visit_data_figshare.xlsx (5.35 MB)
Download file

Visit-level blood pressure and prescription data at a private Ugandan hospital

Download (5.35 MB) This item is shared privately
modified on 2022-04-12, 18:54


This dataset represents a complete yearlong set of visit-level data capturing blood pressure measurements, diagnosis, and treatment at a private Ugandan hospital. Patient IDs, Visit IDs, and Visit Dates have been completely anonymized. Each row in the dataset represents a visit-prescription pair - so should a patient be prescribed multiple medications at a visit, that visit (identified by unique Visit ID) will occupy one row for each medication prescribed. These can be easily aggregated by grouping by Visit ID after data ingestion. 

Field dictionary: 

  • visit_id_anonymous: string, unique visit ID
  • visit_date_anonymized_day: integer, unique integer representing day of visit - these are sequentially correct but cannot be mapped back to the real date of visit
  • patient_id_anon: string, unique patient ID
  • visit_type: string, string description of visit type
  • is_gp: boolean, whether visit is performed by general practitioner/general physician (should only be TRUE in this dataset)
  • is_drug: boolean, whether this line represents a prescription attached to visit
  • is_ht_drug: boolean, whether this line represents an antihypertensive medication
  • prescription_name: string, description of medication prescribed
  • ht_drug_category: string, category of antihypertensive prescribed, if prescribed medication is an antihypertensive
  • gender: string
  • is_htn_dx_code: boolean, whether a hypertension diagnosis code was assigned this visit
  • bp_sys: integer, systolic blood pressure
  • bp_dia: integer, diastolic blood pressure
  • payment_type: string, Cash represents a self-pay patient visit, and Invoicing represents an insured patient visit
  • age: integer, whole number age
  • visit_number: integer, sequential # of visit for this patient - i.e. if this is a patient's 3rd visit during the study period, visit_number = 3 for that visit