Search

The /search endpoint helps find medical concepts (symptoms, risk factors, and conditions) that match a given phrase. In contrast to the /parse endpoint, the /search endpoint can handle short word fragments, even those just a few letters long, and is suitable for building autocomplete-like input forms. Its main function is to convert user-entered phrases into specific medical concept IDs recognized by our API.

Apart from the common attributes name and common_name, /search also has access to a database that has two types of synonyms – ones used by the medical community and ones used by regular users. This enhances its capability to correctly recognize a broader range of user-entered phrases.

ℹ️

Unlike the /parse endpoint, /search supports all languages available in the Infermedica API.

Request

The /search endpoint responds to GET requests and requires two query parameters:

phrase – the phrase to be matched against medical concepts.
age.value – the patient's age, which should be a positive integer between 0 and 130.

An optional age.unit query parameter can also be added to change the default unit from year to month. For example, age.value=5&age.unit=month indicates an age of 5 months. If age.unit is omitted, the default unit of year will be used.

cURL

curl "https://api.infermedica.com/v3/search?phrase=bell&age.value=30" \
  -X "GET" \
  -H "App-Id: XXXXXXXX" \
  -H "App-Key: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX" \
  -H "Interview-Id: d083e76f-3c29-44aa-8893-587f3691c0c5"

Filtering by sex

An optional sex parameter can be used to filter the search results for concepts applicable to a selected biological sex. The allowed values are female or male.

For example, use /v3/search?phrase=lacta&age.value=30&sex=female to show only those medical concepts relevant for one biological sex.

Filtering by concept type

By default, the /search endpoint only returns observations (symptoms and risk factors). You can use the optional types parameter to specify a comma-separated list of medical concept types to be included in the search results. The supported types are condition, symptom, and risk_factor.

For example, to search for both symptoms and conditions, you could use /v3/search?phrase=influ&age.value=30&types=symptom,condition.

Filtering out professional observations

An optional include_pro parameter can be used to include or exclude professional observations from the search results. Professional observations refer to symptoms or risk factors that are difficult for individuals without medical training to self-assess accurately as they require specific knowledge or diagnostic tools to identify. These complex observations are not used in patient-facing interviews and they do not have questions assigned to them.

By default, the include_pro parameter is set to true, which means professional observations are included in the search results. To exclude these observations and only return concepts suitable for patient self-assessment, the include_pro parameter can be set to false.

For example, to exclude professional observations from the search results, use /v3/search?phrase=sign&age.value=30&include_pro=false.

Response

The response is a JSON list of objects, each containing two attributes:

id – the ID of the medical concept,
label – the actual value that matched the phrase. This can be a value of either the name or common_name attributes, or one of their synonyms.

JSON

[
	{
		"id": "s_13",
		"label": "belly pain"
	},
	{
		"id": "s_309",
		"label": "belly feels full"
	},
	{
		"id": "s_1840",
		"label": "belly pain since morning"
	},
	{
		"id": "s_1783",
		"label": "moderate belly ache"
	},
	{
		"id": "s_1852",
		"label": "belly pain from a few days"
	},
	{
		"id": "s_1782",
		"label": "my belly hurts a little"
	},
	{
		"id": "s_1842",
		"label": "belly pain from weeks"
	},
	{
		"id": "s_1860",
		"label": "belly cramps"
	}
]

Medical Concepts NLP