Optimizing the management of electrophysiology labs in Chinese hospitals using a discrete event simulation tool

Cardiac arrhythmias encompass a wide range of heart rhythm and heart rate disorders [1]. Clinical presentation can range from asymptomatic to life-threatening, such as sudden cardiac arrest [1]. Arrhythmias can be managed with antiarrhythmic medication or various electrophysiological (EP) procedures, such as ablation or implantation of pacemakers or defibrillators [2]. The most common type of arrhythmia is atrial fibrillation (AF) [3], which is linked to higher risks of stroke, heart failure, dementia, and death [4‐9]. This significantly impacts healthcare costs owing to hospitalizations and loss of productivity [10].

The prevalence of AF in China is estimated to be 1.6% and increases with age [11]. This poses a formidable challenge to China’s healthcare system owing to its aging population [12]. While the ablation treatment rate among Chinese AF patients is unclear, it is reasonable to assume that over 176,000 ablation procedures may be needed based on the US ablation rate of 0.79% in 2005 [13].

Furthermore, overcrowding at tertiary public hospitals is a major problem in China [14]. While economic growth slows, rapid expansion of hospital operations and infrastructure is not a viable option to meet the demand for EP treatments [15]. Thus, decision-makers must utilize limited medical resources to deliver healthcare services more efficiently [15] in order to meet this growing demand.

To this end, China’s National Health Commission has issued guidelines to improve public hospital operational management and healthcare delivery [16]. The Commission not only recommends optimizing resource allocation and processes but also emphasizes data collection and analysis, as well as improving the quality of decision-making [16]. Public hospitals are urged to collect and manage operational data for analysis, establish an analytical decision-making framework, and implement analysis results [16].

Optimizing EP service delivery efficiency is a complex, multidimensional problem involving factors ranging from patient flow management to novel technology adoption. Discrete event simulation (DES) is an operational research tool that helps decision-makers assess different management strategies, enabling them to evaluate not only the performance of current healthcare delivery systems but also that of hypothetical scenarios [17]. This allows them to choose between different approaches of healthcare delivery to prioritize and pursue without undesirably impacting current systems [17].

This study established a generalized DES model of an inpatient EP treatment process, which can be applied across tertiary hospitals in China. The model examines how cardiology departments under different resource constraints can serve more EP patients by improving the efficiency of each phase of the healthcare delivery process, including pre-operative preparation, operative time, or post-operative recovery. This enables hospital decision-makers to clarify which phases of the EP care delivery process to prioritize in different situations in order to better meet the demand for EP treatment.

In this study, we developed and validated a generalizable DES model based on inpatient EP care delivery processes from two large tertiary hospitals in China. We used the model to simulate hospitalization stays of EP patients in scenarios with fully occupied ward beds and EP labs, and evaluated the effects of accelerating different phases of the delivery process under the two conditions. The model can support hospital decision-makers to identify which phase of the EP care delivery process to prioritize under different resource constraints in order to best satisfy demand for EP treatment. Decision-makers can then consider different methods of EP care delivery to target the relevant phase.

Under the condition of a fully occupied ward, reducing operative time had little to no effect, as this was not the bottleneck. On the other hand, reducing length of stay by 10–30% in paroxysmal and persistent AF ablation patients, which account for approximately 30% of all patients, increased total discharges by 1–7%. There was no difference whether the reduction was applied in the pre-operative or post-operative phase. However, it is imperative that reduction in hospitalization time does not compromise the quality of care. This goal can be achieved by integrating evidence-supported new technology or practices. For instance, the prevailing practice in China involves a pre-operative transesophageal echocardiogram (TEE), typically necessitating a hospital stay of 2 to 3 days prior to the procedure [21]. Intracardiac echocardiography (ICE) has been demonstrated as a safe and effective alternative to TEE, conducted during the procedure, thereby requiring only a 0–1 day pre-operative stay [21]. Pertaining to the post-operative phase, proactive prevention of complications contributes to reducing the length of stay while maintaining the quality of care [22]. This can be realized through careful management of patients with pre-existing medical conditions [23] and the accumulation of physician experience [24]. However, in the context of large tertiary hospitals in China, as in our study, the complication rate is generally well managed, leaving limited room for further improvement. In addition, the post-operative stay in China exceeds that of the US, where the standard practice leans towards an overnight stay [25], or even outpatient procedures in some instances [22]. This disparity may be attributed to the US’s adoption of diagnostic-related groups (DRGs) [26]. It is expected that the length of stay in China may decrease as the DRG payment system gains traction in the Chinese healthcare landscape [27].

When EP labs were fully occupied, reducing operative time by 10–30% in AF ablation patients led to a 3–12% increase in total discharges This improvement is made possible by several technological innovations that can reduce AF ablation operative times. The key to AF ablation is achieving pulmonary vein isolation [28]. Conventional point-to-point radiofrequency (RF) ablation involves creating multiple lesions, which may leave gaps in between [29] and drive long operative times [30]. Better navigation techniques, such as remote magnetic navigation-guided RF ablation, may reduce operative times [31]. The Q-FFICIENCY trial showed that very high-power, short-duration ablation using contact force-sensing RF catheters can reduce operative times by almost 50% [32], which is higher than the 30% reduction used in our analyses. According to the model, this reduction level would result in 330 total discharges (21% increase) using the specifications described previously.

In simulations of fully utilized EP labs, reducing operative time led to a more significant increase in discharges compared to reducing the length of stay in scenarios with fully occupied ward beds. This discrepancy may stem from the fact that the existing length of stay is already brief, allowing limited room for reduction, while there is more potential for improving EP lab efficiency. Our real-world data revealed that, over a two-month period, on 25 out of 40 working days, the working hours in the EP labs exceeded 8 h, with 29 days surpassing 7 h. This underscores the high utilization rate of the EP labs on a substantial number of days. However, it’s essential to note that reducing operative time does not necessarily ensure improved throughput, given the inherent variability in operational processes, including fluctuating patient flows. This variability in real-world settings emphasizes the need for adaptable and versatile modeling approaches. Our model aims not only to analyze the specific hospital from which the data originated but also to demonstrate its broader applicability as a decision-making tool for hospital managers. Recognizing the variability in hospital conditions, we tested two scenarios reflecting significant operational challenges, particularly relevant in China. These scenarios illustrate how our model can adapt to different settings, providing insights into potential operational improvements. The high utilization of the EP labs observed in our data exemplifies one of these challenges, highlighting the necessity for contextually informed strategies to enhance hospital efficiency.

While US practices and technologies provide useful insights as described above, it is crucial to tailor these approaches to align with China’s unique healthcare landscape. For instance, the adaptation from the DRG system to the diagnosis-intervention packet (DIP) system in China arose from challenges faced during a DRG pilot program [33]. The DIP system offered a less technically demanding and more scalable approach [33]. Since 2020, China has been piloting a dual-track arrangement with both DRGs and DIP, acknowledging the diverse regional capacities of local health systems [33]. This initiative underlines the importance of contextualizing global practices to improve resource allocation and potentially reduce hospital stay lengths without compromising care quality, aligning with the overarching objective of enhancing efficiency in healthcare delivery.

A previous study by Kowalski et al. investigated the economic value of reducing AF ablation operative times using DES [34]. However, this study differed from ours conducted in China. Kowalski et al. did not include a hospitalization process [34], which may have been a choice in the model design or because AF ablations in their study were day surgeries. In addition, ablation procedures were arranged in block schedules such that two ablations were performed in an EP lab each day, with the possibility of one additional procedure [34], placing a hard limit on the number of procedures per day. In our model, accounting for competition of ward beds was important because of the almost 100% utilization of hospital beds in large tertiary hospitals in China [35]. Furthermore, EP procedures in China do not follow a block schedule. Procedures can start as soon as the previous procedure is completed and the EP lab is ready. Therefore, there is more potential to increase the number of procedures by reducing operative time.

Although the DES technique is increasingly used in healthcare research, applying it as a generalized decision-making tool is challenging. First, a DES model based on the detailed practices of one hospital may not be suitable for other settings. In addition, employing a DES requires a large amount of data, and more detailed data increases cost and time [36]. Adapting a DES model to a new setting requires new data [37], and new data is again required if the system undergoes a change [38]. Furthermore, the simulated results of specific actions may differ from actual implementation owing to human variation [39].

In complex systems such as healthcare, a decision-making tool may be more helpful to clarify generic activity patterns instead of focusing excessively on specific processes of one particular site [38]. Therefore, we summarized and abstracted the processes at two different hospitals to build a generalized DES model of the inpatient EP care delivery process. Lowering the complexity of the model would also greatly reduce the effort required to collect data and facilitate model updates if system changes are made. This model does not prescribe specific actions but instead identifies the bottleneck of the care delivery process. Knowing which phase of the process to prioritize, the decision-maker can consider several different options to improve the healthcare delivery process and use the model to understand the expected results of these options. Further, they can also select the most worthwhile option by considering the time and effort involved compared to the expected increase in total discharges.

This study has several limitations. First, the generalizability of this model may be constrained as it was based on data from two tertiary hospitals collected over a span of two months. This model might not fully reflect the care delivery processes in other hospitals within China or in other countries. The input data may not account for potential seasonal or procedural variations. Additional research is needed to validate the applicability of the approach in more diverse settings. In addition, our approach involved testing scenarios where hospitalization time was reduced by a specified percentage. However, depending on the technology or practice applied to improve the workflow, not all patients may benefit; identifying the proportion of affected patients would yield a more robust analysis. Due to the absence of such data currently, we opted to test scenarios with varying degrees of hospitalization stay reduction to characterize the uncertainty surrounding this aspect. Furthermore, we did not include scenarios with reduced variability, although this may also impact total discharges. However, we tested several scenarios, and the changes were relatively minor. Moreover, reducing variability without changing the mean in practice is difficult, and therefore these scenarios were not included in the results. Finally, the distributions of the simulated number of discharges in some scenarios are quite wide, implying that even if the simulations are accurate, the implementation results would still be uncertain.

This study developed a generalized DES model to simulate an EP patient’s hospitalization and treatment at a tertiary hospital in China. Using this model, hospital decision-makers dealing with different resource constraints can identify which phase of the care delivery process to optimize in order to better meet demand for EP treatment. Simulations showed that if ward beds are fully occupied, reducing the length of stay of AF ablation patients by 10–30% resulted in a 1–7% increase in the total number of patients discharged. On the other hand, under the condition of fully occupied EP labs, reducing the operative time of AF ablation patients by 10–30% increased discharges by 3–12%.