The NEW ENGLAND JOURNAL of MEDICINE Perspective march 24, 2011 Under Siege — The Individual Mandate for Health Insurance and Its Alternatives Jonathan Oberlander, Ph.D. he battle over the Patient Protection and Af- law, be provided: either the in- dividual pays or the taxpayers Tfordable Care Act (ACA) rages on: in January, pay. A free ride on government is House Republicans passed legislation repealing the not libertarian.”1 The Massachu- ACA, but the measure failed to clear the Democratic- setts mandate has in fact encour- aged healthy people to obtain controlled Senate. Although Re- wise attract disproportionately coverage.2 publicans’ best chance to overturn sicker, costlier enrollees. That But Massachusetts proved to be the law won’t come until 2013 — adverse selection would drive up only an oasis of bipartisanship. and then only if they win a major- premium costs and threaten ex- Among Republicans in Washing- ity in both Congressional houses changes’ stability. ton, pro-mandate arguments about and the presidency in 2012 — Reformers had hoped the man- personal responsibility gave way they may meanwhile pursue tar- date would also confer political to concerns over individual lib- geted repeal of controversial pro- advantages. Some Republicans, erty and the political priority of visions. including former Massachusetts handing the Obama administra- No provision is currently more governor Mitt Romney, previously tion a defeat. beleaguered than the individual supported the policy. In a 2006 The mandate now confronts a mandate to obtain health insur- opinion piece, Romney defended legal and political backlash. Flori- ance or pay a penalty. Many ana- Massachusetts’ decision to im- da’s Roger Vinson recently became lysts view this mandate as crucial pose penalties on people who the second federal judge to deem to ensuring that healthier people didn’t purchase insurance as a it unconstitutional, and the issue join state-based insurance ex- “personal responsibility principle. appears headed for the Supreme changes: since the law prohibits Some of my libertarian friends Court. Some state legislatures are insurers from charging higher balk at what looks like an indi- seeking to block the mandate’s premiums to or turning away vidual mandate,” he wrote. “But implementation. A few Democrat- people with preexisting condi- remember, someone has to pay ic senators, including Claire Mc- tions, exchanges would other- for the health care that must, by Caskill (MO) and Ben Nelson (NE), n engl j med 364;12 nejm.org march 24, 2011 1085 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. PERSPECTIVE The Individual Mandate and Its Alternatives who are up for reelection in 2012, law. Were it to fall, they would when people’s insurance coverage say they’d like to find alterna- not endorse the ACA, but move could shift over time among pri- tives to the mandate. And while on to attack other controversial vate plans in the exchange, pri- conservatives rail against the provisions. vate plans outside the exchange, mandate and politically vulnera- If the mandate cannot be sus- Medicaid, employer-based cover- ble moderates run away from it, tained, alternatives exist. The age, and eventually, Medicare. many liberals support it reluctant- country could adopt single-payer, Another alternative would be ly because of concerns that insur- tax-financed national health in- to automatically enroll people in ance remains unaffordable and surance as a universal entitle- health insurance plans, permit- that the requirement is a gift to ment. There would be no need ting them to opt out. Auto-enroll- the insurance industry. for penalties and no worries about ment would occur primarily at the Increasingly, Democrats may adverse selection, since the gov- workplace but could also happen wonder whether the provision is ernment would operate one risk at state offices such as the Divi- an albatross that should be jetti- pool. Enrollment would be much sion of Motor Vehicles. This mod- soned to save reform. Inasmuch easier to administer and universal el could have bipartisan appeal: as the health care reform debate coverage would be ensured. Yet a 2009 health care reform bill co- is defined by the mandate, Demo- single payer would require com- sponsored by Republican Senators crats have a problem. The policy pulsory participation, new taxes, Richard Burr (NC) and Tom Co- is highly unpopular — 76% of and the political transformation burn (OK) and Republican Con- 3 Americans view it unfavorably necessary to displace private in- gressmen Paul Ryan (WI) and — and makes reform seem puni- surance. It remains infeasible. Devin Nunes (CA) relied on auto- 5 tive. Although Democrats can Staying within the reform’s ex- enrollment. Auto-enrollment can highlight other consumer-friendly isting boundaries, one alternative be combined with a premium provisions that target health in- is to limit enrollment in the ex- penalty for people who opt out surers, in the mandate fight they changes to a fixed period each but later decide to purchase cover- are allied with the insurance in- year and impose premium penal- age. Medicare Part B, which pays dustry. Furthermore, President ties for eligible people who choose for physicians’ services, works Barack Obama could in 2012 be to wait and buy coverage later — this way. in the uncomfortable position of and to make the penalty apply Once again, though, the late- defending a mandate that he ar- not just the first time they pur- enrollment penalty might have to gued against in the 2008 Demo- chase insurance, but across their be substantial for auto-enrollment 4 cratic primary. (The mandate’s lifetimes. This model is employed to effectively induce healthier peo- prominence also complicates by the Medicare Part D program ple to pay for insurance coverage. Romney’s potential bid for the for prescription-drug coverage, Inertia alone may not be a suffi- Republican presidential nomina- which the Bush administration ciently strong force to get younger, tion.) and many Congressional Repub- healthier workers to stay insured, Still, it’s not clear how much licans supported. given high and rising insurance Democrats would gain politically However, to induce healthy un- premiums. Auto-enrollment may by dumping the mandate. Many insured people to sign up, the late also not work as well outside the Republicans also oppose the ACA’s penalty might have to be substan- workplace or for workers’ depen- 4 requirements that employers offer tial, in which case this arrange- dents. coverage or pay a penalty, the ment would be operating similar- Substituting either of these al- Medicaid expansion, increased ly to the mandate. As the health ternatives for the mandate would, government regulatory authority care economist Len Nichols points as economist Jonathan Gruber ar- over the insurance industry, re- out, if we don’t have the political gues, attenuate the ACA’s bene- ductions in projected Medicare will to impose a strong penalty fits, resulting in an increased savings, and the cost of the sub- in conjunction with an individual number of uninsured Americans sidies for the uninsured to pur- mandate, we probably wouldn’t and higher premiums in the chase insurance. Opponents of have the will to impose one as nongroup insurance market (as health care reform are attacking part of a fixed enrollment system. healthier people decline cover- 4 the mandate because it’s the most Furthermore, it’s not clear how a age). Implementing the rest of politically vulnerable part of the lifetime late penalty would work the ACA without any substitute 1086 n engl j med 364;12 nejm.org march 24, 2011 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. PERSPECTIVE The Individual Mandate and Its Alternatives policy would similarly reduce cov- native rationale, especially since Disclosure forms provided by the author erage gains and destabilize insur- any provision meant to ensure calree aatv NaiElaJMble.o wrgi.th the full text of this arti- ance pools. The substantive case broad participation in insurance for the mandate is still strong, pools must include financial From the University of North Carolina, Cha- even if its political and legal penalties. The mandate’s de- pel Hill. foundations are shaken. fenders could again invoke the This article (10.1056/NEJMp1101240) was It’s also unclear whether past rhetoric of personal responsibil- published on February 16, 2011, at NEJM.org. Republican support for alternative ity. Or they could emphasize policies would be sustained or, as that the mandate makes possi- W1. eR’voem fonueny d Ma .w Haye. aWltahl l cSatree tf oJor uernvearl.y oAnperi?l with the individual mandate, evap- ble the insurance reforms that 11, 2006. orate in the heat of the political guarantee the availability of cov- 2. Chandra A, Gruber J, McKnight R. The im- spotlight. Insofar as Republicans erage to sick people.3 Alterna- pdeonrtcaen cfreo mof tMhea sisnadcihviudsueatlt sm. Nan dEantgel —J M eevdi- continue to support alternatives, tively, reformers could appeal to 2011;364:293-5. they will be interested in them as the mandate’s communitarian 3. Kaiser Family Foundation/Harvard School part of a broader conservative foundations, arguing that there oagf ePnudbal ifco rH theea l1t1h2. tTh hCeo npgurbelsics’.s J ahneuaalrtyh 2c0a1r1e. health care reform package, not are some public programs — (http://www.kff.org/kaiserpolls/upload/ a means of bolstering the ACA. such as Social Security and 8134-F.pdf.) Republicans’ good chances of Medicare — that produce in- 4in. dGivriudbuearl mJ. aHnedaaltteh. Wcaares hrienfgotromn ,w DitCh:o Cuetn ttheer winning control of the Senate in valuable social benefits and that for American Progress, February 9, 2011. 2012 further reduce their incen- succeed because everyone par- (http://www.americanprogress.org/issues/ tives to cooperate now on an alter- ticipates in them. Ultimately, 25.0 1C1o/0b2u/rgnr uT.b Ienrd_miviadnudaal taeu.htotm-eln.)rollment: an native plan. the furor over the mandate un- alternative to an individual mandate. (http:// What reformers may need derscores the reality that soli- coburn.senate.gov/public/index.cfm?a=Files even more than a policy alterna- darity remains elusive in U.S. .aSdeer0v4e6&bF8ible8_8i2d.=) e87f06bf-d429-4eac-8e7e- tive to the mandate is an alter- health policy. Copyright © 2011 Massachusetts Medical Society. Early Accelerated Approval for Highly Targeted Cancer Drugs Bruce A. Chabner, M.D. he striking results of recent or rarely, symptomatic relief. for physicians and patients. As Tphase 1 trials of targeted can- These trials are invaluable for es- Miller and Joffe recently empha- cer drugs have provoked serious tablishing the benefits of new sized, the concept of equipoise discussion about shortening the drugs in instances where reason- (uncertainty as to which treat- road to drug approval. A typical able alternative therapies exist. ment carries greater benefit), cancer drug takes 7 years from But phase 3 trials are expen- underlies the ethical basis for entry into human trials to approv- sive and time-consuming, usual- randomization.1 Strong early re- al by the Food and Drug Admin- ly taking at least 2 to 3 years to sults in phase 1 or 2 may tip the istration (FDA), which requires reach survival end points. The balance and argue against delay- proof of efficacy in “well-con- news of a highly successful new ing access by performing phase trolled clinical trials.” In these compound in phase 1 or 2 rap- 3 trials, but Miller and Joffe con- randomized phase 3 trials, a new idly reaches physicians and pa- tend that such trials may none- treatment alone or added to a tients, creating demand for early theless be justified by the need drug combination is compared access. For drugs aimed at dis- to establish new drugs’ long-term with a “standard” drug or com- eases with limited effective treat- efficacy. bination. The most convincing ments, delaying access during New understanding of the end point for such trials is im- phase 2 and 3 trials creates dif- molecular and genetic lesions that proved survival, although the FDA ficult ethical issues for regula- cause cancer has sharpened the has accepted surrogates such as tory agencies and pharmaceutical discussion, however, by enabling tumor progression, response rate, sponsors and agonizing decisions rapid development of drugs that n engl j med 364;12 nejm.org march 24, 2011 1087 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. PERSPECTIVE Accelerated Approval for Targeted Cancer Drugs Mutations in NRAS Activation of PDGFR-β and IGF-1 receptor Dimerization of BRAF and CRAF PLX4032 Increased expression of the inhibits BRAF P13K–AKT signaling pathway BRAF V600E — protein from mutated BRAF Increase in CRAF MEK Increased expression of COT ERK Mechanism of Action of PLX4032 in Melanoma. BRAF dimerizes with its partner CRAF and acts as a relay point for signals for normal cell growth and survival. The BRAF V600E mutation, found in 60% of melanomas, signals constitutively and drives the malignant behavior of the tumor. PLX4032, an inhibitor 1 of BRAF, kills BRAF-dependent tumors and has proved to be highly effective in phase 1 trials. Further mutations in BRAF or activat- ing mutations in NRAS, changes in expression of mitogen-activated protein kinase kinase kinase 8 (MAP3K8, or COT) or elements of the PI3 kinase pathway, or activation of alternative signaling pathways such as platelet-derived growth factor receptor β (PDGFR-β) have led to resistance to BRAF inhibition in experimental settings (labels in red) and in clinical settings (labels in green) and form the basis for new therapeutic strategies. IGF-1 denotes insulin-like growth factor 1. The image of CRAF was produced with the use of the Chimera software and is based on the 3omv crystal in the Protein Data Bank (www.wwpdb.org). specifically target tumors bear- or stable disease for at least 8 zine, a standard agent with a 2 ing aberrant signaling pathways. weeks) exceeded 90% in both response rate of 15%; crossover Inhibitors of the BCR-ABL kinase trials, with minimal toxicity. Be- from the control group to the in chronic myelogenous leuke- cause only a fraction of patients experimental group at the time mia (CML) and inhibitors of the with NSCLC and melanoma have of disease progression was not epidermal growth factor recep- tumors harboring the aberrant allowed, to preserve the integrity tor (EGFR) in EGFR-mutated non– pathways, a biomarker test for of the survival end point. (The 3 small-cell lung cancer (NSCLC) the mutations was required to trial has now completed enroll- have replaced cytotoxic chemo- select appropriate trial subjects. ment and reached its survival end therapy for these indications. Two Drugs such as PLX4032 and point and is being amended to new drugs have produced excel- crizotinib normally undergo a sec- allow crossover.) In its phase 3 lent results in phase 1 trials ond phase 2 trial (involving 30 trial for newly diagnosed NSCLC, against cancers that responded to 100 patients) to confirm their crizotinib is being compared with poorly to standard treatments: activity as second- or third-line standard chemotherapy, which PLX4032 yielded an 81% response therapy. Because of their early has a 30 to 40% response rate, a rate in 38 patients with BRAF promise, both of these drugs en- limited survival benefit, and sub- 4 mutated melanoma (see diagram), tered phase 3 trials directly after stantial toxicity. Patients in the and crizotinib had a 57% re- phase 1. Patients with newly di- control group who have disease sponse rate in 82 patients with agnosed BRAF-mutation–positive progression may receive crizo- 5 the EML4-ALK fusion in NSCLC. melanoma were randomly as- tinib by entering a phase 2 trial. Rates of disease control (response signed to PLX4032 or dacarba- Because we can now define 1088 n engl j med 364;12 nejm.org march 24, 2011 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. PERSPECTIVE Accelerated Approval for Targeted Cancer Drugs patient subgroups with high re- on the basis of surrogate end ated approval if there’s a clear sponse rates in phase 1 trials, points in phase 2 — most com- unmet need. Randomized com- performance of phase 3 trials for monly, a clearly demonstrable tu- parisons with minimally effec- these drugs raises important is- mor response rate (often 20 to tive treatments or placebo should sues. If patients with incurable 30%). Accelerated approval is not be required. Specific end disease who have the right bio- awarded with the stipulation that points for early approval should marker for response are informed definitive trials, with a survival be maximally flexible and ad- of these impressive early results, end point, must be conducted af- justed according to the targeted they will want and perhaps deserve ter approval. Of the 23 oncologic disease and the effectiveness access to the new drug and may drugs given accelerated approval and toxicity of alternative thera- not accept random assignment between 1993 and 2008, two pies. More extensive study would to a modestly effective and toxic were ultimately withdrawn from be required for less effective or standard agent. The phase 3 trial the U.S. market — gemtuzumab more toxic experimental agents. may lack equipoise in the eyes because of toxicity and gefitinib The post–phase-1 approval of of both physicians and patients. because of lack of eff icacy. new agents meeting these goals There are alternatives for ex- A 2008 study of accelerated is highly unlikely to have signifi- pediting access to new drugs for approvals, from the Government cant negative consequences such patients who are not eligible for Accountability Office, concluded as ineffective treatment or un- or accepted into phase 3 trials. that the FDA hasn’t effectively foreseen, overwhelming toxic ef- A phase 2 trial may be opened enforced requirements for post- fects. Early approval would allow simultaneously for patients with approval trials. However, my re- rapid general access to treatment, progressive disease after initial view of accelerated approvals while further evaluation focused therapy. Such trials are open for reveals that most such agents are on defining optimal doses, sched- both PLX4032 and crizotinib, but integral to standard cancer treat- ules, and drug combinations; they accrue limited numbers of ment. Indeed, the more cogent long-term benefits; toxic effects; patients and are conducted in question is whether accelerated and resistance mechanisms. When only a few centers. approval can take place after striking clinical results have been In addition, compassionate-use phase 1. demonstrated in a sizable, readi- protocols allow sponsors to dis- Given trialists’ ability to define ly identifiable patient population tribute experimental drugs on a patient subgroups with responsive in phase 1, the journey to drug case-by-case basis. This mecha- tumors in phase 1 trials, I pro- approval should not be prolonged. nism was widely used to provide pose that for diseases lacking Disclosure forms provided by the author are available with the full text of this arti- early access to experimental AIDS therapies that meaningfully ex- cle at NEJM.org. drugs, with few reports of unex- tend survival, the FDA should set pected adverse drug reactions. flexible standards permitting ac- From the Massachusetts General Hospital Cancer Center, Boston. But sponsors often view compas- celerated approval of new drugs sionate use as competing with after phase 1. These standards 1. Miller FG, Joffe S. Equipose and the di- lemma of randomized clinical trials. N Engl J trial accrual and relieving pres- could be satisfied by the results Med 2011;364:476-80. sure on the FDA for marketing of expanded and targeted phase 1 2. McClellan M, Benner J, Schilsky R, et al. approval. Neither crizotinib nor drug testing, as in the case of An accelerated pathway for targeted cancer therapies. Nat Rev Drug Discov 2011;10:79- PLX4032 was available through PLX4032. This strategy requires 80. an expanded-access protocol when the early evaluation and validation 3. Sequist LV, Martins RG, Spigel D, et al. its phase 3 trial opened, though of a companion biomarker for First-line gefitinib in patients with advanced non-small-cell lung cancer harboring somat- such protocols are now available patient selection. Conceivably, ic EGFR mutations. J Clin Oncol 2008;26: in a limited number of centers. non-targeted agents could also 2442-9. [Erratum, J Clin Oncol 2008;26:3472.] A mechanism for early FDA satisfy criteria for approval after 4. Flaherty KT, Puzanov I, Kim KB, et al. Inhi- bition of mutated, activated BRAF in metastat- approval, “accelerated approval,” phase 1. High response rates ic melanoma. N Engl J Med 2010;363:809-19. was introduced in 1992. Reserved (>50%), high disease-control rates 5. Kwak EL, Bang Y-J, Camidge DR, et al. for serious or life-threatening ill- (>75%), and an acceptable toxic- Anaplastic lymphoma kinase inhibition in non–small-cell lung cancer. N Engl J Med ness not effectively treated by ap- ity profile in a biomarker-defined 2010;363:1693-703. [Erratum, N Engl J Med proved medications, this mecha- population of 75 to 100 subjects 2011;364:588.] nism allows drugs to be registered should be sufficient for acceler- Copyright © 2011 Massachusetts Medical Society. n engl j med 364;12 nejm.org march 24, 2011 1089 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. PERSPECTIVE Tapping the Unmet Potential of Health IT Tapping the Unmet Potential of Health Information Technology Ann S. O’Malley, M.D., M.P.H. ealth information technolo- areas offering glimpses of both sorting and searching of lists and Hgy (HIT) holds promise for the potential and the challenges linking of listed problems to rele- facilitating vast improvements in associated with improved HIT use. vant portions of progress notes, care and, ultimately, in the health Few providers today, for exam- assessments, and treatment plans.3 of Americans,1,2 but achieving ple, can truly coordinate care — In addition, improving care that potential remains a daunting integrating care, in consultation coordination and health outcomes task. A recent article in the Los An- with patients and their relatives requires teamwork. Taking into geles Times described the new phe- and caregivers, across all of a account each team member’s train- nomenon of hiring computer-savvy patient’s conditions, needs, clini- ing, skill set, and expertise when undergraduate “scribes” to take cians, and health care settings.2-4 delegating tasks and defining notes for physicians during patient Outpatient practices and inpatient roles is critical to improving ef- encounters and enter the informa- facilities lack well-developed pro- ficiency, for both primary care tion into electronic health records cesses for exchanging informa- teams within a practice and inter- (EHRs) — a practice that suggests tion, both within their own walls specialty teams sharing patients’ how far we must go to develop and during care transitions. Poor care. HIT can support team-based EHRs that clinicians will embrace. care coordination negatively af- care with tools enabling team Of course, the most highly trained fects patients — particularly those members to identify patient care professional in the room need not with multiple chronic conditions goals and document and monitor be the one to enter data into the who account for an overwhelming progress using a shared care plan. computer, especially during an proportion of U.S. health care In outpatient practices, providers emergency, but the perceived need expenditures. report that electronic messaging for scribes and providers’ expe- HIT, especially if widely imple- and notification of staff about riences using EHRs3 raise impor- mented, can facilitate coordina- patient care tasks facilitate com- tant questions about both the tion by making information elec- munication about delegation and efficiency of care processes and tronically available at the point of task completion. Electronic mes- the usability of current EHRs. care. As clinical care processes saging can also enable real-time Although EHRs laudably pro- become more effective and effi- communication with specialists vide immediate access to patient cient, they can inform new HIT — if they’re on the same system data and electronic messaging capabilities that will better sup- — to determine whether a con- functions, clinicians have been port coordination. For example, sultation is necessary.2,3 frustrated by the difficulty of us- providers need to develop consis- Coordinating care for patients ing them to support care delivery tent notification processes to en- with complex health conditions and coordination.3 Transforming sure timely communication about who see multiple physicians can EHRs into effective clinical tools care transitions. Medical and nurs- also be supported by better HIT rather than a means of captur- ing professional societies could interoperability.3 The primary care ing information primarily for doc- work with HIT vendors to devel- team may be in the best position umentation and billing purpos- op standardized notification pro- to coordinate a patient’s care, but es will require progress on cedures, which could be imple- it will often need information multiple fronts. mented through refined criteria from other providers.4 Most cur- Clinical processes must evolve for the “meaningful use” of HIT.5 rent EHRs don’t adequately sup- so as to improve care and be more HIT can also better support care port data exchange across pro- responsive to patients’ needs, and coordination through the develop- viders and settings, so practices HIT’s capabilities must evolve ment of referral-tracking systems, communicate with outsiders pri- along with them. HIT has par- improved approaches to reconcil- marily on paper.3 To support infor- ticular potential in such areas as ing patients’ medications, and mation exchange, EHRs must pre- coordination of care, workflow expansion of “problem list” capa- sent data in standard ways, and efficiency and use of teams, clin- bilities — to avoid cluttering of separate organizations providing ical decision support, and popu- lists with redundant informa- services for the same patient need lation health management — all tion, for example, and permit to share information securely. 1090 n engl j med 364;12 nejm.org march 24, 2011 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. PERSPECTIVE Tapping the Unmet Potential of Health IT The Health Information Tech- payment reform could be accom- try’s many small primary care nology for Economic and Clini- panied by training for clinicians, practices can be overwhelmed by cal Health (HITECH) Act envi- residents, and medical students in it and will need to be convinced sions the secure exchange of data effective communication with pa- that EHRs are affordable, enhance across providers and settings oc- tients in the presence of an EHR. efficiency, and improve care. Then, curring through the Nationwide HIT can also provide tools to they will need extensive, ongoing Health Information Network, help inform decision making with support. Under HITECH, HIT Re- which will provide a common regard to diagnosis (with clini- gional Extension Centers will pro- 5 platform and protocols. Some cal prediction rules), prevention vide technical assistance, guid- states and communities are de- (reminders), disease management ance, and information on EHR veloping regional health informa- (registries), and treatment (elec- adoption and meaningful use, 2 5 tion exchanges, but most are still tronic prescribing tools). The particularly to such practices. in their infancy. It is hoped that use of computerized medication These centers could also compile these local and national efforts orders, generated with the help clinicians’ feedback for policy- might eventually be linked. Some of decision-support tools, is asso- makers and vendors, providing an have raised concerns, however, ciated with reductions in adverse infrastructure for shared learn- 2 about the sustainability of these drug events. But most current ing. Unfortunately, the United exchanges absent a stronger busi- commercial EHRs don’t provide States currently lacks an adequate- ness case to support them. or link to decision-support sys- ly trained workforce to support Clearly, HIT alone cannot tems, particularly for managing practices in these areas. transform our health care system: chronic care or selecting prefer- Continued research on clinical financial incentives must be re- ence-sensitive treatments. To per- care processes, the design and aligned to reward patient-centered mit the development of adequate use of HIT, and payment reform, care. Current fee-for-service reim- decision-support tools, the evi- as well as ongoing support for bursement encourages EHR use dence base must be expanded clinicians, will be key to the effec- for documentation of billable and actively maintained. tive and meaningful use of HIT. events rather than for tasks im- Finally, primary care practices Today’s EHRs do not sufficiently portant to the quality of care, increasingly must focus not just support aspects of care delivery such as coordination. Payment in- on individual patients but on that are vital to improving care novations such as bundled pay- whole populations, as they strive and controlling costs. ments and accountable care or- to function as medical homes.2 Disclosure forms provided by the author are available with the full text of this arti- ganizations aim to encourage A new orientation and effective cle at NEJM.org. providers to share accountability methods for shifting practices From the Center for Studying Health Sys- for outcomes. Such payment re- from reactive and acute-symptom tem Change, Washington, DC. form would offer clinicians in- care to approaches including pro- 1. Goldzweig CL, Towfigh A, Maglione M, centives to demand HIT capabili- active, planned care for both Shekelle PG. Costs and benefits of health in- ties that better support the clinical healthy and chronically ill popu- formation technology: new trends from the tasks required to improve pa- lations will be needed. Without literature. Health Aff (Millwood) 2009;28: w282-w293. tients’ health and would make HIT, it’s difficult to provide effec- 2. Bates DW, Bitton A. The future of health these activities important to the tive population-based care and information technology in the patient-cen- success of health care organiza- report quality metrics, but most tered medical home. Health Aff (Millwood) 2010;29:614-21. tions. commercial EHRs currently can- 3. O’Malley AS, Grossman JM, Cohen GR, Fee-for-service reimbursement not help identify which patients Kemper NM, Pham HH. Are electronic med- also makes it difficult for clini- in a population may need partic- ical records helpful for care coordination? Experiences of physician practices. J Gen In- cians to take the time to listen ular services. Registries are an- tern Med 2010;25:177-85. to patients’ concerns. Unfortu- other critical tool for population 4. Stille CJ, Jerant A, Bell D, Meltzer D, Elmore nately, adding an EHR to the health management and an area JG. Coordinating care across diseases, set- tings, and clinicians: a key role for the gener- clinical encounter can further where HIT applications could be alist in practice. Ann Intern Med 2005;142: distract clinicians from patients better developed and integrated 700-8. (which is one of the reasons that with EHRs. 5. Blumenthal D, Tavenner M. The “mean- ingful use” regulation for electronic health some emergency departments use Regardless of HIT’s potential records. N Engl J Med 2010;363:501-4. scribes). To counter this tendency, advantages, clinicians in the coun- Copyright © 2011 Massachusetts Medical Society. n engl j med 364;12 nejm.org march 24, 2011 1091 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved. The new england journal of medicine established in 1812 march 24, 2011 vol. 364 no. 12 Tiotropium versus Salmeterol for the Prevention of Exacerbations of COPD Claus Vogelmeier, M.D., Bettina Hederer, M.D., Thomas Glaab, M.D., Hendrik Schmidt, Ph.D., Maureen P.M.H. Rutten-van Mölken, Ph.D., Kai M. Beeh, M.D., Klaus F. Rabe, M.D., and Leonardo M. Fabbri, M.D., for the POET-COPD Investigators* A BS TR AC T BACKGROUND Treatment guidelines recommend the use of inhaled long-acting bronchodilators to From the Hospital of the Universities of alleviate symptoms and reduce the risk of exacerbations in patients with moderate-to- Giessen and Marburg, Marburg (C.V.); Boehringer Ingelheim, Ingelheim (B.H., very-severe chronic obstructive pulmonary disease (COPD) but do not specify whether T.G., H.S.); and insaf Respiratory Research a long-acting anticholinergic drug or a β2-agonist is the preferred agent. We investi- Institute, Wiesbaden (K.M.B.) — all in gated whether the anticholinergic drug tiotropium is superior to the β2-agonist nGoelromgya nAys;s tehses mInesntitt u(ItMe TfoAr) ,M Ereadsimcaul sT eUcnhi- salmeterol in preventing exacerbations of COPD. versity, Rotterdam (M.P.M.H.R.-M.); and Leiden University Medical Center, Leiden METHODS (K.F.R.) — both in the Netherlands; and In a 1-year, randomized, double-blind, double-dummy, parallel-group trial, we com- the University of Modena and Reggio Emilia, Modena, Italy (L.M.F.). Address pared the effect of treatment with 18 µg of tiotropium once daily with that of 50 µg reprint requests to Dr. Fabbri at the Sec- of salmeterol twice daily on the incidence of moderate or severe exacerbations in tion of Respiratory Diseases, Department patients with moderate-to-very-severe COPD and a history of exacerbations in the of Oncology, Hematology, and Pulmonary Diseases, University of Modena and Reg- preceding year. gio Emilia, Policlinico di Modena, Largo del Pozzo 71, I-41124 Modena, Italy, or at RESULTS Th e n e w e ngl a nd j o u r na l o f m e dic i n e hronic obstructive pulmonary dis- ate and severe exacerbations of COPD (hereinafter ease (COPD) is a leading cause of disabil- called exacerbations) in patients with moderate- City and death worldwide.1-3 Exacerbations to-very-severe COPD.17 The study was conducted of COPD indicate instability or worsening of the in accordance with the provisions of the Declara- patient’s clinical status and progression of the dis- tion of Helsinki (1996) and Good Clinical Prac- ease and have been associated with the develop- tice guidelines. All patients provided written in- ment of complications, an increased risk of sub- formed consent before any study procedure was sequent exacerbations, a worsening of coexisting performed. The scientific steering committee conditions, reduced health status and physical (which was made up of two of the academic in- activity, deterioration of lung function, and an vestigators and an external clinical researcher) increased risk of death.4-7 The prevention of ex- and three employees of Boehringer Ingelheim acerbations therefore constitutes a major goal of developed the design and concept of the study, treatment.1,2 approved the statistical plan, had full access to Therapy with a long-acting anticholinergic the data, and interpreted the data. Onsite moni- drug or a long-acting β2-agonist is recommended toring and site management were supported by a as first-line maintenance therapy in patients with contract research organization (PAREXEL). The moderate-to-very-severe COPD,1,2 since both of first draft of the manuscript and subsequent revi- these drugs reduce symptoms, improve quality sions were written by all the authors, and all the of life and lung function, and reduce the risk of authors made the decision to submit the manu- exacerbations and hospitalizations.8-12 However, script for publication. The statistical analysis was treatment guidelines do not specify whether a performed by an employee of the sponsor. All the long-acting anticholinergic drug or a β2-agonist authors had full access to the data and vouch for is the preferred agent.1,2 the accuracy and completeness of the data and the Comparative studies have indicated that tiotro- analyses, as well as the fidelity of the study to the pium is associated with a greater reduction in the protocol. (The protocol, including the statistical risk of exacerbations and exacerbation-related analysis plan, is available with the full text of this hospitalizations than is salmeterol, although the article at NEJM.org.) An independent ethics com- differences were not significant.13,14 These were mittee or institutional review board at each par- short-term studies (3 to 6 months in duration) ticipating center reviewed and approved the pro- and were not designed and powered to detect a tocol before commencement of the study. In difference in the risk of exacerbations. The Preven- addition, an independent data and safety moni- tion of Exacerbations with Tiotropium in COPD toring board and a mortality adjudication com- (POET-COPD) trial was specifically designed to mittee were established (Section 10 in the Sup- directly compare the effects of tiotropium with plementary Appendix, available at NEJM.org). those of salmeterol on the risk of moderate and severe exacerbations. A placebo group was not in- End Points cluded in the study, since there is substantial evi- The primary end point was the time to the first dence of the superiority of both tiotropium and exacerbation of COPD. The time to the first exac- salmeterol over placebo.8,12 Furthermore, a com- erbation was selected as the primary end point parison of two active-treatment groups is in line because it is less likely to be affected by the intro- with the recently growing relevance of compara- duction of additional therapies or by the occur- tive-effectiveness research to guidance regarding rence of multiple exacerbations in some pa- treatment decisions.15,16 tients.17 Secondary and safety end points included time-to-event end points, number-of-event end Me thods points, serious adverse events, and death (Section 2 in the Supplementary Appendix). Study Design and Oversight An exacerbation was defined as an increase We conducted a 1-year, randomized, double-blind, in or new onset of more than one symptom of double-dummy, parallel-group trial at 725 centers COPD (cough, sputum, wheezing, dyspnea, or in 25 countries to compare the effect of tiotropi- chest tightness), with at least one symptom um (Spiriva, Boehringer Ingelheim) with that of lasting 3 days or more and leading the patient’s salmeterol (Serevent, GlaxoSmithKline) on moder- attending physician to initiate treatment with 1094 n engl j med 364;12 nejm.org march 24, 2011 The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.