How Can We Tackle Emergency Department Crowding Once And For All? (Part 3 Of 3)

This article is the third in a three-part series which seeks to dissect the challenge of emergency department (ED) crowding; why it occurs, why it’s a problem and what we need to do to overcome it. The creation of this series is driven by the ever-pressing need to secure the sustainability of our EDs, which move closer and closer to breaking point with each day. This series has been curated with the support of senior Emergency Medicine Consultants from NHS England.

In the previous two articles we’ve defined the problem and outlined its impact on patient outcomes and quality of care delivery. But now, the question remains; how can we tackle this global problem once and for all?

​Tools which provide superficial measurements of ED crowding to aid decision making, such as the NEDOCS Score for Emergency Department Overcrowding and the International Crowding Measure in Emergency Departments (ICMED), are available to emergency medicine leaders, but their limitations and shortcomings are widely recognised.

​As such, other solutions have been tried and proposed with varying degrees of efficacy. These include; increased access to primary care and general practitioners (GPs), alternative models of care, and patient flow modelling via computer simulation.

Let’s evaluate each solution before outlining where we must apply our efforts next if we are to truly alleviate ED crowding.

INCREASED ACCESS TO PRIMARY CARE AND GENERAL  PRACTITIONERS

Increasing access to primary care, as well as the appropriate redirection of low-acuity patients, have both been considered to reduce crowding [1]. One UK study found a reduction in low-acuity presentations as a result of implementing GP-led walk-in centres [2]. Additionally, various studies have shown that increasing GP opening hours has had positive effects on ED crowding [3-6], although the efficacy varies across the studies.

ALTERNATIVE MODELS OF CARE

Alternative models of care, such as the Discharge to Medical Home model, have also aimed to reduce the number of low-acuity patients entering the ED. These models route low-acuity and ambulatory ED patients to a primary care clinic, providing a connection to primary care for the patient. During clinic hours, walk-in patients to EDs can be assessed, and if deemed low risk, can be scheduled a same-day primary care appointment [7].

​However, it is worth noting the position of the Royal College of Emergency Medicine, who have consistently argued that the proportion of low-acuity patients who could be treated in alternative healthcare settings is no more than 15%, so this suggests that the effect of this solution could be limited. [8].

PATIENT FLOW MODELLING VIA COMPUTER SIMULATION

Computer simulations that involve patient flow modelling allow ED leaders to test the impact of changes to care delivery more efficiently, in real time, before implementing them. 

​This may not only have a positive impact on causes related to input, but also throughput and, most importantly, output [9]. Interventions for crowding need to involve more than just the ED [10], so strategies that target input and throughput have had only some positive effects, and these effects are likely to be limited and unsustainable. By predicting demand for hospital care via computer simulation, ED leaders can see when exit block is likely to occur, and therefore respond with more appropriate interventions to address output also [11].

SO, WHERE DO WE GO FROM HERE?

Upon examination it becomes clear that in order for a solution to be effective in tackling ED crowding, it needs to take a holistic approach. Creating an initiative which focuses solely on the ED will be unlikely to stand the test of time. It’s also important to note that, as we explained in the first article in this series, crowding is often affected by many local factors, so a one-size-fits-all solution is likely to prove ineffective [12].

​And it is for these reasons that patient flow modelling via computer simulation seems the most appropriate route forward, as long as the solution is robust and testable, with a measurable outcome.

AN INTERNATIONAL CONSORTIUM TO TACKLE CROWDING ONCE AND FOR ALL

At electronRx, we are working with senior Emergency Medicine Consultants from NHS England and a number of US hospital systems to bring together a consortium of innovative emergency medicine leaders that aims to establish an internationally validated ED crowding solution.

We are marrying our vast knowledge and experience in machine learning with unrivaled clinical expertise to solve this problem; on three levels.

​At a high-level, visualising operational data can enable earlier identification and mitigation of predictable bottlenecks in hospital patient flow. Historical data can also facilitate predictions that allow leaders to gain insights into what will happen in their EDs in 3 hours’ time, as well as what the results of any interventions will be. These holistic predictions allow for more informed and effective resource allocation, and empower leaders with data-driven decision making, rather than leaving them to rely simply on their gut feelings.

​Then, at a mid-level, ED data becomes incredibly valuable when you factor in variables such as length of stay, patient admission information and bed occupancy, allowing predictions to be made at an individual patient-level. 

​However, utilising predictive data is only scratching the surface of what can be achieved when it comes to alleviating crowding. Most exciting is the hitherto unexplored deep-level of insights we can now access due to innovations in digital health tech. Through automated triage and the introduction of patient wearables, ED triage and observations can be streamlined whilst also providing previously inaccessible amounts of patient data. 

​Post-triage, patients can be given wearables which provide a constant stream of observational data. This opens up the opportunity to not only track where they are in the hospital system, but also to capture their vitals in real-time to better inform robust neural networks that allow you to experiment and predict the effect of ED interventions with measurable results, at a level of granularity never seen before.

​ED crowding presents a global threat to the sustainability of our healthcare systems and economies. It’s time to push forward with change.

Increasing access to primary care, as well as the appropriate redirection of low-acuity patients, have both been considered to reduce crowding [1]. One UK study found a reduction in low-acuity presentations as a result of implementing GP-led walk-in centres [2]. Additionally, various studies have shown that increasing GP opening hours has had positive effects on ED crowding [3-6], although the efficacy varies across the studies.

ALTERNATIVE MODELS OF CARE

Alternative models of care, such as the Discharge to Medical Home model, have also aimed to reduce the number of low-acuity patients entering the ED. These models route low-acuity and ambulatory ED patients to a primary care clinic, providing a connection to primary care for the patient. During clinic hours, walk-in patients to EDs can be assessed, and if deemed low risk, can be scheduled a same-day primary care appointment [7].

​However, it is worth noting the position of the Royal College of Emergency Medicine, who have consistently argued that the proportion of low-acuity patients who could be treated in alternative healthcare settings is no more than 15%, so this suggests that the effect of this solution could be limited. [8].

PATIENT FLOW MODELLING VIA COMPUTER SIMULATION

Computer simulations that involve patient flow modelling allow ED leaders to test the impact of changes to care delivery more efficiently, in real time, before implementing them. 

​This may not only have a positive impact on causes related to input, but also throughput and, most importantly, output [9]. Interventions for crowding need to involve more than just the ED [10], so strategies that target input and throughput have had only some positive effects, and these effects are likely to be limited and unsustainable. By predicting demand for hospital care via computer simulation, ED leaders can see when exit block is likely to occur, and therefore respond with more appropriate interventions to address output also [11].

SO, WHERE DO WE GO FROM HERE?

Upon examination it becomes clear that in order for a solution to be effective in tackling ED crowding, it needs to take a holistic approach. Creating an initiative which focuses solely on the ED will be unlikely to stand the test of time. It’s also important to note that, as we explained in the first article in this series, crowding is often affected by many local factors, so a one-size-fits-all solution is likely to prove ineffective [12].

And it is for these reasons that patient flow modelling via computer simulation seems the most appropriate route forward, as long as the solution is robust and testable, with a measurable outcome.

AN INTERNATIONAL CONSORTIUM TO TACKLE CROWDING ONCE AND FOR ALL

At electronRx, we are working with senior Emergency Medicine Consultants from NHS England and a number of US hospital systems to bring together a consortium of innovative emergency medicine leaders that aims to establish an internationally validated ED crowding solution.

We are marrying our vast knowledge and experience in machine learning with unrivaled clinical expertise to solve this problem; on three levels.

​At a high-level, visualising operational data can enable earlier identification and mitigation of predictable bottlenecks in hospital patient flow. Historical data can also facilitate predictions that allow leaders to gain insights into what will happen in their EDs in 3 hours’ time, as well as what the results of any interventions will be. These holistic predictions allow for more informed and effective resource allocation, and empower leaders with data-driven decision making, rather than leaving them to rely simply on their gut feelings.

​Then, at a mid-level, ED data becomes incredibly valuable when you factor in variables such as length of stay, patient admission information and bed occupancy, allowing predictions to be made at an individual patient-level. 

​However, utilising predictive data is only scratching the surface of what can be achieved when it comes to alleviating crowding. Most exciting is the hitherto unexplored deep-level of insights we can now access due to innovations in digital health tech. Through automated triage and the introduction of patient wearables, ED triage and observations can be streamlined whilst also providing previously inaccessible amounts of patient data. 

​Post-triage, patients can be given wearables which provide a constant stream of observational data. This opens up the opportunity to not only track where they are in the hospital system, but also to capture their vitals in real-time to better inform robust neural networks that allow you to experiment and predict the effect of ED interventions with measurable results, at a level of granularity never seen before.

​ED crowding presents a global threat to the sustainability of our healthcare systems and economies. It’s time to push forward with change.

REFERENCES

1. Morley C, Unwin M, Kinsman L. Emergency department crowding: A systematic review of causes, consequences and solutions. PLoS One. 2018; 13(8): e0203316.

2. Arain M, Campbell MJ, Nicholl JP. Impact of a GP-led walk-in centre on NHS emergency departments. Emerg Med J. 2015;32(4):295–300.

3. Dolton P, Pathania V. Can increased primary care access reduce demand for emergency care? Evidence from England's 7-day GP opening. J Health Econ. 2016;49:193–208.

4. Whittaker W, Anselmi L, Kristensen SR, et al. Associations between extending access to primary care and emergency department visits: a difference-in-differences analysis. PLoS Med. 2016;13(9):e1002113.

5. Buckley DJ, Curtis PW, McGirr JG. The effect of a general practice after-hours clinic on emergency department presentations: a regression time series analysis. Med J Aust. 2010;192(8):448–51.

6. Nagree Y, Ercleve TNO, Sprivulis PC. After-hours general practice clinics are unlikely to reduce low acuity patient attendances to metropolitan Perth emergency departments. Aust Health Rev. 2004;28(3):285–91. 

7. Zager K. Discharge to medical home: A new care delivery model to treat non-urgent cases in a rural emergency department. Health (Amst). 2019 Mar;7(1):7-12. 

8. Boyle A, Higginson I. What should we do about crowding in emergency departments? British Journal of Hospital Medicine. 2018;79(9):1750-8460. https://doi.org/10.12968/hmed.2018.79.9.500.

9. Mohiuddin S, Busby J, Savovic J, et al. Patient flow within UK emergency departments: A systematic review of the use of computer simulation modelling methods. BMJ Open. 2017;7:e015007.

10. Forero R, McCarthy S, Hillman K. Access block and emergency department overcrowding. Crit Care. 2011;15(2):216.

11. Mason, S., Knowles, E. and Boyle, A. (2016) Exit block in emergency departments: a rapid evidence review. Emergency Medicine Journal. ISSN 1472-0205. https://doi.org/10.1136/emermed-2015-205201.

12. Javidan AP, Hansen K, Higginson I, Jones P, Petrie D, Bonning J, Judkins S, Revue E, Lewis D, Holroyd BR, Mazurik L, Graham C, Carter A, Lee S, Cohen-Olivella E, Ho P, Maharjan R, Bertuzzi B, Akoglu H, Ducharme J, Castren M, Boyle A, Ovens H, Fang C, Kalanzi J, Schuur J, Thiruganasambandamoorthy V, Hassan T, Bodiwala G, Convocar P, Henderson K, Lang E. White paper from the emergency department crowding and access block task force [Internet]. International federation for emergency medicine. 2020 June. Available from: https://www.ifem.cc/resource-library/.

13. Chapter 39 Bed occupancy Emergency and acute medical care in over 16s: service delivery and organisation. National Institute for Health and Care Excellence. March 2018. Accessed April 30, 2021. https://www.nice.org.uk/guidance/ng94/evidence/39.bed-occupancy-pdf-172397464704.