Documentation
Platform API
Interview types
Triage

Triage

What is Triage?

Triage is an intelligent workflow designed to assess a patient's required level of care. Triage processes a patient’s medical data, such as their risk factors and symptoms, using scientific evidence and the most up-to-date standards of care to provide clear and meaningful information about the patient’s health. Triage can help create a more accountable and accessible health system by offering an initial evaluation. The results of the evaluation can range from self-care to a specialist or physician consultation to urgent emergencies that require an ambulance. At the end of each survey, patients are guided and equipped with basic knowledge about their most probable conditions, regardless of the evaluated triage level.

Possible triage levels include:

  • self care
  • consultation
  • consultation within 24h
  • emergency
  • emergency ambulance

Key steps described within this workflow (survey):

  1. User creation
  2. Patient creation
  3. Survey creation
    a. Demographics
    b. Risk factors
    c. Initial symptoms
    d. Symptoms assessment
  4. Next steps

How to create a Triage workflow

To create a Triage workflow with Platform API, we need to start by creating a user and a patient, which are the fundamental entities necessary to link a survey to a specific user/patient.

Steps to creating a Triage survey:

1. Creating a user

A user represents an individual who interacts with the Infermedica platform.

Example request:

cURL
curl "https://api.infermedica.com/api/mgp/v1/users" \
  -X "POST" \
  -H "Authorization: Bearer <ACCESS_TOKEN>" \
  -d '{
  "email": "user@example.net",
  "phone": "+498999998999"
}'
ℹ️

The user_id from the response to this request is needed for patient creation.

Available attributes:

  • email - user email
  • phone - user phone number

More information on creating a user can be found here.

2. Creating a patient

A patient is an entity representing a person's medical information within the Infermedica system. Each Triage survey is linked to a patient. The data provided during patient creation directly affects the questions that appear in the survey.

Example request:

cURL
curl "https://api.infermedica.com/api/mgp/v1/patients" \
  -X "POST" \
  -H "Authorization: Bearer <ACCESS_TOKEN>" \
  -d '{
  "sex": "male",
  "age": {
	"unit": "year",
	"value": 20
  },
  "dependent": true,
  "user_id": "c599e21b-3795-4f02-a14e-b46884f8a0c9"
}'
ℹ️

The patient_id from the response to this request is needed for survey creation.

Available attributes:

  • dependent - is patient dependent on the user. The value true means that the survey created for this patient is completed by another person, like a parent or caregiver. For underage patients, dependent must be set to true - if provided initially, the question about dependent will be omitted
  • sex - patient sex - if provided initially, a separate question about sex will be omitted
  • age - patient age - if provided initially, a separate question about age will be omitted
  • user_id (required) - related user ID

More information on creating a patient can be found here.

3. Creating a survey

A survey is the interview process in which medical information about a patient is collected.

Example request:

cURL
curl "https://api.infermedica.com/api/mgp/v1/surveys" \
  -X "POST" \
  -H "Authorization: Bearer <ACCESS_TOKEN>" \
  -d '{
  "sections": [
    "intent_survey",
    "risk_factors",
    "dynamic"
  ],
  "expiration_date": "2024-04-22T11:51:14.198593",
  "type": "triage",
  "patient_id": "9702c05b-5871-4109-ae2a-82f8c9dad64c",
  "evidence": [
    {
      "id": "s_98",
      "choice_id": "present",
      "source": "initial"
    }
  ],
}'

Available attributes:

  • sections - Survey question sections that will be displayed during the interview. If no data is provided, all sections will be visible. If data is provided, the interview will include only the specified sections along with mandatory questions such as sex, age, and dependents (if not provided during patient creation). The assignment of questions to specific sections is specified in the table below.
  • expiration_date - the expiry date of the interview, after which it will not be possible to complete it
  • type (required) - survey type - triage
  • patient_id (required) - related patient ID
  • evidence - survey's initial symptoms

More information on creating a survey can be found here.

Demographics

Basic questions that are the same for all patients. Demographic information is important to starting an interview and adapting questions to the appropriate age and sex groups.

  • Dependent
  • Sex
  • Age
  • Risk factors

Symptoms assessment

This is the dynamic interview that is supported with AI-driven question generation. It starts by collecting the evidence. AI-based algorithms then analyze similar symptoms in Infermedica’s Medical Knowledge Base and check the interactions between them.

  • Symptoms
  • Related symptoms
  • Evidence - answers to AI-generated questions during dynamic interview

Summary of all of the questions within the Triage survey:

NameQuestion idSurvey sectionDescription
Patient - Dependentdependent-Determines if the patient is dependent.
Patient - Sexsex-Determine the demographics of the patient.
Patient - Ageage-Determine the demographics of the patient.
Patient - Risk factorsrisk_factorsrisk_factorsSelect risk factors from those suggested based on demographics.
Intent Survey - kind of care plannedwhat_kind_of_care_you_planningintent_surveyDetermines what kind of care the patient is planning to get
Intent Survey - form of contactform_of_contact_considerationintent_surveyDetermines what form of contact the patient prefers
Symptoms - Travel regionstravel_regionsdynamicRegions visited by the patient recently.
Symptoms - Suggested symptomssuggestdynamicSuggested symptoms, based on selected symptoms.
Symptoms - Dynamic interview questionsinterviewdynamicDynamically generated questions, based on the evidence provided.

4. Next steps

This workflow is designed to support patients in deciding what kind of care they need. Evidence assessment can be used to calculate the most probable conditions. The data collected can be integrated with other systems, like patient accounts or doctor panels. Integration of data can differ depending on the solutions used at a given company and should be developed client-side.