Literature review

In recent decades, digitization has spread into every aspect of society, including healthcare, and the field of psychiatry is no exception. Various innovative technologies have pervaded psychiatric practice, revolutionizing the delivery and accessibility of mental health care services [13]. While this process of digitization began much earlier, it was the widespread enforcement of social distancing measures and shortage of healthcare workers resulting from the COVID-19 pandemic that compelled the healthcare community to realize that conventional methods were inadequate to meet the burden of the pandemic-induced surge in mental health disorders [14]. This drove the rapid development and implementation of digital techniques in mental health care services, marking the inception of what we now know as “digital psychiatry [15].”

Globally, digital psychiatry has emerged as an efficient solution for mental health workforce shortages and access disparities. Evidence from high-income countries demonstrates that telepsychiatry can deliver clinical outcomes comparable to face-to-face care, with advantages in cost efficiency and patient satisfaction [16, 17]. Similarly, AI-driven diagnostic support tools and smartphone-based interventions have shown potential to enhance early detection and self-management of mental disorders [18, 19].

In LMICs, where over 75% of individuals with mental disorders remain untreated, digital platforms have been promoted as potential equalizers [4, 20]. Given that Pakistan reports some of the highest rates of anxiety and depression in South Asia [9], coupled with a critical shortage of mental health professionals, limited healthcare infrastructure, and pervasive stigma [10], integrating digitization into psychiatric practice has become increasingly essential.

Studies in Pakistan have shown favourable results on the implementation of digital mental health interventions. For example, a study conducted in rural Sindh found reductions in anxiety and depression scores after delivery of digital counselling interventions by lay health workers [21]. Further studies showed that technology-based mental health interventions were found to be feasible, acceptable, and appropriate by the community when delivered by community health workers in resource-limited areas of Sindh [11]. Surveys of Pakistani psychiatrists indicate growing, but cautious, engagement with telepsychiatry, with less than half of participants indicating telepsychiatry as being a convenient approach to patient care [22].

Indeed, one of the major barriers to the widespread implementation of these technologies in resource-limited countries, including Pakistan, is a hesitancy to adopt these tools by practitioners, stemming from a lack of familiarity with and training in their applications [12].

Many countries lack curricula-specific training in digital methods of psychiatric diagnosis and interventions [12]. According to a survey conducted among medical students and psychiatric trainees, 58.8% of Pakistani respondents stated that they had not received adequate training in telepsychiatry [12]. This lack of awareness has the potential to fuel misconceptions regarding the capabilities, safety, and benefits of these technologies, impeding their integration into the healthcare system. Therefore, in order to introduce digital psychiatry solutions in Pakistan, it is imperative to examine the current level of knowledge and viewpoint of the psychiatric workforce. To the best of our knowledge, literature assessing digital literacy, perspectives, and attitudes of Pakistani practitioners in the context of psychiatry is scarce. This study aims to bridge that gap by gauging their level of knowledge and perceptions of digital psychiatry and of the barriers to its implementation; their familiarity with its associated tools; and their willingness to adopt these services in practice. In doing so, it hopes to lay the foundation for future recommendations and strategies aimed at increasing access to digital mental healthcare in Pakistan.

Methods

This cross-sectional study focused on Knowledge, Attitudes, and Practices (KAP) from public academic hospitals across Pakistan to explore the perspectives of psychiatrists concerning their familiarity with, perceptions of, and willingness towards various digital psychiatry tools. Data were collected via an online questionnaire distributed through Google Forms between January 17 and June 24, 2023.

Participants

Participants included psychiatry residents, fellows, and psychiatry consultants employed in Pakistan’s public healthcare system. Individuals without formal psychiatric training, such as interns, medical students, non-psychiatric nurses, and other non-psychiatric specialists, were excluded.

Survey design

A multidisciplinary core team of psychiatrists, residents, and medical students developed a 48-item survey divided into four sections and 12 subsections. The survey was disseminated via volunteers using social media platforms (Twitter, WhatsApp, LinkedIn). Experts in digital psychiatry conducted internal validation within the core team, focusing on item clarity, content alignment, and thematic coverage. For external validation, a pilot study was conducted at a local public-sector medical university involving n = 23 psychiatrists (consultants and residents). Feedback was obtained through structured debriefing and reviewed for clarity, length, and content relevance. With that, 4 independent experts in digital psychiatry and health education were consulted to assess content validity, leading to minor adjustments in item phrasing and sequencing.

The four sections, telepsychiatry, AI, MHAs, and VR, assessed respondents’ knowledge, perceptions, and willingness to engage with digital psychiatry tools. Knowledge was assessed based on familiarity with relevant terminology and its applicability in Pakistan’s healthcare context. Perception focused on feasibility, accessibility, sustainability, and affordability. Willingness included readiness to collaborate, integrate, and pursue training in digital psychiatry. Participants were also asked about existing frameworks and training opportunities in their institutions, as well as challenges faced during implementation. To prevent duplicate responses, participants were required to submit their email addresses, which were used solely for authentication and were not linked to individual data responses.

Participant recruitment

Participants were recruited through convenience sampling via the Pakistan Psychiatric Society network; departmental heads helped identify consultants, registrars, residents, and medical officers in public hospitals, and questionnaires were distributed using their official contact details.

Survey dissemination

Volunteers facilitated data collection by distributing the Google Forms survey within their hospital networks via email and conducting on-site visits to public hospitals. Use of Google Forms for survey dissemination and self-reported data collection poses inherent limitations. Recruitment through social media and hospital-based volunteers may have introduced self-selection bias, as individuals with a stronger interest or familiarity with digital psychiatry tools may have been more likely to participate. This could limit the generalizability of the findings to all psychiatrists in the public sector. Moreover, the reliance on self-reported data increases the risk of response bias, including social desirability bias and inaccurate recall. To mitigate duplication, email-based authentication was employed, and survey instructions emphasized honest and independent responses. Nonetheless, these methodological constraints are acknowledged as potential influences on data validity and interpretation. To mitigate these concerns, instructions emphasized confidentiality, honest reporting, and independent completion. Respondents provided demographic data, including age, professional rank, years of service, and area of practice.

Data extraction and analysis

Due to the absence of a centralized national registry, the population of public-sector psychiatrists in Pakistan was estimated to be 400 based on consultations with hospital departments. Using Raosoft’s sample size calculator with a 5% margin of error and 95% confidence interval, the required minimum sample size was determined to be 200.

Survey responses were exported from Google Forms to Microsoft Excel 365 and cleaned before statistical analysis in Statistical Package for the Social Sciences (SPSS) version 26. Descriptive statistics (means, standard deviations, and frequencies) summarized participant characteristics and responses across the four domains. Reliability analysis assessed the internal consistency of the questionnaire, and additional correlational analysis explored relationships among key variables.

Correlation analyses were performed to examine relationships among the key domains of digital psychiatry, telepsychiatry, AI, MHAs, and VR in terms of knowledge, perceptions, and willingness. Before choosing the correlation method, we assessed the statistical assumptions necessary for the use of Pearson correlation. Specifically, the normality of continuous variables was tested using the Shapiro–Wilk test, and visual assessments were made via histograms and Q–Q plots. Linearity between variables was evaluated using scatterplots, and multicollinearity among variables (where relevant for regression) was assessed through the variance inflation factor (VIF). Shapiro–Wilk test for normality indicated that several continuous variables, especially those related to perception and willingness towards AI and VR, significantly deviated from normality (W = 0.874–0.941, p < 0.05). Hence, Spearman’s rho was used instead of Pearson’s correlation for these variables. For regression models, multicollinearity was assessed using the VIF, with values ranging from 1.12 to 2.34, well below the cutoff of 5, confirming the absence of problematic collinearity.

Construct validity of the survey was assessed using the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity (BTS). The KMO statistic was 0.812, exceeding the recommended threshold of 0.70, indicating sampling adequacy for factor analysis. BTS was statistically significant (χ² = 2,765.43, df = 1,128, p < 0.001), confirming sufficient inter-item correlations. Internal consistency reliability was evaluated with Cronbach’s alpha, which ranged from 0.78 to 0.87 across the main domains (telepsychiatry, AI, MHAs, VR), reflecting good reliability.

Results indicated that several variables, particularly those related to perception and willingness in less familiar domains like AI and VR, violated the assumption of normality. Consequently, we used Spearman’s rank-order correlation for these non-normally distributed variables. Pearson correlation was retained only for relationships between variables that met the normality and linearity criteria. This hybrid approach ensured the robustness and validity of the statistical findings. All correlation coefficients (Spearman’s ρ and Pearson’s r) were reported with corresponding significance levels in the results section. Factors associated with digital psychiatry readiness were further explored using multiple logistic regression models. All significance tests were two-tailed, with p < 0.05 considered statistically significant.

For the qualitative component, open-ended responses were analyzed using reflexive thematic analysis following the six-phase framework proposed by Braun and Clarke (2006), using NVivo Version no 15. This approach allowed for a flexible, iterative examination of meaning across the dataset while maintaining sensitivity to context. The analysis was conducted using Quirkos software version 3, which facilitated the coding, comparison, and refinement of themes.

Initially, two independent coders conducted inductive coding, generating codes directly from the data without imposing preconceived categories. Codes were then grouped into broader patterns to develop preliminary themes. The team held regular meetings to discuss coding discrepancies, refine theme definitions, and ensure analytic consistency. Themes were reviewed and finalized based on their internal coherence and relevance to the research questions, particularly around barriers to implementation, institutional readiness, and training gaps in digital psychiatry. Inter-coder reliability between the two independent analysts was calculated on a 20% subsample, yielding a Cohen’s kappa of 0.84, which reflects strong agreement. Coding discrepancies were resolved through discussion until consensus was reached. Example codes included: ‘Limited institutional support’ → under the theme ‘Barriers to Implementation’, ‘Need for formal training workshops’ → under the theme ‘Capacity Building and Training Gaps’, ‘Concerns about patient confidentiality in telepsychiatry’ → under the theme ‘Ethical and Privacy Challenges.’ Reflexivity was maintained throughout, with coders documenting their positionality and interpretations during the process. Representative quotes were extracted to illustrate emerging themes. This dual-analyst strategy enhanced the credibility and depth of qualitative findings.

Demographics

We identified 39 psychiatric departments of public hospitals, of which 23 were in Punjab, 6 in Sindh, 9 in Khyber Pakhtunkhwa, and 1 in Balochistan. Data was collected from 18 departments and included departments from all provinces. The sample comprised 55.5% (n = 111) male and 44.5% (n = 89) female participants. Most respondents (56.0%; n = 112) were aged 20–30 years, followed by 36.0% (n = 72) aged 31–40, 6.0% (n = 12) aged 41–50, and 2.0% (n = 4) over 51. A majority (95.0%, n = 190) were Pakistani nationals, while 5.0% (n = 10) were non-Pakistani. In terms of professional rank, 22.5% (n = 45) were consultants, 6.5% (n = 13) registrars, 54.0% (n = 108) residents, and 17.0% (n = 34) medical officers. All participants were Bachelor of Medicine, Bachelor of Surgery (MBBS)-qualified medical doctors. Consultants and registrars had completed the College of Physicians and Surgeons Pakistan (CPSP) postgraduate fellowship in psychiatry (FCPS Psychiatry). Residents were MBBS-qualified doctors currently enrolled in FCPS Psychiatry training, while Medical Officers were MBBS-qualified clinicians providing psychiatric or general services, but not pursuing FCPS Psychiatry. Most participants (95.5%, n = 191) practiced in public hospitals, with only 4.5% (n = 9) in academic or research settings. Regarding participants’ demographics, 26.0% (n = 52) served predominantly urban populations, 60.5% (n = 121) suburban or rural, and 13.5% (n = 27) a mix of both. As for experience, 68.0% (n = 136) had under 5 years, 23.5% (n = 47) had 5–10 years, 5.0% (n = 10) had 11–15 years, and 3.5% (n = 7) had over 15 years in psychiatry (Table 1).

Telepsychiatry

A majority of participants reported basic familiarity with telepsychiatry, 68.5% (n = 137), yet only 42.0% (n = 84) understood its practical applications. Notably, institutional exposure was limited as most respondents 67% (n = 134) stated they had not received any formal training, and nearly half (47.5%; n = 95) indicated that telepsychiatry-related meetings or conferences were not held at their workplace. Over 72.5% viewed it as time-efficient (n = 145), and 75.5% regarded it as a viable option for patient care (n = 151). A large majority (79.0%; n = 158) expressed willingness to consult with larger telepsychiatry centers, and 64.0% (n = 128) were open to using telepsychiatry for real-time observation of procedures like electroconvulsive therapy (ECT).

However, resistance to integration within current health systems was evident: 53.5% (n = 107) disagreed with incorporating telepsychiatry into existing psychiatric services, and only 37.0% (n = 74) supported integration (Table S1).

AI

A majority (68.5%, n = 137) of participants reported unfamiliarity with AI applications in psychiatry. Similarly, 60.5% (n = 121) had not received institutional training on AI usage at their institution, while 39.5% (n = 79) had received some level of AI training. Perceptions were mixed: 46.5% (n = 93) were neutral on AI’s viability in psychiatric services, and 77.5% (n = 155) disagreed with its cost and time-saving potential. Over half (53.5%, n = 107) disagreed that AI could replace conventional psychiatry practices. A significant portion (63.5%, n = 127) reported a lack of understanding of AI algorithm outcomes, though 55.0% (n = 110) showed interest in AI-psychiatry collaborations. While 31.5% (n = 63) were neutral about AI integration into psychiatric care, 43.5% (n = 87) agreed that building trust between clinicians and AI technologies is essential for future development (Table S2).

MHAs

Most respondents (62.0%; n = 124) were unfamiliar with MHA for psychiatric care, and 40.5% (n = 81) disagreed with their clinical utility. More than half (53.0%; n = 106) strongly disagreed that these apps could reduce time and costs within psychiatric systems. Awareness among psychiatrists regarding app recommendations was reported by only 30.0% (n = 60). Willingness to integrate such apps was noted in 35.5% (n = 71), while 53.0% (n = 106) expressed interest in learning about new MHA. However, only 24.5% (n = 49) agreed on the possibility of their integration within Pakistan’s psychiatric care framework (Table S3).

VR

About 32.5% (n = 65) of respondents were familiar with VR, and only 25.0% (n = 50) reported having received institutional training. Only 27.0% (n = 54) agreed that VR could save time and money in psychiatric care. The majority (79.5%; n = 159) cited insufficient digital infrastructure for VR adoption. While 37.5% (n = 75) disagreed with the possibility of VR integration. Over half (51.0%; n = 102) did not view VR as a viable replacement for in-person services. A promising 55.5% (n = 111) expressed willingness to learn about VR, but 85.0% (n = 170) disagreed with their willingness to use it in practice. Furthermore, 40.5% (n = 81) were unwilling to view live procedures such as ECT via VR (Table S4).

Reliability analysis

Reliability analysis showed strong internal consistency for telepsychiatry (Cronbach’s alpha = 0.847) and VR (Cronbach’s alpha = 0.866), suggesting these domains are robust in clinical settings. AI (Cronbach’s alpha = 0.756) showed commendable stability, while MHA exhibited moderate reliability (Cronbach’s alpha = 0.614) (refer to Table 2).

Correlation analysis

Correlation analysis results revealed significant positive correlations between telepsychiatry and AI (r = 0.493), as well as MHA (r = 0.429), indicating their synergistic potential in transforming psychiatric care. Additionally, the significant positive correlations between AI and MHA (r = 0.497) and VR (r = 0.390) highlight the interconnectedness of these innovative technologies in mental health. Notably, VR demonstrates positive correlations across all variables (refer to Table 3).

Qualitative findings

Thematic analysis of the open-ended responses revealed five major themes: lack of infrastructure/resources, lack of education/awareness, unwillingness to adapt, complexity of technology, and confidentiality concerns. Frequencies are presented in Table 4, while illustrative quotations provide context.

Telepsychiatry

Our findings denote that 68.5% (n = 137) of participants were familiar with telepsychiatry. This reflects the significant gap in the understanding of digital psychiatry in Pakistan, where conventional treatment methods dominate. These results echo international concerns. In the U.S., despite its advanced healthcare infrastructure, many psychiatry residents lack hands-on experience with telepsychiatry [23]. Regionally, a study in Saudi Arabia found moderate familiarity (44.6%) but inadequate practical exposure, expert availability, and workplace integration, paralleling the Pakistani context [24]. Similarly, a broader Asia-Pacific study noted that 58.8% of Pakistani respondents lacked theoretical training, with even less practical experience, reflecting a regional pattern [12].

Only 16.5% (n = 33) of participants had received formal training in telepsychiatry, while 25.5% (n = 51) reported unfamiliarity with guidelines or telepsychiatry practice, highlighting the urgent need for capacity-building and national policy implementation. International literature further supports these findings. In Poland, only 15% of psychiatrists had telepsychiatry knowledge [25], whereas in Italy, basic familiarity was offset by limited academic coverage and training [26]. It has been noted that clinicians and patients have shown concerns about ensuring privacy during telepsychiatry sessions. Given the limited digital infrastructure to protect patient information, it is likely that telepsychiatry is not seen as part of common practice with higher risks to patient safety, thereby limiting its exposure and familiarity for Pakistani psychiatrists [27].

Encouragingly, Pakistani practitioners recognize the potential of telepsychiatry: 75.5% (n = 151) agreed it is a viable care model, and 72.5% (n = 145) believed it could reduce costs and save time. As mobile phone penetration increases, telepsychiatry offers a practical solution to reach underserved populations, particularly in the wake of climate-induced disasters in remote settings.

Most (75.0%; n = 150) of the participants confirmed the presence of Information and Communication Technologies (ICT) in their workplaces. Similar findings were reported in the Asian-Pacific region (82.3%) [12] and Saudi Arabia [24]. However, only 37.0% (n = 74) of participants expressed confidence in Pakistan’s digital infrastructure, a concern echoed in Egypt [28]. Additionally, 43.0% (n =86) did not see telepsychiatry replacing in-person services, mirroring findings from India (62%) [27] and the Netherlands, where face-to-face care remains preferred with limited online augmentation [29]. This hesitancy in Pakistan may stem from technological constraints, cultural preferences favoring traditional care models, and system-level unpreparedness due to limited formal training and infrastructure. Furthermore, connecting with patients in their homes through a stable internet connection has its own cultural barriers, particularly in conservative communities where family and religious commitments coupled with societal stigma can potentially limit the therapeutic alliance from being established [27].

Despite infrastructural and cultural barriers, a strong willingness to adopt telepsychiatry emerged: 79.0% (n = 158) were open to collaboration with larger centers via telepsychiatry. This contrasts with Saudi Arabia, where higher readiness was reported: 88% for consultation use, 85.7% for peer adoption, and 84.8% for system readiness [24]. Addressing these challenges within the Pakistani context will require investments in digital infrastructure, enacting supportive policy reforms, clear operational guidelines, and delivering widespread training to educate and familiarize psychiatrists with telepsychiatry’s benefits and implementation strategies.

AI

In our study, only 11.0% (n = 22) knew about its applications in psychiatry. Only 34.0% (n = 68) were aware of AI-based management procedures being used at their workplaces. These findings highlight a significant gap in AI literacy and implementation in Pakistan’s mental health system. A similar Pakistani study found that only 39.0% of postgraduate doctors were familiar with AI, 72.6% were unaware of its practical applications, and 92.9% had no formal AI education during training [30]. In contrast, healthcare professionals in Bahrain reported higher AI knowledge [31], suggesting Pakistan’s gaps stem from outdated curricula, limited resources for AI education, and poor clinical integration of AI. Addressing these issues is crucial to empowering psychiatrists with the skills needed to leverage AI for enhanced diagnostic and therapeutic outcomes in mental health care.

In our study, only 23.5% (n = 47) of participants viewed AI as a viable tool for psychiatric services, but only 13.0% (n = 26) believed it could save time and money. This contrasts with another Pakistani study where 74.4% acknowledged AI’s role in diagnostics [30], and a Bahraini study where 70% believed AI would reduce diagnostic time [31]. Challenges with integrating AI-based solutions as part of reimbursements in patient care continue to be a global challenge, with no clear direction within mental health settings in Pakistan for billing purposes. Cultural and religious influences in Pakistan may contribute to skepticism toward AI in mental health. In Pakistani society, where cultural norms and religious beliefs play significant roles in daily life, concerns about privacy, confidentiality, and the perceived impersonality of digital platforms and AI could deter individuals from their use [32, 33]. Additionally, cultural preferences for face-to-face interactions and traditional healing methods might prioritize in-person consultations over digital alternatives [34].

Only 27.5% of our participants believed the digital infrastructure was adequate for AI services, similar to the 25.6% in Bahrain [31]. Concerns about AI replacing traditional psychiatric practice were expressed by 40.5% of respondents, compared to 75% in Bahrain who feared AI impacting mental healthcare employment [31]. Additionally, 81% in Bahrain doubted AI could deliver empathetic care, echoing global skepticism where most psychiatrists believe AI could augment specific tasks but not replace them [31, 35]. In Pakistan, 53.4% of doctors disagreed with AI’s potential to replace physicians [32]. The lack of confidence in AI integration reflects concerns about job security and infrastructure readiness.

Our study identified challenges in AI implementation, including the complexity of AI algorithms in psychiatric practice. However, financial and technical limitations, insufficiently trained health professionals, absence of data on public perceptions, fear of replacement, societal barriers, confidentiality concerns, and ethical considerations were highlighted as broader challenges in other studies [36–38].

Still, 55.0% (n = 110) of participants were interested in developing AI-based solutions in healthcare and supported integrating AI into the current health infrastructure. These findings align with a systematic review where over 60% favored clinical AI integration [39]. This may indicate a broad openness among healthcare professionals, including Pakistani psychiatrists, towards AI technologies and a potential for successful AI integration in psychiatric healthcare. Respondents emphasized the importance of fostering trust between humans and AI, establishing ethical, legal, and infrastructural frameworks, especially critical in Pakistan, where awareness remains low [12]. A study on AI in LMICs emphasized training and retaining local experts, establishing monitoring systems, system-based implementation, and responsible local leadership, offering broader strategic contrasts to Pakistan’s psychiatry-specific challenges [39].

MHAs

As per our data, MHAs among the psychiatric workforce in Pakistan are markedly low, with only 4.0% (n = 8) reporting familiarity with their usage. In comparison, studies from Germany, Portugal, and France show moderate levels of familiarity and usage of MHAs [40–42]. The lack of standardized guidelines in Pakistan may contribute to this gap, reinforcing the need for independent certification to ensure app quality and reliability [30]. Our findings highlight a significant knowledge gap in MHA use for psychiatric care in Pakistan, calling for targeted educational initiatives and the development of institutional protocols. Addressing these deficits could promote better integration of digital tools, enhancing accessibility and support in mental health services.

Despite low awareness, 30.0% (n = 60) of participants recognized the value of MHA in psychiatric care, and 24.5% (n = 49) considered it time- and resource-saving. In contrast, a German study reported that 68.3% saw the potential for MHA to lower psychotherapy barriers [42], and in Portugal, 70% believed in their efficiency, with 45.5% considering them cost-effective [40]. Additionally, 35.5% (n = 71) of our participants were ready to use MHA for patient care. In Germany, 68.3% were willing to use treatment apps [42], while in Portugal, 22% of psychiatrists and 48% of psychologists were likely to prescribe digital MHA within the year [43]. An important reason for the lack of acceptance for MHA could also be its lack of cultural and language adaptation to local models of care. Such apps are often validated and built on Western datasets without involving psychiatrists from LMICs in the development process. This may widen the digital literacy gap and create further challenges for adoption in Pakistani mental health systems.

These insights reflect a cautiously optimistic outlook on MHAs effectiveness and cost-efficiency in psychiatric care. However, the relatively lower percentages in our study suggest the need for further exploration, awareness, and training to realize the potential of MHA in psychiatric care.

VR

Our findings show moderate awareness and familiarity with VR and its applications within Pakistan’s psychiatric care. Only 32.5% (n = 65) of respondents were familiar with VR. In contrast, studies from Australia and the U.S. show significantly higher exposure, likely due to better funding and earlier adoption of digital tools in mental health care [44, 45]. Bridging this gap through targeted education and training initiatives could facilitate VR integration into psychiatric care, potentially improving treatment outcomes and patient engagement.

Only 24.0% (n = 48) of respondents deemed VR as a viable approach to psychiatric care. Trust in existing infrastructure was strikingly low, with only 4.0% (n = 8) expressing confidence. Comparatively, a survey from the Asia-Pacific region shows that 82.3% of respondents believed digital health could reduce costs, and 68.2% expected digital psychiatry might replace in-person care [12]. Interestingly, a Norwegian study noted improved clinician attitudes toward VR after direct exposure [46], suggesting that hands-on experience can foster acceptance in Pakistani practitioners.

Our findings indicate significant skepticism about VR in Pakistani psychiatry, with doubts about its viability and infrastructure. Although 55.5% (n = 111) of respondents were open to learning about VR, only 12.5% (n = 25) felt ready to apply it in practice—highlighting a critical gap between interest and preparedness. Comparatively, the Asia-Pacific study linked positive attitudes to digital psychiatry with theoretical and practical training, underscoring the need for education [12]. An Australian study cited barriers such as cultural resistance, validation time, lack of guidelines, training, and financial constraints [44]. These findings suggest a moderate willingness to explore VR in psychiatric care, with more enthusiasm for learning than active application. There is also a lack of culturally validated research on VR implementation in Pakistani psychiatric settings, which may possibly decrease the confidence in its uptake and efficacy by psychiatrists. Addressing these barriers and expanding training opportunities could pave the way for broader acceptance and integration of VR in mental health care.

Challenges

The study identified several challenges hindering the adoption of digital psychiatry in Pakistan. Key barriers included the “lack of infrastructure/resources” (44.5%), followed by “lack of education/awareness.” Other concerns included “unwillingness to adapt,” “complexity,” and “confidentiality” issues. These findings highlight the pressing need for infrastructure development and comprehensive education initiatives to support the transition to digital psychiatry in Pakistan.

A parallel study on digital health interventions in Pakistan echoed similar barriers, including limited internet access and phone ownership, network coverage, absence of regulatory frameworks, data security issues, affordability, and reliance on paper-based health records [47]. Both studies emphasize shared systemic obstacles and the urgent need for improved infrastructure, regulatory frameworks, and awareness campaigns to enhance digital health integration in Pakistan’s mental healthcare landscape.

Conclusion and implications

This study provides critical insights into the readiness of psychiatrists in Pakistan to embrace digital psychiatry. While there is growing interest in digital tools such as telepsychiatry, AI, mHealth applications, and VR, actual familiarity and usage remain limited. Bridging this gap will require targeted training, infrastructure development, and inclusive policy frameworks.

At the same time, certain limitations should be noted. The study was limited to public psychiatric hospitals, excluding private and military institutions, which may affect generalizability. The reliance on self-reported questionnaires may introduce recall bias, and the use of English alone could have restricted accessibility for some respondents. Future research should expand to private and military facilities, increase sample size, and consider bilingual surveys to foster inclusivity and enhance data reliability.

Despite these limitations, the findings have important implications for practice and policy. Public education campaigns that highlight the safety, benefits, and cultural relevance of digital mental health tools are recommended, alongside implementation strategies that incorporate local beliefs and sensitivities to improve community acceptance. With tailored, context-sensitive strategies, digital psychiatry holds significant promise for expanding access to care and transforming mental health service delivery in Pakistan.