David Healy's latest blog warns against the credulity of doctors in the face of vast marketing efforts by pharmaceutical firms. In making a general argument about how the medical profession ought to do a better job he makes the following assertion:

Treatments now are far more lethal than they were in [Oliver Wendell] Holmes’ day and treatment induced injury has become one of the leading causes of death and disability.

His focus is probably (from the context) mostly about pharmaceuticals, but might be considered more general. Is the highlighted claim true either restricted to just drugs or if it covers all medical interventions?


The word you may be looking for is "iatrogenic":

In the United States an estimated 225,000 deaths per year have iatrogenic causes, with only heart disease and cancer causing more deaths.[1]

Some iatrogenic effects are clearly defined and easily recognized, such as a complication following a surgical procedure. Less obvious ones require significant investigation to identify, such as complex drug interactions.

Causes of iatrogenesis include negative effects of drugs, chance, medical error, negligence, unexamined instrument design, anxiety or annoyance in the physician or treatment provider in relation to medical procedures or treatments, and the adverse effects or interactions of medications.

"225,000 deaths" seems to be often quoted/requoted when I Google for iatrogenic statistics. The reference in the Wikipedia article is to Is US Health Care really the best in the world (2000) reprinted from The Journal of the American Medical Association (JAMA). Those deaths break down as:

  • 12,000 deaths/year from unnecessary surgery
  • 7,000 deaths/year from medication errors in hospitals
  • 20,000 deaths/year from other errors in hospitals
  • 80,000 deaths/year from nosocomial infections in hospitals
  • 106,000 deaths/year from non-error, adverse effects of medications

In contrast, page 19 of this document from the NHS audit commission reports only about 1,200 deaths/year from medications in the UK.

Allowing for the fact that UK population is 5 times less than that of the USA, I calculate from the above that the death rate due to medications is 106000 / (1200*5) = 18 times higher in the USA than the UK: either the rate of iatrogenic death is much higher than in the States, or there's something wrong or incompatible (not comparable) with the way in which cause-of-death is reported.

Is that a big number compared to, say, heart attacks, road accidents, cancer?

The above Wikipedia article claims "iatrogenic causes" are the 3rd-leading cause of death, "with only heart disease and cancer causing more deaths".

The US Centre for Disease Control gives the following statistics:

Number of deaths for leading causes of death

  • Heart disease: 597,689
  • Cancer: 574,743
  • Chronic lower respiratory diseases: 138,080
  • Stroke (cerebrovascular diseases): 129,476
  • Accidents (unintentional injuries): 120,859
  • Alzheimer's disease: 83,494
  • Diabetes: 69,071
  • Nephritis, nephrotic syndrome, and nephrosis: 50,476
  • Influenza and Pneumonia: 50,097
  • Intentional self-harm (suicide): 38,364

Like the above, this Wikipedia article doesn't list iatrogenic causes at all, as a leading or significant cause of death; however, this other one does (it claims "44,000 to 98,000" deaths from "Preventable medical errors in hospitals", citing To Err Is Human: Building a Safer Health System (2000)).

Page 41 of these CDC statistics from 2010 state 2,490 deaths due to "Complications of medical and surgical care", plus 40,393 deaths due to "Drugs" (which number has a breakdown here and includes illegal drugs and suicides).

In summary, "non-error, adverse effects of medications" is or isn't a big number, depending on the source: I don't see it in the CDC statistics. Has the number been greatly reduced now, since it was reported in the year 2000? Is cause of death ambiguous (e.g. if someone has a heart attack because of an adverse drug effect), which affects the way in which it's reported? I don't know.

Two references (one to another JAMA article and one to a Lancet article) were given in the JAMA article which estimated "106,000" as the number of deaths due to adverse drug effects.

  • The first reference is Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies: it is criticised by Adverse drug reactions in hospitalized patients: A critique of a meta-analysis which says,

    With regard to fatal ADRs, the problem of small numbers of events is likely to introduce large errors in incidence estimates. Simple pooling of fatal event frequencies from only those studies specifically reporting the number of fatal ADRs, as was done in the meta-analysis of Lazarou and colleagues, is likely to dramatically overestimate the death rate.

    CONCLUSION: Meta-analysis was invalid because of heterogeneity of the studies. Most of these studies did not report the data needed for incidence calculations. The methodology used was seriously flawed, and no conclusions regarding ADR incidence rates in the hospitalized population in the United States should be made on the basis of the original meta-analysis.

  • The second reference is Increase in US medication-error deaths between 1983 and 1993. I haven't read it but this letter says says that it "reports an increase in medication-error deaths in the USA from 2876 in 1983 to 7391 in 1993": which is much smaller than 106,000 in 2000. The letter also says (regarding UK data) that "suspected" ADRs are 150 times higher than death-certified ADRs.

In summary the extremely high number is widely reported, but disputed in a subsequent study; and is probably based on estimates rather than officially-reported and certified cause-of-death.

| improve this answer | |
  • 1
    I've never understood how they measure this. If a gunshot victim has a 99% chance of dying through the night, so a surgeon performs an extremely risky procedure that has a 10% chance of saving them and a 90% chance of going wrong and killing them, and the patient dies on the table... what was the cause of death? Iatrogenic or gun-shot? – Oddthinking Dec 30 '13 at 23:33
  • 1
    @Oddthinking In negligence law, these would be considered "independent sufficient causes". However, the surgeon was operating within his duty of care, and thus not liable for the damage. 100% of the responsibility would be attributed to the gun shot. – user5582 Dec 31 '13 at 0:17
  • 1
    @Oddthinking I suppose a iatrogenic cause would be a surgeon performing an extremely risky procedure ...on the wrong organ. – Sklivvz Dec 31 '13 at 12:02
  • 1
    @Oddthinking I guess it's because deaths have more than one cause that statistics are hard to ascertain: for example if someone dies because of a nosocomial infection, is the cause of death filed under "nosocomial" or under "infection". – ChrisW Dec 31 '13 at 16:44
  • 1
    Given the thread of comment I think it worth pointing out that some iatrogenic causes are not ambiguous. Giving the patient the wrong drug is one or the patient dying of an infection acquired in the hospital is another. These are common and traceable to specific mistakes in care. – matt_black May 4 '16 at 12:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .