March 06 jmcp
Editorial Subjects—In This Issue and in Previous Issues
affirms our intuition that (a) patients can discern among specific
meals or rosiglitazone (Avandia) 4 mg twice daily.17 The mean
sensory attributes of intranasal steroid products, (b) some
A1c values at baseline were 9.5% in the INH group versus 9.4%
attributes have higher economic value in their avoidance for the rosiglitazone group. After 12 weeks of the therapy, the(e.g., no aftertaste versus strong aftertaste), and (c) the absolute
INH group (71 of 76 randomized patients completed the trial)
WTP amount is higher for some attributes (i.e., a lot of throat
achieved a mean A1c of 7.2% versus 8.0% for the 63 of 69
rundown and nose runout) for persons with higher household
patients who completed the rosiglitazone treatment arm, yielding
income. It may also be reasonable to extrapolate these findings
64% greater reduction in A1c (-2.3 absolute points vs. -1.4
to the conclusion that physicians and patients should be absolute points) for INH compared with rosiglitazone, with andiscussing product attributes, particularly for patients who will
adjusted treatment group difference of -0.89, 95% confidence
be using intranasal steroids on a regular basis due to persistent
interval [CI], -1.23 to -0.55.18 At 12 weeks of treatment, 82.7%
of the INH group achieved A1c of 8% or less versus 58.2% for
The implications of the research on WTP for sensory attributes
rosiglitazone, adjusted odds ratio (OR), 7.14; 95% CI, 2.48-
beyond these conclusions are less clear. The graphs (Figure 1)
<0.001. By the American Diabetes Association
in the article by Mahadevia et al. are remarkably similar in slope
standard of <7.0% A1c, the proportion achieving the primary
and clustering except for aftertaste (Figure 1C). Second, at $50
end point was more than twice the proportion for rosiglitazone,
per prescription, the sensory attributes have little differentiation.
44.0% for INH vs. 17.9% for rosiglitazone, (adjusted OR, 4.43;
(Readers are encouraged to examine the Choice Set in Table 3
95% CI, 1.94-10.12). According to the guidelines of the
of the article to make their own inferences.) Based upon the
American Association of Clinical Endocrinologists (A1c ≤6.5%),
actual costs of the 7 therapeutic alternatives in 2005 (Table 1
the success rate was even higher for INH versus rosiglitazone,
page 168), there is only an absolute difference of $5 per month
28.0% vs. 7.5% (OR, 5.34; 95% CI, 1.83-15.57).
between budesonide, mometasone, triamcinolone, and fluticasone.
Pharmacy and therapeutics (P&T) committees have an inter-
Members of pharmacy and therapeutics (P&T) committees
esting question to debate in 2006: the relative value and need
are challenged in decision making by the absence of information
for the thiazolidinediones (TZDs) rosiglitazone (Avandia and
about sensory attributes in the labeling for intranasal steroids.
Avandamet) and pioglitazone (Actos) now that clinicians and
But, Kaliner found, in a telephone survey of 503 patients with
patients have access to INH, which has been found to be supe-
seasonal and perennial allergic rhinitis, that 86% of the patients
rior to rosiglitazone in the intermediate outcome of A1c. TZDs
reported that they had not complained to their physician about
have a larger hole to crawl out of when relative safety is
a sensory attribute of their nasal steroid treatment.13 Since the
considered by P&T committees and clinicians. More than
work by Kaliner was performed 6 years ago, this research, if
2 years ago, at the end of 2003, the American Heart Association
conducted today, may find that patients are more outspoken.
and the American Diabetes Association released a joint consensus
On the other hand, perhaps sensory attributes of intranasal
statement warning that patients with moderate-to-severe congestive
steroids are important to only a small proportion of the popula-
heart failure (CHF) should not be prescribed either rosiglitazone
tion that suffers from allergic rhinitis. One wonders how impor-
or pioglitazone.19 Analysis of administrative claims data reported
tant the sensory attributes of intranasal steroids should be in
in November 2003 by Delea found a 55% higher incidence of
clinical and P&T committee decision making. The one apparent
heart failure for 5,441 TZD patients compared with 28,103
opportunity for quality improvement is in increasing, to more
control patients with diabetes (hazard ratio = 1.7, P
than 7%, the proportion of physicians that base their choice of
adjusted incidence of heart failure at 40 months was 8.2% for
a nasal steroid on patient preference14 for sensory attributes.
TZD patients and 5.3% for control subjects (absolute difference2.9%, relative difference 55%).20 The TZDs have a tendency to
■■ What Is the Future of Thiazolidinediones (TZDs)
cause fluid retention and edema, particularly of the feet, a hall-
After Market Introduction of Inhaled Insulin?
mark of CHF. The incidence of fluid retention and edema
The U.S. Food and Drug Administration (FDA) approved
appear to be exacerbated by concomitant use of insulin and
inhaled insulin (INH) under the trade name Exubera on
dose-related for both rosiglitazone21 and pioglitazone.22,23
January 27, 2006, for use in adult patients with type 1 and type
The black cloud over TZDs does not end with fluid retention,
2 diabetes.15 Exubera is an inhalable, powdered form of insulin
edema, and heart failure. The first TZD, troglitzone (Rezulin),
delivered by a device developed by Nektar Therapeutics. Earlier
approved by the FDA for marketing in the Unites States in
in January 2006, Pfizer agreed to acquire the world-wide rights
January 27, 1997,24 was withdrawn in March 2000 after its asso-
to Exubera from sanofi-aventis for $1.3 billion.16
ciation with hepatotoxicity was no longer in doubt. By year-end
One of the clinical trials that favored FDA approval of INH
1998, the FDA had record of a total of 32 reported deaths and
was a multicenter clinical trial that randomized 145 persons
4 liver transplants associated with the use of troglitzone.25 Prior
with uncontrolled type 2 diabetes (with glycosylated hemo-
to the market withdrawal of troglitazone, the manufacturer had
globin [A1c] in the range of 8%-11%) to either INH before
distributed 2 warning letters to physicians, including a notice to
www.amcp.org Vol. 12, No. 2 March 2006 JMCP
Journal of Managed Care Pharmacy 169
Editorial Subjects—In This Issue and in Previous Issues
monitor liver enzymes monthly, and the United Kingdom had
the absence of a pharmacy claim for TZD or insulin plus another
withdrawn troglitazone from the market more than 2 years 12% of patients who did not have continuous eligibility duringearlier, in November 1999.26 For pioglitazone, introduced to
the study period. LaFleur, in a previous issue of JMCP,
the U.S. market in July 1999,27 the first death from liver failure
a comprehensive and useful review of the limitations of admin-
associated with its use was reported in 2004.28 A “Dear
istrative claims data specific to the measurement of costs of care
Prescriber” letter dated April 26, 2002, highlighted changes in
the precautions section of labeling of rosiglitazone regarding the
Third, the time period for the administrative claims used in
potential for hepatic effects, even a rare case of hepatic failure,
the analyses by Kalsekar et al. is confounded by the TZD market
and, in other parts of product labeling, a possibility of weight
disruption caused by the withdrawal of troglitazone in March
2000, in the middle of their 1999-2001 study period. While
Add to peripheral edema, weight gain, and possible hepato-
troglitazone patients were excluded from the 12-month follow-
toxicity an apparent increased risk of macular edema. In
up period, 12.8% of the TZD patients and 13.0% of the insulin
January 2006, the FDA notified health care professionals of
patients in the final analysis received troglitazone in the preperiod
reports of both new-onset and worsening cases of diabetic (Table 3 in Kalsekar et al.). One wonders why the outpatientmacular edema associated with the use of rosiglitazone, albeit
costs, which included office visits and laboratory tests for liver
reversible and apparently dose related.29
enzymes, were not higher in the preperiod for both groups and
With an apparent black eye for safety and inferior to INH in
why the outpatient costs were similar ($177 mean difference,
efficacy for lowering A1c, the cost outcomes for the TZDs may
$1,070 for the insulin group compared with $893 for the TZD
offer a refuge in the storm. Kalsekar et al., in this issue of JMCP
= 0.458 for the comparison) in the postperiod, partic-
found, from administrative claims data in a Medicaid fee-for-
ularly since there was considerable media attention to the market
service population, that patients who started TZD therapy
withdrawal of troglitazone at the time. Consumer groups,
incurred 35% lower costs in a 12-month follow-up period in
represented by Public Citizen, petitioned the FDA in early 2000
2000-2001 for emergency room (ER) visits and hospitalization
to revise the product labeling for all 3 TZDs to describe the
compared with patients who initiated therapy with insulin, an
adverse effects that include liver toxicity and heart failure.33
average of $3,727 versus $5,793, respectively (P
Fourth, a threat to validity seems inherent in the premise
diabetes-related costs only, the 53% higher pharmacy costs in
that higher drug costs for TZDs are offset by lower costs for ER
the TZD patients ($1,678) versus insulin patients ($1,096, visits and hospitalizations, in only 12 months of administrativeP
<0.01) was offset by lower costs for emergency room (ER) claims data. For diabetes-related costs, the insulin group had avisits and hospitalization ($2,855 vs. $5,090, P
higher average number of physician office visits, but the average
in 25% lower total diabetes-related costs for the TZD group
outpatient costs (including physician office visits) were not
compared with the insulin group ($5,425 vs. $7,255, P
different between the 2 groups over 12 months of follow-up.
Limitations in the study by Kalsekar et al. are significant.
However, the contribution of the $177 mean difference ($1,070
Important among these limitations was the inability to measure
for the insulin group vs. $893 for the TZD group, P
= 0.458 for
disease severity. Lacking clinical values for A1c and blood the comparison) may have contributed to the $1,830 differenceglucose, the researchers attempted to assure the comparability
($153 per month) in total diabetes-related costs ($7,255 for the
of the 2 groups through the use of propensity matching.
insulin group vs. $5,425, P
= 0.022 for the comparison).
However, the propensity matching technique would tend to
Stephens et al. provide a useful context for interpreting these
exclude patients with more-severe diabetes who were, therefore,
absolute and relative values in their summary of the literature
using insulin and could not be matched to TZD patients. The
on economic studies in this issue of JMCP
, including their Table 1,
magnitude of this limitation can be gleaned from the authors’
showing the breakdown of total direct costs for managed care
Table 1, which shows that 65% (n = 1,271) of otherwise study-
organization (MCO) enrollees with diabetes.34
eligible patients were excluded in the process of propensity
The place in therapy for TZDs in type 2 diabetes is summa-
matching, leaving only 690 patients in the final analyses: 345
rized clearly and succinctly in the NICE (National Institute for
patients with type 2 diabetes in the TZD group and 345 patients
Clinical Excellence) appraisal released in November 2002.
TZDs are third-line therapy in type 2 diabetes after sulfonyl-
Second, Medicaid patients, particularly those not enrolled in
ureas and metformin have proved unsuccessful when used as
managed Medicaid, are most likely not representative of the
monotherapy and in combination to achieve glycemic goals.35
general population of patients with type 2 diabetes, and discon-
NICE concluded that “the use of glitazone combination therapy
tinuous eligibility contributes to discontinuous care in this with either metformin or sulphonylurea is not likely to be costpopulation.31 A hint of the problem of discontinuous care might
effective when compared with the combination of metformin
be found in the 75% of patients identified with a diagnosis and sulphonylurea.”of type 2 diabetes who were excluded from the study because of
So, much of the market future of TZDs may have to do with
170 Journal of Managed Care Pharmacy JMCP
Editorial Subjects—In This Issue and in Previous Issues
relative cost. In previous research reported in JMCP,
Shetty et al.
12. Drug Facts and Comparisons (Clinisphere version, ISBN 1-57439-036-8).
found that patients with type 2 diabetes maintained at target
St. Louis, MO: Wolters, Kluwer Health, Inc.; December 2005. Respiratoryinhalant products, intranasal steroids. Accessed January 31, 2006.
A1c (≤7%) had 24% lower total diabetes-related medical costs
13. Kaliner MA. Patient preferences and satisfaction with prescribed nasal
compared with patients who were above target A1c (>7%).36
steroids for allergic rhinitis. Allergy Asthma Proc.
2001;22(6 suppl 1):S11-S15.
The actual daily direct drug costs for INH are not yet known.
14. Kaliner MA. Physician prescribing practices: the role of patient preference
What is known is that the TZDs are not inexpensive. Based
in the selection of nasal steroids. Allergy Asthma Proc.
upon MCO pharmacy claims for the fourth quarter of 2005, the
average direct drug cost, including pharmacy dispensing fees,
15. U.S. Food and Drug Administration. FDA approves first-ever inhaled
was $4.34 per day ($130 per month) for Avandia, $4.55 insulin combination product for treatment of diabetes. January 27, 2006.
Available at: http://www.fda.gov/bbs/ topics/news/2006/NEW01304.html.
per day ($136 per month) for Avandamet, and $5.13 ($154 per
month) for Actos. The evidence presents a classic conundrum
16. Loftus P. Pfizer receives FDA approval for inhaled-insulin treatment.
for P&T members: a product—INH—with superior efficacy
Wall Street Journal.
January 28-29, 2006:A6.
and safety but (probable) higher direct drug cost compared
17. U.S. Food and Drug Administration. Statistical review and evaluation.
with TZDs. Still, the market introduction of INH in February
Clinical study no. 110 in NDA/Serial No. 21-868. Available at:
2006 seems likely to knock pioglitazone and rosiglitazone from
http://www.fda.gov/ohrms/dockets/ac/05/briefing/2005-4169B1_02_05-FDA-Clin-Stats-Efficacy.pdf. Accessed February 15, 2006.
their status in the top 12 drugs by expenditure at year-end 2005.
18. DeFronzo RA, Bergenstal, RM, Cefalu WT, et al. Efficacy of inhaledinsulin in patients with type 2 diabetes not controlled with diet and exercise.
Frederic R. Curtiss, PhD, RPh, CEBS
19. Nesto RW, Bell D, Bonow RO, et al. Thiazolidinedione use: fluid reten-
tion, and congestive heart failure. A consensus statement from the AmericanHeart Association and American Diabetes Association. Circulation.
2003;108:2941-48; Diabetes Care.
20. Delea TE. Use of thiazolidinediones and risk of heart failure in people with
1. Buntin MB, Damberg C, Haviland A, et al. Consumer-driven health plans:
type 2 diabetes: a retrospective cohort study. Diabetes Care.
implications for health care quality and cost. June 2005. RAND Corp. for the
21. Raskin P, Rendell M, Riddle MC, et al. A randomized trial of rosigltazone
California HealthCare Foundation. Available at: http://www.chcf.org/docu-
therapy in patients with inadequately controlled insulin-treated type 2 dia-
ments/insurance/ ConsumerDirHealthPlansQualityCost.pdf. Accessed January
betes. Diabetes Care.
22. Rubin C, Egan J, Schneider Ra, et al. Combination therapy with pioglita-
2. Wasserfallen JB, Currat-Zweifel C, Cheseaux JJ, Hofer M, Fanconi S. Parents’
zone and insulin in patients with type 2 diabetes. Diabetes [abstract].
willingness to pay for diminishing children’s pain during blood sampling.
23. Rosenstock J, Einhorn D, Hershon K, et al. Efficacy and safety of pioglita-
3. Whynes DK, Frew EJ, Wolstenholme JL. Willingness-to-pay and demand
zone in type 2 diabetes: a randomized, placebo-controlled study in patients
curves: a comparison of results obtained using different elicitation formats.
receiving stable insulin therapy. Intern J Clin Pract.
Int J Health Care Finance Econ.
24. U.S. Food and Drug Administration. Approvals for January 1997. Available
4. Covey J, Smith RD. How common is the ‘prominence effect’? Additional
at: http://www.fda.gov/cder/da/da0197.htm. Accessed February 5, 2006.
evidence to Whynes et al. Health Econ.
25. Langreth R, Sharpe R. Warner-Lambert diabetes drug set to be reviewed
5. King JT, Tsevat J, Lave JR, Roberts MS. Willingness to pay for a quality-
by FDA panel. Wall Street Journal.
January 18, 1999:B7.
adjusted life year: implications for societal health care resource allocation.Med Decis Making.
26. Winslow R. Warner Lambert bid to prevent diabetes is set back as NIHhalts Rezulin trial. Wall Street Journal.
June 8, 1998:B6.
6. Byrne MM, O’Malley K, Suarez-Almazor ME. Willingness to pay per quality-adjusted life year in a study of knee osteoarthritis. Med Decis Making.
27. Drug Facts and Comparisons
(Clinisphere version, ISBN 1-57439-036-8).
St. Louis, MO: Wolters, Kluwer Health, Inc.; January 2006. Endocrine andmetabolic agents, thiazolidinediones, pioglitazone. Accessed February 5,
7. Gueylard Chenevier D, Lelorier J. A willingness-to-pay assessment of
parents’ preferences for short duration treatment of acute otitis media in children. Pharmacoeconomics.
28. Farley-Hills E, Sivasanka R, Martin M. Fatal liver failure associated withpioglitazone. BMJ.
8. Sadri H, Mackeigan LD, Leiter LA, Einarson TR. Willingness to pay for inhaled insulin: a contingent valuation approach. Pharmacoeconomics.
29. Cobitz AR. Dear Health Care Provider letter. January 5, 2006. Available at:
http://www.fda.gov/medwatch/safety/2006/Avandia_DHCPletter.pdf. AccessedFebruary 5, 2006.
9. Parthiv Mahadevia P, Shah S, Mannix S, et al. Allergic rhinitis patients’ willingness to pay for sensory attributes of intranasal corticosteroids. J Manag
30. Kalsekar I, Iyer S, Mody R, Rajagopalan R, Kavookjian J. Utilization and
costs for compliant patients initiating therapy with pioglitazone or rosiglitazoneversus insulin in a Medicaid fee-for-service population. J Manag Care Pharm.
10. Mahadevia PJ, Shah S, Leibman C, et al. Patient preferences for sensory
attributes of intranasal corticosteroids and willingness to adhere to prescribedtherapy for allergic rhinitis: a conjoint analysis. Ann Allergy Asthma Immunol.
31. Gold M, Felt S. Reconciling practice and theory: challenges in monitoring
Medicaid managed-care quality. Health Care Finance Rev.
11. Keith PK, Haddon J, Birch S. A cost-benefit analysis using a willingness-
32. LaFleur J. Evaluating the relationship between diabetes treatment and
to-pay questionnaire of intranasal budesonide for seasonal allergic rhinitis.
health care costs—measuring the atom with a yardstick? J Manag Care Pharm.
Rhinocort Study Group. Ann Allergy Asthma Immunol.
www.amcp.org Vol. 12, No. 2 March 2006 JMCP
Journal of Managed Care Pharmacy 171
Editorial Subjects—In This Issue and in Previous Issues
33. Adams C. Three diabetes drugs need revisions in labels, consumer grouptells FDA. Wall Street Journal.
March 8, 2000:B10.
Letters to the Editor
34. Stephens JM, Botteman MF, Hay JW. Economic impact of antidiabetic
welcomes letters that serve to clarify subjects published in
medications and glycemic control on managed care organizations: a review
previous issues of the Journal
or regarding subject matter of interest to
of the literature. J Manag Care Pharm.
managed care pharmacists. Letters in JMCP
are not peer reviewed but are
35. National Institute for Clinical Excellence. Appraisal consultation docu-
subjected to editorial review. Letters should be prepared in a word
ment: review: the clinical effectiveness and cost effectiveness of glitazones for the treatment of type 2 diabetes. November 25, 2002. Available at:
processing program, preferably Microsoft Word, and submitted electron-
http://www.nice.org.uk/ page.aspx?o=39365. Accessed February 6, 2006.
ically at jmcp.msubmit.net. See www.amcp.org for details.
36. Shetty S, Secnik K, Oglesby AK. Relationship of glycemic control to total diabetes-related costs for managed care health plan members with type 2 diabetes. J Manag Care Pharm.
172 Journal of Managed Care Pharmacy JMCP
MATERIALS PROVIDED 2. Tramadol test device. The amount of each coated antigen and/or antibody on the strip is less than 1.0 mg for antigen conjugate and is less than 1.0 mg for goat anti-mouse IgG Test zone: contains tramadol bovine protein antigen Control zone: contains Goat anti-mouse IgG antibody. 3. Conjugate pad: contains mice monoclonal anti-drug MATERIALS REQUIRED BUT NOT
Research Papers 1) NiimiC, Goto H, Ohmiya N, Niwa Y, Hayakawa T, Nagasaka T, Nakashima N Usefulness of p53 and Ki67 immunohistochemical analysis for preoperative diagnosis of extremely well-differentiated gastric adenocarcinoma. Am J Clin Pathol,2002 Nov;118(5)683-92 PMID:12428787[PubMed-indexed for MEDLINE] 2) Furata S, Goto H, Niwa Y , Ohmiya N, Kamiya K, Oguri A, Hayakawa T, Mori N