Introduction

Close to 7 million deaths from COVID-19 were reported globally since January 2020, but recent estimates suggest that the real death toll may be as high as 18 million.1 In Costa Rica (population of 5.1 million), the first case was diagnosed in March 2020; and as of July 2023, there have been around 1.24 million cases reported, with close to 9500 deaths.2 An analysis from the RESPIRA cohort described in this manuscript indicates that the true incidence of COVID-19 is four times higher than reported, suggesting that the majority of people in Costa Rica have been infected.3

COVID-19 has a wide spectrum of clinical manifestations,4 mainly in the respiratory tract. Most cases are mild or even asymptomatic, but approximately 20% of patients infected with the more severe variants have inflammatory responses that can lead to acute respiratory distress, multiorgan failure and death.5–7 The SARS-CoV-2 virus has undergone extensive adaptive evolution since the start of the pandemic and multiple new variants with altered transmissibility and virulence impact the efficacy of public health measures, emphasising the need for continued surveillance and research.8 9

Extensive research is underway to better understand the immune response to the virus.10 11 An early innate response can induce resolution of disease, while a late response can trigger a cytokine storm that aggravates clinical evolution.12 Serological tests13 can provide information on the pathogenesis of SARS-CoV-2, the intensity of transmission and the susceptibility to infection after natural infection and vaccination. The nature and evolution of the immune response over time and their role in herd immunity require further investigation.

Costa Rica has a public health system providing medical services to 91.8% of the population14 characterised by a vast network from primary care clinics to university hospitals, with unified statistics and disease surveillance systems. In addition, a clear regulatory framework for research is established by law. For COVID-19, a centralised mandatory surveillance system managed by the Ministry of Health and the Social Security is in place, with generation of detailed statistics. Vaccination against COVID-19 started in December 2020 initiating with people over 58 years old, health personnel and high-risk groups, and progressively extending to other ages.15 By November 2022, 96% of the eligible population older than 12 years had received one dose, 91% two doses and 62% three doses.16 Progression of vaccination by age group in Costa Rica is presented in online supplemental figure S1.

In this context, we are conducting the RESPIRA Study (Evaluation of the immune response to SARS-CoV-2 in Costa Rica) in a collaboration between the Costa Rica Social Security System (CCSS), the Health Ministry (MS), the Costa Rican Agency for Biomedical Research (ACIB-FUNIN), the University of Costa Rica (UCR), the USA National Institutes of Health (NIH) and the German Cancer Research Center (DKFZ), to investigate multiple aspects of the COVID-19 pandemic in the country, with special emphasis on the immune response. In the present manuscript, we discuss details of recruitment and the initial 18 months of a planned 2 years of follow-up.

Cohort descriptionMethodsOverall objective and design

RESPIRA is a population-based cohort study with a 2-year active follow-up period. A cohort of COVID-19 cases and a cohort of matched population-based controls were assembled at recruitment. The main objective is to describe the nature and magnitude of the immune response to SARS-CoV-2 infection and vaccination, its time course, epidemiological and genetic determinants, and protective efficacy against subsequent infection. In addition, RESPIRA will estimate population prevalence of infection over time and investigate determinants of acquisition, clinical presentation, long-COVID and household transmission.

The goal of the study was to enrol and follow 1000 cases confirmed by PCR testing, and 2000 community controls, balanced to the case group on age, sex and geographical area of residence. Controls were selected regardless of their history of COVID-19 to study prevalence of infection in the general population and to evaluate the protective efficacy of past infection and antibodies against subsequent infection. For the cases, we planned a 70% enrolment of participants already recovered from COVID-19 and 30% with a diagnostic PCR sample taken for COVID-19 in the previous 14 days. The latter were included to study the early immune response to the virus.17 RESPIRA includes a 2-year follow-up, with an in-person or telephone interview on COVID-19 symptoms every month and collection of saliva and blood samples at varying times during the study. To evaluate household transmission, a random sample of the RESPIRA cases diagnosed within 30–60 days of recruitment (index cases) was selected, and, if they accepted, members of their household were invited to participate in the household survey.

Study population and recruitment

Costa Rica comprises 7 provinces with 472 districts (2011 census). The recruitment areas of the study were selected considering logistic feasibility and excluded areas with difficult access by car or far from headquarters. Business districts were also excluded. A total of 192 districts distributed in the great metropolitan area (151 districts), Puntarenas province (11 districts) and Upala, hereafter referred to as GAM-Puntarenas-Upala, and the Guanacaste province (30 districts) were included, corresponding to 41% of the districts in the GAM-Puntarenas-Upala region and 50% in Guanacaste. Each district is divided into minimal geostatistical units (MGUs), used to define control selection areas, matched to the MGU of the case (online supplemental figure S2).

Recruitment was conducted through home visits to the selected cases and potential controls by teams comprising doctors/nurses and drivers/field workers. Trained outreach workers conducted the census in the MGUs to identify the controls according to protocol.

Selection of the cohort of cases

A series of stratified random samples of cases diagnosed by PCR from nasopharyngeal swabs (NPS) between 7 March 2020 and 6 July 2021 were selected from the surveillance lists of the CCSS and the Ministry of Health. To assess the time course of the antibody response by comparing the antibody levels among subjects diagnosed several months in the past with cases diagnosed more recently, during the first 12 months of recruitment we selected 50 cases diagnosed in each month since the beginning of the pandemic in Costa Rica. Subsequently, the number of monthly cases was defined according to logistic feasibility. For each random sample, replacement samples were selected concurrently and used when the original case was ineligible or refused participation in the study.

In Costa Rica, definition of COVID-19 cases initially required confirmatory PCR tests, but this was later (July 2020) changed to include cases diagnosed by ‘epidemiologic link’, defined as subjects with COVID-19-compatible symptoms and recent contact with a PCR confirmed case.17 Cases reported using these criteria were not included in RESPIRA.

Selection and recruitment of cases was from 2 November 2020 to 4 September 2021 with a ratio of 70% for GAM-Puntarenas-Upala and 30% for Guanacaste. The samples were stratified to obtain similar numbers in four age groups: 0–19, 20–39, 40–59 and 60 or more. The 0–19 age group was further stratified into 0–9 and 10–19 years.

Two groups of cases were selected: Active cases, included to investigate the early immune response to infection, were randomly selected within each sample to be recruited ≤14 days from the date of collection of the PCR specimen that led to the diagnosis. Recovered cases were randomly selected to be recruited >14 days from the date of collection of NPS.

Our random group of cases was mainly composed of outpatients (95%). To investigate determinants of disease severity, the group of cases was expanded to include a convenience sample of 200 additional hospitalised cases from the same catchment areas as RESPIRA recruited during hospitalisation or at home after discharge. A similar protocol was followed but no additional controls were selected for these cases.

Eligibility criteria for cases included diagnosis by PCR, willingness to donate a blood specimen, plans to remain in the country for the next 12 months, ability to communicate and absence of incapacitating medical conditions.

Selection of cohort of population controls

Two community controls were selected for each participating case from 11 November 2020 to 8 October 2021, matched to cases on age (in the same decade), sex and MGU. A history of COVID-19 was not an exclusion criterion for inclusion as a control, as we aimed to enrol a cohort representative of the underlying population from which the cohort of cases derived. Thus, a fraction (about 21.5%) of the controls were in fact cases, as determined by their self-reported clinical history, documentation in the national surveillance lists or the presence of antibodies against the nucleocapsid (N) protein of the virus, which are present after natural infection but not after application of vaccines based on the S1-RBD antigen, as those used in Costa Rica. Such controls are included as cases or excluded for specific analyses, as appropriate. For control recruitment, the MGU of the case was selected, and trained field workers visited households sequentially in a defined pattern until control subjects with the same age group, sex and MGU as the original case were found. If no potentially eligible control subjects were found, a contiguous MGU within the district was selected. When no potential participants of the same sex or age stratum of the case were identified, the age range was extended to the adjacent decades to obtain the two controls for each case.

Eligibility criteria for controls included willingness to donate a blood specimen, plans to remain in the country for the next 12 months, ability to communicate and absence of incapacitating medical conditions.

Selection of participants for the household transmission study

To investigate household transmission, from the group of cases already included in RESPIRA, we randomly selected a group of participants (index cases) recruited shortly after diagnostic PCR, to allow recruitment of household members within a period of 30–60 days since the diagnosis of the index case. Since ascertainment of disease among family members was done by determining the presence of antibodies against the combination of SARS-CoV-2 nucleocapsid protein (N) and spike protein receptor-binding domain (S1-RBD), this period was chosen to allow antibodies to develop after infection. These cases were asked if they authorised study personnel to invite members of their household to participate in the study.

A household is defined as a group of people who live together and share at least one common space such as the kitchen. To be considered eligible for inclusion, a contact had to have spent at least one night a week in the home since the diagnosis of the case. The participants in this component received a single visit where a questionnaire was administered, and a blood sample was taken for SARS-CoV-2 antibody testing. Characteristics or behaviour of the index cases were considered determinants of transmission, while characteristics of the household members were considered determinants of susceptibility. The results of the household transmission study have been published.18

Enrolment data collection and management

At the enrolment visit, the study staff confirmed the identity of the participant by means of a valid identification document of the subject and legal guardians in case of minors and verified eligibility criteria before proceeding with enrolment. All participants or their legal representatives signed informed consent or assent (for children 12–17 years old) in the presence of a witness, as mandated by Costa Rican law.

Study data were collected and managed using Research Electronic Data Capture with electronic online data entry, adapted for the study and hosted at ACIB-FUNIN. (https://cctsi.cuanschutz.edu/resources/informatics/redcap-resources).

Enrolment visits were conducted at home or at one of the ACIB-FUNIN or CCSS facilities. Items in the interview included sociodemographic data, history of smoking and alcohol consumption, physical activity, influenza vaccination, personal health history, gynaecological and obstetric history for women older than 15 years, clinical presentation of COVID-19 for cases and clinical manifestations since the start of the pandemic for controls, in addition to behaviour and practices during the pandemic and vaccination against COVID-19, including type of vaccine and date of each dose.

For the household survey component, the questionnaire also ascertained behaviours of each household contact during the 15 days before and 15 days after the diagnosis of the index case, as well as questions about COVID-19-like symptoms and use of protective measures during that period. When COVID-19 vaccines became available, vaccination status was ascertained, with timing and number of doses.

Enrolment specimen collection

At the recruitment visit, the participants were given detailed instructions to collect 1 cc of saliva. In the first 5 months of the study, saliva was collected in an OMNIgene ORAL OM-505 vial (OMNIgene Medical Technologies, Malta) containing 1 cc of preservative. Subsequently, saliva was collected in Simport T310-10A vials containing no preservative. After collection, samples were labelled and stored in a cold box at a temperature between 2°C and 10°C, 1 cc aliquots were prepared using jacketed tricode tubes and transferred into a −80°C freezer within 12 hours.

Following saliva collection, a blood sample was collected, stored in cold boxes at a temperature between 2°C and 10°C, aliquoted into its component parts (serum and blood clot from red top tube collection; plasma and buffy coat from purple top tube collections) and transferred into a −80°C freezer within 12 hours. For participants older than 12 years, 10 cc were collected in a red-top tube containing no preservative and 4 cc in a purple-top tube containing EDTA; For children under 12 years of age, only one red-top tube was collected (5 cc if over 2 years of age and 3 cc if under 2 years of age). For participants in the household component of the study, red-top tubes only were used. (online supplemental table S1). Blood components were stored in jacketed tricode tubes (serum, plasma and buffy coat) or Nalgene cryovials with external thread (blood clot).

For saliva, two aliquots of 1 cc each were made, for blood in red tubes up to four aliquots of 1 cc of serum and an aliquot of blood clot of 1.5 cc were made (only for the household component and children under 12). For the blood in a purple-top tube, two aliquots of 1 cc plasma and an aliquot of buffy coat were made. All aliquots except clots were produced in jacketed tricode tubes to allow for sample tracking and automation capabilities. Blood clots were stored in 2 mL Nalgene tubes with external thread. The storage temperature of the aliquots was −150°C at the local repositories and −80°C at the central biobank. Any notes on preanalytical and analytical conditions were noted in the biological samples database (BSI-II system) and reported to the investigators.

Follow-up data collection and management

Two years of follow-up are planned for each participant in the study. All participants will be followed every 6 months with blood and saliva collection and exposure/behaviour and symptoms questionnaire administration. In addition, saliva is collected every 2 weeks for all. Finally, COVID-19 cases recruited during the active phase of their disease (<14 days from diagnosis) have weekly visits in the first month after recruitment. At every visit, participants are administered a COVID-19 symptom questionnaire.

At months 6, 12, 18, 24, an additional questionnaire that asks about persistent COVID-19 is administered, including the 36-item Short Form Survey Instrument (SF-36 questionnaire), with special emphasis on quality of life. The SF-36 questionnaire is a well-validated questionnaire that measures quality of life through 36 items and yields a profile of health status that is applicable to both healthy and ill people.19 It evaluates physical functioning, physical role, body pain, general health, vitality, social functioning, emotional status and mental health.

Follow-up specimen collection and management

The reverse transcriptase (RT)-PCR on NPS is still the gold standard to diagnose COVID-19.20 However, this sample is relatively hard to obtain, is unpleasant for the patient and can put the staff taking the sample at risk. Several studies have shown the sensitivity of saliva to be at least equivalent to the NPS.21 In addition, several studies demonstrate the stability of the saliva over extended periods of time (2–25 days) at different temperatures.22 In those studies that used sterile collection tubes without nucleic acid preservatives, the SARS-CoV-2 RNA was consistently detected at similar levels regardless of the holding time and the temperatures tested.23

Active cases (recruited within 2 weeks of NPS test) provide a blood and saliva specimen every week for 3 weeks. All cases provide blood every month for the first 5 months of the study and subsequently every 6 months until the end of the study. Until 28 June 2021, collection of saliva was performed monthly, but the frequency was thereafter increased to twice a month in order to maximise our ability to detect infections during follow-up, considering that the virus tends to become undetectable 10–15 days after infection.24 The sample obtained 15 days after the monthly visit is self-collected, without media and kept refrigerated at home until the monthly visit, when our staff collect it (online supplemental table S1).

Sample size considerations

The sample size for the control cohort (n2=2000) was calculated to obtain adequate statistical power to study the degree of protection conferred by neutralising antibodies. This power depends on the attack rate in controls during the 2-year follow-up and on the degree of protection, which was unknown. Nevertheless, various scenarios were explored, and the sample size was considered sufficient. In particular, with 80% complete follow-up, the power was at least 80% for proportions with new infections above 5% in the controls and relative risk 0.5 in cases with detectable antibodies compared with controls, or for proportions with new infections above 2% in controls and relative risk 0.25 compared with controls. Introduction of vaccination is likely to change the expected number of outcomes in the different groups. Power was estimated using the two-sample proportions test for unbalanced design (Pearson’s χ2 test). Power calculations were performed using the STATA v.18 function power two proportions with the n1 and n2 options. The 2:1 ratio of controls to cases was selected to gain power and increase the overall number of subjects in the cohort.

Initial antibody testing

Analysis of antibody responses to SARS-CoV-2 and common cold coronavirus proteins was performed at the DKFZ Heidelberg as described.25 Briefly, recombinantly expressed proteins of the SARS-CoV-2 proteome as well as N proteins and S1 protein receptor-binding-domain (RBD) of common cold coronaviruses were linked to the bead surface of fluorescently labelled polystyrene beads (Luminex, Austin, Texas, USA). Antigen-loaded beads were combined to achieve a suspension array that is incubated with serum sample (1:100 serum dilution). A Luminex 200 Analyzer (LuminexA) was used to distinguish the bead sets and their respective antigens, and to quantify the amount of serum antibody bound to the antigen by a reporter fluorescence.

For a subset of sera their functional capacity in terms of virus neutralisation was tested. For this purpose, a lentivirus-based pseudovirus neutralisation assay was employed using a codon-optimised SARS-CoV-2 (Wuhan Hu-1) spike protein for pseudotyping.26 The pseudovirus encoded the firefly luciferase gene used for quantitative measurement of virus entry into (ACE 2) ACE2-transduced A549 lung epithelial cells. Briefly, pseudoparticles were incubated with a 10-step 2-fold dilution series of test sera for 1 hour at 37°C before being added to ACE2-A549 cells. Twenty-four hours later, cells were lysed and luciferase activity was measured. Luciferase signals were standardised to 0% (uninfected cells)–100% (cells infected with pseudovirus without serum) and plotted over serum dilution. The 50% inhibitory titre was determined by logistic regression. For quality control, each batch included positive (recombinant human monoclonal anti-Spike antibody) and negative (prepandemic human serum) controls. The assay was validated by comparing defined sera against an independent authentic virus (SARS-CoV-2) neutralisation test. Technical details of the assay will be published elsewhere.

Initial PCR testing

Extraction of RNA was performed using the Qiasymphony DSP Virus/Pathogen kit on the Qiasymphony automated nucleic acid extraction system (Qiagen, Germany). Saliva specimens were thawed and 200 µL of the specimen was used, RNA was extracted to a final volume of 60 µL. RT-PCR amplification was performed using the Centers for Disease Control and Prevention (CDC) COVID-19 PCR assay described previously Applied Biosystems 7500 Fast Dx Real-Time PCR Instrument (Thermo Fisher Scientific) instrument.27 A cycle threshold value of ≤40 for any of the two targets (N1 or N2) was considered a positive result. A positive result for the RNAse-P target was necessary for a valid call.

For samples that tested positive, we performed a digital droplet PCR to quantify the amount of RNA for both targets in the nucleoprotein region (N1 and N2). This was performed using the 1-step RT-ddPCR Advanced Kit for Probes (Bio-Rad Catalog # 1864022), primers and probes were based on the CDC assay.28

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Findings to date

Recruitment of the original group of cases and controls into RESPIRA began November 2020 and finalised in October 2021, covering the initial period of the pandemic (figure 1). Recruitment of supplemental hospitalised cases continued until August 2022.

Figure 1

Number of cases of COVID-19 reported to the national surveillance system and periods of diagnosis and recruitment of cases and recruitment of controls in the RESPIRA study (Immune response to SARS-CoV-2 in Costa Rica).

The population of the districts included was 2.8 million (online supplemental figure S3), representing 52% of the population of Costa Rica in 2020. A total of 213 000 cases were reported during the case selection period. We selected 3181 PCR-confirmed cases and 1657 were contacted and deemed eligible, and 999 were recruited, for a 61% participation rate.

Of the 999, 279 (27.9%) were recruited as active and 720 (72.0%) as recovered cases. Seventy-five per cent of cases were from the original samples selected, 14% from the first replacement samples and 10% from additional replacement samples. The fraction of eligible subjects and refusals was similar for the different replacement samples. Of the 999 cases, a total of 434 were selected for the household study, of which 304 (90%) agreed to participate. A total of 719 household members from the 304 households were ultimately recruited, representing 80.5% of all eligible household members. In addition, 246 supplemental hospitalised cases were selected, of which 151 were recruited (71,6% response rate). For control selection, 2734 potential participants were selected and 1999 were recruited after exclusion of non-contacted (n=218) and refusals (n=517), for a participation rate of 72.8%. There are 999 cases and 1999 controls because one case had 3 controls recruited (by mistake). Compliance with follow-up to month 18 has been close to 70%.

Recruitment occurred according to the study design, with age and region stratification and case/control matching as planned (table 1). There was no sex stratification because reported incidence of COVID-19 by gender was similar in Costa Rica.

Table 1

Distribution of age, sex, region and date of diagnosis (as applicable) in the case and control cohorts

The median age of cases and controls was 41.3 years (IQR 25–58). The age-stratified study design compensated for the fact that most cases reported in the country were aged 20–59 years. We recruited somewhat lower numbers of subjects under 20 years of age and in the 60+ group. Stratification by region produced the expected distribution with 27% from the more rural (Guanacaste) area and 73% from the GAM-Puntarenas-Upala region, with adequate balance by Province.

Among controls, 180 (9%) had COVID-19 previous to recruitment according to the surveillance lists of the CCSS or to their recruitment interview, of these, only 117 had antibodies. An additional 250 controls (12.5%) had antibodies against the viral nucleocapsid and the S1-RBD protein, indicating previous COVID-19 infection. Exclusion of these groups produced a control group (n=1569) that remained comparable to the cases on the matching variables.

Recruitment of cases in relation to their date of diagnosis is also presented in table 1. It was defined by design, with 28% (active cases) in the first 2 weeks after diagnosis, and approximately 50 cases per month of diagnosis up to a year before recruitment, to evaluate retrospectively the duration of the antibody response. Overall, about 65% of cases were recruited within 6 months of diagnosis.

Approximately 88% of controls were recruited within 3 months after recruitment of the case (table 2). Recruitment of controls in relation to diagnosis of the cases was more distant in time, as the final months of the study were used to finalise recruitment of the control group.

Table 2

Timing of enrolment of controls since enrolment and since diagnosis of the cases

About 47% of cases had complete high school or more education. compared with about 40% of the controls (table 3). The number of people in the household had a comparable distribution between cases and controls, but cases were more likely to be employed. possibly as a result of this group seeking diagnostic confirmation given their need to justify their absences and their access to paid sick leave. Also, controls were searched at home, generating a greater possibility of finding unemployed people. Alternatively, employed people may have been more likely to be infected at work.

Table 3

Sociodemographic characteristics of the case and control cohorts

As shown in table 4, controls were more likely to report active smoking and cases had slightly more comorbidities than controls. The vast majority of cases reported at least one symptom, but 41% of the controls reported at least 1 symptom suggestive of COVID-19 since the beginning of the pandemic. When excluding controls who had a history of COVID-19 infection based on medical history or antibodies, still 32% had symptoms, highlighting the lack of specificity of some of these symptoms.

Table 4

Habits, comorbidities and preventive measures of the case and control cohorts

The 73% prevalence of antibodies detected in cases indicates past exposure to SARS-CoV-2. Seroprevalence among confirmed cases being lower than 100% is likely related to declining antibody levels associated with time since diagnosis.29 The overall prevalence of antibodies at recruitment in the population-based controls was 18% (12.5% among those who did not report past COVID-19 infection). We are conducting detailed analyses of the antibody response, including neutralising antibodies, accounting for age, sex, time since diagnosis and severity of symptoms, among others.

We have initiated PCR testing of saliva in a random sample of 2000 specimens stratified by pandemic wave to help inform design and evaluate cost-effective approaches for subsequent testing. In this context, we are planning to evaluate various approaches to testing of the large RESPIRA saliva biobank, including statistical sampling, staged testing and testing of pooled saliva specimens.

Compliance with the weekly visits by the active cases was close to 80%, with declining participation up to month 5, but again more than 80% at visit 6 for both cases and controls (online supplemental table S2). Monthly visits after the 6-month visit had participation around 60%, but visit 12 had 73% participation. Similarly, participation was lower in monthly visits between 12 months and 18 months and reached 66% at visit 18. These numbers were higher when removing subjects who withdrew from the study, with 91% for the 6 months, 86% for the 12 months and 80% for the 18 months visit (online supplemental table S3).

Among participants who attend their follow-up visits, compliance with blood sample collection has been above 97%, monthly saliva specimens are around 100% and the mid-month saliva samples close to 80% during most of the period except for the 18-month visit when it was 73%, as shown in online supplemental table S4.

In summary, we have completed recruitment and initial follow-up of a cohort of cases of COVID-19 and a control cohort that will be followed to address a series of questions about the pandemic in Costa Rica, with special emphasis on the immune response. The introduction and high coverage of vaccination during the study period presents challenges to evaluate the immune response to natural infection, but presents a great opportunity to study the impact of vaccination on the incidence and clinical characteristics of new infections among vaccinated subjects. Of special interest are the characteristics and effectiveness of immunity after the combination of infection and vaccination (hybrid immunity), particularly considering the emergence of new variants during the study.

The results of the household transmission study have been published.18 We detected a seroprevalence of 53% among household members of our index cases, with an estimated household secondary attack rate of 34%. Mask wearing by the index cases was strongly protective against transmission to the household.

We have also completed an analysis of the seroprevalence of COVID-19 antibodies to estimate cumulative incidence at the population level.3 The estimated total proportion of people infected was 4 times higher than the reported incidence and was 10 times higher among children, implicating that about half the population had been infected by end of 2022. Extrapolating to the current number of cases, the majority of people are likely to have been infected.

Multiple analyses are ongoing, including description of clinical presentation of cases, evaluation of factors for acquisition and disease severity and prevalence of clinical and subclinical infection at the population level. One of the main interests is the assessment of the characteristics of the immune response to infection and vaccination and their evolution over time. The extensive collection of saliva specimen will allow the study of incident infections in the population, their epidemiological correlates and the protection afforded by past infection, vaccination and the combination of both. In addition, we will investigate the incidence and determinants of long-COVID and quality of life after infection as part of this extensive multi-institutional international collaboration.

Collaboration

This study is a collaboration between the CCSS, the MS, ACIB-FUNIN, NIH, DKFZ and the UCR. Future collaborations are appreciated. All research projects using RESPIRA data need to be approved by the CCSS IRB.