Introduction

Telehealth, as defined by the World Health Organization (WHO), encompasses activities, systems, and health services conducted remotely by professionals using Information and Communication Technologies (ICT). Its objectives include providing diagnosis, promoting global health, facilitating the treatment and prevention of diseases and injuries, and advancing health research and education [1].

In 2016, the WHO highlighted the risk of new errors arising from the adoption of telehealth practices, emphasizing the need for telehealth projects to incorporate robust forecasting and planning strategies to effectively address potential harms issues. Addressing patient safety risks in telehealth involves navigating complex and dynamic processes, as these risks are systemic, affecting not only healthcare professionals but also the environment and organizational context [2].

Understanding how errors propagate through the care chain is essential for addressing the specific risks posed by information and communication technologies (ICT) in healthcare. As telehealth continues to expand rapidly, ensuring patient safety becomes increasingly critical. However, despite its central relevance, patient safety remains an underexplored aspect in telehealth, with notable gaps in regulatory frameworks and a lack of comprehensive guidelines in the literature regarding its implementation and monitoring [3]. A systematic review published in 2014 identified the risks associated with patient safety in telehealth within a home care context. However, the present review differs by investigating the risks related to patient safety in telehealth delivered through synchronous care, regardless of the setting [4].

Thus, this research aims to analyze the risks associated with patient safety in real-time interaction between healthcare professionals and patients, known as synchronous telehealth. The following review question guided our study: What are the patient safety risks associated with synchronous telehealth?

Methods

Search strategy

An experienced review specialist (ACPNP and CRF) designed and conducted systematic searches of the literature in three major databases: Medical Literature Analysis and Retrieval System Online (Medline), via PubMed®, Embase® via Elsevier®, and the Cochrane Database of Systematic Reviews. These databases were selected for their comprehensive coverage of clinical and systematic review literature.

The search strategy was based on a combination of health descriptors, entry terms, and free-text vocabulary related to patient safety, telehealth, and systematic reviews. We combined terms for each concept using the Boolean operator “OR” and linked different concepts with “AND”. A list of keywords and synonymous terms used to capture potential studies across the resources is summarized in Chart 1. No restrictions were applied on publication dates, covering all evidence published from inception to September 4^th, 2024. The full search strategy template for each database is provided in S1 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Chart 1. List of keywords and synonyms generated as search terms.

https://doi.org/10.1371/journal.pone.0336992.t003

In addition to electronic searches, we (ACPNP and CRF) screened the references of eligible studies and searched for final publications of reports initially found as protocols or conference abstracts to capture additional studies that may have been missed in the original search.

Eligibility criteria

The eligibility criteria for this scoping review are summarized in Table 1. Thus, we included studies that investigated the types of safety concerns in the real-time interaction process between healthcare providers and patients through any ICT. For the present study, ICT was broadly defined as a set of technological resources, such as telephones and videoconferencing tools, used by health professionals to support patient care.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Eligibility criteria for study inclusion and exclusion.

https://doi.org/10.1371/journal.pone.0336992.t001

Systematic reviews were exclusively included in this scoping review with the aim of mapping consolidated knowledge within the secondary literature, thereby facilitating the identification of well-established areas and persistent gaps. This approach also aimed to ensure a higher methodological quality of the included studies while avoiding the duplication of data already synthesized in primary research. We did not limit the review to specific outcomes, allowing the inclusion of studies that assessed various aspects of patient safety, such as adverse events, medical errors, breaches in privacy, technology failures, or communication breakdowns. Studies that used technology solely as a screening tool or included mixed populations (e.g., synchronous, and asynchronous) without reporting separate results for synchronous participants were excluded. There were no restrictions on the design of primary studies included in the reviews or year of publication.

We restricted our review to full-text articles published in English, Spanish, or Portuguese, as these were the languages of expertise for the authors.

Results

Results of the search

Our search initially retrieved a total of 3,641 titles and abstracts. After removing duplicates, 3,462 abstracts were screened. Of these, 3,421 studies were excluded, justification can be found in S2 Table, leaving 41 studies for further scrutiny. In the end, we included 15 systematic reviews (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Identification of studies via databases.

https://doi.org/10.1371/journal.pone.0336992.g001

Characteristics of included systematic reviews

We included 15 systematic reviews published between 2013 and 2024, covering a total of 315 studies. The number of studies in these reviews ranged from 4 to 80, with most reviews including between 10 and 30 studies. All were traditional systematic reviews of primary studies, except for three: a) one included both primary studies and reviews [4]; b) others included systematic overview of reviews [9,10]. Most reviews focused on randomized controlled trials (RCTs), supplemented by cohort and observational studies. A few also incorporated qualitative or mixed-methods studies Table 2. These reviews covered research conducted in a range of countries, reflecting the global scope of synchronous telehealth.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics, methodological appraisal, and safety results in telehealth systematic reviews.

https://doi.org/10.1371/journal.pone.0336992.t002

The study by Mackintosh et al. (2016) [11] reported data exclusively from the United States. The study by Han et al. (2020) [12] included reviews from Norway, Scotland, the United States, and Denmark. One review included studies from nine countries: the United States, Australia, Kenya, Canada, the United Kingdom, Japan, Singapore, New Zealand, and Sweden [13]. Another review also covered multiple countries, such as the United Kingdom, Sweden, New Zealand, Spain, Canada, Norway, and Denmark [14]. Rush et al. (2018) [15] included Germany, the United States, Northern Ireland, and China. Gleeson et al. (2022) [16] examined research from Sweden, the United States, England, and Finland. In another study, the included countries were Germany, China, Colombia, Malaysia, and the United States [17]. Pathak (2021) [18] focused on studies conducted in the United States. Haveland and Islam (2022) [19] brought together studies from a broad range of countries, including the United Kingdom, Australia, the United States, Norway, Denmark, Spain, and Canada. Finally, Vieira et al. (2022) [20] reported data from Spain, China, Turkey, and Brazil.

The United States was cited in 9 (60%) reviews, followed by 6 Norway, 5 the United Kingdom and Canada each, and 4 Sweden. Other countries represented included Germany, China, Spain, Denmark, Finland, Colombia, Malaysia, Northern Ireland, Brazil, Turkey, Kenya, Japan, Singapore, and New Zealand.

Discussion

This scoping review identified patient safety risks associated with telehealth services. The most frequently reported concern was the patient’s experience, highlighted in 60% (n = 9) of the included studies [9,11,14,16–19,21,22]. The second most cited risk was medication safety, appearing in 40% (n = 6) of the articles [13,16–20]. Other relevant risks included communication failures, patient identification, and gaps in training related to telehealth practices. These findings underscore the complexity of ensuring safe, high-quality care in virtual environments and point to priority areas for system-level improvements.

Given these challenges, it is essential to situate discussion within the broader global movement for patient safety. In 2004, the WHO created the Global Alliance for Patient Safety, which drew up the “Patient Safety Goals” (PSG) to guide global efforts to reduce patient harm [23].

The growing use of telehealth expands access to health services, as it reduces costs and travel time [24]. However, the use of ICT can also introduce new risks to the safety of care [4]. Problems with health IT have been widely documented, such as system errors, poor user interfaces, and lack of proper training, can disrupt the care process and lead to delays in decision-making or errors in clinical tasks, ultimately increasing the likelihood of patient harm [25].

The relevance of the patient’s experience is increasing, requiring physical and digital services to be concerned with users’ overall perception of the care offered. This includes understanding patients’ needs and expectations. Health services must find ways to meet these expectations by offering digital means for health promotion, prevention, rehabilitation, or treatment [26].

For safe and reliable assistance, care results must be aligned with better outcomes, cost reduction, promotion of favorable experience and good patient acceptance. In some cases, the convenience of digital care is linked to commuting, making telehealth a preferred option [14,19,21].

In this context, McLean et al. (2013) [9] found that rural populations tend to be more satisfied with remote care, preferring it to the physical model, and Pathak et al. (2021) [18] observed greater satisfaction in telepharmacies when compared to traditional pharmacies.

Despite the risks, such as limitations in digital diagnosis and treatment without physical attendance, Svendsen et al. (2021) [17] reported that patients and doctors positively evaluated the safety and effectiveness of teleconsultation. In contrast, other studies did not find conclusive evidence. Corso et al. (2022) [22] found no difference in satisfaction between virtual and face-to-face care, while Mackintosh et al. (2016) [11] highlighted the need for more research into patient satisfaction.

The literature highlights, as a safety strategy, the importance of selecting an appropriate group of patients for telehealth. This process of carefully selecting patients for whom telemedicine is most appropriate is fundamental to guaranteeing patient safety and the provision of high-quality care. Providers must identify patients individually, with the active participation of each one, and make additional resources available if Internet access limitations represent barriers to care [27]. As noted in one of the studies included in our review, this strategy is particularly relevant in the context of chronic disease management, where telehealth has shown satisfactory outcomes [9].

In addition, the positive outlook of healthcare professionals appears to influence patient engagement positively [11,19]. Therefore, these findings suggest that the patient’s experience using telehealth tends to be positive. Providers should adopt appropriate patient selection criteria and maintain a patient-centered service that accounts for the limitations of virtual care.

In addition to patient experience, another central concern in this review relates to the safety of medication prescription in telehealth, reported in 40% (n = 6) of the included studies [12,15–19]. In 2017, the World Health Organization launched the Global Patient Safety Challenge with the theme “Medication Without Harm,” reinforcing the international relevance of this topic [28].Rush et al. (2018) [15] observed that video calls showed a reduction in medication errors compared to telephone calls, suggesting that visual interaction may be a positive factor. This finding indicates that the absence of visual interaction makes clinical assessment difficult, increases the risk of prescribing errors, and may compromise safety.

The potential to improve safety in prescribing medications, especially in remote areas, was highlighted by Pathak et al. (2021) [18] who found that telepharmacy showed better performance compared to traditional pharmacies in the detection of medication errors and adherence to treatment. However, the authors highlighted the need for more robust research to minimize possible biases.

Similarly, this observation is consistent with findings by Omboni and Tenti (2019) [29], who emphasized the importance of investigating the impact of telepharmacy and its potential risks to patient safety.

Drug prescribing in remote and face-to-face environments has been investigated to identify possible differences and challenges associated with each modality. Han et al. (2020) [12], when comparing the dosage and choice of antibiotics, found no significant differences. However, the authors emphasized the importance of preventing antimicrobial resistance, a risk present in both models of care. Complementing this analysis, Svendsen et al. (2021) [17] warned of the dangers related to prescribing antibiotics, controlled drugs and drugs with potential side effects, highlighting the need for strict monitoring of prescriptions and possible drug interactions.

This is consistent with findings from Ray et al. [30] who found higher antibiotic prescribing in children seen via telemedicine than in person. Such evidence underscores the risks associated with antibiotic prescribing in remote care environments. These findings reveal that communication is not only a frequent source of risk but also a mediating factor in the safe use of medicines.

Furthermore, other studies have associated effective communication with the safe use of medicines. Two of the included reviews in this research [16,19] emphasized this relationship, especially in contexts where clear interaction between professional and patient is essential. Gleeson et al. (2022) [16], for example, identified language barriers as factors that can cause harm during the prescribing process, reinforcing the need for clarity and accuracy in communication.

While Haveland and Islam (2022) [19] highlighted the importance of effective communication in managing adverse drug reactions in cancer patients, further supporting its role as a critical component of patient safety. The aim of WHO was to strengthen systems to reduce medication errors, since unsafe practices related to medications are one of the main causes of preventable harm [31].

Beyond patient-centered strategies, another critical issue identified was effective communication, the third most cited concern. Our analysis revealed, 33% (n = 5) of the studies included addressed communication as a risk to patient safety [4,10,13,14,17]. The emphasis differed depending on the study, encompassing issues such as connectivity problems, unclear instructions, barriers to professional access, and concerns related to information management in virtual care [4,10].

In this scoping review, we identified the risk of information being lost during virtual consultations, compromising understanding of the clinical picture and potentially leading to incorrect diagnoses [14]. This finding aligns with existing literature, which highlights communication failure as one of the most frequent factors in the occurrence of incidents, a concern that has also been raised in external source. This makes effective communication a crucial goal in various healthcare scenarios [32,33].

Findings from the present review indicate that barriers to accessing or communicating with professionals, as well as risks to confidentiality in teleconsultations, such as leakage of sensitive information, unauthorized access and misuse of personal data, were frequently reported across the included studies [4,10]. These concerns are corroborated by external literature, which points out the fact that violations of privacy may affect patient trust in telehealth services [34]. Furthermore, such breaches may lead to ethical and legal impacts [35].

Previous studies have similarly emphasized that effective communication plays an important role in reducing the likelihood of adverse events, as it minimizes errors related to ambiguities or failures in the communication process [36].

Our results indicate the importance of selecting patients suitable for virtual interventions, such as those in need of chronic disease management or lifestyle modification [19]. Campbell et al. (2023) [13] showed that videoconference consultations can have similar diagnostic accuracy to face-to-face consultations for most conditions. Moreover, videoconferencing has been shown to outperform telephone consultations in terms of probability of errors, diagnostic accuracy, and decision-making, creating barriers to error prevention [15]. This is consistent with external findings showing that videoconferencing can be as effective as in-person visits in certain contexts [37].

Although communication brings benefits, it can also harm patient safety [17]. These results align with prior research, wich indicates that risks can be mitigated when they are considered during the implementation of the service [1,16]. Therefore, safety measures should be included at all stages [16].

Within this framework, in-service training is a strategy adopted by health institutions that uses the work environment itself to enhance professional development and improve the quality of care, aligning practical knowledge with organizational culture [38].

In this context, the inclusion of the topic of “patient safety” in training and education programs within telehealth services was evaluated in 5 reviews (33%) [4,9,17,19,20]. In these reviews, it was recommended that professionals be formally prepared for remote care, ensuring that it is carried out without harm. The use of new technologies in healthcare brings risks to patient safety, requiring continuous educational updating [17].

Corroborating the findings of this scoping review, a recent study on telehealth training in allied health professions reinforces this need by showing that many programs lack pedagogical structure and consistent evaluation criteria, which compromises their effectiveness. It emphasizes that in-service training must be grounded in solid educational methodologies to ensure patient safety in remote care [39].

Additionally, the study by Garber and Gustin (2022) [40] reinforces that most healthcare providers do not receive formal telehealth training during their academic education and are expected to learn on the job. This educational gap highlights the strategic importance of in-service training as a compensatory mechanism for the safe and effective use of telehealth in clinical settings.

Moreover, both studies [39,40] indicate that structured educational approaches, such as online courses and simulation-based training, are significantly more effective than improvised or passive formats, like written instructions. These findings demonstrate that the type of training received influences not only provider competence but also adoption and satisfaction. Ultimately, it affects the quality of the patient’s experience.

Educational action should be broad, ranging from the use of technology and solving operational problems to interaction with patients and colleagues, increasing workers’ knowledge of telehealth [4,17,20].

Vieira et al. (2022) [20] evaluated patient safety and the occurrence of adverse events, classified as mild or moderate, highlighting the importance of trained professionals to adapt therapy with a view to patient safety.

The responsibilities inherent in telehealth work processes should be included in education actions [4,9]. Patients should also be trained to use telehealth safely and confidently [19].

In addition to training, accurate patient identification is a critical component of patient safety, recognized as the first goal proposed by the WHO [40].

This issue was explored in only one of the included reviews. Svendense et al. (2021) [17] highlighted potential risk to patient safety associated with identification failures during telehealth.

This finding is consistent with Riplinger et al. [41], who demonstrated that failure to perform it appropriately can compromise clinical decision-making, reduce treatment effectiveness, jeopardize patient outcomes and privacy.

Taken together, this evidence reinforces the concern that, in virtual settings, there is a risk of misidentification, which may lead to errors in orientation or diagnosis. Therefore, it is essential to implement verification processes to ensure that the patient engaged via videoconference or telephone is correctly identified. To support these safety measures, from accurate identification to effective communication, training emerges as a strategic tool within organizations.

This review has limitations related to the absence of standardized tools for assessing patient safety in telehealth. Different study populations were observed, including age group, diagnosis, setting, and country of origin, which may limit the comparability of findings and the depth of the analysis.

We can adapt good practice guidelines from the traditional care model to telecare. These practices may serve as a starting point for the digital environment, such as the creation of surveillance systems for reporting and monitoring adverse events.

Mapping the main adverse events is relevant not only for continuous process improvement, but also for the development of good practices in telehealth platforms and technologies [18]. Given the diversity of telehealth services and the different risk profiles, an approach based on PSG may constitute a possible approach. The adoption of surveillance actions would help to compose a profile of existing risks, allowing for the creation of more effective strategies to mitigate them.

There is also a lack of evidence on the intrinsic barriers to digital services that can lead to errors, such as connection difficulties, poor sound and image quality, and varying levels of digital literacy. To mitigate potential risks, it is necessary to map stages such as registration, information exchange, sending documents, transmitting test results, guidance and prescriptions [42].

In addition, it is necessary to engage and raise awareness among managers and healthcare professionals about patient safety in telehealth. These actors should develop training and continuing education actions for professionals and patients, reinforcing the importance of a transversal and interdisciplinary safety culture.

Conclusion

This scoping review provides evidence that, while telehealth offers valuable alternatives for healthcare delivery, it also presents specific patient safety risks that warrant careful consideration. The most frequently reported risks were related to the patient’s experience and communication failures, followed by issues concerning medication safety, patient identification, and training gaps.

Corroborating findings from the included studies, this review reinforces the need to strengthen care processes to prevent safety incidents in virtual environments. Further research is essential to expand knowledge about these safety issues and to develop strategies to mitigate them. This discussion contributes to guiding future investigations, promoting structured production and dissemination of knowledge, and enhancing safety in telehealth services. To ensure a positive and secure impact of telehealth implementation, it is critical to invest in professional education, continuous training, and careful planning of care processes that reduce the likelihood of adverse events and promote reliable service delivery.

Supporting information