Access and Authentication

Drupal’s innate authentication system relying on a username and password is used keeping in mind concerns regarding data safety and patient privacy. We plan to have an internal database of patient usernames against patient hospital unique IDs which will be maintained separately from the system. Administrators will be responsible for creating users and assigning them roles and privileges. No personal identifying information will be collected and retained in the system including patient name, age, gender etc.

Preliminary User Experience Testing

We determined that the key user of the system was the patient and we could expect that they would have a diverse range of literacy and proficiency with electronic devices. It was unlikely that a system could be designed such that users who had never used computers and smartphones could use the system without some help. However, from the patient's perspective, it was important to reduce and eliminate roadblocks on the system so that they required minimum assistance. While dedicated testing of the system would need real patients, we proceeded with an initial round of user experience testing such that some data is available as a benchmark for the planned validation studies in future. Physicians and other health care workers would usually be more proficient with the use of electronic devices but would need specific training in the site so that they can use it to maximum benefit.
To get a preliminary idea of the user experience of the website in the patient role, we invited laypersons to do a series of tasks on the website pretending to be a patient and then provide us with feedback on the tasks performed. Social media contacts of authors were invited to participate in this testing. These were a mix of laypersons as well as physicians. We requested all participants to ask their parents and other senior family members to do the tasks if possible. The key tasks to be performed included:
  1. Navigation to the website
  2. Logging in to the website
  3. Checking if changing the language was easy and working.
  4. Find the available questionnaire
  5. Fill the questionnaire
  6. Fill a user experience questionnaire.
Three dummy patient accounts were created (one for each language - English, Hindi and Bengali) and assigned to the cervical cancer disease group. A document with instructions for the user experience testing was prepared in all three languages. 
A dummy questionnaire was made available (a small subset of the European Organization for Research and Treatment in Cancer - EORTC quality of life (QLQ C30) questionnaire) and translated into the two Indian languages. Each question was presented in a single page and each question was accompanied with an audio prompt that allowed users to listen to the question in the language of choice. Time taken to complete the questionnaire was recorded in the system. 
After the questionnaire was completed patients were automatically redirected to a user experience evaluation questionnaire. The questionnaire asked questions on whether there was any difficulty in performing the above tasks and gave the option to the user to give a free text opinion. Additionally, a few demographic questions were asked on patient age, gender, the device used for the task as well as the degree of proficiency with the use of computer or mobile.

Statistical Analysis

Descriptive analysis of the results of the user experience testing is presented. Summaries of continuous variables like time are presented using the median and interquartile range. For categorical variables frequencies are reported. Formal sample size calculations were not done and no group-wise comparisons have been performed.