Open Lab›#30DayMapChallenge 2026›Day 18 · NULL — The registry knows where, not what
Web map & dashboard Day 18 · NULL © CC BY 4.0 18 November 2026

Day 18 · NULL — The registry knows where, not what

Of 2,962 hospital, clinic and centre records, 70.6% have none of seven detail fields filled (bed count, phone, opening hours, emergency flag, operator, specialties, address). Only 0.1% have a bed count and 3.0% a phone number. Only 3.7% have three or more fields.

▦ Open in the dashboard ⛶ Open full screen Share on LinkedIn

Write-up

The question

How much does the national health-facility registry actually know about the facilities it lists, and where is it emptiest?

Data

BD Hospital Dataset registry (Complete_Dataset.csv, synthetic rows removed), BBS Census 2022 districts. NULL means not recorded, not absent.

Methods

  1. Records: hospital, clinic and centre rows of the BD Hospital Dataset registry (Complete_Dataset.csv); synthetic records removed.
  2. Seven fields tested for 'filled': beds greater than zero, phone, opening hours, emergency flag, operator, specialties, street or address.
  3. Each record assigned to a district through its upazila polygon.
  4. Completeness = average share of the seven fields filled; 'empty' = none filled.
  5. Records per 100,000 use BBS Census 2022 district populations.

Findings

The registry lists 2,962 hospitals, clinics and centres across 64 districts.
70.6% have none of seven detail fields filled; only 3.7% have three or more.
Fields filled nationally: Street address 23.5%, Emergency flag 4.8%, Operator 3.9%, Specialties 3.9%, Opening hours 3.2%, Phone 3.0%, Bed count 0.1%.
The most complete district (5+ records) is Narsingdi (29%); the emptiest is Brahmanbaria (0%).
Missing detail is not a district problem alone: the gaps follow who tagged what, and urban places are fuller.

Comments & reviews

What you write appears here once it has been approved.

More from the Open Lab