Phenotype Data Flow
The diagram below illustrates table dependencies and processing steps used to construct the phenotype cohort.
Any change to SQL logic should be reflected here.
flowchart TD
%% -------------------
%% Source Tables
%% -------------------
A[primary_care_patient_demographics]
B[primary_care_encounter]
C[primary_care_medications]
%% -------------------
%% Step 1
%% -------------------
subgraph Step_1["step_1_clean_diabetes_codes.sql"]
D[eligible_patients]
E[diabetes_events]
end
%% -------------------
%% Step 2
%% -------------------
subgraph Step_2["step_2_earliest_clean_diabetes_code_per_patient.sql"]
F[earliest_diabetes_code]
end
%% -------------------
%% Step 3
%% -------------------
subgraph Step_3["step_3_clean_hba1c_measures.sql"]
G[hba1c_events_raw]
H[hba1c_events_clean]
I[hba1c_events_dedup]
end
%% -------------------
%% Step 4
%% -------------------
subgraph Step_4["step_4_glucose_medication.sql"]
J[insulin_medications_raw]
K[non_insulin_medications_raw]
L[insulin_medications_clean]
M[non_insulin_medications_clean]
end
%% -------------------
%% Step 5
%% -------------------
subgraph Step_5["step_5_earliest_hba1c_measure.sql"]
N[earliest_hba1c_per_patient]
end
%% -------------------
%% Step 6
%% -------------------
subgraph Step_6["step_6_earliest_glucose_medication.sql"]
O[earliest_insulin_per_patient]
P[earliest_non_insulin_per_patient]
end
%% -------------------
%% Step 7
%% -------------------
subgraph Step_7["step_7_diabetes_diagnosis_dates.sql"]
Q[earliest_glucose_med_per_patient]
R[diabetes_diagnosis_dates]
end
%% -------------------
%% Step 8
%% -------------------
subgraph Step_8["step_8_diabetes_code_counts_per_patient.sql"]
S[diabetes_code_counts_per_patient]
end
%% -------------------
%% Data Flow
%% -------------------
A --> D
B --> E
B --> G
G --> H
H --> I
I --> N
C --> J
C --> K
J --> L
K --> M
L --> O
M --> P
I --> R
O --> Q
P --> Q
F --> R
E --> S