{"title":"The Family Heart Foundation™ Flag, Identify, Network, Deliver—Familial Hypercholesterolemia (FIND-FH™) Program and Collaborative Learning Network (CL","authors":"Diane MacDougall MS, George Blike MD","doi":"10.1016/j.jacl.2024.04.019","DOIUrl":null,"url":null,"abstract":"<div><h3>Study Funding</h3><p>Funded in part by AMGEN.</p></div><div><h3>Background/Synopsis</h3><p>The Family Heart Foundation (FHF) developed a machine learning model (MLM), FIND-FH, to flag undiagnosed individuals at high risk for familial hypercholesterolemia (FH). The model utilizes structured electronic health record (EHR) data. Past implementation of FIND-FH in a large health system identified 2167 high risk patients appropriate for outreach; 153 (7%) were clinically assessed, 46 (30%) diagnosed with FH. FHF was not involved in developing the approach to patient outreach or patient facing materials in this initial deployment.</p></div><div><h3>Objective/Purpose</h3><p>To characterize interim progress and performance metrics regarding screening, outreach, and diagnosis of patients identified by FIND-FH at 5 health systems participating in FHF led Collaborative Learning Network (CLN).</p></div><div><h3>Methods</h3><p>The FHF CLN team works with CLN members using implementation and quality science tools including expert interviews, patient journey mapping, current state process mapping and rapid cycle tests of change. Patient facing materials are jointly developed by FHF in conjunction with FH patients and CLN members. Performance metrics include: #Identified as High Risk of FH; #Appropriate for Outreach/Assessment; #Completed Assessment; #New Diagnosis Definite/Probable/Possible FH; #Needing Other CV Risk Reduction Intervention(s).</p></div><div><h3>Results</h3><p>Currently ∼1.85M EHRs have been screened by FIND-FH, identifying 3,720 at high FH risk. To date, chart reviews completed on 1278 found 628/1278 (49%) unlikely to have FH. The remaining 650/1278 (51%) were deemed appropriate for outreach/assessment. Of 91/650 (14%) patients assessed thus far, 71/91 (78%) were diagnosed as definite/probable/possible FH. Multiple patients not diagnosed with FH, had conditions requiring intervention to lower cardiovascular risk. Patient facing letters and resources were found to be acceptable to individuals diagnosed with FH.</p></div><div><h3>Conclusions</h3><p>Deployment of FIND-FH through a CLN provides proof-of-concept of the ability of an implementation science framework to improve the diagnosis and care for patients with FH. Preliminary performance metrics are promising, yet difficult to directly compare to prior efforts. Health systems used targeted chart review to avoid outreach and assessment of patients “not likely to have FH.” Patient facing materials developed in conjunction with FH patients and made available to all CLN members prevented duplication of efforts at individual health systems. Insights gained from the CLN are informing the development of more efficient, effective, scalable and sustainable care delivery systems for “FIND”ing individuals living with FH.</p></div>","PeriodicalId":15392,"journal":{"name":"Journal of clinical lipidology","volume":"18 4","pages":"Pages e496-e497"},"PeriodicalIF":3.6000,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of clinical lipidology","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1933287424000667","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"PHARMACOLOGY & PHARMACY","Score":null,"Total":0}
引用次数: 0
Abstract
Study Funding
Funded in part by AMGEN.
Background/Synopsis
The Family Heart Foundation (FHF) developed a machine learning model (MLM), FIND-FH, to flag undiagnosed individuals at high risk for familial hypercholesterolemia (FH). The model utilizes structured electronic health record (EHR) data. Past implementation of FIND-FH in a large health system identified 2167 high risk patients appropriate for outreach; 153 (7%) were clinically assessed, 46 (30%) diagnosed with FH. FHF was not involved in developing the approach to patient outreach or patient facing materials in this initial deployment.
Objective/Purpose
To characterize interim progress and performance metrics regarding screening, outreach, and diagnosis of patients identified by FIND-FH at 5 health systems participating in FHF led Collaborative Learning Network (CLN).
Methods
The FHF CLN team works with CLN members using implementation and quality science tools including expert interviews, patient journey mapping, current state process mapping and rapid cycle tests of change. Patient facing materials are jointly developed by FHF in conjunction with FH patients and CLN members. Performance metrics include: #Identified as High Risk of FH; #Appropriate for Outreach/Assessment; #Completed Assessment; #New Diagnosis Definite/Probable/Possible FH; #Needing Other CV Risk Reduction Intervention(s).
Results
Currently ∼1.85M EHRs have been screened by FIND-FH, identifying 3,720 at high FH risk. To date, chart reviews completed on 1278 found 628/1278 (49%) unlikely to have FH. The remaining 650/1278 (51%) were deemed appropriate for outreach/assessment. Of 91/650 (14%) patients assessed thus far, 71/91 (78%) were diagnosed as definite/probable/possible FH. Multiple patients not diagnosed with FH, had conditions requiring intervention to lower cardiovascular risk. Patient facing letters and resources were found to be acceptable to individuals diagnosed with FH.
Conclusions
Deployment of FIND-FH through a CLN provides proof-of-concept of the ability of an implementation science framework to improve the diagnosis and care for patients with FH. Preliminary performance metrics are promising, yet difficult to directly compare to prior efforts. Health systems used targeted chart review to avoid outreach and assessment of patients “not likely to have FH.” Patient facing materials developed in conjunction with FH patients and made available to all CLN members prevented duplication of efforts at individual health systems. Insights gained from the CLN are informing the development of more efficient, effective, scalable and sustainable care delivery systems for “FIND”ing individuals living with FH.
期刊介绍:
Because the scope of clinical lipidology is broad, the topics addressed by the Journal are equally diverse. Typical articles explore lipidology as it is practiced in the treatment setting, recent developments in pharmacological research, reports of treatment and trials, case studies, the impact of lifestyle modification, and similar academic material of interest to the practitioner. While preference is given to material of immediate practical concern, the science that underpins lipidology is forwarded by expert contributors so that evidence-based approaches to reducing cardiovascular and coronary heart disease can be made immediately available to our readers. Sections of the Journal will address pioneering studies and the clinicians who conduct them, case studies, ethical standards and conduct, professional guidance such as ATP and NCEP, editorial commentary, letters from readers, National Lipid Association (NLA) news and upcoming event information, as well as abstracts from the NLA annual scientific sessions and the scientific forums held by its chapters, when appropriate.