Study of Association Between Different Coronary Artery Disease Presentations and Its Effect on Short-Term Mortality, Readmission, and Cost in Patients Undergoing Percutaneous Coronary Interventions

Introduction Atherosclerotic coronary artery disease (CAD) is the major cause of mortality in the USA. CAD requiring percutaneous coronary intervention (PCI) can have a wide spectrum of presentations. We reviewed the cost of admission and PCI at the tertiary care center stratified for different CAD presentation types. Methods We performed a retrospective study of 7,389 patients undergoing coronary angiogram at our facility from 2015 to 2017. Patients were selected from CathPCI registry. Chart review was done for readmission and death data. Cost data were provided by the finance department. Patients going for coronary artery bypass surgery (CABG) were excluded. We split the patients based on their need for PCI. Cost analysis was based on CAD presentation types (No symptoms, atypical symptoms, stable angina, unstable angina, NSTEMI [non-ST segment elevation myocardial infarction], STEMI [ST-segment elevation myocardial infarction]). Adjusted linear regression was run for the outcomes. Primary outcomes were 30-day readmission and death. The secondary outcome was cost of admission. Results The final sample size was 6,403. The mean age was 65.6 years (SD: 12.5; male: 63.8%). 2444 required PCI (38%; p < 0.001). PCI group had lower mean age (62.5 years; SD: 12.3, p<0.001) with lower BMI (30.6 vs 31.1, p=0.015). PCI group had significantly lower odds for 30-day readmission (OR: 0.63; CI: 0.45-0.89; p=0.009) and 30-day mortality (OR:0.60; CI: 0.41-0.89; p = 0.011). A severe presentation increased the odds of getting PCI. Cost of admission was higher in all groups receiving PCI. Conclusions PCI group had better 30-day readmission and mortality rates. PCI increases the cost of admission in all CAD types.


Introduction
Atherosclerotic coronary artery disease (CAD) is major cause of death in United States [1]. Percutaneous coronary artery intervention (PCI) remains the most common treatment modality among the treatments of coronary artery disease. More than 600,000 PCIs are performed in the United States each year, the cost associated with these procedures is estimated at around $12 billion [2]. Although the mortality associated with the procedure is low, the readmission rate after the procedure is indicative of the quality of care and determines the cost of care [3]. The main determinants of cost-effectiveness are age and class of angina. PCI is most successful in the case of left ventricular disease and in younger patients [4].
For patients with stable angina COURAGE trial compared medical therapy with PCI and found the cost to be higher for the intervention arm [5]. However, a subsequent analysis done for stable angina patients has emphasized that the acceptable cost threshold for improvement in symptoms is difficult to achieve [6]. In patients with unstable angina and non-ST elevation myocardial infarction (NSTEMI) studies comparing the cost-effectiveness of early intervention versus conservative management revealed increased cost associated with the intervention [7]. In STEMI patients the cost-effectiveness of the PCI is reported in terms of length of stay (LOS), 30-day readmission, and mortality associated with the procedures, and these outcomes depend on the location of the lesion. Studies have shown that early discharge (3 days) after PCI is associated with decreased 30-day readmission rate and cost, especially in non-anterior wall myocardial infarction [8].
Based on varying outcomes in different CAD presentation types, it is difficult to calibrate the factors affecting the cost of admission. In this study, we looked at the admission cost associated with different types of coronary artery disease presentations with and without PCI.
The abstract from this article has been published in the National Meeting of Cardiology. SCAI 2019. https://onlinelibrary.wiley.com/doi/10.1002/ccd.28864.

Patient population and study design
We performed a retrospective analysis of 7,389 patients who had undergone coronary angiogram at our facility, between January 2015 and December 2017. Institutional Review Board approval was obtained from the office of Human Research at the University of Illinois Chicago at Peoria, IL. Initial patient number was 7,389. We excluded patients with incomplete records, those with missing labs and patients who required CABG. The final sample size was 6,403. We used retrospective data from chart review, and every patient had an electrocardiogram (ECG) available before the procedure. Clinical variables evaluated are listed in Table-1. We split the patients in two categories based on whether the patient received PCI or not. Additionally, we divided patients based on their CAD presentation type at the time of the coronary angiogram. The CAD presentation classes included No symptoms, Atypical symptoms, Stable angina, Unstable angina, NSTEMI, and STEMI.

Outcomes
Primary outcomes were 30-day readmission and mortality. The secondary outcome was cost of admission.

Statistical analysis
Continuous data were reported as mean ± standard deviation (SD) and categorical data as proportions. T-test was utilized to compare continuous variables and the Chi-square test was used for categorical variables. Adjusted statistical analyses were conducted to compare clinical variables. 30-day readmission, 30-day mortality, and admission cost required a multivariate analysis. We used Stata software, v12 for statistical analysis and p-value of less than 0.05 was marked as significant.

Sample demographics
Of the total 6,403 patients, 2444 (38%) patients received PCI compared with 3959 (62%) who did not receive PCI (No-PCI group) ( Table 1). Patients in the PCI group were comparatively younger (mean age of 64.5 vs. 66; p < 0.001) and had lower BMI (30.6 vs. 31.1; p = 0.015). There was a significantly higher proportion of male patients in the PCI group, compared with the No-PCI group (70.9% vs. 59%; p < 0.01).

Clinical variables and PCI treatment
Patients in the PCI group demonstrated a larger proportion of smokers (32.6% vs 23.1%; p < 0.001) and those with history of CAD (11.8% vs 9.3%; p = 0.002) (  (Figure 1). The PCI group had a significantly higher number of patients with NSTEMI (31.7% vs 16%), and STEMI (29.8% vs 3.5%) compared with the No-PCI group.

FIGURE 1: Percutaneous coronary intervention comparison in different coronary artery disease presentation types.
Percentage bar graph comparing PCI vs No-PCI ground among CAD presentation classifications.

Predictors of PCI treatment
On linear regression analysis (  Comparison of CAD presentation and PCI rate demonstrated increasing odds of PCI with increased severity of CAD presentation (Figure 2). For example, odds of receiving a PCI were 0.74 in patients with no symptoms at presentation, compared with 27.9 in patients presenting with STEMI ( Table 2).

Discussion
CAD has a spectrum of presentations and mortality risk depends on which end of the spectrum the patient belongs. Regardless, all attempts are made to make sure the diagnosis of coronary artery disease is not missed, particularly with atypical presentations. This leads to a whole gamut of workup which may end up with percutaneous coronary intervention. CAD does have significant social and economic implications.
The overall cost of hospitalization depends, predominantly, whether the patient receives any kind of intervention. Many other factors further contribute to the cost including geographical location, availability of an intervention, which tercile the hospital belongs to, proficiency in complex interventions, etc. Various cost-effective analyses have been done comparing medical management and interventions. A systemic review done by Gholami et al. showed that percutaneous coronary interventions had good results in terms of quality-of-life measures and intermediate initial costs, however, required a higher need for further revascularizations [9]. On the other hand, medical therapy was associated with lower initial costs with lower quality of life measures. Caruba et al. included fifteen randomized control trials and did a network metanalysis and concluded that significant savings in the healthcare expenditure can be achieved by using medical therapy in stable angina patients [10].
So far there has been no study discussing the cost of hospitalization with varied presentations of CAD. Based on our analysis, we had 22.9% of the population with no or atypical symptoms who underwent a coronary angiogram, and only 3.5% received an intervention. The odds for intervention increased consistently with the rest of the presentations, stable angina through STEMI. This was reflected with respect to the cost in our economic analysis. There was no significant difference in the cost for the patients without symptoms or with atypical symptoms whether they received the intervention or not. However, with typical symptoms, there was a definite increase in the cost of the group receiving an intervention. Part of the reason for the increase in cost in the intervention group includes the duration of the stay in the hospital, which is increased as those patients required monitoring post-intervention.

Limitations
Our study is a retrospective analysis with its inherent limitations. Although the percentage of PCI goes up with the more intense presentation of CAD, we had about 1.8% patients in the "No symptoms" group who received PCI; these are likely the patients with occult CAD, presenting for diagnostic coronary angiogram, however because of lack of complete data on the indication, we cannot make this assumption with certainty.
In addition, the difference in cost of care was highest for patients with no symptoms ($6253), when they received a PCI; we have limited data to explain the reason for this difference as well.
30-day readmission was significantly high in the NSTEMI group, we do not have any explanation for this finding. In addition, we did not have enough data points for readmission in patients with atypical symptoms.

Conclusions
The primary outcome of mortality and 30-day readmission rates were better in the PCI group compared with the No-PCI group. The addition of PCI to coronary angiogram was associated with an increased cost of admission. Notably, PCI treatment among patients that presented with stable angina, unstable angina, NSTEMI, and STEMI was associated with significantly higher cost.