Ethical implications of artificial intelligence integration in nursing practice in arab countries: literature review

This systematic review has adopted the systematic identification, selection, and synthesis of all literature related to ethical issues regarding the integration of AI into nursing within Arab countries. The focus is very evident on the presentation of ethical implications, cultural considerations, and regulatory landscaping important to understand the challenges those technologies are uniquely presenting. This review was carried out using the PRISMA guidelines to ensure that the review is clear and transparent. The guidelines introduce the frame of conducting systematic reviews and meta-analyses, emphasizing the necessity of a comprehensive literature search strategy and well-defined research questions with detailed reporting [51, 52], hence reducing bias and enhancing the reproducibility of the results.

A systematic search strategy for retrieving all relevant studies from the year 2010 to April 2024 has been developed for this paper through databases such as PubMed, IEEE Xplore, Web of Science, Scopus, and the Journal of Medical Ethics-AI in Healthcare. The retrieved studies included empirical quantitative and qualitative studies, systematic reviews, meta-analyses, and theoretical articles that described the ethical implications of AI in nursing.

The inclusion criteria were empirical studies, reviews, theoretical articles, and professional guidelines related to the ethics of AI in healthcare. Searches combined terms such as “artificial intelligence,” “nursing,” “ethics,” and “Arab countries” using Boolean operators; filters were applied for studies in both English and Arabic.

The exclusion criteria were a non-peer-reviewed article, a study unrelated to nursing, and a study that did not address ethical issues. Grey literature, editorials, commentaries, and opinion pieces were excluded. The ethical considerations presented in the review were based on an operational framework covering some of the key dimensions: patient privacy, data security, accountability, transparency in AI decision-making, fairness, informed consent, and cultural sensitivity. This systematic approach thus ensured a comprehensive and relevant selection of literature on ethical dimensions of integrating AI into nursing practice, with a key focus on Arab countries. The selected articles provided a broad view of challenges and opportunities for the ethical adoption of AI in healthcare.

CINAHL was not included in the search strategy to have an increased focus on databases like PubMed, Scopus, and Web of Science offering broader, interdisciplinary coverages that are necessary for catching the many different aspects of AI research in health, technology, and ethics. It is important to mention that although CINAHL is specific to nursing and allied health literature, these selected databases would provide greater relevance in terms of reviewing studies with a broader perspective, which is required for the comprehensive analysis of ethical implications concerning the integration of AI into nursing practice.

To ensure a rigorous selection process, the titles and abstracts were screened for relevant studies by two independent reviewers. The above operational framework was consistently applied in the training of these reviewers. Full-text screening was done for articles that, from the title and abstract, appeared to fulfill the inclusion criteria to confirm explicit discussions on one or more of the identified ethical dimensions. Disagreements between the two reviewers were resolved through a structured discussion process. When no consensus was reached, a third reviewer was consulted for a final decision. Finally, 50 articles were selected based on their fulfillment of the inclusion criteria.

The quality of the included studies was appraised using appropriate appraisal tools, such as the Joanna Briggs Institute Checklist for Analytical Cross-sectional Studies, for quantitative research [53]. All methodological items have been evaluated in this review, applying the adapted Newcastle Ottawa Checklist for cross-sectional studies, covering the following fundamental attributes: well-defined research question, appropriate study design given the research question, participant selection, sampling method, measuring exposure and outcome, and analyzing and interpreting the data. Each of these factors is scored as “yes,” “no,” or “unclear,” yielding a summary judgment about the quality of the study.

Scores on this checklist can also be summed to provide a quantitative summary of the quality assessment: each “yes” is usually given one point, and “no” and “unclear” receive zero points. The overall score is then expressed as a percentage of the highest possible score and is used to determine studies with higher or lower quality based on predetermined cut-offs. It is taken into consideration that not all the papers cited in this review were cross-sectional studies. The JBI Checklist was applied to the quantitative studies; other types of publications discussed, such as qualitative studies, systematic reviews, and theoretical articles, were assessed using different quality assessment tools that were specific to their methodologies. The JBI Checklist for Qualitative Research was used to appraise qualitative studies, and PRISMA guidelines for systematic reviews and meta-analyses were applied. Subsequently, all the papers were independently reviewed by two reviewers, and discrepancies were resolved by consensus or, if necessary, consultation with a third reviewer.

Data from included studies were extracted using a standardized form, capturing study characteristics, AI application details, ethical issues discussed, and main findings. A thematic synthesis of the ethical considerations was performed and discussed in the literature. For quantitative data, where available, a narrative summary was provided due to the anticipated heterogeneity of the studies. The synthesis aimed to filter out the primary ethical concerns and recommendations pertinent to AI in nursing practices across Arab countries.

Ethical problems in this review, coming especially from Floridi et al. [2], Kumar and Upadhyay [5], Alahmed et al. [10], Polok et al. [12], Rony et al. [13], Atalla et al. [21], ElHassan, & Arabi [36], Al-kfairy et al. [43], Ghaly [44], Bayan [48], and Ibrahim et al. [49] related to biasing nature of the lack of transparency and accountability within the AI system. As AI continues to evolve, there is an increasing need to generate ethical guidelines that will stipulate the use of AI in health facilities to prioritize patient autonomy and safety, and hence create an enabling environment where ethical considerations start from the design process to the application of AI technologies.

The studies of Floridi [2], Solaiman et al. [3], Lu et al. [4], Uraif [16], Bendary & Rajadurai [17], Rahman [24], Alsufyani [25], Al-Samarraie [41], Al-kfairy [43], El-Gazar [45] emphasized that the implementation of AI in health care is thwarted by deficiencies in the regulations. The complexity of AI requires that legal systems be all-encompassing and sensitive to culture. Arab policymakers must engage the professional community of health in designing regulations mindful of the specific ethical and cultural contexts in which AI will be used in nursing. In such a way, valid guidelines on the safe and responsible integration of AI into healthcare could also emerge.

The potential loss of patient autonomy due to AI decision-making, highlighted by Lu et al. [4], Jabarulla, & Lee [35] raised significant ethical concerns. Developing AI systems that enhance, rather than diminish, patient involvement in their care is crucial. Incorporating patient perspectives into AI development processes can promote respect for autonomy and ensure that technology empowers patients rather than replaces their decision-making capabilities.

As Lu et al. [4], Abduljaber [7], Hussein Mohamed et al. [11], Aladwani et al. [15], Serbaya et al. [25], demonstrated the positive impact of educational programs reinforces the importance of continuous education in nursing. By equipping nurses with knowledge and skills related to AI, healthcare institutions can adopt a culture of informed practice where ethical considerations are at the forefront. Ongoing training can also facilitate discussions around the ethical use of AI, helping nurses navigate the complexities of patient care in a technologically advanced environment.

As Chen et al. [9], Rahman [24], Almarzooqi [26], Albahri et al. [47], noted, the emphasis on trust highlights the need for transparency and accountability in AI applications. Trust is a fundamental component of the nurse-patient relationship and is crucial for successfully integrating AI into clinical practice. Ensuring that AI systems are designed with ethical considerations and that their functions are transparent can enhance trust among healthcare professionals and patients alike.

The privacy concerns identified by Rony et al. [13], Sarabdeen & Moonesar [29], illustrated the delicate balance between leveraging AI for improved healthcare delivery and protecting patient confidentiality. Clear policies and practices must be established to safeguard sensitive patient information. This can include implementing stringent data protection measures and ensuring that nurses are trained in ethical data management practices.

Cultural influences also play a very important role in the ethical implications of AI in the health care system. Most Arab societies are family-oriented, with most decisions made by the family itself, which complicates the issue when AI systems are designed to make autonomous decisions in healthcare. For instance, in the context of palliative care, when family members have often been integrated into critical decision-making processes, AI systems that may give preference to medical data above family discussions might be opposed. Any development of AI technologies should therefore be sensitive to the cultural role of the family in decision-making processes, even as such systems are aligned with ethical principles of patient autonomy [54]. It is also important to consider religious convictions, especially of Islamic origin, towards the reception of AI in health. The doctrine of “tawhid”-one single deity-and the concept of “maslahah” regarding public interest provide a framework through which AI might be ethically advanced in healthcare in a manner that does not confront religious dogma. Integration of AI systems through these cultural and religious considerations contributes to better acceptability and ethical stand [55].

Globally, AI adoption in healthcare faces challenges such as privacy, transparency, and bias. Western countries, like those in the EU, have frameworks like the General Data Protection Regulation (GDPR) to address these issues [56]. However, Arab countries face additional challenges due to cultural and religious values influencing privacy and decision-making, which require region-specific regulations [57]. The UAE’s AI Strategy reflects efforts to integrate AI while respecting cultural values [58]. This highlights the need for tailored approaches in Arab countries to align AI adoption with both global standards and local contexts.

Based on this review’s findings, future research should explore several key areas. Longitudinal studies could assess the long-term impact of AI on nursing practice and patient outcomes, particularly shifts in awareness, attitudes, and ethical considerations. Interventional studies can evaluate educational programs that enhance nurses’ knowledge and attitudes toward AI, offering insights into effective training practices. Emphasizing cultural sensitivity in AI development is essential to address the unique needs of diverse populations in Arab countries, ensuring equitable healthcare. Investigating frameworks that incorporate patient feedback into AI tools could enhance patient engagement and autonomy. Analyzing current regulatory frameworks will help address AI’s challenges in healthcare, guiding policy reforms. Developing culturally sensitive AI systems, using methodologies like Design Science Research (DSR), ensures technical and ethical integration. Participatory Action Research (PAR) involving healthcare professionals, patients, and families can align AI technologies with societal values, addressing ethical concerns effectively.

Ethical integration of AI into nursing improves the care provided for patients but calls for nurses to acquire competencies in AI technologies, overcome possible biases, and act in defense of patients’ rights. Continuous education is necessary to ensure responsible and effective use of AI by nurses. AI developers should consider system design aspects that ensure respect for a patient’s autonomy, privacy, and cultural diversity. In all these, the developers should work hand in glove with health professionals to ensure data security and remove bias. The said approach calls for comprehensive policy-level strategies balancing AI developments against the protection of patients’ rights. Policies should entail standards for data protection, ensuring equity in access and supporting health professionals to adapt to new models of care driven by AI. Organizations like the ANA and ICN emphasize ethical AI use, patient safety, and informed consent. AI in end-of-life care should respect and not replace human judgment, remaining respectful and attentive to the dignity and preference of the patient. Nursing education needs to adapt to include AI training, thus preparing future nurses for a digital healthcare landscape. The curricula have to provide education both in technical aspects of AI as well as on ethical decision making and be related to lifelong learning from current practitioners in that area, harmonizing benefits of AI for the patients and also social values within WHO and AI Act EU.

This study is limited to English publications in databases such as PubMed, Scopus, and Web of Science, potentially introducing language bias and excluding relevant non-English studies. The focus on Arab countries may limit the generalizability of findings to other regions. Selection bias could arise from excluding studies not directly focused on nursing ethics, potentially omitting important perspectives. Variability in the methodological quality, sample size, and rigor of included studies affects the reliability of conclusions. The rapid evolution of AI technology may render some findings quickly outdated. Many studies rely on self-reported data, which can introduce bias as participants may overestimate their AI knowledge. Ethical issues in AI are complex and multifaceted, making it difficult to capture all implications in a limited review. The lack of longitudinal studies assessing the long-term impact of AI integration on nursing practice and patient outcomes restricts deeper insights into evolving attitudes and practices, underscoring the need for continuous research to address these gaps.