Differences Between Patients Enrolled Early and Late During Clinical Trial Recruitment

HomeCirculation: Cardiovascular Quality and OutcomesVol. 11, No. 7Differences Between Patients Enrolled Early and Late During Clinical Trial Recruitment Free AccessLetterPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessLetterPDF/EPUBDifferences Between Patients Enrolled Early and Late During Clinical Trial RecruitmentInsights From the HF-ACTION Trial Haider J. Warraich, MD, Christopher O'Connor, MD, Angie Wu, MS, Adrian Coles, PhD, William E. Kraus, MD, Ileana L. Piña, MD, MPH, David J. Whellan, MD, MHS and Robert J. Mentz, MD Haider J. WarraichHaider J. Warraich Haider J. Warraich, MD, Department of Medicine (Cardiology Division), Duke University Medical Center, 2301 Erwin Rd, DUMC 3485, Durham, NC 27710. E-mail E-mail Address: [email protected] Duke Clinical Research Institute, Durham, NC (H.J.W., A.W., A.C., R.J.M.). Department of Medicine (Cardiology Division), Duke University Medical Center, Durham, NC (H.J.W., W.E.K., R.J.M.). , Christopher O'ConnorChristopher O'Connor Inova Heart and Vascular Institute, Fairfax VA (C.O.). , Angie WuAngie Wu Duke Clinical Research Institute, Durham, NC (H.J.W., A.W., A.C., R.J.M.). , Adrian ColesAdrian Coles Duke Clinical Research Institute, Durham, NC (H.J.W., A.W., A.C., R.J.M.). , William E. KrausWilliam E. Kraus Department of Medicine (Cardiology Division), Duke University Medical Center, Durham, NC (H.J.W., W.E.K., R.J.M.). , Ileana L. PiñaIleana L. Piña Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, NY (I.L.P.). , David J. WhellanDavid J. Whellan Thomas Jefferson University Hospital, Sidney Kimmel Medical College, Philadelphia, PA (D.J.W.). and Robert J. MentzRobert J. Mentz Duke Clinical Research Institute, Durham, NC (H.J.W., A.W., A.C., R.J.M.). Department of Medicine (Cardiology Division), Duke University Medical Center, Durham, NC (H.J.W., W.E.K., R.J.M.). Originally published6 Jul 2018https://doi.org/10.1161/CIRCOUTCOMES.117.004643Circulation: Cardiovascular Quality and Outcomes. 2018;11:e004643Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: July 6, 2018: Previous Version of Record The duration of clinical trials has increased with time given increasing trial complexity, recruitment challenges, and lowering event rates.1 Although inclusion and exclusion criteria of clinical trials are used in part to homogenize enrolled patients, limited data are available whether patient characteristics and outcomes change over the course of a clinical trial. We hypothesized that significant changes in patient characteristics and response to intervention occurred over the course of enrollment in a clinical trial.We analyzed patients in the HF-ACTION trial (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training; URL: http://www.clinicaltrials.gov. Unique identifier: NCT00047437), which recruited from April 2003 to February 2007, randomizing outpatients with heart failure (HF) with reduced ejection fraction to exercise training versus usual care.2 There was a nonsignificant difference in the prespecified primary outcome (all-cause mortality and hospitalization) between the exercise and usual care arms in the original trial.2 In the present analysis, prespecified primary and secondary outcomes were compared between the first half of enrolled patients (n=1166, April 2003 to March 2005, 23.2 months) and the second half (n=1165, March 2005 to February 2007, 23.4 months). Adjusted models included the treatment arm and prespecified HF-ACTION adjustment variables used consistently in prior analyses.3 All data and materials have been made publicly available at the NHLBI's BioLINCC (Biologic Specimen and Data Repository Information Coordinating Center) and can be accessed at https://biolincc.nhlbi.nih.gov/studies/hf_action/. HF-ACTION was approved by the institutional review boards of all participating sites.Early enrollees were less likely to be white (60% versus 66%), from outside the United States (8% versus 15%), and had marginally lower median ejection fraction (24% versus 25%) with longer baseline cardiopulmonary exercise test times (10 minutes versus 9 minutes; P<0.01; Table 1). United States sites were activated earlier in the trial thus more blacks were enrolled early. There were no other differences in key baseline demographics, disease severity, or functional measures. Late enrollees received more evidence-based therapies such as β-blockers (96% versus 94%), mineralocorticoid receptor antagonists (48% versus 42%), implantable cardioverter defibrillators (50% versus 31%), cardiac resynchronization therapy (21% versus 15%; all P<0.05). Almost all characteristics were similar between randomized arms by time of enrollment except for higher implantable cardioverter defibrillator prevalence and increased burden of depressive symptoms in the control arm as compared with the intervention arm in patients enrolled late. Late enrollees had greater adjusted 1 to 3 month exercise adherence (effect estimate +9.81 minutes/week [95% confidence interval, 0.7–19.0]; P=0.04) and exercise duration (+40.3 minutes/week [95% confidence interval, 11.8–68.8]; P<0.01). There was no difference in the primary outcome of all-cause mortality or hospitalization between those enrolled early and late, and no treatment-by-subgroup interaction was noted for the primary outcome (Table 2). However, late enrollees experienced comparatively greater mortality benefit with exercise training than early enrollees.Table 1. Baseline Patient Characteristics by Time of Recruitment and Randomized TreatmentEarly vs Late RecruitmentEarly RecruitmentLate RecruitmentRandomized TherapyRandomized TherapyCharacteristicAll Patients (N=2331)Early Recruitment (N=1166)Late Recruitment (N=1167)P ValueTreatment Arm (N=584)Control Arm (N=582)P ValueTreatment Arm (N=575)Control Arm (N=590)P ValueDemographics Age, y59 (51–68)59 (51–68)59 (51–68)0.35459 (51–68)59 (51–68)0.54460 (52–69)59 (51–68)0.586 Male1670 (71.6%)821 (70.4%)849 (72.9%)0.187422 (72.5%)399 (68.3%)0.117436 (73.9%)413 (71.8%)0.426 Hispanic or Latino ethnicity928 (40.1%)502 (43.4%)426 (36.7%)<0.001251 (43.5%)251 (43.4%)0.959207 (35.2%)219 (38.3%)0.276 Black772 (33.1%)431 (37.0%)341 (29.3%)<0.001213 (36.6%)218 (37.3%)0.796169 (28.6%)172 (29.9%)0.634 White1468 (63.0%)694 (59.5%)774 (66.4%)<0.001346 (59.5%)348 (59.6%)0.961403 (68.3%)371 (64.5%)0.172 Other82 (3.6%)44 (3.8%)38 (3.3%)0.51526 (4.5%)18 (3.1%)0.21213 (2.2%)25 (4.4%)0.037 Country<0.0010.8990.904 United States2068 (88.7%)1077 (92.4%)991 (85.1%)537 (92.3%)540 (92.5%)504 (85.4%)487 (84.7%) Canada188 (8.1%)89 (7.6%)99 (8.5%)45 (7.7%)44 (7.5%)48 (8.1%)51 (8.9%) France75 (3.2%)0 (0.0%)75 (6.4%)0 (0.0%)0 (0.0%)38 (6.4%)37 (6.4%)Clinical characteristics BMI30 (26–35)30 (26–35)30 (26–35)0.78130 (26–35)30 (26–35)0.52830 (26–35)30 (26–35)0.705 NYHA class0.9070.2220.489 II1477 (63.4%)743 (63.7%)734 (63.0%)373 (64.1%)370 (63.4%)381 (64.6%)353 (61.4%) III831 (35.6%)411 (35.2%)420 (36.1%)206 (35.4%)205 (35.1%)203 (34.4%)217 (37.7%) IV23 (1.0%)12 (1.0%)11 (0.9%)3 (0.5%)9 (1.5%)6 (1.0%)5 (0.9%) CCS angina class0.0070.3510.374 No angina1950 (83.8%)949 (81.4%)1001 (86.1%)480 (82.5%)469 (80.3%)514 (87.3%)487 (85.0%) I200 (8.6%)117 (10.0%)83 (7.1%)59 (10.1%)58 (9.9%)36 (6.1%)47 (8.2%) II178 (7.6%)100 (8.6%)78 (6.7%)43 (7.4%)57 (9.8%)39 (6.6%)39 (6.8%) Ischemic cause of HF1197 (51.4%)586 (50.3%)611 (52.4%)0.290291 (50.0%)295 (50.5%)0.861308 (52.2%)303 (52.7%)0.866 LVEF25 (20–30)24 (20–30)25 (21–30)0.00624 (20–30)24 (20–30)0.66725 (20–31)25 (21–30)0.625 Diabetes mellitus748 (32.1%)376 (32.2%)372 (31.9%)0.870189 (32.5%)187 (32.0%)0.868181 (30.7%)191 (33.2%)0.353 Previous MI979 (42.0%)474 (40.7%)505 (43.3%)0.187235 (40.4%)239 (40.9%)0.849264 (44.7%)241 (41.9%)0.329 Hypertension1388 (59.9%)708 (61.0%)680 (58.7%)0.256345 (59.6%)363 (62.5%)0.312331 (56.5%)349 (61.0%)0.118 Atrial fibrillation or flutter488 (20.9%)227 (19.5%)261 (22.4%)0.080111 (19.1%)116 (19.9%)0.733130 (22.1%)131 (22.8%)0.771 Moderate or severe MR256 (12.0%)133 (12.4%)123 (11.6%)0.54272 (13.4%)61 (11.4%)0.33661 (11.3%)62 (11.8%)0.785 Beck Depression Inventory Score8 (4–15)8 (4–15)8 (4–15)0.4458 (4–15)8 (5–15)0.8438 (4–14)9 (5–15)0.029 Systolic blood pressure, mm Hg111 (100–126)112 (100–126)110 (100–126)0.277112 (100–128)112 (102–126)0.764110 (100–125)110 (100–126)0.482 Diastolic blood pressure, mm Hg70 (60–78)70 (60–80)70 (60–78)0.87170 (60–80)70 (60–78)0.70470 (60–78)70 (62–78)0.578 Sodium, mmol/L139 (137–141)139 (137–141)139 (137–141)0.673139 (137–141)139 (137–141)0.316139 (137–141)139 (137–141)0.784 BUN, mg/dL20 (15–28)20 (15–27)21 (15–29)0.19120 (15–28)20 (15–27)0.96321 (15–29)21 (15–29)0.719 Serum creatinine, mg/dL1.20 (1.00–1.50)1 (1.00–1.50)1 (1.00–1.50)0.8741.20 (1.00–1.50)1.20 (1.00–1.50)0.6091.20 (1.00–1.50)1.20 (1.00–1.40)0.183Baseline meds, devices ACEI-ARB1736 (74.5%)889 (76.2%)847 (72.7%)0.050440 (75.6%)449 (76.9%)0.607421 (71.4%)426 (74.1%)0.295 β-blocker2203 (94.5%)1090 (93.5%)1113 (95.5%)0.029549 (94.3%)541 (92.6%)0.242563 (95.4%)550 (95.7%)0.850 Aldosterone receptor antagonist1051 (45.1%)494 (42.4%)557 (47.8%)0.008250 (43.0%)244 (41.8%)0.685278 (47.1%)279 (48.5%)0.632 Loop diuretic1816 (77.9%)915 (78.5%)901 (77.3%)0.509467 (80.2%)448 (76.7%)0.143454 (76.9%)447 (77.7%)0.747 Digoxin1046 (44.9%)576 (49.4%)470 (40.3%)<0.001304 (52.2%)272 (46.6%)0.053243 (41.2%)227 (39.5%)0.552 Nitrate559 (24.0%)296 (25.4%)263 (22.6%)0.112150 (25.8%)146 (25.0%)0.762127 (21.5%)136 (23.7%)0.385 ICD938 (40.2%)358 (30.7%)580 (49.8%)<0.001177 (30.4%)181 (31.0%)0.830271 (45.9%)309 (53.7%)0.008 Biventricular pacemaker419 (18.0%)177 (15.2%)242 (20.8%)<0.00193 (16.0%)84 (14.4%)0.448110 (18.6%)132 (23.0%)0.070Functional measures Six-minute walk distance, m371 (299–435)372 (302–434)369 (293–435)0.519374 (304–430)369 (301–439)0.989373 (298–439)366 (282–435)0.366 Cpex time, min10 (7–12)10 (7–12)9 (7–12)0.00610 (7–13)10 (7–12)0.93210 (7–12)9 (7–12)0.229 Peak HR, min120 (104–134)120 (105–134)119 (104–134)0.400120 (104–134)120 (106–134)0.931120 (103–133)118 (104–136)0.933 Peak O2 consumption, mL/kg per minute14 (12–18)14 (11–18)15 (12–18)0.42814 (12–18)14 (11–18)0.43315 (12–18)14 (12–18)0.449 VE/Vco2, mL/kg per minute33 (28–39)32 (28–39)33 (28–38)0.91332 (28–38)32 (28–39)0.94233 (28–38)33 (28–39)0.825 KCCQ summary score68 (51–83)69 (52–83)67(50–83)0.16569 (53–84)69 (51–83)0.95768 (51–84)66 (49–82)0.134Baseline characteristics by time of enrollment and randomization arm. Values are presented as n (%), or median (interquartile range). ACEI indicates angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index; BUN, blood urea nitrogen; CCS, Canadian Cardiovascular Society; HF, heart failure; HR, heart rate; ICD, implantable cardioverter defibrillator; KCCQ, Kansas City Cardiomyopathy Questionnaire; LVEF, left ventricular ejection fraction; MI, myocardial infarction; MR, mitral regurgitation; and NYHA, New York Heart Association.Table 2. Association Between Randomized Treatment Group and Clinical Outcomes by Time of Enrollment (Control Is Reference Group)Clinical OutcomeIncidence Rate (Per 365 Patient Days)AdjustedIntervention ArmControl ArmRate Ratio (95% CI)P ValueAll-cause mortality/hospitalization* Interaction timing of treatment arm and enrollment0.579 Intervention vs control in early enrollers0.43 (0.39–0.47)0.46 (0.42–0.51)0.88 (0.76–1.02)0.088 Intervention vs control in late enrollers0.46 (0.42–0.52)0.50 (0.45–0.56)0.94 (0.79–1.11)0.461All-cause mortality* Interaction timing of treatment arm and enrollment0.012 Intervention vs control in early enrollers0.07 (0.06–0.09)0.06 (0.05–0.08)1.16 (0.90–1.48)0.253 Intervention vs control in late enrollers0.05 (0.04–0.07)0.07 (0.06–0.09)0.66 (0.46–0.95)0.023CV mortality† Interaction timing of treatment arm and enrollment0.053 Intervention vs control in early enrollers0.05 (0.04–0.06)0.05 (0.04–0.06)1.20 (0.86–1.67)0.287 Intervention vs control in late enrollers0.04 (0.03–0.05)0.06 (0.04–0.07)0.67 (0.42–1.09)0.107Clinical outcomes based on time of enrollment and randomized therapy. CI indicates confidence interval; and CV, cardiovascular.*Adjustment model includes: Weber class, Kansas City Cardiomyopathy Questionnaire symptom stability score, country, sex, mitral regurgitation grade, ventricular conduction, blood urea nitrogen, left ventricular ejection fraction, and β-blocker dose.†Adjustment model includes: exercise duration, creatinine, body mass index, sex, loop diuretic dose, left ventricular ejection fraction, Canadian Cardiovascular Society angina classification, and ventricular conduction.Our data demonstrate that patient characteristics and treatment adherence can change over the course of a large clinical outcomes trial. Utilization of evidence-based therapies for HF improved over time, suggesting that despite strict inclusion/exclusion criteria, the enrolled population reflected changes in management similar to those experienced by the general population during this time period.4 However, baseline characteristics were largely similar between patient groups by randomization arm and time of enrollment. Late enrollees were also more adherent with the exercise intervention potentially because of increased focus on adherence by investigators and overall improvement in performance of study staff with time. All of these changes, including increased adherence in the late enrolled group, may have contributed to the mortality benefit observed with exercise in late enrollees, a benefit not noted in the overall trial.2These findings have important implications for trial design and interpretation. Based on our analysis of the ClinicalTrials.gov database, phase 3 to 4 cardiovascular trials completed between 1991 and 2016 (n=1282) lasted for 3.4 years on average, enrolling 964 subjects, compared with 4.9 years for HF-ACTION. Thus, with an average duration of 3 to 4 years, temporal changes in the state-of-the-art management of conditions such as cardiovascular disease may affect trial findings despite randomization and blinding. Although our findings are hypothesis generating and we cannot rule out residual confounding factors, they suggest that time of enrollment could be included in the variables used for covariate adjustment within the outcome analysis of long duration trials during which background therapy, clinical characteristics or adherence to therapy may change. These findings are important particularly given that many trials implement strategies and protocols that may be subject to variable implementation and adherence similar to the exercise intervention in HF-ACTION.Our results point to the importance of timely enrollment and the importance of maintaining intervention adherence, both to avoid such confounding interactions and to render the overall findings more appropriately applicable to the contemporary population at the time of publication. However, given the exploratory nature of this analysis, additional analyses exploring this concept in other data sets would help support the potential implications of time of enrollment over the duration of a long-term outcome trial.DisclosuresDr Mentz has received research support from Gilead and honoraria from Thoratec, HeartWare, BMS, and Novartis. Dr O'Connor has received research support and is a consultant to Merck, ResMed, Actelion, and Biscardia. The other authors report no conflicts.Footnoteshttps://www.ahajournals.org/journal/circoutcomesHaider J. Warraich, MD, Department of Medicine (Cardiology Division), Duke University Medical Center, 2301 Erwin Rd, DUMC 3485, Durham, NC 27710. E-mail haider.[email protected]eduReferences1. Martin L, Hutchens M, Hawkins C, . Trial watch: clinical trial cycle times continue to increase despite industry efforts.Nat Rev Drug Discov. 2017; 16:157. doi: 10.1038/nrd.2017.21.CrossrefMedlineGoogle Scholar2. O'Connor CM, Whellan DJ, Lee KL, Keteyian SJ, Cooper LS, Ellis SJ, Leifer ES, Kraus WE, Kitzman DW, Blumenthal JA, Rendall DS, Miller NH, Fleg JL, Schulman KA, McKelvie RS, Zannad F, Piña IL, ; HF-ACTION Investigators. Efficacy and safety of exercise training in patients with chronic heart failure: HF-ACTION randomized controlled trial.JAMA. 2009; 301:1439–1450. doi: 10.1001/jama.2009.454.CrossrefMedlineGoogle Scholar3. O'Connor CM, Whellan DJ, Wojdyla D, Leifer E, Clare RM, Ellis SJ, Fine LJ, Fleg JL, Zannad F, Keteyian SJ, Kitzman DW, Kraus WE, Rendall D, Piña IL, Cooper LS, Fiuzat M, Lee KL, . Factors related to morbidity and mortality in patients with chronic heart failure with systolic dysfunction: the HF-ACTION predictive risk score model.Circ Heart Fail. 2012; 5:63–71. doi: 10.1161/CIRCHEARTFAILURE.111.963462.LinkGoogle Scholar4. Fonarow GC, Heywood JT, Heidenreich PA, Lopatin M, Yancy CW, ; ADHERE Scientific Advisory Committee and Investigators. Temporal trends in clinical characteristics, treatments, and outcomes for heart failure hospitalizations, 2002 to 2004: findings from Acute Decompensated Heart Failure National Registry (ADHERE).Am Heart J. 2007; 153:1021–1028. doi: 10.1016/j.ahj.2007.03.012.CrossrefMedlineGoogle Scholar Previous Back to top Next FiguresReferencesRelatedDetails July 2018Vol 11, Issue 7 Advertisement Article InformationMetrics © 2018 American Heart Association, Inc.https://doi.org/10.1161/CIRCOUTCOMES.117.004643PMID: 29980656 Originally publishedJuly 6, 2018 Keywordsheart failureprevalenceexercise testoutpatientscardiac resynchronization therapyPDF download Advertisement SubjectsClinical StudiesExerciseHeart Failure

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.