Informing the design of a digital intervention to support sexually transmissible infection care in general practice: a qualitative study exploring the views of clinicians

Introduction

Sexually transmissible infections (STIs) cause significant harms in Australia.¹ Rates of bacterial STIs are increasing; between 2011 and 2019, Chlamydia trachomatis (chlamydia) by 32% (n = 107 286 cases in 2019), Neisseria gonorrhoea (gonorrhoea) by 188% (n = 34 780 cases in 2019) and Treponema pallidum (syphilis) by 344% (5912 cases in 2019).² These STIs are often asymptomatic and may not be detected, but can cause serious complications. Chlamydia and gonorrhoea can ascend to the upper genital tract to cause pelvic inflammatory disease (PID) and tubal infertility for women^3,4 and epididymitis for men.⁵ Anti-microbial-resistant gonorrhoea is of global concern as it introduces the prospect of untreatable infection.⁶ Furthermore, rising heterosexual syphilis transmission between 2011 and 2020 saw a >400% increase in the number of syphilis cases among women and 58 cases of congenital syphilis.²

Timely diagnosis and optimal management are crucial for reducing STI transmission and their associated adverse impacts. In Australia, specialist STI care is provided in sexual health and family planning services, but they are at capacity and not widely available outside of metropolitan areas. General practice is Australia’s mainstream primary care setting. Visited by >80% of Australians at least annually,⁷ it is the most accessible option for STI testing and care for most Australians. Barriers to general practitioners (GPs) routinely offering STI testing and providing optimal STI care have been identified and include knowledge/skill deficits (e.g. who to test, what tests to offer), time constraints (e.g. STI screening in addition to the patient’s reason for the visit), and lack of recall systems.^8,9 Strengthening STI testing and management in general practice is prioritised in Australia’s Fourth National STI strategy.¹

Computerised clinical decision support systems (CDSSs) are gaining increasing attention in health care and may offer a mechanism to support general practice in STI care provision. Furthermore, digital infrastructure and clinical decision support is a priority for Australian primary care going forward.¹⁰ Recent advances have seen CDSS development with algorithms that integrate and query data within the general practice electronic medical record (EMR) system to identify at-risk patients who can be targeted for intervention.^11–13 Such tools can operate both in and out of clinical consultations to provide clinicians with computer prompts for patient-specific recommended actions and links to clinical guidelines.^11–13 To date in Australia, CDSSs have largely been developed and evaluated for prescribing or chronic disease. A CDSS to support antibiotic prescribing was viewed as acceptable by Australian GPs and assisted appropriate prescribing in a simulation study.¹² A cardiovascular disease CDSS supported improvements in risk factor screening in Australian primary care; however, recommended medication prescribing did not change.¹⁴ In the United States, CDSSs for STI care have shown some success, leading to increased chlamydia and STI screening in paediatric settings.^15,16

To date, there is no CDSS for STIs care provision in Australian general practice and no research within this context to inform the development of one. In this study we aimed to: (1) understand clinicians’ views on how health professionals would use a proposed STI CDSS; and (2) identify barriers and enablers to the use and implementation of a STI CDSS in primary care. Through determining end-user viewpoints, the results of this study will be used to inform the design and development of a suitable STI module for an existing CDSS that we have developed for chronic disease in general practice.¹³

Methods

We undertook a qualitative study involving one-to-one semi-structured interviews to explore the perspectives of GPs and sexual health physicians (SHPs) working in the state of Victoria, Australia, on a proposed CDSS for STI care in general practice, including potential barriers and enablers to its use and implementation. Our approach was informed by the Theoretical Domains Framework (TDF), an integrative behaviour change framework, proposed by Michie et al. in 2005¹⁷ and revised by Cane et al. in 2012.¹⁸ The TDF comprises 14 domains created from 84 theoretical constructs that may explain health-related behaviour change and provides an empirical method for assessing implementation problems¹⁸ and key behaviours important for intervention implementation. It has been successfully used as a framework to facilitate identifying barriers and enablers in implementation research.^19–21 In this study, the TDF was used to understand aspects of health professional behaviour related to using a theoretical CDSS and providing STI care, and to identify potential barriers and enablers that may influence intervention development and use, to facilitate broader implementation. It further provides a theoretical scaffold for addressing such determinants if and when such a CDSS is developed and implemented.

We drew on our existing CDSS, called Future Health Today (FHT),¹³ to explore how a CDSS for STI care might look and operate. In brief, FHT is a software with embedded algorithms that query data in the EMR to automatically identify patients meeting specific criteria who can be targeted for care. Triggered by these algorithms, FHT’s two main components are: (1) the dashboard, used outside of consultations to create lists of patients who can be proactively targeted for care (e.g. recall); and (2) the point-of-care, which integrates evidence-based clinical decision support, guidelines, and recommendations for management to guide care during consultations. First developed for chronic disease, FHT was co-designed with general practice staff and health consumers to ensure it was acceptable, met clinical needs and fitted into general practice workflow.¹³ Specifically exploring the usefulness and appropriateness of FHT for sexual health care is important given the unique sensitivity and likely nuances of the subject area compared to chronic conditions like cardiovascular disease.

Results

A total of nine participants (SHPs = 2, GPs = 7) were interviewed (eight on Zoom, one by telephone) with interview duration ranging from 23 to 42 min. All practiced in a metropolitan setting, and most were female (67%). In regard to current practices in providing STI care, GPs described many challenges including following-up positive cases, raising sexual health matters, offering testing in non-sexual health consultations, and keeping abreast of guidelines. GPs emphasised challenges and knowledge gaps in providing syphilis care, including diagnosing, interpreting serology, and providing treatment, which were largely related to the complexity of syphilis pathophysiology and seeing infrequent syphilis cases.

Ten subthemes regarding the use and implementation of a CDSS for STIs emerged across the six TDF domains (detailed in Table 1) and a range of enablers and barriers to implementation of a STI CDSS were identified. The TDF domains ‘Beliefs about consequences’, ‘Knowledge’, and ‘Skills’ captured enablers to a CDSS for STI care and the TDF domains ‘Optimism’, ‘Environmental context and resources’, and ‘Memory, attention and decision processes’ captured a mix of enablers and barriers. The emergent subthemes are grouped as enablers or barriers under their relevant TDF domain and are described below with example quotes in Table 2.

Discussion

To our knowledge this is the first study exploring clinician views on a CDSS to support STI care in Australian general practice. As such, our findings provide important considerations for developing such a tool. We found that GPs and SHPs viewed the concept of a CDSS for STI care favourably, suggesting a possible role in Australian general practice. Interviewees were optimistic about a digital tool’s potential to improve STI care processes, address some challenges in providing STI care, and facilitate STI education. For such a tool to be appropriate and useful, it must meet end-user needs, align with clinical context, and integrate key resources, such as testing and treatment guidelines. Insight was provided into envisaged uses of such a tool. More complex conditions (e.g. syphilis) were seen as areas a digital tool may most benefit.

Our study drew on the features of FHT, specifically the dashboard and point-of-care components, to explore views on the use and usefulness of a CDSS for STI care. Our sample of clinicians considered they would be less likely to use the dashboard than the point-of-care function and attributed this to concerns around accuracy of recall processes and the risk of recall notifications being misinterpreted or otherwise negatively impacting patients. Co-design is essential when considering these concerns. One possible solution that could promote patient autonomy could be an approach used in an Australian sexual health clinic whereby men who have sex with men (MSM) patients could opt-in for automated computer-generated reminders for STI screening.²⁷ FHT’s point-of-care component was viewed positively for STI care for which preferences included targeted and actionable pop-ups that were also modifiable (e.g. frequency, type), direct links to appropriate resources to guide clinical actions (e.g. clinical guidelines); and information presented unobtrusively to align with workflow. These desirable components are already available within FHT, suggesting that co-design for an STI module should focus on the displayed recommendations, integrated resources and underlying algorithms.

Interviewees emphasised their preferred areas of care and content for a STI CDSS including how it is presented and engaged with. Prompts for testing to detect new infections and for timely re-testing for re-infection were identified as areas that may benefit due to current gaps in STI detection and re-testing.²⁸ Syphilis care was viewed as an area that may benefit due to it being infrequently cared for by some GPs, and PID as an area a CDSS might support decision-making. GP and SHP interviewees suggested many resources that covered testing, infection management, patient information and emphasised the need to incorporate trusted sources. A 2021 Australian study found that trust and confidence in the content influenced GPs’ intention to engage with a CDSS for antibiotic management.²⁹ Co-design is imperative to ensure integrated resources reflect end-user needs and expert input, as positive user attitudes are important facilitators of CDSS adoption.^30,31

Key attributes emphasised for a successful CDSS design included simplicity (e.g. guidance that is quick and easy-to-follow, and actionable), useability (e.g. alignment with clinical processes), flexibility (e.g. potential to customise resources and prompts), and consideration of time constraints. These attributes are largely consistent with preferred features of other CDSSs. Useability is an important facilitator of CDSS engagement.³² Simplicity for design of a proposed colorectal cancer CDSS was viewed positively, whereas complexity was viewed as a barrier to adoption.³³ Our interviewees viewed the notion of an inflexible CDSS negatively, which may reflect preferences for clinical autonomy. Time to use a CDSS was viewed as a hindrance if it created extra work or an enabler if it streamlined care. This dichotomy was highlighted in a systematic review of medication-related CDSSs in which time constraints were also perceived as a barrier or facilitator to adoption.³⁴

The patient context was an important consideration. A CDSS’s potential insensitivity to contextual factors necessary for STI-related clinical decision-making was perceived as a possible barrier to use. For example, general practice EMRs may not contain necessary data (e.g. sexual orientation, number of sexual partners) in an extractable form that can be incorporated into CDSS algorithms. As such, a CDSS may be limited in available data to query to guide care, with a risk of imprecise prompts that may compromise both clinicians’ trust in the tool and delivery of optimal care.

Pop-ups were viewed positively and negatively. On the positive side, they were seen as a mechanism to communicate knowledge, guide clinical behaviours, and an avenue that could facilitate improvements in knowledge and skills via engagement with recommendations and integrated resources. Suggestions for their use included re-testing reminders, identifying high-risk patients for testing, and recommending clinical decisions and management. Concerns included that pop-up relevance depends on correct data in the EMR and that they can be excessive or perceived as unimportant, leading to pop-up fatigue.³⁵

Our findings show that the concept of a CDSS for STI care was viewed favourably and provides a valuable starting point for intervention design. Whether the preferred features as described can translate into care improvements is unknown. Despite considerable enthusiasm for the potential of CDSSs, two recent systematic reviews have shown variable impact on outcomes. A meta-analysis investigating provider uptake among 55 studies found that only one-third of clinicians used the CDSS and that quality of patient data underlying the CDSS was positively associated with uptake.³⁶ Another meta-analysis (122 studies involving over 1 million patients and 10 000 clinicians) found that CDSSs increased the proportion of patients receiving desired care by 5.8%.³⁷ To hypothesise the potential impact of a STI CDSS, we translate this proportion to a quantifiable gap in STI care. Australian guidelines recommend a chlamydia retest around 3 months after treatment to check for reinfection; however, retesting rates are around 25%.³⁸ A 5.8% increase would only increase retesting rates to around 26.5%. The potential harms of undiagnosed infections or sub-optimal care are key considerations for any CDSS for STI care. In view of increased heterosexual and congenital syphilis, and updated guidelines to include syphilis in routine STI testing, clinical decision support to prompt syphilis testing could be a key area of focus and potential benefit. Consideration will also need to be given to other factors such as development and availability of resources and education for GPs to access the necessary support for additional infections diagnosed and requiring management.

Use of the TDF in this study facilitated us to identify key considerations for developing a CDSS for STI care. Our findings have informed the subsequent co-design and technical work to develop a prototype tool that provides prompts for syphilis testing that we will soon pilot in general practice. Our understanding of the barriers and enablers to implementation and adoption will draw directly on this analysis, whereby each TDF domain can be mapped to the COM-B framework (e.g. ‘Knowledge’ maps onto ‘Capability’; ‘Environmental context and resources’ maps onto ‘Opportunity’; ‘Beliefs about consequences’ maps onto ‘Motivation’), and subsequently to the Behaviour Change Wheel to facilitate understanding of the actions to support behaviour change in regards to the implementation and use of CDSSs for sexual health care.¹⁸

A strength of this study includes interviewing both SHPs and GPs, allowing exploration of expert and end-user views into STI care and the general practice environment. As noted earlier, our analysis was guided by the TDF, providing a theory-based approach to understanding behaviour change components that can inform intervention development. Furthermore, the TDF is complementary to frameworks for assessing intervention implementation and behaviour change. The main limitation is our small sample size and our sample being limited to metropolitan clinicians despite efforts to recruit from rural and regional areas. However, our findings were very consistent across all those interviewed. As we only interviewed GPs as end-users, our results cannot be generalised to other potential end-users of a STI CDSS, such as practice nurses. Despite this, our findings provide valuable baseline information as a first step in STI CDSS design. The co-design process will involve ongoing engagement and input from a wider range of end-users to capture a greater breadth of views. It is also acknowledged that as participants were asked to reflect on a hypothetical tool, responses are seen as views on a concept. It is possible such views may change once participants see or engage with a STI CDSS in practice.

Conclusion

In this qualitative study, we found that the concept of a STI CDSS based on the FHT software was viewed favourably by our sample of GPs and SHPs, indicating a potential role for such a tool in Australian primary care. Although such a tool might support improvements in STI care processes, this can only occur if end-users’ needs are met and barriers are addressed. A STI CDSS should be considerate of the time-pressured environment, the patient context, and include appropriate resources and pop-ups that reflect end-users’ clinical needs and workflow. During 2022, we commenced work on developing a prototype in FHT that provides prompts for syphilis testing, with this study providing an important first step in the tool’s design.

Data availability

Data for this analysis are not publicly available.

Conflicts of interest

Jane Hocking is an Editor of Sexual Health, but was blinded from the peer review process for this paper. All other authors declare no conflicts of interest.

Declaration of funding

This research did not receive any specific funding.