Perceptions and attitudes regarding delayed antibiotic prescription for respiratory tract infections: a qualitative study

Abstract

Background

Antibiotics are overprescribed for respiratory tract infections (RTIs). However, the decision to prescribe is often complex. Delayed antibiotic prescription (DAP), a strategy designed to promote more rational antibiotic use, is still not widely used. The aim of this study was to explore perceptions and attitudes in primary care professionals, regarding antibiotic use and different DAP strategies for uncomplicated RTIs.

Methods

We conducted a qualitative study, using an inductive thematic approach to generate themes, based on focus group discussions and semi-structured interviews with professionals, recruited from 6 primary care centres (Barcelona metropolitan area, Spain).

Results

26 professionals (25 family physicians and one nurse) were included in four focus group discussions and three semi-structured interviews. Participants commented that RTIs were a main reason for consultation, motivated often by patient anxiety and fear of possible complications, and this was associated with the patients’ poor health-related education. Acknowledging inappropriate antibiotic use in the health system, participants attributed this, mainly to defensive medicine strategies. DAP was used when in doubt about the aetiology, and considering factors related to patient-physician interactions. The main perceived advantage of DAP was that it could reduce the need for additional visits, while the main disadvantage was uncertainty regarding proper use by the patient.

Conclusions

DAP was used by participants in cases of doubt, in specific situations, and for specific patient profiles. Weak points were detected in our primary care system and its users that affect the proper use of both antibiotics and DAP, namely, time pressure on professionals, poor patient health-related education, and the lack of a patient-physician relationship in some scenarios.

Introduction

Respiratory tract infections (RTIs), the most frequent infections encountered in primary care [1], are mostly self-limiting and are caused by viruses. While antibiotics may slightly modify course [2, 3], they tend to be overprescribed [4, 5]. Overuse of antibiotics is closely related with antimicrobial resistance [6,7,8], by now a major global public health challenge [9] that entails an increased risk of adverse effects for patients [10] and increased beliefs of the need to consult for similar episodes [11, 12]. In a context of optimal use of antibiotics, the decision to prescribe is complex, as, in some cases, symptoms are unclear; furthermore, the decision also depends on factors related to patient-physician interactions [4, 13], such as pressures from the patient [14, 15] and the patient-physician relationship [15].

One approach to reducing inappropriate antibiotic use for RTIs is delayed antibiotic prescription (DAP) [16], a strategy designed to promote a more rational antibiotic use in situations of uncertainty, regarding the need of immediate antibiotic prescription (IAP). DAP consists of the patient only using the antibiotic prescription if the RTI has not improved or has worsened, some days after consultation. A recent systematic review [17] comparing DAP, IAP, and no antibiotic prescription (NAP), reported that RTI symptom severity was similar for the 3 strategies, that symptom duration was slightly shorter for IAP versus DAP, and that re-visit and complication rates were lower and patient satisfaction was higher for DAP versus NAP.

DAP is still not widely deployed by professionals, as reported by several qualitative studies that have investigated views and experiences of DAP for RTIs among professionals in northern Europe [13, 18, 19], United States [18], Australia [20,21,22], and New Zealand [23]. While a study conducted by our group suggests that under 50% of primary care professionals in Spain use the DAP strategy [24], to our knowledge, no qualitative research evidence is available regarding this issue in Spain. Our objective was, therefore, to explore perceptions and attitudes of professionals regarding use of antibiotics and of different DAP strategies for noncomplicated RTIs.

Methods

Study design

Qualitative study using focus group discussions (FGDs) and semi-structured interviews [25,26,27], performed in primary care centres in the Barcelona metropolitan area (Spain).

Participants and recruitment

Family physicians were recruited from six primary care centres, five of which had previously participated in the DAP-Trial [11]. This trial, conducted with adults with uncomplicated RTIs in a primary care setting, assessed the efficacy and safety of IAP versus NAP, and versus two DAP strategies: a delayed patient-led strategy (the patients receives the prescription, with instructions to only use it if the RTI worsens or fails to improve), and a delayed collection strategy (the patients collects their prescription from the primary care centre 3 days after the first visit if they consider they need it).

Sampling was purposive, with participants selected according to a strategy in which sample design was based on a theoretical construct [25, 27]. The criteria to define professional profiles that reflected possibly different discourses, were as follows: (a) the professional’s previous participation in the DAP-Trial (yes/no); and (b) the socioeconomic level of the professional’s primary care centre’s catchment population (medium-low/medium-high). Socioeconomic level was taken as a proxy for the education level of patients [28], as previous studies have shown that professionals do not consider DAP to be appropriate for less educated patients [21, 23]. For this reason, professionals were selected according to the deprivation index of the primary care centre’s catchment area [29]. The sample was demographically as heterogeneous as possible in terms of gender (women/men) and age (junior: <45 years/ senior: ≥45 years).

Candidate participants for this study were recruited by the DAP-Trial centre coordinators. A sample size of 24–36 participants was estimated as necessary (4 FGDs based on 6–9 participants); however, the final number of included participants was determined once data saturation was reached.

Data collection

The study was conducted in two phases: first, focus group discussions, aimed at fostering interaction between participants [25,26,27]; and individual semi-structured interviews afterwards [25, 27]. In phase I, the participants completed a questionnaire about sociodemographic data and use of DAP strategy in their clinical practice.

Focus group discussions

FGDs were profiled according to sampling criteria as follows: FGD1, DAP-Trial participants and medium-low socioeconomic area; FGD2, DAP-Trial participants and medium-high socioeconomic area; FGD3, DAP-Trial non-participants and medium-low socioeconomic area; FGD4, DAP-Trial non-participants and medium-high socioeconomic area. In relation to FGD2, not enough family physicians were recruited. Thus, one nurse participating in the DAP-Trial was included. Even though her role with the antibiotics was different, we considered that her opinion could also be relevant because in many centres, the nurses carry out triage consultations. Similarly, their educational work and their experience in the trial was deemed relevant. A script was prepared for this study (Appendix 1) that was sufficiently flexible for participants to suggest new topics. FGDs, run with a moderator and an observer, were conducted in a meeting room in the coordinating centre (Hospital de la Santa Creu i Sant Pau (HSCSP) in Barcelona, Spain). All FGDs were digitally audiorecorded, and recordings were transcribed verbatim. Notes taken by the moderator and observer were also used, to check and complement the transcriptions’ data.

Individual semi-structured interviews

Semi-structured interviews were guided by a specifically designed script for this study (Appendix 2). They were carried out in order to further explore key issues that emerged in the FGDs. Interviews were conducted in the primary care centres where the participants worked with professionals drawn from the FGDs. The same researcher who moderated all FGDs also conducted the semi-structured interviews.

Data analysis

The transcriptions were cross-checked against the digital recordings and inductive thematic analysis was performed as described by Braun and Clarke [30]. The analysis was conducted by 3 researchers. Two of them independently analysed the transcription of FGD1 and agreed a preliminary coding frame. The analysis of the other transcriptions was conducted by one researcher and a second researcher reviewed the coding. The discrepancies about emergent themes and codes were resolved by consensus between the researchers. We used ATLAS.ti (version 8) software for data coding and analysis. Quotations from the FGDs and interviews were translated from Catalan or Spanish to English. Investigator triangulation and search for negative cases were undertaken to improve rigour of the analysis [31].

Results

We conducted 4 FGDs and 3 individual interviews, with a total of 26 participants, 25 physicians and 1 nurse, with a mean (SD) age of 46.81 (8.56) years, 13 (50%) worked in a primary care centre in a medium-low socioeconomic area and 12 (46.15%) previously participated in the DAP-Trial (Table 1).

Table 1 Sociodemographic characteristics of study participants (N = 26)

The FGDs were conducted between September 2013 and June 2014. Mean duration was 90 min, except for FGD2, which lasted 60 min. Note that 2 physicians in FGD1 and 1 physician in FGD2 belonged in primary care centres with a different socioeconomic level from the rest of participants in their groups. The semi-structured interviews were conducted between October and December 2018 and lasted approximately 60 min.

We identified 4 main themes arising in the 4 FGDs and the 3 interviews: (1) Characteristics of RTI visits; (2) Expectations and adequacy of antibiotic treatment; (3) DAP, how and for whom; and (4) DAP-Trial and primary care research barriers. Example quotes are shown in Tables 2 and 3.

Table 2 Illustrative quotes:Characteristics of RTI’s visits and Expectations and adequacy of antibiotic treatment

Table 3 Illustrative quotes: DAP, how and for whom; and DAP-Trial and primary care research barriers

Characteristics of RTI visits

The concept of RTI

Most uncomplicated RTIs were considered banal and self-limiting. Physicians commented that, with some patients, once informed that the infection was caused by a virus, they perceived this as the physician’s incapacity to determine the diagnosis or as not having any disease. RTIs were one of the main reasons for scheduled and unscheduled visits in winter, and patients tended to consult at very early RTI stages seeking a rapid cure. Some patients reconsulted every year and several times for each episode, and this despite previous experiences and having received appropriate information.

Despite RTIs being considered mostly banal, visits required a time investment in examining, informing, and educating patients, and in establishing a relationship of trust (if not previously established). However, this time investment was often not possible due to work loads and the structure of the healthcare system. RTI consultations were mainly motivated by patient self-perceptions of poor health, anxiety, and a fear of possible complications. These feelings varied depended on their own or acquaintances’ previous experiences and were often attributed to hearsay. According to some participants from centres with a medium-low socioeconomic level, poor health-related education was linked to a low socioeconomic status, while other participants considered that health knowledge among the general population had decreased from previous generations. It was considered that more education was needed, via primary care centres, the media, and schools.

Consultation often reflected the patient’s age, with young patients consulting because they were not used to being sick, and elderly patients consulting because they were concerned about their comorbidities. Another reason for consultation were requests for sick leave from work. Some participants were of the opinion that healthcare human resources were misused when patients consulted for mild cases of RTIs, with this misuse attributed to a lack of responsibility for self-care by patients. It was suggested that there was a need for patient empowerment, and also that access to rapid tests would be useful visit aid.

Expectations and adequacy of antibiotic treatment

Physician-indicated treatment

Some patients expected a drug prescription to feel reassured or considered antibiotics to be an effective and fast-acting cure. Some patients felt that they were not being treated properly when recommended symptomatic medication (e.g., analgesics, antihistamines, mucolytics, and antitussives) while other patients even explicitly expressed dissatisfaction that antibiotics were not prescribed. There was a general opinion that the patient’s satisfaction was often greater when they were prescribed an antibiotic. Patients who accepted the non-prescription of antibiotics were those who had previously recovered without antibiotics, had experienced some adverse effects of antibiotics, or who were better informed about antibiotics, such as pregnant women.

Inappropriate antibiotic use

Inappropriate use of antibiotics in the healthcare system was acknowledged. This inappropriate use in primary care was attributed mainly to defensive medicine based on low-cost drugs, most especially when there were time pressures or when there was no patient-physician relationship (e.g., unscheduled visits). Another reason was the presentation of some antibiotics does not fit with prescription patterns, meaning that patients typically have medication left over. Some participants suggested that this may be due to potential financial interests from pharmaceutical industry.

Despite recognizing the inappropriate use of antibiotics, there was a generalized opinion that primary care professionals make every effort to use them rationally, and even a trend in both hospital and primary care settings towards prescribing fewer antibiotics. This was attributed to better training and incentives for professionals, although such strategies were still considered to be insufficient. The trend to reduce antibiotic use was considered not to occur in the private health sector. In the opinion of the participants, inappropriate use of antibiotics was due to a range of opinions regarding both indication and choice of antibiotics. Private health sector physicians tended to prescribe more antibiotics and they are more often non-generic and expensive than public health sector physicians. Finally, it was acknowledged that doubts existed regarding the use of antibiotics because the criteria were always not clear.

Patient self-medication

When patients visited, they had often already started symptomatic medication, and a typical recommendation was to follow the same treatment for a few more days. Occasionally, patients had already taken an antibiotic, typically left over from a previous prescription.

DAP, how and for whom

Context

DAP was used in cases of doubts regarding aetiology, and was mainly used for pharyngitis in adults, and for acute middle-ear infection in paediatric patients. Taking into account information obtained in the patient-physician interaction, DAP was typically deployed in the following circumstances: before a weekend, travel, or an event; in unscheduled visits without follow-up; when enough time was available to appropriately inform the patient; and when the patient refused to leave without an antibiotic prescription even if not clinically indicated.

Patient profiles

DAP may be indicated for specific patients, considering, most importantly, the patient’s capacity to understand the strategy. Candidates were also patients who were considered trustworthy, those with greater common sense (they probably would not use the antibiotics immediately), those with a relationship of trust with their physician, and those with chronic conditions who were knowledgeable about their pathology.

There was no consensus as to whether it was more difficult to implement DAP in young people who probably did not have a physician-patient relationship, or in older people with comorbidities or cognitive difficulties. It was agreed that DAP would not be indicated for patients experiencing anxiety, frequent healthcare users, or patients who insist on an antibiotic prescription.

DAP advantages and disadvantages

Avoiding the need for a further visit was considered the main advantage of DAP, although the extent of the advantage was perceived to vary. A second advantage was that the DAP strategy generally satisfied both patient and physician. DAP also meant that patients had a safety net, in that they had the prescription if the condition deteriorated or failed to improve. DAP also represented an opportunity to educate patients that antibiotics are not always needed for RTIs and empowered them with greater decision-making autonomy. Finally, DAP as an alternative was useful when pursuing more rational use of antibiotics.

The main concern was uncertainty regarding patients’ proper use of the DAP strategy, mainly that they might use the antibiotic immediately. Some participants proposed that the prescription should not be available until a date recommended by physician. Related to this uncertainty, some physicians who did not use DAP stated that they preferred to take responsibility for the final clinical decision, despite the possibility of an additional visit. Two other physicians who did not use DAP considered that the patient had to be properly informed prior to being offered DAP and one physician considered that, with DAP, there was a possibility of antibiotics being prescribed despite not being indicated.

Some physicians who used DAP confirmed that it required a greater investment in time and effort, mainly in assessing whether the patient was a suitable candidate and then issuing instructions for use of the prescription. Possible professional responsibility in the event of a complication was expressed as a concern regarding the DAP strategy by one physician who used it.

Patient-led DAP versus DAP collection

While DAP collection rather than patient-led DAP was considered by some to be a better strategy because immediate use was avoided, a recognized advantage of patient-led DAP was that it avoided a return visit by the patient.

Improvement proposals

It was proposed that DAP use should be rewarded with incentives. It was also pointed out that deployment of DAP required more time and would need the health system’s educational role to be enhanced. Another proposal was to involve nurses and pharmacies in deployment of the DAP strategy.

DAP-Trial and primary care research barriers

DAP-Trial

While some advantages to carrying out the DAP-Trial were commented, the main focus was on barriers. The main barrier perceived by both DAP-Trial participants and non-participants was the lack of time for the work implied by research. Perceived barriers by the DAP-Trial non-participants were the lack of suitable candidate patients and the disruption implied by DAP inclusion in routine practice. Perceived barriers by the DAP-Trial participants were the lack of support and a lack of agreement with recommendations to patients allocated to the DAP strategies. Some DAP-Trial participants found the study useful in making them more aware of and familiar with DAP, and interesting in that the study was implemented independently of the pharmaceutical industry. Also expressed was a feeling of belongingness, resulting from the follow-up emails periodically sent by the coordinating centre.

Primary care research

It was recognized that research in the primary care compared to the hospital setting was scant, with a lack of time and poor rewards stated as the main barriers. Proposed in addition to involving nurses and residents in research, were incentives such as reducing work burdens and healthcare pressures, and the provision of financial rewards and additional holidays.

Discussion

Main findings

We identified a vicious circle between poor health-related education in patients with RTIs and time pressures in primary care centres. Time-consuming RTI consultations of poorly educated patients feeling anxious and fearful of possible complications, led to healthcare pressures that constrained physicians in terms of educating patients.

Physicians generally acknowledged inappropriate use of antibiotics in the health system, but also considered that they made every effort to prescribe them rationally, attributing inappropriate use to defensive medicine with low-cost drugs, based on a perceived trade-off between short-term negative consequences of non-prescription (i.e., complications) and long-term negative consequences of prescription (antimicrobial resistance).

DAP was therefore deployed in cases of doubt, in specific situations, and to specific patient profiles. The main advantage of DAP was considered to be the reduction in additional visits, while the main disadvantage was perceived to be uncertainty as to proper patient use. Regarding the DAP-Trial and primary care research, a lack of time was considered to be the main barrier to research in primary care settings. We did not find major differences between DAP-Trial participants and non-participants possibly because most of them used DAP in their practice.

Results in context

The decision to prescribe antibiotics for some RTIs depends not only on medical factors but also on patient-physician interaction factors [4, 13]. The results of our study corroborate previous studies in that the DAP strategy was considered useful for this kind of complex decision-making scenario [13, 18,19,20,21,22,23, 32].

Our study participants deployed DAP in cases of uncertainty and, as in previous studies, in specific situations, e.g., before the weekend or holidays [13, 19,20,21, 23, 32], as a negotiation strategy when patients insisted on antibiotics [13, 18,19,20,21,22, 32], and for certain patient profiles [13, 19,20,21,22,23, 32]. The main characteristics of DAP candidates that emerged in our study, consistent with previous studies, were patients capable of understanding the strategy [21, 23, 32], patients considered trustworthy [20, 32] and having common sense [32]. The patient-physician relationship was another key aspect to consider in deploying DAP, according to the results of our study. An issue that did not emerge in our study, unlike other studies, was that DAP was considered to strengthen this relationship [19,20,21, 23].

DAP was used by our study participants in apparently contradictory situations: (a) for patients who demanded antibiotics and refused to leave without a prescription, and for patients who were trusted not to immediately use the prescription (as in Hoye et al. [32] and Sargent et al. [21]); and (b) for patients consulting in unscheduled visits, in which the patient-physician relationship considered fundamental to this strategy was lacking. These apparently contradictory deployments of DAP, highlight the complexity of physician decision-making regarding antibiotic prescription.

DAP strategy advantages and disadvantages, in our study as in previous studies, are associated with the fact that DAP is a more patient-centred approach [18, 23]. Thus, while DAP provides the patient with a safety net [13, 18,19,20,21,22] since the prescription can be used if needed [32], and also empowers the patient by making them responsible for the final decision [18,19,20,21, 32], control is lost by the physician [13, 19, 20, 23].

Our study identified some important health system barriers to appropriate antibiotic use and DAP deployment, primarily the lack of a patient-physician relationship in unscheduled visits, poor patient health-related education, and the lack of professional time. These latter issues could be simultaneously addressed by nurses and pharmacists becoming more involved in educating patients regarding RTIs and their treatment. DAP was perceived, as in previous studies, as a golden opportunity for educating people about antibiotics [20, 21, 23, 32].

Limitations and strengths

The main limitation of our study is that the participants mostly came from primary care centres participating in the DAP-Trial, and most used DAP in their clinical practice. While the advantages of DAP may therefore be considered to be overestimated, our results are nonetheless consistent with the extant literature. A second limitation is that our study did not include participants from rural settings, although Fletcher et al. [22] found no differences between rural-urban contexts in their study. A third limitation is that, due to the few professionals participating in the DAP-Trial, two FGDs were not homogeneous in terms of the socioeconomic level of the centre’s population. Furthermore, a nurse who participated in the DAP-Trial was included in a FGD. Including this participant granted the feasibility of one of the groups. The researchers involved in conducting and analysing the FGD assessed that the dynamics were not negatively affected, and, indeed, the nurse’s contributions were particularly enriching.

A major strength of our study is that it included professionals who had deployed DAP and so were well aware of the positive and negative aspects of DAP. A second strength is that, as far as we are aware, this is the first qualitative study of professionals and DAP conducted in a country in southern Europe, where antibiotic use is comparatively higher than in northern Europe [6]. Finally, our study complements several other studies published by our group [11, 17, 24, 33] aimed at raising awareness and improving implementation of the DAP strategy.

Implications for practice and research

Our findings highlight the fact that time pressures, poor health-related education of patients, and the lack of a patient-physician relationship in unscheduled visits were important barriers to optimal antibiotic use and to deployment of the DAP strategy in primary care. Policy-makers may therefore consider strategies, such as the following to overcome these challenges: (i) the provision of RTI health-related education and self-care, and the encouragement of proper use of healthcare services supported by primary care nurses and pharmacists; (ii) improved access to rapid streptococcal testing; and (iii) reorganization of physician agendas so that RTI consultations are attended by the referring physician whenever possible.

Another implication of our findings is that they point to a lack of consensus about some of the criteria to be considered by physicians in deploying the DAP strategy. This suggests that clinical guidelines on RTI management in primary care need to better specify criteria for deployment of DAP, including patient and contextual factors which should be considered when using DAP strategies, as well as the standardization of prescription use recommendations for patients. Finally, the poor health-related education of patients was one of the main themes that emerged in this study. A recent systematic review showed that educational interventions were one of the most efficacious and safe strategies for optimal antibiotic prescribing for RTIs [16]. Given the need for further studies to evaluate RTI educational interventions for patients, our group is conducting a multicentre factorial trial of two educational interventions, targeting both parents and professionals.

Conclusions

DAP was used by participants in cases of doubt, in specific situations, and for specific patient profiles. Weak points were detected in our primary care system and in its users that affect the proper use of both antibiotics and DAP, namely, time pressures on professionals, poor patient health-related education, and the lack of a patient-physician relationship in certain scenarios. Proposed to overcome these challenges are educational interventions regarding RTIs and optimal use of healthcare resources and the formulation of better DAP-related recommendations in guidelines.