JCEM: EMR-based mixed diagnosis model improves the diagnosis of familial hypercholesterolemia

Background: familial hypercholesterolemia (FH) is the most common genetic lipid disorders, is an autosomal dominant disorder characterized by low-density lipoprotein cholesterol (LDL-C) raised serious is A major public health problem in the United States
.

Although FH can remain asymptomatic for decades, it is estimated to affect more than 25 million people worldwide

An autosomal dominant genetic disease, characterized by a severe increase in low-density lipoprotein cholesterol (LDL-C).

Results: Among the 264 records analyzed, according to the mixed diagnosis model, 794-1571 patients were confirmed to have FH, and the prevalence was 1:300-1:160
.

Compared with the general population (1.

Figure 1 Using EMR data, patients were divided into 5 groups according to different combinations of existing FH diagnostic criteria (modified DLCNS, AHA, and/or genotype)
.

The first group-FH genotype confirmation, if not, confirm that DLCNS≥6, if not, confirm at least one low-density lipoprotein based on a personal or family history of early-onset cardiovascular disease (AHA-2)- C-≥190 mg/dL; The second group-DLCNSAHA-6, if not, then ≥-C≥190 mg/dL more than once, and there is also a personal or family history of early-onset cardiovascular disease (AHA-2) ( AHA-2)

Figure 1 Using EMR data, patients were divided into 5 groups according to different combinations of existing FH diagnostic criteria (modified DLCNS, AHA, and/or genotype)

Figure 2 (A) Venn diagram showing the total number and the interaction between different features; (B, C, and D) Stacked Venn diagram showing various mixtures using existing FH diagnostic criteria (modified DLCNS , AHA) The incremental detection rate of FH and the proportional contribution of each feature to the final composition of each research group; (E) A table showing the total number of each variable (highlighted in yellow)
.

The DLCNS 5 (adopted) feature is not used to contribute to the total count of group 1, because it does not add more content than ha -1 and ha -2

Figure 2 (A) Venn diagram showing the total number and the interaction between different features; (B, C, and D) Stacked Venn diagram showing various mixtures using existing FH diagnostic criteria (modified DLCNS , AHA) The incremental detection rate of FH and the proportional contribution of each feature to the final composition of each research group; (E) A table showing the total number of each variable (highlighted in yellow)

Figure 3 According to various diagnostic criteria, FH phenotypic characteristics and treatment characteristics (using the data in Supplementary Table 3) (24)
.

In each panel, the X-axis represents the research group and the Y-axis represents the research variable

Figure 3 According to various diagnostic criteria, FH phenotypic characteristics and treatment characteristics (using the data in Supplementary Table 3) (24)

Figure 4 Interaction between FH genotype and phenotypic characteristics
.

*Cascade: Individuals who initially did not meet the clinical indications of genetic testing, but were tested based on family members who were FH mutation-positive (positive or VUS)

Figure 4 Interaction between FH genotype and phenotypic characteristics
.
*Cascade: Individuals who initially did not meet the clinical indications of genetic testing, but were tested based on family members who were FH mutation-positive (positive or VUS)
.
†AHA-1: FH diagnosed with AHA standard LDL-C≥190 mg/dL at least once
.
‡Ha -2: FH LDL-C ≥190 mg/dL diagnosed using AHA criteria at least twice
.
§LDL-C≥190 mg/dL, which does not meet ha-2, ha-1 or DCLNS standards
.
||Unique mutation: The only mutation not reported in the ExAC dataset (https://gnomad.
broadinstitute.
org/)
.

Figure 5 Treatment characteristics of the group without comorbidities (using the data in Supplementary Table 4)
.
(24) In each group, the X-axis represents the research group, and the Y-axis represents the research variable
.

Figure 5 Treatment characteristics of the group without comorbidities (using the data in Supplementary Table 4)
.
(24) In each group, the X-axis represents the research group, and the Y-axis represents the research variable
.

Conclusion: Compared with the standard diagnosis alone, using the mixed model, we found that FH has a higher clinical and genetic detection rate
.
The use of statins and other LLTs is sub-optimal and lower than the guideline recommendations
.
Insufficient diagnosis and treatment of FH is the difficulty in implementing the existing diagnostic criteria in the primary care setting.
This mixed model can potentially improve the diagnosis of FH and provide appropriate treatment at an early stage
.

Compared with the standard diagnosis alone, using the mixed model, we found that FH has a higher clinical and genetic detection rate
.
The use of statins and other LLTs is sub-optimal and lower than the guideline recommendations
.
Insufficient diagnosis and treatment of FH is the difficulty in implementing the existing diagnostic criteria in the primary care setting.
This mixed model can potentially improve the diagnosis of FH and provide appropriate treatment at an early stage
.

Original source:

Eid WE, Sapp EH, Wendt A, Improving Familial Hypercholesterolemia Diagnosis Using an EMR-based Hybrid Diagnostic Model .
J Clin Endocrinol Metab 2021 Dec 06

Improving Familial Hypercholesterolemia Diagnosis Using an EMR- based Hybrid Diagnostic Model in this message