Current issues of ACP Journal Club are published in Annals of Internal Medicine


Cardiogenic shock after acute myocardial infarction

ACP J Club. 1992 Mar-April;116:58. doi:10.7326/ACPJC-1992-116-2-058

Source Citation

Goldberg RJ, Gore JM, Alpert JS, et al. Cardiogenic shock after acute myocardial infarction. Incidence and mortality from a community-wide perspective, 1975 to 1988. N Engl J Med. 1991 Oct 17;325:1117-22.



To examine short-term and long-term mortality rates in patients with acute myocardial infarction (MI) complicated by cardiogenic shock.


Cohort study of patients hospitalized with acute MI in 1975, 1978, 1981, 1984, 1986, and 1988.


16 hospitals in the Worcester, Massachusetts, Standard Metropolitan Statistical Area.


The 4762 study entrants included patients with a primary or secondary discharge diagnosis of acute MI and a random sample with related disorders in whom acute MI might have been diagnosed. Discharge diagnoses were reviewed and validated according to study criteria.

Assessment of prognostic factors

Cardiogenic shock was diagnosed when systolic blood pressure was below 80 mm Hg in the absence of hypovolemia and was associated with cyanosis, cold extremities, mental status changes, oliguria, or congestive heart failure.

Main outcome measures

In-hospital and long-term case fatality rates, with deaths during follow-up identified through review of hospital records and a search of death certificates.

Main results

For the 6 study periods between 1975 and 1988, crude unadjusted incidence rates for cardiogenic shock complicating acute MI remained relatively constant at 7.5%. Patients with cardiogenic shock were older and included a greater proportion of women and patients with recurrent and Q-wave type MI. More patients with cardiogenic shock had other clinical complications of MI. More patients with shock than without received digoxin, lidocaine, and other antiarrhythmic drugs. β-blockers and thrombolytic drugs were used less often in patients with shock. The in-hospital case fatality rate was higher among those with cardiogenic shock than without it (78% vs. 14%; P < 0.001). The risk for death among patients who had shock was higher than among those without shock for all study periods combined (adjusted relative risk, 11.28; 95% CI, 8.22 to 15.48; P < 0.001). Survival during the 14-year follow-up period was worse among patients who survived cardiogenic shock during hospitalization than among those who had not had it (P < 0.001).


Neither the incidence nor the prognosis of cardiogenic shock resulting from acute myocardial infarction has improved over time. Both in-hospital and long-term survival remains poor for patients with this complication.

Source of funding: National Heart, Lung, and Blood Institute.

Address for article reprint: Dr. R.J. Goldberg, Department of Medicine, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655.


Have earlier treatment and reperfusion therapies improved the picture with cardiogenic shock in MI? This provocative report suggests not. Further, its multicenter design means that referral bias could confound the results only if some Worcester-area hospitals became transfer centers for very ill patients from outside the 16-hospital network.

Assuming the study is valid, why has the incidence of shock not changed? We know that thrombolytic therapy is generally underused. Only 20.2% of Worcester patients with acute MI received it in 1988. It is, moreover, unclear whether the significant association of shock and reduced use of thrombolysis represented cause, effect, or both. We also speculate that early presentation and modern treatments may prevent cardiogenic shock in some, but keep others alive to develop it.

Turning to outcomes, MI caseloads in some Worcester hospitals were low (an average of 794 patients per year across 16 centers), raising questions of volume-outcome relationships. More use of thrombolysis might have helped, but in GISSI-I, in-hospital mortality among Killip class IV patients was 70% in both streptokinase and control groups (1). Observational studies of emergency angioplasty have been more encouraging (2). However, in Worcester, angioplasty was used in only 1.7% of patients with shock.

We suspect that more liberal use of thrombolytic drugs and emergency angioplasty is needed in these desperately ill patients. Clinical researchers, however, must clearly take a harder look at the mechanisms, incidence, management, and outcomes of cardiogenic shock in the modern era of reperfusion therapy.

C. David Naylor, MD, DPhil
Paul W. Armstrong, MD University of Toronto Toronto, Ontario