Article type
Year
Abstract
Background: We have previously demonstrated that the publication rate of randomized clinical trials (RCTs) sponsored by the National Cancer Institute (NCI) and performed by Cooperative Oncology Groups (COGs) is high (Soares HP et al, JAMA 2005 and ASCO proceedings 2004). However, many questions arose from our findings such as 'How long it takes to get a trial published' and 'Are negative trials associated with delayed publication?'. The latter, time-lag bias, has been investigated in a Cochrane methodology review which showed that this bias might lead to a couple of years difference between the publication of trials with positive and negative findings.
Objectives: To: a) identify the length of time taken to publish phase III oncology trials and b) assess the possible association between the length of time and the trials conclusions.
Methods: We assessed all completed phase III trials conducted by Radiation Therapy Oncology Group (RTOG), one of the NCI COGs in United States. There are 59 such trials until 2000, 56 of which were published. We calculated the time to publication using the date of trial initiation, closure to accrual and the dates of first and final publication related to primary outcomes. For each trial we identified the standard and the experimental intervention(s). We assessed the original trialists preferences between experimental and standard interventions based on their reports (using their claims of which intervention was judged 'better').
Results: The mean duration of the trials was 5.0 years (range 0.5 to 9.5 years). The average time to the final publication from the date of closure of the trial was 5.3 years (range 0 to 14.3 years). Table 1 summarizes the results. There was no statistical significant difference in the average time to publication between the trials that concluded that experimental treatments were better than standard ones or vice versa.
Conclusions: Although conducting a trial in the oncology setting is very time demanding, it does not appear that the length of time taken to publish the findings of the trials conducted by this NCI COGs was affected by the direction of those findings.
Objectives: To: a) identify the length of time taken to publish phase III oncology trials and b) assess the possible association between the length of time and the trials conclusions.
Methods: We assessed all completed phase III trials conducted by Radiation Therapy Oncology Group (RTOG), one of the NCI COGs in United States. There are 59 such trials until 2000, 56 of which were published. We calculated the time to publication using the date of trial initiation, closure to accrual and the dates of first and final publication related to primary outcomes. For each trial we identified the standard and the experimental intervention(s). We assessed the original trialists preferences between experimental and standard interventions based on their reports (using their claims of which intervention was judged 'better').
Results: The mean duration of the trials was 5.0 years (range 0.5 to 9.5 years). The average time to the final publication from the date of closure of the trial was 5.3 years (range 0 to 14.3 years). Table 1 summarizes the results. There was no statistical significant difference in the average time to publication between the trials that concluded that experimental treatments were better than standard ones or vice versa.
Conclusions: Although conducting a trial in the oncology setting is very time demanding, it does not appear that the length of time taken to publish the findings of the trials conducted by this NCI COGs was affected by the direction of those findings.