Skip to main content

Table 3 Summary of main findings for each of the eight overarching human factors principles assessed by modified I-MeDeSA [21, 31, 32]

From: Alerts for community pharmacist-provided medication therapy management: recommendations from a heuristic evaluation

Human Factors Principle (number of associated heuristics on modified I-MeDeSA)

Main Findings a

Principles generally met

 Visibility (3)

Across alert categories, all visibility heuristics (3i,3ii,3iii) were consistently assessed affirmatively (i.e., alerts were rated as distinguishable from the background, having appropriate color contrast, and font.)

 Color (4)

All heuristics (5i-5iv), except the use of color coding (5i) to indicate specific MTP categories (e.g., minimal use of colors with clear meanings for each color), were consistently assessed affirmatively.

Principles with Improvement opportunities and/or unable to be assessed

 Alarm philosophy (1)

MTM vendor platforms did not appear to consistently have a catalog of MTPs indicating associated alerts’ priority level and expected consequences if not followed (1i.)

 Prioritization (5)

Across alert categories, colors, shapes, icons, and signal words were sometimes used, but these did not clearly indicate priority (4i-4iv.) For most alert categories, color, when used, was not a redundant cue (4ii.) For patients with multiple alerts, the order of alerts was found to not clearly indicate priority (4v.)

 Text-based information (10)

Heuristics pertaining to the inclusion of text to explain why the alert was shown (6ii) and the appropriateness of language for the end user (6vii) were consistently rated affirmatively. Across all alert categories, however, signal words, if used, were rated as insufficient for indicating priority (6ia.) The need for a clear consequence statement (6iv) and minimization of text (6v) were noted as enhancement opportunities for most alert categories.

 Proximity of task components being displayed (1)

For most alert categories, alerts did not consistently include the information needed to support decision-making within or in close proximity to the alert. (7i)

 Corrective actions (4)

Alerts did not consistently include “intelligent” corrective actions (8ia.) Alert examples could not be consistently assessed on whether the systems monitored and alerted the user to follow through with corrective actions (8ii.) For most alert categories, improvements would be needed to help prevent usability-related errors (8iii.)

 Placement (4)

For many alert categories, the layout of the alert was rated as insufficient for facilitating quick information uptake by the user (2iv.) For most alert categories, alert examples could not be consistently assessed on whether the alerts appeared at appropriate times (2iii.)

  1. aRoman numerals refer to specific heuristics on the modified I-MeDeSA (Additional file 1)