- Email: [email protected]
95 Issues in running a sequential randomized trial in pressure sore prevention
101s
Abstracts 95 ISSUES IN RUNNING A SEQUENTIAL RANDOMIZED TRIAL IN PRESSURE SORE PREVENTION Julia Brown, Damien McElvenny, Jane Bridel-Nixon and Su Mason Yorkshire Clinical Trials and Research Unit Leeds, England
A sequential double blind triangular design was used to compare the incidence of pressure sores following major elective surgery among patients lying on a standard foam mattress with those on a dry polymer gel pad during their operation. 446 patients were recruited into the trial between 1994 and 1996. Interim analyses were carried out after 181 patients were entered into the trial and then subsequently after approximately every 100 patients recruited. The trial unexpectedly reached a stopping boundary at the first interim, however the Independent Data Monitoring Committee recommended continuation of the trial. They were concerned that there was a need for a larger definitive trial, were concerned about an apparent treatment center interaction, required a substudy be undertaken to further validate the subjective endpoint and that further sensitivity analyses of the main trial endpoint should be carried out in the second interim analysis. The trial was stopped at the third interim analysis when again a stopping boundary was crossed indicating that the gel pad was associated with significantly fewer pressure sores than the standard mattress (odds ratio 0.46, (95% CI, 0.26, 0.82), p = 0.01). The design, monitoring and analysis of this trial will be presented including the problems encountered for the local hospitals, for the trials unit and for the funding committee. This trial demonstrates that the use of a sequential design to obtain results in nursing research is feasible, if not without practical difftculties. 96 EVALUATION OF PREDICTORS AND IMPACT OF LOSSES TO FOLLOW-UP IN THE DOMAIN OF HIV INFECTION John P.A. Ioannldls, Roland Bassett, Michael D. Hughes, Paul A. Volberdllg, Henry S. Sacks and Joseph Lau National Institute of Allergy and Infectious Disease Bethesda, Maryland
We aimed to study predictors for and the impact of losses to follow-up in long-term HIVrelated trials. For the analysis, we selected ACTG 019, a large (n = l&09), long-term trial of immediate vs. deferred therapy in asymptomatic HIV-l infected patients with more than 500 CD4/mm3, because it had the longest follow-up among all HIV-related trials and thus the largest number of patients whose vital status was unknown at study end. In Cox models, young age, a history of drug abuse and the specific clinical site were the major predictors of loss to followup. Patient retention was significantly better in sites that enrolled many participants, with 25% of enrollees lost to follow-up in sites enrolling more than 100 patients and 44% in sites enrolling less than 33 patients each. There was a suggestion that the recorded primary endpoints were fewer after the second year in sites with many losses compared to sites with more complete follow-up (p=O.O2). Although CIs overlapped, sites with best follow-up suggested a 30% increase in the risk of disease progression by deferring therapy (CI, -22 - 119%), while no difference between arms was seen in sites with many losses. Lost patients had steeper declines of CD4 cell counts and a significantly larger
Abstracts 95 ISSUES IN RUNNING A SEQUENTIAL RANDOMIZED TRIAL IN PRESSURE SORE PREVENTION Julia Brown, Damien McElvenny, Jane Bridel-Nixon and Su Mason Yorkshire Clinical Trials and Research Unit Leeds, England
A sequential double blind triangular design was used to compare the incidence of pressure sores following major elective surgery among patients lying on a standard foam mattress with those on a dry polymer gel pad during their operation. 446 patients were recruited into the trial between 1994 and 1996. Interim analyses were carried out after 181 patients were entered into the trial and then subsequently after approximately every 100 patients recruited. The trial unexpectedly reached a stopping boundary at the first interim, however the Independent Data Monitoring Committee recommended continuation of the trial. They were concerned that there was a need for a larger definitive trial, were concerned about an apparent treatment center interaction, required a substudy be undertaken to further validate the subjective endpoint and that further sensitivity analyses of the main trial endpoint should be carried out in the second interim analysis. The trial was stopped at the third interim analysis when again a stopping boundary was crossed indicating that the gel pad was associated with significantly fewer pressure sores than the standard mattress (odds ratio 0.46, (95% CI, 0.26, 0.82), p = 0.01). The design, monitoring and analysis of this trial will be presented including the problems encountered for the local hospitals, for the trials unit and for the funding committee. This trial demonstrates that the use of a sequential design to obtain results in nursing research is feasible, if not without practical difftculties. 96 EVALUATION OF PREDICTORS AND IMPACT OF LOSSES TO FOLLOW-UP IN THE DOMAIN OF HIV INFECTION John P.A. Ioannldls, Roland Bassett, Michael D. Hughes, Paul A. Volberdllg, Henry S. Sacks and Joseph Lau National Institute of Allergy and Infectious Disease Bethesda, Maryland
We aimed to study predictors for and the impact of losses to follow-up in long-term HIVrelated trials. For the analysis, we selected ACTG 019, a large (n = l&09), long-term trial of immediate vs. deferred therapy in asymptomatic HIV-l infected patients with more than 500 CD4/mm3, because it had the longest follow-up among all HIV-related trials and thus the largest number of patients whose vital status was unknown at study end. In Cox models, young age, a history of drug abuse and the specific clinical site were the major predictors of loss to followup. Patient retention was significantly better in sites that enrolled many participants, with 25% of enrollees lost to follow-up in sites enrolling more than 100 patients and 44% in sites enrolling less than 33 patients each. There was a suggestion that the recorded primary endpoints were fewer after the second year in sites with many losses compared to sites with more complete follow-up (p=O.O2). Although CIs overlapped, sites with best follow-up suggested a 30% increase in the risk of disease progression by deferring therapy (CI, -22 - 119%), while no difference between arms was seen in sites with many losses. Lost patients had steeper declines of CD4 cell counts and a significantly larger