According to the European Medicines Agency, medication errors account for roughly two million deaths per year worldwide.

The global price tag of medication errors has been estimated at $42 billion annually – however healthcare professionals only report 14 per cent of them.

Kyle Malone MPharms MPSI, Ph.D. Researcher, School of Pharmacy, University College Cork, tackles this data and shares how a platform for improvement to patient outcomes has emerged via the concept of a Learning Healthcare System.

Internationally, the study of healthcare systems has become increasingly common. Annually, the World Health Organisation (WHO) alone compiles data on dozens of healthcare metrics—from life expectancy to financial fairness, to overall health expenditure. At a glance, cross-country comparisons between nearly 200 nations can be made.

Nevertheless, lost amongst the decimal points and the soccer-style league tables are the answers to the most fundamental questions.

  • Why do we collect healthcare system data?
  • What, if anything, can we do with the results?
  • Are there improvements being made and better patient outcomes as a result?

Despite the many differences in the provision of healthcare globally, overall some stark realities cannot be escaped.

In 2018, the worldwide demand for healthcare services continued to rise. At current growth rates, experts predict that by 2050 many countries will spend more than 20% of gross domestic product (GDP) in this sector.

When we speak of such increased demands on healthcare systems, what we really mean is an increased call for scarce resources. Rarely do the demands on healthcare systems shift from short-term, transactional goals (more significant investment, faster access to new therapies) to more long-term issues of structure (quality, innovation, efficiency).

The bottleneck, of course, is that the primary function of all healthcare systems—from single organisations to entire nations—is to meet the day-to-day needs of patients. These needs are increasing, but the resources to meet the requirements are decreasing in many cases.

If at the same time, a healthcare system wishes to root out inefficiencies and operate according to best practice, it needs to doggedly comb its past in search of lessons for its future.

In short, a healthcare system needs to learn.

The Learning Health System

The concept of a Learning Healthcare System, or more broadly, a Learning Health System (LHS) is itself not new.

First advanced in academic circles by the U.S. Institute of Medicine in 2007, Learning Health Systems have since gained global traction.

Unsurprisingly, the rapid evolution of the LHS concept has occurred against the backdrop of a decade where we’ve witnessed an explosion in health-related data.

After all, information (be it biomedical or otherwise), is a central tenet in the Learning Health System philosophy. Through a series of loops (commonly termed “learning cycles”), information gathered on a health-related issue can be transformed into improvements on current practice.

The first step of the learning cycle (see figure below) assembles data and converts it into knowledge (D2K).

The Learning Cycle

This knowledge is then used to influence performance (K2P). If a change in performance then occurs, the new data is documented to kickstart the next round of the learning cycle.

In this way, even small amounts of initial information can lead to significant improvements in a system.

By definition, a Learning Health System has been described as “sociotechnical systems for continuous improvement and innovation” that result in “higher quality, safer, and more efficient care”.

In essence, though, any health system (at any scale) could become a learning health system, so long as it committed to routinely study and improve itself through the model described above.

That said, several fundamental properties distinguish Learning Health Systems from their more conventional counterparts:

  1. The characteristics and experience of every participant are available to learn from (Data to Knowledge)
  2. Best practice knowledge is immediately available to support decisions (Knowledge to Performance)
  3. Improvement of the system is continuous through ongoing study
  4. Infrastructure enables improvement to happen routinely and with an economy of scale
  5. All the above is part of the culture

A vital cornerstone of the Learning Health System philosophy is the availability of a shared infrastructure.

Without one, each learning cycle requires its own technology, policies, staffing and analytics.

In an LHS, however, these building block services are instead shared across learning cycles. As a result, a Learning Health System can function at a fraction of the overall cost, while still addressing a variety of unique health problems.

The multitude of other benefits Learning Health Systems offers only become apparent when we consider one of these issues in more detail.

The Science of Health Learning

The World’s first medical school academic department dedicated to the Learning Health Sciences department was created at the University of Michigan in the U.S.

Professor Charles Friedman, the Chair of the Department of Learning Health Sciences and Editor-in-Chief of the Learning Health Systems Journal, believes the principles of the LHS approach can be applied to accelerate reduction health problems such as medication errors.

“If we continue to rely on journal articles as the means of dissemination, there will continue to be a 10, 15, 17-year latency before learning makes its way into practice. We need to find a way to interpret findings and codify knowledge. Automated systems could codify knowledge with machine executable guidelines and representations of information. These are potential methods that could be used to translate knowledge back into the efferent aspect of the cycle.”

Source: The Learning Healthcare Project: Professor Charles Friedman Interview

According to the European Medicines Agency, medication errors (preventable events that may lead to inappropriate medication use or patient harm) account for roughly 2 million deaths per year worldwide.

Similarly, in the paper, Medication Errors: Technical Series on Safer Primary Care, published by the World Health Organization in 2016, it states that in the UK it is estimated that “38% of all primary care patients over 75 may be affected by a medication error”. [Note that link will open full PDF publication.]

Over 500 million medication errors occur in Europe every year, making the issue an unwanted yet common reality. As well as being a major source of patient harm, medication errors are also associated with significant financial costs.

[click_to_tweet tweet=”Over 500 million #medicationerrors occur in Europe every year! Learn More Here.” quote=”Over 500 million medication errors occur in Europe every year” theme=”style3″]

Globally, the price tag of medication errors—including the costs of litigation and hospital admission—has been estimated at $42 billion annually.

Evidently, the issue of medication errors could benefit from the Learning Health System approach.

However, despite the staggering statistics noted above, healthcare professionals only report a paltry 14% of medication errors, citing time required as the main barrier.

This vastly reduces the amount of data entering the learning cycle.

Where data is gathered, limited analysis occurs, often because the information is only available in paper form. For the small amounts of data that do reach the interpretation stage, their final destination will almost certainly be the pages of a medical journal.

Yet, in the U.S. alone, more than 800,000 medical studies are published each year. As a result, information which should be immediately applied in clinics becomes lost in a sea of ever-growing biomedical knowledge.

On average, it will take the same information seventeen years to swim the path between publication and practice.

A Cycle of Continuous Medication errors

For a Learning Health System to successfully tackle the issue of medication errors, a fully-integrated infrastructure is required.

In terms of policy, several leaps forward have already been made, most notably the World Health Organization’s commitment to halve medication-related errors by 2020.

Now, the stage is set for the introduction of technological solutions on a global scale.

The World’s First Global Learning Health System

Pharmapod has developed the first Global Learning Health System for reducing medication errors.

[click_to_tweet tweet=”Pharmapod has developed the first Global Learning Health System for reducing medication errors. #WePh ” quote=”Pharmapod has developed the first Global Learning Health System for reducing medication errors.” theme=”style3″]

With over 10,000 pharmacies using its cloud-based platform internationally, it is providing the necessary infrastructure to enable improvement to happen routinely and with economies of scale – a key requirement of a true Learning Health System.

The experiences of all the pharmacy teams are available to learn from.

They are empowered to identify risks in their practice, share their learnings and prevent recurrence of patient harm.

In terms of medication errors, the platform offers an immediate advantage over traditional methods.

Currently, where online systems exist on a national basis, no feedback on errors is given to the reporter, and the added value for busy healthcare professionals is negligible.

Pharmapod is engaging and captures information in a format that allows for highly efficient, interrogable data analysis and interpretation that can be used to detect national and international trends and system risk.

Rather than gathering data to satisfy regulatory requirements alone, healthcare professionals can now engage in the Learning Health System approach.

Information about medication errors and near-misses—from single pharmacies to multiple sites—can be regularly collated. From there, the platform can carry out effective root-cause analysis and offer pharmacists tailored feedback on what aspects of practice could be improved.

Finally, when actions are taken, the platform can evaluate the impact on practice, thus completing the learning cycle.

In this way, the Pharmapod platform can get the right information to the right person(s) at the right time.

More importantly, Pharmapod is creating a Global Learning Health System of the future.

Pharmapod’s Founder, Leonora O’Brien, is a pharmacist. This is key to its success.

One of the reasons IT systems in Healthcare have limited benefit is because when they are developed commercially, they are very rarely developed by clinicians.

Dr. Charles Safran, Chief Division of Clinical Informatics at the Beth Israel Deaconess Medical Centre and Associate Professor of Medicine at Harvard Medical School explains that:

‘In the US, about 85% of the benefit of using IT systems tend to accrue to others in the system, not to frontline staff. Historically these systems are designed to help with the financing of healthcare, such as facilitating billing. There’s very little attention to clinicians and nurse’s workflow – what are the tasks that clinicians and nurses need to carry out and what do they need to know’.

On the purpose of Pharmapod, Leonora O’Brien explains

“our global platform gets the data to those who need it – the first step has been to surface insights to practitioners and leaders in their organisation responsible for standards and process improvement.

Our vision is to build a community and learning exchange for as many diverse stakeholders within the healthcare network as possible and facilitate more real time guidance.

We simply provide the tools and infrastructure which enable the community to share, learn and grow. We are also linking data to researchers and academics as a true LHS needs to bridge the gap between research and patient care.

In our mission to shorten the cycle between learning and positive change, we are now involved in development of a Global Research Advisory Board to harness the high degree of collaboration and motivation to solve this problem that’s out there internationally.”


Melissa Sheldrick and Andrew SheldrickMelissa Sheldrick, a mother who tragically lost her 8-year-old son through medication error in 2016 has been involved in the Pharmapod training in Ontario, Canada.

Melissa explains how the right culture must be present to enable a Learning Health System to flourish.
“The objective must be to adopt a ‘systems approach’ and not a ‘person approach’ when dealing with incidents. A ‘person approach’ focusses on what the person did wrong and blames them whereas a ‘systems approach’ concentrates on examining the conditions the person works under, the training that is in place and the supports they have.

This system aims to build preventative measures into the healthcare organisation’s processes so that the same errors do not recur. Learning from errors can only be effective when there is a culture at work that does not apportion blame but instead strives to support your development as a professional. Analysing and collaborating on how to prevent incidents from occurring or reoccurring is an essential piece of a pharmacy team’s work.

As Pharmapod is an engaging technology dedicated to solving this problem, healthcare teams can use it to its fullest capabilities. IT systems must be this ‘user-friendly’ so that professionals and colleagues are encouraged to share learnings quickly and simply, as well as share their ideas about how to improve their practice. It’s only then, that professionals can make long lasting changes that prevents future incidents.”

Follow this link to learn more about the Pharmapod incident management platform