[Skip to Navigation]
February 1986

Heroin vs Morphine for Cancer Pain?

Author Affiliations

From the Departments of Clinical Epidemiology and Biostatistics and Medicine, McMaster University, Hamilton, Ontario (Drs Levine and Sackett); the Department of Radiation Oncology, University of Western Ontario, London, Ontario (Dr Bush); and the Ontario Cancer Foundation Hamilton (Dr Levine) and London (Dr Bush) Regional Treatment Centers, Canada. Dr Levine is a Clinical Trials Scholar of the National Cancer Institute of Canada, and Dr Sackett is a National Health Scientist.

Arch Intern Med. 1986;146(2):353-356. doi:10.1001/archinte.1986.00360140183027

• Narcotic analgesics are the mainstay of pain control in patients with cancer. A controversy has been raging in the United States and Canada as to the legalization of heroin. We have reviewed the literature in order to determine the relative efficacy of heroin and morphine in cancer pain. We applied the following methodologic criteria: (1) Was the assignment of patients to the different opiates randomized? (2) Were all clinically relevant outcomes reported? (3) Were the patients recognizable? (4) Were both clinical and statistical significance considered? (5) Was the opiate regimen feasible in routine clinical practice? (6) Were all patients who entered the study accounted for at its conclusion? Two trials satisfied our first standard. The first, a double-blind cross-over trial, failed to meet standard 4 (the negative conclusion may represent a type 2 error) and only 21% of patients completed both treatment periods. The second study, which compared intramuscular heroin and morphine among patients with postoperative pain, failed to meet standards 3 (patients not described in sufficient detail and only tangentially related to chronic cancer pain) and 4 (type 2 error). Thus the relative efficacy of heroin and morphine in the relief of chronic cancer pain remains unknown. Randomized trials that meet all six methodologic standards must therefore be carried out for this controversy to be resolved.

(Arch Intern Med 1986;146:353-356)

Add or change institution