Healthcare Economics of Effective Anticoagulation: ARoundtable Discussion Among Experts
Faculty members engaged in discussion at the held April 2, 2004, in San Francisco. Below are the highlightsof their conversation.
Moderator Wittkowsky: Our discussion willcover (1) the economic consequences ofstroke in atrial fibrillation [AF] and deepvein thrombosis [DVT] or pulmonaryembolism [PE], (2) the cost consequences ofadhering to guidelines versus nonadherence,(3) the primary causes of underuse of anticoagulants,(4) ways to improve adherenceto anticoagulation guidelines, and (5) monitoringoutcomes in patients who are receivinganticoagulants to find out if there is acost benefit in providing prophylaxis forvenous thromboembolism [VTE] or in providingappropriate stroke prevention inpatients with AF.
With those broad topics as a guide, let'sput on our economist hats and think aboutthe healthcare system economics associatedwith patients who develop stroke orDVT or PE as a consequence of inadequatetreatment.
The Cost of Stroke
Dr Morrow: I'll start with a case study aboutthe costs of one stroke. A 79-year-old manwas diagnosed with AF last year. He wasdiagnosed on a Monday, given a prescriptionfor warfarin on a Wednesday [that he didn'tfill until a day later], and on that Friday hehad a stroke. Obviously, he was inadequatelyanticoagulated. The stroke was not amotor stroke; it was a short-term memorystroke, which took out his ability to driveand to function as a human being. What arethe economic consequences? The familywill need to sell their home and find a placefor him and his wife.
Moderator Wittkowsky: What could haveprevented that consequence?
Dr Morrow: Several things. He had at least3 of the high-risk issues. He was over 75; hehad a history of uncontrolled hypertension,with a blood pressure at that time of about180 to 190 systolic; and he has coronarydisease.
Dr Caro: The American Heart Associationestimate of $54 billion for stroke costsincludes indirect costs, but only in terms oflost productivity and so on. I don't believe itincludes the costs to the family. It was interestingthat that was your first reaction to theeconomics of a stroke. These economics—from the personal point of view—are rarelyconsidered in any of the analyses of stroke.
In terms of actual hospital costs, currentlyin the United States it costs on averageabout $7000 for a stroke admission, butthere is an enormous range. A person thatcomes in and dies very quickly is relativelyinexpensive. You'll also see admissions inthe million-dollar range. So it's a hugerange, but the average is around $7000.
Dr Morrow: The next largest cost is the subacutecost for rehabilitation. The averagerehabilitation is 2 weeks. But depending onthe neurological loss it could be severalmonths or longer in rehab or a step-downnursing home. And then you have homecare, readmissions for side effects, and thelong-term care facilities. These and so manyother costs—for meals on wheels, the socialservices, the transportation services, assistedliving, the cost to the family for lost productivity—become practically immeasurablebecause they're coming from so many differentareas.
Moderator Wittkowsky: So, who pays for allof this?
Dr Balfour: The government pays for abouttwo thirds of all the healthcare costs now, andI'm sure the family pays for some, but itdepends on the situation. If they're in a nursinghome the federal government generallypays for 100 days, and then after that it'seither onto Medicaid or to the family.Nursing homes cost anywhere from $3000 to$4500 a month now, and that doesn't includeother services such as physical therapy.
The Cost of Prevention
Moderator Wittkowsky: The next logicalquestion, then, is what is the comparativecost of stroke prevention? What does it coston an annual basis to prevent stroke? Perhapsone of our pharmacists can comment.
Journal ofManaged Care Pharmacy
Mr Nelson: A small study in the looking at thecosts of managing warfarin therapy in anoutpatient pharmacy anticoagulation cliniccalculated the per-patient per-month costsfor the drug, monitoring, and labor to be$51.1 So the cost of prevention with warfarinis actually low for just those 3 elements.
Mr. Rice: The view of the managed careorganization may vary based on the contractualresponsibility and the flow of money. Insome medical contracts, the hospitalization,the home health, and the rehabilitation maybe separate contracts from one another.
Moderator Wittkowsky: So whose responsibilityis it to pay for stroke prevention?
Mr. Rice: From a pharmacy standpoint, werecognize that physicians need to put morepatients on the only available product thatcan be taken orally [warfarin], but we alsorecognize the complexities associated withthe therapy itself. There are titrationsrequired and changes can be handled overthe telephone, such as the dose that ischanged by the physician verbally to thepatient.
Dr Schaecher: From a health plan's perspective,I believe anticoagulation is an issuebetween the patient and the physician,regardless of the payment mechanism, capitatedor not. Managed care's obligation inthis arena is to create as few barriers as possibleto allow for adequate anticoagulation,and perhaps even assist them to do whatthey need to do. In terms of costs, our healthplan is a commercial product so we don'thave nearly as much AF as you're going tosee necessarily in a Medicare product.
Ms Simonson: Our integrated system isquite large and we lose money on our anticoagulationclinic when you look at the costversus reimbursement for visits. But in thepharmacy-run clinic, patients stay in range70% of the time.
Dr Menzin: The costs depend partly also onthe model of care. We studied anticoagulationservices provided by 3 different managedcare plans, tracking staff time,laboratory costs, and overhead. If you add indrug costs, it was around $600 a year foranticoagulation clinic costs.
Dr Go: At Kaiser Permanente of NorthernCalifornia, which is fairly large, fully integrated,and serves more than 3.2 millionlives, there's been a major commitment toanticoagulation. We have 21 pharmacistswho run anticoagulation clinics and we'veachieved some efficiencies. For us it's all thesame bottom line, and so we've taken awaysome of the financial disincentives thatexist in other different healthcare systems.
Barriers to Cost-effective Anticoagulation
Moderator Wittkowsky: What are some ofthe barriers to cost-effective anticoagulation?
Dr Schaecher: Utah is fairly rural. Though asignificant proportion of our population iscongregated along the 90-mile corridor of theWasatch Mountains, we have many patientsspread out over large geographic areas.Thus, we have some obvious geographic barriersin getting these patients to anticoagulationclinics. In addition, I've observed in ourarea that clinician providers often don'tknow the optimal way to treat and are afraidof losing patients. Anticoagulation clinics canhelp doctors get over these barriers.
Dr Balfour: I had anticipated that physicianswould not want to turn their patientsover, but we've actually had the oppositeexperience. It was an avalanche. We had tostop patients from entering the anticoagulationclinic. We're in an urban area, and thatmay account for the lack of physician resistance.Our main barrier has always beencost. It's very frustrating that we treatMedicare patients and they receive all thebenefit of the telephonic anticoagulationclinic, but we cannot charge them unlessthere's a face-to-face visit.
Dr Simonson: It's very frustrating. We tele-manageabout 120 patients a day with about1 pharmacist and 2 medical assistants. Wetry to do face-to-face visits whenever possible(about 30-40 a day). Cost is the onlyreason these visits are not available formore groups. You really benefit with thismodel but because of Medicare reimbursementand budget silos, it just doesn't lookthat way on paper.
Causes of Anticoagulation Underuse
Moderator Wittkowsky: Changing the focusof our discussion to DVT prophylaxis, theNABOR [National Anticoagulation Benchmarkand Outcomes Report] program hasrecently addressed some of the general causesof underuse of DVT prophylaxis in hospitalsettings.
Dr Colgan: In our research, hip fracturerepair patients, who are most probably thefrail elderly and at the highest risk of anevent, were least likely to receive VTE prophylaxis.We guess that it's the patient frailtyand age that causes the fear of a bleedingevent, which overrides the aversion to a possibleDVT or a pulmonary emboli. The durationof treatment is also problematic. About3 of 10 of the total knee and total hipreplacement patients are receiving only 3 to4 days of therapy.
Moderator Wittkowsky: What are the keyfactors that influence underprescribing, underuse,and shorter durations?
Dr Schaecher: From a physician point ofview, it all gets down to fear. The fear of thecomplications of overanticoagulation faroutweigh, in the physician's mind, the fearof underanticoagulation.
Dr Menzin: But bleeding risk is probably themajor factor regarding who does and doesnot receive anticoagulation therapy. Thereare also patient-related factors, education,and health literacy. It's a complicated regimenrequiring a lot of communication withthe nurses and the pharmacists over thetelephone, so that adds to the burden onpatients. And of course there are people whoare not candidates for warfarin at all, such asthose with dementia or alcoholism. It can behard to identify those patients.
Mr. Rice: Patients are also fearful. If the pharmacyand physician do a good job counselingpatients, they understand that there's a riskof bleeding. Also, many patients are dischargedon the low-molecular-weightheparins that are still not routinely stockedin retail pharmacy stores. They may haveproblems getting it filled at a pharmacy, andthen they may also be surprised by the costs.
Mr Kus: Yes, the retail economics may alsoplay a factor with patients in terms of anticoagulationunderuse. We're in the city ofDetroit with a very high population of uninsured,so we've made every effort to workwith the drug company for a replacement-typeprogram.
Dr Morrow: There are really 2 issues relatedto anticoagulation underuse. One is notstarting a medication, but the other is notcontinuing or having falloff in the persistency.One of the big problems with continuationis simply getting the tests done.
Focus on DVT and "The Handoff"
Moderator Wittkowsky: Are the issues ofunderutilization in the prevention groupsthat relevant to patients who are beingtreated for acute DVT or PE?
Dr Simonson: Within our setting, we have amuch more organized system for treatingthose patients. We know that we need totreat them and get them out of the hospitalquicker. They're not on anticoagulation longterm (longer than 3-6 months), even thoughthere is some data that shows that theyprobably should be on it longer.
Dr Go: Even 3 to 6 months of treatment ismore tolerable than chronic therapy. It'salso symptomatic, unlike something likehypertension where it's so hard to get folksto take a drug chronically when they feelfine or don't sense the long-term risks. Sowhen patients are brought into the hospitalbecause of something highly symptomaticlike an acute PE and DVT, it's much easier totreat. We also have a transition system fromPE to outpatient, which makes it easierbecause they get education and they see alight at the end of the tunnel.
Moderator Wittkowsky: What can we dowithin our healthcare systems to increasethe number of patients who receive appropriateDVT prophylaxis when they areadmitted to the hospital for surgical proceduresor for acute medical illnesses?
Dr Simonson: We've helped a number ofinstitutions develop programs to improveDVT/PE prophylaxis. Most people don'trealize the numbers of patients that diefrom PE. The healthcare providers areshocked when they get that information.So hospital-wide education about screeningpatients when they come in is the firstpriority.
Dr Schaecher: If you look at the dynamicsof how they go about admitting patients andtaking care of patients, they're in a hurry,they're running. So prophylaxis is oftentimesskipped.
Mr Kus: In our institution, inpatient pharmacydirectors are now charged by the chiefmedical officer to take more accountabilityfor anticoagulation. When prescriptionscome into the pharmacy with a diagnosispossibly requiring anticoagulation, it's theirresponsibility to determine if the patient is acandidate for anticoagulation at discharge.
Dr Colgan: Vanderbilt has presented theirprogram as an example of excellent practice.They have an electronic medical record thatprints a list of patients at risk, for example,if a patient has-- been immobile for 3 days ormore. This type of sophisticated systemwould certainly help in making sure morepatients were on prophylaxis therapy.
Dr Balfour: Another problem is that hospitalsuse different types of electronic medicalrecord systems. We're not able to talk toeach other.
Dr Caro: It's hard to say anything negativeabout education, but considering we've nowhad 20 years of experience in trying toimprove anticoagulation, at least with AF,I'm wondering if more of the educationinterventions we've been discussing willreally change the picture?
Mr. Rice: We've also got new drugs comingout that should simplify some of theseprocesses. It will require a combination,but I think technology and new drugs aregoing to significantly positively impactthis area.
Mr Wilson: If you consider the day-to-daypractice of a physician, and if you look athow little time they have to spend with anygiven patient, you'll see that even with electronicmedical records, starting anticoagulanttherapy and monitoring those folkssuccessfully is more difficult than managingthe average patient. So any new therapy thatmakes the patient easier to deal with provides a significant step forward in helping toget more patients on therapy.
Dr Colgan: When we looked to practice sitesin terms of following guidelines, we found thatwith the exception of AF the VA [VeteransAdministration] was about 14 points betterthan everybody else, and the academics wereabout 10 to 15 points better than other practicesites. So there is no question that place ofpractice and local education has a great effecton how patients are treated.
Impact of New Treatments
Dr Colgan: Dr Morrow and Mr Rice bothtalked about new drug technology, and I'mcurious about how ximelagatran will impactanticoagulation. What will be this drug'seffect on nontreatment, duration of treatment,and bridging?
Mr. Rice: We won't know until the productsare available, but the physicians' andpatients' fear of bleed will be reducedenough so that it's prescribed more. Wethink of older drugs as being simplistic,but a drug like warfarin has to be prescribedand monitored, it's a very complexdrug. I'm anxious for us to have otheroptions that are oral. We have injectableoptions right now, but they have their owncomplexities. I think the newer nonwarfarinoral products will have an impact.
Dr Morrow: They will have somewhat of animpact because there are fewer side effectsand not as much monitoring. But we still willneed an automated mechanism of followingthese patients or evaluating them.
Dr Simonson: I think the new drug, ximelagatran,will make a significant difference fora patient coming out of a hospital. Physicianswill probably want to use it right awaybecause it is therapeutic within 4 hours.Within the hospital, that is what's going tohappen unless the patient is contraindicatedfor ximelagatran.
Dr Go: Providers will take their time integratingthe new products into practice. Inthe late 1980s and 1990s, we saw a massiveincrease in the use of warfarin only becausethey had more experience and they hadbeen hounded by guidelines. It takes timebecause there's a doubt about whether ornot trials translate to practice.
Mr Wilson: I think this product will help interms of the initial prescription, but I thinkyou're going to have falloff as you do withevery chronic medication, as with theabysmal persistence rates for medicationsfor chronic diseases like high blood pressure,lipids, or diabetes.
Dr Menzin: Even within a specialized anticoagulationclinic, it's hard to keep peopleon therapy. So, yes, getting them on therapyis one thing, keeping them on is another.
Dr Colgan: Measurement might also be animportant tool in terms of bridging the gapwith guidelines.
Moderator Wittkowsky: Based on the discussionthat we've had today, what one thingwould you like to implement in your systemtoday to improve adherence to guidelines?
Dr Balfour: I think the electronic medicalrecord will give you better adherence toguidelines, protocols, and pathways and leadto better care. I'm sure it can be used to generatepatient information, but there areother systems that can do that also.
Dr Colgan: I have to agree that electronicsystems and the ability to remind people isthe first and foremost step toward improvedcare.
Dr Caro: The costs of stroke and DVT/PEprevention are complex, and the data maynot be readily available. So, from theresearch perspective I think we need towork on clarifying these cost data.
Dr Go: Combining the emerging informationtechnology with real-time feedback willbe helpful. Capturing data from your ownhealth system will be key to changing practiceand producing better outcomes. Mostproviders care about outcomes rather thanthe costs per se, and with real-time feedbackI think we can make a persuasive argumentfor increasing anticoagulation use.
Mr Kus: In our system, we are alreadyinvesting a lot of money in our electronicmedical records to get those prompts toremind physicians, nurses, and pharmacistsabout what they should be doing.
Dr Menzin: How do we know when we'vereached a limit to what we can do withcurrent anticoagulation? You can pushtechnologies and so forth, but at somepoint there are the tough issues of physician—patient interactions, education aboutthe condition, issues of health literacy, andmaybe health disparities.
Dr Morrow: Computerized medical recordswill help but even manual check-off listswould be useful. Public awareness is obviouslydriving a whole lot of medical care rightnow. We also need systematic reminders topatients using highly automated mechanisms,perhaps in a disease management approach,which does not cost much using the Net.
Mr Nelson: There is already a recognition ofthe liabilities of warfarin as an anticoagulant,but unfortunately not a lot is done about it. Iagree that we may have reached a pointwhere we have done all we can to improvethe situation with existing technology. So I'mhoping there will soon be new technologiesand new drugs available, and that I can helpmy organization understand their potentialvalue and the opportunities for improvingthe management of these patients.
Mr. Rice: I can simplify what I'd like to do into3 words: value, empowerment, and accountability.As we add technology and otherprocesses, we have more opportunities to besuccessful, but I still think you need to createthe value proposition so that it will be done.You also need to create the empowerment forpeople to do it. And then you create accountability,with report cards or other processes.
Dr Schaecher: As medical director for ahealth plan, I can provide easier access to theanticoagulant medications. And I can provideeducational opportunities based on claimsfor medications to providers to help remindthem of what's out there and perhaps feedbackon how poorly they're doing. Unfortunately,we're somewhat limited other thanproviding easy access and trying to reducebarriers. So mainly I can be an agent forchange with the control I have with the drugs.
Dr Simonson: We've already implemented anumber of things, so our main goal at thispoint is measuring how we are doing in ourintegrated system.
Dr Wilson: As a research organizationinside a large health plan, the most importantthing that we do every day is put informationin the hands of the decision makersat the point they're making the decision,and in a format that they understand. Wecan't forget that there are a lot of differentdecision makers even in this one diseasestate. It's a P&T [pharmacy and therapeutics]committee that regulates access to certainproducts; it's physicians who decidewhether or not to prescribe a product; andit's a patient who decides whether or notthey're going to continue to take that product.The choice of technologies for deliveringinformation to these diverse audiences iscertainly important, but we need to rememberto communicate our messages in a languagethat each audience understands.
J Manag Care Pharm.
1. Anderson RJ. Cost analysis of a managed care decentralizedpharmacy anticoagulation services. 2004;10:159-165.