self-reported differences between cardiologists and heart failure specialists in the management of...

Self-reported differences between cardiologistsand heart failure specialists in the management ofchronic heart failure David Bello, MD, Nihir B. Shah, MD, Martin E. Edep, MD, Ida M. Tateo, MS, and Barry M. Massie, MD San Francisco, Calif

Background Heart failure (HF) is responsible for considerable mortality morbidity rates and resource utilization.Recently, several studies have reported improved outcomes when patients are managed by special HF clinics, but it is uncer-tain whether this improvement reflects differences in physician practices or other aspects of the operation of these clinics.

Objectives This study was designed to identify differences in HF management practices between general cardiologistsand cardiologists specializing in the treatment of patients with HF.

Methods A survey examining diagnostic and treatment practices in patients with HF was sent to a sample of cardiolo-gists derived from the American Medical Association Masterfile and to HF specialists who were members of the Society ofTransplant Cardiologists or principal investigators in HF trials. Responses were examined in relation to guidelines issued bythe Agency for Health care Policy and Research released 9 months previously.

Results In general both groups practice in conformity with published guidelines. However, there were important differ-ences between the practice patterns of general cardiologists and HF specialists. For instance, in patients being evaluated forthe first time, cardiologists reported using a chest radiograph to assist in the diagnosis more than did HF specialists (47% vs12%), whereas HF specialists were more likely to use an echocardiogram (73% vs 48%). Both groups were likely to evaluatetheir patients for ischemia and possible revascularization, even in patients not having angina. However, HF specialists tendedto use coronary angiography as the initial diagnostic test, whereas cardiologists were more likely to use stress testing. HF spe-cialists more often used angiotensin-converting enzyme inhibitors as part of their initial therapy in patients with mild to moder-ate HF (94% vs 86%) and during maintenance therapy (91% vs 80%). Also, HF specialists were more likely than cardiologiststo titrate angiotensin-converting enzyme inhibitors to higher doses (75% vs 35%), even in the presence of renal dysfunction.

Conclusion Cardiologists and HF specialists generally manage their patients in conformity with guidelines. However, inmany areas, such as angiotensin-converting enzyme inhibitor use, HF specialists do so more aggressively. These approachesmay, in part, explain the success of the HF clinic model and raise the possibility that some portion of the HF population maybe more optimally managed by cardiologists with a special interest in and additional training or experience with this condi-tion. (Am Heart J 1999;138:100-7.)

Chronic heart failure (HF) continues to be a majormedical burden, affecting more than 4 million patientsin the United States and with an estimated 465,000 newcases being diagnosed each year.1 Unlike most cardio-vascular conditions, HF is increasing in prevalence,causing more deaths, and consuming more health careresources; these trends are expected to continue as the

population ages. HF is responsible for nearly 1 millionhospitalizations in the United States annually,1,2 and theestimated costs for managing HF range from $15 to $30billion. Despite recent advances in the therapy, survivalof patients with HF remains poor, with only approxi-mately 50% of patients remaining alive after 5 years;morbidity rates also remain high.1 These discouragingstatistics have made improved outcomes and more cost-effective management a priority for both health authori-ties and health maintenance organizations.

One contributing factor for these grim figures has beenthe incomplete penetration of effective managementapproaches, such as the early recognition of HF, use ofangiotensin-converting enzyme (ACE) inhibitors, andintensive follow-up, into clinical practice. To address this,guidelines have been developed and disseminated.3,4

Nonetheless, compliance with key recommendations

From the Department of Medicine and Cardiovascular Research Institute of the Uni-versity of California, and the Cardiology Section of the Department of VeteransAffairs Medical Center, San Francisco. Submitted July 7, 1998; accepted October 12, 1998.Reprint requests: Barry M. Massie, MD, Cardiology Section (111-C), VeteransAffairs Medical Center, 4150 Clement St, San Francisco, CA 94121.0002-8703/99/$8.00 + 0 4/1/96976

See related Editorial on page 14.

American Heart JournalVolume 138, Number 1, Part 1 Bello et al 101

such as treatment with ACE inhibitors remains subopti-mal.5-10 Several studies have indicated that cardiolo-gists’ practices may conform more closely to those rec-ommended by guidelines than those of primary carephysicians,5,6 raising the possibility that specialty caremay improve outcomes, although this has yet to beconvincingly demonstrated. Another approach that isbeing increasingly used is the establishment of multifac-eted HF clinics under the guidance of cardiologists withspecial interest and expertise in treating patients withcomplex or severe HF, many of which are connected toheart transplantation programs.11-14 In uncontrolledstudies, these programs appear to have improvedpatient management and outcomes.

However, little information is available about how thepractices of these HF specialists differ from those ofother cardiologists. Therefore we surveyed a cohort ofHF specialists contemporaneously with a random sam-ple of cardiologists regarding their approaches to thediagnosis, assessment, and treatment of patients withHF and compared these responses with the guidelinespublished by the Agency for Health Care Policy andResearch 9 months previously.3

MethodsStudy population

The HF specialists cohort consisted of 118 cardiologists inthe United States who had participated as principal investiga-tors in published HF trials or were cardiologists who weremembers of the Society of Transplant Cardiologists. The datafor the cardiologists were obtained as part of a previouslypublished study comparing practices of cardiologists, familyand general practitioners, and internists.5 For that survey, 750cardiologists had been randomly selected from the AmericanMedical Association Physician Masterfile after eliminating allphysicians older than 65 years, those within 5 years of med-ical school graduation, or practitioners in communities offewer than 100,000 inhabitants.

Survey methodsA 7-page, 32-question, self-administered survey instrument

was specifically designed for this study. Questions were keyedto the major points of the Agency for Health Care Policy andResearch guidelines3 and dealt with the evaluation and diag-nosis of patients with possible HF; the treatment of patientswith HF; and characteristics of the respondents and their prac-tices, including board certification, year of training comple-tion, number of patients with HF treated, and practice setting.Questions dealing with medical therapy were focused onpatients with HF caused by left ventricular systolic dysfunc-tion (ejection fraction <40%), with separate questions dealingwith patients considered to have mild to moderate or severeHF. Many of the questions have been previously published5;additional questions dealing with the evaluation of patients forcoronary artery disease and ischemia were included in the sur-veys distributed to the cardiologists and HF specialists.

The survey was distributed from March through May 1995.The initial mailing included the survey and a letter describ-ing the purpose of the study and the confidential nature ofthe responses. A reminder postcard followed 10 days later,and all physicians who had not responded were contactedafter 3 weeks by telephone to confirm their addresses andtheir willingness to complete the survey. Four weeks laterthe survey was mailed again to those who had not yetreturned the first copy.

Response ratesOf the 750 Cardiologists, 126 had died or retired or their

addresses were incorrect and they could not be located bytelephone or through the American College of Cardiologydirectory. Questionnaires were returned by 340 (54%) ofthe remaining 624, but 10 of these were excluded becausethe physician did not treat at least 1 patient with HF perweek (a predetermined criterion) and 3 were excludedbecause the form was not fully filled out. Therefore therewas a final cohort of 327 physicians for the final analysis. Ofthe 118 HF specialists, 10 were not actively practicing orcould not be located. Ninety-four (87%) of the remaining

Cardiologists HF specialists

Responders 327 (54%) 94 (87%)*Median year training completed 1980 1981Board certified 90% 97%*Median number of HF patients/week† 6-10 11-20*Primary practice setting

Office 23% 0%Community hospital 45% 5%University (or affiliated) hospital 24% 85%*Government hospital 4% 6%Other hospital 5% 3%

Awareness of AHCPR guidelines 81% 94%

AHCPR, Agency for Health Care Policy and Research.*P < .05, cardiologists vs HF specialists.†Possible responses to number of patients with HF/week seen were: <1, 1-5, 6-10,11-20, >20.

Table I. Characteristics of responding physicians


Primary modality to diagnose HFRadiograph 47% 12%*Echocardiogram 48% 73%*Nuclear LV function test 1% 8%Other 4% 7%

LV function assessed 92% 97%*Use of diagnostic tests

Echocardiogram 83% 70%Chest radiograph 82% 76%ECG 91% 90%Nuclear EF 10% 28%*Right heart catheterization 7% 22%*Left heart catheterization 20% 36%*Exercise capacity measurement 17% 37%*Endomyocardial biopsy 2% 8%

LV, Left ventricular; EF, ejection fraction; ECG, electrocardiogram.*P < .05, cardiologists vs HF specialists.

Table II. Diagnostic evaluation of new patients with HF

American Heart JournalJuly 1999Bello et al102

108 returned a completely filled out questionnaire, all ofwhom indicated that they were treating at least 1 HF patientper week.

Data analysisAll responses were entered into a database with a code num-

ber for individual physicians. The database was checked foraccuracy by 1 of the investigators. Differences between thephysician groups were analyzed by chi-square testing for cate-goric variables and the Kruskal-Wallis test for continuous vari-ables with a nonnormal distribution. To determine whether thephysician group was an independent predictor of responses,other possible covariates, such as the year of training comple-tion, board certification, number of patients with HF seen, andawareness of published guidelines, were entered into multivari-ate models. Multiple logistic regression was used for binary vari-ables, and a general linear model was used for continuous vari-ables. Analyses were performed with SAS version 6 statisticalsoftware (SAS Institute).

ResultsPhysician characteristics

The characteristics of the physicians are shown inTable I. The median year of highest training completionand the proportion of physicians certified in their areaof practice was similar for both groups. HF Specialiststended to see more patients with this diagnosis than car-diologists. Although the great majority of HF specialists’and cardiologists’ practices were hospital based, mostHF specialists practiced in university/government hospi-tals and most cardiologists practiced in community hos-pitals. The majority of HF specialists and cardiologists(93% and 81%, respectively) indicated awareness of theguidelines dealing with the management of HF.

Diagnosis and evaluation of patients with HFThe use of tests is shown in Table II. In patients being

evaluated for the first time for HF, cardiologists weremuch more likely to use the chest radiograph than HFspecialists were, whereas the HF specialists were morelikely to use an echocardiogram. The large majority ofboth groups measured left ventricular function as part

of the evaluation. The use of other testing proceduresin patients with HF symptoms in the absence of anginais also shown. HF specialists reported using severaltests more frequently than other cardiologists, includ-ing nuclear medicine quantification of ejection fraction,right heart catheterization, measurements of exercisecapacity, cardiac catheterization and coronary angiogra-phy, and endomyocardial biopsy. It is uncertainwhether this pattern reflects a higher proportion ofpatients being evaluated for cardiac transplantation.

A majority of both HF specialists and cardiologistsindicated that they believed that coronary revasculariza-tion in selected patients could improve HF symptomseven in patients without angina (77% and 75%, respec-tively) and the prognosis of patients with HF (72% and62%, respectively). Therefore both groups were quiteaggressive in evaluating patients for ischemia if theirgeneral state of health made this appropriate. Table IIIshows the reported proportions of patients with HF andcurrent angina, HF and prior myocardial infarction with-out current angina, and HF without clinical evidence ofcoronary artery disease who were evaluated and the ini-tial testing approach used. The proportion of patientsevaluated by the 2 groups was nearly identical, but theHF specialists tended to use coronary angiography asthe initial diagnostic test more frequently. A substantialproportion of the patients with clinical coronary diseasemanaged by both groups underwent revascularization,but the percent was lower among those treated by HFspecialists. Revascularization was relatively uncommonamong those without overt evidence of coronary dis-ease. Coronary bypass surgery was reported to be theprocedure in the majority of cases by both groups, butthe percentage undergoing percutaneous coronaryrevascularization was higher in patients managed by car-diologists (37% vs 21%, P < .01).

TreatmentInformation was specifically sought about both initial

therapy and long-term maintenance therapy. For thispurpose, physicians were asked separately about

Patient presentation

HF and angina HF and prior MI No clinical CAD

Cardiologists HF Specialists Cardiologists HF Specialists Cardiologists HF Specialists

Percent of patients with each presenta-tion evaluated for CAD

Initial test (%) 92 93 79 78 61 60Coronary angiography 55 65* 20 36* 13 47*Stress testing 41 31* 76 61* 80 51*

Percent revascularized 55 45* 33 25* 18 16

CAD, Coronary artery disease; MI, myocardial infarction.*P < .01, specialists vs cardiologists.

Table III. Evaluation of patients for coronary revascularization


patients with mild to moderate symptoms (defined as astable ambulatory outpatient with symptoms only onsome degree of exertion) and those with severe symp-toms (defined as patients with symptoms occurringsometimes at rest). “First line” therapy was defined as asingle drug or combination of drugs that may be initi-ated together or in planned sequence. The survey find-ings are shown in Table IV.

The majority of both physician groups reported usingACE inhibitors as part of their initial treatmentapproach, although in patients with mild to moderateHF, the HF specialists reported doing this more often.In both mild to moderate and severe HF, the HF spe-cialists were more likely to initiate a 3-drug regimenconsisting of a diuretic, an ACE inhibitor, and digoxin;indeed, this was the most commonly used treatmentapproach by this group.

During the maintenance therapy, most patients weregiven combination therapy. In patients with mild tomoderate HF, specialists indicated that they used ACEinhibitors more often than cardiologists did (91% vs80%). However, among patients with severe HF bothgroups of physicians reported using ACE inhibitors in asimilar proportion of patients. HF specialists were morelikely to use digoxin, hydralazine, and β-blockers inpatients with HF, regardless of the severity of symp-toms. Calcium channel blocker use was low.

In addition to the somewhat greater use of ACEinhibitors among the HF specialists, there were signifi-cant differences in the way they were used (Table V).The usual approach reported by 75% of HF specialistsbut only 35% of cardiologists (P < .001) was to titrate to

high doses (defined as >75 mg/day of captopril, 15mg/day of enalapril, or equivalent doses of other agents)if tolerated. Cardiologists were more likely to titrate tosymptomatic responses or to a lower dose (65% vs 25%,P < .001). This resulted in 50% of patients of HF special-ists being maintained with these high doses comparedwith 30% of cardiologists’ patients (P < .001). ReportedACE inhibitor use in patients with asymptomatic leftventricular dysfunction was also higher among HF spe-cialists (88% vs 75%, P < .01).

One apparent reason for the greater use of ACEinhibitors by HF specialists was their greater willing-ness to institute and continue these drugs in patientswith renal dysfunction, either preexisting or arisingduring ACE inhibitor therapy. Thus 87% of HF special-ists but only 52% of cardiologists reported that theywould initiate an ACE inhibitor in a patient with aserum creatinine >3.0 mg/dL (P < .001). Similarly, 64%of HF specialists but only 45% of cardiologists (P <.001) indicated that they would continue ACE inhibitortherapy if it was associated with a rise in serum creati-nine >1.0 mg/dL.

Patient follow-upBoth physician groups reported that their usual

approach to monitoring stable patients with HF wasclinical assessment by history and physical examination(46% and 44%). Cardiologists were more likely to userepeat chest radiographs (12% vs 4%, P = .026) orechocardiograms (30% vs 22%, P = .14), whereas HFspecialists were more likely to use serial measurementsof exercise capacity (3% vs 18%, P < .001).

Mild to moderate HF Severe HF

Cardiologists HF Specialists Cardiologists HF Specialists

Initial therapy (% of physicians)Diuretic alone 9 2 4 1ACE inhibitor alone 20 22 4 5Diuretic and ACE inhibitor 37 27* 19 9*Diuretic, ACE inhibitor, and digoxin 21 33* 71 81*ACE-inhibitor and digoxin 8 12 1 2Diuretic and digoxin 4 0 0 0Any regimen with ACE inhibitor 86 94* 95 97

Maintenance therapy (% of physicians)Diuretic 75 78 96 97ACE inhibitor 80 91* 88 92Digoxin 60 71* 83 92*Nitrate 25 25 40 44Hydralazine 9 9 12 20*β-Blocker 9 19* 9 19*Calcium channel blockers 7 7 7 5

*P < .01, HF specialists vs cardiologists.

Table IV. Approach to therapy of patients with congestive HF in sinus rhythm


DiscussionNeed for improved HF management

HF has emerged as one of the most prevalent chronicdiseases, particularly among older individuals.1 It isassociated with high mortality rates, considerable mor-bidity, and a great deal of resource utilization. As aresult, HF has been the target of both research initia-tives and evidence-based practice guidelines. Clinicaltrials have established the value of a number of thera-pies in the management of this condition, including theuse of ACE inhibitors to prolong survival and preventhospitalizations, hydralazine and nitrates to prolong sur-vival, digoxin to prevent worsening of HF, and morerecently, β-blockers to reduce morbidity and mortalityrates.3,4,15,16 Nonetheless, there has been no apprecia-ble improvement in prognosis of patients with HF andno reduction in HF-related hospitalizations.1,17 Thesedisappointing trends have been ascribed, in part, tosuboptimal implementation of management strategiesrecommended by guideline panels and, especially, lowrates of utilization of effective therapies such as ACEinhibitors.5-14

Role of specialty care A number of strategies have been suggested to

improve the management of patients with HF. Becausethe majority of patients with HF are managed by pri-mary care physicians, one approach is the more wide-spread use of specialty care by cardiologists. In a previ-ous report of data collected with the same surveyinstrument used in this study,5 we found that cardiolo-gists indicated that their practices were consistentlymore concordant with guideline recommendationsthan those of internists and family or general practition-ers. Specifically, they more frequently measured leftventricular function (92% vs 69% and 61%, respec-tively), used ACE inhibitors in more of their patientswith mild to moderate HF (80% vs 71% and 60%,respectively) or severe HF (89% vs 84% and 76%,respectively), and more often titrated their patients tothe higher recommended doses (35% vs 12% and 3%,respectively).

Other studies have also suggested that cardiologists

are more likely to follow recommended practices inmanaging HF and other cardiac diseases.6,18-21 Althoughwhether these variations in practice translate intoimproved outcomes in patients with HF is unclear,such differences are associated with better survival inpatients with acute myocardial infarction.22 Also ofnote is that cardiologists are more aggressive in usingdiagnostic testing and invasive procedures in bothpatients with HF and patients with myocardial infarc-tion,22-25 with resultant higher costs. Some of thisincreased resource utilization is likely responsible forthe improvement in outcomes.22,25 Clearly this continuesto be an important area for cost-effectiveness studies.

Role of the HF specialistAnother approach that has generated considerable

interest is the development of comprehensive HF pro-grams and clinics overseen by cardiologists with a spe-cial interest and extensive experience in the manage-ment of patients with HF.11-14 These often consist of amultidisciplinary program with attention to patienteducation, diet, and medication compliance; carefulpatient monitoring; and intensification of previousmedical therapy. Although there have been no prospec-tive randomized trials of this type of intervention,reported experiences from these comprehensive HFclinics suggest that they can provide improvement inpatients’ clinical status and symptoms, often to thepoint of preventing or delaying the need for cardiactransplantation, and can reduce HF hospitaliza-tions.11-14 This appears to be the case regardless ofwhether the patient has been previously managed byprimary care physicians or cardiologists.

Because these programs are multifaceted, whetherclinical improvement is the result of differences inmedical management or the other interventions or thecombination is uncertain. This study was, therefore,undertaken to identify practice differences betweencardiologists and HF specialists that may play a role inthe success of these HF clinics. Our findings indicatethat there are several significant differences in the man-agement approaches of the 2 groups.

In the evaluation of patients with HF, specialists were


% Of patients with asymptomatic systolic dysfunction treated with ACE inhibitors 75% 88%*% Who usually titrate to high doses 35% 75%†

% Of patients on high ACE inhibitor doses 30% 50%†

% Who initiate ACE inhibitors in patients with creatinine >3.0 mg/dL 52% 87%†

% Who continue ACE inhibitors despite creatinine rise >1.0 mg/dL 45% 64%†

LV, Left ventricular.*P < .01, †P < .001, HF specialists vs cardiologists.

Table V. Approach to ACE inhibitor use


more likely to report ordering echocardiograms andless likely to rely on chest radiographs to make the ini-tial diagnosis. They also reported more frequent use ofother tests, including nuclear procedures to measureejection fraction, right heart catheterization, left heartcatheterization with coronary angiography, quantitativemeasurements of exercise capacity, and endomyocar-dial biopsy. By report, HF specialists were also morelikely than cardiologists to use coronary angiographyand less likely to use stress testing as the initialapproach to evaluating patients with HF for potentialcoronary revascularization. Although some of thesetests may be used as part of an evaluation for cardiactransplantation, it is noteworthy that these differencesappear to be an extension of our previous observationsin comparing primary care physicians with cardiolo-gists.5 For instance, echocardiography was indicated tobe the initial test to confirm the diagnosis of HF by 15%of family/general practitioners, 22% of internists, 48%of cardiologists, and 73% of HF specialists.

HF specialists also appear to be more aggressive intheir treatment of patients with HF. In patients withmild to moderate symptoms, a higher proportiondescribed including an ACE inhibitor in their initial regi-men and maintaining more of their patients on ACEinhibitors. HF specialists also indicated that they useddigoxin and β-blockers more frequently than cardiolo-gists, both in patients with mild to moderate symptomsand with severe symptoms. Again, these differences inmedication use followed the trends observed when car-diologists were compared with primary care physi-cians. Thus family and general practitioners reportedthat 60% of their mild to moderate patients with HFwere taking ACE inhibitors; internists reported 71% tak-ing ACE inhibitors, and cardiologists reported 80% tak-ing ACE inhibitors, compared with 91% by HF special-ists. HF specialists were also more likely to titrate ACEinhibitors to higher target doses that have been used inclinical trials and that cause a further reduction in mor-bidity and mortality rates than the lower doses com-monly used. Similar trends toward greater use of med-

ications with greater specialization were described fordigoxin, hydralazine, and β-blockers.

A major reason for the greater use of ACE inhibitors byHF specialists appears to be their willingness to use thesedrugs in the presence of renal dysfunction and to con-tinue treatment in patients who exhibit a rising serumcreatinine level. Eighty-seven percent of HF specialistsindicated that they would initiate an ACE inhibitor in apatient with a serum creatinine level >3.0 compared with52% of other cardiologists.

Of note is that by report, HF specialists were also lesslikely to use percutaneous revascularization techniquesin these patients than were other cardiologists.

Importantly, the practices reported by the HF special-ists conformed more closely with the major recommen-dations of published HF guidelines (Table VI) thanthose of other cardiologists. And, in each case our pre-vious study had indicated that cardiologists had con-formed more closely than internists and family or gen-eral practitioners. Some of these differences may be atleast in part responsible for the success of HF clinics.

Case management: A possible alternative to specialty care?

Another approach that has been used to improve themanagement of patients with HF is case management by aspecialized nurse working under protocols that incorpo-rate guideline recommendations.26-29 This is an attractivesolution because most patients with HF are not managedby specialists. Indeed, reported experience with theseprograms, including a prospective randomized study,26

has been favorable. However, most of these initial studieshave been conducted in clinics specializing in HF man-agement. It is not at all certain that similar programswould be as effective in primary care practices or even incardiology practices. The findings of this study suggestthat HF specialists are substantially more aggressive thancardiologists in both their use of diagnostic tests and intheir medical treatment of HF. The gulf between HF spe-cialists and primary care practices would be even greaterand may not be easily bridged by nurse case managers.

A. Diagnosis and evaluation 1. All patients with suspected HF should have an assessment of left ventricular function.

B. Treatment2. All patients with HF caused by left ventricular systolic dysfunction should be treated with an ACE inhibitor unless contraindicated or not tolerated.3. ACE inhibitors should be titrated to dosages used in clinical trials.4. Patients with asymptomatic left ventricular ejection fraction <35% to 40% should be treated with an ACE inhibitor.5. Therapy with ACE inhibitor should be attempted in patients with renal dysfunction if the serum creatinine is <3.0 mg/dL.6. Digoxin should be used in patients with severe HF and should be added to the medical regimen of patients with mild to moderate HF who

remain symptomatic after optimal management with ACE inhibitors and diuretics.D. Follow-up

7. Patients with HF should be followed up primarily by clinical assessment rather than by serial tests.

AHCPR, Agency for Health Care Policy and Research.

Table VI. Major recommendations of the AHCPR HF guidelines


LimitationsSeveral limitations of this study should be noted. This

report is based on a survey, and it is known that actualpractices may differ substantially from self-reported prac-tices. The data from the cardiologists were derived froma randomly selected sample of specialists from the Ameri-can Medical Association Masterfile, whereas the HF spe-cialists were selected from the membership of the Soci-ety of Transplant Cardiologists and physicians whoparticipated in HF trials. The 54% response rate by thecardiologists was far lower than that 87% rate in the HFspecialists, raising the possibility that the 2 groups ofrespondents were not similarly representative. However,even the 54% figure is comparable or higher than inmany previous reports. If anything, one might anticipatethat the respondents are more knowledgeable than thosewho did not respond, so the observed differences maybe underestimates.

Another issue is that individual physicians may haveanswered the questions from different frames of refer-ence. As already mentioned, HF specialists are ofteninvolved in selection of patients for cardiac transplanta-tion. Some of their chosen diagnostic tests, particularlyexercise testing and endomyocardial biopsy, may havebeen done for transplant evaluation. Their patients mayalso have more severe HF but otherwise be younger andhave less comorbidity, which could explain the moreaggressive medical therapy. To some extent this prob-lem was addressed by clearly indicating the questionsdealt separately with patients who had mild to moderateor severe symptoms. It should also be recognized thatthis survey was conducted in the spring of 1995. Thetreatment of HF has changed since then, particularlyregarding the use of β-blockers, although only a minor-ity of patients are currently receiving these drugs.

The most important limitation, however, is that thisstudy did not compare patient outcomes or costs andtherefore cannot draw conclusions regarding whethermanagement by one group was more effective than theother. A prospective, randomized trial would be the bestapproach to address this question, but such studies maybe difficult to accomplish and survey results have previ-ously been concordant with outcome studies.18,22,23

ImplicationsThis study extends our previous observations drawn

from a random sample of physicians who indicated thatthe self-reported practices of cardiologists conformmore closely to guideline recommendations than thoseof primary care physicians. HF specialists appear to fol-low these recommendations even more closely. This ismost evident in their more frequent and more intensivereported use of ACE inhibitors and their use of othermedications. These findings are not surprising becausecardiology is a diverse specialty and many practitionersare subspecialized regarding the specific diseases they

treat and the procedures they perform. An interven-tional cardiologist or electrophysiologist may see fewerpatients with HF than an internist, for instance. In thatregard, these findings do provide a rationale for the HFclinic model and raise the possibility that some portionof the HF population may be more optimally managedby cardiologists with a special interest and either train-ing or experience with this condition.

References1. Massie BM, Shah NB. Evolving trends in the epidemiology of heart

failure: rationale for preventive strategies and comprehensive diseasemanagement. Am Heart J 1997;133:703-12.

2. Graves EJ. National hospital discharge survey: annual summary,1993. Vital and health statistics. Series 13: data from NationalHealth Survey. 1995;121:1-63.

3. Konstam MA, Dracup K, Baker DW, et al. Heart failure: Evaluationand care of patients with left ventricular systolic dysfunction. Publi-cation no. 94-0612. Rockville (MD): Agency for Health Care Policyand Research; 1994.

4. American College of Cardiology/American Heart AssociationCommittee on Evaluation and Management of Heart Failure.Guidelines for the evaluation and management of heart failure.J Am Coll Cardiol 1995;26:1376-98.

5. Edep ME, Shah NB, Tateo IM, Massie BM. Differences betweenprimary care physicians and cardiologists in management of con-gestive heart failure: Relation to practice guidelines. J Am Coll Car-diol 1997;30:518-26.

6. Reis S, Holubkov R, Edmundowicz D, Mc Namara, Zell K, Detre K,et al. Treatment of patients admitted to the hospital with congestiveheart failure: specialty-related disparities in practice patterns andoutcomes. J Am Coll Cardiol 1997;30:733-8.

7. Chin MH, Friedman PD, Cassel CK, Lang RM. Differences in gener-alist and specialist physician knowledge and use of angiotensin-converting enzyme inhibitors for congestive heart failure. J GenIntern Med 1997;12:523-30.

8. Stafford RS, Saglam D, Blumenthal D. National patterns of angiotensin-converting enzyme inhibitor use in congestive heart failure. ArchIntern Med 1997;157:2460-4.

9. The Large State Peer Review Organization Consortium. Heart Fail-ure treatment with angiotensin-converting enzyme inhibitors inhospitalized Medicare patients in 10 large states. Arch Intern Med1997;157:1103-8.

10. Philbin EF. Factors determining angiotensin-converting enzymeinhibitor underutilization in heart failure in a community setting.Clin Cardiol 1998;21:103-8.

11. Fonarow GC, Stevenson LW, Walden JA, Livingston NA, Steimle AE,Hamilton MA, et al. Impact of a comprehensive heart failure manage-ment program on hospital readmission and functional status of patientswith advanced heart failure. J Am Coll Cardiol 1997;30:725-32.

12. Houghton AR, Cowley AJ. Managing heart failure in a specialistclinic. J Royal Coll Physicians (London) 1997;31:276-9.

13. Smith LE, Fabbri SA, Pai R, Ferry D, Heywood JT. Symptomaticimprovement and reduced hospitalization for patients attending acardiomyopathy clinic. Clin Cardiol 1997;20:949-54.

14. Hanumanthu S, Butler J, Chomsky D, Davis S, Wilson JR. Effect of aheart failure program on hospitalization frequency and exercise tol-erance. Circulation 1997;96:2842-8.

15. Massie BM. Heart failure 1997: a time to take stock. Curr OpinCardiol 1997;12:209-17.


16. Heidenreich PA, Lee Tina T, Massie BM. Effect of beta-blockade onmortality in patients with heart failure: a meta-analysis of randomizedclinical trials. J Am Coll Cardiol 1997;30:27-34.

17. Ho KK, Anderson KM, Kannel WB, Grossman W, Levy D. Survivalafter the onset of congestive heart failure in Framingham HeartStudy subjects. Circulation 1993;88:107-15.

18. Ayanian JZ, Hauptman PJ, Guadagnoli E, Antman EM, Pashos CL,Mc Neil BJ. Knowledge and practices of generalist and specialistphysicians regarding drug therapy for acute myocardial infarction.N Engl J Med 1994;331:1136-42.

19. Borowsky SJ, Kravitz RL, Laouri M, Leake B, Partridge J, Brook RH.Effect of physician specialty on use of necessary coronary angiog-raphy. J Am Coll Cardiol 1995;26:1984-91.

20. Calenda P, Jain P, Smith LG. Utilization of echocardiography byinternists and cardiologists: a comparative study. Am J Med 1996;101:584-91.

21. Stein JH, Uretz EF, Parrillo JE, Baron JT. Cost and appropriatenessof radionuclide exercise stress testing by cardiologists and non-car-diologists. Am J Cardiol 1996;77:139-42.

22. Jollis JG, DeLong ER, Peterson ED, Muhlbaier LH, Fortin DF, Califf RM,et al. Outcome of acute myocardial infarction according to the spe-cialty of the admitting physician. N Engl J Med 1996;335:1880-7.

23. Ayanian JZ, Guadagnoli E, McNeil BJ, Cleary PD. Treatment and

outcomes of acute myocardial infarction among patients of cardiol-ogists and generalist physicians. Arch Intern Med 157:2570-6.

24. Fleg JL, Hinton PC, Lakatta EG, Marcus FI, Smith TW, Strauss HC, et al.Physician utilization of laboratory procedures to monitor outpatientswith congestive heart failure. Arch Intern Med 1989;149:393-6.

25. Frances CD, Go AS, Dauterman KW, Massie BM, Newman J,Browner W. The quality of care for patients with acute myocardialinfarction [abstract]. Circulation 1997;96(suppl I):I-505.

26. Rich MW, Beckmam V, Wittenberg C, Leven CL, Freedland KE, Car-ney RM. A multidisciplinary intervention to prevent the readmissionof elderly patients with congestive heart failure. N Engl J Med 1995;333:1190-5.

27. Kornowski R, Zeeli D, Averbuch M, Finkelstein A, Schwartz D,Moshkovitz M, et al. Intensive home-care surveillance preventshospitalization and improves morbidity rates among elderly patientswith severe congestive heart failure. Am Heart J 1995;129:762-6.

28. West AJ, Miller NH, Parker KM, Senneca D, Ghandour G, ClarkM, et al. A comprehensive management system for heart failureimproves clinical outcomes and reduces medical resource utiliza-tion. Am J Cardiol 1997;79:58-63.

29. Shah NB, Der E, Ruggerio C, Heidenreich PA, Massie BM. Preven-tion of hospitalizations for heart failure with an interactive homemonitoring program. Am Heart J 1998;135:373-8.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS

Bound volumes of American Heart Journal are available only to subscribers from the Publisher at a cost of $112.00 fordomestic, $142.31 for Canadian, and $133.00 for international subscribers for Vol. 137 (January-June) and Vol. 138(July-December), shipping charges included. Each bound volume contains subject and author indexes, and all advertisingis removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durablebuckram, with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all

orders. Contact Mosby, Inc., Subscription Services, 11830 Westline Industrial Dr., St. Louis, MO 63146-3318, USA;(800)453-4351, or (314)453-4351.

Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.

self-reported differences between cardiologists and heart failure specialists in the management of...

Documents