JMIR Publications

ABSTRACT

Background:

When caring for patients with chronic conditions like Chronic Obstructive Pulmonary Disease (COPD), healthcare professionals (HCPs) rely on data from a range of sources to make decisions. Collating and visualising this data, for example on clinical dashboards, holds potential to support timely and informed decision-making. Most studies about data supported decision-making (DSDM) technologies for healthcare have focused on their technical feasibility or quantitative effectiveness. While these studies are an important contribution to the literature, they do not further our limited understanding of how HCPs engage with these technologies and how they can be designed to support specific contexts of use. To progress our knowledge of this area, we must work with HCPs to explore this space and the real-world complexities of healthcare work and service structures.

Objective:

This research aimed to qualitatively explore how DSDM technologies could support HCPs in their decision-making about COPD care. We achieved this through creating a scenario-based research tool, called Respire, that visualises HCPs’ data needs about their COPD patients and services. We used Respire with HCPs to uncover rich and nuanced findings about human-data interaction in this context, focusing on the real-world challenges that HCPs face when carrying out their work and making decisions.

Methods:

We engaged nine respiratory HCPs from two collaborating healthcare organisations to design Respire. We then used Respire as a tool to investigate human-data interaction in the context of decision-making about COPD care. The study had three stages and spanned a two-year period. The first stage involved five workshops with the HCPs to identify data-interaction scenarios which would support their work. Based on these insights, the second stage involved creating Respire, an interactive scenario-based web app that visualised HCPs’ data needs with feedback the HCPs. The final stage involved 11 one-to-one sessions with HCPs to use Respire, focusing on how they envisaged it could support their work and decisions about care.

Results:

We found that: (1) HCPs trust data differently depending on its source and author; (2) ‘sporadic’ and ‘subjective’ patient-generated data can have important effects on HCPs’ decision-making; and (3) accountability concerns can impact how, and if, HCPs engage with certain types of data.

Conclusions:

Our study uncovers important lessons for the design of DSDM technologies to support healthcare contexts. We show that while DSDM technologies have valuable potential to support patient care and healthcare delivery, there are important sociotechnical and human data interaction challenges which influence how these technologies should be designed and deployed in practice. Exploring these considerations during the design process can ensure DSDM technologies are designed with a holistic view of how decision-making and engagement with data occurs in healthcare contexts.