Direct admission versus interhospital transfer for revascularisation in non‐ST‐segment elevation myocardial infarction

Abstract Background The differences in outcomes and process parameters for NSTEMI patients who are directly admitted to an intervention centre and patients who are first admitted to a general centre are largely unknown. Hypothesis There are differences in process indicators, but not for clinical outcomes, for NSTEMI who are directly admitted to an intervention centre and patients who are first admitted to a general centre. Methods We aim to compare process indicators, costs and clinical outcomes of non‐ST‐segment elevation myocardial infarction (NSTEMI) patients stratified by center of first presentation and revascularisation strategy. Hospital claim data from patients admitted with a NSTEMI between 2017 and 2019 were used for this study. Included patients were stratified by center of admission (intervention vs. general center) and subdivided by revascularisation strategy (PCI, CABG, or no revascularisation [noRevasc]). The primary outcome was length of hospital stay. Secondary outcomes included: duration between admission and diagnostic angiography and revascularisation, number of intracoronary procedures, clinical outcomes at 30 days (MACE: all‐cause mortality, recurrent myocardial infarction and cardiac readmission) and total costs (accumulation of costs for hospital claims and interhospital ambulance rides). Results A total of 9641 NSTEMI events (9167 unique patients) were analyzed of which 5399 patients (56%) were admitted at an intervention center and 4242 patients to a general center. Duration of hospitalization was significantly shorter at direct presentation at an intervention centre for all study groups (5 days [2–11] vs. 7 days [4–12], p < 0.001). For PCI, direct presentation at an intervention center yielded shorter time to diagnostic angiography (1 day [0–2] vs. 1 day [1–2], p < 0.01) and revascularisation (1 day [0–3] vs. 4 days [1–7], p < 0.001) and less intracoronary procedures per patient (2 [1–2] vs. 2 [2–2], p < 0.001). For CABG, time to revascularisation was shorter (8 days [5–12] vs. 10 days [7–14], p < 0.001). Total costs were significantly lower in case of direct presentation in an intervention center for all treatment groups €10.211 (8750–18.192) versus €13.741 (11.588–19.381), p < 0.001) while MACE was similar 11.8% versus 12.4%, p = 0.344). Conclusion NSTEMI patients who were directly presented to an intervention center account for shorter duration of hospitalization, less time to revascularisation, less interhospital transfers, less intracoronary procedures and lower costs compared to patients who present at a general center.


| INTRODUCTION
Non-ST-segment elevation myocardial infarction (NSTEMI) is defined as angina symptoms with elevated cardiac biomarkers in the absence of persistent ST-segment elevation. 1 Based on numerous randomized trials and several meta-analyses, a routine invasive strategy is recommended by the European society of Cardiology (ESC) guideline in most NSTEMI patients. 1 Since 2015, the ESC has increasingly endorsed an early invasive strategy and in 2020, the most recent guidelines for myocardial revascularisation recommend same day transfer of patients with an established NSTEMI diagnosis from a general to an intervention center (class I, level of evidence A). 1 However, in the Netherlands and other countries in Europe, only generic regional arrangements are made between centers and emergency medical services toward the location of patient presentation of NSTEMI patients. [2][3][4][5] The expert opinion of the paramedic in conjunction with telephonic consultation with the attending cardiologist (if deemed necessary), leads to a decision where to present a patient. In practise this means low and high risk NSTEMI patients are transferred to the nearest hospital (intervention or nonintervention center) and very high risk patients (i.e., haemodynamic instability or cardiogenic shock, recurrent or ongoing chest pain refractory to medical treatment, lifethreatening arrhythmias, mechanical complications of MI, acute heart failure, or recurrent dynamic ECG changes) are, in accordance with the ESC guideline and in analogy to the STEMI population, directly admitted to an intervention center. 6,7 Although we know that these recommendations are not always followed in clinical practise. 4 Furthermore, in the Netherlands, intracoronary angiography (ICA) is performed at both general and intervention centers; however only the latter has on-site revascularisation facilities (either PCI or CABG) due to national legislation. General centers, in contrast, are limited to diagnostic invasive facilities, without PCI or CABG capabilities. As a result, NSTEMI patients who are initially admitted at a general center and are candidates for revascularisation inevitably require interhospital transfer. The current situation in the Netherlands therefore lends itself to evaluate the incremental value of the ESC guideline recommendation of same day transfer of patients with an established NSTEMI diagnosis from a general to an intervention center. The primary objective of the current study is to conduct a comparative analysis of process indicators, clinical outcomes and costs of NSTEMI patients who are directly admitted at an intervention center and patients who are first admitted at a general hospital.

| Study design and study population
The design of the current study was a retrospective cohort study using aggregated hospital claim data from CZ (Centraal Ziekenfonds) which is one of the largest health insurance companies in the Netherlands with 3.7 million insured persons in 2019 and a market share of~16% of the total population and 50% of the South of the Netherlands (in the Netherlands basic medical insurance is mandatory and all NSTEMI care is covered). 8 The study was approved by the local medical ethical committee, which waived the need for informed consent (W20.310).
The study population consisted of adult patients who presented at an emergency department and were diagnosed by a physician with a NSTEMI (ICD-10 I21.4) between 2017 and 2019 and were insured at CZ. Each unique NSTEMI event was included. Patients who switched to another health insurer within 30 days after the event were excluded (~6% of the insured switch annually). Patients who had a NSTEMI diagnosis registration during an existing admission for another reason were excluded because the aim of the current study focusses on logistics in NSTEMI hospitalizations.
Alongside an analysis of the total study population, a division was made for each NSTEMI event into the following cohorts: (1) NSTEMI events for which patients underwent PCI during initial admission; (2) NSTEMI events for which patients underwent CABG during initial admission; (3) NSTEMI events for which patients did not undergo revascularisation during initial admission (noRevasc).
For the total population and each cohort, a distinction was made between presentation at an intervention center and presentation at a general center. A hospital was considered an intervention center if facilities for performing PCI and/or CABG were available on site. A general center was defined as a center that only has diagnostic invasive facilities, and does not have PCI or CABG facilities. An admission was defined as the uninterrupted, cumulative clinical episodes in intervention and general centers.

| Baseline characteristics
Given the utilization of a database based on insurance claim data, a comprehensive characterization of the study population is not anticipated. This stems from the fact that not all findings corresponding to an NSTEMI are reported to the insurance provider.
Nevertheless, to facilitate a scoping understanding of the patient population under study, several characteristics can be inferred from the available claim and diagnosis records. As such, the following baseline characteristics were assessed for each NSTEMI event in the study period: age, gender, diabetes mellitus, obstructive pulmonary disease (COPD/asthma), renal insufficiency, previous PCI and previous CABG. Data was based on historical claim data from CZ and was available from 2014 onward. Diabetes mellitus (DM) was defined as a diagnosis registration on a hospital reimbursement claim for DM (or its complications) or use of glucose-lowering medication.
Obstructive pulmonary disease was defined as a diagnosis registration on a hospital reimbursement claim for COPD/asthma or use of any inhalation medication. Renal insufficiency, previous PCI and previous CABG were defined as a historical registered diagnosis for the disease or treatment.

| Study outcomes
The primary outcome for this study was duration of hospitalization.
This outcome reflects the NSTEMI care pathway most accurately and may act as a surrogate for time to invasive diagnostic work-up and revascularisation. The latter are of importance to improve outcomes of NSTEMI patients. 1 Secondary outcomes were process indicators (time from admission to ICA and revascularisation and total number of coronary diagnostic sessions performed during admission), costs and MACE at 30 days. MACE was defined as a composite of all-cause mortality within 30-day after the NSTEMI event, myocardial infarction (STEMI or NSTEMI) within 30 days after discharge and cardiac readmission rate (defined as all readmission for cardiovascular causes excluding readmissions for myocardial infarction) at 30 days using claim data from CZ. These outcomes were also reported individually.
Mortality was subdivided in mortality between 0 and 5 days and 6-30 days because differences in mortality between these periods may be the result of triage that we cannot identify from the baseline data.

| Statistical analyses
General descriptive statistics of included NSTEMI events were used to describe baseline characteristics and primary and secondary outcomes as stratified by the three treatment modalities and subdivided per hospital of presentation. Data was expressed as mean ± standard deviation (SD) for continuous normal distributed data, median with interquartile range (IQR) for continuous nonnormal distributed data and as absolute and relative frequencies for categorical data. Student's t test, Mann-Whitney U test or Fisher exact-test/χ 2 test respectively were used to make a comparison between cohorts where appropriate. A p < 0.05 was considered significant and all analyses were performed using SPSS 25 (SPSS Inc). VAN Figure 1). In the total population, no significant difference in baseline characteristics was found between study groups (Table 1).

| Primary outcome
For the total study population, length of stay was shorter for patients directly admitted to an intervention center in comparison to initial admission to a general center (5 days
Costs were lower for patients directly transferred to an intervention F I G U R E 1 Flowchart of included patients and allocation to study groups. CABG, coronary artery bypass surgery; NSTEMI, non-ST-segment elevation myocardial infarction; PCI, percutaneous coronary intervention.

| DISCUSSION
The principal finding of the current study was that patients who were directly admitted to an intervention center had shorter duration of hospitalization compared to patients first admitted to a general center regardless if these patients undergo revascularisation during admission or not. This prolonged duration of hospitalization appeared driven by longer time between admission and intracoronary angiography (ICA) in PCI patients while the duration between admission and the PCI and CABG also contributed to these differences. Additionally, secondary outcomes revealed that patients requiring revascularization and initially admitted to a general center had significantly more interhospital ambulance rides and hospital claims, leading to higher overall costs. The current logistics in  T A B L E 2 Primary and secondary outcomes related to unique NSTEMI events stratified by center of presentation and treatment strategy during initial admission.     | 1003 compared to direct presentation at a PCI center. 10  Furthermore, the general efficiency of workflow and regional arrangements also yield improvement potential. Attributes to the primary findings are most probably ad hoc scheduling issues at the intervention center combined with heart team deliberation in selected patients (e.g., postponement of treatment decision due to missing information). Improvement potential could potentially lie in streamlining work processes with all regional actors and tailored scheduling. The application of information technologies for fast sharing of ICA data (which is available for most centers in the Netherlands) is a great step in this direction. Applying bundled payment models for the entire care chain associated with a NSTEMI event could also stimulate optimization of regional logistics and make providers financially accountable for the quality and costs of an entire NSTEMI episode. Also, these models could prevent duplicate claims by multiple providers. 21,22

| Limitations
The current study is subject to general shortcomings associated with retrospective studies and relies on the correct diagnosis registration. Nevertheless, there is vast experience with using declaration or diagnosis registration data to evaluate outcomes for various conditions or treatments and this data is considered highly accurate. 23 However, important confounding factors that are relevant for risk classification and adjustment were not available (i.e., hemodynamic parameters, findings on the ECG, lab results), thus impeding more in-depth analysis such as regression analysis for clinical outcomes and risk stratification. It is worth noting that, aside from age, all other variables crucial to the GRACE risk scorea primary risk stratification tool for estimating mortality risk post-NSTEMI-were unattainable. 24 Future research with these variables at hand is warranted.

| CONCLUSION
In conclusion, our study demonstrates that NSTEMI patients who are directly admitted to an intervention center experience shorter hospitalization durations, fewer interhospital transfers, and reduced costs compared to those first admitted to a general center. Despite these advantages, transferring all NSTEMI patients directly to intervention centers may pose logistical challenges and may not yield consistent clinical benefits across all patient subgroups.
Exploring alternative approaches, such as enhancing prehospital triage and streamlining regional arrangements, is crucial for optimizing healthcare resources and improving patient outcomes. Further research is needed to validate these findings and investigate innovative strategies in various healthcare settings.