This study demonstrates how a systems approach was used to develop eight user requirements for an intervention to increase PR uptake that could be delivered in primary care and address identified needs of patients and clinicians. The user requirements are mutually supportive, for example a good understanding of PR by a clinician will enable them to better allay patient concerns about attending PR and link potential benefit to the patient’s own circumstances. The user requirements are evidence-based, derived from barriers and enablers to PR uptake identified in prior research13, and are solution-independent, so can inform the development of a range of interventions in primary care settings.

Engineers have long recognised the tendency to rush into ‘cutting metal’ early in the mistaken belief that this will advance a project. This may provide a comforting illusion that something useful is being done but avoidance of a detailed understanding of the problem and clear statements of requirements is known to lead to serious consequences later in the life of a system, product, or intervention22. While the subject of requirements is well established in fields such as engineering design, systems engineering and software engineering, it is new to healthcare research. Of the limited examples in healthcare, information technology appears the most common area23,24,25. An example in primary care investigated requirement capture issues in the design of primary health facilities in the United Kingdom26. The study found poor management of design because of inappropriate processes with inconsistent requirements management. In contrast, clear and consistent requirements ensure that the right intervention is designed and can be appropriately tested. Requirements have played a crucial role in the design and delivery of successful projects in systems and software engineering27.

Requirements emerge from a clear understanding of user needs, and state in clear terms what a system or intervention should do in order to meet those needs. The process has been defined as: “… the iterative process by which the needs, preferences and requirements of individuals and groups – stakeholders – significant to the product development are researched and identified”; overall, requirements capture defines customers’, users’ and market requirements, design requirements and technical requirements28. Developing requirements can therefore be an extensive task as is the case in systems engineering22,29. It is important that efforts invested in developing requirements are appropriate to the scale of the system or proposed intervention30. In systems and software engineering, requirements are treated to an extent that is beyond the scope of the current paper. The few examples in healthcare are limited in scope; however, some highlight the variety of ways that requirements can be fully explored. For example, Zillner et al. studied user needs and requirements analysis for big data applications in healthcare. They distinguished between requirements that were business-related, technical related and those that were both business and technical related25.

The more comprehensive the requirements are, then the more fully the system or intervention is defined, and the chances of success improved. However, Davis has argued for what may be described as context-specific requirements by stressing the idea of “just enough” requirements30. In line with Davis, we consider the findings reported in this study a sufficient starting point in requirement development in a field for which the concept is new. The user requirements developed here were derived from clinicians’ and patients’ needs to understand PR and its benefits and to address patient anxiety. These factors have been frequently identified as important barriers to PR uptake. The need for interventions targeting these areas has been highlighted elsewhere31.

While Requirement 1 (help the patient understand what happens on a PR programme), requirement 2 (help the patient feel positive about attending a PR programme) and requirement 8 (help HCPs, patients and carers to feel they have had a positive conversation about PR) were voted as the three most important to prioritise, there was very little difference between the importance of all eight. In cases of intervention design where there is a large number of requirements and several constraints to be met then trade-offs may be needed, and differences in importance could provide a basis for agreeing trade-offs. In our case, there was no evidence from the stakeholder workshop that any of the requirements were rejected by participants or had to be traded-off for any system constraint.

The practical implications of the user requirements are consistent with the conceptual framework of access to healthcare described by Levesque et al. who argue that access to optimal care requires the person to be fully engaged in care32. They define access as the opportunity to reach and obtain appropriate services in situations of perceived need for care. This involves expectations, health literacy, knowledge about services and their usefulness, health beliefs and fit between services and patient need. An intervention that met our eight user requirements would address these factors.

The user requirements are solution-independent. They do not specify how an intervention might satisfy the requirements or what any intervention content might be. Rather, they provide intervention developers with a set of validated, evidence-based end goals that could be met through a variety of solutions. Examples are suggested here to illustrate a range of potential approaches. Two user requirements (1 and 5) address patient and clinician needs to understand what happens on a PR programme. Interventions could provide clear, printable information about PR so that clinicians can accurately describe to patients what happens at PR and/or patients could receive first-hand testimonies from previous PR participants about the PR experience. Three user requirements (3, 6 and 7) address patient and clinician needs to understand and value the benefits of PR as a treatment and the ways in which it could personally benefit the patient. An intervention could provide tailored resources that enable the clinician and patient to generate a personalised understanding of why PR is an important part of the patient’s treatment and how it will support them to overcome their particular symptoms and challenges. Three user requirements (2, 4 and 8) address patient and clinician needs for reassurance about patient anxieties and to have a positive conversation about PR. An intervention could provide tailored resources to address the patient’s anxieties so that they feel confident to participate in PR and/or could support clinicians in the skills needed to have sensitive conversations about patient concerns and anxieties.

Strengths of our approach are twofold. First, the EBC framework ensured that the user requirements were underpinned by a rigorous and systematic approach. The likelihood of laying the foundations for interventions that are fit for purpose was improved by ensuring a thorough understanding of the problem and exploring what mattered to a wide range of stakeholders. Second, setting user requirements is an early step in intervention development and the user requirements are solution-independent. By not predetermining the format or content of any subsequent intervention, intervention developers have maximum flexibility to develop a solution that is appropriate for their own context. The user requirements validated here therefore offer a broad contribution to intervention development in this field because they could be applied to the development of a range of interventions that share similar goals.

Developing, specifying and managing requirements is not a simple process and can take up to 30% of the total project time, particularly for complex products that are heavily regulated such as medical devices33. A potential limitation of our approach is that the study focused on the early stage of a full design process. Further work will be required to develop a comprehensive requirements list beyond user requirements, e.g. requirements also comprise technical, business and regulatory requirements. Additionally, our stakeholder validation workshop included people living with COPD, their carers/families, practice nurses, GPs and PR providers who were physiotherapists. PR is a multi-disciplinary programme and so inclusion of other clinical stakeholders who contribute to PR (e.g. occupational therapists, dieticians, nutritionists) would have broadened the stakeholder perspective at this stage. However, we were encouraged by the close agreement achieved on the requirements at this stage and other stakeholder perspectives can be incorporated as the design progresses. Lastly, requirements tend to be refined during the design process and this principle will apply to our user requirements. For example, further detail, such as performance limits that form the basis of testing, tends to be added as the design progresses. Similarly, as the form of the solution takes shape, this tends to dictate new requirements and may lead the refinement of existing requirements. For example, new stakeholders may join the design process at later stages and this brings opportunities to ensure the continued validity of the requirements and to make any refinements as required. Thus, our user requirements should be seen as a comprehensive first attempt, to be further developed during the later stages of intervention design.

In conclusion, we have argued that effectively increasing uptake to pulmonary rehabilitation from primary care while meeting the needs of a wide range of stakeholders is a systems challenge. We employed the Engineering Better Care systems approach to address the challenge of developing user requirements for an intervention to support uptake of pulmonary rehabilitation that could be delivered in primary care. We developed eight validated, evidence-based and solution-independent user requirements to address evidence-based user needs. We found few studies on systems approaches to requirements definition in respiratory medicine. This study demonstrates there may be potential in taking a systems approach to more challenges within this discipline.