Digest.net.cn

Annals of Internal Medicine
COMMENTS AND RESPONSES
Potential Financial Conflicts of Interest: None disclosed.
References
Is Patient Cost-Sharing the Best Way to Protect the Medical
1. Wharam JF, Galbraith AA, Kleinman KP, Soumerai SB, Ross-Degnan D, Landon Commons?
BE. Cancer screening before and after switching to a high-deductible health plan. AnnIntern Med. 2008;148:647-55. [PMID: 18458277]2. Mahajan AP, Brook RH. High-deductible health plans and better benefit design TO THE EDITOR: Wharam and colleagues (1) reported that a high-
[Editorial]. Ann Intern Med. 2008;148:704-6. [PMID: 18458284] deductible insurance plan providing first-dollar coverage for fecal 3. Sabin JE, Cochran D. Confronting trade-offs in health care: Harvard Pilgrim Health occult blood tests (FOBTs) but not colonoscopies did not change Care’s organizational ethics program. Health Aff (Millwood). 2007;26:1129-34.
colorectal cancer screening rates but did result in somewhat higher use of FOBTs and lower use of colonoscopies. The article and ac- 4. Emanuel EJ. The cost-coverage trade-off: “it’s health care costs, stupid”. JAMA.
companying editorial (2) imply that it might be unsound to cover FOBTs but not colonoscopies. However, clinical practice guidelines 5. Hiatt HH. Protecting the medical commons: who is responsible? N Engl J Med.
at the time of the study recommended both tests. These guidelines did not consider either method to be unequivocally superior, butpointed out that they had different characteristics that might matter IN RESPONSE: We thank Dr. Fletcher and colleagues for their thought-
to individual patients. Very recent guidelines judge colonoscopy and ful letter describing HPHC’s decision-making process and rationale for other structural examinations of the colon to be preferable to stool providing first-dollar coverage of FOBT, but not colonoscopy, in its tests because they can better detect adenomas and so prevent cancer.
high-deductible health plan. The HPHC’s ethics advisory group, which In their editorial, Mahajan and Brook (2) wonder whether Har- engages a range of stakeholders, is laudable.
vard Pilgrim Health Care (HPHC) gave careful consideration to the Dr. Fletcher and colleagues defend HPHC’s decision to provide consequences of providing first-dollar coverage for FOBT, the less first-dollar coverage for FOBT but not colonoscopy by citing evi- expensive test, but not colonoscopy. Since 1996, an ethics advisory dence-based guidelines, from the time when the coverage decision group has advised HPHC on ethical issues, including tough alloca- was made, that regarded the 2 screening interventions as equally tion decisions and confronting contemporary health plans (3). Two effective. In light of new consensus guidelines recommending dozen participants represent a range of stakeholders, such as HPHC colonoscopy over FOBT because of the added benefit of early detec- staff, community physicians, consumers, purchasers, and ethicists. At tion and removal of polyps (1), we reviewed the high-deductible a time of national backlash against insurer- and physician-led man- health plan information available to potential enrollees on HPHC’s aged care, the ethics advisory group was asked to explore the ethical Web site and found that colonoscopy is still subject to the deductible dimensions of insurance that was more affordable and included de- (2). Given the plan’s goal to design benefits in a way that encourages ductibles that encouraged greater consumer participation in their the use of high-value preventive services, we hope that the HPHC is own health care decisions. The group devoted 6 meetings to these in the process of changing its high-deductible health plan policy to issues and did take into account research evidence of test effective- provide first-dollar coverage for colonoscopy.
ness as well as evidence that, in some settings, deductibles reduce care Citing the need to control rising health care costs, Dr. Fletcher without regard to its effectiveness. Harvard Pilgrim Health Care also and colleagues justify the use of cost-sharing to promote use of less- encouraged studies of the effects of its policies, and the study by expensive care that produces outcomes equivalent to those of more- expensive care. Although “protecting the medical commons” is in- We agree that it would be wonderful if “patient cost-sharing deed an important obligation that we share, it is not clear from were not needed to control costs and patients and physicians instead existing evidence that an isolated focus on promoting preventive worked together to eliminate waste and equivocal . . . or ineffective screening will achieve significant cost reductions in the long run. The services” (2). But the cost of heath care is an urgent, practical prob- HPHC high-deductible health plan does not provide first-dollar cov- lem with widespread consequences. For example, Emmanuel (4) has erage for highly effective care, such as prescription medications for argued that cost is the underlying cause of so many uninsured Amer- chronic disease (2); increased cost-sharing for medications reduces icans. So pipe dreams are not enough. Harvard Pilgrim Health Care their use for such conditions as hypertension, diabetes, asthma, and believes that one of its most important obligations is “protecting the depression (3). In addition, cost-sharing reduces the use of clinically medical commons” (5). Should we not applaud, rather than criticize, effective services and less-effective or ineffective services in roughly efforts to live with less-expensive care at a time when there was not equal proportions (4). If HPHC is interested in protecting the med- evidence-based consensus that more-expensive care produced better ical commons, why resort to a cost-sharing policy that will adversely affect the well-being of some patients? Wouldn’t it be better to firstattempt to eliminate wasteful and inappropriate care (in which risks to the patient exceed the potential benefit) (5) by implementing appropriateness criteria and methods (4, 5) and systematically edu- cating its providers and enrollees in how to use the criteria? Disclaimer: The authors are affiliated with Harvard Pilgrim Health
David Geffen School of Medicine at University of California, Care, whose policies were studied in the article (1) and commented on in 2008 American College of Physicians 435
Veterans Affairs Ann Arbor Healthcare System Disclaimer: Dr. Brook’s wife, Dr. Jacqueline Kosecoff, is CEO of Pre-
Potential Financial Conflicts of Interest: None disclosed.
Reference
Potential Financial Conflicts of Interest: None disclosed.
1. Glasziou PP, Irwig L, Heritier S, Simes RJ, Tonkin A. LIPID Study Investigators.
Monitoring cholesterol levels: measurement error or true change? Ann Intern Med.
2008;148:656-61. [PMID: 18458278] References
1. Levin B, Lieberman DA, McFarland B, Smith RA, Brooks D, Andrews KS, et al.
American Cancer Society Colorectal Cancer Advisory Group. Screening and surveil-
TO THE EDITOR: The study by Glasziou and colleagues (1) finds
lance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint that “noise” random variations during serial cholesterol level moni- guideline from the American Cancer Society, the US Multi-Society Task Force on toring may be greater than the change in cholesterol levels due to a Colorectal Cancer, and the American College of Radiology. CA Cancer J Clin. 2008; therapeutic effect. This points to the importance of what exactly has been tested in cholesterol-lowering studies. Although guidelines and 2. Harvard Pilgrim Health Care. Six Facts About the HPHC Insurance Company Best clinical practice often focus on titrating lipid treatment to obtain Buy HSA PPO—Massachusetts. Accessed at https://www.harvardpilgrim.org/pls/portal/ specified goal levels (2), the intervention that has actually been tested docs /PAGE/MEMBERS/COVERAGE/BESTBUYHSAPPO/CC2389_MA_PRE in many of the important statin trials is the administration of a fixed medium or high dose of a cholesterol-lowering medication. For ex- 3. Goldman DP, Joyce GF, Escarce JJ, Pace JE, Solomon MD, Laouri M, et al.
ample, the WOSCOPS (West of Scotland Coronary Prevention Pharmacy benefits and the use of drugs by the chronically ill. JAMA. 2004;291:2344- Study) and LIPID (Long-Term Intervention with Pravastatin in Ischaemic Disease) studies compared pravastatin, 40 mg/d, with pla- 4. Newhouse JP. The Insurance Experiment Group. Free for All? Lessons from the cebo (3, 4), and the PROVE-IT (Pravastatin or Atorvastatin Evalu- RAND Health Insurance Experiment. Cambridge, MA: Harvard Univ Pr; 1993.
ation and Infection Therapy) and REVERSAL (Reversal of Athero- 5. Schuster MA, McGlynn EA, Brook RH. How good is the quality of health care in sclerosis with Aggressive Lipid Lowering) studies compared fixed the United States? 1998. Milbank Q. 2005;83:843-95. [PMID: 16279970] high-dose atorvastatin with fixed medium-dose pravastatin (5, 6).
Thus, the treatment that has been evaluated in each case is a fixeddose of a cholesterol-lowering medication, not titration to a specified Monitoring Cholesterol Levels: Understanding Variance and
goal. Evidence-based cholesterol treatment should then focus on pro- Finding the Most Useful Data
viding patients with an appropriate statin dose based on trial datarather than on a less-studied dose-titration strategy.
TO THE EDITOR: I read with great interest the article by Glasziou
and colleagues (1), which provided useful variance parameters for a University of Wisconsin School of Medicine and Public Health simulation study I am currently conducting. I believe the authors made 2 errors, however, and although neither error alters the con-clusions of their study, researchers using the reported estimates forother work might come to erroneous conclusions depending on Potential Financial Conflicts of Interest: None disclosed.
First, the variance estimates reported on page 659 (column 2, References
paragraphs 3 and 4) are in the wrong units. The units should be 1. Glasziou PP, Irwig L, Heritier S, Simes RJ, Tonkin A. LIPID Study Investigators.
mmol2/L2 (mg2/dL2) and not mmol/L (mg/dL).
Monitoring cholesterol levels: measurement error or true change? Ann Intern Med.
Second, the variance estimates given on page 658 (column 2, paragraph 4) are incorrect. To convert a variance estimate from 2. Grundy SM, Cleeman JI, Merz CN, Brewer HB Jr, Clark LT, Hunninghake DB, mmol2/L2 to mg2/dL2, one must take the square of the conversion et al. National Heart, Lung, and Blood Institute. Implications of recent clinical trialsfor the National Cholesterol Education Program Adult Treatment Panel III guidelines.
factor because variance is not a linear operator: Var(cX) ϭ Circulation. 2004;110:227-39. [PMID: 15249516] c2 ϫ Var(X). Thus, the variation in initial response to treatment is 3. Influence of pravastatin and plasma lipids on clinical events in the West of Scotland not 21.8 mg2/dL2, but: (1 mg/dL Ϭ 0.02586 mmol/L)2 ϫ 0.56 Coronary Prevention Study (WOSCOPS). Circulation. 1998;97:1440-5. [PMID: The same error applies to the variance estimates reported on 4. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study page 659. The authors report the correct estimates for the SD of Group. Prevention of cardiovascular events and death with pravastatin in patients with changes in low-density lipoprotein cholesterol throughout the paper, coronary heart disease and a broad range of initial cholesterol levels. N Engl J Med.
probably because the authors made the conversion from mmol/L to mg/dL on the scale of SDs (and not variance), in which squaring the 5. Ridker PM, Morrow DA, Rose LM, Rifai N, Cannon CP, Braunwald E. Relative conversion factor is not needed. For this reason, any reader interpret- efficacy of atorvastatin 80 mg and pravastatin 40 mg in achieving the dual goals of ing the study’s conclusions by using the SD estimates will be correct low-density lipoprotein cholesterol Ͻ70 mg/dl and C-reactive protein Ͻ2 mg/l: an regardless of the units, but when interpreting the variance estimates, analysis of the PROVE-IT TIMI-22 trial. J Am Coll Cardiol. 2005;45:1644-8.
will be correct only for the variance estimates reported in mmol2/L2.
436 16 September 2008 Annals of Internal Medicine Volume 149 • Number 6
www.annals.org
6. Nissen SE, Tuzcu EM, Schoenhagen P, Crowe T, Sasiela WJ, Tsai J, et al. Reversal tion or adjustment, but we do not believe practice must exactly echo of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) Investigators. Statin the trials, which are designed for maximum power rather than opti- therapy, LDL cholesterol, C-reactive protein, and coronary artery disease. N Engl J Like Dr. Hong and colleagues, we would like to understand the factors that could explain and reduce the variability. Some of the TO THE EDITOR: As Glasziou and colleagues (1) report, after the
variation is irreducible, such as the analytic variation from the labo- initial decrease in cholesterol level in response to treatment, subse- ratory, which has a coefficient of variation of 2.7% compared with quent cholesterol level monitoring may be much less frequent than is the within-person coefficient of variation of 7.8% (1). Only some of currently recommended. They show that much of current testing the short-term biological variation is explained by the other factors will detect only false-positive results, which are related to either short-term biological variation or analytic error (1). However, they Unfortunately, we generally do not know what these factors are.
don’t clearly show the factors inducing false-positive results or a way In the LIPID trial analysis, this variability was minimized by both the design (run-in periods and fixed-dose therapy) and analysis meth- We would like to ask some questions. First, what are the statis- ods that accounted for patients changing therapy. In clinical practice, tical factors that induce the differences observed? Except for prava- the variability may be greater. Therefore, our results may not hold if statin treatment, the variation in cholesterol levels may result from a patient changes medication or substantially changes diet.
other factors, such as other medications, level of physical activity,eating habits, smoking status, depression, and others (2). For exam- ple, the patients may receive another medication that may affect cholesterol levels and enhance (or weaken) the effect of pravastatin.
Second, did the authors try to avoid controllable biological variation,such as not allowing the patients to drink alcoholic beverages when detecting serial cholesterol concentrations? Third, what methods did the authors use to decrease false-positive results in their study? Fi- nally, this study is based on participants from Australia and NewZealand, so the results may not be applicable to Asians. Could the Potential Financial Conflicts of Interest: None disclosed.
authors comment on the effect of racial diversity on the variation incholesterol levels? Reference
1. Choudhury N, Wall PM, Truswell AS. Effect of time between measurements on
within-subject variability for total cholesterol and high-density lipoprotein cholesterol in women. Clin Chem. 1994;40:710-5. [PMID: 8174241] Daiming Fan, PhDInstitute of Digestive Diseases, Xijing Hospital, Fourth Military CLINICAL OBSERVATION
Potential Financial Conflicts of Interest: None disclosed.
Sildenafil-Induced Thrombocytopenia
Background: Drug-induced thrombocytopenia is a serious side effect that is typically due to platelet destruction caused by drug- References
1. Glasziou PP, Irwig L, Heritier S, Simes RJ, Tonkin A. LIPID Study Investigators.
induced antibodies (1, 2). Sildenafil is an approved treatment for Monitoring cholesterol levels: measurement error or true change? Ann Intern Med.
both erectile dysfunction and pulmonary arterial hypertension and has not been previously associated with thrombocytopenia (3). Re- 2. Bogers RP, Bemelmans WJ, Hoogenveen RT, Boshuizen HC, Woodward M, ports of any adverse reaction to this medication may have important Knekt P, et al. BMI-CHD Collaboration Investigators. Association of overweight with consequences because more than 23 million patients were prescribed increased risk of coronary heart disease partly independent of blood pressure and cho- lesterol levels: a meta-analysis of 21 cohort studies including more than 300 000 per- Objective: To describe a case of sildenafil-induced thrombocy- sons. Arch Intern Med. 2007;167:1720-8. [PMID: 17846390] Case Report: A 53-year-old woman was admitted with dyspnea IN RESPONSE: We thank Dr. Timbie for the 2 corrections. All our
and volume overload. Her medical history included coronary artery calculations and writing were done in mmol/L, but we added the bypass surgery that resulted in recurrent pleural effusions and fibro- U.S. units (mg/dL) in our revision and did this incorrectly for vari- thorax. Other previous conditions included pulmonary hyperten- ances (although SDs and means are correct).
sion, diastolic dysfunction, hypertension, diabetes, chronic kidney Dr. Cayley points out that most of the statin trials have used a disease, and obstructive sleep apnea.
fixed dose (4S [Scandinavian Simvastatin Survival Study] is an ex- On admission, we administered furosemide and metolazone as ception) rather than monitoring-based titration or adjustment. As well as the patient’s home medications of lisinopril, atenolol, aspirin, our work demonstrates, trials with a fixed dose allow the opportunity simvastatin, omeprazole, acetaminophen, and insulin. We adminis- to assess how cholesterol values increase over time. In patients who tered deep venous thrombosis prophylaxis with unfractionated hep- truly have a substantial increase, a change in treatment can be con- arin. Her clinical condition continued to deteriorate, and she devel- sidered. Inferences can then be made about monitoring-based titra- oped respiratory distress. We transferred her to the intensive care www.annals.org
16 September 2008 Annals of Internal Medicine Volume 149 • Number 6 437
Figure. Daily platelet count and the relationship to dosing of medications administered during hospitalization.
Sildenafil,
Sildenafil
Sildenafil, simvastatin,
omeprazole restarted
Sildenafil
Platelets,
Simvastatin and
Sildenafil stopped
omeprazole stopped
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 2324 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Hospital Day
unit. Because of the severity of her condition, we started empirical count subsequently decreased from 174 ϫ 109 cells/L to 74 ϫ 109 treatment with sildenafil, 25 mg 3 times daily, even though the cells/L over 24 hours. We permanently discontinued sildenafil, and patient did not meet established indications for this therapy (3).
her platelet count again normalized. This confirmed the diagnosis of One week into her hospitalization, we found that the patient sildenafil-induced thrombocytopenia, and we eliminated the contin- had acute thrombocytopenia with a decrease in platelet count from ued use of sildenafil in this patient.
158 ϫ 109 cells/L to 61 ϫ 109 cells/L over 24 hours. She had no Discussion: This case of sildenafil-induced thrombocytopenia history of thrombocytopenia, and a peripheral smear revealed no meets standardized criteria established for drug-induced thrombocy- platelet clumping or schistocytes. We discontinued all heparin prod- topenia (2). Sildenafil was the only medication started before the ucts and ordered both heparin-induced thrombocytopenia (HIT) an- development of thrombocytopenia, and withdrawal resulted in sus- tibody and serotonin-release assay testing. Platelet count decreased tained recovery of platelet levels (Figure). Sildenafil-induced throm-
further, and we stopped other medications previously reported to bocytopenia has not been previously described, possibly because sil- cause thrombocytopenia (acetaminophen, simvastatin, furosemide, denafil is infrequently used with sustained daily dosing. The relatively small number of participants in studies of sildenafil for Four days after the initial diagnosis of thrombocytopenia, the pulmonary arterial hypertension makes it possible that a rare side patient’s platelet count continued to decrease to 17 ϫ 109 cells/L.
effect could be missed; we note that our patient did not meet criteria We did not inadvertently give any heparin products to the patient.
for pulmonary arterial hypertension, which would have fully justified We then discontinued sildenafil, a new medication started during the use of the drug. Furthermore, no data are available on platelet counts admission. The next day, her platelet count increased to 72 ϫ 109 in patients using sildenafil regularly for erectile dysfunction. Some of cells/L and her HIT antibody result was positive. We then treated these patients may develop undetected thrombocytopenia after recur- the patient as having an atypical case of HIT and started argatroban Conclusion: Sildenafil therapy was the likely cause of this pa- We restarted sildenafil, simvastatin, furosemide, and omeprazole tient’s thrombocytopenia. Whether this is a class effect of phospho- treatments. Forty-eight hours later, the platelet count decreased to diesterase type 5 inhibitors cannot be determined. However, these 4 ϫ 109 cells/L. We again confirmed that no heparin products were medications should be considered as a possible cause in patients given. At this time, the HIT serotonin-release assay result was nega- being evaluated for new-onset thrombocytopenia.
tive, bringing into question the diagnosis of HIT. Repeated HITantibody and serotonin-release assay tests at this time yielded nega- Her respiratory status transiently improved while taking silde- nafil, but we again discontinued the medication because of her re- current thrombocytopenia. The patient’s platelet counts normalized during the next 4 days, but her respiratory status worsened. Without the confirmed diagnosis of drug-induced thrombocytopenia, we re-challenged the patient with a 25-mg dose of sildenafil. Her platelet Potential Financial Conflicts of Interest: None disclosed.
438 16 September 2008 Annals of Internal Medicine Volume 149 • Number 6
www.annals.org
References
stead of 81) and 16 patients treated with standard triple therapy 1. Aster RH, Bougie DW. Drug-induced immune thrombocytopenia. N Engl J Med.
(instead of 75). Pooling these data with pertinent data from the study of Vaira and colleagues (3) showed eradication rates of 83.9% 2. George JN, Raskob GE, Shah SR, Rizvi MA, Hamilton SA, Osborne S, et al.
and 35.1% with sequential and standard triple therapy, respectively Drug-induced thrombocytopenia: a systematic review of published case reports. Ann (difference between treatments, 48.7% [95% CI, 25.7% to 64.8%]).
Intern Med. 1998;129:886-90. [PMID: 9867731]3. Galie` N, Ghofrani HA, Torbicki A, Barst RJ, Rubin LJ, Badesch D, et al. Sildenafil References
Use in Pulmonary Arterial Hypertension (SUPER) Study Group. Sildenafil citrate 1. Jafri NS, Hornung CA, Howden CW. Meta-analysis: sequential therapy appears therapy for pulmonary arterial hypertension. N Engl J Med. 2005;353:2148-57.
superior to standard therapy for Helicobacter pylori infection in patients naive to treat- ment. Ann Intern Med. 2008;148:923-31. [PMID: 18490667] 4. Jackson G, Gillies H, Osterloh I. Past, present, and future: a 7-year update of Viagra 2. De Francesco V, Margiotta M, Zullo A, Hassan C, Troiani L, Burattini O, et al.
(sildenafil citrate). Int J Clin Pract. 2005;59:680-91. [PMID: 15924597] Clarithromycin-resistant genotypes and eradication of Helicobacter pylori. Ann Intern 5. Arepally GM, Ortel TL. Clinical practice. Heparin-induced thrombocytopenia. N Engl J Med. 2006;355:809-17. [PMID: 16928996] 3. Vaira D, Zullo A, Vakil N, Gatta L, Ricci C, Perna F, et al. Sequential therapyversus standard triple-drug therapy for Helicobacter pylori eradication: a randomizedtrial. Ann Intern Med. 2007;146:556-63. [PMID: 17438314] CORRECTIONS
Correction: Seeding Trials: Just Say “No”
Correction: Sequential Therapy Appears Superior to
In a recent editorial on seeding trials (1), Drs. Sox and Rennie Standard Therapy for Helicobacter pylori Infection in
said that they received a letter to the editor from Dr. Amos Egilman.
Patients Naive to Treatment
The doctor’s name was actually David Egilman. The online version The Results section of the meta-analysis of trials comparing of the editorial has been corrected.
sequential and standard triple therapies for treatment of Helicobacterpylori infection contained errors (1). Data regarding clarithromycin- Reference
resistant strains derived from the study by De Francesco and col- 1. Sox HC, Rennie D. Seeding trials: just say “no” [Editorial]. Ann Intern Med.
leagues (2) included 22 patients treated with sequential therapy (in- PERSONAE PHOTOGRAPHS
Annals of Internal Medicine invites submissions of Personae photo-graphs for our cover and offers a $500 prize for the best photographsubmitted each year. Personae photographs are pictures that catch peo-ple in the context of their lives and that capture personality. We preferblack-and-white print submissions but will accept color, slides, or digitalfiles. Please submit photographs or questions to Dr. Christine Laine([email protected]).
www.annals.org
16 September 2008 Annals of Internal Medicine Volume 149 • Number 6 439

Source: http://www.digest.net.cn/UpFiles/jylw/Hong_L-AIM1.pdf