Sezione Medicina

da Leadership Medica n. 9 del 2006

Abstract: In our health system the medical errors problem has reached a high level of awareness not only because of the economical relapse on the insurance costs. The errors in the health system are often amenable to system errors, preventable if a policy is implemented in order to identify them and analyze why, how, when, where the error happens and actions are proposed to prevent other errors embracing the “learning from the errors” culture.

Key words: clinical risk management, error

The health system is a complex, adaptable and open system where many elements act together in a continuous and dynamic interaction along with other multiple heterogeneous factors. These are represented by (1) human elements, the single patients with their own specificity, expectations and pathology, the health workers, represented by a variety of professional roles (health-technical and financial-administrative) in a continuous evolution and by a plurality of specialized competences, the relatives, the volunteers, the stakeholders; (2) medical interventions represented by simple or extremely complex and risky procedures, interactions with health instruments and technology in a continuous renewal; (3) a connective, peculiar of each system, that characterizes it and is poorly deciphered and categorized, constituted by histories, unwritten rules, behaviours, affections and emotions of persons who live together for many years; (4) the strategic management with local and non local rules and regulations who governs over all the elements that have to integrate themselves and guarantee the achievement of the only objective: “taking care of the patients”.
The health system, like many other complex systems such as the aviation and nuclear energy industry, has to foresee that errors and consequent injuries are possible and not erasable events even though most of them could be checked by the system.
In some of these systems engineering – mathematical models have been adopted and perfected to control the risks coming from errors, these models however are difficult to integrate into the health system because of the particular characteristics involved, for example, the clear preponderance of the “human factor” and the interpersonal relations compared to the “mechanical and technical” components.
One of the true specific characteristics of the health system is represented by adaptability, that is the necessity of acting promptly by new behaviours in front of  an unexpected situation; this creates an instable climate where the emotional component is high and where positive actions are often generated, although sometimes these are critical and could even favour the onset of a medical error.
Within the health system new and specific risk control models have to be consequently projected whose aim is to create those conditions that would prevent errors or at least reduce the damaging consequences.
The numerous definitions that might be found in the medical literature or glossary under the headline “adverse events” or “medical error” have in common some characteristics: the error is explained as a deficiency in the system which results in the failure of the planned activities.
It represents an “unsafe action” or an “omission” with potential negative consequences on the outcomes of the treatment process.
An action is considered an “error” when the opinion of well-experienced and skilled experts (“peer review”) is unanimously expressed and judged as an inadequate behavior, independently of the consequences for the patient. Not all medical errors cause adverse events, the dimension of the medical error within the health system is bigger than the one of the adverse events. The medical error research was focused for many years within the hospitals for obvious reasons, predominantly dictated by the availability of clinical documentation and analysis of the events delimited in a well defined period.
At the beginning of the 90’s the first papers on medical error subject were written; they created an overall awareness of the problem and helped develop a methodology for individualization, analysis and subsequent control of the errors.
Harvard Medical Practice Study is the first systematic study related to this topic, and it is considered the reference epidemiologic study for the estimation of the medical errors. Approximately 30.000 case records of inpatients discharged in 1984 from 51 hospitals for acute patients in New York, were randomly selected and screened.
The psychiatric patients were excluded. Adverse events were highlighted in 3,7% out of the analyzed records; among these 2,6% had permanent disability and 13.6% caused death. The probability of adverse events increased with the age of the patients and the thoracic and vascular surgery were the most critical areas. A third of the events were represented by pharmaceutical complications and infections transmitted through the surgical cut.
On the other hand another study in Australia on a sample of 14.000 case histories proved that 16, 6% of inpatients presented an adverse event that caused disability or increased the length of stay; of these adverse events 51% were considered preventable, 13, 7% provoked permanent disability and 4,9% caused the patient’s death. Another review of the case records proved that an error was evident in a percentage variable from 3 to 7% and of these 17% was defined avoidable by the revisers (tab.1). Errors are more frequent and avoidable in Medical Units and in drug therapy. A conceptual common matrix was outlined in order to prevent and avoid medical error and adverse event,  independently of the different percentages shown by the AE (Adverse Events) studies, which are not comparable because of the health system heterogeneity and different survey methods.

Adverse event (in patients) 3-7% Surgery 47,7%
    Medicine 52,3 %
    Drug 19,4 %
Adverse event avoidable 17% Surgery 17,7%
    Medicine 37,2%
    Drug 17,7%
Errors ?    

Tab . 1  E.A. avoidable in specific area. Modified  by Marlock L.
Clinical Risk management . BMJ 1999 – Leape ed al N.Engl J Med. 1991 – 324 – 377 - 84

The execution of an ultrasound tomography on a wrong patient is an error, but there are no such harms or consequences that categorize it as an adverse event. However the error could cause an adverse event that is a negative and undesirable event, an injury or a lesion not related to the clinical conditions of the patient, but caused by the medical intervention. The adverse event is, by its nature, undesirable, non intentional, harm to the patient; if the adverse event is caused by an error it should be considered “preventable”. Errors consequent to some procedures (therapeutic, surgical or invasive) are more easily defined as preventable and the consequences are identifiable (foreign bodies left on the surgical site, reinterventions); in other fields the borderline with the complications is sometimes more uncertain, predictable is not always preventable.



% A.E Preventable / total AE

% AE severe/ death/AE

USA NEJ 1991




USA MedCare 2000




USA IOM 2002




Australia Med.JAus 1995




PCISME (int.)MJA 2002




Danimarca Uges Laeg 2001




N.Zelanda Minis.Health 2001




UK BMJ 2001




USA  BMJ 2000 pz < 65 aa




USA  BMJ 2000 pz > 65 aa




Tab. 2 : sintesi dei principali studi

In recent years are reported studies (tab.2) that demonstrate that the adverse events are evident also from the clinical documentation. They show that AE have different frequency (2,8% – 16,6%), are preventable, avoidable, and show a high percentage of deathly harms. The publication of this data has allowed  to modify health policies and to experiment new models in many health systems, not only Occidental, that are based on a new paradigm: “we learn from our errors”: when they occur we can prevent them, if we can recognize them, analyze them and approach them as errors of the health system that have to be  managed by a generative culture.
The research and experimentation of these models and instruments for preventing the errors and managing the clinical risks, involve a lot of professionals (not only health workers), and their skills. The necessary resources to approach such a complex problem are numerous and have to act synergistically with a holistic vision. Also in the last few years there has been some studies published which focused on quantifying the error outside the hospital, associated to adverse events caused by the incorrect consumption of medicines.
In Italy there are no studies published related to that topic; there are many and various experiments conducted by hospitals and regions over the years, and there are numerous press releases that estimate the harms and the deaths in the health system, using data coming from foreign health systems.
Nowadays, it seems of no interest promoting studies to define the size of the phenomenon for economical and methodological difficulties and, above all, for ethical reasons. In July, 2005, the Ministry of Health promoted an experiment, which is actually running at national level, that compels to declare and analyze the “sentinel events”. The knowledge of this data is expected to generate awareness of the phenomenon at national level, even if they have some bias.

Sentinel  events

  1. Procedure on the wrong patient
  2. Procedure on the wrong body part (side, organ or section)
  3. Inpatients suicide
  4. Medical instruments left on the surgical site which require a subsequent surgery or additional procedures
  5. Transfusion reaction resulting from ABO incompatibility (ICD9CM: 999.6 code)
  6. Death, coma or severe functional variations resulted from therapeutic errors linked to drugs
  7. Maternal death or severe illness correlated to labour or/and birth
  8. Abuse on inpatient
  9. Death within 48 hours from the birth of healthy newborn  whose weight is over 2500 g.
  10. Any other adverse event that causes death or serious injury related to the abnormal functioning of the system which provoke the loss of the citizen trust towards the health system

The risk identification is an important priority phase of the clinical risk management policy that could be approached in a proactive way with inexpensive, well-timed and effective instruments and methods.
In the error analysis within the health system the aim is the analysis of the system; we can distinguish between active error and latent error. Active error is an human error, physically made by the health worker who finds himself with the last “lit match” on the verge of the happening of an adverse event; it is related to an incorrect action, a memory deficit, an accident, for example a defibrillator abnormal functioning or an uncontrolled monitor.
Latent errors are organizational deficiencies, wrong planning, system managerial problems, lack of human resources, all errors that can be unnoticed but favour the conditions for an active error when there are fewer safeguards in the system.
So that an adverse event take place it is enough the change of an health worker with a worker with less experience or more stressed one or not well trained for those activities, or the introduction of a new device without training. The latent and active conditions are in continuous movement, dynamic, changeable, almost never reproducible. The administration of a wrong drug  is an active error, made by one or more health workers, easily esteemed as an incorrect behaviour that could cause injuries and sometimes even death.
But it is fundamental to retrace all the phases and the work processes that led to the drug administration error in order to identify all the circumstances that directly or indirectly caused the error. For example, in the case of an incorrect administration of a drug, we could identify as latent insufficiencies of the system a lack of communication when taking care of the patients, a manual prescription – transcription of the drug  name written by the health workers, a mix – up of medicines, an insufficient training of the health workers, a scarce attention given to the activities necessary at the preparation of the medicines, that should be protected from interferences and noises of the surrounded environmental factors. Such analysis consents to move the attention from the health worker to the processes, from the individual to the system, to make it possible to identify the causes of the errors and to introduce process rectification mechanisms that could reduce the probability that the same error happens again. The active error is simply detected and represents the sum of all latent errors that only by chance haven’t previously expressed themselves as adverse events because there are “protections” in the medical organizations which prevent the error situations. When an error or a nearmiss happens it becomes very complex to promptly and efficiently individuate all the latent insufficiencies of the system because the sequence of the latent errors could be so speed, differently severe and complex that it becomes difficult to detect the primary and/or secondary errors or the precise order. Very often the analysis of an error, harmful for a patient, shows a
number of latent and active errors higher than 1 because a cascade of progressive events haven’t been “miraculously” blocked and this fact is classified as an example of the “normalization of the deviance”.
There are  a lot of these examples of analyzed errors published on international “agencies” and “institutions” web sites and also local experiences confirm these analyses.

Important methods from other systems are imported and adapted to the health reality in order to perform a correct error analyses within the complex systems.
Clinical risk management is more and more interesting for the National and Regional Health System where a safety culture is developing and spreading. The importance of this problem is so high that from 2002 the Ministry of Health, as well as other international institutional organizations, started to promote numerous initiatives at national level, oriented in creating a “no blame culture” within the health organizations. The first document, (published on March, 2004) gave cultural and organizational guidelines, soliciting hospitals to undertake clinical risk management activities and educational programs.
Also in Italy medical error issue is gaining growing attention for a lot of reasons: the main aims are: the increment of the efficacy of error prevention activities and the reduction of the predictable and preventable adverse events, the comeback of the patients’ trust towards the health organizations and, of course, the economical decrease of the insurance costs and damages.
The Risk Management has the purpose to orientate the risk management policies that have to be pointed both to the reduction of the avoidable errors and the containment of their possible harmful effects, and therefore, finally, to the higher level of patients’ safety. In the health context there are planned new and specific risk control models whose aim is the creation of environments where an error is hard to be made and where the harmful consequences are limited.
An effective risk management activity is extremely complex and includes:

  1. the detection / mapping of the risk (surveys on the events and errors through system report, case histories revision, analysis of  data  and safety indicators)
  2. the development of analyses method in order to identify the causes and the insufficiencies (Root Causes Analysis, Clinical Audit, Failure Mode and Effect Analysis – FMEA, Failure Mode and Effect Critical Analysis)
  3. the elaboration of corrective acts  that have to be adopted, actively supported, supervised and verified periodically.

Fundamental components in the patient safety planning are: the estimation of the resource “time” for health operators, so that they could assimilate the “no blame culture” and spend the “right time” in the cure, the creation of specialized university courses on risk management, and the growing of an alliance net with patients’ associations.
The evaluation of risk management strategies in other health systems suggests that maintaining continuity in safety programs, independently from political change, is another winning element in reaching the efficacy; of course it is necessary to destine sufficient resources proficiently. The problem of medical errors has reached a high level of awareness also in our health system as a consequence of its economical relapse on the insurance costs, that creates a “not coming back” condition; the issue is extremely complex and cannot be tackled in a simplistic way if we wish to avoid a “defensive medicine” where deployments among the different parts are created, inhomogeneity among the various technical roles enhances as well as the interpretative variability of the issue, the cognitive asymmetry between health workers and patients worsens as well as their distance; all these relapses can only provoke a cost increase in favour of the profit of lawyers, forensic doctors, brokers and other associations.
On the contrary the involvement of all the heath components in all phases of the risk management process could be the key to reach the wished efficacy.
This could have a role not only at preparatory and motivational level but also to the culture growth of the health system.

Luciana Bevilacqua
Direttore S.C. Qualita’ M.C.Q. A.O. Niguarda
Giuseppe Vighi
Resposanbile S.S. Farmacoviglanza A.O. Niguarda


- Health System “Risk management in the Health System. Medical errors problem"
- Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al. Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I. 1991, New Engl J Med;324:370-6.
- Cohen, M. R. (ed.) 1999, Medication errors (AphA publications, Washington DC)
- Crossing the Quality Chasm: A New Health System for the 21st Century (2001)
-Institute of Medicine (IOM Reason, J. 1997, Managing the risks of organizational accidents, (Ashgate, London)
-Institute of Medicine; Assessing Medical Technologies; 1995; Washington, D.C.; National Academy Press.
-Department of Health; Building a safer NHS for patients: implementing an organisation with a memory; July 2000; Department of Health.
-Leape LL. Error in medicine; 1994; JAMA;272:1851-57.
-Reason J; L’errore umano – “Human error”; Il Mulino; Bologna; 1990
-Reason J; Human error: model and management; British Medical Journal; 2001; 320: 768-70
-Towle A., Physician and Patient Communication Skills: Competencies for Informed Shared Decision-making, Informed Shared Decision-making Project.Vancouver, Canada, University of British Columbia, 1997
-Kohn, L. T., Corrigan, J.M., Donaldson, M. S. (eds.) 2000, To err is human buiding a safer health system, (National Academy press, Washington DC)
-Vincent, C. 2001, Clinical risk management, (BMJ press, London)
-Vincent CA, Coulter A, Patient Safety: what about the Patient?, Quality in Healthcare, 11, 1, 2002,76-80