A population approach using cholesterol imputation to identify adults with high cardiovascular risk

A report from AHRQ's EvidenceNow initiative

Samuel Cykert, Darren A. DeWalt, Bryan J. Weiner, Michael Pignone, Jason Fine, Jung In Kim

Research output: Contribution to journalArticle

Abstract

Objective: Large practice networks have access to EHR data that can be used to drive important improvements in population health. However, missing data often limit improvement efforts. Our goal was to determine the proportion of patients in a cohort of small primary care practices who lacked cholesterol data to calculate ASCVD risk scores and then gauge the extent that imputation can accurately identify individuals already at high risk. 219 practices enrolled. Patients between the ages of 40 and 79 years qualified for risk calculation. For patients who lacked cholesterol data, we measured the effect of employing a conservative estimation strategy using a total cholesterol of 170 mg/dl and HDL-cholesterol of 50 mg/dl in the ASCVD risk equation to identify patients with 10%, 10-year ASCVD risk who were eligible for risk reduction interventions then compared this to a rigorous formal imputation methodology. 345 440 patients, average age 58 years, qualified for risk scores. 108 515 patients were missing cholesterol information. Using the good value estimation methodology, 40 565 had risk scores 10% compared to 43 205 using formal imputation. However, the latter strategy yielded a lower specificity and higher false positive rate. Estimates using either strategy achieved ASCVD risk stratification quickly and accurately identified high risk patients who could benefit from intervention.

Original languageEnglish (US)
Pages (from-to)155-158
Number of pages4
JournalJournal of the American Medical Informatics Association
Volume26
Issue number2
DOIs
StatePublished - Jan 1 2019

Fingerprint

Cholesterol
Population
Risk Reduction Behavior
HDL Cholesterol
Primary Health Care
Health

Keywords

  • Cardiovascular disease
  • Informatics
  • Registry
  • Risk scores

ASJC Scopus subject areas

  • Health Informatics

Cite this

A population approach using cholesterol imputation to identify adults with high cardiovascular risk : A report from AHRQ's EvidenceNow initiative. / Cykert, Samuel; DeWalt, Darren A.; Weiner, Bryan J.; Pignone, Michael; Fine, Jason; Kim, Jung In.

In: Journal of the American Medical Informatics Association, Vol. 26, No. 2, 01.01.2019, p. 155-158.

Research output: Contribution to journalArticle

@article{297fefeffbeb4382a46e6fdc1a8f7045,
title = "A population approach using cholesterol imputation to identify adults with high cardiovascular risk: A report from AHRQ's EvidenceNow initiative",
abstract = "Objective: Large practice networks have access to EHR data that can be used to drive important improvements in population health. However, missing data often limit improvement efforts. Our goal was to determine the proportion of patients in a cohort of small primary care practices who lacked cholesterol data to calculate ASCVD risk scores and then gauge the extent that imputation can accurately identify individuals already at high risk. 219 practices enrolled. Patients between the ages of 40 and 79 years qualified for risk calculation. For patients who lacked cholesterol data, we measured the effect of employing a conservative estimation strategy using a total cholesterol of 170 mg/dl and HDL-cholesterol of 50 mg/dl in the ASCVD risk equation to identify patients with 10{\%}, 10-year ASCVD risk who were eligible for risk reduction interventions then compared this to a rigorous formal imputation methodology. 345 440 patients, average age 58 years, qualified for risk scores. 108 515 patients were missing cholesterol information. Using the good value estimation methodology, 40 565 had risk scores 10{\%} compared to 43 205 using formal imputation. However, the latter strategy yielded a lower specificity and higher false positive rate. Estimates using either strategy achieved ASCVD risk stratification quickly and accurately identified high risk patients who could benefit from intervention.",
keywords = "Cardiovascular disease, Informatics, Registry, Risk scores",
author = "Samuel Cykert and DeWalt, {Darren A.} and Weiner, {Bryan J.} and Michael Pignone and Jason Fine and Kim, {Jung In}",
year = "2019",
month = "1",
day = "1",
doi = "10.1093/jamia/ocy151",
language = "English (US)",
volume = "26",
pages = "155--158",
journal = "Journal of the American Medical Informatics Association : JAMIA",
issn = "1067-5027",
publisher = "Oxford University Press",
number = "2",

}

TY - JOUR

T1 - A population approach using cholesterol imputation to identify adults with high cardiovascular risk

T2 - A report from AHRQ's EvidenceNow initiative

AU - Cykert, Samuel

AU - DeWalt, Darren A.

AU - Weiner, Bryan J.

AU - Pignone, Michael

AU - Fine, Jason

AU - Kim, Jung In

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Objective: Large practice networks have access to EHR data that can be used to drive important improvements in population health. However, missing data often limit improvement efforts. Our goal was to determine the proportion of patients in a cohort of small primary care practices who lacked cholesterol data to calculate ASCVD risk scores and then gauge the extent that imputation can accurately identify individuals already at high risk. 219 practices enrolled. Patients between the ages of 40 and 79 years qualified for risk calculation. For patients who lacked cholesterol data, we measured the effect of employing a conservative estimation strategy using a total cholesterol of 170 mg/dl and HDL-cholesterol of 50 mg/dl in the ASCVD risk equation to identify patients with 10%, 10-year ASCVD risk who were eligible for risk reduction interventions then compared this to a rigorous formal imputation methodology. 345 440 patients, average age 58 years, qualified for risk scores. 108 515 patients were missing cholesterol information. Using the good value estimation methodology, 40 565 had risk scores 10% compared to 43 205 using formal imputation. However, the latter strategy yielded a lower specificity and higher false positive rate. Estimates using either strategy achieved ASCVD risk stratification quickly and accurately identified high risk patients who could benefit from intervention.

AB - Objective: Large practice networks have access to EHR data that can be used to drive important improvements in population health. However, missing data often limit improvement efforts. Our goal was to determine the proportion of patients in a cohort of small primary care practices who lacked cholesterol data to calculate ASCVD risk scores and then gauge the extent that imputation can accurately identify individuals already at high risk. 219 practices enrolled. Patients between the ages of 40 and 79 years qualified for risk calculation. For patients who lacked cholesterol data, we measured the effect of employing a conservative estimation strategy using a total cholesterol of 170 mg/dl and HDL-cholesterol of 50 mg/dl in the ASCVD risk equation to identify patients with 10%, 10-year ASCVD risk who were eligible for risk reduction interventions then compared this to a rigorous formal imputation methodology. 345 440 patients, average age 58 years, qualified for risk scores. 108 515 patients were missing cholesterol information. Using the good value estimation methodology, 40 565 had risk scores 10% compared to 43 205 using formal imputation. However, the latter strategy yielded a lower specificity and higher false positive rate. Estimates using either strategy achieved ASCVD risk stratification quickly and accurately identified high risk patients who could benefit from intervention.

KW - Cardiovascular disease

KW - Informatics

KW - Registry

KW - Risk scores

UR - http://www.scopus.com/inward/record.url?scp=85060379097&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85060379097&partnerID=8YFLogxK

U2 - 10.1093/jamia/ocy151

DO - 10.1093/jamia/ocy151

M3 - Article

VL - 26

SP - 155

EP - 158

JO - Journal of the American Medical Informatics Association : JAMIA

JF - Journal of the American Medical Informatics Association : JAMIA

SN - 1067-5027

IS - 2

ER -