Role of AI in reducing medical error – Result of YouDiagnose Survey

Sep 24, 2020 | Clinical AI

Whether it is the case of a miscalculated dose of anaesthetics that killed a patient or an instance of wrong side surgery leading to limb loss on both sides, medical error is emerging as a serious public threat. We carried out a mixed input survey – comprising of a telephonic survey, face to face interview and web-based survey – to study people’s perception about safety in the healthcare industry and compared it with other industries. Out of 123 invitees, 93 participants took part in the survey in 2018.

Observation – 1 To a question, “Which industry is most hazardous?” most participants rated the nuclear industry as the most hazardous one. 69% of the participants labelled the nuclear industry as the most dangerous, followed by aviation (21%) and healthcare (10%).

Observation – 2 When people were asked about the reason behind what they think, most people mentioned that nuclear explosion could wipe out the human race from the face of the Earth (~16%) or other such statements about the potential harm and damage from a possible nuclear explosion.

Observation – 3 Repetitive Work Driving Medical Error – One of the critical aspects which emerged as a startling observation was repetitive tasks, which increasing error proneness. Three participants mentioned that when someone is mindlessly repeating the tasks, the brain often switches off assuming full control over the situation. This may often lead to inaccurate decision-making. All participants mentioned that no-brainer work, when given in high volume, may lead to sluggish performance and vulnerability to more errors.

Observation – 4 How AI can help? – With regards to finding a solution to error-making, a large percentage of the people favoured the use of augmented intelligence or artificial intelligence in patient care to assist and support in the decision-making process. The joint model of working offers a human validation to a machine-led decision and machine validation to a human decision, thereby creating a man-machine synergy.

Observation – 5 Survey Results on AI’s potential role in medical error – 46% of the participants had a strong feeling that futuristic technologies will play a decisive role in solving the problems of medical error, 24% participants said they neither favour nor disfavour futuristic technologies and the remaining 30% of participants thought AI might play a decisive role.

Risk Perception Vs Ground Reality

However, ground realities are far away from people’s perception. The below-mentioned figures show that the most hazardous industry is the healthcare industry.

  • Aviation – 1 in a million chance of a person being harmed
  • Nuclear – 400 deaths per 20,000 years of reactor operation
  • Healthcare – 250,000 deaths annually in the USA

Most of the respondents who had chosen differently were utterly surprised when they heard the real figures, which are as follows:

Incidence of Medical Error

  • 750,000 harm-inflicting medical errors per year in the EU
  • 3.2 million days of hospitalisation due to medical error in EU
  • 260,000 fewer incidents of permanent disability in the EU
  • 95,000 fewer deaths per year in EU
  • 850,000 adverse events in a year in the UK
  • 10% of hospital admissions in the UK were medical errors and health-care related adverse events
  • WHO says that there is 1 in 300 chance of a patient being harmed when receiving healthcare.

Patient Survey on Medical Error Experience

  • 23% of European Union citizens claim to have been directly affected by a medical error
  • 18% claim to have experienced a serious medical error in a hospital and 11% claim to have been prescribed a wrong medication
  • 50% to 70.2% of such harm is preventable.

What are the ways to reduce medical errors?

Attention is the most valuable asset an individual has. In today’s complex and demanding healthcare environment, time is no longer an organisation’s or an individual’s most valuable asset; it’s attention.

In non-medical professions, attention is defined as what we each choose to focus on and what we decide to ignore. However, in a healthcare organisation, that is certainly not a huge possibility, given the sensitivity of the environment. Mental triaging of the circumstances and tasks is dependent on one’s experience and environmental factors. Thus, reducing medical error and maintaining patient safety standards needs a tremendous amount of attention which is always in limited supply.

Categories of Medical Errors

For simplification, we have categorised medical errors into the following groups:

  1. Intentional Error – A person wilfully commits a mistake due to exceptional circumstances.
  2. Human Errors
    • Executive Errors – Unintentional Errors, e.g. slips or lapses in memory or action
    • Mistakes
      • Gap in knowledge
      • Lack of Consistent Exposure or Training
      • Systemic Issues – a gap in rota, high patient waiting time, communication issues, lack of induction
      • Personal Stress – Illness or stress
      • Equipment Issues – Faulty instrument or inadequate instructions or training to operate
      • Heavy Work-Load

Root Cause Analysis of Medical Error

Every step involved in patient care requires active attention and decision-making, which reduces in a subtle way when the tasks are repeated, and the work is overwhelming. The medical errors are attributed to the following causes:

  1. Intensity of work (10%)
  2. Multiple parallel tasks (7%)
  3. Chronic staff deficit and a consequent need to cover the job of other people (14%).
  4. Gap in knowledge (8%)
  5. Repetitive work (15%)
  6. Chain reaction and cascade effect (7%)
  7. Career burn-out (12%)
  8. Lack of communication or coordination(5%)
  9. Technical failure of devices or systems(4%)
  10. Human error of judgement(18%)

When we analyse the responses, we find that the work-pressure to human ratio contributes about 31%, while the human error of judgement was the second most common cause (18%). Out of the rest, repetitive work and workplace frustration, for example, career burnout were the third and fourth most common cause behind medical errors.

How do we aim to reduce the incidence of medical errors?

The medication error was one of the critical issues affecting patient safety in nursing practices. The medication administration error was significantly reduced when the “double-checking of medication” practice was introduced. In this practice, two humans cross-check the medication content, expiry date and route of administration before administration. This is a routine practice in the National Health Services of the United Kingdom. If this practice can be adopted in patient care and decision making, it has a vast potential to reduce medication error.


Dr Arti Garg, Consultant Surgeon, Bart’s Health NHS Trust, says,

“Double-checking is a good practice to reduce the incidence of human errors in patient care. This is the main reason behind the cancer care MDT (multi-disciplinary team decision making) practice in cancer care in UK hospitals. But there is a paucity of doctors. NHS is already struggling with a staff shortage and therefore, getting additional human resources for double-checking is not possible.

The shortage of trained healthcare professionals is a global issue. The World Health Organisation predicts, “The global economy is projected to create around 40 million new health sector jobs by 2030, mostly in middle-income and high-income countries. Despite the anticipated growth, there will be a projected shortage of 18 million health workers needed to achieve the UN Sustainable Development Goals (SDGs) in low-income and lower middle-income countries, fuelled in part by labour mobility, both within and between nations.”

Prof Bijen Patel, Academic Surgeon from Queen Mary University and University College Hospital, London, says,

“It takes 15-20 years to produce a fully qualified and experienced consultant. We can’t resolve the huge staff shortage issue at this pace. The only way we can move forward is to make efficient use of technologies to develop a synergy between man and machine. This will not only address the issue of chronic shortage of healthcare staff but will also improve the healthcare practice standards by reducing medical errors.”

YouDiagnose and Medical Error

YouDiagnose is the world’s first artificial intelligent cancer-care solution which will provide a review of cancer diagnosis and treatment process, and thus, “double-check” the decision-making process. This will validate the decision taken, offer new options and alternatives, empower the patient, improve patient compliance and reduce the incidence of medical errors.