Letters · Correspondance

CMAJ 1999;160:1551-6

• The evidence for insulin lispro
  L.D. Grossman; response: J. McCormack, K. Bassett
• Adjudicating ethics in research
  G. Crelinsten; J.V. Frei; response: F. Rolleston
• Chiropractic and orthodoxy
  C.A. Gauthier
• PMAC Code of Marketing Practices
  J. Lexchin
• Calling up the troops
  E. Berger
• Angry scientists fight university's attempt to affiliate with chiropractic college [Correction]

In their critique of the review article by Anuradha L. Puttagunta and Ellen L. Toth [abstract / résumé]¹ on insulin lispro therapy, James McCormack and Ken Bassett [letter]² have inaccurately interpreted the existing clinical data on our product. They write, "Postprandial blood glucose levels were lower in the insulin lispro groups, but the clinical significance of this difference must be questioned, given that it did not translate into a reduction in the incidence of hypoglycemic episodes or levels of hemoglobin A_1c." In fact, Anderson and colleagues³ showed a significant difference between insulin lispro and regular insulin therapy in the mean number of hypoglycemic episodes (3.18 v. 3.43 episodes per 30 days per patient, p < 0.02). Overnight hypoglycemia was reduced significantly as well (0.47 episodes per 30 days per patient for insulin lispro v. 0.73 for regular insulin, p < 0.001).

Zinman and colleagues⁴ also demonstrated a significant reduction in hypoglycemia with insulin lispro compared with baseline (12.7 episodes per 30 days at baseline v. 8.6 after 3 months of insulin lispro, p = 0.035). This was associated with improved levels of hemoglobin A_1c. A recent meta-analysis by Brunelle and colleagues⁵ also clearly demonstrated a 30% reduction in severe hypoglycemia with insulin lispro compared with regular insulin (3.3% v. 4.4%, p = 0.024). Insulin lispro is the first insulin product capable of improving the hemoglobin A_1c levels with fewer hypoglycemic episodes, rather than more, as occurred in the Diabetes Control and Complications trial.⁶

The issue of the effect of insulin lispro therapy on hemoglobin A_1c also needs clarification. McCormack and Bassett quote the meta-analysis by Davey and colleagues⁷ as evidence that there is no significant difference between insulin lispro and regular insulin in this respect. However, they fail to point out that Davey and associates alluded to the fact that the studies evaluated for the meta-analysis were not necessarily designed to demonstrate significant changes in hemoglobin A_1c levels. More recent studies, which were specifically designed to examine the effect of insulin lispro on hemoglobin A_1c and which directed proper attention toward adjusting the basal insulin and meal plan in conjunction with insulin lispro therapy, have demonstrated the ability of insulin lispro to provide for a lower hemoglobin A_1c level with fewer hypoglycemic episodes.^8-11

Finally, a number of the abstracts that McCormack and Bassett dismiss as not providing sufficient detail have now, in fact, been published as full papers.^5,8

Numerous studies of insulin lispro therapy have demonstrated lower postprandial glycemic excursions, fewer hypoglycemic episodes overall, and fewer severe and overnight hypoglycemic episodes. More recent reports have demonstrated lower hemoglobin A_1c levels without an increased risk of hypoglycemia.

Loren D. Grossman, MD
Associate Vice-President
Clinical Research
Eli Lilly Canada Inc.

Competing interests: Dr. Grossman has financial interests in Eli Lilly Canada Inc.

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Puttagunta AL, Toth EL. Insulin lispro (Humalog), the first marketed insulin analogue: indications, contraindications and need for further study. CMAJ 1998;158(4):506-11.
2. McCormack J, Bassett K. The evidence for insulin lispro [letter]. CMAJ 1998;159(11):1353-4.
3. Anderson JH Jr, Brunelle RL, Keohane P, Koivisto VA, Trautmann ME, Vignati L, et al. Mealtime treatment with insulin analog improves postprandial hyperglycemia and hypoglycemia in patients with non-insulin-dependent diabetes mellitus. Multicenter Insulin Lispro Study Group. Arch Intern Med 1997;157:1249-55.
4. Zinman B, Tildesley H, Chaisson JL, Tsui E, Strack T. Insulin lispro in CSII: results of a double-blind crossover study. Diabetes 1997;46:440-3.
5. Brunelle RL, Llewelyn J, Anderson JH Jr, Gale EAM, Koivisto VA. Meta-analysis of the effect of insulin lispro on severe hypoglycemia in patients with type 1 diabetes. Diabetes Care 1998;21:1726-31.
6. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977-86.
7. Davey P, Grainger D, MacMillan J, Rajan N, Aristides M, Gliksman M. Clinical outcomes with insulin lispro compared with human regular insulin: a meta-analysis. Clin Ther 1997;19:656-74.
8. Ronnemaa T, Viikari J. Reduction of snacks when switching from conventional soluble to lispro insulin treatment: effects on glycemic control and hypoglycemia. Diabet Med 1998;15:601-7.
9. Ebeling P, Per-Anders J, Smith U, Lalli C, Bolli GB, Koivisto VA. Strategies toward improved control during insulin lispro therapy in IDDM. Diabetes Care 1997;20:1287-9.
10. Jansson P, Ebeling P, Smith U, Conget I, Coves M, Comis R, et al. Glycemic control in IDDM is improved by optimized combination of insulin lispro and basal insulin. Diabetes 1997;46(1 Suppl):162A.
11. Sindaco P, Ciofetta M, Lalli C, Torlone E, Perriello G, Brunetti P, et al. Importance of basal insulin treatment of IDDM with lispro insulin. Diabetes 1997;46(1 Suppl):331A.

The table included as part of our critique [letter]¹ of the review article on insulin lispro therapy by Anuradha L. Puttagunta and Ellen L. Toth [abstract / résumé]² does not contain any inaccuracies. In their Table 2, Anderson and colleagues³ reported the risk of hypoglycemia at end point as virtually identical for regular insulin and insulin lispro therapy: 2.79 and 2.67 episodes per 30 days for regular insulin and insulin lispro respectively (p = 0.31). In the text, they did report that the mean hypoglycemic rates, which were inexplicably different from the end-point hypoglycemic rates, were different between the 2 treatment groups. However, the absolute difference in hypoglycemic episodes between the 2 groups was almost identical to the absolute differences between the groups at baseline. Our concern with this study is not with the number of hypoglycemic episodes reported, but with its validity. It used an open study design to determine the frequency of subjectively reported mild hypoglycemic episodes.

Blinded trials are needed to assess hypoglycemic episodes as well as more objective end points such as hemoglobin A_1c levels. Blinding requires giving both forms of insulin at the same time before meals. Zinman and colleagues⁴ have shown that blinded trials are safe and therefore ethical. They did not find a statistically significant difference in rates of hypoglycemic episodes between the regular insulin and insulin lispro groups. It is difficult to extrapolate from their findings because patients used continuous insulin infusion technology and intensive assessment.

The strength of a meta-analysis depends on the strength of the trials on which it is based. The primary problem in the insulin lispro literature lies in determining the strength of the individual trials. The meta-analysis by Davey and colleagues, for example, included 8 unpublished trials involving 2500 patients and 3 published trials involving 30 patients.⁵ An additional problem is that the published trials are methodologically weak. The meta-analysis by Brunelle et al,⁶ which at the time was not available to us or to Puttagunta and Toth when they wrote their article, was based on open trials. Accepting or rejecting its conclusions, which showed a decrease in the number of severe hypoglycemic episodes (1.3% absolute difference) in favour of insulin lispro but no benefit for insulin lispro with respect to hemoglobin A_1c, must await full publication of valid, randomized blinded studies.

Loren Grossman lists 4 additional references to defend his statement that "more recent studies ... have demonstrated the ability of insulin lispro to provide for a lower hemoglobin A_1c level with fewer hypoglycemic episodes." However, the studies by Ebeling and colleagues⁷ and Ronnemaa and colleagues⁸ rank among case series, the methodologically weakest form of evidence. They are trials of switching from regular insulin to insulin lispro therapy. Ebeling and colleagues showed an 0.8% reduction in hemoglobin A_1c levels and no difference in hypoglycemic events. Ronnemaa and colleagues found no difference in hemoglobin A_1c or hypoglycemic events until they looked at a subset of patients. As Ebeling and colleagues correctly state, "it is not possible to estimate how much of the improvement in glycemia and HbA_1c was due to insulin lispro ... or how much is due to the intensive attention the patients were given during the study." Grossman's last 2 references are to abstracts, and 1 of these describes research involving only 6 patients.

It is important to reiterate that our original letter was meant to criticize CMAJ's decision to publish Puttagunta and Toth's article as much as it was intended to question the value of insulin lispro itself. The article was neither systematic in its data gathering nor scientific in its appraisal of the evidence. Systematic review and critical appraisal methodology is now widely known and vigorously endorsed by medical educators and health care administrators. Publishing methodologically weak systematic review articles such as this one significantly undermines these collective efforts. Although selective reporting of data and references to trials that are not well designed can often appear to be "evidence based," closer examination often reveals that the emperor has no clothes.

James McCormack, PharmD
Ken Bassett, MD, PhD
Therapeutics Inititative
University of British Columbia
Vancouver, BC

Competing interests: None declared.

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. McCormack J, Bassett K. The evidence for insulin lispro [letter]. CMAJ 1998;159(11):1353-4.
2. Puttagunta AL, Toth EL. Insulin lispro (Humalog), the first marketed insulin analogue: indications, contraindications and need for further study. CMAJ 1998;158(4):506-11.
3. Anderson JH Jr, Brunelle RL, Keohane P, Koivisto VA, Trautmann ME, Vignati L, et al. Mealtime treatment with insulin analog improves postprandial hyperglycemia and hypoglycemia in patients with non-insulin-dependent diabetes mellitus. Multicenter Insulin Lispro Study Group. Arch Intern Med 1997;157:1249-55.
4. Zinman B, Tildesley H, Chiasson JL, Tsui E, Strack T. Insulin lispro in CSII: results of a double-blind crossover study. Diabetes 1997;46:440-3.
5. Davey P, Grainger D, MacMillan J, Rajan N, Aristides M, Gliksman M. Clinical outcomes with insulin lispro compared with human regular insulin: a meta-analysis. Clin Ther 1997;19:656-74.
6. Brunelle RL, Llewelyn J, Anderson JH Jr, Gale EAM, Koivisto VA. Meta-analysis of the effect of insulin lispro on severe hypoglycemia in patients with type 1 diabetes. Diabetes Care 1998;21:1726-31.
7. Ebeling P, Per-Anders J, Smith U, Lalli C, Bolli GB, Koivisto VA. Strategies toward improved control during insulin lispro therapy in IDDM. Diabetes Care 1997;20:1287-9.
8. Ronnemaa T, Viikari J. Reduction of snacks when switching from conventional soluble to lispro insulin treatment: effects on glycemic control and hypoglycemia. Diabet Med 1998;15:601-7.

Miriam Shuchman [full text / in brief]¹ indicates that a national body should have been available to investigate, evaluate and adjudicate the issues raised by the controversy surrounding Dr. Nancy Olivieri and The Hospital for Sick Children.

The National Council on Ethics in Human Research is a multidisciplinary organization with a mandate to advance the protection and promotion of the well-being of human participants in research and to foster high ethical standards for the conduct of research involving humans. The council has the responsibility to assist research ethics boards in interpreting and implementing the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (1998). The council has established an ongoing mechanism to assess the functions of research ethics boards and has the capability and expertise to address substantive and procedural issues, and it will respond to questions coming from research ethics boards and from organizations such as granting councils or professional organizations.

I am surprised that the ethics director of the Medical Research Council of Canada (MRC), Dr. Francis Rolleston, did not point out that the National Council on Ethics in Human Research is the natural place to refer all issues, contentious or otherwise, that have to do with human participation in research. Canada does have the ability to conduct independent, competent reviews of controversial ethical issues involving human research.

Gordon Crelinsten, MD
Montreal, Que.
gcrelinsten@qc.aibn.com

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Shuchman M. Independent review adds to controversy at Sick Kids. CMAJ 1999;160(3):386-8.

I am quite disturbed by Dr. Francis Rolleston's statement, quoted in Miriam Shuchman's article [full text / in brief],¹ in which he comments on the ethical responsibilities of the MRC: "These [issues] are institutional responsibilities. If you have big brother in Ottawa looking after these things, that's not healthy." I think it is generally accepted that activities under the liberal laissez-faire theory in business must be supervised and to varying degrees regulated by governments and their agencies. I think also that the MRC has unjustifiably abdicated its responsibility to the public to regulate medical research in abandoning individual researchers such as Dr. Olivieri to potentially unscrupulous industrial supporters of their research. The pharmaceutical industry has an obvious vested interest in outcomes to its liking. Somebody, if not the MRC, must set up national rules spelling out appropriate freedoms of enquiry and publication for research supported by industry that are not at the mercy of self-interest or of the "intellectual property" bugbear.

J.V. Frei, MD, PhD
Toronto, Ont.

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Shuchman M. Independent review adds to controversy at Sick Kids. CMAJ 1999;160(3):386-8.

The quotation in Miriam Shuchman's article [full text / in brief]¹ was accurate but incomplete. My view is that conflicts between researchers, institutions and companies should be resolved by the protagonists, not by national organizations such as the MRC or the National Council on Ethics in Human Research (NCEHR). However, as Gordon Crelinson and J.V. Frei point out, the MRC and the NCEHR can, should and do help to set standards. Further, the NCEHR, which was founded and is mainly funded by the MRC, is playing a vital role in supporting research ethics boards and institutions in implementing the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans.

Frei's antipathy to industry disturbs me. Industry is essential to health and health care. Effective collaboration between industry and academe is of great advantage to both and to the health of Canadians. Should we not work together so that, despite different sub-objectives, we achieve our common long-term goals of helping patients while maintaining our principles?

Since the publication of the Tri-Council policy statement I have visited all 16 Canadian medical schools to discuss issues and concerns surrounding processes for research ethics. Interactions with industry were frequently raised. I also established a task force on research ethics boards and clinical trials to address issues that inhibit collaboration between industry and academe.

At a recent workshop entitled "Research Ethics: Maximizing Effectiveness," the almost 100 participants from industry and academe strongly supported the already initiated Working Group on Best Practices in Industry­Academe Interactions. This working group will help to develop principles and approaches with respect to such issues as consent forms, the submission of protocols for ethics review, fees, incentives, compensation, liability and publication. In this way, and through the active, collaborative implementation of the Tri-Council policy statement, the MRC will continue to promote the highest standards of ethics by all involved.

Francis Rolleston, DPhil
Director, Ethics and International Relations
Medical Research Council of Canada

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Shuchman M. Independent review adds to controversy at Sick Kids. CMAJ 1999;160(3):386-8.

The article by Terry Johnson concerning the affiliation of the Canadian Memorial Chiropractic College (CMCC) with York University was particularly biased [full text].¹ The "academic nuptials" of these 2 institutions would not make York University a "laughing-stock within the world's science community," as Michael De Robertis is reported to have said, but rather would enhance York's reputation in the scientific and health fields.

Johnson writes that the article by Balon and colleagues² "marked the first time a leading journal has published a study by chiropractic researchers." In 1985 Canadian Family Physician published an article coauthored by J.D. Cassidy,³ a chiropractor. The British Medical Journal, Spine and other "leading" journals have also published articles by chiropractors. Johnson could have also reviewed the Journal of Manipulative and Physiological Therapeutics, a peer-reviewed, not "unrecognized," journal where scientists, physicians and chiropractors have written quality articles through the years.

De Robertis is reported as having said that chiropractic "metaphysical doctrines" and "unorthodox practices" are not well known. After reviewing the literature and the profession's guidelines⁴ he would probably not call chiropractic a "metaphysical doctrine" (a list of references is available from cagkiro@infonet.ca). The effectiveness of spinal manipulation therapy, the main treatment procedure of chiropractic, is well documented. It is not "unorthodox."

With the synergy of improved collaboration and communication among all professional health care providers and researchers and the scientific community, we can produce studies that will help us better understand human physiology and health to the benefit of all. Change often provokes anxiety and fear, but without change nothing can progress.

Claude A. Gauthier, DC
Aylmer, Que.

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Johnson T. Angry scientists fight university's attempt to affiliate with chiropractic college. CMAJ 1999;160(1):99-100.
2. Balon J, Aker PD, Crowther ER, Danielson C, Cox PG, O'Shaughnessey D, et al. A comparison of active and simulated chiropractic manipulation as adjunctive treatment of childhood asthma. N Engl J Med 1998;339(15):1013-20.
3. Kirkaldy-Willis WH, Cassidy JD. Spinal manipulation in the treatment of low back pain. Can Fam Physician 1985;31:535-40.
4. Henderson D, et al. Clinical guidelines for chiropractic practice in Canada. J Can Chiropractic Assoc 1994;38(1 Suppl).

Both the CMA and the Pharmaceutical Manufacturers Association of Canada (PMAC) have guidelines regarding travel and accommodation arrangements for physicians attending industry-sponsored continuing medical education (CME) events. The CMA prohibits industry payment for travel or lodging costs for physicians or their spouses.¹ PMAC also prohibits payment for spouses but allows it for physicians provided that the CME event meets all 6 criteria as laid out in the latest revision of its Code of Marketing Practices.²

The CMA relies on moral suasion to ensure compliance with its policy; the PMAC tries to enforce its code by investigating complaints and fining companies found guilty of violations. However, it appears that both physicians and drug companies sometimes break the guidelines of their respective organizations. Here are 2 recent examples.

A 1996 internal memo from Bayer Inc. contained the following remark: "For instance, if PMAC were to find out that we sponsor doctors to ... meetings by paying for their flights and or hotel accommodations, we could be in a lot of trouble." In February 1998 Boehringer Ingelheim sent an invitation to psychiatrists to attend the launch of a new antidepressant, bupropion. This "consultative forum" was held at the Grandview Resort in Muskoka, Ont. The invitation contained an offer to pay for all travel and accommodation expenses for attendees and their spouses. At a maximum, Boehringer's meeting fulfilled 3 of the 6 provisions of the PMAC code allowing payment for physicians. The code also states that social functions must not take precedence over the educational component. However, the forum ran from a Friday evening to about Sunday noon, with the educational component taking place on Saturday from 8 am to 12:45 pm. The rest of the time was set aside for a welcome reception, recreation, dinner and brunch.

The CMA publishes its guidelines but does little else to promote them. There is good evidence that merely distributing guidelines is not enough to ensure their adoption.³ A complaints system, such as the one that PMAC uses, means that unreported violations avoid detection. Herxheimer and Collier⁴ speculated that a considerable number of violations of the British industry's code escaped detection because few health professionals bothered to complain.

Both the CMA and the PMAC will have to do better at ensuring compliance with their codes or lose the trust of the public and much of the profession.

Joel Lexchin, MD
Toronto, Ont.

Competing interests: None declared.

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Canadian Medical Association. Physicians and the pharmaceutical industry (update 1994) [policy summary]. CMAJ 1994;150:256A-C.
2. Pharmaceutical Manufacturers Association of Canada. Code of marketing practices. Ottawa: The Association; 1997.
3. Raisch DW. A model of methods of influencing prescribing: part I. A review of prescribing models, persuasion theories, and administrative and educational methods. DCIP Ann Pharmacother 1990;24:417-21.
4. Herxheimer A, Collier J. Promotion by the British pharmaceutical industry, 1983-8: a critical analysis of self regulation. BMJ 1990;300:307-11.

In his article on the sorry state of medical staffing in the Canadian armed forces, Patrick Sullivan reports that training a new medical doctor for the military costs $300 000 [full text].¹ Why not use the pool of forcibly retired military doctors to provide medical services on a 6- to 12-month basis? This group of physicians has more mobility than younger members of the profession, concerned as they are with raising a family and building up a practice. Discrimination on the grounds of age is unlawful in Canada, except in the military. This leads to such inanities as having to pay civilian doctors with no military experience top dollars for services that former regular and reserve medical officers are much better qualified to provide but are prevented from doing so by the obsolete compulsory retirement age regulation.

Col. Emile Berger, MD, CD (retd.)
President, Defence Medical Association of Canada
Montreal Chapter

[Contents]

Send a letter to the editor Envoyez une lettre à la rédaction

1. Sullivan P. Military set to offer large signing bonuses, higher pay in face of unprecedented MD staffing crisis. CMAJ 1999;160(6):889-91.

A recent article stated incorrectly that York University in Toronto offers a program in physiotherapy.¹ We apologize for this error.

The online version is correct.

[Contents]

1. Johnson T. Angry scientists fight university's attempt to affiliate with chiropractic college. CMAJ 1999;160(1):99-100.