Lab investigations

Your investigations are what a doctor orders or advises: laboratory tests, scans, panels and packages. Each row is one orderable item, identified by your own id. Use type=lab_investigation with the API, or pick Investigations in the console. An investigation is not the value that comes back. The values a lab reports, such as hemoglobin, are lab results. A named group of results that a doctor fills in, such as CBC, is a lab panel. Complete Blood Count can exist as an investigation (ordered) and as a panel (filled in). These are separate records with their own ids.

Mandatory columns: investigation_id and name. Every other column is optional; leave the cell empty if you don’t have the value.

Columns

Column	Required	What it means	Allowed values	Example
`investigation_id`	Yes	Your own id for the investigation. Writing a row with an id that already exists updates it.	Text, up to 200 characters	`LAB-CBC`
`name`	Yes	The investigation name. Used for search.	Text, up to 200 characters	`Complete Blood Count`
`display_name`	No	The name shown to doctors and patients. Defaults to `name`.	Text, up to 200 characters	`CBC`
`aliases`	No	Other names people search by.	Several values separated by `\|`	`CBC\|Hemogram`
`category`	No	The group the investigation belongs to.	Text, up to 64 characters	`laboratory`
`type`	No	What sort of item this is. Stored for your own use.	Text, up to 64 characters	`test`, `panel`
`specimen`	No	The sample the investigation runs on.	Text, up to 100 characters	`Blood`
`unit`	No	Unit, if you track one. Stored for your own use. A result carries the unit that is served.	Text, up to 32 characters	`%`
`ekaid`	No	Eka’s own id for the same investigation, from Eka’s reference list. You map your id to Eka’s; leave it empty if you don’t have it. See Mapping your records to Eka ids.	Text, up to 64 characters	From Eka’s list
`is_active`	No	Set to `false` to keep the investigation but hide it. Defaults to `true`.	`true` or `false`	`true`

Example file

lab-investigations.csv

investigation_id,name,ekaid,display_name,aliases,category,type,specimen,unit,is_active
LAB-CRP-Q,"C-Reactive Protein (CRP), Serum (Qualitative)",,CRP (Qualitative),CRP,laboratory,test,Blood,,true
LAB-CBC,Complete Blood Count,,CBC,CBC|Hemogram,laboratory,panel,Blood,,true

Put quotes around any value that contains a comma, like "C-Reactive Protein (CRP), Serum (Qualitative)". Spreadsheet programs do this for you when they save as CSV.

The smallest valid file has just the two mandatory columns:

minimal.csv

investigation_id,name
LAB-CBC,Complete Blood Count

Example JSON

The same records sent to Create or update records. In JSON, multi-value fields are lists.

[
  {
    "investigation_id": "LAB-CRP-Q",
    "name": "C-Reactive Protein (CRP), Serum (Qualitative)",
    "display_name": "CRP (Qualitative)",
    "aliases": ["CRP"],
    "category": "laboratory",
    "type": "test",
    "specimen": "Blood",
    "is_active": true
  },
  {
    "investigation_id": "LAB-CBC",
    "name": "Complete Blood Count",
    "display_name": "CBC",
    "aliases": ["CBC", "Hemogram"],
    "category": "laboratory",
    "type": "panel",
    "specimen": "Blood",
    "is_active": true
  }
]

Common mistakes

Putting result values here. Hemoglobin, with its unit, belongs in lab results. Add an investigation only for what is ordered.
Expecting a panel to be filled in from here. Setting type to panel is only a label. To give doctors a CBC to fill in, upload a lab panel.
A wrong ekaid. Eka stores the id you send without checking it. A wrong id can attach your investigation to the wrong Eka item, so check each mapping against Eka’s reference list.
Reusing an id for a different investigation. An existing investigation_id updates that record; it never creates a second one.

Authorization

Webhooks

Encryption

HealthAI

ABDM Connect

Datasets

Doctor Tool

Medical Knowledge-bases

Medical Records

User App

Medical Database

Columns

Example file

Example JSON

Common mistakes

​Columns

​Example file

​Example JSON

​Common mistakes

Columns

Example file

Example JSON

Common mistakes