The factors that affect GPs’ deprescribing decisions for older adults and frail patients in primary care in Crete, Greece were investigated, and five themes were developed that were subsequently mapped to the relevant TDF domains: recognition of an undertreated issue, knowledge and skills, beliefs and fears, professional role, interprofessional and social factors, and environmental context and resources. Initially, it was established that deprescribing is underutilized owing to both physicians’ lack of motivation and the absence of reinforcement from any deprescribing initiative or prescribing regulations. Numerous GPs acknowledged their insufficient knowledge surrounding deprescribing and frailty while simultaneously expressing their criticism toward the minimal coverage of these issues in disease-specific guidelines. The importance of the bond between doctors and patients was emphasized, with many participants agreeing that they need more training in communication skills. Besides reporting various beliefs and fears that can arise when medications are discontinued, our study also found a specific fear related to deprescribing psychotropic drugs. One significant issue identified by participating GPs was the lack of recognition of their professional role as the primary healthcare provider. The involvement of other specialties in primary care, the promotion of OTC products by pharmacists, and influences from pharmaceutical representatives were also viewed as barriers. In terms of resources, the lack of easily accessible electronic health records that include patient clinical data and the absence of an efficient interprofessional communication system pose obstacles to deprescribing. Nonpharmacological interventions were considered a facilitator, but limited availability often hinders their integration. Lastly, the majority of GPs stressed that the limited time and excessive workload make it impractical to engage in deprescribing in daily clinical practice.

This research contributes to the existing evidence on factors influencing deprescribing decisions of GPs in primary care. Our findings align with previous international studies [14,15,16, 21, 22]. However, this study adds to literature with its specific focus on factors related to frail patients.

In our study, GPs viewed deprescribing as a beneficial procedure, exerting greater effect in patients with frailty. Nevertheless, they were inclined to maintain the status quo and avoid regimen modifications unless an adverse event occurred. Likewise, in the LESS study [16], although most Swiss GPs indicated their readiness to deprescribe in case vignettes involving frail individuals, they also reported low rates of actually carrying out deprescribing. A systematic review has shown that deprescribing is both safe and effective in frail patients [27]. The attitude of repeat prescriptions and the lack of initiatives on a national level were viewed as barriers by GPs in our study. The dangers associated with repeat prescriptions have been described in existing literature [28] and the absence of polypharmacy initiatives in Greece shifts the focus away from deprescribing [29]. In addition, Greek GPs have criticized the e-prescribing service for its lack of tools to support them in decision-making [30].

Our study revealed a specific gap in knowledge regarding deprescribing tools. Another Greek study investigated the effectiveness of START/STOPP in primary care and found it to be a valuable tool for identifying inappropriate prescribing [31]. Most participants in our study emphasized their desire for the inclusion of recommendations specifically for deprescribing and frail patients in clinical guidelines. The representation of frail adults in clinical drug trials is relatively low, prompting drug regulatory agencies to advocate for an increase in their inclusion [32]. Previous studies have emphasized the importance of a strong bond between patients and doctors [21, 22]. GPs in our study agreed that it is challenging to build this relationship with frail patients owing to cognitive impairments, limited access to primary care, and reliance on caregivers. Research conducted in Poland has identified various obstacles to accessing healthcare, such as challenges with transportation, fragmented care, and a shortage of social care services [33]. Literature also highlights the importance of effective communication and caregiver engagement in decision-making for better care [34]. During our interviews, participants expressed that they lacked the communication skills needed to effectively interact with frail patients or their caregivers.

GPs in our study raised concerns that the fear of adverse outcomes leads patients to become dependent on their medication, and, at the same time, doctors feel that they will be accountable for any potential harm. Similar concerns have been reported in studies from other countries [35, 36]. We also revealed that doctors would be more reluctant to deprescribe psychotropic drugs in contrast with other drug classes. A recent study conducted in the UK displayed that GPs lacked the confidence to manage antipsychotic medications by themselves [37]. In addition, we found that there is a notable fear surrounding benzodiazepines among doctors owing to the withdrawal syndrome. However, an analysis of the national prescription database revealed that older adults in Greece have a high rate of benzodiazepine use, primarily prescribed by internists and GPs [38].

There was no doubt among the participants that general practice is the most suitable specialty for engaging in deprescribing. However, in Greece, the GP has not yet been established as the primary provider. It is indisputable that in complex healthcare systems [39], when GPs do not function as gatekeepers and the patient is free to consult any specialty by themselves, as in Greece, then the complexity and the number of possible interactions increase, and, consequently, the possibility of adverse outcomes expands [40, 41].

In our research, the influence of pharmacists and the promotion of OTC drugs were recognized as factors that contribute to polypharmacy. The participating GPs in this study found pharmacists to be a barrier to deprescribing. This contrasts with previous research that highlighted the positive influence of pharmacists in facilitating deprescribing [14, 21]. In the UK, pharmacy services offer more advanced services, such as medication use reviews, which prioritize medication optimization. Greek pharmacists have a more restricted role, which mainly involves dispensing medications and providing basic advice [42]. The limited participation of Greek pharmacists in clinical services, as compared with pharmacists in other countries, may explain why they are less involved in medication optimization and are seen as a hindrance to deprescribing. Older individuals who are frail are at a higher risk of overusing OTC products [43]. Furthermore, the availability of OTC medications through online pharmacies has led to a rise in self-medication, frequently influenced by advice from acquaintances and online sources [44]. Previous studies in Greece have underlined that influence from family and friends [45, 46], as well as from pharmacists [47], are key factors that may give rise to misuse of nonprescribed products. Our study also highlighted the potential influence of pharmaceutical representatives. Previous research has shown that interactions between physicians and pharmaceutical sales representatives can affect prescribing behavior and result in irrational prescribing [48]. The demand for a uniform electronic health record that would facilitate interprofessional collaboration between different parts of the healthcare systems was consistently mentioned in our discussions. Several Greek articles have brought attention to this issue [49, 50].

In relation to nonpharmacological treatments, it was pointed out that, while doctors and patients may be in favor in certain situations, they often resort to pharmacotherapy owing to the restricted availability. As of today, in Greece, physiotherapists, speech therapists, and occupational therapists are mainly private practitioners, and depending on the diagnosis, the National Organization for Health Care Services compensates patients with a fixed fee. Nonetheless, because of the scant compensation set by the state, providers often ask patients for additional charges [51].

Most participants in our study emphasized that the excessive workload, along with the limited time frame that a GP has to cope with, often results in omitting medication reviews. The data we gathered support this claim; participants reported an average of 32 daily visits, with 15 min allocated for each individual. Current literature suggests that older patients are also displeased with the consultation time [52].

There is a clear need for the implementation of unified national deprescribing policies that would provide clear guidance. In Canada, the Canadian Deprescribing Network has played a crucial role as a leader in deprescribing practices, leading to numerous positive outcomes. The use of patient education strategies, such as educational materials and online media, can help raise public awareness [53]. Educating patients about the advantages and safety of deprescribing could help alleviate their concerns and resistance towards the process. Considering the study’s findings on GPs’ reluctance to discontinue psychotropic drugs, as well as the widespread use of benzodiazepines in Greece [38], it would be advantageous to introduce educational programs against benzodiazepine misuse aimed at GPs. A previous study in Spain demonstrated the efficacy of such interventions in decreasing chronic users [54].

The e-prescription service could benefit from improvements such as restricting repeat prescriptions, incorporating computerized prescribing alerts, and integrating deprescribing tools. However, it is important to first implement initiatives that provide thorough training on deprescribing and the utilization of tools such as the Beers Criteria and STOPP/START. To fill the gaps in knowledge related to frail patients, it is important to promote the inclusion of these individuals in clinical trials and guideline recommendations. Experts in geriatrics and frailty should be involved in the design of these studies, as suggested by Denkinger et al. [32].

To enhance the bond between GPs and their patients, it is necessary to improve accessibility to primary care. This is important for frail patients who may have trouble visiting on their own. Establishing social care services that closely collaborate with primary care and providing transportation services for older adults could be beneficial. Regarding communication skills, integrating formal training in such skills during residency and undergraduate programs can aid in fostering strong connections. In the past, communication skills programs have been implemented during residency and have showed significant effectiveness in enhancing confidence in communication abilities [55].

It is vital to establish the role of GPs as gatekeepers while simultaneously aiming for continuity of primary care [56]. GPs should be acknowledged as the cornerstone of healthcare, with the responsibility of coordinating care and acting as the primary point of contact. However, to facilitate effective communication between GPs and other physicians, there is still a need for a reliable and interoperable electronic health record system [49].

To manage overuse of OTC medications, it is necessary to introduce interventions that educate patients about the potential risks associated with their misuse and to minimize the spread of misinformation through strict monitoring of advertisements [57]. A study across five European countries found that educational intervention can reduce irrational prescribing of OTC medications by GPs [58]. To ensure proper regulation, it is necessary to register all dispensed medications and adopt an interprofessional collaborative approach between physicians and pharmacists. A proposed solution to eliminate any potential bias caused by pharmaceutical representatives is to advocate for and potentially mandate the use of generic names when prescribing medications.

To promote nonpharmacological treatments, it is essential to recruit healthcare professionals, including physiotherapists, occupational therapists, and psychologists in primary care. Comprehensive coverage of service costs could eliminate financial barriers. It is essential to increase the number of doctors as well to enhance primary care. This would lead to a decrease in the number of patients assigned to GPs, allowing them to provide more thorough care. Past literature has also portrayed the longing of GPs for enhanced recruitment [59].

The study benefits from a strong data collection method, as it allows for in-depth information gathering through one-on-one interviews. A stratified sampling method was used during participant recruitment, allowing for a more diverse exploration of the views of GPs, with participants selected from both rural and urban regions. Another advantage is that the study examined factors related specifically to frail patients, addressing a gap in current research [16]. Moreover, the TDF provides a theoretical foundation for implementation research [20]; hence, by using it during data analysis, we present results that could be used to guide future interventions.

The study specifically targeted GPs who exclusively practice in primary care in Heraklion, Crete. While interviews were carried out until data saturation was reached, it is important to note that the transferability of these results to different settings or populations might be restricted. In addition, the study did not consider the perspectives of healthcare professionals practicing in remote regions without access to specialized healthcare. Another potential limitation of the study is that GPs may face challenges in distinguishing frail older adults from the broader population of older adults. Therefore, some responses provided might pertain to older adults in general, rather than specifically to frail individuals. A possible limitation is that the interviewer was a GP resident, which might have influenced the data collection process by introducing potential biases into the respondents’ answers. However, it is important to note that the interviewer had no prior relationship with the participants. The study did not explore the beliefs of other stakeholders, including nurses, pharmacists, hospital physicians, and other healthcare professionals. Prescribing decisions often require interprofessional collaboration. Therefore, excluding these perspectives could limit the understanding of the various factors that influence deprescribing behavior.