Epidemiology of medical errorBMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7237.774 (Published 18 March 2000) Cite this as: BMJ 2000;320:774
All rapid responses
Editor - Weingart et al have written an excellent review on the
epidemiology of medical error (1). We would like to point out additional
ways of addressing patient and consumer safety, both analytically and
For healthcare consumers, it is important to avoid not only adverse events
related to specific medical errors, but also adverse outcomes where a
causal relationship to errors in practice cannot necessarily be
established. Clinical decision-making is complicated and often includes an
element of "normal" risk taking. The level of risk regarded as acceptable
depends on the clinical situation, and may vary from doctor to doctor,
patient to patient, and between a patient and his or her doctor (2).
Information about the factual risks associated with clinical interventions
is a necessary prerequisite for rational decision-making.
The possibilities to find facts about the risks of adverse events depend
on the features of the healthcare system generating the events. In Sweden,
a national Patient Insurance Scheme gives economic compensation to
patients who have incured physical or mental injury as a consequence of
medical treatment. Compensation is granted regardless of medical
responsibility or malpractice. Other Nordic countries have established
parallel insurance schemes. Since the Swedish insurance began in the
1970s, more than 100.000 claims have been filed and approximately 40
percent of them have been compensated economically. Data about these cases
are available in an extensive numerical database. Although we do not know
how large a proportion of all patient injuries are reported to the
insurance, this vast database offers possibilities to analyse, eg, injury
profiles for different types of healthcare units and the severity of
consequences of different types of errors. As opposed to studies of
hospital data, the database also permits analyses of outpatient care,
comparisons between different levels of care, and studies of trends over
Previous analyses have addressed, eg, specific types of medical error (3),
injuries associated with the use of selected healthcare technologies (4),
and gender disparities in the quality of care (5). A more novel approach
is to study reports from patients with selected chronic diseases, such as
diabetes, and to use the results in informing and educating both
practitioners and patients.
As the risk of adverse medical outcomes will never fall to zero, the least
we can do for consumers is to provide them with available information
about the relative safety of their options in everyday health care.
Pia Maria Jonsson,
Karolinska Institutet, Dept. of Public Health Sciences, S-171 76
The Swedish Patient Insurance Fund, PSR, Box 17830, S-118 94 Stockholm
Karolinska Institutet, Dept. of Public Health Sciences, S-171 76 Stockholm
1. Weingart SN, Wilson RM, Gibberd RW, Harrison B. Epidemiology of
medical error. BMJ 2000;320:774-777.
2. Gafni A, Charles C, Whelan T. The physician-patient encounter: the
physician as a perfect agent for the patient versus the informed treatment
decision-making model. Soc Sci Med 1998;47:355-6.
3. Johansson H, Räf L. A compilation of "diagnostic errors" in
Swedish health care. Missed diagnosis is most often a fracture.
Lakartidningen 1997;94:3848-50. (Swedish.)
4. Räf L, Claes G. Complications are frequent after surgery for
excessive hand sweating. Patients should be informed about the risks.
Lakartidningen 1999;96:930-2. (Swedish.)
5. Jonsson PM, Räf L. Is quality of health care for women worse than
for men? Two out of three insurance claims concern women. Lakartidningen
Competing interests: No competing interests