Using Normalisation Process Theory to Develop a Complex Intervention: A Qualitative Case Study of Screening for Homelessness in Healthcare Settings Authors: Nilakshi Gunatillaka, Tze Lin Chai, Suzanne Nielsen, Belinda Hammond, Kimberley Norman, Helen Skouteris, Elizabeth Sturgiss Published: 2026 ABSTRACT ABSTRACTBackground and AimsNormalisation Process Theory (NPT) seeks to describe and explain how a complex intervention can be embedded into an existing health and social care system. NPT is frequently used to evaluate interventions, and less frequently to design them.We demonstrate how NPT can be applied to exploratory qualitative data to understand the potential barriers and facilitators to the implementation of a new intervention in a health setting using the NPT framework. We illustrate this application using a case study of screening for homelessness and housing instability in Australian emergency department and primary care settings.MethodsThis qualitative study comprised semi‐structured interviews of healthcare workers and people with lived experience of homelessness. Our hybrid analysis involved an initial inductive reflexive thematic approach followed by deductive qualitative analysis guided by a theoretical framework, that is, the four domains of NPT: Coherence, Cognitive Participation, Collective Action and Reflexive Monitoring.Results39 interviews were conducted with healthcare workers (n = 30) and people with lived experience of homelessness (n = 9). Healthcare workers differentiated screening for homelessness from their current work (Coherence) but had mixed views on the utility of screening (Cognitive Participation) and were reluctant to do the extra work (Collective Action). Healthcare workers from Aboriginal Community Controlled Health Organisations and community health centres demonstrated higher commitment and readiness to do this work than their counterparts in hospital or fee‐for‐service primary care settings. People with lived experience of homelessness wanted their healthcare workers to understand how their housing situation affected their health, and to adapt care accordingly.ConclusionThis study demonstrates that NPT can be used in the very early planning of interventions. This application of NPT may support the intervention design that is better aligned to the expectations of people tasked with implementing the intervention, thus increasing uptake and facilitating normalisation.Keywords: implementation research, intervention design, normalisation process theory, qualitative research Introduction Complex interventions in health settings are more likely to be feasible, widely adopted and effective when the settings in which they are implemented are well understood and considered during intervention design [1, 2]. Often, however, decisions about an intervention's purpose, content, format and delivery are made by people who will not be involved in doing the work. This can lead to intervention failure from poor uptake, wasted time and misused resources [1]. There is also the potential for far more serious consequences, including unintended harms to people using health services, particularly those experiencing multiple or complex disadvantage [1, 3]. Therefore, involving the healthcare workers who will be tasked with implementing an intervention, and the health care users who are expected to benefit from it, is foundational to intervention design. This paper seeks to demonstrate how Normalisation Process Theory can be used to augment an exploratory qualitative study of frontline health workers' and patients' perspectives on discussing housing in a healthcare consultation. This approach could offer a better understanding of the context in which an intervention will be set, and the type of intervention that would suit the setting. It is particularly appropriate when the change (i) is outside the usual scope of practice (e.g. a social intervention in a health setting), (ii) is new or untested rather than a change to an existing process and (iii) requires multiple groups of people with different roles to work together to be implemented. This method has utility in complex resource‐ and funding‐ restricted environments where implementation failure is more likely. Normalisation Process Theory (NPT) [4] is often used to evaluate the implementation of complex interventions in health care settings [5, 6, 7, 8, 9, 10] using both qualitative (interviews, focus groups) and mixed methods (surveys, interviews and focus groups) [8, 9]. NPT could be a useful tool for understanding how underlying cognitive, relational, and contextual factors affect the extent to which an intervention is embedded into routine practice, before the intervention has been developed and implemented. Such an application would allow for potential barriers to implementation to be identified and addressed as part of the intervention, and for facilitators to be better leveraged to support uptake. Our method involves examining exploratory qualitative data from clinicians who would be tasked with changing their ‘normalised’ behaviour to adopt an intervention. We gather clinicians' perspectives on whether the proposed intervention would be feasible within their health service context, and if not, what actions could be taken as part of intervention design and implementation to modify the context to be more favourable. We use the example of a homelessness/housing instability screening intervention in Australian primary care and hospital settings (Box 1) to illustrate this approach. Homelessness and housing instability screening intervention –Homelessness can be defined as the lack of access to minimally adequate housing, which usually refers to rooflessness or sleeping rough. A broader view of housing instability encompasses houselessness (a temporary place to sleep such as shelters or crisis accommodation), insecure housing (informal tenancies, difficulty paying rent/mortgage, threatened eviction) and inadequate housing (overcrowded or unfit housing) [11, 12]. Homelessness can be defined as the lack of access to minimally adequate housing, which usually refers to rooflessness or sleeping rough. A broader view of housing instability encompasses houselessness (a temporary place to sleep such as shelters or crisis accommodation), insecure housing (informal tenancies, difficulty paying rent/mortgage, threatened eviction) and inadequate housing (overcrowded or unfit housing) [11, 12]. –People experiencing homelessness have poorer health outcomes than their housed counterparts. They are at increased risk of complex chronic disease, infectious disease, mental health conditions and substance use [13, 14]. People experiencing homelessness have poorer health outcomes than their housed counterparts. They are at increased risk of complex chronic disease, infectious disease, mental health conditions and substance use [13, 14]. –They are frequent users of health services, particularly hospital emergency departments [15]. They are frequent users of health services, particularly hospital emergency departments [15]. –Healthcare settings are critical points for contact for early identification of homelessness or risk of homelessness [16]. Healthcare settings are critical points for contact for early identification of homelessness or risk of homelessness [16]. –Screening involves separating people who are homeless or at higher risk of homelessness from the broad patient population. The purpose of screening is to offer early intervention and improve health or other outcomes for screened individuals [17]. We define screening for homelessness or housing instability as the process by which all patients are asked about their housing situation before or during a healthcare consultation in a hospital emergency department or general practice clinic setting. Screening involves separating people who are homeless or at higher risk of homelessness from the broad patient population. The purpose of screening is to offer early intervention and improve health or other outcomes for screened individuals [17]. We define screening for homelessness or housing instability as the process by which all patients are asked about their housing situation before or during a healthcare consultation in a hospital emergency department or general practice clinic setting. –Knowledge of patients' housing status allows clinicians to adapt clinical care to better suit their patients and engage additional services that can assist with material needs and housing. These can lead to significant improvements in health and well‐being [18, 19]. Knowledge of patients' housing status allows clinicians to adapt clinical care to better suit their patients and engage additional services that can assist with material needs and housing. These can lead to significant improvements in health and well‐being [18, 19]. –Clinicians' and patients' views on discussing housing are not well understood, thus limiting the potential feasibility and acceptability of interventions to improve screening in healthcare settings. Clinicians' and patients' views on discussing housing are not well understood, thus limiting the potential feasibility and acceptability of interventions to improve screening in healthcare settings. We aimed to demonstrate that NPT can be applied when planning an intervention design, and to offer an example for how this may be done. This paper does not seek to describe our intervention as it is currently under development and will be discussed elsewhere. Instead, we describe our approach to intervention design. Methods Rationale for Selection of NPT The Normalisation Process Model is a precursor to NPT, which posits that collective actions (how people are involved in implementing an intervention) can affect whether the intervention becomes routine [4]. NPT adds three further constructs to enhance our understanding of how interventions become embedded into routine. They are: coherence (how people make sense of an intervention and how it is different from current practice), cognitive participation (how relational work enables people from different groups or roles within an organisation to work together to change practice) and reflexive monitoring (how people individually and collectively assess how well the intervention was implemented and whether it can be sustained) [4, 20]. NPT was chosen as it best aligned with our aim of understanding how clinicians make sense of a new intervention and how they work to integrate it into the context of their usual clinical and non‐clinical routines. We further applied NPT to understand how an intervention could be designed to avoid potential problems with implementation. Study Design A descriptive qualitative approach was used to gain an in‐depth understanding of the experiences of frontline health professionals when discussing housing with patients. Our use of a qualitative research design, with semi‐structured interviews to explore the applicability of NPT to intervention design, is in line with current approaches in this field. A recent review by Huddlestone and colleagues [8] found that 25 of 31 included studies used qualitative methods, with 10 of these using semi‐structured interviews for data collection. This study also identified the emerging use of NPT for intervention design. We follow the SRQR reporting guideline [21]. Researcher Characteristics and Reflexivity The research team included early career (N.G., T.L.C., K.N.) and established researchers (E.S., S.N., H.S.), with expertise in qualitative research (E.S., S.N., H.C., K.N., N.G., T.L.C.) and implementation science (E.S., S.N., H.C.). Several team members have a medical (E.S.) or allied health training (S.N., H.S., T.L.C.) with experience in healthcare delivery. E.S. is a general practitioner who works with people experiencing homelessness. One team member was from an Aboriginal and Torres Strait Islander background with expertise in Aboriginal and Torres Strait Islander health promotion and policy (B.H.). The researchers do not have lived experience of homelessness, however, we conceptualise stable housing as essential to good health, and that health professionals have a responsibility to assist their patients with social needs such as housing. None of the team members involved in data collection had pre‐existing relationships with participants. Participants and Setting There were three groups of participants: (i) general practice staff, including general practitioners, practice nurses and administrative staff, (ii) emergency department staff including doctors, nurses and administrative staff, and (iii) people with lived experience of homelessness (conceptualised as people who are currently homeless or living in unstable or unsuitable housing or people who have been homeless or lived in unstable housing previously). General practice staff from any Australian state or territory, and any type of practice including mainstream GP clinics, community health centres (CHCs), and Aboriginal Community Controlled Health Organisations (ACCHOs) were eligible to participate. Emergency department participants were drawn from two hospitals in the Frankston‐Mornington Peninsula region of Melbourne. People who self‐reported their lived experience of homelessness were recruited from a health service and from a non‐government organisation advocating for and representing people with lived experience of homelessness. Recruitment A recruitment flyer with information about the project, the study coordinator's (N.G.) contact details and an embedded secure survey link to capture interested participants' contact information was created. General practice staff recruitment: the flyer was distributed via email, newsletter and social media platforms to the professional network of a GP‐academic (E.S.). Emergency Department staff recruitment: departmental heads of emergency medicine and nursing were asked to circulate the flyer to all staff. Participants indicated their interest in the study by contacting the study coordinator directly or providing their details via the survey link. All interested participants were contacted by the study coordinator, who explained the study, screened the participant for eligibility (English‐speaking, working in one of the two study settings) and provided study materials. If the participant provided informed consent and interview date and time were arranged. Recruitment of people with lived experience of homelessness: a study flyer was circulated to staff at health services and a consumer advocacy peak organisation focused on homelessness/housing instability, and they were asked to explain the project to people with lived experience of homelessness and attain consent to share their contact details with the researcher if they were interested in finding out more and participating. All interested participants were contacted by the study coordinator (NG), who explained the study, screened the participant for eligibility (English‐speaking, aged over 18 years) and provided study materials. If the participant provided informed consent and interview date and time were arranged. Data Collection Semi‐structured interviews were conducted via telephone or videoconference by the study coordinator between August and December 2022. The interview schedule (Appendix 1) was developed by the research team and included open‐ended questions designed to elicit participants' perspectives on discussing housing in a healthcare setting or consultation, individual and systemic barriers to discussing housing and suggested changes to clinical routines and systems that could facilitate discussions about housing. Interviews lasted between 30 and 60 min. Interview progress was discussed during regular research team meetings. Interviews ceased when (i) all interested participants had been interviewed and (ii) no new data were emerging. All interviews were transcribed verbatim using Otter.ai, checked and deidentified before being imported into the Atlas.ti software [22] for analysis (N.G., T.L.C.). Data Analysis Our hybrid approach [23] to analysis involved inductive reflexive thematic analysis [24] followed by further deductive analysis using NPT to develop guidance on how to embed discussions about housing into healthcare consultations. Two team members were responsible for analysis, which included immersion through data collection (N.G.), transcription and verification (N.G., L.C.). Our thematic analysis [24] involved systematically reading through our data to identify key ideas (codes), then grouping those ideas into broader concepts that could address the research questions (themes). Initial codes were guided by the interview guide. Coding was done independently and discussed extensively with the team. Codes were then combined into themes, and themes were further analysed using the 4 constructs and 16 subconstructs of NPT. The themes and their relationship to the NPT constructs and subconstructs were reviewed by the broader team before finalisation. Negative case analysis was used to examine disconfirming cases and qualify our findings. Results were written up using the NPT framework, explanatory text and illustrative quotes. Our overall narrative presents the viewpoints of the majority of our participants, with disconfirming views explored where applicable. Ethical Approval and Consent to Participate Approval for this study was provided by the Monash University Human Research Ethics Committee (Approval Number 2024‐35179‐107482) and the Peninsula Health Human Research Ethics Committee (Approval Number LNR/93042/PH‐2023‐363038). Participants provided written or verbal informed consent before data collection. Participant codes are used for confidentiality. Results Participants A total of 39 interviews were conducted (see Table 1). Of the 30 healthcare workers, 53% (16/30) worked in primary care settings and 47% (14/30) worked in a Victorian hospital emergency department. Our sample comprises mostly women (77%, 23/30), aged between 25 and 34 years (40%, 12/30). Our primary care participants worked across fee‐for‐service private or mixed billing clinics (50%, 8/16) and free‐to‐user community health clinics (50%, 8/16). They were recruited from across five of the eight Australian states and territories. Our sample also included ACCHOs (19%, 3/16) that provide services to Aboriginal and Torres Strait Islander peoples. Nine people with lived experience of homelessness participated in the study. One person is Aboriginal and Torres Strait Islander. As described above, we used a convenience sampling method; we cannot comment on the representativeness of our sample. Table 1Participant demographics (N = 39).(a) Healthcare workersParticipant IDGenderAgeProfessional roleWorkplaceState of residencePH01Woman25–34Emergency medicineEmergency departmentVictoriaPH02Woman35–44NursingEmergency departmentVictoriaPH03Man25–34Emergency medicineEmergency departmentVictoriaPH04Man25–34NursingEmergency departmentVictoriaPH05Woman35–44NursingEmergency departmentVictoriaPH06Woman18–24NursingEmergency departmentVictoriaPH07Woman35–44NursingEmergency departmentVictoriaPH08Woman18–24NursingEmergency departmentVictoriaPH09Woman25–34NursingEmergency departmentVictoriaPH10Woman35–44Emergency medicineEmergency departmentVictoriaPH11Woman18–24NursingEmergency departmentVictoriaPH12Woman25–34NursingEmergency departmentVictoriaPH13Woman35–44NursingEmergency departmentVictoriaPH14Woman35–44NursingEmergency departmentVictoriaP001Woman35–44GPCommunity Health CentreQueenslandP002Woman45–54Administrative staffSolo GP clinicVictoriaP003Man65–74GPCommunity Health CentreVictoriaP004Woman45–54Practice nurseNon‐government organisation, primary health care clinicNew South WalesP005Woman25–34GPAboriginal Community Controlled Health ServiceNorthern TerritoryP006Woman35–44Administrative staffGroup or corporate GP clinicNew South WalesP007Man55–64GPAboriginal Community Controlled Health ServiceQueenslandP008Man35–44GPGroup or corporate GP clinicQueenslandP009Man35–44GPAboriginal Community Controlled Health ServiceNorthern TerritoryP010Woman25–34Administrative staffGroup or corporate GP clinicAustralian Capital TerritoryP011Woman55–64GPGroup or corporate GP clinicVictoriaP012Man65–74GPGroup or corporate GP clinicAustralian Capital TerritoryP013Woman35–44GPGroup or corporate GP clinicNew South WalesP014Woman45–54Practice nurseNot for profit drug and alcohol serviceAustralian Capital TerritoryP015Woman45–54Practice nurseGroup or corporate GP clinicAustralian Capital TerritoryP016Woman35–44GPCooperative clinicAustralian Capital Territory(b) People with lived/living experience of homelessnessParticipant IDGenderAgeState of residenceC01Woman45–54VictoriaC02Woman45–54VictoriaC03Woman55–64VictoriaC04Man45–54VictoriaC05Man55–64VictoriaC06Woman35–44VictoriaC07Man55–64VictoriaC08Man55–64VictoriaC09Man45–54Victoria Participant demographics (N = 39). We defined the intervention as “a process by which all patients are asked about their housing situation before or during a healthcare consultation in a hospital emergency department or general practice clinic setting”. This is distinct from routinely collected address information. The participants were clinicians (doctors, nurses, some allied health professionals), administrative staff and people attending the health service. Our results are structured by the four core NPT constructs that were used to interpret our qualitative findings. A complete description of how our data maps to the NPT constructs/subconstructs can be found in the Appendix 2. Coherence: How Do Participants Understand and Make Sense of the Intervention Healthcare workers from both emergency department and general practice settings reported that patients are routinely asked for their address in healthcare settings. This task usually falls to an administrative staff member (e.g. ward clerk in the ED, receptionist in a GP clinic). The purpose is to confirm a patient's identity and ensure they can be contacted following their appointment, rather than to identify housing instability. “it is part of the triage … because we have to ask for like an address, essentially, or forwarding address” – PH01 The notation “No Fixed Address” or “NFA” is often used in hospital emergency department settings as a proxy measure of homelessness or housing instability. Participants working in general practices in our study did not describe an equivalent in their setting. “We do get quite a lot of patients that do come in with no fixed address… It kind of starts right at the start of the clerical stage, when they go in and confirm all their details, you know, confirm their address, they'll say no fixed address” – PH08 Many health care workers felt that current practice was more likely to miss people experiencing housing instability or those who did not fit a stereotypical view of homelessness. “Oh, I guess it just depends on the context. Like if the person comes in, and they look homeless… like if they look homeless, which I know is quite… putting people in a box, but quite often they can appear to be unkempt, you'd just say, you know, do you have safe housing at the moment?” – PH02 One participant with lived experience of homelessness confirmed this perspective. “Probably because of the way I presented to them, they didn't think it was a concern. They didn't ask any questions specifically about my situation. They just, just went on their job, being a medical service rather than a housing service, so to speak.” – C08 We found that almost all healthcare workers understood the difference between the current practice and the proposed intervention. However, there were discordant views on the purpose of screening for housing instability. All clinicians working in CHCs or ACCHOs viewed asking about housing as integral to holistic care. This may be due in part to personal beliefs about healthcare and in part to the nature of their workplaces, which are more likely to provide services to people experiencing homelessness and housing instability. “If you don't have solid housing, how are you ever going to make progress with good health? So that's fundamental to me … There can't be a siloing between housing and healthcare. That just doesn't work.” – P004 Healthcare workers in ED settings or fee‐for‐service/mixed billing GP practices felt that housing was not relevant to discuss in healthcare consultations, citing lack of clinical relevance, limited understanding of the complexities of housing instability, limited time and limited ability to assist as key barriers. “The question is why? Unless the patient is bringing that up, or it's relevant clinically, I'm not sure whether that would filter through the type of things that might, could be … my time could be useful, I suppose.” – P008 ACCHOs were the only setting with established approaches to screening for housing insecurity. This may have been due to important differences in the model of care of ACCHOs compared to mainstream services, as described in subsequent sections. “We attempt to do Aboriginal and Torres Strait Islander health checks for every regular patient. And as part of that, we ask two questions. One about, ‘Do you feel that your housing is crowded? Do you live in crowded housing?’ And also, ‘Are you homeless?’ So that would be the‐ that's the time when we would ask systematically…. Well, it's meant to be annual. So yeah, people come in for their annual health check. And we would ask those questions.” – P006 Cognitive Participation: How Committed Are Participants to Changing Their Usual Routines Most healthcare workers in our study felt that discussing housing with patients was peripheral to their role of delivering health care. There was low overall commitment from mainstream healthcare workers to routinely discuss housing with patients. Participants were more likely to see value in asking patients about housing if this information was clinically relevant, for example, people presenting with mental health issues, substance use issues, family and domestic violence or frailty. “We certainly don't [ask] every patient. Certain patient groups, so, older patients, or patients with some mental health concerns, would be more likely to be [asked] about their housing or social situation” – PH10 There was a strong expectation in mainstream healthcare settings that patients were responsible for disclosing any concerns about their housing. “There are so many things we could be doing, that if a patient hasn't said, ‘This is a problem for me’. We probably don't have the capacity to kind of keep going and search for it more.” – P005 Healthcare workers from CHCs or ACCHOs did not share the views of their mainstream counterparts. “Normally … we do check with them [about housing]. And when we see them again, whether they're still sleeping rough; whether they want to be linked in with housing; whether they've had any forward movement” – P004 People with lived experiences of homelessness had conflicting views about discussing housing issues in health settings, too. Many were reluctant to tell healthcare workers about their lack of housing due to stigma, prior negative experiences or belief in the inability of healthcare workers to assist with housing needs. “Depending on the person, but me personally, I wanted to hide everything I was going through. I didn't want to tell the health service everything I was going through.” – C06 There were a few people with lived experience of homelessness who were open to sharing their housing situation, in the hope of being better understood and having their clinical care adapted to suit their living situation. “I would tell them about my van and my situation. And I think sometimes they misunderstood how difficult that was.” – C04 There was a general consensus among from healthcare workers in similar settings about which profession would be responsible for discussing housing with patients. In the ED, it was agreed that nurses would be well‐suited to the role, and in a primary care setting, either the GP or nurse could do the work. “It would have to be during the initial assessment from the nursing staff. That it would set the standard, added in question.” – PH14 “It is part of a GP's work, it's part of a remote area nurse's work, it's part of a child health nurse's work. We don't have to have a special social worker to do that.” – P005 Healthcare workers across a range of settings suggested ways to integrate screening for homelessness and housing insecurity into their work. Within ED settings, there was broad support for embedding a standardised question or question(s) about housing into ED intake and screening processes. Within a GP setting, there was support for incorporating housing‐related questions into health care plans or as a prompt within GP electronic medical recordkeeping software. “I think also if there were some systems, systems support in place, like a standardised questionnaire, or things that‐, questions that we just have to ask or boxes on our documentation that we have to fill out, then that also makes it easier because you have to do it. It's not just you being the one doctor that asks everyone.” – PH01 Healthcare workers in both settings highlighted the importance of additional training to support implementation, with a focus on what to do to help people who disclose housing issues. Lastly, many participants alluded to changes to organisational policy, culture and funding that could support the routinisation of discussions around housing. “I think it would be good if we actually had either some education from social work as to how to ask the questions, [and] what happens if someone does say to you they are experiencing homelessness, or they are at risk; what we can offer?” – PH06 Collective Action – How Will the Work of the Intervention Affect Participants' Current Role(s), and How Will It Affect Interactions between People and Technology As we did not seek to implement an intervention, our findings here relate to exemplar models such as ACCHOs, where discussions about housing occur routinely. We explore participants' perceptions about what works in their setting and what could work in other settings given their knowledge of health services. Participants from CHCs and ACCHOs had a strong understanding of the importance of asking about housing routinely (Coherence) and commitment to doing so (Cognitive Participation). Most participants from CHCs and ACCHOs were routinely asking patients about their housing situation. Many participants also described having more experience in managing complex presentations such as homelessness than their mainstream counterparts. In my perspective, it's completely embedded, like, you can't separate [housing from] any of it. It's all interrelated. That's the difference of … having the experience and … [experience with] this type of complex presentation with complex trauma and, and housing instability and food insecurity and chronic disease and substance use. – P001 Participants from ACCHOs also described how adaptations to software systems, alternate models of funding, access to a multi‐disciplinary teams and reliable referral pathways supported routine discussions around housing. “Being able to have the assistance from the social worker and spend that time and talk to them about, ‘Okay, so we need this report from here, this report from there’, has proved wonders. And also with the social worker … we have a referrals pathway … [for] every single situation you can think of: homelessness, domestic violence, social prescribing, grief, loneliness, mental health, social prescribing activities. So it's this huge guide that just keeps growing. – P007 Participants from mainstream services such as the emergency department and fee‐for‐service general practice identified components of interventions that would allow screening for homelessness to be embedded into health settings. They spoke about the need to address knowledge and skill gaps by improving health professionals' understanding of how housing affects health, making the benefits of asking about housing clear and relevant to health professionals, and building health professionals' skills in asking about housing. “One of the first things [is] just to raise awareness of housing as an issue and what you could potentially do for people and how to potentially ask about it.” – PH11 Mainstream participants also spoke about strengthening collaboration between professional roles, for example, by clarifying referral pathways from health services into social work or housing support services. “There's not like a natural referral pathway to go through, and it seems to be somewhat haphazard and somewhat, as I said, I need to, at least for me, need to relearn it every time.” – P008 Lastly, healthcare workers in mainstream settings described technological and procedural aids that would help make discussions about housing routine such as embedding housing questions into a health assessment or intake forms. So, in the health assessments … [it's] embedded in it. So it comes very naturally, in a way … So we also ask, you know, ‘Who do you live with?’, ‘Where do you live?’, and ‘Do you feel safe at home?’ So it kind of‐, it just rolls into all the other social questions because they, the patients, have an expectation that we're asking, you know, about their alcohol use, whether they smoke, or all those other health parameters that patients are used to doctors asking about. And it works very well. – P011 Participants from specialist settings such as CHCs and ACCHOs spoke about how funding models in their organisations did not rely on time based billing and that this enabled them to discuss housing with patients. Conversely, clinicians in mainstream settings that are more reliant on patient throughput may find it harder to justify spending this additional time with their patients. “Because we're not reliant on the funding. So, that's, that's a really big difference in our model…. So, even though we'd like to maximise our billing, it's, our GPs are all salaried. So I think, you know, that makes a huge difference. If you're not‐ if the work that you're doing is not reliant on you kind of really pumping through the numbers, you can spend an hour with a client and not feel that going to lose out financially for that.” ‐P004 Reflexive Monitoring – What Feedback Do Participants Offer on the Intervention Following Implementation As an intervention was not performed, we are not able to comment on the way routine screening for housing impacts healthcare professionals' work and others around them. This NPT construct did not inform the findings of this study. Discussion NPT is a useful theoretical framework for carefully planning a complex intervention in healthcare. The current literature suggests that NPT is used in qualitative evaluations of feasibility trials to allow potential pitfalls for normalisation to be identified and addressed before full implementation [25, 26, 27]. Another application of NPT is in planning for scale up and translation where evidence from prior studies is used to adapt the intervention to be more easily embedded into daily practice [28, 29]. A recent systematic review [8] identified 11 studies that used NPT to develop interventions, with nine using NPT to plan for full implementation or scale up as described above. Only three papers reported using NPT to examine how a proposed intervention could be embedded into existing routines. This paper adds to the existing evidence base by demonstrating that NPT can be used to test whether the principles underpinning an intervention are acceptable to people who will be tasked with doing the work, and people who will be affected by the work. We also show that this novel application of NPT at this early stage allows intervention developers to better understand what intervention components could improve uptake and facilitate normalisation. Our case study illustrates the application of NPT as an early planning tool. Our findings in the Coherence domain suggest that most healthcare workers understand the benefits of knowing about a person's housing situation for providing holistic clinical care, particularly when mental health, family violence and substance use issues are present. They also had a good awareness of the difference between routinely asking about housing and their current practice. However, Cognitive Participation may be limited, with healthcare workers in mainstream hospital and primary care settings reluctant to change their current routines. We also found challenges within the Collective Action domain as mainstream healthcare workers were unsure about (i) how to ask about housing, (ii) what to do when housing issues are identified and (iii) whether discussing housing will have any adverse effects for themselves or their patients. We were unable to comment on Reflexive Monitoring as it is not part of this stage of early intervention planning. Our case study also demonstrates the synthesis of qualitative data into evidence to support intervention design. We show that Cognitive Participation can be strengthened by improving healthcare workers' knowledge and self efficacy [30], e.g. understanding of the different forms of homelessness and housing insecurity, how housing issues impact on health, how to sensitively ask about housing and how to respond when people disclose housing issues, including tailoring healthcare to an individual's social situation even if housing cannot be immediately addressed. Exemplar screening questions are available within Aboriginal and Torres Strait Islander health assessment forms. Collective Action may be supported with clear guidance on how professional groups could work together. Our data suggest doctors or nurses are best placed screen for housing issues [31] and that they are more likely to do this if they have reliable referral pathways to engage health professionals with housing expertise e.g. social work or housing support services. Technological aids, such as screening prompts in electronic medical record software, could also support uptake of screening for housing issues. We recognise that an intervention to screen for homelessness, like most others in health settings, would be introduced within the context of structural barriers such as limited time, limited staffing, and limited funding. It is important to recognise that mainstream healthcare workers will be unable to add this new task to their usual routines without amending or redistributing other tasks. If additional investment is not possible, intervention planners would need to consider how tasks may be reallocated within health professional teams to create more time within workflows for discussions about housing while minimising the impact of this extra work on individual workloads. Intervention planners would also need to consider what additional training and support healthcare workers require to manage the professional discomfort of knowing their patient does not have stable housing without being able to fix it. Lastly, intervention planners would need to develop clear, evidence‐informed guidance on how health professionals could adapt usual clinical care to improve treatment adherence and patient well‐being to make clear the potential benefits of continuing to discuss housing with patients. Limitations This paper focuses on the perspectives of healthcare workers on screening for homelessness and housing instability as they pertain to intervention design and implementation. Healthcare workers who believe that healthcare delivery to people experiencing homelessness and housing instability could or should be improved may have been more likely to participate in this study (selection bias). The researchers have sought to present different viewpoints and qualify findings to minimise the potential impact of this. The perspectives of people with lived experience of homelessness are only discussed in this paper as they relate to intervention design. A second paper is planned to more fully explore how people with lived experience of homelessness sought, accessed and utilised health care. Conclusion This study shows that NPT can be applied to the exploratory analysis of qualitative data to gain insights into the contextual factors at the organisational, professional and patient levels that will affect the implementation of an intervention. Our approach can be used alongside conventional intervention design principles to guide the development of complex healthcare interventions. Author Contributions Nilakshi Gunatillaka: conceptualization, data curation, formal analysis, methodology, project administration, writing – original draft, writing – review and editing. Tze Lin Chai: data curation, formal analysis, writing – original draft, writing – review and editing. Suzanne Nielsen: conceptualization, formal analysis, funding acquisition, methodology, writing – original draft, writing – review and editing. Belinda Hammond: formal analysis, methodology, writing – original draft, writing – review and editing. Kimberley Norman: formal analysis, methodology, writing – original draft, writing – review and editing. Helen Skouteris: conceptualization, funding acquisition, methodology, writing – original draft, writing – review and editing. Elizabeth Sturgiss: conceptualization, data curation, formal analysis, funding acquisition, methodology, project administration, supervision, writing – original draft, writing – review and editing. Conflicts of Interest The authors have no conflicts of interest to declare. Declarations All authors have read and approved the final manuscript. Nilakshi Gunatillaka has full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis. Transparency Statement The lead author, Elizabeth Sturgiss, affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. Supporting information Appendix 1A: Interview guide for people with lived experience of homelessness or unstable housing. Appendix 1B: Interview guide for clinicians and administrative staff working in general practice (from anywhere in Australia). Appendix 2: Complete analysis by NPT component and subdomain. Acknowledgments Funding for this work was received through the National Centre for Healthy Ageing (NCHA). The NCHA acknowledges the support provided through the Commonwealth Department of Health and Aged Care through the Community Health and Hospitals Program. The funder had no involvement in study design, collection, analysis or interpretation of data. Open access publishing facilitated by Monash University, as part of the Wiley ‐ Monash University agreement via the Council of Australasian University Librarians. Gunatillaka N., Chai T. L., Nielsen S., et al., “Using Normalisation Process Theory to develop a complex intervention: a qualitative case study of screening for homelessness in healthcare settings,” Health Science Reports 9 (2026): e72011. 10.1002/hsr2.72011. Nilakshi Gunatillaka and Tze Lin Chai are joint first authors. Data Availability Statement The datasets generated and/or analysed during the current study are not publicly available as our qualitative data includes personal narratives which, if made available broadly, may compromise individual privacy. Data can be made available from the corresponding author on reasonable request and with suitable ethical approvals. References Associated Data Supplementary Materials Appendix 1A: Interview guide for people with lived experience of homelessness or unstable housing. Appendix 1B: Interview guide for clinicians and administrative staff working in general practice (from anywhere in Australia). Appendix 2: Complete analysis by NPT component and subdomain. Data Availability Statement The datasets generated and/or analysed during the current study are not publicly available as our qualitative data includes personal narratives which, if made available broadly, may compromise individual privacy. Data can be made available from the corresponding author on reasonable request and with suitable ethical approvals.