A Value-based Healthcare Approach to Emergency Care

By Seb Kerr and Neil Milner

The Problem

A hospital needed to meet patient demand, reduce waiting times and reduce hospital admissions but had no more money to spend to address these issues. This is a problem that is common the world over.

Emergency Departments in the UK are under more pressure than ever before. A reduction in real-time funding over the last ten years has seen staff burnout, longer waiting times and a service firefighting every day to cope with the demand on the service. With less accessibility to primary care resources, Emergency Departments are overwhelmed with more and increasingly complex demands. Yet, in modelling terms, the UK’s National Health Service’s (NHS’s) Patient Level Information and Costing System (PLICS) approach to understand Accident and Emergency (A&E) services has delivered no tangible benefit in helping to address this problem.

Challenges from Using PLICS

All UK hospitals are required to submit detailed costs at the patient level, by resource and activity, on an annual basis to the UK regulator. This is the largest collection of detailed patient costs in the world. An average hospital submission is many millions of rows of data. This submission includes each A&E attendance. You would expect with all this data the problem could be addressed. Unfortunately, that is not the case. Key reasons why PLICS does not deliver value include:

• The PLICS model is primarily designed to track a tariff.

• The key objective of the submission is to inform a tariff mechanism for payment. All attendances are clinically coded and grouped into an HRG (Healthcare Resource Group).

• These HRGs form the basis for payment and within A&E there are less than ten HRGs that form the basis for such payments.

• Additionally, price and cost are two quite different things, and price has no bearing on the cost of providing healthcare.

• For these reasons, the PLICS “rules” for costing are overly simplified.

PLICS rules indicate that A&E cost should be traced to patients based on treatment time. While superficially this makes some sense and appears to follow Time Driven ABC principles, it obscures the opportunity to address the challenge of delivering services more effectively and efficiently.

The drivers of cost in PLICS only focus on patient demand. With treatment time and the number of patients forming the basis for understanding cost, there are only two levers the hospital can pull to appear to improve their performance, namely by either:

• Seeing more patients and reducing the unit cost of average treatment minute of the resources used OR

• Increasing the amount of time spent treating each patient, effectively increasing the number of demand minutes and therefore reducing their unit cost per minute.
  

Neither of these improve productivity or help the patient!

Cost must be balanced with non-cost metrics yet PLICS focusses on cost information to the exclusion of non-cost information. Simply looking at the cost effect presents no context by which it can be understood. The Information side of PLICS should be about bringing in non-cost metrics to provide further understanding and to help address the problem.

More data and detail do not always mean a better model. PLICS delivers on data and detail, yet this wealth of data delivers little of value to a hospital that wishes to improve performance and deliver better patient care. PLICS therefore becomes yet another data liability for a hospital already drowning in regulatory and operational data reporting requirements.

PLICS views cost as being “spent” by the patient. Hospitals spend money on employing resources to meet patient demand. Cost at the patient level is merely the effect of these resourcing decisions.

The focus of PLICS is on patient attendance and not the patient. As the tariff payment for A&E is based on attendances, the focus of the PLICS submission is on each attendance and not the patient - so the “P” of PLICS is actually an “A”.

At the root of all these shortcomings is one aspect: the PLICS model is not based on causal principles and is therefore limited in its use at best and is misleading at worst. Essentially the PLICS model is not a digital twin of what physically happens in the A&E department.

Understanding Supply and Demand

It is imperative that both resource supply and patient demand are understood in sufficient, actionable detail in order to understand causality. This data is already available and can be understood at a level where it can be used to effect tangible change and start to help to solve the problem, especially as the hospital can control and influence supply more than demand. 

In PLICS the term “cost driver” is overly simplified as it typically focuses on patient demand.  Whilst in many cases reducing the time each activity takes is a good thing, this is not always the case and particularly in healthcare this can lead to a reduction in quality, put more pressure on healthcare professionals, and ultimately lead to significantly more cost and patient demand being generated further down the line.  In practice, there are supply drivers and demand drivers, or in ABC terms, resource drivers and activity drivers. 

Operational variation versus clinical variation. Considerable research and analysis focusses on removing clinical variation.  Clinical variation often focusses on homogenising care and identifying variances in clinical practice.  For example, why for the same condition were more tests requested for Patient A versus Patient B?

Healthcare is complex and because of this doctors and nurses are some of the most skilled professionals on the planet. While pressure to reduce clinical variation may yield some benefits, it can over-simplify healthcare and devalue the operational decisions that these immensely talented people have to make. 

The good news is that focussing on operational variation has the potential to help improve performance and patient care and begin to solve the problem. With the wealth of data now available we can begin to understand and identify operational variation and obtain the benefits this understanding can reveal.

Costing Supply & Demand Presents Opportunity

As already discussed, in PLICS the basis for allocating cost based purely on patient demand presents one perspective and this limits the operational benefits that a hospital can gain from PLICS.  Figure 1 illustrates the current limitations of costing based on patient demand only, while Figure 2 demonstrates how, by bringing in resource supply and expanding the patient demand data to include such resources, the opportunities for operational improvement begin to present themselves.

Figure 1. Costing Based on Patient Demand

Figure 2. Costing Based on Resource Supply & Patient Demand

In practice, operational variation occurs hour by hour, day by day.  By costing supply and demand, causality begins to be reflected, leading to the potential for action and change to occur. This can be done in three steps:

Step 1 – Understand Supply

The general ledger helps us to identify the total spend on A&E Nursing in the period. Using the HR roster, we can see who worked when. Using this data, we can see the actual supply minutes available, and their cost, for each hour of each day.

Step 2 – Understand Patient Demand

Demand is made up of the following three components which form the basis of understanding the patient demand for the A&E services.

A.      Attendances - How many attendances were treated and for which patients? This defines the universe of patients demanding treatment from A&E in each period.

B.      How long did the treatment take? - The longer a patient was treated, the more demand they placed on the A&E service – The same as the recommended approach for costing A&E attendances

C.      How many nurses were present at each treatment? – Acuity of care is not considered in the national PLICS approach, yet it has a huge bearing on demand. A cut finger may take time to be treated by a single nurse, whilst a patient requiring resuscitation may be stabilised quickly but require an army of resources to make this happen.

Patient Demand = A * B * C

We also know the start and end date and times of each treatment so we can begin to understand the aggregate patient demand each hour of the day.

Step 3 – Understand Utilisation

By understanding supply and demand each hour of the day we can begin to understand and identify where the opportunity is for change. What are the results?

• Supply: 23.6m minutes; Cost per supply minute: 56p

• Demand: 17.1m minutes; Average cost per demand minute:  77p

• Utilisation: 72.4%; Opportunity to explore:  £3.68m (28%)

As demand increases, pressure on the service increases as demand outpaces supply.

Figure 3. Supply and Demand Minutes by Calendar Month

Supply, Demand and Utilisation by Weekday

Demand is lower on weekends, but utilisation remains largely consistent.

Figure 4. Supply, Demand and Utilisation by Weekday

In the example below, A&E services are under pressure from 4 pm onwards with two spikes at 7 am and 7pm indicating when shifts changes occur. One occurs at the quietest time for the department, and the other nearly at the peak of demand.

Figure 5. Supply, Demand and Utilisation by Hour

Each of these data points are also fully costed. Where utilisation is low, the effect of this is increased costs for patients treated in that hour. Where utilisation is high the reverse is opposite. This also indicates pressure on the service is high at different points in the day.

“The closer you come to the capacity of a resource the greater the damage is done by a disruption”.

Source:  Eliyahu M. Goldratt

With such high levels of demand and inconsistent supply levels, the chances for disruption become higher when A&E is at its busiest. The utilisation chart (Figure 6) indicates consistent utilization over all days of the week.

Figure 6. Utilisation Across Days of the Week (x axis = Hour of the day, y axis = hourly utilisation)

Actions:  This approach gives insight that can be acted upon and provides a chance to improve both productivity and patient care. By changing resource levels, utilisation can be smoothed out and patient demand for medical resources used more effectively.

Using such data to inform future staffing patterns offers an opportunity to reduce operational variation without negatively impacting the care given to patients by more efficiently using resources.

Summary

Understanding Healthcare is a complex task, but logical causal modelling at the patient level can shine new light on improving healthcare for all and improve hospital productivity.

The approach we have described is not for the faint hearted, with thousands of resource drivers and millions of activity drivers across a complex organisation such as an NHS hospital. However, the benefits of such analysis have the potential to far exceed its related cost, turning existing PLICS data from a data liability into a true data asset.

Next Steps

More work is needed to further understand Nursing within Accident & Emergency departments. This is the first step on the journey to understanding A&E Nursing, and the cause and effect of supply decisions on cost. The following key areas will be explored further as a part of the next case study:

Understanding Supply further: Currently supply is all nursing in A&E – This effectively gives us the same cost per supply minute across the year. However, nursing supply is made up of a mix of resources.  Substantive staff at different bands employed by the hospital, overtime worked, bank and agency staff.  

Figure 7. Understanding Resource Mix

By understanding the staff mix per hour, a cost differential can be an observed which also highlights key areas where the hospital must rely on more expensive variable resources to meet patient demand.

Understanding patient demand further: Demand is currently being understood in the context of one activity. By understanding how time is spent managing the flow of patients through A&E by hour, a greater understanding of the dynamics of patient demand can be better understood.

Figure 8. Understanding Patient Flow

Improving waiting times: Currently patient demand shows when patients are treated. However, if we want to reduce waiting times we also need to look at when that demand presented itself. By offsetting each patients’ treatment time to when they first presented to A&E and when staffing was available to meet this demand, patients can be treated more quickly and the pressure on the A&E service minimised. 

Understanding admissions: By identifying which attendances led to admission, we will be able to better understand if pressure on the system is a contributing factor that leads to such admissions. However, targets, measures, and decisions are too often made in a siloed way that may improve performance in one area but cause a larger problem to be managed later downstream, an issue avoided by seeing the bigger picture.

Understand the true drivers of cost: staff satisfaction, absence rates and sickness rates Staff welfare and work satisfaction have a huge bearing on the running of a hospital, and it is vital to understand these measures in conjunction with cost and utilisation.

Figure 9. Understanding Staff Absence

Staff are the healthcare system’s most precious resource. Change should not negatively impact staff. A&E is one of the highest-pressure work environments in the world, with life and death decisions made daily. It is the responsibility of all support functions within a hospital to support those on the front-line in caring for patients. By aligning supply and demand more closely and ensuring safe levels of staff are in place to meet the demand, some pressure can be removed.

Managing Demand: Whilst more difficult to do, the opportunity to reduce patient demand for A&E services and consider more appropriate ways to care for patients can now be better understood and inform actionable insights.

Finally, although the study focusses on A&E the use of these techniques is equally possible throughout the Hospital, Outpatients, Theatres, Wards, etc. will all benefit from this approach.

About the Authors

Seb Kerr and Neil Milner possess extensive expertise spanning 25 years in the data, analytics, consulting and costing fields, providing software and implementation support in Activity Based Costing for over 100 organisations. Over the last 15 years, Neil and Seb have focussed exclusively on the Healthcare sector, particularly within the UK National Health Service (NHS). Drawing on their experience, they established CareCosting in 2022 with the goal of elevating standards and uncovering tangible value and benefits through Patient Level Costing and data.