HBP 3024 Codebook

In support of Labs for 432

The hbp3024 data

The hbp3024.xlsx file is available for download on the 432 data page.

The (simulated) data in the hbp3024.xlsx file describe 3024 adults living with hypertension (high blood pressure) diagnoses who receive primary care in one of seven practices.

• In each of the seven practices, 432 (different) individuals (who I’ll call subjects in what follows) were sampled at random from all eligible subjects.
• The data are based on real electronic health record (EHR) data, but with some noise added.
  • The practices are named after streets that appear in The Simpsons.

Eligibility Criteria

The data are cross-sectional and describe results from a one-year reporting window. To be eligible for the study, a subject had to meet all of the following criteria:

• have an EHR-documented hypertension diagnosis which applied during the one-year reporting window,
• cared for at one of the seven practices in this study, and by one of the 55 participating providers in this study
• age 25 or older at the start of the one-year reporting period (note that all subjects with ages 80 and higher are listed as age 80 in the data)
• between 1 and 12 primary care office visits in the one-year reporting period
• between 2 and 24 primary care office visits combined across the reporting period and the previous year
• fall into one of two biological sex categories (female or male)
• fall into one of four primary insurance categories, specifically Medicare, Commercial, Medicaid or Uninsured.
• have a most recent systolic BP between 80 and 220 mm Hg and most recent diastolic BP between 40 and 140 mm Hg, where the systolic BP is at least 15 and no more than 130 mm Hg larger than the diastolic BP.

Codebook

Variable	Description
record	unique code for each subject (six digits, first digit is 9, last indicates practice)
practice	primary care practice, of which there are seven in the data
provider	primary care provider (each practice has multiple providers)
age	subject’s age as of the start of the reporting period
race	subject’s race (4 levels: Asian, AA_Black, White, Other)
eth_hisp	is subject of Hispanic/Latino ethnicity? Yes or No
sex	subject’s sex (F or M)
insurance	subject’s primary insurance (Medicare, Commercial, Medicaid, Uninsured)
income	estimated median income of subject’s home neighborhood (via American Community Survey, to nearest $100)
hsgrad	estimated percentage of adults living in the subject’s home neighborhood who have graduated from high school (via American Community Survey, to the nearest tenth of a percent)
tobacco	tobacco use status (Current, Former, or Never)
depr_diag	does subject have depression diagnosis? Yes or No
height	subject’s height in meters, rounded to two decimal places
weight	subject’s weight in kilograms, rounded to one decimal place
ldl	subject’s LDL cholesterol level, in mg/dl
statin	does subject have a current prescription for a statin medication? Yes or No
bp_med	does subject have a current prescription for a blood pressure control medication? Yes or No
sbp	subject’s most recently obtained systolic blood pressure, in mm Hg
dbp	subject’s most recently obtained diastolic blood pressure, in mm Hg
visits_1	subject’s number of visits for primary care in reporting period (one year)
visits_2	subject’s visits for primary care in the past two years
acearb	does subject have a current prescription for an ACE-inhibitor or ARB? Yes or No
betab	does subject have a current prescription for a beta-blocker? Yes or No

Notes on Specific Variables

• The list of medications included in bp_med is: ACE-inhibitor, ARB, Diuretic, Calcium-Channel Blocker, Beta-Blocker, Alpha-1 Blocker, Centrally acting Alpha-2 Agonist, Vasodilator or other antihypertensive agents. A subject with a current prescription for any of these will have a Yes in bp_med.
• For the acearb, betab, bpmed, statin and depr_diag variables, a No response includes all subjects where there’s no evidence in the EHR of meeting the Yes criterion, so that there are no missing values (a missing value is interpreted there as No.)
• For the height, weight and ldl results, implausible values were treated as missing in preparing the data for you.
• The race and eth_hisp values are self-reported, and some subjects refused to answer one or both of the relevant questions.
• The income and hsgrad values are imputed from the subject’s home address, usually at the census block level, but occasionally at the level of the zip code.
  • When a subject’s home address could not be geocoded, these values are noted as missing.
  • Geocoded estimates of income below 6500 are reported as 6500, and estimates above 130000 are reported as 130000.
  • For hsgrad, geocoded estimates below 40 are reported as 40, and estimates above 99.9 are reported as 99.9.