‘No ‘rules’, no service transformation
All service transformation is based on consistently delivering – at scale – better patient outcomes and experience at lower cost*. Yet the transformation of the health service, without employing software to assist, is progressively becoming more and more difficult. In the current healthcare climate we face new problems, that of an ever-growing and aging population with increasing demands*. As the need for healthcare rises, so too does the variation in the ability of services to cope with these demands*. The Atlas of Variation provides a solid grounding in the widespread disparities of service provided to different areas of the country*. Resources are scarce, and as highlighted by the NHS Five Year Forward View, ‘in a tax funded health-service, every of pound of waste is a pound that can be reinvested in new treatments and better care for the people of England.’* If we are to tackle these issues, change in services are desperately needed to face the tide of changing requirements of an evolving health economy.
But how can services effectively implement change?
Change means disruption, confusion and inevitably, resistance*. For service transformations to be executed successfully, standards need to be developed and upheld. Without clear articulation, such consistent standards cannot be adhered to; particularly when one considers the immense complexity of the healthcare system. No longer can we rely on human intuition and policies hidden in paper binders to meet the requirements of society. The diversity of the population means that there may be a wealth of patient characteristics, specialties, reasons for diagnoses and based on these areas, appropriate referral pathways that must be decided on. Consider the amount of information an individual must maintain in their minds, to act in line with all the requirements of these interconnected facets of the health service, let alone the difficulty in adjusting these concepts as policies change.
To drive change, coherent and systematic articulation of ‘rules’ and processes to monitor the implementation of such rules or guidelines is essential* . In many cases, a patient referred for x, should be directed down pathway/service y based on z. ‘It is not that simple’, some might say, ‘it is about our clinicians and administrators using their judgment and understanding of the local health economy’.
We are the first to acknowledge that it is not that simple…human decision making can be both subtle and complex. However, humans are also subject to limits which prevent them from consistently performing at the same, high level. Humans get tired and have off days –all of which ultimately means that decision making is impaired by the human condition in a way which algorithms and rules-based models are not*. The fallibility of human judgement has been amply demonstrated. One example is the Oregon study, which proves that human decision making is not always consistent. In the study, clinicians were asked to describe the cues they use to predict the risk of cancer, before being shown pictures and asked to assess the presence of cancer (at times they were given duplicates). Clinicians frequently contradicted themselves, assigning different risks to identical x-rays by basing their decisions on different cues. Using the cues mentioned by clinicians, the researchers then created a simple rules-based model which they used to predict clinicians’ assessments , proving that decisions which may at first glance appear to be complex, can be mapped out and modelled. In a second experiment, the authors went on to demonstrate how simple algorithms could also outperform the doctors, unlike the clinicians’, the algorithm could be configured to consistently account for all the relevant cues in each decision, rather than only consider those that stood out to them at the time of diagnosis*.
In many cases, the decisions made by administrators and clinicians require little critical thought but are simply based on adhering to service policies. Take the decision to reject a referral if the patient is a child referred to an adult health service, then the decision to reject that patient back to the referrer is not subjective, but a hard and fast rule. Yet, in health services all around the UK, staff are making these decisions, performing seemingly mindless, repetitive tasks with little opportunity to upskill. The nature of such tasks leaves individuals open to error . This not only prevents individuals from progressing but also reduces establishments ability to meet quotas. In such cases, the waste of resources in health services is immense.
The software solution ART (=Advanced Referral & Triage) is designed to automate almost all admin tasks in referral management centres (RMC) or single point of access (SPA). After a bedding in period, the software routinely achieves automation success rate (ASR) of 95% or higher – effectively eliminating all but 5% of admin work.
At the heart of the ART software is a rules engine. A rules engine designed to mimic and uphold standards provided at input, and a rules engine that can be configured to cope with change, simply by deciding on a new rule. Together with our customers we leverage this rules engine to codify pathway policies to suit a service’s demands in a structured and auditable way. These rules then act as pre-made decisions which can be adhered to. Should a change be necessary, it can be implemented as soon as the decision is approved, removing the bulk of tedious and complex decisions, whilst also allowing services to respond flexibly to new requirements.
Consistent human processing without the support of software rules is near impossible: An ART customer (typically an operator of an RMC or SPA) will use 200 to 400 rules relating to triage and booking. Consider the cost of training staff and keeping them up to date with such complexity! Also, think about the room for human error and the challenge of auditing such decisions.
Not only does a rule-based software approach deliver better outcomes; it delivers a safer service for patients and it lends itself to deliver successful service transformation at scale and at lower cost.
No ‘rules’, no successful service transformation.