1.Introduction

This page intends to describe functionalities in general of the product PatientAid that the project has been working to deliver, in its initial release, i.e. MVP.

2.Services and Users

PatientAid as a product, consists of a set of services that primarily serve two types of users:

1. Individual patient or potential patient, as the patient
2. Healthcare coordinator user, as the coordinator We aim to deliver 2 sets of webpage services for each of them to access.

3.High level Requirements and Scope

3.1.The patient gets help with acute symptoms

Adhering to the core principle of PatientAid to help healthcare providers to save the most lives amid the COVID-19 pandemic, individuals who have similar symptoms or suspect they might have COVID-19 will use PatientAid to get medical help, this includes patients who are in critical conditions that require immediate hospitalisation. After going through the list of acute symptoms if the patient confirms to have one or many of acute symptoms, s/he is directed to complete the session and to be scheduled an ambulance appointment for. Because its severity this is the first step of the user journey once the session starts.

3.2.The patient goes through COVID-19 self-assessment

If the patient doesn't have any acute symptoms, through the self-assessment the patient completes three steps for different questions, which include COVID-19 relevant symptoms, underlying diseases and demographic data that are described as medical metadata. Medical metadata comes from the actual healthcare provider that adopts PatientAid, by an import procedure to load them into PatientAid database at deployment stage.

3.2.1.COVID-19 related symptoms

At this step the patient is displayed with the list of COVID-19 symptoms identified by the healthcare provider. In addition to choose having one or many of them, the patient is also required to indicate severity of each symptom at a scale from 1(mildest) to 5(the most severe).

3.2.2.Underlying diseases

Following the symptoms the patient is presented with the list of multiple choices of underlying diseases that the healthcare provider believes to factor the diagnosis. In similar aside from confirming to have one or many of them, the patient is also required to indicate severity of each underlying disease at a scale from 1(mildest) to 5(the most severe).

3.2.3.Demographic info

In the last step the patient is then presented a few demographic questions which are also from healthcare providers, we anticipate them mostly to be single choice of two questions: age group and gender.

3.3.The patient provides necessary contact information for help to deliver

As indicated by the wireframe the patient then is asked to provide contact information that is mandatory for the help to be delivered by the healthcare provider.

3.4.The patient completes the session and is presented with a triage decision

Within few seconds based on data collected during the assessment steps the patient is informed with one of below triage decisions:

• Continue to self-isolate: The patient is also able to view more content/guidance by clicking the link offered by the healthcare provider;
• To be tested through an appointment: The patient is advised to check email for further notification of the appointment once it's confirmed;
• To be diagnosed via a telemedicine session: The patient is advised to check email for further notification of the appointment once it's confirmed;
• To be taken into the hospital ASAP: The patient is told to wait for further email about the ambulance appointment.

3.5.The healthcare coordinator users make appointments for the patient

The coordinator user works in dedicated groups/roles to make appointments on patient's behalf, this is done by going through different queues of patient assessment records.

3.5.1The coordinator user makes and confirms an appointment for the valid patient record

Provided valid contact information once the coordinator manages to make the appointment and confirms with corresponding service provider, the coordinator is able to update the patient record by filling the appointment details via a pre-canned template and confirms the update.

3.5.2The patient record is invalid

If the coordinator is unable to make the appointment due to erroneous key contact information, i.e. home address is invalid for ambulance, or invalid email address for test or telemedicine appointment, the coordinator indicates the appointment is unsuccessful and opts the reason.

3.6.The patient receives the notification via an email and/or text message

Based on pre-canned templates, the system delivers the notification to the patient via an email and/or text. This covers both successful and unsuccessful cases. Below are two sample templates with parameters in square bracket.

3.7.Contextual info

We want to make our services self-explanatory and comprehensive to the user, a few features help to achieve so.

3.7.1.Display Policy and acquire consents before the patient starts

The full policy document and terms and conditions drafted by our legal team will be displayed via a link clicked by the user, the user is required with consent before the session starts.

3.7.2.Display text explanatory text in webpages that collect patient's data

Similar to the NHS example in below, the patient should be given an explanation of why information in context has been asked from the patient.

3.8.Data import

As PatientAid is designed upon broad range of metadata to be prepared or tailored by the healthcare provider, a facility for a user to be able to load those data into database at deployment stage and later as needed is required.

3.9.Scope

Below epics are identified to form the scope of PaitentAid MVP:

1. Provide the landing page with T&Cs consent
2. The patient to consider acute symptoms and the first step
3. The patient to consider COVID-19 related symptoms
4. The patient to consider underlying diseases
5. The patient to provide demographic info
6. The patient to provide minimal and necessary contact info
7. The patient is presented with a triage result
8. The patient is able to discard the session
9. The coordinator is able to update the patient record with the confirmed appointment
10. The coordinator is able to indicate an invalid patient record
11. The system sends a notification to the patient
12. An admin user is able to upload medical metadata at anytime required

4.Design concerns

4.1.Patient data protection compliance

PatientAid needs to be designed in a way to be as much configurable as possible, for a number of reasons:

• Assessment/triage criteria may vary from individual healthcare providers, or change over time when the outbreak develops, corresponding questions and answers, as well as the model to carry out the assessment process, will need to be customised by each healthcare provider adopter whenever needed.
• Although patient data protection has been relaxed for COVID-19 by several countries, we want to design PaitentAid to collect and keep as little patient data as possible. This needs to demonstrated and proved by both persistence and process of the patient data collected.

Therefore we have been considering below design options:

• Metadata driven' approach: this will make almost all customisable content is designed as metadata, e.g. symptoms, underlying diseases, notification templates etc. See some sample data in the spreadsheet.Metadata sheet
• Process and calculation of collected patient medical data is performed on the frontend without any communication to the backend or to persist. Only till the assessment is complete to work out the triage decision, the result along with patient's minimal contact info is passed on to the backend to persist.