Family physicians’ perceptions of digital health technologies as enablers of personalised medicine in primary care: a cross-sectional survey in Moldova

Introduction

Personalised medicine, considered one of the major directions in the development of modern medicine, integrates multiple types of patient data, including the molecular-genetic profile, offering new possibilities for the personalization of prevention, diagnosis, and treatment [1-3]. These data include -omics data reflecting the patient’s unique biology (metabolomics, proteomics, transcriptomics, epigenomics), together with data on lifestyle, individual characteristics, and anthropometric measurements, and  constitute a central element of personalised medicine [3, 4]. Digital technologies are required if this large volume of data is to be organised and applied in clinical practice [3, 5, 6]. In this context, electronic health records (EHRs), information systems, and other digital solutions are essential components of personalised medicine [4]. The integration of all patient data through these technologies is one of the pillars of the implementation of personalized medicine, and the increasingly widespread adoption of EHRs, as well as the development of population biobanks integrated with these systems, creates significant opportunities for applying personalised medicine in clinical practice [5, 7].

Implementing the third pillar of personalised medicine – healthcare system reform – requires investment in centralized and interoperable digital infrastructure, multidisciplinary collaboration, and new economic models that ensure the sustainability of personalized medicine and equity of access to it [5, 8-10]. At the same time, effective use of personalised medicine in routine clinical practice requires physicians to have competencies in both personalised medicine and digital health, as well as access to user-friendly and clinically useful decision-support systems integrated into the clinical workflow [5, 11, 12].

Furthermore, healthcare professionals play an important role in populating patients’ electronic health records and sharing them for secondary use in research [13]. The family physician therefore occupies a strategic position, collecting and integrating into the patient’s electronic health record information on molecular-genetic characteristics, clinical parameters, family history, as well as social, behavioural, and environmental determinants [13].

Although digitalization is an essential prerequisite for the implementation of personalised medicine and is one of the priorities of the National Programme for Digitalization and Innovation in Health 2025–2030, the adoption of digital technologies is accompanied by a number of concerns expressed by family physicians. These include the perception that intensive use of digital technologies may diminish the quality of the physician-patient relationship and that the large volume of data that must be entered into the Electronic Health Record may undermine one of the fundamental principles of primary care – patient-centred care [14, 15]. In addition, some studies have reported an association between the use of digital technologies and burnout among healthcare professionals [14]. Such concerns can act as barriers to physicians’ adoption of digital technologies and may slow the implementation of personalised medicine. Recent national publications have recognised the need to involve all stakeholders in the process of digitalising the healthcare system [16].

Therefore, the aim of the study was to assess family physicians’ perceptions and practices regarding the use of digital health technologies in clinical practice and to identify potential barriers to the implementation of digitalization as a pillar of personalised medicine.

Materials and methods

The study was an observational, descriptive, national cross-sectional survey. It is part of the doctoral research entitled „Implementation of personalised medicine in the Republic of Moldova from the perspective of family physicians and decision-makers”.

The present article focuses specifically on the digital health component of the doctoral research and explores family physicians’ perceptions regarding the implementation of digital health technologies in the context of personalised medicine in primary care. The assessed domains included the perceived usefulness of digital technologies, integration of electronic health data, patient access to electronic health information, and concerns related to the security of electronically stored medical information. Given the exploratory cross-sectional design of the study, no primary or secondary outcomes were predefined.          

Study population and inclusion criteria. The target population consisted of family physicians currently practicing in primary healthcare institutions in the Republic of Moldova. Eligible participants were those who were actively practicing at the time of the study and who provided written informed consent to participate. No additional exclusion criteria were applied.

         Sampling and sample size calculation. The target population comprised all family physicians practicing in the Republic of Moldova (N = 1536). The minimum required sample size (n = 310) was estimated using Cochran’s formula with finite population correction, assuming a 95% confidence level (Z = 1.96), a 5% margin of error (d = 0.05), and a conservative expected proportion of 50% (p = 0.50), which maximizes the required sample size in the absence of prior estimates. To account for an anticipated non-response rate of 25%, the planned sample size was increased accordingly.

To ensure geographical coverage across the entire Republic of Moldova, a proportionate stratified sampling approach was applied, using the 35 second-level administrative units as strata. The required number of family physicians from each stratum was determined proportionally according to the number of family physicians practicing in each administrative unit. Within each stratum, a sampling frame of family physician positions was constructed based on the number of positions available in primary healthcare institutions. As an individual-level list of family physicians was unavailable, each physician position was assigned a unique identifier linked to the corresponding institution. Random selection was performed at the physician-position level using the WHO STEPS sampling tool. Questionnaires were subsequently mailed to the management of the selected institutions according to the number of positions selected, with institutional representatives facilitating questionnaire distribution within the institution. Healthcare institutions were not sampling units but were used only for questionnaire distribution. Questionnaires were distributed through 124 institutions corresponding to the selected physician positions, and responses were received from 97.

Data collection. Paper-based, self-administered questionnaires were sent by mail to the selected institutions. The selected institutions were contacted, and questionnaires were distributed through institutional representatives to eligible family physicians. During the initial telephone contact, institutional representatives were informed about the objectives of the study and the voluntary nature of participation. The information sheet accompanying the questionnaire also explained that participation was entirely voluntary and that physicians could decline or withdraw without any consequences.

To maximize coverage across all strata, data collection incorporated a responsive survey approach. Return rates were monitored throughout the data collection period, and an additional 38 questionnaires were distributed to administrative units in which the number of returned questionnaires was insufficient to achieve the planned sample size. This approach allowed the planned proportional allocation across strata to be maintained despite differential response rates. During data collection, an additional 38 questionnaires were distributed to administrative units in which the number of returned questionnaires was insufficient to achieve the planned sample size. Data collection was conducted between March and June 2024. The detailed methodology of the data collection process has been described previously [18].

Overall, 454 questionnaires were distributed. A total of 357 questionnaires were returned (return rate 78.6%). Of these, 38 questionnaires were excluded because informed consent was not provided (n = 30), key sociodemographic information was missing (n = 4), more than 50% of questionnaire items were incomplete (n = 2), or duplicate responses were identified within the same institution (n = 2). Consequently, 319 questionnaires were included in the final analysis, representing 70.3% of all distributed questionnaires (319/454) (Figure 1).

Research instrument. Data were collected using a questionnaire developed specifically for this research, entitled “Exploring the level of knowledge and perceptions of family physicians regarding personalised medicine”. The development process, expert content validation, pilot testing, and subsequent modifications of the questionnaire have been described in detail in a previous publication [17]. Briefly, the questionnaire was developed on the basis of a literature review and expert assessment. Content validity was evaluated using a Delphi approach involving two experts with experience in public health and primary healthcare, representing both academic and clinical settings. Subsequently, the instrument was pilot-tested among nine family physicians from both urban and rural settings through individual structured interviews to assess the clarity, comprehensibility, and acceptability of the items. The questionnaire was administered in Romanian, and the pilot participants were not included in the final study sample.

For the present article, a predefined subset of questionnaire items focusing on digital health technologies as enablers of personalised medicine was analysed. These items assessed the use of the electronic medical record (EMR) system in routine clinical practice, physicians’ perceptions of its usefulness for clinical activities, confidence in the security of electronically stored medical information, attitudes towards patient access to electronic medical records, awareness of electronic health data integration as a component of personalised medicine, and perceptions regarding the role of broader digital health technologies, including remote patient monitoring and digital solutions for improving healthcare quality. Although the study assessed physicians’ perceptions of EMR use in primary care, these findings are discussed in the broader context of electronic health record (EHR) implementation and digital health transformation reported internationally. While EMR systems represent a more limited form of electronic documentation compared with interoperable EHR platforms, both rely on the effective integration and use of digital health information in clinical workflows. Consequently, physicians’ attitudes towards EMR use may serve as an indicator of professional readiness and acceptance of future digital health initiatives, including more advanced EHR-based solutions that support personalised care.

Statistical analysis. Data were analysed using Microsoft Excel and IBM SPSS Statistics version 28.0.0.1 (IBM Corp., Armonk, NY, USA). Descriptive statistics, including absolute frequencies (n) and relative frequencies (%), were used to summarize categorical variables and questionnaire responses. For proportions, 95% confidence intervals (95% CIs) were calculated using the standard approximation formula for proportions:

with calculations performed in Microsoft Excel. Age was reported as mean ± standard deviation (SD). Participants’ professional experience was categorized into three groups (≤10 years, 11–20 years, and >20 years) for comparative analyses. Associations between categorical variables were assessed using Pearson’s chi-square test (χ²). Statistical significance was defined as a two-sided p < 0.05. For significant χ² tests, post-hoc pairwise comparisons of column proportions were performed with Bonferroni correction to identify categories contributing to differences in response distributions. Differences in age between urban and rural physicians were assessed using descriptive statistics. Ordinal logistic regression was used to assess whether differences in Likert response distributions by practice setting remained after adjustment for age. Likert responses were analysed as ordinal outcomes, with practice setting and age included as predictors. Professional experience was not included because of its expected strong correlation with age. Sex was not included because of the highly unbalanced sex distribution in the study sample, reflecting the predominantly female composition of the family physician workforce in the Republic of Moldova; however, the possibility of residual confounding cannot be excluded. Results are reported as odds ratios (ORs) with 95% confidence intervals (CIs). The proportional odds assumption was assessed using the test of parallel lines. For overall descriptive presentation of individual items in the text, adjacent positive response categories (e.g., agree and strongly agree) were occasionally combined to indicate the overall proportion of agreement. However, when comparing groups (e.g., urban versus rural physicians), the complete five-category Likert distribution was retained and used for statistical analyses. Missing data were handled using available-case analysis; each item was analysed based on the number of valid responses available for that item.

Ethical considerations. The research protocol and data collection instrument were approved by the Research Ethics Committee of Nicolae Testemițanu State University of Medicine and Pharmacy (minutes no. 6 from 18.05.2022 and no. 2 from 18.12.2023). All participants received written information about the study objectives, the voluntary nature of participation, confidentiality of responses, and their right to decline participation without any consequences. Written informed consent was obtained from all participants who agreed to participate before they completed the questionnaire.

Results

The respondents’ demographic characteristics. Of the 318 respondents for whom data on sex were available, the majority – 288 (90.6%; 95% CI: 86.5–93.1)  were women. Among the participants, 126 (39.5%; 95% CI: 34.1–44.9) worked in urban areas and 193 (60.5%; 95% CI: 55.1–65.9) in rural areas. The mean age of respondents working in urban areas was 47.57 (SD 9.74) years, compared with 52.34 (SD 12.34) years in rural areas (p < 0.001).

Perception of the components of personalised medicine and associated digital technologies. The assessment of family physicians’ perceptions of the components of personalised medicine, including the role of digital technologies, revealed that only 106 respondents (33.2%; 95% CI: 28.0–38.4) identified the electronic recording and integration of patient data through decision-support systems as a component of personalised medicine. In comparison, a considerably higher proportion of respondents (Figure 2) recognised the other components, such as individual patient characteristics, genetics and multi-omics technologies, gene-environment interactions, lifestyle, biomarkers, and anthropometric data.

Use of digital technologies. The majority of respondents – 310 (97.2%; 95% CI: 95.4-99.0) – reported using a primary care electronic medical record system, with no statistically significant differences according to practice setting (χ² = 1.15, df = 1, p = 0.281).

Physicians’ perceptions regarding comfort in using the electronic medical record. Most family physicians reported feeling comfortable using the electronic medical record for entering patients’ medical data. A total of 220 respondents (69.0%; 95% CI: 63.9–74.1) of respondents agreed or strongly agreed with this statement, including 140 (43.9%; 95% CI: 38.5–49.4) who strongly agreed and 80 (25.1%; 95% CI: 20.4–29.9) who agreed. Fifty-seven participants (17.9%; 95% CI: 13.7–22.1), selected the neutral response category, while 42 respondents (13.2%; 95% CI: 9.5–16.9) expressed disagreement, including 20 (6.3%; 95% CI: 3.6–9.0) who disagreed and 22 (6.9%; 95% CI: 4.1–9.7) who strongly disagreed.

Comparative analysis based on practice setting revealed statistically significant differences. Physicians in urban areas more frequently reported strong agreement regarding comfort in using the electronic medical record compared with those in rural areas, χ² = 23.52, df = 4, p < 0.001 (Figure 3).

In the age-adjusted ordinal logistic regression model, practice setting remained significantly associated with responses to this item (OR = 2.85; 95% CI: 1.83–4.44; p < 0.001), whereas age was not significantly associated with response distribution (OR = 0.995; 95% CI: 0.98–1.01; p = 0.596).

The perceived usefulness of digital technologies. This set of questions explored family physicians’ perceptions regarding the usefulness, acceptability, and clinical relevance of digital health technologies, including electronic medical records, remote monitoring systems, and the potential role of digital tools in improving the quality of care. The corresponding results are presented in Table 1.

Perceived usefulness of electronic medical records. About two-thirds of family physicians perceived electronic medical records as useful in their clinical practice. A total of 209 respondents (65.5%; 95% CI: 60.3–70.7) agreed or strongly agreed that electronic medical records facilitate their clinical work, while 53 respondents (16.7%; 95% CI: 12.6–20.8) disagreed or strongly disagreed with this statement.

Urban physicians reported a more favourable response distribution compared with rural physicians, with a higher proportion of positive responses. The difference in response distribution across the five Likert categories was statistically significant (χ² = 25.10; df = 4; p < 0.001). Bonferroni-adjusted pairwise comparisons indicated that this difference was mainly explained by a higher proportion of urban physicians selecting the strongly agree category, whereas rural physicians more frequently selected the neutral response category.

After adjustment for age, urban practice setting remained significantly associated with higher Likert response categories (OR = 2.53; 95% CI: 1.64–3.89; p < 0.001), whereas age showed no significant association (OR = 0.99; 95% CI: 0.97–1.01; p = 0.247), respectively.

Perceived impact of digital technologies on the quality of healthcare. Family physicians also reported a favourable perception regarding the contribution of digital technologies to improving the quality of care. Agreement or strong agreement was reported by 204 respondents (64.0%; 95% CI: 58.7–69.3), whereas 45 respondents (14.1%; 95% CI: 10.3–17.9) expressed disagreement.

A significant difference in the distribution of responses was observed between urban and rural physicians (χ² = 22.02; df = 4; p < 0.001). Post-hoc comparisons with Bonferroni correction revealed a higher proportion of strong agreement among urban physicians, whereas rural physicians had a greater proportion of responses indicating disagreement. After adjustment for age, urban practice setting remained significantly associated with more favourable Likert responses (OR = 2.14; 95% CI: 1.40–3.27; p < 0.001), whereas age showed no significant association (OR = 0.99; 95% CI: 0.97–1.01; p = 0.178).

Perceived usefulness of remote patient monitoring systems. The perceived usefulness of remote patient monitoring systems was rated favourably, with 200 family physicians (62.7%; 95% CI: 57.4–68.0) expressing agreement or strong agreement, whereas 51 respondents (16.0%; 95% CI: 12.0–20.0) reported disagreement. Comparative analysis by practice setting revealed statistically significant differences (χ² = 16.05, df = 4, p = 0.003). Physicians in urban areas generally reported a more favorable perception of the usefulness of remote monitoring systems, with 87 respondents (69.1%; 95% CI: 61.0–77.2) expressing agreement or strong agreement compared with 113 rural physicians (58.5%; 95% CI: 51.6–65.5). The proportion of physicians who strongly agreed was higher in urban areas than in rural areas. In contrast, rural physicians more frequently indicated a neutral position and reported disagreement slightly more often. The distribution of responses differed statistically significantly between groups (Pearson’s χ² = 16.05; df = 4; p = 0.003). The observed urban–rural difference remained significant after adjustment for age, with urban practice setting associated with higher Likert response categories (OR = 1.71; 95% CI: 1.12–2.60; p = 0.012). Age showed an independent association with response distribution, with older age associated with lower likelihood of selecting higher response categories (OR = 0.98; 95% CI: 0.96–1.00; p = 0.032).

Perceptions and attitudes regarding patient access to digital data. In the context of personalised medicine, in which patient access to their own health data represents an essential element, family physicians’ perceptions regarding the benefits and impact of such access on the quality of care were explored. The corresponding results are presented in Table 2.

Perceptions regarding patients’ access to their own digital health data. Overall, there was a slight predominance of favourable views regarding patient access to electronic medical records, with 142 respondents (44.5%; 95% CI: 39.1–50.0) expressing agreement or strong agreement. However, a considerable proportion of physicians reported disagreement (n = 105; 32.9%; 95% CI: 27.7–38.1) or neutral views (n = 72; 22.6%; 95% CI: 18.0–27.2), indicating a heterogeneous pattern of perceptions and uneven acceptance of patients’ access to their own medical data.

Comparative analysis by practice setting revealed statistically significant differences in perceptions regarding patient access to electronic medical records (χ² = 12.42; df = 4; p = 0.014). Urban physicians showed a more favourable response pattern, with a higher proportion of strong agreement responses (Table 2). The observed difference remained significant after adjustment for age, with urban practice setting associated with higher Likert response categories (OR = 1.79; 95% CI: 1.18–2.70; p = 0.006), whereas age showed no significant association (OR = 1.00; 95% CI: 0.98–1.01; p = 0.612).

Perception of the impact of patient access to electronic medical records on the quality of care. Family physicians’ opinions regarding the effect of patient access to electronic medical record data on the quality of care were divided, showing a relatively balanced distribution between favourable and unfavourable evaluations. The proportion of respondents who perceived a positive impact (n = 114; 35.8%; 95% CI: 30.5–41.1) was close to that of those who expressed disagreement (n = 123; 38.6%; 95% CI: 33.3–44.0), while 82 participants (25.7%; 95% CI: 20.9–30.5) adopted a neutral position, indicating a lack of consensus regarding the direct benefits of patient access to medical data within electronic medical records.

Comparative analysis by practice setting revealed statistically significant differences (χ² = 17.30; df = 4; p = 0.002). The proportion of strong agreement was higher among urban physicians (n = 34; 27.0%; 95% CI: 20.0–35.3) compared with rural physicians (n = 22; 11.4%; 95% CI: 7.6–16.7), while disagreement and neutral responses were more frequent among rural physicians. After adjustment for age, urban practice setting remained associated with higher Likert response categories (OR = 1.66; 95% CI: 1.10–2.50; p = 0.016), while age was not independently associated with response distribution (OR = 1.00; 95% CI: 0.98–1.02; p = 0.839).

Perception of data security in electronic medical records. Family physicians' perceptions regarding the security of data stored in electronic medical records were predominantly favourable. Overall, 207 respondents (64.9%; 95% CI: 59.7–70.1) expressed agreement or strong agreement, including 103 participants (32.3%; 95% CI: 27.2–37.4) who agreed and 104 (32.6%; 95% CI: 27.5–37.7) who strongly agreed. A total of 76 respondents (23.8%; 95% CI: 19.1–28.5) selected the neutral response category. Only 36 participants (11.3%; 95% CI: 7.8–14.8) expressed disagreement, including 21 (6.6%; 95% CI: 3.9–9.3) who disagreed and 15 (4.7%; 95% CI: 2.4–7.0) who strongly disagreed.

Comparative analysis by practice setting did not reveal statistically significant differences (χ² = 5.388; df = 4; p = 0.250), indicating a relatively consistent perception among urban and rural physicians regarding the security of electronic medical data.

Discussion

Implementation of personalised medicine faces multiple barriers, one of the most important being the insufficient digitalization of healthcare systems, given that the use of digital technologies is an essential condition for the application of personalised approaches in clinical practice [19, 20]. While the international literature emphasises the integration of digital tools as a necessary condition for implementing personalised medicine, our study reveals a discrepancy in local perception: only one third (n = 106; 33.2%; 95% CI: 28.0–38.4) of surveyed physicians considered electronic medical records and digital technologies to be a component of personalised medicine [5]. This limited perception may reflect several factors, including possible limitations in the functionality of current electronic medical records, as well as insufficient awareness of how advanced electronic health records can integrate large and complex datasets, and pharmacogenomics data, or support clinical decision-making.

By contrast, the high rate of use of digital technologies among family physicians in the Republic of Moldova (97.2%; n = 310; 95% CI: 95.4-99.0) represents a favourable premise for the integration of personalised medicine principles into clinical practice. This finding aligns with reports of widespread digital record use among primary care physicians in other healthcare systems, where electronic health records are used by approximately 96% of family physicians in the European Union and 93% in Canada [21, 22]. However, these comparisons should be interpreted with caution, as the present study assessed the use of an EMR system, whereas international reports on EHRs generally refer to longitudinal, interoperable patient records.

Although the overall perception of digital technologies was predominantly favourable, with approximately two-thirds of physicians expressing agreement or strong agreement regarding their usefulness in clinical practice (65.5% for electronic medical records facilitating clinical work, 64,0% for digital technologies improving quality of care, and 62.7% for remote patient monitoring facilitating clinical work), a considerable proportion of respondents either remained neutral or did not express agreement with these statements. This finding may indicate that, despite widespread exposure to digital technologies, their perceived value and integration into clinical workflows may remain variable. The reasons underlying this pattern, including possible usability issues, workflow incompatibilities, or training needs, were not evaluated directly in this study and require further assessment. Consequently, to support the implementation of personalised medicine, health information systems must be designed to be intuitive, interoperable, and adapted to clinical workflows [16]. Such systems should directly address physicians’ practical needs and facilitate their daily work, ensuring that digital tools are perceived as genuinely useful for patient management.

Accordingly, the development of the digital infrastructure required for personalised medicine must go beyond simple technical implementation. Family physicians should be regarded as partners in the development of digital technologies and actively involved in their design, testing, and evaluation. Integrating their feedback on clinical usefulness, usability, and encountered difficulties may contribute to the development of information systems better adapted to clinical practice and may prevent technology from becoming an additional source of administrative burden [23]. Although the level of use of information systems is very high, full utilization may remain limited by user experience, as reflected by the finding that only about two-thirds of respondents (n = 220; 69.0%; 95% CI: 63.9–74.1) reported feeling comfortable using electronic systems for entering patient data indicates that, despite widespread adoption, challenges related to the effective use of these systems may persist. While the present study did not directly assess the specific factors underlying these challenges, they should be considered when further developing digital health infrastructure. Addressing these challenges requires the implementation of high-performance, integrated, and interoperable systems that support clinical workflows and reduce unnecessary administrative burden, as also emphasized in previous national publications [16].

An important finding is the discrepancy between a generally favourable attitude towards digitalization and the relatively lower level of support for granting patients access to information contained in their medical records: 44.5% (n = 142; 95% CI: 39.1–50.0) of respondents agreed that patients should have access to their electronic records, while 35,7% considered that such access could improve the quality of care.

This finding is particularly important, as one of the key perspectives of personalised medicine involves informed and empowered citizens – partners in health data management and in decision-making regarding their own health [5]. The literature emphasizes that patient involvement is an essential condition for healthcare system reform, and its development requires tools such as digital platforms for data access, participation in health-related decision-making, and patient-oriented informational resources on personalised medicine [24-27].

In this context, patient access to electronic health data remains limited in the Republic of Moldova (below 10%), in contrast to European policy directions aiming to ensure universal access to electronic health records by 2030 [28, 29]. The limited proportion of family physicians supporting patient access to their medical data may therefore represent a potential barrier to advancing a patient-centred care model and the implementation of personalised medicine principles.

In the process of implementing personalised medicine, the family physician plays a strategic role in facilitating understanding of its benefits and encouraging patient engagement in managing their own health. However, this role may be constrained by reluctance towards patient access to medical data, highlighting the need for information and training measures for family physicians regarding the benefits of patients’ access to their own digital health data, including in the context of the national implementation of the Electronic Health Record (2025-2030).

The study results also indicate that a substantial minority of physicians do not consider data in electronic medical records to be adequately protected: 64.9% (n = 207; 95% CI: 59.7–70.1) expressed confidence in the security of these data, while 23.8% were neutral and 11.3% disagreed.. These concerns are also reported in the literature, being recognized as important barriers to the adoption of digital health technologies, particularly in the context of personalised medicine, where sensitive data such as genetic and predictive information are used [30, 31]. Addressing these concerns requires strengthening the legal and ethical framework for data protection, as well as incorporating components related to digital technologies, cybersecurity, and associated ethical and legal aspects into personalised medicine training programmes.

The findings further highlight differences between urban and rural physicians. After adjustment for age, practice setting remained significantly associated with Likert response distributions across all analysed items, suggesting that observed urban–rural differences were not fully explained by age differences between groups. Urban physicians reported higher agreement with statements reflecting comfort with digital systems and more positive perceptions of digital technologies, while rural physicians showed greater reluctance both regarding patient data access and perceived usefulness of these technologies. These differences may be related to inequalities in digital infrastructure, access to resources and technical support – also highlighted in the National Digital Transformation Programme, as well as differing needs for information and training. In the context of personalised medicine, these disparities may contribute to inequities in access to modern healthcare services and to uneven implementation of personalised approaches between urban and rural settings.

Strengths and limitations

This study has several strengths and limitations. A major strength is the national stratified sampling design, which enabled the inclusion of family physicians from all second-level administrative units of the Republic of Moldova. Additional strengths include the relatively high questionnaire return rate (78.6%), the use of a structured questionnaire developed specifically for this research, ethical approval, and the reporting of 95% confidence intervals for the main estimates.

Several limitations should also be considered. First, the cross-sectional design precludes causal inference and allows only the identification of associations. A strength of the study was the random selection of family physician positions from each administrative unit. However, the facilitation of questionnaire distribution within healthcare institutions by institutional representatives does not completely exclude a potential risk of selection bias. During the initial telephone contact and in the accompanying invitation letter, institutional representatives were informed that participation was entirely voluntary and that physicians should decide independently whether to participate. Physicians also received an information sheet stating that participation was voluntary, responses would remain confidential, and withdrawal from the study was possible at any time without consequences. Questionnaires were completed individually and returned in sealed pre-addressed envelopes to help preserve confidentiality. Nevertheless, the involvement of institutional representatives in questionnaire distribution may have introduced a degree of perceived institutional influence or social desirability bias. This potential influence is likely to have been limited, as the study explored self-reported perceptions, attitudes, and experiences regarding digital health technologies rather than physicians’ professional performance, institutional compliance, or individual competencies. In addition, non-response bias cannot be excluded because information on non-respondents was unavailable. Although the national stratified sampling design supports the external validity and national relevance of the findings, the potential for selection bias, possible residual dependence of responses at the institutional level, should be considered when interpreting and generalizing the results to all family physicians in the Republic of Moldova.

As with all questionnaire-based studies, self-reported responses are subject to reporting bias. Although the questionnaire underwent expert content validation and pilot testing before implementation, additional psychometric evaluation, including construct validation, would further strengthen the measurement properties of the instrument. Furthermore, the adjusted analyses included age but did not account for all potential confounding factors; therefore, residual confounding cannot be excluded. Finally, although the findings provide valuable national evidence on family physicians’ perceptions of digital health technologies, the distinction between the locally implemented electronic medical record (EMR) system and interoperable electronic health record (EHR) platforms should be considered when interpreting international comparisons and generalizing the findings.

Conclusions

Among surveyed family physicians, use of primary-care electronic medical record systems was nearly universal, whereas recognition of digital data integration as a component of personalised medicine and support for patient access to records were less frequent. Urban-rural differences in perceptions of digital technologies persisted after adjustment for age. These findings support further evaluation of physicians’ user needs and the development of user-centred digital technologies tailored to clinical practice, as well as targeted education on digital health, interoperability, data sharing, and patient access.

Competing interests

None declared.

Authors’ contributions

IG contributed to data collection, data analysis, and manuscript drafting. NZ contributed to study conceptualization, coordination and overall research management, and critical revision of the manuscript. Both authors reviewed and revised the manuscript and approved the final version for submission.

Acknowledgements and funding

The authors express their gratitude to the leadership of the research project Pilot Testing the Application of Personalised Medicine Principles in the Management of Patients with Chronic Non-Communicable Diseases for the support provided, as well as to all family physicians who responded positively to the invitation to participate in the study.

Data availability statement

The dataset generated and analyzed during the current study is not publicly available due to ethical and confidentiality considerations regarding participant data. Aggregated data supporting the findings of this study may be available from the corresponding author upon reasonable request, subject to applicable ethical and institutional requirements.

Ethics statement

The research protocol and data collection instrument were approved by the Research Ethics Committee of Nicolae Testemițanu State University of Medicine and Pharmacy (minutes no. 6 from 18.05.2022 and no. 2 from 18.12.2023). All participants received written information about the study objectives, the voluntary nature of participation, confidentiality of responses, and their right to decline participation without any consequences. Written informed consent was obtained from all participants who agreed to participate before completing the questionnaire.

Provenance and peer review

Not commissioned, externally peer reviewed.

Authors’ ORCID IDs

Ilenuța Gușilă – https://orcid.org/0000-0003-1326-5342

Natalia Zarbailov – https://orcid.org/0000-0003-0120-3072

References

1. Molla G, Bitew M. Revolutionizing personalised medicine: synergy with multi-omics data generation, main hurdles, and future perspectives. Biomedicines. 2024;12(12):2750. doi: 10.3390/biomedicines12122750.
2. Council of the European Union. Council conclusions on personalised medicine for patients [Internet]. Brussels: The Council; 2015 [cited 2026 Jun 12]. Available from: https://op.europa.eu/en/publication-detail/-/publication/f416ce37-a48c-11e5-b528-01aa75ed71a1
3. Badr Y, Abdul Kader L, Shamayleh A. The use of big data in personalised healthcare to reduce inventory waste and optimize patient treatment. J Pers Med. 2024;14(4):383. doi: 10.3390/jpm14040383.
4. Jaskulski S, Nuszbaum C, Michels KB. Components, prospects and challenges of personalised prevention. Front Public Health. 2023;11:1075076. doi: 10.3389/fpubh.2023.1075076.
5. International Consortium for Personalised Medicine; Vicente AM, Ballensiefen W, Donertas D, Eklund M, Ivask A, Jönsson J, et al. The ICPerMed vision for 2030: how can personalised approaches pave the way to next-generation medicine? [Internet]. Cologne: ICPM; 2019 [cited 2026 Jun 12]. Available from: https://www.icpermed.eu/wp-content/uploads/Vision_Paper_2019-1.pdf
6. Khan SN, Danishuddin, Khan MWA, Guarnera L, Akhtar SMF. Multi-modal AI in precision medicine: integrating genomics, imaging, and EHR data for clinical insights. Front Artif Intell. 2026;8:1743921. doi: 10.3389/frai.2025.1743921
7. Abul-Husn NS, Kenny EE. Personalised medicine and the power of electronic health records. Cell. 2019 Mar 21;177(1):58-69. doi: 10.1016/j.cell.2019.02.039.
8. Litvinova Y, Lafolie P, Panteli D. Personalised medicine for healthier populations: Key considerations for policy-makers [Internet]. Copenhagen (Denmark): European Observatory on Health Systems and Policies; 2025 [cited 2026 Jun 12]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK618018/
9. Ginsburg GS, Phillips KA. Precision medicine: from science to value. Health Aff. 2018;37(5):694-701. doi: 10.1377/hlthaff.2017.1624.  
10. Blasimme A, Fadda M, Schneider M, Vayena E. Data sharing for precision medicine: policy lessons and future directions. Health Aff. 2018;37(5):702-9. doi: 10.1377/hlthaff.2017.1558.
11. Pearce A, Terrill B, Alffenaar J, et al. Pharmacogenomic testing: perception of clinical utility, enablers and barriers to adoption in Australian hospitals. Intern Med J. 2022;52(7):1135-43. doi: 10.1111/imj.15719.
12. Borges do Nascimento IJ, Abdulazeem H, Vasanthan LT, et al. Barriers and facilitators to utilizing digital health technologies by healthcare professionals. NPJ Digit Med. 2023;6(1):161. doi: 10.1038/s41746-023-00899-4.