Changing HCW attitudes: a case study of normalizing HIV service delivery in emergency departments

In the field of HIV, several well designed and funded interventions have failed to scale up and make the impact they planned due to the lack of buy-in from the health system and healthcare workforce [1]. While numerous evidence-based public health interventions and clinical treatment strategies exist, it can be challenging to implement and sustain them outside of controlled settings. Changing clinical practices and norms is challenging as it requires disturbing the status quo. Much clinical practice occurs in stable healthcare contexts and can be assumed to be habitual, wherein healthcare workers (HCW) have been described as “creatures of habit” and “resistant to change,” thus rendering existing clinical behaviors unlikely to be spontaneously reconsidered [2, 3]. Beliefs and perceptions held by HCWs have the potential to undermine the implementation and successful integration of interventions. A fear of the unknown and perceptions around how the status quo will be impacted by the changed behavior likely drives attitudes that hinder its acceptance. Additionally, in many cases the implementation of evidence-based interventions (EBIs) into practice aren’t supported with sufficient training or via supported process integration practices. Establishing a new normal also involves unlearning current practices and beliefs, and developing an understanding of the dynamic between new evidence-based practices and the existing context [4, 5].

South Africa faces the highest burden of HIV infection globally, with 7.5 million people (~ 20% of the global estimate) living with HIV and prevalence ranging from 12.6% to 27% across the country [6, 7]. As such, numerous measures have been taken by the national government to deliver EBIs focusing on HIV prevention and treatment, including routinely offered provider-initiated HIV counselling and testing (HCT) to all persons attending healthcare facilities as a standard component of medical care, including trauma, casualty, and specialty clinics, as per the South African National Strategic Plan on HIV, sexually transmitted infections and tuberculosis [8, 9]. However, much of the program’s focus has been directed to primary healthcare centers and antenatal facilities or to high-risk populations such as sex workers, men who have sex with men, injection drug users, and prisoners [10‐12]. As a result, high-volume venues, such as EDs, are neglected, which remains the sole point of contact with the healthcare system for 28% of the country’s population [13].

The integration of HIV testing into health facilities has been pivotal to the goal of universal provision of testing. In the United States of America, routinization of HIV screening as a standard of care in EDs has been critical in shaping the national strategy for addressing HIV and has now been widely adopted across the country [14‐16]. Previous studies in the Eastern Cape, South Africa, have also demonstrated a high burden of HIV in the ED with a prevalence of 25% and a new diagnosis rate of 7% [10, 17]. However, EDs are complex clinical environments, providers do not have formal HIV training, and routine HIV service delivery is not supported in this setting. Qualitative studies, including work by this team, has identified several potential barriers to HIV testing in the ED; concerns around confidentiality, increased burden of work for ED staff; high stigma among healthcare workers towards people with HIV; and lack of resources to provide social support or facilitate linkage to care [18]. We argue, the ED is a key clinical venue where patients missed by current HIV testing and treatment programs may be accessed, however, the listed barriers resulting in a lack of HCW buy-in impedes successful implementation and scale-up of ED-based testing.

Routinization of ED-based HCT testing will require overcoming cited barriers and a more reflective ‘normalization’ of practices [19]. There are several theoretical frameworks that seek to identify determinants of intervention adoption. This paper draws on the Normalization Process Theory (NPT), a theoretical model, which emphasizes that people change when their environment, or interactions with processes, change. NPT identifies, characterizes, and explains key factors that promote and inhibit the implementation and integration of complex healthcare interventions into everyday practice [20]. The theory has four core constructs; coherence (or sense-making); cognitive participation (or engagement); collective action (work done to enable the intervention to happen); and reflexive monitoring (formal and informal appraisal of the benefits and costs of the intervention) [21]. Drawing on these constructs NPT allows us to explore the processes through which complex interventions can be integrated and embedded as routine elements of clinical and organizational work in health care, making it applicable to assess if and why HCW attitudes shift after exposure to clinical research.

In this study we sought to understand the role of implementation research on provider attitudes to HCT and hypothesize that the process of bystander participation in research may normalize the practice and thus improve adoption of the intervention [21]. We proposed that controlled exposure to a new practice may also somewhat mitigate the fear of the unknown and perhaps decrease HCW stigma towards HCT service delivery. Here we use a validated survey instrument to quantify the attitudes of providers, pre-and-post exposure to the HIV research intervention, towards their acceptance and adoption of HIV service delivery within the ED and draw on NPT constructs to examine current thinking and approaches to inform large-scale behavior change.

A total of 132 surveys (66 surveys each at baseline and after 6 weeks) were completed across the three sites [Table 1]. A majority of respondents were female (n = 93, 70.5%), between the ages of 20 and 29 years (n = 50, 37.9%), of African race (n = 107, 81.1%), and had been practicing medicine for 0 to 4 years (n = 50, 37.9%). Of the respondents, 52 (39.4%) were registered nurses, 45 (34.1%) were physicians, 19 (14.4%) were nursing assistants/practitioners, and 16 (12.1%) were case managers. Given the similarities in the overall sample population and the mean baseline attitude score by site, future analyses were pooled.

In our study, we saw slight improvements in both the providers’ assumed added burden of offering HCT, and their comfort and confidence with implementing ED-based HCT, following exposure to a research study that sought to implement HCT as part of the ED’s routine clinical care pathway.

Translation of national HIV testing guidelines to practice in the ED has been challenging worldwide [24‐27]. Significant resistance to embedding HCT as routine practice in EDs has been raised by some in the emergency medicine community who are reluctant to re-negotiate existing practices in an already complex clinical setting [28, 29]. Despite the dissent among providers, there is a reassuringly high acceptance of ED-based HIV testing among patients [18, 30, 31]. Our previous work in this study population identified perceived operational barriers that an ED-based testing program would bring to the current work environment as a significant concern among providers [18]. Among the concerns raised, limited resources, discomfort with disclosing results, concerns with maintaining the confidentiality of results, patients’ reactions to a positive HIV test, ED financial struggles, and ED crowding, predominate [32, 33]. Many of these negative perceptions are anchored in providers’ existing knowledge of HIV testing provision across other health care facilities and their previous experiences in implementing interventions in the ED.

It is increasingly recognized that successful implementation of any complex intervention in a clinical setting begins with the willingness and commitment of providers to adopt such a program, followed by a strong intention to modify habitual practices [2]. And while provider support has been previously demonstrated as a key facilitator for the implementation of public health interventions in the ED (e.g., domestic violence and alcohol screening) [34‐36], inherent in many social cognitive theories is also the assumption that providers have their beliefs, preferences, and intentions, but there are always additional social factors that promote or constrain particular expressions of agency [37]. Our study confirmed providers’ support and buy in as key stakeholders in the ED was crucial to the successful routinization of HCT, and additionally the NPT framework identified which actions facilitate or prevent the adoption of the newly proposed practices, and which key actors contributed and how, to its implementation, embedding, and integration. Furthermore, by exposing providers to the HCT program we were able to support their ability to make better sense of the true operational barriers to HCT implementation as well as enable engagement without making it mandatory or dictatory.

As difficult as it is for individuals and organizations to adopt new practices, equally important but considerably slower and more challenging may be the process of ‘de-innovation’, i.e., eliminating entrenched practices that previously made sense but are now less valuable, due to new evidence or competing approaches [38]. In our study, our aim was not only the implementation of a new practice, but also the de-implementation of deeply embedded and integrated patterns of interactions, expectations, and practices. The processes required to remove such established practices are not simply the reverse of those needed to initiate and adopt new ones. In more recent literature, the terminology of ‘de-innovation’ has become increasingly synonymous with unlearning [39]. There are numerous established strategies that can promote unlearning, such as education and persuasion, strategies to mitigate cognitive rigidity, systems engineering approaches to support alternative decision making, and use of social media [40, 41]. However, results of our study support the notion that controlled exposure to a new practice with an iterative design and continuous stakeholder buy-in can facilitate the process of unlearning and improve the adoption of a new intervention.

Longitudinal studies have already demonstrated the impact of policy change and normalization on longitudinal change in perceived barriers and benefits to testing in both patients and providers [42‐44]. We have demonstrated that even short, discrete investments can have a significant role in not only changing the attitudes of providers, but also in improving the feasibility of adopting and scaling up new interventions.

In South Africa, a country with a complicated history around HIV/AIDS and significant investment in HIV service delivery, we suggest that research interventions may have a dual benefit. Not only is research necessary to understand the epidemiology of HIV and implementation of HIV service delivery, but also exposure to interventions within a research context may normalize HCW perceptions to HIV interventions. Research activities may thus have a significant role in supporting the acceptance, adoption, and scale-up of innovative HIV interventions that are required to curb the HIV epidemic.

This study has a limited sample size, capturing a small number of providers at the beginning and end of HIV testing intervention in the ED, and unfortunately, the participants were not paired due to the fluid nature of shift workers and their unpredictable schedules, thus it could mean that our results are representative of two different sample populations, opposed to a true change in attitudes of an individual provider. In addition, while the survey instrument utilized is validated, and normalization process theory was incorporated as part of the analytical framework, the survey itself could have discerned the impact of research activities across those constructs if the framework was part of the survey design, and provides an opportunity for future refinement of the survey tool.

The results of this study support the theory that controlled exposure to a new practice, one that is learned and implemented iteratively, can facilitate the process of unlearning, where required, and improve the adoption of a new intervention. Understanding stakeholder attitudes and perceptions as well as structural factors of a given context through the lens of theory and practice can explain the failure of an apparently widely adopted and encouraged practice to become routinely incorporated, even in circumstances where professionals are favorably disposed to them and where significant material and political support was committed to them. Similar to other preventive programs, HCT cannot be easily layered onto current ED practices. The NPT model helps think about not just how its constructs can be introduced into planning and designing implementation processes, but also in asserting control over these practices which are influenced by other aspects of study design and context with uncontrollable outcomes. Implementation of small research projects not only benefit scientific inquiry but may go some way to changing perceptions, de-stigmatizing diseases and interventions, and prepare the system for change.