JACC F~hnIiI~ I’ J’ JX ol gender bias in tho management ol HF; 2) Caregiver SpOCialty may lnfk~eflce gander blas in managoment of HF. I 1044 34 Are Womon Wlth Heart Failure lkeated Differently? C.A. Sueta, A. Lulla, D. Slggs, SC. Smith Jr., R.J. Simpson Jr. Medical Revlow of North Carolina. Inc.: Mcm?k, Co.: Unlversify of Norfh Cflrullns. Chapel HI/I, NC, USA Sovorat studlea havo raportod dlffarancos In ths treatment 01 women with cardlovaaoult\r dlseaaa comparad to man. We ravlowod an admlnlatratlva data sat oontalnlng chall audlts an 16,576 oulpallsnts (pta) with a haart fallura (HF) dlagnosln from 138 US cardiology and multl~speclalty practlcas to dotarmlne whothor there wore gander dlffarenaaa In the uao of ACf3s. Pte had an ICD-9 coda 429, wore ovor lha n9a of 21 years (yra) and had at loaot 2 vlslts wlthln tho provlous 12 months, the last vislt occurrlng bolwoon 7/l/95 nnd 1011199. Tho moan ago was 73 f 12 yrs and 47% were womon. Comparod to mon, womon wore oldor, had moro hypartanalon, 49% vs 39o.i) and less lachemic dlsaaae 32% va 4A%. Tho mcan LVEF dotormlnad In 10,623 pta (94%) was algnlllcantly hlghar In wamon aamparad la man, 49% t- If va 39% -L 19, p m R.I)FOl. ACE.1 lhompy was pranorlbad In 46% ol womon vs 63% 01 mon. tlowovor, of Iho pts with nn LVEF c 4O?h, 98% al women vs OQO/R of men recalvad nn ACE-I. Tho recommandod tnrgot doso of ACE.1 was proacrlbod In 3596 of woman end 97% of man. Tho pmacrlptlon at ACE-Is romnlnn auboptlmnl In pts with HF. Women opponr to havo bollal prosolved LV lunctlon nnd mom dlaatollc dyalunctlon compared to man. However, In pta with an LVEF .~ 4096, ACE-I proscnptton and doslng appaars slmllnr In mon and woman. r 1044-3, The PrognosticSlgnlllcanceof Valvular Abnormalities In PatlentaWlth Severe Lafl Ventricular Dyafunctlan J.R. Teerllnk, T.S. Newman, E. Foster. N.B. Schlllar. UCSF Medlcill Confer. Snn Fran&o, CA, USA Backgmund: Loll vontrlcular dystunctlon Is an lncroaslngty common health problem wlth poor survival rates despite recant advances In treatment. How. ever, even patients with markadly reduced function have slgnllicantly dls. parate survlvnl. We examined the prognostic value of valvular disease In an unselected group 01 patients wlth systolic dysfunction (LV ejection frac. tlon ~35%) referrod for echocardiography between 1999-1990. excludmg patients with congonltal heart disease. Methods: Severity codes for valvular nbnormalltles were linked to a more talky database derived from state and national vital statistics and Cox propor. tional hazards models were used to measure their association with survival. Resub: 01 6605 patients referred for echo during this time period, 676 patlams (moan age = 61 t 3 years; M:F 2.2.1) had LV ejeclion fraction8 -35%: In this cohort. stgnlficant unlvarista pradtctors 01 death were aortic stenosis (,y’ 45. p < 0.0001: Severa AS relative hazard 3.9. C.I. 2.168) ana mitral annular calcification (ts 19. p c 0.001; Savere MAC relative hazard 2.8, Cl. 1.1-7.6). with no significant predictive value 01 mitral regurgitation nor mitral valve prolapse severity: there ware insufficient cases of aorlic msufflclency and mitral stenosis for analysis. Unexpectedly. in a multivan- ate modal including valvular abnormalities, ejection lraction and age, mitral annular calcification remained an Independent predictor along with aortic stenosis. Conclusion: Although the impact of aortic stenosis on survival is well rec- ognized. the age-independent effect 01 mitral annular calcilication in pationts with severely reduced left ventricular function is provocative, warranting lur- thar investigation with particular attention to the relationship between MAC and cerebrovascular events. 11044_361 Do Cllnlcal Events PredictSubsequentMode of Death In Patients With Advanced Chronic Heart Failure? G. Neubarg. A. Miller. C. O’Connor, Ft. Selkin. P. Carson, A. Cropp. 0. Frid. M. Presslor. J. Wertheimer. For zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA the Pf?AlSE Investigators: Cofumbia Unrversrfy New York NY USA Background: Prevention of sudden death (SD) in heart failure (CHF) pts is a priority, but prediction of SD is difficult. Efforts to dekne SD pre Victors have examined baseline pt data but not interourrent events. PRAISE monitored nonfatal Oven18 in pts with advanced CHF over 14 months, so we evaluated their relation to cause-specific mortality. MerhO@? We compared tutal mortality (TM). SD, pump failure death (PF). other deaths & SDRM in 1) all pts; 2) pts with adverse events (AE) of wors- ening CHF; & pts with “life-threatening” events judged by an Endpoints Com- mittee (EC). iw ‘*lding 3) hemodyrlamic (HD) events (severe hypr?eriusion. pulmonary edema) 8 4) nonHD ovonts (sustained ventricular arrhythmias or myocardlal Infarction). n TM% SD”. PF% olhor% SDrrM 1) All PRAISE pta 11% 30 1s 14 6,2 04s 2) AECHF 451 60 15 27 80 0.30 3) EC.HD 71 65 5.6 37 13 a.10 4) EC.nonHD 4s 49 22 13 13 0.45 Resulfs: Nonfatal cardiac ovonts Increased marfallty rtsk In PRAISE. NonHD eventa corrlacf tha most SD rlak. but SD/TM waa simllar 10 Iha overall population. HD ovanta carrlad the hIghost TM risk, but predlcted lowar lnoldanca and pmporllan al SO. Conclusion: SD romalns prevalent In advanced CHF, but recent HD tnsta- blllly strangly pradlots PF death, so HCI event-lrue pls have greater potential aurvlval banollt lram antI-arrhylhmlc drugsldovlces, Asaumlng davlces may roduco Sll by fig%, thsy would have roducod TM by 22% In group 1 but only by 15% In group 2. or 596 In group 3. iIII-3 1044-37 Decreased Left Ventricular Ralatlve Wall --_* Thlcknoas PredlctaDecreased Event-Free Sur vival In African Amerlcans With Dilated Cardiomyopathy M,Q. Koono. M.G. St. John Sutton, 0, DoNofrio, E. Lob. Universrfy of Rnnsylvank School ol Medicine, Phtlndolphra PA. l/S4 Background: LV wall thtcknoss Increases to match LV cavity dtamoter. pm- sorvlng normal wall straas and contractila Iunction. The ratio of wall thickness to cavity dlametor, rolabve wall lhtcknaas (RWT). ia therefore maintamed con. slant In tho normal heart, Irrespoctlva of heart size. In dilated cardiomyopathy @CM). fallure to malntaln RWT has been aS8OCiated wlth poor prognosis in whites. The Impact of RWT dlffomnccs In African Americans (AA) with DCM has not bean studied. Methods: LV dimensions and RWT warn assessed on baseline echoes of all AA patients presenting to our hoart lailure service from 1991 to 1997. Muth- ple cllnlcal covarlates were assessed. obtained from a prospecllve dalabasa. Measured outcome was survival to a combined clinical endpoint. mcludmg death and cardiac transplantation. Resuffe: $4 AA pts (4OM. 14F) were identified, wllh mean follow-up a? 20.1 f 16.4 mos. RWT .’ 0.3 (250 below the population mean), NYtiAclass. and history of hypertension (HTN) were signllicant predictors of decreased evenblree survival in unlvariate analyses. Pts with low RWT had a mean event-free survival of 13.6 + 2.2 mcnths. compared to 46.2 f 4.0 months in pts with normal RWT (p i 0.001). There was no correlatton between RWT and HTN. RWT ..0.3 ramainad a signihcant predictor (p = 0.053) m Cox multwariate regressIon, while NYHA class and HTN were no longer associated. Conclusion: Inappropriately low RWT independently predicts death and/or the neod for cardiac transplantation in AA wifh DCM. \1044-381 Did Cllnical Features Dlstingulsh Low From Normai EjectionFraction in Chronic Heart Failure In the DIG Trial? E.F. Phllbin. S. Hunsberger. R. Garg. S. Kouz. E. Lader. U. Thadani. A Weeks, F. McSherry. M.A. Silver. Henry Ford Hospital Delro~t, MI. NHLBI. Bethesda. MD. USA Background: Whether cknlcal data In chronic HF distinguish pahents with a low vers$ normal L\1 FF is controversial. Methods: Hall of the patients in the DIG Trial database were randomly assigned to a Derivation Set, which was used to develop a linear regression model for EF based on medical history, physical exam and laboratory data kept on file. The remainder, a Vakdation Set, were used to assess this equation’s predictive validity. Resulfs: The best model (6 = 0.23, p . 0.00001, Mallow’s Cp = 24.5) retained these predictors: age, sex, previous MI. functional class. hyperten- sion. angina. systolic and diastolic SP, heart rate, body mass index, elevated