We have reviewed more than 100 articles and works and all of them show error statistics. Even so, it is very difficult to reach firm numbers. For those of us who are used to the engineering sciences, it is difficult to understand so many differences between some works and others. There´s still a lack of clarity about what we are measuring, what process, in what condition, etc.

Many times, we find numbers that seem exaggerated and we find it difficult to present them in our works, because it will generate a discredit to all the figures presented.

However, the numbers of errors are so high, when compared to other industries, that it makes no sense to discuss whether the correct value is 6 or 8%, when the desired value should be 100 times lower.

According to IATA data, during 2014, 3,100 million passengers flew and 265 died. (Small airplanes were not considered because they are not a company or industry itself).

This gives us 0.085 deaths per million of passengers (265 / 3,100 = 0.085)

If it were in the health sector, the index would be 3/1000, deaths by people treated in a hospital. (Manifesto Err is Human, page 1: 98,000 deaths for 33.6 million cases).

What gives us for the health sector, 3,000 deaths per million.

So, the comparison would be: 0.085 vs 3.000.

The ratio between the two would be: 3,000 / 0.085 = 35,300 (35,300 times more)

When a method of observation (analysis of records) gives us a value 10 times lower than another (direct observation), or an objective method like the Global Trigger Tool gives us 10 times more AE than the reported ones, we must stop arguing if value is 8 or 22. That doesn´t matter anymore, the values ​​are also exaggeratedly high in relation to other industries. That is, the value that should be and that maybe we can achieve in the future. Numbers cannot be the goal of a study, as they are equally disproportionately exorbitant.


These would be the most consensual numbers.

  • 10% of the attendance have an AE (data from ANVISA AND WHO).
  • 50% of these is avoidable (there is a great coincidence among researchers on this value)
  • 30% of hospital AEs occur in the Medication process (D. Bates)
  • 30% of medications have an error with damage.
  • 50% of drug administration AEs are severe or moderate, the rest are of light incidence.

We have found a lot of disparity of concern with Patient Safety and Quality in many countries. Even inside Europe.

Many values ​​are obtained from self-denunciations that remains a resisted topic throughout the world. Self-denounced values ​​vary a lot from one culture to another. Statistics copied this disparity. It should not surprise us.

It is considered that the indicators in developing countries would be 3 times bigger than in the developed ones.

We mean by this that indexes do not matter as much as we think. Values ​​are also very high. We must change procedures and offer to health professionals all means to help them to do their job well. Perhaps because Health is a right recognized in almost all the world and many times sustained by the governments, it does not receive as much investment as other industries.

It is incredible that many health institutions do not have a strong economic motivation to improve Quality. In many cases, if the patient must remain 10 more days due to an AE, the hospital continues to receive a payment for each busy day. In other words, you earn more with AE.



To Err Is Human: Building a Safer Health System. The title of this report encapsulates its purpose. Human beings, in all lines of work, make errors. Errors can be prevented by designing systems that make it hard for people to do the wrong thing and easy for people to do the right thing. Cars are designed so that drivers cannot start them while in reverse because that prevents accidents. Work schedules for pilots are designed so they don’t fly too many consecutive hours without rest because alertness and performance is compromised.

In health care, building a safer system means designing processes of care to ensure that patients are safe from accidental injury. When agreement has been reached to pursue a course of medical treatment, patients should have the assurance that it will proceed correctly and safely so they have the best chance possible of achieving the desired outcome.

This work, launched in 1999, was the beginning for all the great work on quality that happened over the next 15 years.

But this work also said: to err is human and natural and we cannot change it. Then we must prepare ourselves for when the error happens. We must accept that we are going to err and then we must create systems that solve that incident.

James Reason: We cannot change the human being, but we can change the conditions in which he works, to improve the results.

We emphasize the use of automated or computer systems, conducting processes, but always with the supervision and vigilance of the people.

We must give people all the necessary conditions to do a good job.