the patient's consent, but, often there isn't time to inform the patient, as when a life-saving procedure is required but the patient is unconscious, confused or irrational. An example is treatment for toxic psychosis. About 95% of induced abortions are being provided not because of medical indications but because patients so choose. By recommending or performing an abortion only because it is a woman's choice, physicians are demeaning themselves and their profession. If physicians deliver babies or destroy them in utero when and how patients choose, the patient becomes a customer and the physician a salesperson or a mechanic. Philip G. Ney, MD 1958 Fort St. Victoria, BC
Reference 1. Ney PG, Wickett AR: Mental health
and abortion: review and analysis. PsY'chiatr J Univ Ottawa 1989; 14: 506-516
Resuscitation of the terminally ill T | ahe articles and letters on this topic recently published in CMAJ are put into perspective by the case pending in Cincinnati, Ohio, in which an 84-year-old man is suing a hospital for resuscitating him against his specific directives to the contrary (Toronto Star, Mar. 19, 1990: 2). Although the hospital claims that the event and the outcome were "an act of God", it appears that the man's case is very strong legally. More important, he has made a point about the "right to die" in a humane and gentle manner. Medical technology is being misused especially for cardiopulmonary resuscitation in people with terminal illnesses and in frail 924
CAN MED ASSOC J 1990: 142 (9)
aged people requiring institutional care because of multiple complex and debilitating medical problems. The fact that the technology is available should not mean that it is used indiscriminately, which at present seems to be the rule rather than the exception. Michael Gordon, MD, FRCPC Medical director Baycrest Centre for Geriatric Care North York, Ont.
Making hockey safer I have one objection to the recommendations of the Canadian Academy of Sport Medicine (CASM) in their Position Statement on Violence and Injuries in Ice Hockey, as reported by Patrick Sullivan in his timely and excellent article "Sports MDs seek CMA support in bid to make hockey safer" (Can Med Assoc J 1990; 142: 157-159). I do not agree that body checking should be banned at the peewee level (ages 12 and 13) and below. I believe that elite players (such as those on all-star teams), who account for perhaps 15% or 20% of all players, should be
taught body checking techniques from the day they begin to play elite hockey. All other players, those taking part in what I call recreational hockey, should not be allowed to body check from the time they first play hockey through all levels into old-timers" hockey. We know from the statistics so well reported by Dr. Charles Tator and colleagues' that catastrophic hockey injuries occur at an average age of 17 years and an average weight of 77 kg. Waiting to introduce body checking at age 14 or, as Dr. James Sproule, chairman of the CASM's Hockey Safety Committee, suggests, 16 or 17 years is hazardous. Suddenly elite players entering the age bracket in which spinal cord inju-
ries occur are asked to change their style of play. Would it not be wiser for these players to have been taught proper body-checking techniques from the start? I was a dissenter when the body-checking policy was being formulated at the CASM meeting in Banff in March 1989, and I still am. I suggest that the CMA speak to Murray Costello, president of the Canadian Amateur Hockey Association, and John Gardner, president of the Metropolitan Toronto Hockey League, before supporting this one clause in the otherwise excellent CASM proposal. Thomas J. Pashby, MD, CRCSC 215-20 Wynford Dr. Don Mills, Ont.
Reference I.Tator CH, Edmonds VE, Duncan EG et al: Danger upstream: catastrophic sports and recreational injury in Ontario. Ont Med Rev 1988; 55: 7-12
Antibiotic therapy for acute otitis media T n he validity of the study reported by Dr. William Feldman, Ms. Theresa Sutcliffe and Dr. Corinne Dulberg (Can Med Assoc J 1990; 142: 115118) in their article "Twice-daily antibiotics in the treatment of acute otitis media: trimethoprimsulfamethoxazole versus amoxicillin-clavulanate" is questionable since no bacteriologic specimens were taken. It is very difficult to prove the efficacy of antibiotic therapy without culture results. The article is vague as to how the diagnosis of otitis media was made. Also, the reliability of the nurse's examination is questionable since the accuracy of the diagnosis was not measured by, for example, comparing tympanometry or reflectometry results with tympanocentesis results. The 95% rate of agreement be-
Third, it is not important to know what the bacteria are at the outset and after treatment. What it is important to know is whether clinical improvement is apparent subjectively and objectively. There is no reason why our nurse clinician should have compared the reflectometry results with tympanocentesis results. As we pointed out in the article, others, using tympanocentesis as the gold standard,' have already shown reflectometry to be a highly sensitive and specific instrument. Dr. Schloss errs in saying that we found cure rates of 93% and 82% after 10 days of treatment with trimethoprim-sulfamethoxazole and amoxicillin-clavulanate Melvin D. Schloss, MD, FRCSC respectively. What we said was Chairman that those rates were for "cure or Department of Otolaryngology improvement", improvement McGill University being diagnosed mainly in chilMontreal, PQ dren who felt better and whose Reference tympanic membranes looked better but in whom reflectometry still 1. Teele D, Klein JO, Rosner DA: Epide- showed fluid. It is well known that miology of otitis media in children. Ann for most children residual fluid Otol Rhinol Laryngol 1980; 89: 5-6 spontaneously resolves. As to the high incidence of diarrhea with amoxicillin-clavul[Dr. Feldman replies.] anate, we were merely reporting About 12 years ago, when we the facts. began a series of randomized controlled trials of antibiotic therapy William Feldman, MD, FRCPC for otitis media we decided not to Professor Departments of Pediatrics, Epidemiology obtain culture specimens, for and Community Medicine Head of pediatric ambulatory care three reasons. First, obtaining such speci- Children's Hospital of Eastern Ontario mens requires sticking a needle Ottawa, Ont. into the middle ear - a painful, Reference frightening procedure that we could not support on ethical 1. Teele DW, Teele J: Detection of middle grounds. ear effusion by acoustic reflectometry. J Pediatr 1984; 104: 832-838 Second, most primary care physicians do not do tympanocentesis in diagnosing and treating childhood otitis media. In fact, I would question the validity Placebos: some ethical of studies that do involve this considerations procedure, since the parents who D_ r. Eike-Henner Kluge's would give informed consent editorial (Can Med Assoc must be a small minority; thus, J 1990; 142: 293-295) the results could hardly be related to the average child seen in a prompts me to pose a few quesprimary care practice. tions.
tween pediatrician and nurse could be questioned too. The cure rates of 93% and 82% after 10 days of treatment with trimethoprim-sulfamethoxazole and amoxicillin-clavulanate respectively are rather surprising, for others have shown that in 50% of children the middle ear fluid has not cleared 1 month after treatment.' The high incidence of diarrhea (78%) with amoxicillin-clavulanate is rather surprising too. This figure is much higher than that reported by the pharmaceutical company producing this antibiotic (C.D. Bluestone: personal communication).
In the first paragraph Kluge defines a placebo as "any substance, agent or procedure that is causally ineffective for the diagnosed condition but that nevertheless is used in such a way as to allow a patient to believe that it is specific for the condition". My questions: Is the defined condition the total explanation of what the patient is suffering from? Should the patient be made to believe that the doctor and medical science have the entire explanation of that condition? Is an illness not associated with some unknowns? If the doctor communicates this to the patient, does he not maintain his integrity and that of the profession better than if such doubts and questions are dismissed as nonexistent? Placebos may provide scientific and patient knowledge, but their use is an experiment and a violation of the bond of integrity that is supposed to be part of ethical medical practice. In the final paragraph Kluge states that "there is no final word in ethics or in medical practice; at best, there is an ever-closer approximation to an ideal". My question: Is there an ideal outlook for the patient based purely on the knowledge available through medical science or behavioural science? I doubt it. In our society there are immense pressures, collective ones, that appear to be pushing for the presence of a medical Napoleon to conquer and rectify everything in a revolutionary sweep. Some of this pressure may arise from the absence of basic integrity in communication on medical matters at all levels of our society, a failing that could be undermining the confidence of the patient in the doctor or of the public in the medical profession. The patient facing an illness needs the physician's integrity as a support; this he can feel. Although enormous moral and ethical questions facing our profession have CAN MED ASSOC J 1990; 142 (9)