Retrospective cohort study of a community-based primary care program's effects on pharmacotherapy quality in low-income Peruvians with type 2 diabetes and hypertension
Data files
Sep 27, 2023 version files 87.26 KB
-
care_period_data_dict_rev_2023-09-26.csv
-
care_period_rev_2023-09-26.csv
-
pre-post_data_dict_rev_2023-09-26.csv
-
pre-post_rev_2023-09-26.csv
-
README.md
Oct 16, 2023 version files 84.26 KB
-
care_period_data_dict_rev_2023-09-26.csv
-
care_period_rev_2023-09-26.csv
-
pre-post_data_dict_rev_2023-09-26.csv
-
pre-post_rev_2023-09-26.csv
-
README.md
Abstract
A door-to-door survey was conducted to enumerate all household members by age and sex in a low-income community in Peru. 856 adults 35 years and older were eligible to participate in screening for type 2 diabetes and hypertension. 709 (83%) participated in screening. 130 (18.3%) were diagnosed with hypertension and/or type 2 diabetes of which 109 (84%) participated at program onset and 22 were added later from earlier non-participants in screening or program onset to form the cohort of 131 patients with diabetes and/or hypertension. The primary care program had components of the Chronic Care Model, community health workers, and freely accessible visits and medications. The program operated between September 2011 and May 2014, and consisted of two care periods (separated by a six-month hiatus), first a 10-month home-care period, then a 17-month clinic-care period. The dataset is two files corresponding to two exposures: the 27-month program overall (post- versus pre-) (N=262 observations, 131 pairs with patients as self-controls) and care period (clinic versus home), N=211 (109 home and 102 clinic observations, >131 because 80 patients participated in both care periods). Exposures were evaluated for their effects on guidelines-based pharmacotherapy standards: hypoglycemic and antihypertensive medications, low-dose aspirin, and first-line angiotensin converting enzyme inhibitor (ACEi) treatment of diabetes with elevated blood pressure.
README: Retrospective cohort study of a community-based primary care program's effects on pharmacotherapy quality in low-income Peruvians with type 2 diabetes and hypertension
There are two data files in text format (comma-delimited). Pre-post....csv contains the 262 observations for the program exposure. Care period....csv contains the 211 care period observations. Each file has a data dictionary, also in comma-delimited non-UTF-8 text format. "Pre-post data dict....csv" describes the variables in program exposure study. "Care period data dict....csv" describes the variables in the care period exposure study.
Usage notes
There are four files (two data and two data dictionaries), all in comma-delimited text format (non-UTF-8) and a README.md file.
Sharing/Access information
There are no other ways to access the data. The data is not derived from any other source.
Code/Software
No coding or scripts are included with the data.
Methods
From 2011 to 2014, data was collected prospectively, during weekly (home visits) or monthly (clinic visits), on paper encounter forms that were entered into Microsoft Excel as part of the standard operation of the community-based program. In January 2020, the University of Arizona institutional review board approved the use of the de-identified data for a study of the program's effects on clinical outcomes. Time-series data (fasting glucose and blood pressure) was collapsed on the median of monthly average fasting glucose and blood pressure values during the program (27 months) and the respective care periods, home (10 months) and clinic (17 months). Antihypertensive and hypoglycemic agents were collapsed on the highest dose ever received, angiotensin-converting enzyme inhibitors (ACEi) and aspirin on whether any dose was ever received, by treatment-eligible groups, and within program and care period time intervals. Retention in care was obtained by counting visits and elapsed months (from first to last patient encounters) during the program and care periods. Treatment-eligible groups were low-dose aspirin candidates (10-year cardiovascular disease (CVD) risk >=10% by the Framingham alternate model that uses clinical factors only, no laboratory factors; blood pressure (BP) treatment candidates (BP >=130/80 mm Hg if diabetic or >=140/90 mm Hg if non-diabetic); hypoglycemic agent candidates (patients with diabetes); and diabetic ACEi candidates (diabetes with BP >=130/80 mm Hg). Data has been transformed into two files corresponding to two exposures: 1) program, post- versus pre- (referent), N=262 observations; and 2) care period, clinic versus home (referent), N=211 observations.
There are two data files in text (comma-delimited) format. Pre-post....csv contains the 262 observations for the program exposure. Care period....csv contains the 211 care period observations. Each file has a data dictionary also in comma-delimited format. "Pre-post data dict....csv" describes the variables in the program exposure study. "Care period data dict....csv" describes the variables in the care period exposure study.
Usage notes
There are four files (two data and two data dictionaries), all in comma-delimited text format and a README.md file.