Energy Efficiency and Usability of Web-Based Personal Health Records. In: Rocha Á, Adeli H, Reis LP, Costanzo S, Orovic I, Moreira F, editors. Trends and Innovations in Information Systems and Technologies. Cham: Springer International Publishing; 2020. pp. 25-35. [DOI: 10.1007/978-3-030-45697-9_3]

Investigating the Impact of Usability on Energy Efficiency of Web-based Personal Health Records

Usability plays an important role in eHealth applications for their widespread adoption. These software systems have been studied in depth in the literature from this perspective. However, the energy consumption of information systems in the eHealth domain lacks comprehensive research. In this paper, the relationship between the energy consumption of the main components of a PC and the usability evaluations of graphical user interfaces (GUI) were studied in client applications. For this purpose, personal health records (PHR) were used as a case study. A set of 4 web-based PHRs were evaluated with the performance of 20 common tasks in the aforementioned systems. The usability evaluations were carried out by 4 experts. A total of 8 heuristics particularly designed for electronic health record systems and based on the Nielsen usability heuristics were employed. The instantaneous energy values of the power supply and the monitor were collected with an energy consumption measuring equipment. As a result, the following significant correlations were found. In NoMoreClipboard between the Memory heuristic and the power supply (r_s = 0.495, p = 0.044), and the Minimalist heuristic and the monitor (r_s = - 0.513, p = 0.035). In PatientsLikeMe between the Minimalist heuristic and the power supply (r_s = - 0.479, p = 0.083), as well as in the Error heuristic and the power supply (r_s = - 0.638, p = 0.014). Finally, in Health Companion between the Match heuristic and the power supply (r_s = - 0.481, p = 0.043), and the Error heuristic and the power supply again (r_s = - 0.602, p = 0.008). Apart from that, the tasks were divided into those with excellent and fair usability. Those tasks with high usability scores were evaluated to detect whether they also had low energy consumption during task performance. Significant differences were found in PatientsLikeMe with respect to the Error heuristic and the power supply (212.39 ± 18.85W). Significant differences were also found in Health Companion with respect to the Match heuristic and the power supply (199.91 ± 7.19W). The results were discussed to reveal the fundamentals of how implementing usability can impact on energy costs when running client applications.