Article Text

Home-based management of hypoxaemic COVID-19 patients: design of the Therapy@Home pilot study
  1. Josi A Boeijen1,
  2. Alma C van de Pol1,
  3. Rick T van Uum1,
  4. Roderick P Venekamp1,
  5. Karin Smit1,
  6. Karin A H Kaasjager2,
  7. Robert van den Broek3,
  8. Wilma Bijsterbosch3,
  9. Lisette Schoonhoven1,
  10. Frans H Rutten1,
  11. Dorien L M Zwart1
  1. 1Department of General Practice and Nursing Science, Julius Center for Health Sciences and Primary Care, Utrecht, Netherlands
  2. 2Department of Epidemiology and Health Economics, Julius Center for Health Sciences and Primary Care, Utrecht, Netherlands
  3. 3Department of Internal Medicine, University Medical Center Utrecht, Utrecht, Netherlands
  1. Correspondence to Ms Josi A Boeijen; J.A.boeijen{at}


Introduction During the COVID-19 pandemic, hospital capacity was strained. Home-based care could relieve the hospital care system and improve patient well-being if safely organised.

We designed an intervention embedded in a regional collaborative healthcare network for the home-based management of acutely ill COVID-19 patients requiring oxygen treatment. Here, we describe the design and pilot protocol for the evaluation of the feasibility of this complex intervention.

Methods and analysis Following a participatory action research approach, the intervention was designed in four consecutive steps: (1) literature review and establishment of an expert panel; (2) concept design of essential intervention building blocks (acute medical care, acute nursing care, remote monitoring, equipment and technology, organisation and logistics); (3) safety assessments (prospective risk analysis and a simulation patient evaluation) and (4) description of the design of the pilot (feasibility) study aimed at including approximately 15–30 patients, sufficient for fine-tuning for a large-scale randomised intervention.

Ethics and dissemination All patients will provide written, informed consent. The study was approved by the Medical Ethics Review Committee of the University Medical Center Utrecht, the Netherlands (protocol NL77421.041.21). The preparatory steps (1–4) needed to perform the pilot are executed and described in this paper. The findings of the pilot will be published in academic journals. If we consider the complex intervention feasible, we aim to continue with a large-scale randomised controlled study evaluating the clinical effectiveness, safety and implementation of the complex intervention.

  • COVID-19
  • Telemedicine
  • Feasibility Studies
  • Organisation of health services
  • Patient-Centered Care

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • We describe five essential elements (‘building blocks’) that comprise complex interventions for home-based management of acutely ill COVID-19 patients.

  • These generic building blocks could also be applied to the development of home-based management for other acutely ill patients.

  • The design of the intervention was iteratively developed and extensively evaluated by a multidisciplinary expert panel.

  • Informed consent will be asked from acutely ill patients, which may lead to a high participation barrier.

  • The current study will give important information on the feasibility of the intervention implementation, but it will not yield data on the formal efficacy of the intervention.


During the COVID-19 pandemic, hospital capacity was often strained by a high influx of critically ill COVID-19 patients.1 2 To relieve pressure on hospitals, early discharge initiatives were successfully introduced to manage hypoxaemic COVID-19 patients at home with oxygen treatment after clinical improvement during hospital admission, under remote monitoring.3–7 Ideally, future home-based management of COVID-19 patients is organised acutely, at the start of presentation, without initial hospital admission, directly from the emergency department (ED) or from home. Evidence for feasibility and safety of acute home-based management is, however, lacking.

If found to be feasible and safe, acute home-based management may improve patients’ self-efficacy, recovery and mental well-being, notably in older adults at higher risk of delirium when hospitalised. The development and evaluation of acute home-based management have been prioritised by the Dutch Ministry of Health,8 the Dutch College of General Practitioners9 and the Dutch Society of Internal Medicine.10

Acute home-based medical management should be according to current practical guidelines, but it should also consist of the timely delivery of equipment, for example, an oxygen concentrator and a pulse oximeter, and an initial check by a healthcare professional on the adequate use of such equipment by patients or relatives at home. Furthermore, adequate remote monitoring should be established in order to follow-up on the patient’s health status during treatment. Remote monitoring entails periodical measurement of, for instance, peripheral oxygen saturation (SpO2), heart rate and shortness of breath score for detecting deterioration in a timely manner.11 If performed regularly, these spot-(self)measured health data should provide sufficient follow-up of the patient’s oxygenation status and circulatory circumstances. The envisioned intervention for home-based management should be well coordinated, making sure that building blocks interact seamlessly, thus leading to safe, effective and patient-centred care. For this, the adaptive behaviour of healthcare workers from diverse disciplines is needed. All in all, such an interprofessional and multicomponent intervention qualifies as complex.12 To be successfully implemented in general practice, the intervention should be feasible and as lean as possible.

The UK Medical Research Council (MRC) published guidance on developing complex interventions and created a framework that facilitates working towards an effective and implementable design.13 Participatory action research (PAR) is a research method that fits this framework well. The PAR approach actively involves healthcare professionals and other stakeholders to design and implement complex interventions, considering local needs, barriers and facilitators. PAR has been widely used in the design of interventions that address complex and multifactorial healthcare problems,14 such as antibiotic resistance and stewardship interventions,15 and telemonitoring for chronic conditions,16 but has not been explicitly used in the context of acute COVID-19 care.

In this paper, we describe the study design according to MRC guidance for developing and evaluating a home-based complex care intervention for acutely ill COVID-19 patients who require oxygen. We describe the essential elements—five ‘building blocks’—that comprise the intervention, as well as the design of the pilot study used for evaluation. The feasibility data generated in this pilot can be used to set up and implement large-scale acute home-based management initiatives for critical episodes of COVID-19 or other acute respiratory tract infections requiring oxygen treatment, for example, a severe influenza infection or community-acquired pneumonia.

Methods and analysis


Guided by the MRC framework, we designed the intervention in four steps: (1) literature search plus evaluation of regional protocols with the assessment of knowledge gaps in clinical practice and the establishment of a multidisciplinary expert panel; (2) design of the intervention ‘building blocks’ through consensus meetings with the expert panel; (3) a prospective risk analysis and test case with a simulation patient and (4) design of the pilot study. During these successive steps, the PAR approach was followed, as further specified below.

PAR approach

PAR is a research process that requires the active involvement of different stakeholders. PAR aims to (1) trigger a change process for solving a practical problem, (2) be a learning process among those directly involved and (3) help further develop scientific knowledge. It simultaneously facilitates change in daily practice and contributes to the scientific debate. The PAR approach works through an iterative process of planning, action and reflection together with those who will experience the envisioned change intervention.17 Plan-Do-Study-Act (PDSA) is the iterative problem-solving model that is used as a method to structure the research process throughout (online supplemental figure 1).

Step 1: literature search and establishment of the multidisciplinary expert panel

Literature search

We performed a literature search. We did not find studies evaluating the development and effect of acute (preadmission) home-based interventions for COVID-19 patients in acute respiratory distress. Several studies, however, reported positive experiences with home-based use of pulse oximeters in COVID-19 patients.3 4 11 18–29

The multidisciplinary expert panel

We established a multidisciplinary expert panel to deliver an intervention that is tailored to managing acute home-based care and to make sure the intervention can eventually be evaluated appropriately. This panel is involved in the design, development and evaluation of the intervention throughout all phases. Relevant stakeholders in the Utrecht region were approached by the research team and recruited for representation in the panel: general practitioners (GPs) from four regional primary care groups, pulmonologists and acute care internists from four hospitals in the region, home care nursing organisations and an office for nursing care mediation, a monitoring centre, a regional care coordination centre and a patient representative from the Utrecht Elderly Care Network (see online supplemental table 1 for a list of stakeholder organisations).

Step 2: defining the intervention building blocks

In this step, the home-based intervention was established. The expert panel defined five essential components (‘building blocks’) for acute care at home during weekly consensus meetings. Each ideated building block was then organised and protocolised, using existing regional care structures where possible and inventing new or redirected care paths where necessary. To endeavour equivalency to hospital care, the intervention design details a plan for (1) acute medical care, (2) acute nursing care, (3) remote monitoring, (4) equipment and technology and (5) organisation and logistics. The designed concept of each building block is shown in tables 1–3. Pilot-phase implementation strategies for each building block are shown alongside.

Table 1

Concept and implementation of intervention building blocks 1 and 2: acute medical care and acute nursing care

Table 2

Concept and implementation of intervention building blocks 3 and 4: remote monitoring and equipment and technology

Table 3

Concept and implementation of intervention building block 5: organisation and logistics

Figure 1

Five building blocks for acute home management: (1) acute medical care (supervised by a general practitioner or hospital physician), (2) acute nursing care, (3) remote monitoring, (4) equipment and technology and (5) organisation and logistics.

Step 3: safety assessments

After the initial intervention was established, we subjected the intervention design to two practical safety assessments: (1) a prospective risk analysis to identify weaknesses in the protocols and develop safety net strategies and (2) a test case with a simulation patient to trial the logistics of the complex intervention.

Prospective risk analysis

Prospective risk analysis enables evaluation, followed by improvement of the healthcare processes and may prevent incidents.30 We subjected the intervention to a prospective risk analysis using the Health Failure Mode and Effect Analysis (HFMEA) method.31 In this method, a multidisciplinary team describes a healthcare process in detail and identifies all possible failure modes. The potential severity of the consequences and estimated frequency for each failure mode are then assessed and scored on a scale from 1 to 5. After the identification of potential failure mode causes, actions and controls can be implemented to eliminate failure modes or mitigate their effects.31 32 The prospective risk analysis aided in identifying protocol omissions and formulating explicit safety net strategies. A delay in the delivery of the oxygen concentrator, for example, was identified as a potential risk of the complex intervention. Possible causes identified included healthcare professionals’ time constraints to organise delivery and a lack of clear agreements on responsibilities. To overcome the latter potential risk, all stakeholders agreed that the responsibility for oxygen delivery was with the Care Coordination Centre. We made a coordination protocol summarising the tasks needed to organise timely delivery and this was made available to Care Coordination Centre employees.

Test case with a simulation patient

We tested the logistics of the intervention with a simulation patient, a healthy 66-year-old woman not involved in the study and without work experience in healthcare. During the test case, all care providers were informed about the simulation setting, but the timing of the test was not announced beforehand. Logistics were trialled, and the encountered hurdles were discussed in the following expert panel meetings to mitigate the encountered problems. The test case showed that there was a lack of helicopter view over the logistics, partly due to communication delays. This was mitigated by the more active involvement of a ‘first-day case manager’ and explicit feedback loops between building blocks. Moreover, we encountered a lack of acute nursing care availability in the short term due to a lack of personnel, a problem that is difficult to mitigate for future scenarios. The simulation patient provided valuable feedback from a patient’s perspective. She stressed that the information and instructions given to her were extensive and recommended repeated instruction, preferably with an informal caregiver present.

Step 4: feasibility pilot study protocol

In this upcoming step, the intervention and its feasibility will be piloted in 15–30 patients, while we will fine-tune the intervention further, aiming to learn from each patient’s trajectory using a PAR approach. Here, we outline the design of the pilot phase.


An observational pilot study in the province of Utrecht (1.3 million inhabitants) was executed between November 2021 and December 2023.

Inclusion and exclusion criteria

Two different categories of patients can participate in the study: type 1—nonfrail patients in respiratory distress who would normally be admitted to the COVID-19 ward (not the intensive care unit) for oxygen support and type 2—frail COVID-19 patients in respiratory distress for whom hospital admission is not desirable due to anticipated treatment limitations, as a result of, for example, frailty, terminal phase care or patients’ preferences. Inclusion and exclusion criteria are described in online supplemental table 4.

Intervention description

The intervention is described in detail in tables 1–3 (implementation).

Data collection

Evaluation cycles: short and large

In PDSA cycles, data gathered during the pilot study will be iteratively collected, analysed and discussed by the expert panel and adjustments will be implemented in the intervention and/or amendments will be made to the research protocol of the pilot study. There is a short-cycle evaluation in which the expert panel will convene every 1–2 weeks to evaluate the care trajectories of already included patients. In addition, the research team will evaluate the intervention in a larger evaluation cycle, every 6–8 weeks, to assess the predefined outcome measures; the ‘key elements’ are summarised in box 1. These key elements were defined by the expert panel prior to the initiation of the pilot study as crucial steps in the intervention in the pilot phase and for future implementation in a larger randomised controlled trial (RCT).

Box 1

‘Key elements’ (outcome measures) of the intervention which are assessed to help define successful implementation.

Other data collected

We will collect data on patient characteristics, disease course (days alive out of hospital) and healthcare use through a combination of inclusion forms, GP and hospital medical file extractions and patient monitoring diaries. We will follow up with patients for 30 days. We use WHODAS 2.0 patient questionnaires to facilitate future RCT development. The coordinating researcher will contact patients on days 2 and 30 to fill out the questionnaires. Lastly, we will collect qualitative data using semistructured interviews. We will separately interview involved GPs and participants with their informal caregivers. Key topics of these interviews will be (1) the feeling of safety, (2) the experienced advantages and disadvantages (or perceived risks) of home management and (3) recommendations for improving the intervention. Of note, in this pilot study, we will not assess the clinical effectiveness of the piloted intervention.

Data analysis

During and after the completion of the pilot phase, we will assess the completion of the key elements for each patient and, if not, the reasons for noncompletion. This facilitates further tailoring of the intervention and designing a final intervention that may be evaluated in a larger real-world setting. Furthermore, we will register patients’ healthcare use for the benefit of a health technology assessment to estimate the potential financial impact and advantage of the home management intervention. This assessment will provide a basis to facilitate a broader discussion for future implementation (ie, with healthcare insurance companies and policymakers). Lastly, qualitative data from the interviews will be coded and then analysed thematically.

Sample size calculation

In this pilot, a formal sample size calculation is not indicated. Fifteen to a maximum of 30 patients are expected to be sufficient for data saturation to go through the evaluation cycles to fine-tune the intervention and evaluate whether the intervention is suitable for use on a larger scale.

Patient and public involvement

Patient representatives are involved in all stages of our pilot. A patient representative from the Utrecht Elderly Care Network helped with designing, formalising and continuously evaluating the intervention and was a participant in the multidisciplinary expert panel. We had a simulation patient as a test case to trial the intervention logistics (and hence safety) prior to the inception of the pilot. Furthermore, the experience of patients participating in the pilot will be used in the iterative PAR approach to further develop the intervention. For this, we will use patients’ experiences as collected from both questionnaires and qualitative interviews.

Ethics and dissemination

This study will be conducted according to the principles of the Declaration of Helsinki (10th version)33 and the Dutch Conduct Code of Health Research.34 The Medical Ethics Review Committee of the University Medical Center Utrecht, the Netherlands, approved the study (protocol NL77421.041.21) according to the Medical Research Involving Human Subjects Act (WMO) and will monitor any amendments made. We have registered the pilot in the Dutch Trial Register (LTR) under number 22 655 ( The Dutch Trial Register (LTR) is the official data provider of the International Clinical Trial Registry Platform. Patients will be informed about the study through their GP. If they are interested, the coordinating researcher will visit the patient at home and obtain written informed consent. Study results will be published in international peer-reviewed medical journals and will be used to design a larger RCT to properly assess the effectiveness and safety of this complex home-based care intervention.

Ethics statements

Patient consent for publication


We would like to acknowledge the valuable contributions of all expert panel participants: E Vreede-Chabot, B Botma, Q Wagenaar, A Beetsma, S Komijn, M Hamers, I Bulsink, M Breteler, H Hardeman, R Vorselaars, J van den Brand, E-J Oudijk, A Bressers, S Sankatsing, F Hirsch, M Weenen, I Looijmans, R Wennekes, N Tak, S Balgit, M Mulder, S Hullegie, A Oostlander, A Ahmad and G-J Geersing.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors Conceptualisation: DZ; methodology: JB, AP, RU, RV, KS, LS, FR and DZ; resources: DZ, FR and RV; writing—original draft preparation: JB, AP, FR and RU; writing—review and editing: JB, AP, RU, RV, KS, KK, WB, RB, LS, FR and DZ. All authors read and approved the final manuscript.

  • Funding This work was supported by Zilveren Kruis, UNICUM Huisartsenzorg, RegiozorgNU, Huisartsen Utrecht Stad, Huisartsen Eemland (innovation fund S3).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.