Online postal self-sampling (OPSS) is a digital health intervention that allows service users to screen for sexually transmitted infections (STIs) and blood-borne viruses (BBVs) by ordering a self-sampling kit online, taking their own samples, returning them to a laboratory for testing, and receiving their results remotely [1]. OPSS differs from self-testing in that samples are sent off for processing in a laboratory, rather than being processed within the test kit [1-3]. Results are usually communicated electronically, and users in need of treatment or partner notification are typically referred to clinic-based sexual health services [4].

Many settings globally are developing and implementing OPSS, with its availability and uptake having been accelerated during the COVID-19 pandemic [5-8]. The proportion of tests conducted via OPSS within England’s National Chlamydia Screening Programme, for example, rose from 19% in 2019 to 43% in 2023 [9]. This transition is occurring within the context of wider efforts to digitize health care; there is a national recommendation that OPSS should be provided as part of all sexual health services in England, while the World Health Organization recommends that self-sampling be provided alongside clinic-based sexual health services to improve the uptake of chlamydia and gonorrhea testing [10,11].

In England, public health commissioners are responsible for the process of planning, agreeing, purchasing, and monitoring health services [12]. The commissioning of free, open-access STI or BBV testing has primarily been the responsibility of local authorities since the implementation of the Health and Social Care Act 2012 [13]. Commissioners are responsible for identifying a preferred model of service provision and instigating a tendering process to select preferred sexual health services that can deliver the model within budgetary constraints. Although OPSS is increasingly part of sexual health service models, it has been adopted in different regions of England at different times, with different models of delivery. For example, OPSS is delivered by National Health Service (NHS) trusts in some areas and outsourced to sexual health clinical services, laboratories or online testing platforms provided by the private sector in others [14-17]. There is a lack of evidence as to why certain delivery models were decided on, how they have been implemented, or how implementation strategies have influenced the sustainability of OPSS into sexual health service delivery.

There is extensive evidence about factors influencing the adoption and embedding of digital health interventions in health care [18]. However, there are fewer studies that have considered the sustainability of their implementation, and this evidence gap has been noted specifically in relation to OPSS and other remote testing services for STI or BBVs [3,19]. In addition, it is unclear how applicable pre–COVID-19 findings (before March 2020) are to the current context of digital health implementation. This is important, given the impact of COVID-19 on normalizing online health care delivery [20].

This study is nested within the assessing the impact of OPSS for STIs on health inequalities, access to care and clinical outcomes in the United Kingdom (ASSIST) study, a comprehensive evaluation of OPSS within the UK context, that aims to assess the impact of these services on health inequalities, access to care, and clinical and economic outcomes and to identify the factors that influence the implementation and sustainability of these services [21]. This study focuses on the implementation of OPSS in 3 case study areas (CSAs) of England. In so doing, it addresses important evidence gaps in digital STI or BBV testing.

Evaluation of OPSS implementation from staff and stakeholder perspectives formed 1 workstream within ASSIST, a wider mixed methods realist evaluation of OPSS, with the published protocol documenting the full study methods and design [21]. Separate papers covered the findings from other workstreams, for example, on service user experiences [22], health inequalities, access to care, and clinical and economic outcomes (Gibbs J, unpublished data, April 2025 and Jackson L, unpublished data, April 2025).

In this study, a combination of qualitative data collection methods were used: interviews, document analysis, and contextual observation. Methods and findings are reported in line with COREQ (Multimedia Appendix 1). Each data source had a distinct primary purpose and was initially analyzed separately, with preliminary findings from each compared, triangulated, and explanatory detail added where available. Details of implementation models, impacts, and sustainability were drawn primarily from interviews. Observations were used to understand how OPSS featured in clinic-based settings and to complement experiences of OPSS drawn from interviews. Documentary sources were primarily used to construct a timeline for implementation.

The design and methods were informed by normalization process theory (NPT), a substantive middle-range theory that has been used extensively for the study of the work required for initiating, integrating, and embedding (normalizing) digital health innovations into practice [23].

We evaluated OPSS implementation in 3 CSAs in England, labeled CSA 1, CSA 2, and CSA 3 for anonymity. The CSAs were purposively selected based on them having undergone a distinct commissioning process for OPSS, with implementation at different times and the continuation of clinic-based services alongside OPSS. This enabled exploration of different delivery models at different stages following OPSS implementation. Each CSA was predominantly urban and had a population of at least 500,000, which was diverse in terms of socioeconomic circumstances and the proportion of people who are: ethnic minorities; lesbian, gay, bisexual, transgender, queer, and similar minorities; and young. These populations may be disproportionately affected by STIs or BBVs [21]. The focus of data collection was clinic-based and OPSS service settings, with 1 clinic-based service in each CSA chosen as the setting for contextual observation. These clinics were all level-3 sexual health services—offering the most complex care—and situated near the center of each CSA.

Interviews were the dominant source of data in our evaluation, and we first coded these sources using a coding framework based on NPT to analyze them. This initial analysis was used to develop early “pen portraits” of our CSAs and to inform analysis of documentary sources (ie, to identify pivotal contextual events or circumstances and the key implementation milestones in each site). Extraction from documentary sources was used primarily to provide the “hard” data on when key events happened and also, where available, to triangulate interviewees’ recollections.

As our analysis progressed, we inductively identified that the implementation of OPSS in all CSAs was part of—or took place alongside—substantial transformation of sexual health services. Therefore, we used other frameworks to inform our coding to complement NPT and added CFIR (Table 1), guided by the Dissemination and Implementation Models in Health webtool [26]. We applied a conceptual framework relating to major system change, developed by Fulop et al [27], to structure the account of how OPSS was conceived and implemented. We were also guided by the implementation outcome measures proposed by Proctor et al [30], which provided a structured framework for reporting implementation outcomes.

We used these frameworks to develop a table that enabled us to make an initial comparison of the implementation in each CSA, from adoption to outcomes. We continuously refined the table categories and content as analysis progressed until we created the version and accompanying narrative presented in the results. Documents were used as a comparison with the interview data to explore where the documents confirmed, contradicted, or provided further explanation to the interview data. Contextual observation enabled us to test our findings derived from analyzing the interview data using the NPT framework, examining whether OPSS appeared in staff workflows as described in the interviews.

Ethics approval was granted by the NHS South Central—Berkshire B Research Ethics Committee (21/SC/0223). All participants provided written informed consent to be interviewed or observed as relevant to their role in the study. Data were deidentified by not naming the CSAs, removing any names from transcripts, and listing job categories rather than individual roles. Participants took part without compensation.

Two contextual factors determined the implementation approaches available to service leaders and had a substantial impact on the implementation outcomes in all CSAs. The influence of each factor was further impacted by a third contextual factor, the COVID-19 pandemic.

This study of the implementation of OPSS for STI or BBVs across 3 diverse CSAs found that budgetary pressures and organizational relationships determined the implementation strategies available to decision makers and how these strategies were enacted. Our study additionally illustrates the profound but divergent effects the COVID-19 pandemic had on OPSS implementation in different contexts.

Our study has several strengths. Its multisite design allowed comparison of implementation strategies, and our combination of documentary sources over 12 years, with contemporary interviews and observations, enabled us to examine outcomes, including sustainability. Combining interviews with documentary sources and contextual observations provided richer insights than interviews alone. For example, the observations generated a more nuanced understanding of the collective action in “the real world,” where OPSS was less dominant than in an interview specifically about OPSS implementation. It also provided unintended insights into sustainability, for example, how CSA 3 implemented a daily cap on OPSS testing. Our use of multiple theoretical frameworks allowed these to complement each other and develop a more comprehensive understanding of implementation outcomes—and the factors that led to them—at different levels. The major system change framework helped to standardize the narrative across the 3 CSAs, enabling us to compare implementation approaches and outcomes, while NPT was particularly helpful for surfacing the implications on staff work (captured under “collective action”). CFIR enabled us to capture wider contextual factors, such as financing, policies, and laws, that influenced our findings using the other 2 frameworks.

Our study faced some limitations. The CSAs in this study were very different in size and complexity. On the one hand, this was a strength of the study in surfacing contrasting implementation approaches and outcomes within the same sexual health care system. On the other hand, it affected the research processes we were able to undertake. For example, we were not able to source key documents from all CSAs, such as tender specification and contracts. This was in part because nearly 10 years had elapsed since OPSS had launched in 1 CSA, and staff turnover meant no one in post still had access to documentation from the time of adoption. This was partly due to different implementation approaches; those that followed more formal processes had extensive documentation, whereas others documented fewer aspects. This made it harder to ascertain certain aspects of implementation and compare between CSAs and to corroborate some of the interview data. The time elapsed since the launch of OPSS in some CSAs may have introduced recall bias in certain interviews. In addition, some of the study team worked for, or closely with, sexual health services included in our evaluation. However, the data collection and analysis were conducted by researchers who did not have these relationships.

The coherence of OPSS to staff in this study aligns to some extent with the findings of other studies. In their review of home-based sexual health care, Goense et al [3] reported that staff believed “home-based STI testing” could improve access, especially for clients vulnerable to STIs. The interview study staff in the study by Chabot et al [33] articulated an expectation that home-based STI testing would provide increased convenience for users. However, there was less evidence of staff conflict or a lack of coherence in these studies than we identified in our study. The research by Ryu et al [34] on the experiences of sexual health services during COVID-19 lockdowns in Ontario in Canada did identify some concerns among staff. They were primarily concerned about the risks to service users not having sufficient privacy at home to access services remotely and staff finding it difficult to provide emotional support and education outside of face-to-face interactions. While staff concerns in this study did include risks to service users, the ambivalence or opposition to OPSS in some CSAs may have been driven by wider sexual health transformation changes, which also affected staff employment and service viability [35].

Ryu et al [34] identified that the pandemic catalyzed the adoption and implementation of remote models of sexual health care, which persisted beyond the pandemic. This aligns to some extent with experiences in 1 of the 3 CSAs in our evaluation, where expanding the scope of OPSS improved its acceptability among staff in CSA 2. However, our findings illustrate how COVID-19 could have contrasting effects. It severely undermined OPSS in CSAs 1 and 3, although for different reasons. The interaction of the organizational context and OPSS implementation approaches helps us to understand why the pandemic had such contrasting effects. These findings contribute toward addressing an evidence gap articulated by Goense et al [3] by providing insights on what affects the sustainability or maintenance of OPSS services. They also address a wider gap in implementation science literature, where context has been well established as an important factor in implementation, but there is limited evidence on how it can dynamically affect implementation approaches, particularly over time, or its relationship with implementation outcomes [19,25,28,32].

Existing research has observed that financial resources are one of the most commonly cited contextual factors affecting implementation, and as Nilsen et al [32] note, implementation research so far has focused predominantly on the need for sufficient financial resources to enable implementation to happen [36]. However, our study affirms a contrasting perspective, in that OPSS was implemented precisely because of concerns about insufficient financial resources [37]. Commissioners and services appeared to be faced with an unenviable choice: implement this “new, untested” service or face a failure of sexual health services altogether [38]. This distinct context, where implementation is initiated because of financial pressures, may help explain some of the surprising findings of this study and could have implications for other digital health innovations. For example, NPT posits that embedding of an innovation is more likely when staff perceive its benefits [23]. In this study, clinic staff acceptability did not predict implementation success; instead, implementation success—or otherwise—influenced acceptability.

We initially approached OPSS as a digital health innovation and sought to evaluate it as such. However, our findings suggest that implementation of OPSS was part of a process of large-scale service transformation [39] and as such has more in common with digital transformation, in that it involved several organizations and was associated with changes in funding and staffing models across both clinic-based and internet-based services. The rapid pace of this transformation in 2 of our CSAs—which each implemented OPSS within 6 to 12 months—offers insight into why they faced challenges with sustainability, especially when compared to the more gradual implementation in CSA 2 and other digital sexual health innovations internationally, such as the 3-year implementation of Get Checked Online in British Columbia in Canada [40]. This rapidity may have hindered the capacity to obtain sufficient local evidence to effectively guide implementation and suggests that future studies of the implementation of OPSS or similar intentions should consider, as Chabot et al [33] have done, how OPSS is embedded in the wider social and structural processes of the public health system.

In this multisite case study, OPSS implementation was part of systems change to address a wider problem of insufficient funding to deliver sexual health care. This systems change, alongside organizational relationships, determined the implementation strategies available to decision makers and the outcomes they achieved, as evidenced by stark differences in sustainability during COVID-19. Implementation of OPSS may have progressed before sufficient evidence to effectively guide implementation was available.

On the basis of our findings, we provide 3 actionable recommendations for implementing OPSS that could contribute to successful and sustainable implementation of OPSS services into sexual health systems.

First, those planning to introduce OPSS should allow sufficient time in the preparatory phase to iterate plans, informed by stakeholder feedback, to build coherence among staff and service users. Our study documented that the process of introducing OPSS varied significantly between CSAs in terms of timelines, stakeholder engagement, and how formalized the process was. Our findings suggested that very rapid implementation may affect later sustainability and that local authorities commissioning as collectives are able to deliver OPSS at lower cost, particularly when outsourcing. Sexual health system leaders considering introducing OPSS would benefit from considering the implications of their business and delivery models before implementation.

Second, professionals delivering OPSS and those commissioning OPSS should build flexibility into implementation so that delivery and models can be altered to respond to changes in context. Our findings illustrate how implementation of OPSS is not a one-off event but an ongoing process that is sensitive to the context into which it is introduced and sustained. In our CSAs, important contextual issues included public service budget cuts, the sexual health service transformation agenda, and the COVID-19 pandemic, which affected the process and outcomes of implementation, particularly CSAs’ capacity to respond to changes in demand for OPSS. The nature of contextual changes may not be predictable, but building in the capacity to increase or reduce delivery of OPSS—and rebuilding staff coherence around any changes—could enhance sustainability of implementation.

Third, professionals delivering OPSS and those commissioning OPSS should reflexively monitor changes to staff workload across the sexual health system following introduction to OPSS. Our study found that introducing OPSS into sexual health systems changes the work of different staff groups in heterogeneous ways. Some of these changes could not have been anticipated, but continually monitoring and meaningfully responding to such changes could enhance the acceptability of OPSS to staff, enhancing its adoption and sustainability.