Group To cite: Pittet LF, Messina NL, ABSTRACT Gardiner K, et al. BCG Strengths and limitations of this study Introduction BCG vaccination modulates immune responses vaccination to reduce the to unrelated pathogens. This off-­target effect could reduce impact of COVID-­19 in ►► By including healthcare workers in five countries the impact of emerging pathogens. As a readily available, healthcare workers: Protocol across three continents, this randomised controlled inexpensive intervention that has a well-­established safety for a randomised controlled trial is large enough to assess the effect of BCG vac- trial (BRACE trial). BMJ Open profile, BCG is a good candidate for protecting healthcare cination on the incidence of severe COVID-­19, as 2021;11:e052101. doi:10.1136/ workers (HCWs) and other vulnerable groups against well as symptomatic COVID-­19 and other infections. bmjopen-2021-052101 COVID-­19. ►► The trial includes robust clinical data collection on a Methods and analysis This international multicentre phase ►► Prepublication history and daily basis during illness episodes to enable precise III randomised controlled trial aims to determine if BCG additional supplemental material measurement of severity, as well as real-­time track- vaccination reduces the incidence of symptomatic and severe for this paper are available ing of missed opportunities for SARS-­CoV-­2 testing online. To view these files, COVID-­19 at 6 months (co-­primary outcomes) compared with to trigger individualised reminders. please visit the journal online no BCG vaccination. We plan to randomise 10 078 HCWs ►► By including regular blood sampling, the trial will (http://​dx.​doi.​org/​10.​1136/​ from Australia, The Netherlands, Spain, the UK and Brazil in also provide information on the BCG-­induced mo- bmjopen-​2021-​052101). a 1:1 ratio to BCG vaccination or no BCG (control group). The lecular and immunological changes associated with participants will be followed for 1 year with questionnaires Received 07 April 2021 protection against off-­target infections. and collection of blood samples. For any episode of illness, Accepted 02 September 2021 ►► Limitations include the difficulty in blinding partic- clinical details will be collected daily, and the participant will be ipants to group allocation due to the reaction and tested for SARS-­CoV-­2 infection. The secondary objectives are scar induced by BCG in most people, therefore re- to determine if BCG vaccination reduces the rate, incidence, quiring careful choice of objective outcomes as well and severity of any febrile or respiratory illness (including as blinded assessment of measures where possible. SARS-­CoV-­2), as well as work absenteeism. The safety of BCG ►► A further limitation is that aspects of the trial may vaccination in HCWs will also be evaluated. Immunological require adjustment as the pandemic evolves and analyses will assess changes in the immune system following knowledge about COVID-­19 expands. vaccination, and identify factors associated with susceptibility to or protection against SARS-­CoV-­2 and other infections. Ethics and dissemination Ethical and governance approval will be obtained from participating sites. Results will be to the COVID-­19 pandemic.1 With no pre-­ published in peer-­reviewed open-­access journals. The final existing immunity and in the absence of a cleaned and locked database will be deposited in a data vaccine, it was predicted that up to 60% of © Author(s) (or their sharing repository archiving system. the global population could be infected.2 As employer(s)) 2021. Re-­use Trial registration ​ClinicalTrials.​gov NCT04327206 a result of their close contact with patients, permitted under CC BY. Published by BMJ. healthcare workers (HCWs) are at particularly Correspondence to INTRODUCTION high risk.3–5 By September 2020, over 7000 Professor Nigel Curtis; In the twilight of 2019, the novel human HCWs worldwide had succumbed to COVID-­ ​nigel.​curtis@​rch.​org.​au pathogen SARS-­ CoV-­ 2 emerged leading 19.6 This susceptibility is consistent with the Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 1 Open access SARS epidemic in 2003, when HCWs comprised 21% of compared with no BCG vaccination: (1) Prolongs the all cases.4 HCW absenteeism or quarantine requirements time to first SARS-­CoV-­2-­proven symptomatic infection; due to COVID-­19 impair healthcare delivery during the (2) Reduces the incidence and/or severity of febrile or pandemic. respiratory illness, including COVID-­ 19; (3) Reduces Preventive interventions to protect against emerging absence from work and (4) Is safe in HCWs (including pathogens are needed, particularly for HCWs. The tuber- revaccination). Exploratory objectives are to determine culosis (TB) vaccine BCG has beneficial off-­target effects if BCG vaccination compared with no BCG vaccination: that protect against unrelated infections.7–15 These effects (1) Reduces oral herpes simplex virus reactivation in the have been most extensively studied in high-­ mortality subgroup of adults with recurrent cold sores; and (2) settings in Africa, where trials have shown a 38% reduc- Changes immune function and whether these changes tion in all-­ cause neonatal mortality in infants vacci- are associated with protection against non-­tuberculous nated with BCG-­Denmark compared with unvaccinated infectious diseases, including COVID-­19. The study will infants.16 17 This protection, observed within days of vacci- also investigate factors that influence immune responses nation, is proposed to be attributable to reduced deaths and infection risk (including COVID-­19). from infections other than those caused by Mycobacterium tuberculosis, particularly respiratory tract infections and sepsis.7–9 16 The beneficial off-­target effects of BCG are METHODS AND ANALYSIS proposed to result from BCG-­induced immunomodula- Trial design and setting tion.10 18–22 In adults, BCG vaccination increases innate In this multicentre, phase III, randomised controlled immune responses to unrelated pathogens, an effect trial, HCWs will be randomised in a 1:1 ratio to receive termed trained immunity,23 that is sustained for at least a or not receive BCG vaccination. The protocol is avail- year after vaccination.24 able in the online supplemental material. Recruitment In a human challenge model, prior vaccination with started 30 March 2020 and is divided into two stages. In BCG-­ Denmark reduced viraemia by over 70% and stage 1 (March to April 2020), HCWs in Australia were improved antiviral immune responses to yellow fever randomised during the Australian influenza season in an virus vaccine,10 a single-­ stranded, positive-­sense RNA open-­label design to receive BCG and influenza vaccine, virus (as is SARS-­CoV-­2) compared with no BCG vaccina- or influenza vaccine alone. In stage 2 (from May 2020), tion. In three randomised controlled trials in adults, BCG HCWs in five countries are randomised in a blinded vaccination reduced the incidence of acute upper respira- fashion to receive BCG vaccine or placebo saline intra- tory tract infections by 70%–80% compared with no BCG dermal injection. Participants will be followed-­up for 12 vaccination (although this was not necessarily the primary months. Data will be combined from both stages in a outcome of these trials).11–13 Several studies have shown preplanned meta-­analysis. that BCG can also reduce symptoms in human papilloma virus and herpes simplex virus infections in adults.14 15 Participants and eligibility criteria Another study in adults showed that BCG-­Bulgaria altered Up to 10 078 HCWs in Australia, The Netherlands, the clinical and immunological responses to malaria.25 In Spain, the UK and Brazil will be recruited across animal models, numerous studies have shown that BCG both stages of the trial. Potential participants will protects against disease and mortality caused by a wide receive information about the trial via email, health- range of pathogens, including single-­stranded, positive-­ care facilities notice board, and/or website/social sense RNA viruses.15 26–28 media. This will include a short description about BCG vaccination potentially offers a readily available, the study, a link to a website with further informa- safe, and low-­cost way to reduce the incidence and/or tion and contact details for questions. Potential severity of COVID-­19, as well as other emerging patho- participants will be able to evaluate their eligibility gens that might arise in the future.29 30 This would be of online, and access the site-­s pecific participant infor- particular benefit among high-­risk groups in whom the mation and consent form (see online supplemental disease has the greatest impact, such as HCWs. material) prior to attending the clinic for enrolment. Eligibility will be ascertained by study staff during the baseline visit where participants will give informed STUDY AIMS consent. HCWs are eligible if working in healthcare Primary objective settings during the COVID-­1 9 pandemic or having To determine in HCWs if BCG vaccination compared with face-­t o-­f ace contact with patients. Stage 1 participants no BCG vaccination reduces the incidence of (1) Symp- were required to receive influenza vaccination on tomatic and (2) Severe COVID-­19 during the 6 months the day of randomisation, regardless of group alloca- following randomisation. tion. Exclusion criteria are: previous positive test for SARS-­ C oV-­ 2 , contraindication to BCG vaccine (eg, Secondary objectives immunosuppression, pregnancy, serious underlying To determine in HCWs, during the 6 months and 12 illness, history of active TB), previous adverse reac- months following randomisation, if BCG vaccination tion to BCG vaccine (eg, significant local reaction 2 Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 Open access such as abscess or suppurative lymphadenitis), BCG Outcome vaccine administered within the last year, any other Primary outcomes live-­a ttenuated vaccine administered within the last Two primary outcomes have been chosen for this study: month or indicated in the next month, any COVID-­ incidence of symptomatic COVID-­19 and incidence of 19-­specific vaccine administered, or involvement in severe COVID-­19 during the 6 months after randomis- another COVID-­1 9 prevention trial. ation. In light of the lack of knowledge about this new virus, we deemed it important to have sufficient power Intervention to detect the potential effect of BCG vaccine compared Participants randomised to BCG will receive a single dose with no BCG (control group) on both of these outcomes. of BCG-­ Denmark (AJ Vaccines, Copenhagen), 0.1 mL Our hypothesis is that BCG vaccine will reduce both (corresponding to 2–8×105 colony forming units of Myco- the number of cases of symptomatic COVID-­19 and the bacterium bovis, Danish strain 1331) as an intradermal number of severe cases of COVID-­19. injection over the region where the deltoid muscle inserts Symptomatic COVID-­ 19 will be defined as an episode into the humerus, using a 1 mL syringe fitted with a short of illness with fever or at least one symptom of respiratory (10 mm) bevel needle (25G to 30G). disease (including sore throat, cough and shortness of In stage 2, participants randomised to not receive BCG breath) plus a positive SARS-­CoV-­2 test (PCR, antigen or will be given a saline placebo intradermal injection using serology). the same procedure described for BCG. Severe COVID-­19 will be defined as an episode of illness with a positive SARS-­CoV-­2 test plus at least one of the Randomisation process following as a consequence of COVID-­19: (1) Death, (2) Randomisation to BCG or non-­BCG groups will be done Hospitalisation or (3) Non-­ hospitalised severe disease, using a web-­ based randomisation procedure on the defined as being non-­ ambulant or unable to work for Research Electronic Data Capture platform (REDCap),31 three consecutive days or more. Non-­ ambulant will be provided by an independent statistician. Randomisation defined as being ‘pretty much confined to bed (meaning will be in randomly permuted blocks of variable length finding it very difficult to do any normal daily activities)’, (2, 4 or 6) stratified by stage of the study (stage 1 or 2), and unable to work as ‘not feeling physically well enough study site, age (<40 years; 40–59 years; ≥60 years) and to go to work’. presence of comorbidity (any of diabetes, chronic respi- ratory disease, cardiac condition, hypertension). Secondary outcomes Secondary outcomes will be assessed over the 6 months Blinding and 12 months following randomisation and are: any In stage 1, participants will be recruited in an open-­label febrile or respiratory illnesses, duration of symp- setting, meaning that only the trial statisticians will be toms, number of days of absence from work, number blinded. of days confined to bed, complications (eg, pneu- In stage 2, only those preparing and administering monia, need for oxygen therapy, admission to critical the intervention (BCG or placebo) will be unblinded; care, need for mechanical ventilation, outcome), and participants, investigators, statisticians and other trial asymptomatic SARS-­CoV-­2 infection. Vaccine-­related staff involved in follow-­up and data collection, will be adverse reactions (frequency, severity and duration) blind to the randomisation group throughout the trial. will be compared between groups, and between The code breaking procedure is available in the online participants who are BCG-­ naïve and those who are supplemental material. BCG-­revaccinated. Figure 1 Study flow chart. Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 3 Open access Exploratory outcomes randomisation, during the 2 weeks postvaccination The impact of BCG vaccination on herpes simplex virus (vaccine diary) and 3-­monthly during follow-­up, using reactivation will be evaluated using the time to first recur- the REDCap platform.31 A summary of the data collected rence, as well as the number, duration and severity of is provided in table 1. To verify eligibility, and for stratifi- recurrences. The impact of vaccinations on the immune cation prior to randomisation, the baseline questionnaire system will be evaluated using serology, immunoprofiling will collect data on demographics, workplace exposure and cytokine responses.21 22 The influence of host and and medical history. The vaccine diary will be used to external factors on the immune response and infection document common reactions in the first 2 weeks after will also be evaluated. vaccination, with severity categorised using a toxicity grading scale.32 At 3 months, 6 months, 9 months and Data and sample collection 12 months after randomisation, follow-­up questionnaires Participants will be followed-­up for 12 months as illus- will be used to collect medical data outcome measures trated in figure 1, using questionnaires and collection and potential modulating factors. of blood samples. Additional information on severe disease will be obtained from hospital medical records. Illness questionnaires Questionnaires Participants will be asked weekly if they have been unwell Web-­based questionnaires will be completed by since the last contact using a smartphone application participants or study staff at the time of recruitment, designed for the trial (Trial Symptom Tracker, WeGuide) Table 1 Data collected from the questionnaires Eligibility Baseline / Vaccine 3-­mly Episode Participant questionnaires check randomisation diary* FU† FU Demographics, medical history            Inclusion and exclusion criteria X          Age, sex, BMI, comorbidities (diabetes, cardiovascular disease, chronic   X       respiratory disease, hypertension), alcohol, smoking  Medication: use of hydroxychloroquine, azithromycin, lopinavir-­ritonavir,   X   X   oseltamivir or antihypertensive drugs  Vaccination: timing of administration of other vaccine(s)       X   Mycobacterium spp exposure: prior BCG vaccination, previous positive TST,     X   X   latent TB, or stay in high TB prevalence country  Cold sores recurrences and impact on quality of life   X   X   COVID-­19 exposure          Workplace: profession, department, amount of direct patient contact, contact   X   X   with COVID-­19 cases  Household: composition, COVID-­19 exposure   X   X   BCG vaccination site          Photograph of BCG vaccination site   X X X    Vaccine site reaction (pain, redness, swelling, tenderness)     X      Enlarged lymph node     X      BCG scarring and complication     X X   Illness episode            Presence of fever, cough, shortness of breath/difficulty breathing, sore throat,       X X runny/blocked nose, fatigue, muscle or joint pain, headache, nausea or vomiting, diarrhoea, loss of smell or taste  COVID-­19 test result       X X  Days absent from work       X X  Days confined to bed         X  Medical consultation, ED presentation       X X  Hospital admission       X X  Complications       X X *Completed during the 2 weeks following vaccination. †Questionnaire available in the online supplemental material. BMI, body mass index; ED, emergency department; Episode, illness episode; FU, follow-­up; mly, monthly; TB, tuberculosis; TST, tuberculin skin test. 4 Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 Open access and/or by contacting the participant by telephone, text the interim analysis. Allowing for a 16% loss to follow-­up, message or email. With each episode of illness, the partic- we plan to recruit 7244 HCWs in stage 2. ipant’s symptoms will be recorded daily. At the end of the In the preplanned meta-­analysis, there will be a total episode, a short survey will record COVID-­19 test results, of 10 078 participants (8062 allowing for 20% loss to the severity of the episode, its management and its impact follow-­up). Assuming a 4% rate of severe COVID-­19 by on ability to work. 6 months in the control group, a sample size of 8062 (4031 per group) will provide 90% power with two-­tailed COVID-19 testing 0.04 significance level for the Pearson’s χ2 test to detect a When symptomatic, participants will be asked to risk ratio of 0.67, equivalent to an absolute difference of undergo testing for SARS-­CoV-­2 infection with a vali- -1.3%. dated test as required by their institution or local health authority. Participants who report fever or Statistical analysis respiratory symptoms but have not been tested will be The primary analysis will be an intention to treat anal- identified rapidly through the daily illness question- ysis on all eligible participants randomised during stage naires and the participant will be called by the study 2 who did not have a SARS-­CoV-­2 positive test result staff to help arrange testing. at time of randomisation. The proportions of partici- pants that meet each of the primary outcomes will be Blood sampling compared between the two groups using an absolute Blood will be collected at recruitment and 3 months, 6 risk difference, estimated using a time-­to-­event anal- months, 9 months and 12 months after randomisation ysis, where the survival curve for each combination of for measurement of anti-­SARS-­CoV-­2 antibodies. This will strata and randomised group will be calculated using a enable us to identify participants who were seropositive flexible parametric survival model (Royston-­Parmar33 prior to randomisation and those who seroconvert during model), presented with 95% CIs calculated with boot- follow-­up. Blood samples will also be used for the explor- strap standard errors. atory objectives related to vaccine-­ induced changes in Secondary analyses are planned using similar models, the immune system and prediction of risk of COVID-­19. adjusting for the following covariates: sex, number and Interferon γ release assays will be included in sites with a type of comorbidities, history of previous BCG vaccina- high TB prevalence. tion. Subgroup analysis based on the same covariates are planned. The survival analysis strategy enables all Active tracking of missing data participants with any follow-­up data to be included in the Automated reports will identify in real time any missing analysis. data or missed opportunity for COVID-­ 19 testing, Secondary outcomes will be analysed and reported enabling individualised reminders by email, text message according to their nature (binary, continuous or cate- and telephone call by the study staff. gorical). Survival analysis will be used to analyse time to event outcomes with censoring of participants at their last Sample size follow-­up. The sample size for stage 2 of the study and the A meta-­ analysis combining participants from both preplanned meta-­ analysis was chosen to provide stages will be done, using the same analysis as described adequate power for the two primary outcomes. above, adjusted for the stage of the study. Control of type I error will be managed by splitting Rate, severity, time to onset and duration of adverse the significance level between both outcomes and reactions will be described using proportion, median and the preplanned interim analysis. The calculations are IQR, and expressed according to the number of partici- summarised in online supplemental table 1. pants answering the safety questions, presented by treat- It is estimated that 55% of participants will have symp- ment group. The safety outcomes in the BCG group will tomatic COVID-­19 in the control group by 6 months; be compared between participants who were revaccinated with 1:1 randomisation, a sample size of 2016 will provide and those who were BCG-­naïve using the χ2 test and the 95% power with two-­tailed 0.005 significance level for the Mann-­Whitney U test. Pearson’s χ2 test (with continuity correction) to detect an Further exploratory analyses will evaluate the associa- absolute difference of -10% (to 45%) between the BCG tion between various factors and immune function, using vaccine group and the control group. both clinical and in vitro measures. For severe COVID-­19 by 6 months, the study is powered The full details of the interim and final analysis will be to identify a risk ratio of 0.67 in the BCG compared provided in the statistical analysis plan which will be final- with the control group. Assuming a 4% rate of severe ised prior to unblinding and database lock. COVID-­19 by 6 months in the control group, a sample size of 6076 will provide 80% power with two-­tailed 0.04 Interim analysis significance level for the Pearson’s χ2 test to detect a There will be a single interim analysis of severe COVID-­19 risk ratio of 0.67, equivalent to an absolute difference of once 100 events have occurred.34 The analysis will follow -1.3%. The remaining 0.005 was planned to be spent on the same methodology as the primary analysis. The Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 5 Open access stopping rule will be based on an alpha spending func- of BCG at the time of hospital admission for COVID-­19 tion, where the threshold to identify efficacy is based on did not raise any safety concerns.48 In addition, there was the amount of data available at the time of the interim no evidence of increased severity of COVID-­19 among analysis. participants who participated in BCG trials prior to the pandemic.49 ETHICS AND DISSEMINATION Limitations Ethics As a papule develops at the injection site around The trial will be run in accordance with the ethical prin- 2 weeks after BCG vaccination in most people, it is chal- ciples of the Declaration of Helsinki, and ethical and lenging to blind participants to their group allocation, governance approval sought for all participating sites. even with a placebo. We have chosen objective primary The primary Human Research Ethics Committee is the outcomes to decrease the risk that awareness of alloca- Royal Children’s Hospital Melbourne (No. 62586), and tion biases the trial results. Members of the research the protocol was approved by all participating sites. An team following up the participants will be blinded to independent data safety monitoring committee (DSMC) group allocation, as well as those doing the analysis (by will oversee trial conduct, safety and the interim analysis; the removal of all variables related to BCG from the the DSMC charter is available in the online supplemental data set) until data cleaning is complete and the statis- material. tical analysis plan has been signed by all investigators. The BRACE trial is designed in two stages, from which Risk data will be combined in a preplanned meta-­analysis. The BCG vaccine has a well-­established safety profile in Stage 2 was initiated when extending the recruitment healthy individuals. Infant BCG administration is near outside Australia in April 2020. As it was spring in the universal in many countries, and therefore many HCWs northern hemisphere, the influenza vaccine could have previously received this vaccine. The risk of an not be administered to the control group. In stage 2, earlier and accelerated local reaction is increased for participants are randomised to BCG vaccination or a HCWs who have previously had the vaccine (BCG revac- saline placebo injection, with the placebo contributing cination), compared with those receiving it for the first to increasing the retention rate and lowering the risk time (BCG-­naïve).35–42 However, passive surveillance in of response bias. countries recommending revaccination has not raised any The evolution of COVID-­19 epidemiology and avail- safety concerns.43 44 Also, rates of serious adverse events ability of COVID-­ 19-­ specific vaccines are unpredict- among BCG-­revaccinated participants were not increased able, and we may not be able to detect any difference in large randomised controlled trials in Africa compared between the groups if the numbers of symptomatic with those who had not previously received BCG.11 45 46 and severe COVID-­19 are too low. It is possible that Participants in the current study will not be tested for HCWs may become less likely to contract SARS-­CoV-­2 latent TB prior to inclusion as this does not predict the as a result of improved preventive practices, reduced development of local skin reactions, abscesses or axillary community transmission and/or following the avail- lymphadenitis.36 Of relevance for participants enrolled ability of COVID-­ 19 vaccines. In addition, once during stage 1, there is no additional risk for co-­adminis- COVID-­19-­specific vaccines become available, interest tration of influenza and BCG vaccines.38 47 in participating in the current trial might wane. We While BCG has not been shown to cause fetal damage, have therefore chosen diverse recruitment settings the use of live-­attenuated vaccines is contraindicated in and included secondary outcome measures to eval- pregnancy. Therefore, women of childbearing potential uate the impact of BCG vaccination on other illnesses who think they could be pregnant or are planning to and infections (febrile and/or respiratory symptoms, become pregnant within the next month are not eligible. herpes simplex virus reactivation) and on the immune Participants are asked to do a pregnancy test if they have system overall (immunological studies). any doubt, and encouraged to do so in the privacy of their homes. In some regions, completing a pregnancy test will Perspective be an eligibility requirement and test kits will be made This trial is designed to determine whether BCG vaccina- available at recruitment. tion can reduce the incidence and/or severity of illness There is a hypothetical risk that BCG-­induced trained caused by SARS-­CoV-­2 infection. It also aims to provide immunity could increase symptoms in those who contract data on the ability of BCG vaccination to reduce the SARS-­CoV-­2, leading to a higher incidence of symptom- overall rate and/or severity of febrile and respiratory atic COVID-­ 19 in the BCG-­ vaccinated compared with illnesses in adults. This is particularly important for viral the non-­vaccinated group. However, even if this occurs, outbreaks that coincide with the winter influenza season, BCG might still be associated with a lower risk of severe and could help to reduce the overall strain on the health- COVID-­19 and hospitalisation as a result of a trained care system. If the hypothesis of a beneficial effect of BCG immune response reducing the viral load or clearing vaccination is correct, then this vaccine could be imple- the infection faster. In a recent report, administration mented as an early intervention to protect HCWs and 6 Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 Open access other high-­risk groups in future novel respiratory virus Dominguez Castellano; Glauce Dos Santos; Joyce dos Santos Lencina; Débora dos outbreaks. Santos Silva; Mark Douglas; Ross Dunn; Andrew Dunn; Jemma Dunnill; Georgina Eden; Harriet Edmund; Nat Eiffler; Hannah Elborough; Sonja Elia; Olivia Elkington; Michelle England; Wellyngthon Espindola Ayala; Maria Esteve; Nick Evans; Sue Dissemination Evans; Krist Ewe; María Carmen Fariñ Álvarez; Kieran Fahey; Jill Fairweather; The trial protocol is registered at ​ ClinicalTrials.​ gov Denise Faustman; Erica Fernandes Silva; Monique Fernandez; Galina Fidler; P.M.G. (NCT04327206) and is available in the online supple- Filius; Adam Finn; Carolyn Finucane; Stephanie Firth; Emily Fletcher; Catherine Flynn; Lorraine Flynn; Liam Fouracre; Sarah Fowler; Thamires Freitas; Ana Carolina mental material. Dissemination of the findings is Furtado; Maria Gabriela Oliveira; Anna Gabriela Santos; Leandro Galdino Cavalcanti planned, regardless of the results, through the WHO, Gonçalves; Laura Galletta; Larissa Gama; Dinusha Gamage; Radhika Ganpat; Carlos in peer-­ reviewed journals and at scientific confer- García; Mariana Garcia Croda; Kaya Gardiner; Evangeline Gardiner; Grace Gell; Aline ences. Once the database is cleaned and locked, it will Gerhardt de Oliveira; Susie Germano; Michael Gibbons; Camille Gibson; Alison Gifford; Teresa Giménez Poderos; Ann Ginsberg; Jet Gisolf; Bojana Gladanac; Penny be deposited in a data sharing repository archiving Glenn; Vanessa Godinho; Mayara Góes dos Santos; Josune Goikoetxea; Telma system. Access to the data will follow the rules of the Goldenberg; Adriano Gomes; Susana Gonzalez Marcos; Claudia González Rico; repository system. Casey Goodall; Louise Goodchild; Victoria Gordon; Frances Greven; Ana Greyce Capella; Liddy Griffith; Christina Guo; David Gutierrez Campos; Manuel Gutierrez Cuadra; Amanda Gwee; Richard Hall; Lydia Hall; Kate Hamilton; Matthew Hannan; Public involvement statement Houda Harbech; Alex Harding; Neil Harker; Robert Harrison; Robert Jan Hassing; The trial participants comprise only HCWs. The Thaynara Haynara Souza da Rosa; Zaheerah Haywood; Christine Heath; Nadine BRACE trial investigators include a number of HCWs Henare; Paulo Henrique Andrade; Susan Herrmann; Erin Hill; Sam Hilton; Danique and therefore representatives of the target popula- Huijbens; Heidi Hutton; Jane James; Tenaya Jamieson; Axel Janssen; Bruno Jardim; Tyane Jardim; Lance Jarvis; Narelle Jenkins; Jane Jones; Jan Jones; Karen Jones; tion were heavily involved in the design, management Leticia Jorge; Maria Jose Rios Vilegas; Sri Joshi; Rosemary Joyce; Joel Junior; and conduct of the trial. Most of the trial steering Rama Kandasamy; Anushka Karunanayake; Hana Karuppasamy; Tom Keeble; committee and the data safety and monitoring board Jennifer Kent; Paul Kloeg; Jan Kluytmans; Bridget Knight; Tobias Kollmann; Tony (DSMB) members are also HCWs. The results of the Korman; Ann Krastev; Meredith Krieg; Nathan La; Marcus Lacerda; Alicia Lacoma; Renier Lagunday; Debbie Lalich; Erin Latkovic; Irene Latorre; Paulo Leandro Garcia trial will be disseminated to the trial participants and Meireles Junior; Katherine Lee; Donna Legge; Toos Lemmers; Titia Leurink; the HCW community. Katherine Lieschke; Kee Lim; Gemma Lockhart; Cíntia Lopes Bogéa; Karla Lopes dos Santos; Reyes Lopez Marques; Michaela Lucas; David Lynn; Miriam Lynn; Correction notice This article has been corrected since it first published. Members Maria Luciana Silva De Freitas; Norine Ma; Sam Macalister; Cristiane Machado; of 'the BRACE trial Consortium Group' have been updated. Matheus Machado Ramos; Francesca Machingaifa; Ivan Maia; Bernardo Maia; Twitter Laure F Pittet @PittetLaure, Nicole L Messina @immunity_lab, Samantha Richard Malley; Laurens Manning; Sarah Manton; Jose Manuel Carrerero; Ana Bannister @samibannister, Marc Bonten @MarcBonten, John L Campbell @ Maria Barriocanal; Cíntia Maria Lopes Alves; Rosa Maria Plácido Pereira; Bianca profjcampbell, Julio Croda @JolioCroda, Susie Germano @GermanoSusie, Amanda Maria Silva Menezes Arruda; Adriana Marins; Angela Markow; Helen Marshall; Gwee @GweeAmanda, Bruno Jardim @BrunoJardim89, Katherine J Lee @ Christopher Martin; Katya Martinez Almeida; Wayne Mather; Megan Mathers; Fábio Katherine_J_Lee, David J Lynn @DavidJohnLynn, Ellie McDonald @EllieMcDonald, Mauricio Nogueira Gomes; Mariana Mayumi Tadokoro; Nadia Mazarakis; Kelry Kirsten P Perrett @PerrettKirsten, Cristina Prat Aymerich @crisprat2010, Peter Mazurega; Sonia McAlister; Amy McAndrews; Ellie McDonald; Fiona McDonald; Richmond @PeterCRichmond and Nigel Curtis @nigeltwitt Rebecca McElroy; Mark McMillan; Brendan McMullan; Nick McPhate; Lee Mead; Andrea Meehan; Bob Meek; Rosangela Melo; Guillermo Mena; Daniella Mesquita; Acknowledgements The authors thank the large number of people involved Nicole L Messina; Isabella Mezzetti; Hugo Miguel Ramos Vieira; Skye Miller; Kirsten in establishing the BRACE Trial (see BRACE Trial Consortium Group list in the Mitchell; Marcus Mitchell; Jesutofunmi Mojeed; Kitty Molenaar; Gemma Molina; supplementary material), as well as Kathryn North, Ann Ginsberg and the Orygen Barbara Molina; Lisa Montgomery; Cecilia Moore; James Moore; Simone Moorlag; team. Thilanka Morawakage; Julie Moss; Will Moyle; Kim Mulholland; Craig Munns; Collaborators The BRACE trial Consortium Group: Carolinne Abreu; Veronica Elizandra Nascimento; Nicolas Navarrette; Mihai Netea; Juliana Neves; Georgina Abruzzo; Lynne Addlem; Sophie Agius; Adelita Agripina Refosco Barbosa; Ahmed Newman; Belle Ngien; Jill Nguyen; Khanh Nguyen; Fran Noonan; Wendy Norton; Alamrousi; Ayla Alcoforado da Silva dos Santos; Yasmeen Al-­Hindawi; Samyra Melissa O’Donnell; Jess O’Bryan; Abby O’Connell; Sasha Odoi; Liz O’Donnell; Almeida Da Silveira; Lais Alves da Cruz; Jeremy Anderson; Christina Anthony; Roberto Oliveira; Marilena Oliveira; Thais Oliveira; Ingrid Oliveira; Nadia Olivier; Ligia Andrea Antonia Souza de Almeida dos Reis Pereira; Francisco Arnaiz de las Revillas Olivio; Benjamin Ong; Jaslyn Ong; Joanne Ong; Jakob Onysk; Isabelle Ooi; Frances Almajano; Annabelle Arnold; Beth Arrowsmith; Kristy Azzopardi; Cristina Badia Oppedisano; Francesca Orsini; Belinda Ortika; Orygen Group; Arthur Otsuka; Kristen Marti; Twinkle Bahaduri; Samantha Bannister; Sarah Barney; Lydia Barrera; Anabel Overton; Rosie Owens; Rayssa Paes; Pamela Palasanthiran; Virginia Palomo Barriocanal; Dayanne Barros; Simone Barry; Adam Bartlett; Lilian Batista Silva Jiménez; Girlene Pandine; Kimberley Parkin; Alvaro Pascual Hernandez; Nienke Muranaka; Therese Baulman; Morgan Bealing; Justin Beardsley; Ana Belen Martin Paternotte; David Paterson; Ana Paula Conceição de Souza; Lisa Pelayo; Casey Pell; Gutierrez; Jason Bell; Saoirse Benson; Vicki Bennett-­Wood; Nikki Bergant; Fabiane Sille Pelser; Handerson Pereira; Gabrielle Pereira; Glady Perez; Cristina Perez; Bianca Barbosa; Wouter Bijllaardt; Patricia Bimboese; Camila Bitencourt de Tomás Perez Porcuna; Susan Perlen; Kirsten Perrett; Amandine Philippart De Floy; Andrade; Stephen Blake; Kitty Blauwendraat; Wim Boersma; Pilar Bohedo Garcia; Sigrid Pitkin; Laure F Pittet; R.C. Pon; Ines Portillo Calderón; Jeffrey Post; Catherine Rhian Bonnici; Marc Bonten; Anne Boon; Anna Bourke; Kirsty Bowes; Larissa Brasil; Power; Christiane Prado; Endriaen Prajitno; Cristina Prat-­Aymerich; Lieke Preijers; Clare Brophy; Rochelle Botten; Sandy Buchanan; Jess Bucholc; Alison Burns; Emma Marco Puga; Evelyn Queiroz; Lynne Quinn; Ashleigh Rak; Leticia Ramires Burrell; Natalia Bustos; Bridie Byrne; Anthony Byrne; Esther Calbo; Jorge Calvo Figueiredo; Encarnacion Ramirez de Arellano; Pedro Ramos; Karla Regina Montes; Beatriz Camesella; John Campbell; Atsegiñe Cangas; John Carlin; Maria Warszawski de Oliveira; Jack Ren; Stephanie Reynolds; Shelley Rhodes; Carmen Roque; Roberta Carolina Pereira Diogo; Estela Carvalho; Irma Casas; Erika Claudinalva Ribeiro dos Santos; Chris Richards; Peter Richmond; Holly Richmond; Castro; Ramon Castro; Helen Catterick; Rodrigo Cezar Dutra Escobar; Joyce Chan; Ana Rita Lopes Souza; Jorge Rocha; Teresa Rodrigues; Laleyska Rodrigues; Bebeto Jo Cheah; Tee Yee Chern; Thilakavathi Chengodu; Marianna Ciaverella; Sharon Rodrigues; Iara Rodrigues Fernandes; Jesús Rodríguez-­Baño; Nienke Roescher; Clark; Marie-­Alix Clement Espindola; Annie Cobbledick; Clinton Colaco; Simone Sally Rogers; Anke Rol; Jannie Romme; Antoni Rosell; Maria Roser Font; Domenic Collopy; Patricia Comella; Mary Corbett; Gabriela Correa E Castro; Erlane Costa; Sacca; Sonia Sallent; Vanderson Sampaio; Nuria Sanchez; Blanca Sanchez; Daniel Raquel Coya; Nigel Crawford; Julio Croda; Alda Cruz; Nigel Curtis; Jac Cushnahan; Santos; Tilza Santos; Ariandra Sartim; Amber Sastry; Alice Sawka; Nikki Schultz; Anna Czajko; Renato da Costa Silva; Bouchra Daitiri; Margareth Dalcolmo; Karen Clare Seamark; David Seamark; Engelien Septer-­Bijleveld; Raquel Serrano; Frank Dalton; Aiken Dao; Andrew Davidson; Phoebe Dawe; Diane Dawson; Miriam de Shann; Ketaki Sharma; Margaret Shave; Lisa Shen; Kate Sidaway-­Lee; Adrian Siles Jesus Costa; Karina De La Cruz; Almudena de la Serna; Fabiani de Morais Batista; Baena; Rafaela Silva; Juliana Silva; Emanuelle Silva; Mariana Simão; Ronita Singh; Adriely de Oliveira; Rocio del Alba Rey Morillo; Maria Desylva; Helga Dijkstra; Rachel Marilda Siqueira; Marciléia Soares D.Allão Chaves; Thijs Sondag; Enoshini Dixon; Maria Dolores del Toro Lopez; Jose Dominguez; Catriona Doran; Angel Sooriyarachchi; Antonny Sousa; Leena Spry; Sarah Statton; Andrew Steer; Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 7 Open access Dyenyffer Stéffany Leopoldina dos Santos; Katrina Sterling; Leah Steve; Luke ORCID iDs Stevens; Natalie Stevens; Carolyn Stewart; Lisa Stiglmayer; Lida Stooper; Josephine Laure F Pittet http://​orcid.​org/​0000-​0002-​2395-​4574 Studham; Kanta Subbarao; Eva Sudbury; Astrid Suiker; Lorrie Symons; Esther Taks; Nicole L Messina http://​orcid.​org/​0000-​0001-​8404-​4462 Niki Tan; Bruna Tayara Leopoldina Meireles; Menno te Riele; Jaap ten Oever; Kaya Gardiner http://​orcid.​org/​0000-​0001-​9796-​4567 Guilherme Teodoro de Lima; Rob ter Heine; Jhenyfer Thalyta Campos Angelo; Helen Samantha Bannister http://​orcid.​org/​0000-​0001-​7507-​3704 Thomson; Ryan Toh; Alexandre Trindade; Harry Tripp; Enriqueta Tristán; Darren John L Campbell http://​orcid.​org/​0000-​0002-​6752-​3493 Troeman; Alexandra Truelove; Daniel Tsuha; Marlot Uffing; Fernando Val; Olga Julio Croda http://​orcid.​org/​0000-​0002-​6665-​6825 Valero; Ester Valls; Chantal van de Ven; Leo Van Den Heuvel; Sigrid van der Veen; Margareth Dalcolmo http://​orcid.​org/​0000-​0002-​6820-​1082 Marije van der Waal; J.H. van Leusen; Linda van Mook; H. van Onzenoort; Sonja Elia http://​orcid.​org/​0000-​0002-​6368-​189X Marjoleine van Opdorp; Miranda van Rijen; Nicolette van Sluis; Adria Vasconcelos; Susie Germano http://​orcid.​org/​0000-​0001-​9924-​7545 Noelia Vega; Sunitha Velagapudi; Louise Vennells; Tamsin Venton; Harald Verheij; Amanda Gwee http://​orcid.​org/​0000-​0003-​4016-​8986 P.M. Verhoeven; Caroliny Veron Ramos; Paulo Victor Rocha da Silva; Sandra Vidal; Bruno Jardim http://​orcid.​org/​0000-​0003-​3465-​7556 Patricia Vieira; Matheus Vieira de Oliveira; Rosario Vigo Ortega; Paola Villanueva; Tobias R Kollmann http://​orcid.​org/​0000-​0003-​2403-​9762 Raquel Villar; Amanda Vlahos; Ushma Wadia; Mary Walker; Kate Wall; Rachael Marcus Vinícius Guimarães Lacerda http://​orcid.​org/​0000-​0003-​3374-​9985 Wallace; Justin Waring; Ruth Warren; Adilia Warris; Emma Watts; Michelle Katherine J Lee http://​orcid.​org/​0000-​0002-​8334-​0291 Wearing-­Smith; Daniel Webber-­Rookes; Jamie Wedderburn; Ashleigh Wee-­Hee; Michaela Lucas http://​orcid.​org/​0000-​0001-​8881-​9990 Steve Wesselingh; Jia Wei Teo; Bethany Whale; Phoebe Williams; Beatrijs Wolters; David J Lynn http://​orcid.​org/​0000-​0003-​4664-​1404 Nick Wood; Ivy Xie; Angela Younes; Angela Young; Felipe Zampieri Vieira Batista; Ellie McDonald http://​orcid.​org/​0000-​0003-​2295-​4641 Carmen Zhou; Vivian Zwart. Laurens Manning http://​orcid.​org/​0000-​0003-​4334-​5351 Craig F Munns http://​orcid.​org/​0000-​0001-​5898-​5808 Contributors NC is the lead investigator and is responsible for study conception, Kirsten P Perrett http://​orcid.​org/​0000-​0002-​5683-​996X design and funding acquisition. NC, AD, KS, LFP, NLM, KG, FS, KPP, AG, MB, JCa, Cristina Prat Aymerich http://​orcid.​org/​0000-​0001-​6974-​9165 JCr, MD, BJ, TRK, MVGL, ML, DJL, LM, CFM, CPA, PR, NJW, were involved in study Peter Richmond http://​orcid.​org/​0000-​0001-​7562-​7228 design and outcome definition. SB, ES, PV, KPP, LFP, NC critically reviewed the Frank Shann http://​orcid.​org/​0000-​0002-​9899-​1804 literature for assessing the safety of BCG revaccination. KLi, LFP, NLM, FO, KLe, Paola Villanueva http://​orcid.​org/​0000-​0002-​6504-​2979 CG, EM, SB, PV, ES, KG, TJ, KPP, AD, NC prepared the ethics application and all Kanta Subbarao http://​orcid.​org/​0000-​0003-​1713-​3056 other authors provided critical evaluation and revision. KG, VA, CG developed the Nigel Curtis http://​orcid.​org/​0000-​0003-​3446-​4594 recruitment methods. EM, CG, VA, LFP developed the questionnaire, smartphone application and trial database. SE, DL developed procedures for administration of the investigational product. NLM, SG designed the sample collection, processing methods and sample database. FO, KLe, LFP drafted the statistical analysis plan. LFP drafted the manuscript, coordinated manuscript preparation and revision. REFERENCES All authors provided critical evaluation and revision of the manuscript, and have 1 World Health Organisation. Report of the WHO-­China joint mission approved the final version of this manuscript. on coronavirus disease 2019 (COVID-­19). World Health Organisation, 2020: 40. Funding This trial is supported by the Bill & Melinda Gates Foundation (INV-­ 2 Anderson RM, Heesterbeek H, Klinkenberg D, et al. How will country-­ 017302), the Minderoo Foundation (COV-­001), Sarah and Lachlan Murdorch, the based mitigation measures influence the course of the COVID-­19 Royal Children's Hospital Foundation (2020-­1263 BRACE Trial), Health Services epidemic? Lancet 2020;395:931–4. Union NSW, the Peter Sowerby Foundation, the Ministry of Health Government 3 Gómez-­Ochoa SA, Franco OH, Rojas LZ, et al. COVID-­19 in health-­ of South Australia, the NAB Foundation, the Calvert-­Jones Foundation, the care workers: a living systematic review and meta-­analysis of prevalence, risk factors, clinical characteristics, and outcomes. Am J Modara Pines Charitable Foundation, the UHG Foundation Pty Ltd, Epworth Epidemiol 2021;190:161–75. Healthcare and individual donors. NC is supported by the National Health and 4 Chou R, Dana T, Buckley DI, et al. Epidemiology of and risk factors Medical Research Council (NHMRC) (Investigator Grant GNT1197117). LFP is for coronavirus infection in health care workers. Ann Intern Med supported by the Swiss National Science Foundation (Early Postdoc.Mobility 2020;173:120–36. grant number P2GEP3_178155). SB is supported by the University of Melbourne 5 Nguyen LH, Drew DA, Graham MS, et al. Risk of COVID-­19 among Research and Training Program scholarship and the Clifford Family scholarship. front-­line health-­care workers and the general community: a PV is supported by the Australian Government Research Training Program (RTP) prospective cohort study. Lancet Public Health 2020;5:e475–83. 6 Amesty International. Global: amnesty analysis reveals over 7,000 Scholarship administered by the University of Melbourne and Murdoch Children’s health workers have died from COVID-­19, 2020. Available: https:// Research Institute PhD Top-­Up Scholarship. The Melbourne WHO Collaborating www.​amnesty.​org/​en/​latest/​news/​2020/​09/​amnesty-​analysis-​7000-​ Centre for Reference and Research on Influenza is supported by the Australian health-​workers-​have-​died-​from-​covid19/ [Accessed 24 Sep 2020]. Government Department of Health (Health Protection Program Agreement # 7 Pollard AJ, Finn A, Curtis N. Non-­specific effects of vaccines: 4-­4GFBE88). plausible and potentially important, but implications uncertain. Arch Dis Child 2017;102:1077–81. Competing interests None declared. 8 Hollm-­Delgado M-­G, Stuart EA, Black RE. Acute lower respiratory Patient consent for publication Not applicable. infection among Bacille Calmette-­Guérin (BCG)-­vaccinated children. Pediatrics 2014;133:e73–81. Provenance and peer review Not commissioned; externally peer reviewed. 9 de Castro MJ, Pardo-­Seco J, Martinón-­Torres F. Nonspecific (heterologous) protection of neonatal BCG vaccination against Supplemental material This content has been supplied by the author(s). It hospitalization due to respiratory infection and sepsis. Clin Infect Dis has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have 2015;60:1611–9. been peer-­reviewed. Any opinions or recommendations discussed are solely 10 Arts RJW, Moorlag SJCFM, Novakovic B, et al. BCG vaccination those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability protects against experimental viral infection in humans through the and responsibility arising from any reliance placed on the content. Where the induction of cytokines associated with trained immunity. Cell Host content includes any translated material, BMJ does not warrant the accuracy and Microbe 2018;23:89–100. reliability of the translations (including but not limited to local regulations, clinical 11 Nemes E, Geldenhuys H, Rozot V, et al. Prevention of M. tuberculosis guidelines, terminology, drug names and drug dosages), and is not responsible Infection with H4:IC31 Vaccine or BCG Revaccination. N Engl J Med 2018;379:138–49. for any error and/or omissions arising from translation and adaptation or 12 Wardhana D, Datau EA, Sultana A. The efficacy of Bacillus Calmette-­ otherwise. Guerin vaccinations for the prevention of acute upper respiratory Open access This is an open access article distributed in accordance with the tract infection in the elderly. Acta Med Indones 2011;43:185–90. Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits 13 Giamarellos-­Bourboulis EJ, Tsilika M, Moorlag S, et al. Activate: randomized clinical trial of BCG vaccination against infection in the others to copy, redistribute, remix, transform and build upon this work for any elderly. Cell 2020;183:315–23. purpose, provided the original work is properly cited, a link to the licence is given, 14 Moorlag SJCFM, Arts RJW, van Crevel R, et al. Non-­specific and indication of whether changes were made. See: https://​creativecommons.​org/​ effects of BCG vaccine on viral infections. Clin Microbiol Infect licenses/​by/​4.​0/. 2019;25:1473–8. 8 Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 Open access 15 Pittet LF, Curtis N. Does Bacillus Calmette-­Guérin vaccine prevent Info​rmation/​Guidances/​Vaccines/​ucm091977.​pdf [Accessed 8 Apr herpes simplex virus recurrences? A systematic review. Rev Med 2020]. Virol 2021;31:e2151. 33 Royston P. Flexible parametric alternatives to the COX model: 16 Biering-­Sorensen S, Aaby P, Lund N. Early BCG-­Denmark update. Stata J 2004;4:98–101. and neonatal mortality among infants weighing. Clin Infect Dis 34 BRACE statistical analysis plan for interim analysis. Available: 2017;65:1183–90. https://​doi.​org/​10.​25374/​MCRI.​14721309.​v1 [Accessed 13 Oct 17 Higgins JPT, Soares-­Weiser K, López-­López JA, et al. Association 2021]. of BCG, DTP, and measles containing vaccines with childhood 35 Australian Government Department of Health. The BCG mortality: systematic review. BMJ 2016;355:i5170. vaccine: information and recommendations for use in Australia. 18 Novakovic B, Messina N, Curtis N. Chapter 6 - The Heterologous communicable diseases intelligence. Effects of Bacillus Calmette-­Guérin (BCG) Vaccine and Trained 36 Lotte A, Wasz-­Höckert O, Poisson N, et al. BCG complications. Innate Immunity. In: Faustman DL, ed. The value of BCG and TNF in estimates of the risks among vaccinated subjects and autoimmunity. 2nd ed. Academic Press, 2018: 71–90. statistical analysis of their main characteristics. Adv Tuberc Res 19 Kleinnijenhuis J, Quintin J, Preijers F, et al. Bacille Calmette-­Guerin 1984;21:107–93. induces NOD2-­dependent nonspecific protection from reinfection via 37 Bothamley GH, Cooper E, Shingadia D. Tuberculin testing before epigenetic reprogramming of monocytes. Proc Natl Acad Sci U S A BCG vaccination. BMJ 2003;327:243–4. 2012;109:17537–42. 38 Australian Technical Advisory Group on Immunisation (ATAGI). 20 Kleinnijenhuis J, Quintin J, Preijers F, et al. BCG-­induced trained Australian immunisation Handbook. Canberra: Australian immunity in NK cells: role for non-­specific protection to infection. Clin Government Department of Health, 2018. Immunol 2014;155:213–9. 39 Hendry AJ, Dey A, Beard FH, et al. Adverse events following 21 Freyne B, Donath S, Germano S, et al. Neonatal BCG vaccination immunisation with Bacille Calmette-­Guérin vaccination: baseline influences cytokine responses to Toll-­like receptor ligands and data to inform monitoring in Australia following introduction heterologous antigens. J Infect Dis 2018;217:1798–808. of new unregistered BCG vaccine. Commun Dis Intell Q Rep 22 Freyne B, Messina NL, Donath S, et al. Neonatal BCG vaccination 2016;40:E470–4. reduces interferon-γ responsiveness to heterologous pathogens 40 New Zealand Medicines and Medical Devices Safety Authority. BCG in infants from a randomized controlled trial. J Infect Dis vaccine SSI data sheet 14 March 2018. Available: https://www.​ 2020;221:1999–2009. medsafe.​govt.​nz/​profs/​datasheet/​b/​BCGCSLinj.​pdf [Accessed 24 23 Netea MG, Domínguez-­Andrés J, Barreiro LB, et al. Defining trained Dec 2019]. immunity and its role in health and disease. Nat Rev Immunol 41 Pilgrim S. Administration of the Bacillus Calmette Guerin (BCG) 2020;20:375–88. vaccination in neonates, 2019. Available: https://www.​meht.​nhs.​ 24 Kleinnijenhuis J, Quintin J, Preijers F, et al. Long-­lasting effects of uk/​EasysiteWeb/​getresource.​axd?​AssetID=​14766&​type=​full&​ BCG vaccination on both heterologous Th1/Th17 responses and servicetype=​Attachment [Accessed 24 Dec 2019]. innate trained immunity. J Innate Immun 2014;6:152–8. 42 Bannister S, Sudbury E, Villanueva P, et al. The safety of BCG 25 Walk J, de Bree LCJ, Graumans W, et al. Outcomes of controlled revaccination: a systematic review. Vaccine 2021;39:2736–45. human malaria infection after BCG vaccination. Nat Commun 43 Fine P, Carneiro I, Milstien J. Issues relating to the use of BCG 2019;10:874. in immunization programmes: a discussion document. Geneva: 26 Freyne B, Marchant A, Curtis N. BCG-­associated heterologous Department of Vaccines and Biologicals, World Health Organization, immunity, a historical perspective: intervention studies in animal 1999. models of infectious diseases. Trans R Soc Trop Med Hyg 44 Rodrigues LC, Pereira SM, Cunha SS, et al. Effect of BCG 2015;109:287. revaccination on incidence of tuberculosis in school-­aged children 27 Floc'h F, Werner GH. Increased resistance to virus infections of in Brazil: the BCG-­REVAC cluster-­randomised trial. Lancet mice inoculated with BCG (Bacillus Calmette-­Guérin). Ann Immunol 2005;366:1290–5. 1976;127:173–86. 45 Randomised controlled trial of single BCG, repeated BCG, or 28 Old LJ, Benacerraf B, Clarke DA, et al. The role of the combined BCG and killed Mycobacterium leprae vaccine for reticuloendothelial system in the host reaction to neoplasia. Cancer prevention of leprosy and tuberculosis in Malawi. Karonga prevention Res 1961;21:1281–300. trial group. Lancet 1996;348:17–24. 29 Netea MG, Giamarellos-­Bourboulis EJ, Domínguez-­Andrés J, et al. 46 Roth AE, Stabell Benn C, Ravn H, et al. Effect of revaccination with Trained immunity: a tool for reducing susceptibility to and the BCG in early childhood on mortality: randomised trial in Guinea-­ severity of SARS-­CoV-­2 infection. Cell 2020;181:969–77. Bissau. BMJ 2010;340:c671. 30 Curtis N, Sparrow A, Ghebreyesus TA, et al. Considering 47 Leentjens J, Kox M, Stokman R, et al. BCG vaccination enhances BCG vaccination to reduce the impact of COVID-­19. Lancet the immunogenicity of subsequent influenza vaccination in healthy 2020;395:1545–6. volunteers: a randomized, placebo-­controlled pilot study. J Infect Dis 31 Harris PA, Taylor R, Thielke R, et al. Research electronic data capture 2015;212:1930–8. (REDCap)--a metadata-­driven methodology and workflow process 48 Padmanabhan U, Mukherjee S, Borse R. Phase II clinical trial for providing translational research informatics support. J Biomed for evaluation of BCG as potential therapy for COVID-­19, 2020. Inform 2009;42:377–81. Available: https://www.​medrxiv.​org/​content/​medrxiv/​early/​2020/​11/​ 32 Food and Drug Administration. Guidance for industry: toxicity 03/​2020.​10.​28.​20221630.​full.​pdf grading scale for healthy adult and adolescent volunteers enrolled 49 Moorlag SJCFM, van Deuren RC, van Werkhoven CH, et al. Safety in preventive vaccine clinical 2007. Available: https://www.​fda.​gov/​ and COVID-­19 symptoms in individuals recently vaccinated with downloads/​Biol​ogic​sBlo​odVa​ccines/​Guid​ance​Comp​lian​ceRe​gula​tory​ BCG: a retrospective cohort study. Cell Rep Med 2020;1:100073. Pittet LF, et al. BMJ Open 2021;11:e052101. doi:10.1136/bmjopen-2021-052101 9