Author, Date	Study Purpose(s) Specific to Trial Factors and Accrual	Type(s) of Cancer	Sample Description, Size	Focused within NCI sponsored cooperative group setting?	Phase(s) of CTs	Database	Trial-related Factors	Results Specific to Trial Factors
Bennette et al., 2016	Evaluate associations and predictors between trial-level factors and low accrual in adult cooperative group cancer CTs (clinical trials)	Multiple	787 interventional, late phase, cooperative group adult oncology CTs that started in 2000–2011	Yes	II, III	Aggregate Analysis of ClinicalTrials.gov (AACT), Drugs@FDA Database, Surveillance, Epidemiology, and End Results (SEER) Program	Number of competing trials, treatment setting, intervention modality, therapeutic, targeted therapy, new investigational agent, priority status, metastatic setting, clinical setting, sample size, randomized design, phase, placebo, number of interventions, more than one condition, blinded, number of participating sites, eligibility limited by performance status, eligibility limited by age	-Predictors of low accrual included the following: higher number of competing trials, phase III, higher enrollment percentage of eligible population, non-targeted therapy, radiation therapy, lower annual incidence of clinical condition, tissue sample required to assess eligibility, non-new investigational drug, metastatic setting, sample size, more than one condition, and common solid cancer. -Other factors associated with low accrual were multimodality, surgery, arduous eligibility criteria, randomization, and trial complexity including number of interventions, number of study locations, and more than one disease. -There were no associations between low accrual and placebo use, length of follow-up, fast track review, blinding, and eligibility limited by performance status.
Cheng et al., 2010	Investigate trial development time on accrual to oncology CTs	Multiple	419 therapeutic, non-pediatric oncology CTs activated between 2000 and 2004 and sponsored by National Cancer Institute (NCI) Cancer Therapy Evaluation Program (CTEP)	Yes	I, I/II, II, III	CTEP Protocol and Information Office database with input from Clinical Data Update System and Clinical Trials Monitoring Service	Trial development time	-CTs developed in <12 months were significantly more likely to meet accrual targets than those developed in 12–18 months. -CTs developed in >24 months were significantly less likely to meet accrual targets.
Duma et al., 2019	Identify comorbidities that adversely impact recruitment of patients with breast, colorectal, or lung cancers in early phase CTs	Breast, colorectal, lung	1103 early phase therapeutic cancer CTs from 2000 to 2015	No	I, Ib/II, II	ClinicalTrials.gov	Trial phase, target disease, anticancer therapy, line of therapy, location, sponsor, inclusion and exclusion criteria (age limits, comorbidities, organ function)	-The CTs had the following exclusion criteria: age >75 years (6%), history of prior malignancies (86%), autoimmune disease with exceptions of vitiligo and alopecia (48%), any central nervous system (CNS) metastasis (38%), symptomatic CNS metastasis (34%), human immunodeficiency virus (31%), hepatitis B or C (21%), and atrial fibrillation (20%). -Renal and hepatic eligibility criteria were prevalent such as creatinine <1.5 of the upper limit of normal (ULN) (35%).  = Compared to targeted therapy CTs, chemotherapy CTs were more likely to have exclusion criteria pertaining to CNS metastasis and history of other malignancies. -Trials sponsored by industry were more likely to have liver function exclusion criteria than those with other types of sponsors.
Gerber et al., 2014	Determine prevalence of prior cancer-related exclusion criteria and their impact on lung cancer CT accrual	Lung	51 lung cancer CTs sponsored or endorsed by the Eastern Cooperative Oncology Group (ECOG) thoracic committee	Yes	I/pilot, II, III	ECOG thoracic committee website; linked Surveillance, Epidemiology, and End Results (SEER)-Medicare database	Eligibility criteria related to prior cancer and its treatment	= 41 (80%) of ECOG -affiliated lung cancer CTs excluded prior cancer diagnosis: active cancer (16%), any prior cancer (14%), within 5 years (43%), within 2–3 years (7%)). -Estimated proportion of excluded prior lung cancer patients was up to 18% (>5% for 2/3 of CTs and>10% for approximately 1/3 of CTs). -Exclusion criteria related to prior cancer treatment were present in 20 (39%) of CTs, with 15 (29%) excluding chemotherapy or other therapy and 5 (10%) excluding both that and radiotherapy.
Gross et al., 2005	Ascertain the effect of protocol factors on enrollment of older patients in cancer CTs	Lung, breast, colorectal, prostate	36,167 patients enrolled in 33 National Cancer Institute (NCI)-sponsored cooperative group cancer CTs in 1996–2002	Yes	Unspecified	NCI Clinical Trial Evaluation Program database; NCI Physician Data Query (PDQ) clinical trial database	Cancer type, performance status, comorbidities excluded stage	-Cancer type (early stage) and performance status in exclusion criteria were significantly associated with enrollment of older persons.
Hernandez-Torres et al., 2020	Determine if exclusion criteria are associated with low accrual of older adults to cancer CTs	Multiple	69 Canadian Cancer Trials Group studies that started in 1990–2010	No	III and randomized phase II	Canadian Socioeconomic Management System database	CT start date, cancer type, and exclusion criteria	= The following CT factors were associated with lower accrual of older adults: start date prior to 2003, breast cancer indication, and exclusion criteria related to renal dysfunction.  = Central nervous system CTs were associated with higher accrual of older adults.
Khunger et al., 2018	Ascertain the frequency and factors associated with withdrawal and early termination of oncology CTs, focusing on immune checkpoint inhibitor (ICI) trials	Multiple	12,875 adult, interventional, randomized oncology trials; 350 ICI trials (2011–2015)	No	I, I/II, II, II/III, III	ClinicalTrials.gov	Type of cancer, type of treatment, sponsor, phase, accrual goal	-Low accrual was the most common reason for early termination for all trials. −5% of CTs were early terminated, and 3.5% were withdrawn. -4% of ICI trials were early terminated, and 1.4% were withdrawn. -ICI trials were less likely to early terminate compared with all other oncology drug trials, but the results were not statistically significant. -Institution-sponsored trials were significantly more likely to early terminate compared with industry sponsored trials. -Phase II and phase III trials were significantly less likely to early terminate compared with phase I trials. -The accrual goal was higher for completed trials with a median 47 compared with terminated trials with a median 9.
Kim et al., 2015	Investigate implications of eligibility criteria in phase I to III molecular trials	Multiple	67 CTs conducted by Novartis Oncology in the United States from 2006 to 2013	No	I, II, III (only II and III in final analysis)	Use of ClinicalTrials.gov was not successful; Manual review of trials	Number and characteristics of eligibility criteria	Overall, the total number of eligibility criteria did not affect enrollment duration. However, it was significantly associated with the enrollment period's duration in trials that had at least 35 patients.
Korn et al., 2010	Examine accrual for National Cancer Institute (NCI) Cooperative Group phase III CTs between 2000 and 2007	Multiple	191 CTs activated in 2000–2007 *includes 42 pediatric CTs	Yes	III	Unspecified	Disease site, use of randomization, use of investigational new drug	An estimated 22.0% of all adult and pediatric CTs would be terminated due to inadequate accrual, with 1.7% (2991) of the total enrolled accrued patients being on these CTs. Fewer breast cancer CTs terminate due to inadequate accrual. 2 of 42 pediatric trials had poor accrual. None of the pediatric nonrandomized CTs had inadequate accrual. There was no significant difference in inadequate accrual between CTs that involved an investigational new drug and those that did not.
Lemieux et al., 2008	Identify protocol characteristics of breast cancer CTs associated with poor recruitment	Breast	688 CTs opened between 1997 and 2002 in Ontario	No	I, II (or I and II), III (or (II and III)	Questionnaires to cooperative groups and pharmaceutical companies; missing data obtained from publications (ClinicalTrials.gov and websites for cooperative groups and pharmaceutical companies were used only to verify if trials should be included if no completed questionnaire received)	Phase, randomization, control group, blinding, intervention, intervention available outside the study, sponsor, location, number of participating sites, menopausal status, metastasis, minimal age limit, maximal age limit, number of eligibility criteria, premature dosing, maximum interval between diagnosis/surgery/end of therapy and enrollment, extra baseline tests, extra follow-up tests	The following protocol factors were associated with better recruitment: no placebo vs. placebo, nonmetastatic vs. metastatic, and allowed 12 week or more interval vs. less from diagnosis, surgery, or end of previous therapy for nonmetastatic CTs.
Lyss & Lilenbaum, 2009	Ascertain accrual patterns among cooperative group non-small cell lung cancer CTs	Non-Small Cell Lung	16 randomized CTs sponsored by the main cooperative groups in North America that closed accrual between 2000 and 2005	Yes	II, III	Community Oncology and Prevention Trials Research Group; National Cancer Institute of Canada	Extent of disease, trial phase, # of modalities	-Accrual was poorer for Radiation Therapy Oncology Group trials than other cooperative groups and for multimodality trials that did not primarily include systemic treatment. -Accrual was better for trials that involved advanced disease. -CTs involving standard therapy regardless of the inclusion of a new therapy had better accrual.
Massett et al., 2016	Determine reasons for slow accrual in early phase trials sponsored by the National Cancer Institute	Multiple	135 corrective action plans from 2011 to 2013 *11 (8%) were pediatric trials and 5 (4%) were for trials for both adults and children	Yes	I, II	Corrective action plans and NCI Cancer Therapy Evaluation Program (CTEP) database	Study design/protocol, eligibility	-The main reported reasons for slow accrual for phase I CTs were safety/toxicity (48%), design/protocol issues (42%) and eligibility criteria (41%). The main reasons for phase II CTs were eligibility criteria (35%) and design/protocol issues (33%).
Nguyen et al., 2018	Compare characteristics of completed and incomplete randomized CTs in radiation oncology and identify predictors of trial failure	Multiple	134 trials that were registered from 2007 to 2010	No	I, II, III	ClinicalTrials.gov	Cooperative group involvement, sponsor, PI location, number of open institutions, international study, PI's h-index, disease site, age, sex, main comparators, number of study arms, masking, blinding, primary purpose, anticipated enrollment, final enrollment, primary outcome	-Lack of accrual (57.5%) was the main reason for trial failure -Significantly more trials failed with each consecutive time period (11.8% before 2007, 34% in 2007–2008, and 39.5% in 2009–2012). -Predictors of failure were surgical comparator, government sponsorship, safety endpoint, and studies starting after 2006 via univariate analysis. -Via multivariate analysis, predictor of failure was surgical trials, and predictor of trial success was behavioral trials.
Paul et al., 2019	Determine predictors of adequate accrual in urological and nonurological solid cancer trials	Prostate, colorectal, kidney, bladder, testicular, breast, lung	326 trials in 2000–2006	No	III and IV	ClinicalTrials.gov; International Standard Randomized Controlled Trial Number Registry (United Kingdom based); online databases such as PubMed and Google Scholar	Age group, nonrandomized vs randomized, funding source, sex, intervention model, therapeutic vs nontherapeutic, masking vs open label, primary purpose, specialty, phase	−63% of trials reported sufficient accrual. -There was no significant difference in adequate accrual between urological and nonurological trials. -Kidney cancer trials accrued the best whereas bladder cancer trials accrued the worst. -Compared to government funded trials, industry sponsored trials were significantly more likely to attain adequate accrual. -No other factors (e.g. age group, nonrandomized vs randomized, intervention model, therapeutic vs nontherapeutic, masking vs open label, primary purpose, specialty, phase) were significantly associated with sufficient accrual.
Ruther et al., 2015	Determine accrual speed in published phase III oncology CTs across geographical locations and identify its influential factors	Multiple	546 phase III oncology therapeutic CTs published in 2006–2010 *included 4% pediatric/young adult CTs		III	OVID-Medline	Country, type of cancer, funder, arms, and result	-The fastest accruing CTs were those that had the following characteristics: multinational, breast cancer indication, industry sponsorship, and equivalency. -There were no significant differences in accrual time between placebo and non-placebo CTs and those CTs conducted in the United States versus Europe.
Stensland et al., 2014	Evaluate study factors associated with trials that fail to complete	Multiple	7776 adult interventional cancer trials	No	I/II, II, III	ClinicalTrials.gov	Number of sites, sponsor, location	-The most common reason for CTs to fail to complete was poor accrual (39%). - The following trials were more likely to not complete: - Single center versus multicenter trials - Industry-sponsored versus federally funded trials -Trials performed outside of the United States or both within and outside of the United States were more likely to complete than those conducted solely in the United States.

CT = clinical trials.