Article Text
PDF
Abstract
Background Familial hypercholesterolaemia (FH) increases propensity for premature atherosclerotic disease. Knowledge of inpatient outcomes among patients with FH admitted with acute myocardial injury (AMI) is limited.
Objectives Our study aimed to identify myocardial injury types, including type 1 myocardial infarction (MI), type 2 MI and takotsubo cardiomyopathy, assess lesion severity and study adverse short-term inpatient outcomes among patients with FH admitted with AMI.
Setting Our study retrospectively queried the US National Inpatient Sample from 2018 to 2020.
Population Adults admitted with AMI and dichotomised based on the presence of FH.
Study outcomes We evaluated myocardial injury types and complexity of coronary revascularisation. Primary outcome of all-cause mortality and other clinical secondary outcomes were studied.
Results There were 3 711 765 admissions with AMI including 2360 (0.06%) with FH. FH was associated with higher odds of ST-elevation MI (STEMI) (adjusted OR (aOR): 1.62, p<0.001) and non-ST-elevation MI (NSTEMI) (aOR: 1.29, p<0.001) but lower type 2 MI (aOR: 0.39, p<0.001) and takotsubo cardiomyopathy (aOR: 0.36, p=0.004). FH was associated with higher multistent percutaneous coronary interventions (aOR: 2.36, p<0.001), multivessel coronary artery bypass (aOR: 2.65, p<0.001), higher odds of intracardiac thrombus (aOR: 3.28, p=0.038) and mechanical circulatory support (aOR: 1.79, p<0.001). There was 50% reduction in odds of all-cause mortality (aOR: 0.50, p=0.006) and lower odds of mechanical ventilation (aOR: 0.37, p<0.001). There was no difference in rate of ventricular tachycardia, cardioversion, new implantable cardioverter defibrillator implantation, cardiogenic shock and cardiac arrest.
Conclusion Among patients hospitalised with AMI, FH was associated with higher STEMI and NSTEMI, lower type 2 MI and takotsubo cardiomyopathy, higher number of multiple stents and coronary bypasses, and mechanical circulatory support device but was associated with lower all-cause mortality and rate of mechanical ventilation.
- myocardial infarction
- cardiomyopathy
- lipid disorders
- thromboembolism
Data availability statement
No data are available. Data included in this study can be found on the public website of the Healthcare Cost and Utilization Project (HCUP). Information about NIS database can be obtained from the HCUP website (https://hcup-us.ahrq.gov).
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
We used a large weighted and pooled database that is nationally representative of the US hospital population.
The use of multivariate analysis for confounders and the large sample size limited sampling bias.
Administrative databases like the National Inpatient Sample (NIS) are subject to coding errors leading to the risk of misclassification bias.
The NIS has no outpatient data.
The NIS has no details on medications and how they affect observations.
Introduction
Background
Familial hypercholesterolaemia (FH) is a genetic disorder that causes markedly elevated low-density lipoprotein cholesterol (LDL-C), which can lead to early atherosclerotic disease. The global prevalence of FH is estimated to be 1 in 313 in the general population, with higher rates among high-risk groups such as those with premature myocardial injury and LDL-C >190.1 FH can be diagnosed either genetically or clinically, with several genes identified that follow a Mendelian pattern and could be autosomal dominant, such as ApoB receptor, LDL receptor and PCSK9 related, or autosomal recessive, such as LDLRAP1.2 3
Patients with FH have a disproportionately higher burden of morbidity, especially at a younger age. In a French lipid registry study, patients with probable FH who were admitted with acute coronary syndrome (ACS) were 12 years younger than those without FH but had double the rates of 5-year mortality and atherosclerotic complications.4 While long-term adverse outcomes following atherosclerotic events are well established in patients with FH, the impact of recent advances in treatments of atherosclerotic disorders on inpatient mortality during admission for acute myocardial injury is still unclear. There is scarcity of data from the USA on FH among hospitalised patients admitted with ACS. There are currently no data on non-type 1 myocardial infarction (MI) among patients with FH including takotsubo or type 2 MI.
In this study, we hypothesise that patients with FH, despite having higher morbidity, are likely to have lower inpatient mortality rates during admission for acute myocardial injury due to recent advances in treatment of atherosclerotic coronary lesions and novel therapies for hyperlipidaemia and diabetes. To our knowledge, there is currently only one large cohort study on this topic, based on the National Inpatient Sample (NIS) database.5
Objective
Our study aims to expand on this previous research and provide a more comprehensive understanding of the relationship between FH and inpatient mortality rates during admission for acute myocardial injury. We also provide the first data on non-type 1 MI among patients with FH admitted with acute myocardial injury including takotsubo cardiomyopathy and type 2 MI.
Method
Study design
This is a retrospective cohort study that used data from the NIS database from 2018 to 2020. We identified adult patients (age >18 years) with acute myocardial injury, including acute MI, non-ischaemic acute myocardial injury and takotsubo cardiomyopathy, using the International Classification of Diseases, 10th revision (ICD-10) codes. Patients were dichotomised based on the presence of FH. Using the procedure code of the ICD-10, we determined the type and complexity of coronary revascularisation based on the number of stents and grafts. We then determined the odds of all-cause mortality as well as other secondary outcomes among patients with FH and those without FH. We also examined the association between FH and the complexity of coronary revascularisation procedures required for treatment.
Data source
The data source for this study is the NIS. The NIS is a large administrative database provided by the Agency for Healthcare Research and Quality (AHRQ), which contains data from approximately a 20% sample of inpatient hospital admissions in the USA. The NIS includes information on the principal diagnosis, which is the main reason for hospitalisation identified by the primary ICD-10 code, as well as secondary diagnoses recorded during the hospitalisation. This database provides a valuable resource for researchers to examine patterns of care and outcomes in a nationally representative sample of hospitalised patients in the USA.6 7
Patient population and outcomes of study
Our study population included all patients who were hospitalised between 2018 and 2020 and recorded in the NIS database. We focused on patients who had been admitted for acute myocardial injury. We used specific ICD-10 codes to identify these diagnoses, which are listed in online supplemental table 1. We collected and analysed various patient characteristics such as demographics, hospital-level features and relevant medical comorbidities. Our primary outcome measure was all-cause mortality which was defined as any death reported during admission for any acute myocardial injury. We also analysed a range of secondary outcomes including length of stay, total hospital charge, ventricular tachycardia, cardioversion rate, new implantable cardioverter defibrillator (ICD) insertion, intracardiac thrombus, cardiac arrest, cardiogenic shock, mechanical circulatory support device and mechanical ventilation.
Supplemental material
Statistical analysis
We performed statistical analyses using Stata, V.17 (Stata Corp, College Station, Texas, USA) standard edition. A univariate logistic regression analysis was conducted using all available variables and comorbidities to calculate unadjusted ORs for the primary outcome. We included all variables with p values less than 0.1 in a multivariate logistic and linear regression model to control for confounders. To estimate the effect sizes, we used ORs and mean differences. The independent sample t-test was used to compare group differences in continuous variables, while group differences in categorical variables were compared using Pearson’s χ2 analysis and Fisher’s exact test. All tests were double-sided. We considered outcomes with p values less than 0.05 and a 95% CI to be statistically significant. To adjust for the comorbidity burden, we used the Charlson Comorbidity Index. In a secondary analysis, we stratified the sampled population into three levels: admissions with type 1 MI (ST-elevation MI (STEMI) and non-ST-elevation MI (NSTEMI)), admissions with any acute MI including type 1 MI and type 2 MI, and admissions with other acute myocardial injury types except for type 1 MI. We performed a similar analysis for each stratum to investigate any differences in the associations between the variables and outcomes.
Patient and public involvement
This study used secondary, de-identified data from the NIS, which is a publicly available all-payer inpatient database in the USA. This is a part of the Healthcare Cost and Utilization Project which is administered by the AHRQ. No direct human subjects were involved.
Results
The study identified 3 711 765 admissions with acute myocardial injury, out of which 2360, comprising 0.06%, had FH. These 2360 FH cases were further categorised as follows: 695 (29.45%) cases with STEMI, 1340 (56.78%) cases with NSTEMI, 265 (11.23%) cases with type 2 MI, 40 (1.69%) cases with takotsubo cardiomyopathy and 20 (0.85%) cases with non-ischaemic acute myocardial injury.
Baseline characteristics
Details of patients’ baseline characteristics and univariate analysis are presented in table 1. Overall, there was male predominance (58% vs 42%) without significant difference based on FH (42.37% vs 41.84%, p=0.811). Patients with FH were younger by 8 years (p<0.001), were more white (79% vs 72%, p=0.007) and had lower comorbidity index. Patients with FH had higher comorbid coronary artery disease (CAD) (13.98% vs 6.19%, p=0.001), lower obesity (19.28% vs 24.15%, p=0.007), lower chronic kidney disease (CKD) (17.16% vs 31.17%, p<0.001), lower HIV (1.91% vs 4.47%, p=0.007) and lower chronic obstructive pulmonary disease (8.69% vs 17.30%, p<0.001). Other risk factors, including history of stroke, heart failure and diabetes, were not different between patients with and without FH.
Baseline characteristics and univariate analysis among patients admitted with acute myocardial injury
Myocardial injury types and revascularisation complexity
FH was associated with higher odds of STEMI (adjusted OR (aOR): 1.62, p<0.001) and NSTEMI (aOR: 1.29, p=0.008) but lower odds of type 2 MI (aOR: 0.39, p<0.001) and takotsubo cardiomyopathy (aOR: 0.36, p=0.004). There were higher odds of primary percutaneous coronary intervention (PCI) (aOR: 2.15, p<0.001) and coronary artery bypass graft (CABG) (aOR:2.45, p<0.001) among patients with FH. There were higher odds of revascularisation complexity among patients with FH, one-stent PCI (aOR: 1.69, p<0.001) and multistent PCI (aOR: 2.36, p<0.001). Similar observation is made regarding CABG with rate of one-vessel bypass (aOR: 2.22, p<0.001) and multivessel bypasses (aOR: 2.65, p<0.001). There was reduced length of stay by 1 day (p<0.001) without a difference in total charge. Figure 1 below describes a plot of ORs associated with various myocardial injury types and revascularisation pattern and complexities based on comorbid FH.
A plot of acute myocardial injury type and complexity of revascularisation among patients admitted with acute myocardial injury with versus without familial hypercholesterolaemia. Acute myocardial injury: acute myocardial infarction (MI) or takotsubo cardiomyopathy or non-ischaemic acute myocardial injury. ORs are calculated after adjusting for race, sex, age, hospital size, hospital location and teaching status, Charlson Comorbidity index, stroke, heart failure, diabetes mellitus, CKD, obesity and HIV. CABG, coronary artery bypass graft; CKD, chronic kidney disease; NSTEMI, non-ST-elevation MI; PCI, percutaneous coronary intervention; STEMI, ST-elevation MI.
All-cause mortality
FH was associated with lower odds of all-cause mortality (aOR: 0.50, p =0.006). In a three-tier stratified population secondary analysis, odds of all-cause mortality remained low among FH admitted with type 1 MI (3.40% vs 8.03%; aOR: 0.56, p=0.041) or any acute MI (3.85% vs 8.67%; aOR: 0.55, p=0.019) and with a trend towards lower all-cause mortality in those admitted with acute myocardial injury without type 1 MI (6.67% vs 10.20%; aOR: 0.53, p=0.298). Table 2 below shows details of multivariate analysis of various clinical outcomes after adjusting for potential confounders including race, sex, age, hospital size, hospital location and teaching status, Charlson Comorbidity Index, hypertension, stroke, heart failure, diabetes mellitus, CKD, obesity and HIV.
Myocardial injury pattern, complexity of coronary revascularisation and adverse clinical outcomes among patients admitted with acute myocardial injury with or without familial hypercholesterolaemia (FH)
Secondary outcomes
After adjusting for relevant confounders including race, sex, age, hospital size, hospital location and teaching status, Charlson Comorbidity Index, hypertension, stroke, heart failure, diabetes mellitus, CKD, obesity and HIV, there were higher odds of intracardiac thrombus (aOR: 3.28, p=0.038) and a need for mechanical circulatory support device (aOR: 1.79, p<0.001) among patients with FH. On the contrary, there were lower odds of mechanical ventilation for all causes (aOR: 0.37, p<0.001). There was no difference in ventricular tachycardia, cardioversion, need for new ICD implantation, cardiogenic shock and cardiac arrest. Figure 2 below depicts frequencies of various adverse clinical outcomes based on comorbid FH.
Frequencies of adverse clinical outcomes among patients admitted with acute myocardial injury with versus without familial hypercholesterolaemia. Acute myocardial injury: acute myocardial infarction or takotsubo cardiomyopathy or non-ischaemic acute myocardial injury.
Discussion
Baseline characteristics
In our study of patients admitted with acute myocardial injury, we found 0.06% had FH, like findings by Elbadawi et al from a study of patients with acute MI who found a prevalence of 0.07% using the NIS.8 Our study is the second NIS-based study to estimate prevalence of FH among hospitalised patients in the USA. Data on all-comers or patients with ACS with FH are lacking in the USA. Reported prevalence is very varied globally and is dependent on the population, the setting and clinical criteria used. For instance, Schmidt et al estimated that among patients with ACS, prevalence of FH was 0.35%,9 but a similar analysis by De Luca et al stated a prevalence of 3.3%.10 We observed a higher proportion of males which is likely due to the higher prevalence of ischaemic heart disease among men. Patients with FH were 8 years younger but had a higher prevalence of CAD which is similar to findings in most observational studies.8 11
Myocardial injury pattern
Patients with FH had a higher prevalence of STEMI and NSTEMI, consistent with what is widely reported in literature. Prevalence of non-type 1 MI among patients with FH has not been previously reported. It is widely reported that hyperlipidaemia is a risk factor for takotsubo cardiomyopathy.12 13 There are no previous studies that specifically investigated an association of takotsubo cardiomyopathy or type 2 MI with FH. Our study observed a lower prevalence of type 2 MI, takotsubo cardiomyopathy and other forms of acute myocardial injury among patients with FH compared with those without. Our stated rate of takotsubo cardiomyopathy of 1.69% agrees with the incidence rate of takotsubo cardiomyopathy among patients with acute MI in the HORIZONS-AMI trial of 0.5–2.1%.14 The lower comorbidity burden and younger age are likely protective against the occurrence of demand ischaemia and takotsubo among patients with FH.15 16 This favourable outcome of lower odds of non-type 1 MI persisted after accounting for age and comorbidity burden. Studies have reported a better lifestyle among patients with FH, starting from an early age compared with the general population. This could explain a potentially robust cardiometabolic reserve in times of stress compared with the general population.17 It has been hypothesised in many translational studies that there are other factors such as enhanced thrombogenicity that plays a role in the pathogenesis of coronary disease in patients with FH besides elevated LDL.18 This likely makes patients with FH prone to type 1 MI compared with others such as type 2 MI or takotsubo.
Primary and secondary clinical outcomes
Like observations made by Kheiri et al, we also observed increased utilisation of mechanical circulatory support among patients with FH.19 Additionally, we studied the rate of intracardiac thrombus and observed increased odds among those with FH, a problem that was previously not studied. The concept of enhanced platelet activation has been studied and reported to be higher in patients with elevated serum cholesterol. This, together with a higher rate of transmural infarction, may explain these observations.20
We also believe that higher prevalence of STEMI with transmural ischaemia and myocardial stunning is the likely cause of increased mechanical circulatory support device use as well as higher risk of intracardiac thrombosis.
In our study, we observed a paradoxically lower all-cause mortality among patients with FH admitted with acute myocardial injury and this reduction remained significant among patients with type 1 MI after stratification. Our findings agree with the findings of prior similar studies. For instance, in an analysis of a readmission database between 2016 and 2018, Elbadawi et al found that patients with FH admitted with acute MI had a non-significant trend towards lower mortality.8 Similarly, an NIS-based study that looked at outcomes of acute MI between 2016 and 2018 found lower odds of inpatient mortality among patients with FH.5 Several studies have found a negative correlation between serum LDL and short-term mortality after MI.21 22 The relationship between serum LDL and mortality is however not present linearly across all spectra of hypercholesterolaemia. For example, in a Danish population-based study, Johannesen et al found a U-shaped relationship between serum LDL and mortality following acute MI, with mortality reduction occurring between LDL level of 70 mg/dL and 140 mg/dL. They added that these observations applied only to patients who were not on statin therapy.23 Explaining this well-known lipid-ACS paradox has however been challenging. Some authors believe that the negative relationship between moderately elevated serum LDL and mortality may be a simple case of lead-time bias. It is thought that early screening offers patients with FH earlier treatment with statins which are known to improve outcomes during acute MI.24 25 There have been some postulates about the mitigating effect of increased remnant lipoproteins. They are often elevated in patients with FH and their levels have been shown to correlate negatively with mortality and adverse outcomes after MI.26
Additionally, Ravnskov et al, in a systematic review, found no significant difference in mortality between cohorts with and without elevated LDL-C and confirmed a similar U-shaped association as stated above. They reported that this observation was present independent of statin therapy.27 Studies have found a better lifestyle, including higher physical activity and less smoking among patients with FH which may portend an overall improved cardiovascular health. Patients with FH, though at increased risk of coronary atherosclerosis, may have a more robust cardiometabolic reserve to survive acute stress such as MI.17
Finally, ischaemic preconditioning, a concept widely proven to be protective in animal models but without much empirical backing in humans, has been considered as well.28 29
We also found a higher rate of revascularisation among patients with FH. This may be due to more STEMI and high-risk NSTEMI in this group. Evidence from previous studies supports this observation. For instance, in a case series of 45 young patients who underwent coronary angiography, those with FH had more complex CAD compared with those without FH.30 Similarly, a French registry study with 5147 patients found a higher prevalence of triple-vessel CAD among those with FH.31 Another review of the RICO database in 2021 by Yao et al found that definite FH, as determined by a Dutch lipid clinic network score >6, was associated with more complex coronary lesions, including a higher syntax score, more bifurcation lesions and more multivessel disease.32 In addition to the clinical outcomes, it is worth noting the healthcare utilisation among patients with FH. We observed a reduced length of stay in the hospital but no significant difference in total charges. However, a study by Patel et al, based on electronic record reviews, found high annual total revenue among patients with FH.33
Study strengths and limitations
Our study was based on large, pooled data, which increased its power. With FH being an uncommon condition, our study design and large database of the NIS provide a good basis for hypothesis generation. The NIS is generated at the hospital level and weighted to reflect the US population thus a reliable source to make hypotheses about population-wide problems in the inpatient setting. One limitation is that the NIS may be subject to coding errors which puts it at risk of misclassification bias. Using a specific and sensitive endpoint such as all-cause mortality as a primary outcome minimises such risk. Our study includes data from 2018 to 2020 during which the NIS relies on ICD-10 codes that have improved specificity and are universally comparable. Additionally, most diagnoses of FH are made in the outpatient setting, often by experts in their field and often may require further testing including genetic study and thus are less likely to be misclassified.
Another limitation is that the NIS does not include information on outpatient care. An important limitation pertaining to this study is that the NIS does not contain codes for medical therapeutics. Nonetheless, the observations made raise pertinent hypotheses that may need future robust controlled trials to ascertain.
Conclusion and future research
Our study found that patients with FH who were hospitalised with acute myocardial injury had a higher prevalence of STEMI and NSTEMI, but a lower prevalence of type 2 MI and takotsubo cardiomyopathy. Additionally, they required a higher number of multiple stents and bypasses and had greater need for mechanical circulatory support devices. FH was associated with lower all-cause mortality and mechanical ventilation rates, and there were no significant differences in cardiac arrest, cardiogenic shock, ventricular tachycardia, cardioversion or new ICD implantation. Moving forward, longitudinal studies are needed to further investigate the long-term outcomes, including major adverse cardiovascular events, among patients with FH. These studies should consider the evolving diagnostic testing, clinical criteria and novel therapies available for FH. We also look forward to reviewing this concept using a database in which the impact of statin therapy on outcomes in patients with FH can be assessed.
Data availability statement
No data are available. Data included in this study can be found on the public website of the Healthcare Cost and Utilization Project (HCUP). Information about NIS database can be obtained from the HCUP website (https://hcup-us.ahrq.gov).
Ethics statements
Patient consent for publication
Ethics approval
It is important to note that all patient data in the NIS are de-identified and the data are publicly available. Therefore, we did not seek Institutional Review Board approval for this study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors DK—project administration, supervision, software, visualisation, data curation, formal analysis, writing original draft, review and editing, and guarantor. JTN—data curation, formal analysis, investigation, validation, methodology, resources, review and editing. SMO—investigation, validation, methodology, review and editing. AO—investigation, validation, methodology and resources. RG—validation, review and editing. EG-T—investigation, validation, methodology, review and editing. SK—methodology, review and editing. AA—methodology, review and editing. SF—formal analysis, investigation, validation, methodology, resources, review and editing. TA—supervision, validation, review and editing.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.