Seasonal Variations of Arrhythmias and Their Impact on Mortality in Cancer Patients with Health Disparities: A Propensity Score Adjusted Machine Learning Analysis of Over 100 Million Hospitalizations Across 3 Years

Abstract

Various cardiovascular diseases and arrhythmias undergo seasonal variations and are known to increase during high in uenza activity season (HIA) 1 , de ned as December through February.Within these periods of increased prevalence of cardiac arrhythmias, there is a direct correlation to the number of related hospitalizations, morbidity, and mortality 2 .Although several studies have demonstrated this association between seasonal variation and cardiac disorders 3 , there are few studies evaluating these outcomes in cancer patients.Cancer patients are at increased risk for arrhythmias in the setting of unique electrophysiologic issues and treatment therapies 4 , such that their outcomes may be optimized through closer monitoring, prevention strategies, and vaccination during HIA.Arrhythmia-related morbidities and mortalities in cancer patients may also vary based on race, income, geography, and even primary malignancy for which our study aimed to analyze and better quantify.Our goal is to better understand the impact of socioeconomic disparities and seasonal variations on arrhythmia prevalence and arrhythmiarelated mortality in cancer patients.Methods

Data Source
The National Inpatient Sample (NIS) is the largest publicly available U.S. all-payer inpatient healthcare administrative dataset spanning approximately 4,500 hospitals in 50 states, and was the data source for this study.The dataset includes demographic, comorbidity, procedural, complication, mortality, length of stay, total cost, and hospital characteristics for each hospitalization.The 2016, 2017, and 2018 NIS datasets were selected for this study as they are among the latest available datasets and the rst to use International Classi cation of Disease-10 (ICD-10) coding and so better re ect current clinical trends in diagnoses, treatments, and outcomes compared to prior years.Study inclusion criteria included all NIS hospitalizations for adults aged 18 years or older during the above index time periods.Per the US Department of Health and Human Services (DHHS) and National Bureau of Economic Research, no review by an Institutional Review Board (IRB) is required for the NIS under the HIPPA Privacy Rule since the NIS is a limited data set (in which 16 direct identi ed speci ed by the Privacy Rule have been removed) 5,6 .This study used de-identi ed data and was conducted according to the ethical principles in the Declaration of Helsinki.

Study Design
To conduct a more comprehensive analysis more broadly and practically applicable within current healthcare systems, the primary analysis consisted of AI-driven Computational Ethics and policy analysis (AiCE) according to its rst empirical (clinical then economic) step then the second ethical-policy step [7][8][9] .
The rst empirical step featured a nationally representative retrospective longitudinal multicenter cohort analysis of inpatient mortality and total cost among all hospitalized adults including by active cancer, arrhythmia, and season (by High In uenza Activity Season [HIA] de ned by the US Centers for Disease Control as December to February) 10 .It additionally utilized Machine Learning-augmented Propensity Score adjusted multivariable regression (ML-PSr) and deep learning (DL) arti cial neural network to assess inpatient mortality and cost.A cost analysis was then conducted using the above clinical results.This empirical step was followed by the nal ethical-policy step in which the above AI-augmented empirical results informed a pluralistic-based global bioethical analysis to optimize equitable care for the above patient populations.

Descriptive and Bivariable Statistical Analysis
Descriptive statistics were performed for the full sample to de ne the arrhythmia among all adult hospitalizations.Bivariable analysis was then performed according to active cancer (yes/no) and HIA (yes/no) across the full 2016-2018 duration and within each year separately (2016, 2017, and 2018).For continuous variables, independent sample t-tests were performed to compare means, and Wilcoxon rank sum tests were performed for medians.For categorical variables, Pearson Chi-square tests or Fisher exact tests were performed to compare proportions as applicable.
Demographics included age, sex, race, income, insurance, urban density, and region.Comorbidities were selected for analysis (and identi ed in the dataset by their ICD-10 codes) based on their clinical and/or statistical signi cance identi ed in prior published studies and current clinical practice.They included cancer status (active, prior, and metastatic), hypertension, diabetes, coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease, cirrhosis, and chronic kidney disease.The 26 primary malignancies investigated included brain and nervous system, head or neck, thyroid, breast, lung, esophagus, stomach, pancreas, liver or bile system, rectum or anus, colon, peritoneum, bone or connective tissue system, hematological malignancies (including Hodgkin lymphoma, non-Hodgkin lymphoma, leukemia, and multiple myeloma), melanoma, non-melanoma skin, uterus, cervix, ovarian, prostate, testes, bladder, and renal.

Regression Statistical Analysis, Machine Learning Analysis, and Model Optimization Overview
The primary outcome was inpatient mortality (yes/no), and the secondary outcomes were arrhythmia (yes/no; de ned as non-sinus rhythm and non-eucardia) and total hospitalization cost (in U.S. dollars [$]).
The regression statistical analysis, machine learning analysis, and model optimization methods have been explained in our previously published study 11 .

Bayesian machine learning-augmented propensity score translational (BAM-PS) statistics
Regression analysis featured the particular technique within BAM-PSr performed for the above NIS dataset 10,[12][13][14] .This technique integrates ML-PSr (Machine Learning-augmented Propensity Score adjusted multivariable regression), in which the traditional statistical methodology of causal inferencebased propensity score analysis is augmented (b) by machine learning capable of handling higher dimensional, more complex, and faster data streams, and then translates its results as informative priors for (c) Bayesian regression.Regression was conducted on the above outcomes and strati ed by active cancer and HIA (and additionally included sub-group analysis by primary malignancy among patients with active cancer).The propensity score (modi ed as a disease risk score) for the likelihood of presenting during HIA was rst created (utilizing the same above variables used in the nal regression model given the double propensity score adjustment method), a balance was con rmed among blocks, and then the propensity score was included in the nal regression models as an adjusted variable 15,16 .

Health Disparity Analysis
Mortality disparity analysis was conducted using ML-PSr to assess possible signi cant disparities independent of clinical confounding.

Cost-Effectiveness and Cost Bene t Analysis
Cost analysis was conducted by calculating the signi cant and independent mortality disparities (if any) from ML-PSr multiplied by the 2022 in ation adjusted statistical value of a human life as utilized by the US federal government for health and public policy 17 .

Computational Ethical and Policy Analysis
The second or ethical-policy step within AiCE was then conducted by integrating the above quantitative analyses with ethical analysis using the pluralistic global bioethical framework of the Personalist Social Contract (PSC) [18][19][20] .The PSC is a novel integration of modern ethics (principally utilitarianism-informed Rawlsian social contract of political liberalism, bounded by Kantian deontology and informed by feminist, Marxist, deconstructionist, and ecological ethics) and classical ethics (principally Thomistic-Aristotelian virtue ethics, articulated by William Carlo's esse/essence revision of Norris Clarke's Strong Thomistic Personalism, a derivative formulation of Thomism as a development of classical Aristotelianism metaphysics) [21][22][23][24][25][26][27] .It uniquely articulates the philosophical foundation and framework of the United Nation's 1948 Universal Declaration of Human Rights, founded on the primary metaphysical principle of human dignity and resultant rights and duties, which has since united the world's diverse belief systems and 193 nations in what has become the dominant modern ethical framework and foundation of international law.
2.9.Quality Control, Result Reporting, and Analytic Software An academic physician-data scientist, biostatistician, and ethicist (DJM) con rmed that the nal analytic models were su ciently supported by the existing literature and related theories.Mean values are reported with standard deviation (SDs).Fully adjusted regression results were reported with 95% con dence intervals (CIs) with statistical signi cance set at a 2-tailed p-value of < 0.05.Statistical analysis was performed with STATA 17.0 MP edition (STATACorp, College Station, TX, USA), and ML and DL analyses were performed with Java 9 (Oracle, Redwood Chores, CA, USA).

ML-PS multivariable regression of mortality disparities
In ML-PSr among patients with active cancer and arrhythmia, the only signi cant mortality disparities across all 3 years were for African Americans who had signi cant greater mortality than Caucasians (OR 1.13, 95%CI 1.03-1.23,p = 0.013), independent of the above socio-economic and clinical confounders otherwise.Based on the related population averaged (predictive margins) increased mortality likelihood multiplied by the number of African Americans and Caucasians with cancer and arrhythmia and the 2022 in ation adjusted statistical value of a human life, this translates into an excess annual mortality of 406 African Americans and $5.16 billion among adult patients with cancer and arrhythmia.

Computational ethical and policy analysis
The above health and economic results then informed the nal or focused computational ethical-policy analysis step of AiCE.The primary material object of this ethical analysis was arrhythmia, the primary context was inpatient healthcare delivery to patients with active cancer and arrhythmia, and the primary formal object or ethical analytic framework is the PSC.Applied to this concrete ethical situation, the formal PSC argument is as follows.(Premise 1) Arrhythmia in active cancer diagnosed inpatient carries a high mortality risk, which may be signi cantly increased during HIA.(Premise 2) Among this patient population, there appears to be signi cant disparities for African Americans who are more likely to die inpatient than their Caucasian counterparts, even when adjusting for clinical severity and other such relevant factors.(Premise 3) Life and equal societal protection are fundamental individual and state rights logically derivative from the human person's dignity and are politically enshrined across the United Nations, multiple other international institutions, and the majority of nations' constitutions and legal statutes.(Premise 4) Respect for dignity at the individual level requires respecting the person's rights to goods (beginning with the primary good of life) necessary for the person to develop through just and stable commitment to the common good and thus the community in reciprocal care for the individual.(Premise 5) Respect for dignity at the communal level requires respecting another cultures as the communal manifestations of their constitutive individuals seeking through justice the common good (as the objective good of the community, entailing the objective good of individual ourishing, and subjectively experienced as the ultimate individual good of self-actualization through justice).(Premise 6) Racial disparities in hospitalization outcomes for patients with arrhythmia and cancer can produce disproportionate morbidity and even related mortality in those clinical and social sub-communities, resulting in disproportionate threat to the preservation of those persons and related cultures, leading to the global human community's impoverishment with the loss or diminishment of those individuals and cultures.(Premise 7) The reduction of such disparities may result in over hundreds of additional patients lived and $5 billion saved annually.(Premise 8) Disparities in the effective and equitable chronic and acute treatment of patients with arrhythmia and cancer undermines respect for the rights of patients and respect for their cultures, which is critical to the wellbeing of societies that encompass all peoples and cultures (Premise 9) (Conclusion) Therefore, clinical, economic, and ethical justi cation supports greater healthcare policy and healthcare system investment reducing disparities in the health and cost burden of arrhythmia particularly in those with active cancer to ensure equitable, value-based healthcare for all peoples regardless of socio-demographics.

Discussion
Our study found that cancer patients experienced increased arrhythmia-related mortality during HIA, with notable difference by primary malignancy.Our ndings are consistent with prior studies describing worsening mortality and occurrence of arrhythmia-related hospitalizations during winter months [28][29][30][31][32][33][34][35] .Lower temperatures and viral illnesses are considered the driving mechanisms for higher rates of mortality related to arrhythmias during HIA.Cooler, drier weather has been linked with exacerbations in atrial and ventricular arrhythmias 36,37 .It is believed that as temperatures fall, heat loss is minimized through peripheral vasoconstriction and redistribution of blood to the core which results in increased cardiac output 38 .Related elevations in metabolic production can also be quanti ed by higher rates of norepinephrine and epinephrine measured in the plasma and urine during wintertime 39 .For cancer patients who undergo radiation, chemotherapy, immunotherapy and may be more sensitive to uctuations in temperature, these processes can become more pronounced.Through rise in sympathetic drive, central blood volume, and left ventricular end diastolic pressure, increases in atrial distension can incline the atria to brillate 38,40 .
Viral illness is often accompanied by uid loss, fever, elevated heart rate and hypoxia which are known to worsen arrhythmias.We must promote the administration of the in uenza vaccines as a routine public health measure during HIA for cancer patients.Such preventative measures hope to reduce the rates of HIA arrhythmia mortality seen in this analysis.
As the eld of cardio-oncology grows, the socioeconomic disparities in care are being further recognized.Amongst patients with arrhythmia and active cancer in HIA, African Americans had signi cantly greater mortality than Caucasians independent of socio-economic and clinical confounders which translated to more than $5.16 billion dollars in cost.Disparities in care are multifactorial and have been speculated to stem from a combination of structural racism, higher prevalence of cardiovascular risk factors, and reduced access to specialty care 41 .The American Heart Association (AHA) Council on Epidemiology and Prevention note that African Americans have worse cardiovascular health and signi cantly higher rates of fatal coronary artery disease compared to their non-Hispanic White counterparts with increased prevalence of hypertension, diabetes, and obesity.A cross-sectional analysis of the Multi-Ethnic Study of Atherosclerosis (MESA) performed by Heckbert et al. 42 detailed the differences by race/ethnicity in the prevalence of clinically detected and monitor-detected atrial brillation and found that the prevalence of clinically detected atrial brillation was substantially lower in African American than in white participants.The lower rates of detection of arrhythmia may re ect differences in symptom perception, clinical atrial brillation recognition, or health care access.These data in conjunction with our ndings may even represent an underrepresentation of the morbidity and mortality burden that arrhythmias may have in the African American patient population.In addition to the cardiovascular disparities experienced by African Americans, there has been a disproportionate cancer burden, including the highest mortality and the lowest survival of any racial/ethnic group for most cancers reported by the American Cancer Society 43 .
African American/Black people have higher mortality than any other broadly de ned racial/ethnic group for most cancers and other leading causes of death, including heart disease, stroke, and diabetes with these disparities being driven by lower socioeconomic status 44 .We hope that our ndings will enable the cardio-oncologic community to recognize existing racial healthcare imbalances and encourage collective steps towards providing equitable care.
The most common cancers with increased mortality and prevalence of arrhythmias during HIA were gastrointestinal (GI), leukemia and lung.These cancers have been previously linked to arrhythmias in the literature 45 , connections which can be attributed to overlapping risk factors, symptomology, chemotherapy and post-surgical complications.In GI cancers, patients frequently experience anemia and electrolyte abnormalities from malabsorption (vomiting, diarrhea, bowel resection).The presence of common cancer risk factors such as obesity, smoking and alcohol can further worsen arrhythmia outcomes.Amongst hematologic cancers, acute leukemia is shown to predispose patients to cardiac disease even before initiation of chemotherapy 46 .Leukemia chemotherapy regimens involve alkylating agents, anthracyclines, tyrosine kinase inhibitors and arsenic trioxide; medications all regularly associated with sinus bradycardia, AV block, atrial brillation, ventricular tachycardia/ brillation and QT interval prolongation [47][48][49] .The increased risk of arrhythmias from leukemia chemotherapy regimens in conjunction with our ndings of worsened mortality in HIA-associated arrhythmias for leukemia is concerning.
Finally, lung cancer is often accompanied by hypoxia, demand ischemia, and is more common in smokers, all clinical factors that exacerbate arrhythmias.Patients who undergo lung resection may experience arrhythmias from direct myopericardial irritation or as a post-operative complication of intrathoracic surgery [50][51][52] .The same can be said for any other major surgeries such as colectomy or esophageal surgery in GI cancers 53 .With respect to all cancers, our study found increased prevalence of arrhythmias during HIA for cancer patients for 2017, 2018, but not 2016.The ML-PS analysis did not nd signi cantly increased odds for arrhythmia during HIA for cancer patients using these ndings, but inclusion of additional years may reveal more precise results.Current literature supports that a diagnosis of cancer and even a history of cancer after active treatment predisposes patients to increased arrhythmias [54][55][56][57] .Our ndings of worsening mortality with HIA-associated arrhythmias in cancer patients should prompt physicians to practice caution in this vulnerable population.

Conclusion
In conclusion, our nationally representative study shows that there is increased arrhythmia-related mortality during high in uenzas activity season in patients with cancer.Cancer presents a unique challenge to the management of cardiac arrhythmias and through investigations of these disparities, we hope to more appropriately address these barriers to treatment and better establish standards of cardiac care in the cardio-oncology patient population.

Clinical Perspectives
According to our ndings, physicians should educate and encourage cancer patients to adopt greater preventative measures during HIA to avoid arrhythmias-related mortality.These include emphasizing the importance of regular in uenza in the vulnerable cancer population.Closer monitoring is recommended in the outpatient and inpatient setting, speci cally for patients with lung, GI and hematologic malignancies who were found to experience higher mortality and prevalence of arrhythmias during HIA.Lastly, recognizing racial disparities within medical outcomes is crucial for taking incremental steps towards equitable patient care.

Limitations
Firstly, we conducted a large population case-control study for which a de nitive case-effect conclusion may not be established.A consequence of using the NIS database is the lack of a comparative matched control group, but this was mitigated through our propensity score adjustment and machine learning statistical approach.Secondary, our ndings pertain mainly to the inpatient setting with lack of long-term follow-up.However, our results lay the groundwork for further outpatient studies to investigate similar themes for HIA associated arrhythmias in cancer.Lastly, we were not able to con rm the increased odds of arrhythmias occurring during HIA seen in the general population, for cancer patients.Inclusion of additional NIS data beyond 2018 may reveal more precise results.

Figures Figure 1
Figures

Table 1 ML
-PS multivariable regression of arrhythmia, mortality, and cost by cancer and HIA* p = 0.003) and non-HIA seasons ($3,474.68,p<0.001).3.4.ML-PS multivariable regression by sub-group analysis among primary malignanciesAmong patients with arrhythmia and active cancer during HIA, the mortality rate increased the most in different primary malignancies in different years as follows: GI in 2016 (OR 1.15, 95%CI 1.01-1.32,p =