def extract_focus_patients(self, patients_data):
with_dm_or_htn = [
p for p in patients_data
if (p["Diabetes"] == "true" or p["Hipertensión"] == "true")
]
with_dm_or_htn_dir = INTERMEDIATES_DIR / self.community
with_dm_or_htn_file = with_dm_or_htn_dir / HAS_DM_OR_HTN_FILENAME
ezcsv.write_dicts(with_dm_or_htn,
with_dm_or_htn_file,
mkdir=True,
silent_fail=True)
print("Out of {} total patients, found {} with DM or HTN.".format(
len(patients_data), len(with_dm_or_htn)))
print(
"Writing list of these patients to {}".format(with_dm_or_htn_file))
return with_dm_or_htn
def extract_col_consults(self, focus_pts):
focus_pts_cesids = set(p["CesID"] for p in focus_pts)
consults_data = ezcsv.read_dicts(INPUT_DIR / self.community /
CONSULTS_CSV)
consults_for_focus_pts = [
c for c in consults_data if c["CESid"] in focus_pts_cesids
]
print()
print("Found {} consults for those patients.".format(
len(consults_for_focus_pts)))
consults_for_focus_pts_file = (INTERMEDIATES_DIR / self.community /
CONSULTS_FOR_FOCUS_PTS_FILENAME)
print("Writing list of these consults to {}".format(
consults_for_focus_pts_file))
ezcsv.write_dicts(
consults_for_focus_pts,
consults_for_focus_pts_file,
mkdir=True,
silent_fail=True,
)
with_hdl_data = []
for c in consults_for_focus_pts:
hdl = c["HDL"] if c["HDL"] != "" else _extract_hdl_from_note(c)
col = (c["Colesterol"]
if c["Colesterol"] != "" else _extract_col_from_note(c))
if hdl: # check merely that it's not None
c["HDL"] = hdl
c["Colesterol"] = col
with_hdl_data.append(c)
with_hdl_data_file = INTERMEDIATES_DIR / self.community / WITH_HDL_DATA_FILENAME
print()
print("Found {} consults mentioning 'hdl'.".format(len(with_hdl_data)))
print(
"Of these, extracted HDL for {} of them, and Total Cholesterol for {} of them."
.format(
len([1 for l in with_hdl_data if l["HDL"] != ""]),
len([1 for l in with_hdl_data if l["Colesterol"] != ""]),
))
print("Writing these consults to {}".format(with_hdl_data_file))
ezcsv.write_dicts(with_hdl_data,
with_hdl_data_file,
mkdir=True,
silent_fail=True)
return with_hdl_data
def extract_output_fields(self, patients, consults):
output_data = []
patient_by_cesid = {p["CesID"]: p for p in patients}
for c in consults:
patient = patient_by_cesid[c["CESid"]]
output_line = {f: c[f] for f in CONSULT_OUTPUT_FIELDS}
output_line.update({f: patient[f] for f in PATIENT_OUTPUT_FIELDS})
output_data.append(output_line)
pruned_output_data = self.extract_last_records_per_patient(output_data)
output_data_file = OUTPUT_DIR / self.community / OUTPUT_FILE
print()
print("Writing output data to {}".format(output_data_file))
ezcsv.write_dicts(pruned_output_data,
output_data_file,
mkdir=True,
silent_fail=True)
return pruned_output_data
def extract_missing_data_patients(self, patients, consults):
cesids_with_col_data = set([c["CESid"] for c in consults])
no_col_data = [
p for p in patients if p["CesID"] not in cesids_with_col_data
]
print()
print(
"Found {} patients with DM or HTN but no Cholesterol data".format(
len(no_col_data)))
(OUTPUT_DIR / self.community).mkdir(parents=True, exist_ok=True)
stats_file = OUTPUT_DIR / self.community / OUTPUT_STATS
with open(stats_file, "w", encoding="utf-8") as stats:
stats.write(
"Patients with DM or HTN but no cholesterol data: {}\n".format(
len(no_col_data)))
no_col_data_file = OUTPUT_DIR / self.community / OUTPUT_NO_COL_DATA_FILE
print("Writing this list of patients to {}".format(no_col_data_file))
ezcsv.write_dicts(no_col_data,
no_col_data_file,
mkdir=True,
silent_fail=True)
return no_col_data
def do_refs(censo_parto, raw_refs_data, communities):
print()
print("Computing refs matches")
with_parto_and_refs_matches = match(censo_parto, raw_refs_data, OUTPUT,
REFS)
refs_community_matches = get_by_community(with_parto_and_refs_matches,
raw_refs_data, REFS, communities)
censo_parto_refs = with_parto_and_refs_matches + refs_community_matches
all_keys = sorted(
list(
set((list(censo_parto[0].keys()) +
[PREFIXES[REFS] + m for m in raw_refs_data[0].keys()] +
list(ADDITIONAL_COLUMNS.keys())))))
output = []
for l in censo_parto_refs:
output.append(dict(ADDITIONAL_COLUMNS, **l))
ezcsv.write_dicts(output, OUTPUT_PATH, mkdir=True, fieldnames=all_keys)
def do_partos(raw_censo_data, raw_partos_data, communities):
print("Computing parto matches")
with_parto_matches = match(raw_censo_data, raw_partos_data, CENSO, PARTOS)
parto_community_matches = get_by_community(with_parto_matches,
raw_partos_data, PARTOS,
communities)
censo_parto = with_parto_matches + parto_community_matches
all_keys = sorted(
list(
set(([PREFIXES[CENSO] + l for l in raw_censo_data[0].keys()] +
[PREFIXES[PARTOS] + m for m in raw_partos_data[0].keys()] +
list(ADDITIONAL_COLUMNS.keys())))))
ezcsv.write_dicts(censo_parto,
CENSO_WITH_PARTO_LINES,
mkdir=True,
fieldnames=all_keys)
return censo_parto
def main(community):
COMMUNITY = community
input_file = OUTPUT_DIR / COMMUNITY / ASCVD_INPUT_FILENAME
input_data = ezcsv.read_dicts(input_file)
output_data = []
for p in input_data:
try:
ascvd_10yr = ascvd.compute_ten_year_score(
isMale=(p["Sexo"] == "1"),
isBlack=False,
smoker=False,
hypertensive=(p["Hipertensión"] == "true"),
diabetic=(p["Diabetes"] == "true"),
age=(datetime.now().year - int(p["FN_Ano"])),
systolicBloodPressure=int(p["PA Sistólica"]),
totalCholesterol=int(p["Colesterol"]),
hdl=int(p["HDL"]),
)
p["ASCVD 10 year"] = ascvd_10yr
except ValueError: # expected for patients who are missing some data
pass
output_data.append(p)
output_data_file = OUTPUT_DIR / COMMUNITY / OUTPUT_FILENAME
ezcsv.write_dicts(output_data, output_data_file)