WAYRA (the word for air in the Quechua language) was a randomized, controlled, open-label, two-arm clinical trial that evaluated the efficacy of a 6-week home-based respiratory physiotherapy and telephone-based psychological program compared to usual posthospital care provided by a public Peruvian hospital, aimed at improving exercise tolerance, lung function, mental health, and quality of life (QoL) outcomes. A baseline assessment was conducted at hospital discharge, with two follow-ups at week 7 and 12 (Figure 1). Due to the nature of the rehabilitation interventions, blinding of participants and the researchers was not possible; however, the data scientist conducting the data analysis was blinded. The trial was registered at ClinicalTrials.gov (NCT04649736).

This study took place in the COVID-19 ward at Hospital Nacional Cayetano Heredia (HNCH), a public tertiary-care hospital and a major referral center that serves approximately 3 million people from underserved neighborhoods in the northern metropolitan area of Lima, the capital of Peru. HNCH has been one of the main COVID-19 national referral centers throughout the pandemic, with 241 hospitalization beds and 23 intensive care unit beds [14]. During the pandemic, more than half of Lima’s population reported depressive symptoms, with increased rates of symptoms reported among young people with low income and without higher education [15].

The study population consisted of confirmed cases of COVID-19, diagnosed by serological or molecular tests, including members of both sexes, between 18 and 75 years of age, discharged from hospitalization, who received high-flow oxygen (>6 liters/minute) treatment at any point during hospitalization. Participants with prior respiratory pathology or psychiatric diagnoses were excluded. The complete eligibility criteria are provided in Textbox 1.

A trained nurse made the initial contact and prescreening, starting with a daily review of the list of hospitalized patients during the enrollment period, to identify potentially eligible participants among newly discharged patients or those close to discharge in the following days. Potential participants were informed about the study and its procedures and were invited to participate. Those who accepted received a detailed explanation of the nature of the study, randomization, study procedures, their potential risks and benefits, their rights as participants, and the study timelines. Potential subjects were explicitly told that participation was not mandatory, that there was no penalty for refusing to participate, and that their clinical treatment at the hospital would not be compromised in any way if they refuse to participate or opt out of the study at any time. All interested patients received the screening informed consent form (Multimedia Appendix 1) for their review, and they provided a signature once all their questions and concerns were addressed.

After obtaining participant consent and signature, a trained nurse performed the screening assessments, including review of eligibility criteria compliance and baseline measurements. These included cognitive status (assessed using the Montreal Cognitive Assessment, visual impairment version [MoCA-BLIND] [16]), sociodemographic and clinical data, baseline respiratory symptoms (using the St. George’s Respiratory Questionnaire [SGRQ]), depressive symptoms (using the Nine-Item Patient Health Questionnaire [PHQ-9]), anxiety symptoms (using the Generalized Anxiety Disorder Assessment [GAD-7]), posttraumatic stress disorder symptoms (Impact of Event Scale-Revised [IES-R]), and QoL (using the 36-Item Short Form Survey [SF-36] and EuroQuol-5D [EQ-5D]). Subsequently, a trained physician assessed the participants’ exercise tolerance with the 6-minute walk test (6MWT) and pulmonary function with a spirometry test. Participants were screened for severe depressive symptoms. Those with a GAD-7 score ≥15 or IES-R score ≥37 were referred to a psychiatrist for evaluation and treatment. Participants with a PHQ-9 score ≥20 and self-reported suicidal ideation during the psychological evaluation were immediately referred to mental health services for treatment and were excluded from the study. Patients who completed the baseline assessments and met eligibility criteria were enrolled into the study. The recruitment team communicated with the study coordinator who performed a baseline data quality verification, the allocation to one of two arms (control or intervention), and assigned a participant ID code.

We used a blinded list preuploaded in REDcap [17], which was created by a data administrator who was not involved in study recruitment. To achieve balance in the number of participants assigned to each study arm, the treatment arm was randomized at a 1:1 ratio using permuted blocks of variable size, randomly varying from 2, 4, and 6 participants. The blocks were randomized using the Stata v16.0 ralloc command (StataCorp). Finally, all participants were scheduled for follow-up visits as described in Table 1. Sample selection and enrollment took place between the day before hospital discharge and up to 3 days after discharge.

We will perform an intention-to-treat analysis of primary and secondary outcomes at 7 and 12 weeks after hospital discharge. The final analysis will be performed when all 108 participants completed the trial. We will compare the 6MWD (meters), FVC (liters), FEV1 (liters), and questionnaire scores at 7 and 12 weeks after hospitalization discharge between the intervention and control arms. In the secondary analyses, we will compare the proportion of participants with respect to restrictive spirometry pattern, depression, anxiety, and posttraumatic stress disorder at 7 and 12 weeks after hospital discharge between the intervention and control arms. In addition, following the recommendation by Cocks and Torgerson [32], we will use a one-sided 80% CI to determine if this trial should proceed to a phase III trial. As a sensitivity analysis, the association between the number of therapies received and outcomes will be examined, adjusting for potential confounders that will be identified by a causal acyclic diagram. There will be no formal interim analysis of the data.

We used STATA Version 16 (StataCorp) to calculate the ideal sample size. We attempted to enroll 108 participants in total, allocating 54 to each treatment arm. We estimate that a sample size of 86 participants would be sufficient to detect a mean difference between arms of 55.1 meters in the 6MWT according to the study by Liu et al [11] considering a 95% CI and 90% power. Assuming a potential 20% rate of rejection or loss to follow-up, we aimed to achieve a sample of 108 participants.

Ethical approval of the study protocol and informed consent form was obtained from the Institutional Review Board at Universidad Peruana Cayetano Heredia (#202852) and HNCH (#085-2020). Likewise, any changes in the protocol were reported to the mentioned ethics committees. Written informed consent was requested from all participants prior to any data collection. To protect participants’ confidentiality, all collected data of clinical assessments were linked to a unique ID code. Data were registered in paper form and electronically. The electronic information was stored, backed-up, and secured by password protection in the RedCap server. Paper forms were archived in secure locked cabinets in the HNCH. All confidential information, including participants’ contact details or other sensitive data, was only accessible to authorized staff from the project. Long-term data management complied with Universidad Peruana Cayetano Heredia research policy. The Peruvian National Health Institute reviewed the protocol and determined that it was exempt from review by the Ethics Committee due to the nonpharmacological nature of the intervention.

Both control- and intervention-arm participants benefited from the study. These benefits included education and counseling on respiratory disease risk factors, reporting of pulmonary function test results by a specialist physician, and referral to trained personnel if necessary.

Potential risks of respiratory rehabilitation were oxygen desaturation, palpitations, sweating, arrhythmias, chest tightness, shortness of breath, and muscle aches. Potential risks of psychological support were symptom substitution, dependence on the physiotherapist, stigmatization, problems in social relationships or even separation, as well as alcohol or drug abuse, deliberate self-harm, and suicidal ideation or attempts. Adverse events of the interventions were collected after participants provided informed consent and were duly enrolled in the study. If adverse events were detected, the field staff immediately notified the study coordinator and the management protocol was activated.

WAYRA is the first randomized controlled trial evaluating integrated pulmonary and mental health rehabilitation for hospitalized COVID-19 survivors in a low- and middle-income country setting. Through this trial, we gained experience in implementing a remote rehabilitation program in a resource-limited country and will generate evidence on pulmonary, mental health, and QoL alterations following COVID-19 the first 3 months after hospitalization. We hypothesize that participation in the combined program will improve pulmonary, physical, and mental health outcomes. We hope that our study will provide a reliable basis for further clinical trials focused on comprehensive rehabilitation in patients hospitalized for severe COVID-19 and other lung diseases.

The burden of long-term complications of COVID-19 will increase because, although several countries have achieved broad vaccination coverage, newly emerging variants may be resistant to vaccines [33]. The most frequently affected systems are respiratory and physical, although psychological symptoms are also common [3]. Despite this, there are still few published studies that have proposed strategies for COVID-19 survivors with long-term complications and far fewer that include mental health support. There are approximately 200 trials registered with ClinicalTrials.gov evaluating rehabilitation programs for COVID-19 survivors; most include pulmonary rehabilitation, whereas only a few have a psychological component [34].

The main strength of this study is that the calculated sample size has the power to detect a clinically significant difference between the groups. Nevertheless, this study has some limitations to be considered. First, mental health and QoL outcomes will be assessed with screening tools. Second, this trial has an open-label design due the nature of the intervention.

In conclusion, the COVID-19 pandemic will result in millions of people prone to develop respiratory and mental health sequelae worldwide, and health systems are not prepared to respond adequately to this situation. Little evidence exists on rehabilitation strategies, and most of the existing evidence does not come from low- and middle-income countries that face additional challenges compared with developed countries. The clinical trial described in this protocol will generate useful evidence to develop effective interventions to benefit patients after COVID-19, and will serve as a basis to help plan future studies evaluating pulmonary rehabilitation strategies and mental psychological support.