**4. Artificial Intelligence advancements**

A person working in partnership with an information resource is "better" than that same person unassisted. This is the "Fundamental Theorem" of Biomedical Informatics [45]. Computer Science is making rapid progress and affecting all aspects of human activities. Artificial intelligence (AI) is an interesting domain. It utilizes computers and technology to simulate intelligent behavior and critical thinking comparable to a human being. Its applications for patient safety need to be explored.

The increasing availability of data and emerging technologies needs to be best converged and utilized for better healthcare. A conceptual framework for this leading to AI Patient Safety applications is proposed (**Figure 1**). The AI applications are broadly classified into five categories:


*Prevention Strategies for Patient Safety in Hospitals: Methodical Paradigm, Managerial… DOI: http://dx.doi.org/10.5772/intechopen.106836*

**Figure 1.**

*Conceptual framework for AI patient safety applications.*

#### **4.1 AI for proper assessment**

Clinical assessment involves history taking, clinical examination, investigations leading to diagnosis. Diagnostic error as an area of patient safety has had insufficient research in spite of the negative health outcomes it can lead to [1]. Evidence is accumulating that computer-based trigger algorithms may reduce delayed diagnosis and improve diagnostic accuracy [46].

#### *4.1.1 Future progress*

Diagnostic error causation is very complex and typically occur from the convergence of multiple contributing factors [47]. Lack of medical data can lead to inefficient or inappropriate practice [48]. There are opportunities for improvement using Electronic Health Record (EHR) data sources and AI. ML could help to reduce the frequency of diagnostic errors by improving upon limitation factors of clinicians, namely pattern recognition, bias, and limited capacity [49].

The focus should be on the 'Big Three diseases' which account for about threefourths of serious misdiagnosis-related harms, namely vascular events, infections, and cancers [50].

#### **4.2 AI for pertinent treatment**

Selecting right medications, rightly planned surgeries, right counselling are all benefitted by technology, ensuring patient safety.

Computer-generated prescriptions have many benefits, including links to software that highlights risks from drugs or drug-drug combinations [51]. AI can lead to further improvements and alerts for mistakes.

3D printing is enabling precision surgery. Virtual surgical planning using information regarding patient anatomy and medical devices to be used in surgery increase confidence and knowledge before surgery for better outcomes [52]. Further AI applications can be built with real time image processing capabilities which alert surgeons of any deviations from precision surgery.

#### *4.2.1 Future progress*

AI can analysis vast data at lightening fast speed. All EHRs need to be analysed for side-effects of medicines, including correlation with the doses prescribed. Refinements are always needed to minimize side-effects [53].

AI for correct prescriptions should first focus on medications that commonly result in errors (**Table 3**).


#### **Table 3.**

*Medications commonly resulting in errors.*
