The Opportunities and Challenges of Data Analytics in Healthcare

(USA) Healthcare faces unique obstacles that set it aside from other sectors and that make it difficult for the benefits of AI to be realised. These include:


- Most healthcare organisations have no strategies for data governance or analytics.


- The nature of healthcare decisions. (They concern sensitive information, require timely information and action, and can have life or death consequences. Additionally, sometimes patients refuse what is considered the clinically best medical decision and so systems need to keep these preferences in mind. This all means that there is a very high standard for any data-analytics tool to meet, before patients or physicians consider consenting to their use or deployment.)


- Problematic data conventions. (Healthcare data is difficult to access: they are split across various entities and in different formats; and they are often not available to researchers or providers. Healthcare has resisted making the information open and widely accessible for fears of privacy violation and because markets typically undersupply public goods. There is a need not only for accessible data, but good quality data that is representative.)


- Institutionalised practices in care delivery. (Data tool designers are often ignorant of the healthcare context, particularly how decisions are made by several people and following institutional guidelines. Uptake has also been slow because of the workflow disruption it causes and data entry burden it has placed on physicians.)


- The misaligned incentives of various actors in the industry. (Claims data are held by the insurer while clinical data are held by care providers, meaning not only that the two agents fail to make use of the other’s data, but also that neither agent, nor the system, performs optimally for the patient. Both agents also have differing incentives and needs, meaning that patient treatment and pursuit of data analytics will take on different guises and directions. Patients also have privacy concerns, and a fear of data sharing with insurers that might impact their claim or policy, as such, they oppose data sharing arrangements that might benefit healthcare for all.)


The solution suggested here is that government policy should emphasise interoperability of health data and financially incentivise providers to develop data analytics capabilities.



#Health #HealthCare #Data #Analytics #Care #AI

©2019 by AITHICS