Table 3 Examples of how AI can address IS challenges in health systems and public health settings
IS challenge | AI example |
|---|---|
Speed | NLP was used for qualitative analysis coding and compared to a traditional, non-AI-enhanced approach. The authors found that NLP can identify major themes, but that traditional approaches were best at identifying more nuanced details [33] |
AI-enhanced chatbots were designed to identify and address barriers to chronic medication adherence with messages tailored to patient characteristics and needs [34] | |
Sustainability | Case study showed how AI can improve data analysis and improve the efficiency of clinical processes and, thereby, improve sustainability [35] |
Equity | AI-enhanced chatbots were created to provide culturally relevant education and support for new mothers and to address health disparities [36] |
Demonstrates how AI can be used to identify ongoing clinical trials that historically underrepresented patients are eligible for [37] | |
Generalizability | AI was applied to social media data to identify adverse events and public sentiment associated with immunizations in a large and heterogeneous population [38] |
Assessing context and context-outcome relationships | AI was used to identify contextual reasons for clinician non-adherence to guideline-recommended practices, including identification of previously unrecognized issues [38] |
AI was used to predict coronary artery disease and quantify death risk using electronic health records and genetic data [38] | |
Assessing causality and mechanisms | Describes how AI was used to create actionable and individualized causal treatment effect predictions for patients with Alzheimer’s disease [39] |