Study Design Overview

This study is a prospective, observational cohort study enrolling people living with and without HIV within 4 weeks of first SARS-CoV-2 infection. The study also enrolled people with and without HIV who believed they never had SARS-CoV-2 infection as controls. The study collected health information and specimens from participants at predetermined time intervals through questionnaires, cognitive assessments, and blood draws (Figure 1). This study was designed to enroll participants soon after SARS-CoV-2 infection so that the time course of postacute symptoms and sequelae in the first 12 months could be prospectively investigated and because biomarkers or immune markers of long COVID may only be present early in the postacute period [21]. The study recruited people into the following 4 study arms:

Open in a separate window

Figure 1

Schedule of study events. *Optional for HIV−COVID− participants; **Only for COVID+ participants.

1. HIV+COVID+ arm: people living with HIV with first SARS-CoV-2 infection within the previous 4 weeks.

2. HIV−COVID+ arm: people without HIV with first SARS-CoV-2 infection within the previous 4 weeks.

3. HIV+COVID− arm: people living with HIV who believed they never had SARS-CoV-2 infection.

4. HIV−COVID− arm: people without HIV who believed they never had SARS-CoV-2 infection.

Study Eligibility

We recruited people with and without HIV, who either had first SARS-CoV-2 infection within the previous 4 weeks or who believed they had never had SARS-CoV-2 infection. Adults aged ≥18 years living in the 48 contiguous United States who had the ability to provide consent, who believed they could complete the study events, and who could communicate by telephone in either English or Spanish were eligible. Further eligibility criteria are presented in Table 1.

Recruitment

Participants living with HIV came to know about this study through health care providers, social media, and peers in a number of ways. The study coinvestigators and study team emailed and called HIV clinic directors across the 48 contiguous United States and requested permission to hang study flyers in their clinics. The study principal investigator (PI) and coinvestigators established a network of academic investigators and other people interested in HIV and COVID-19 across the United States and kept them abreast of updated flyers and recruitment updates through regular contact. HIV providers who were colleagues of these people were informed about the study by emails or announcements at provider meetings. Emails were sent to the United States HIV Medicine Association and the Infectious Diseases Society of America’s Exchange listserves (primarily subscribed to by infectious diseases physicians and providers) every 2-4 months with updated flyers and information about the study to increase study referrals. HIV providers who had heard about the study informed their patients with recent SARS-CoV-2 diagnosis about the study and sought permission for the study team to contact them. This study also reached out to established HIV and substance use study cohorts with information on this research study to disseminate to their study cohort participants. Importantly, this study and people living with HIV aiding this study engaged people living with HIV in the community in numerous ways including social networks, email, and social media. The study worked with the Network for Long COVID Justice to amplify dissemination efforts around this study to several social media sites and web-based media sites popular among people living with HIV. The study worked with several community members to disseminate information about this study through social media channels. The study staff asked the participants who were enrolled to ask their friends and contacts who were eligible to contact our study if they were interested; there was no compensation for this recruitment. Advertisements for the study were posted on Facebook and Grindr. Finally, the study investigators received permission to identify people living with HIV with recent positive SARS-CoV-2 tests within the Johns Hopkins Health System via weekly electronic medical record reports. The study staff called these people with an invitation to hear more about the study.

Participants without HIV heard about this study in a number of ways. Advertisements for the study were posted on Facebook and Grindr. The study worked with several community members to disseminate information about this study through social media channels. Participants were recruited through the Johns Hopkins Institute for Clinical and Translational Research recruitment registry, Hopkins Opportunity for Participants Engagement (HOPE) registry. The HOPE registry is a centralized, institutional review board–approved database that contains the name and contact information of people who have taken a SARS-CoV-2 test within the Johns Hopkins Health System and have expressed interest in participating in research studies. The registry is housed in Research Electronic Data Capture (REDCap; Vanderbilt University), and people who have consented to participate in the registry enter their own data into the registry’s web-based intake form, or they have their data collected by recruitment navigators via telephone and entered into the database by HOPE registry navigators. Eligible participants are then matched to each participating study based on the information they have provided and study eligibility criteria. This study’s staff then reached out to the people matched to this study to assess their willingness and eligibility for this study. Physical flyers were posted on the Johns Hopkins medical campus.

A study website was developed to provide information for interested people to self-refer via a web-based REDCap Health Insurance Portability and Accountability Act (HIPAA)–compliant form or by contacting the study directly (via email or phone). This website was advertised on social media platforms such as Facebook (including special interest groups such as for lesbian, gay, bisexual, transgender, queer, and others [LGBTQ+] communities in different cities) and Grindr.

Consent

A study coordinator based at Johns Hopkins University held a consent conversation with potential participants by telephone or video call, and consent was signed using DocuSign or paper consent forms. All consent activities were conducted in either English or Spanish. The DocuSign process was approved by the institutional review board and is 21 Code of Federal Regulations Part 11–compliant to obtain a secure electronic signature. This process involved sending the consent form to the participant via an email from DocuSign and providing a participant-specific access code that was required when the participant accessed and signed the consent form. Participants were given adequate time to consider the study and ask questions before signing the consent form. When ready to sign, each participant entered their code, verifying that the person signing the consent was the person whom we spoke with previously, and signed the consent within the DocuSign system. Once the participant had electronically signed, the study team member obtaining informed consent would be notified that the electronic form was ready for their signature. Once signing was completed by all parties, both the study team and the participant could download the signed consent as a PDF. The study team also had access to the audit log and the Certificate of Completion. Documentation capturing details of the consent process was maintained on REDCap.

In instances where DocuSign could not be executed or the potential participant declined to use it, the consent conversation happened in person or by telephone or video call and a paper consent was signed. In these cases, participants were provided with a copy of the paper consent document before the consent conversation either via email, fax, mail, or in person. After the consent conversation, the participants were given adequate time to consider the study and ask questions. The participant would sign, date and time the paper informed consent document. The document was then handed, mailed, emailed, or faxed to the consent designee for their signature, and a copy of the double-signed consent document was handed, mailed, emailed, or faxed back to the participant.

Surveys

Multimedia Appendix 1 [22-48] and Multimedia Appendix 2 provide detailed information about the surveys and the timing of study events. At enrollment, participant demographics, social history, substance use, medical history, medication use, SARS-CoV-2 vaccination status, and HIV history (for people living with HIV only) were recorded. Participants in the COVID+ arms also completed a comprehensive survey on symptoms, mental health, and quality of life via telephone or on the web in the month before having SARS-CoV-2 infection. At each time point after COVID-19 diagnosis (for the COVID+ arms) or after enrollment (for the COVID− arms), participants completed a comprehensive symptom, mental health, and quality of life survey (Multimedia Appendix 1). This survey consisted of several validated tools, which are detailed in the following paragraphs.

Participants indicated the presence and severity (on a scale of “Not at all,” “A little bit,” “Somewhat,” “Very much,” or “Quite a bit”) of 49 long COVID symptoms. Overall, 34 symptoms derive from the FLU-PRO Plus, and 15 additional symptoms were added based on their frequency of report in a Long COVID Patient Led Research Collaborative survey with input from the patient-researcher authors [5,22]. Dyspnea was assessed using the Modified Medical Research Council (mMRC) Dyspnea Scale, which is used to assess the degree of baseline functional disability because of dyspnea [23]. The Breathless, Cough, and Sputum Scale (BCSS) was used to evaluate the respiratory symptoms common in COVID-19 [24]. The survey also included the Insomnia Severity Index (ISI), which is a 7-item questionnaire assessing the participant’s perceived insomnia symptoms and resulting disruption to their lives [25,26]. Participants self-reported the average daily sleep duration in the previous week. Participants completed the Fatigue Severity Scale (FSS), a 9-item questionnaire assessing their perception of the impact of fatigue on their quality of life and activities of daily living [27]. The FSS global fatigue question was used to measure the participants’ global fatigue. At each study event, the participants were asked whether they had received a COVID-19 vaccine or had a new infection with SARS-CoV-2 since the previous study event.

The survey included the Patient Health Questionnaire Depression 8-item (PHQ-8) and Generalized Anxiety Disorder-7 (GAD-7) tool to assess the severity of depressive and anxiety symptoms, respectively [28-30]. Participants also completed the computerized adaptive testing for mental health (CAT-MH), which draws upon a bank of >1000 test items aligned with the Diagnostic and Statistical Manual of Mental Disorders (DSM)-5 criteria.[31-33]. The modules completed included the computerized adaptive diagnostic test for major depressive disorder (CAT-MDD; screener for major depressive disorder), the computerized adaptive diagnostic test–Depression Inventory (CAT-DI; assesses depression severity), and computerized adaptive diagnostic test–Anxiety Inventory (CAT-ANX; assesses anxiety severity). The survey also assessed self-reported cognitive problems using the self-assessment section of the General Practitioner Assessment of Cognition (GPCOG). This tool consists of 6 yes or no items relating to participants’ perceived difficulty with certain activities owing to cognitive limitations compared with a prior time point [34]. We asked participants to evaluate their current ability to complete the tasks compared with their ability before they had COVID-19 or before enrollment.

Quality of life was assessed using the Short Form Survey (SF-36) and the EuroQuol EQ-5D-5L. The SF-36 contains 36 Likert and binary items relating to the participant’s activities and perceptions of health status [35,36]. The EQ-5D-5L assesses general quality of life through 5 dimensions that include mobility, self-care, usual activities, pain and discomfort, and anxiety and depression [37]. The EQ-5D-5L includes a general health question that asks the participant to provide 1 number from 0 to 100 on a visual analog scale representing their overall health [37]. The Brief Resilience Scale was used to assess the perceived ability to bounce back or recover from stress [38]. Stressor-related questions were also asked to assess general stress levels from daily life during the COVID-19 pandemic (Multimedia Appendix 1).

Participants in the COVID+ arms were administered an additional acute COVID-19 survey at 1 or more months after COVID-19 diagnosis in which the participant was asked to recall the presence and severity of the same 49 COVID-19 symptoms described earlier within the first 14 days of symptom onset or positive testing (if asymptomatic). The survey also queried the type of COVID-19 test taken, sample source, COVID-19 exposure, treatment for acute COVID-19, and whether hospitalization or oxygen was required during acute COVID-19. Reinfection was not uncommon in this cohort, and each participant was asked about reinfection (COVID+ arm) or new SARS-CoV-2 infection (COVID− arm). Each individual with a new SARS-CoV-2 infection since enrollment was administered the same acute COVID-19 survey described earlier one or more months after COVID-19 diagnosis to collect information on the acute course of COVID-19. Further information about the surveys used in the study can be found in Multimedia Appendix 1.

Neurocognitive Assessments

Participants in all arms underwent a telephone-based neurocognitive assessment at 1 and 4 months after acute SARS-CoV-2 symptom onset or diagnosis (COVID+ arm) or after enrollment (COVID− arm). Telephone assessment has been shown to be feasible and provides a valid assessment of cognitive function relative to in-person neurocognitive testing [49-51]. This telephone assessment has been successfully implemented in diverse patient populations, including post–COVID-19 populations [52-54].

A total of 11 cognitive assessments were administered in this study. The Hopkins Verbal Learning Test–Revised (HVLT-R) is a test of new auditory-verbal learning (VL), verbal memory (VM), and recognition discrimination memory (RM) [39,40]. The Oral Trail Making Test part A (OTMT-A) and Oral Trail Making Test part B (OTMT-B) are brief motor-free tests of mental processing speed (part A) and executive functioning requiring sequential set shifting (part B) [39-41]. The Wechsler Adult Intelligence Scale–Fourth Edition (WAIS-IV) Digit Span Forward (DSF) and Digit Span Backwards (DSB) assessments test auditory attention (forward) and working memory (backward) [42-44]. The Calibrated Ideational Fluency Assessment (CIFA) assesses letter-cued and category-cued verbal fluency (VF). Category-cued VF assesses rapid access to semantic information, and letter-cued VF assesses speeded word retrieval in response to phonetic cues [45-47]. The WAIS-IV Information (WAIS-IV IN) test assesses the general fund of knowledge and was only administered once in this study [42-44]. Performance on the WAIS-IV IN is relatively resistant to neurological damage, and scores correlate highly with overall IQ, thereby allowing scores to serve as a proxy for lifelong intellectual functioning [48].

Mobile Phlebotomist Visits and Clinical Laboratory Testing

Participants in the COVID+ arms had mobile phlebotomy visits at a location of their choice at 1 and 4 months after post–acute COVID-19 symptom onset (or after diagnosis if asymptomatic). Participants in the HIV+COVID− arm had 1 mobile phlebotomy visit at either month 1 or 4, provided that they had not had SARS-CoV-2 infection before the visit. Participants in the HIV−COVID− arm could optionally undergo specimen collection at either month 1 or month 4, provided that they had not had SARS-CoV-2 infection before the visit. Anti–SARS-CoV-2 nucleocapsid immunoglobulin G was measured in all participant specimens, and data from participants in the COVID− arm were excluded from analyses if their specimens tested positive for antinucleocapsid antibodies and they were unaware of SARS-CoV-2 infection.

A mobile phlebotomy company that operates in every state of the United States was engaged to (1) measure participant height and weight, (2) measure orthostatic vital signs, and (3) draw ≤50 mL blood. A blood draw kit was shipped to the participants before the date of the blood draw. The kit included a vital sign sheet, a sodium citrate tube, serum separator tubes, EDTA tubes, and a PAXgene blood RNA tube. All tubes were plastic and labeled with the participant identifier and a space for the date and time of collection. In addition, absorbent sleeves to separate tubes, approved shipping materials including an outer cardboard box, and an inner 95-kPa specimen transport bag with appropriate biohazard labeling were included in the blood draw kit.

For the measurement of orthostatic vital signs, the participants were asked to lie down for 5 minutes before the first blood pressure and heart rate measurement. Participants were then asked to stand up, and after 1 and 3 minutes, additional heart rate and blood pressure measurements were recorded. After the measurement of orthostatic vital signs, blood was drawn from the participant by the phlebotomist. After filling, all tubes were inverted gently 6-8 times. The serum separator tubes were kept upright for 30-60 minutes before centrifugation for 15 minutes at 1100-1300 g by the mobile phlebotomist. The tubes were placed in absorbent sleeves and were shipped overnight using International Air Transport Association–approved procedures to the study laboratory (Rush Clinical Retrovirology Research Laboratory, Chicago). When specimens arrived at the laboratory, designated specimens were sent to the study Clinical Laboratory Improvement Amendments of 1988 (CLIA)–certified testing laboratory (Rush Medical Laboratories, Chicago) for testing. If unforeseen shipping delays occurred and blood was received >48 hours after collection, then some of the clinical testing was not performed owing to stability issues (eg, CD4, complete blood count, lipid panel, and D-dimer). The remaining blood was processed on the same day, and serum, plasma, peripheral blood mononuclear cells (PBMCs), and the PAXgene blood RNA tubes were appropriately frozen and stored. An inventory of the biorepository was created and maintained by the study laboratory using the Laboratory Data Management System (Frontier Science) [55]. The Johns Hopkins REDCap team developed a novel method to directly input biorepository inventory files from the Laboratory Data Management System into the study’s REDCap database for seamless and rapid updating of the biorepository in the central database.

Biospecimen Testing

The CLIA-certified study testing laboratory ran the following panel of tests: SARS-CoV-2 antinucleocapsid IgG (Abbott Architect), complete blood count with differential, high-sensitivity C-reactive protein, creatine phosphokinase, D-dimer, lipid panel, HIV fourth-generation antigen and antibody testing (for those who self-reported as HIV-seronegative), and CD4 T cell percentage and absolute count (for those who self-reported as living with HIV). Plasma, PBMCs, and serum were also obtained from all samples. If the blood sample was received by the Rush Clinical Retrovirology Research Laboratory >48 hours after collection, the sample was not sent for complete blood count, CD4 panel, lipid panel, or D-dimer according to the clinical laboratory protocol. The blood sample was still processed for plasma and PBMCs, but in cases where the sample did not separate, plasma and PBMCs were not saved. However, serum was obtained from these samples.

Coronavirus antibodies and select hormones and cytokines were quantified in a subset of specimens. The panel of coronavirus antibodies that were quantified included HCoV-229E Spike, HCoV-HKU1 Spike, HCoV-NL63 Spike, HCoV-OC43 Spike, SARS-CoV-1 Spike, SARS-CoV-2 Nucleocapsid, SARS-CoV-2 NTD. SARS-CoV-2 RBD, SARS-CoV-2 Spike, AY.4, B.1.351, BA.2, BA.2+L452M, BA.2+L452R, BA.2.12.1, BA.3, BA.4, BA.5, and CoV-2 Spike. The hormones and cytokines that were tested were based on what might be important in long COVID from the current literature at the time: interferon (IFN)-β, interleukin (IL)-1RA, IL-1α, IFN-λ1, IL-5, IL-8, interferon gamma-induced protein-10 (IP-10), chemokine (C-X-C motif) ligand 9 (CXCL9), macrophage inflammatory protein (MIP)-1β, vascular endothelial growth factor (VEGF), alpha-1-acid glycoprotein 1 (ORM1), MIP-5, T cell immunoglobulin and mucin-domain containing (TIM)-3, Factor D, lipocalin 2 (LCN2), glial fibrillary acidic protein (GFAP), pentraxin 3 (PTX3), Cortisol, IFN-λ2, transforming growth factor (TGF)-β1, ADAMTS13, complement component 5a (C5a), IFN-γ, IL-1β, IFN-λ3, IL-6, C-C motif chemokine ligand 20 (CCL20), C-C motif chemokine ligand 23 (CCL23), and tumor necrosis factor (TNF)-α [21,56-59]. Most of these analytes were measured with validated assays using commercial plates manufactured by Meso Scale Discovery. Results were analyzed using Discovery Workbench (version 4.0.12; Meso Scale Discovery). The antibody kits were provided by the company.

Data Collection and Management

Database services were provided by the Johns Hopkins University REDCap, a secure, HIPAA–compliant web application for building databases and managing web-based data for research [60]. Almost all data were collected by Johns Hopkins University study staff. A limited number of data points were generated by the mobile phlebotomy company (eg, orthostatic vital signs, height, weight, and status of collection and shipment) and by the study laboratories at Rush University (eg, clinical laboratory test results, status of shipment receipt, and status of sample processing and storage). Study staff at Johns Hopkins University or Rush received data from the mobile phlebotomy company via the phlebotomy secure internet portal and entered it into the REDCap database. The PI of the study was notified of grossly abnormal vital signs or laboratory results and about moderate to high depression or anxiety PHQ-8 and GAD-7 scores. The PI was also notified if the mobile phlebotomist noted any concerning event during the mobile phlebotomy visit. In each case, the PI reached out to the participant to provide basic health advice and guidance and resources for referral to the appropriate health provider.

Outcomes

Compensation

All participants were offered compensation for their time and efforts in participating in the study. A compensation of US $25 was offered per blood draw and US $25 per cognitive assessment completed, for a total of up to US $100 (for COVID+ arm) or US $75 (for COVID− arm) if all study events were completed. Participants were compensated in the form of an Amazon or VISA gift card.

Power Calculation

The target enrollment was as follows: 105 participants in the HIV+COVID+ arm, 105 participants in the HIV−COVID+ arm, 45 participants in the HIV+COVID− arm, and 50 participants in the HIV−COVID− arm. This resulted in a total target sample size of 305 participants overall, 210 of whom would be in the post–COVID-19 arms. Using a 2-tailed α of .05 and 10% estimated proportion of persistent symptoms in the comparator group, we estimated that there was 80% power to detect a difference of 15% in new or worsened symptom prevalence between people living with and without HIV after SARS-CoV-2 infection using a chi-square test of proportions.

Strengths

This study was designed to meet the needs of people potentially experiencing long COVID who may be juggling fatigue and other symptoms and sequelae with life, work, and family responsibilities. The major strength of this study is its remote and participant-centric design. Participants could complete every study event from any location of their choice, including their home. This also allowed for the recruitment of a diverse participant population from around the United States who spoke either English or Spanish. In addition, the development of a central biorepository for specimen sharing will serve as a resource for investigators interested in the intersection of HIV and long COVID. Another strength of the study was that our surveys were developed in consultation with patient researchers who were experiencing long COVID.

The authors would like to thank all study participants for their time and samples. Major funding for this study was provided by the American Foundation for AIDS Research grant 110180-69-RSCV. Additional support was provided by Johns Hopkins University Center for AIDS Research (P30AI094189). AARA is also supported by National Institute of Allergy and Infectious Diseases K08AI143391. The authors would like to thank all the community members, providers, staff, and other individuals who helped disseminate information about the study, including the Long COVID Justice Network; AIDS Action Baltimore; SisterLove; POZ magazine; Somos Baltimore Latino; the COVID Advocates Advisory Board; The Cranky Queer Newsletter; the Johns Hopkins Institute for Clinical and Translational Research Hopkins Opportunity for Participants Engagement registry; the DC HIV Cohort; and the numerous physicians, researchers, and other individuals who referred participants to the study. The authors would like to thank Hannah Davis and other members of the Patient Led Research Collaborative for Long COVID for providing survey data to inform symptom questions and for reviewing surveys before use. The authors would also like to thank all study staff at all participating study sites. The authors would like to thank Dr Yukari Manabe, Dr Richard Moore, Ms Jeanne Keruly, and Ms Zoe Demko for discussion and support during initial protocol design. The authors would like to thank the Johns Hopkins REDCap (Research Electronic Data Capture) team, specifically Mr Scott Carey, who developed a novel method to directly input biorepository .csv files in the Laboratory Data Management System format into the study’s REDCap database.