Limits...
On the Use of Local Assessments for Monitoring Centrally Reviewed Endpoints with Missing Data in Clinical Trials.

Brummel SS, Gillen DL - Open J Stat (2013)

Bottom Line: When formalized, monitoring is typically implemented using group sequential methods.This results in a missing data problem wherein a surrogate measure of response may provide useful information for interim decisions and future monitoring strategies.We illustrate the proposed methodology in the case of binary endpoints under various missingness mechanisms including missing completely at random assessments and when missingness depends on the IRC's measurement.

View Article: PubMed Central - HTML - PubMed

Affiliation: Harvard School of Public Health, Center for Biostatistics in AIDS Research, Boston, USA.

ABSTRACT

Due to ethical and logistical concerns it is common for data monitoring committees to periodically monitor accruing clinical trial data to assess the safety, and possibly efficacy, of a new experimental treatment. When formalized, monitoring is typically implemented using group sequential methods. In some cases regulatory agencies have required that primary trial analyses should be based solely on the judgment of an independent review committee (IRC). The IRC assessments can produce difficulties for trial monitoring given the time lag typically associated with receiving assessments from the IRC. This results in a missing data problem wherein a surrogate measure of response may provide useful information for interim decisions and future monitoring strategies. In this paper, we present statistical tools that are helpful for monitoring a group sequential clinical trial with missing IRC data. We illustrate the proposed methodology in the case of binary endpoints under various missingness mechanisms including missing completely at random assessments and when missingness depends on the IRC's measurement.

No MeSH data available.


Effects of shifting information time for the first three of fouranalyses on information time on ASN and maximal sample size evaluated under thealternative hypothesis ψ = −0.43. The x-axis is thel value in Π = {0.25 + l, 0.5 +l, 0.75 + l, 1}. (a) Effect on ASN; (b)Effect on maximal sample size.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4273501&req=5

Figure 2: Effects of shifting information time for the first three of fouranalyses on information time on ASN and maximal sample size evaluated under thealternative hypothesis ψ = −0.43. The x-axis is thel value in Π = {0.25 + l, 0.5 +l, 0.75 + l, 1}. (a) Effect on ASN; (b)Effect on maximal sample size.

Mentions: To illustrate the impact of changing the timing of analyses we consider ashift parameter l so that П = {0.25 + l,0.5 + l, 0.75 + l, 1}. Under the alternativehypothesis, Figure 2 depicts the maximal samplesize and the average sample number (ASN) for the symmetric O'Brien-Flemingand Pocock designs as the timing of analyses shifts away from the originally desiredequally spaced setting (l = 0). Figure 2(a) shows that the minimum ASN attained by theO’Brien-Fleming design occurs at values of l between−0.1 and 0.1, while the minimum ASN for the Pocock design occurs atapproximately l = −0.06. In addition, Figure 2(b) shows that the maximal sample size for theO’Brien-Fleming design is fairly robust to the timing of analyses. It isclear that the ASN and maximal sample size for the Pocock design is more sensitiveto shifts in the analysis timing when compared to the O’Brien-Flemingdesign. This is because the Pocock is far less conservative at early analyses whencompared to the O’Brien-Fleming design.


On the Use of Local Assessments for Monitoring Centrally Reviewed Endpoints with Missing Data in Clinical Trials.

Brummel SS, Gillen DL - Open J Stat (2013)

Effects of shifting information time for the first three of fouranalyses on information time on ASN and maximal sample size evaluated under thealternative hypothesis ψ = −0.43. The x-axis is thel value in Π = {0.25 + l, 0.5 +l, 0.75 + l, 1}. (a) Effect on ASN; (b)Effect on maximal sample size.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4273501&req=5

Figure 2: Effects of shifting information time for the first three of fouranalyses on information time on ASN and maximal sample size evaluated under thealternative hypothesis ψ = −0.43. The x-axis is thel value in Π = {0.25 + l, 0.5 +l, 0.75 + l, 1}. (a) Effect on ASN; (b)Effect on maximal sample size.
Mentions: To illustrate the impact of changing the timing of analyses we consider ashift parameter l so that П = {0.25 + l,0.5 + l, 0.75 + l, 1}. Under the alternativehypothesis, Figure 2 depicts the maximal samplesize and the average sample number (ASN) for the symmetric O'Brien-Flemingand Pocock designs as the timing of analyses shifts away from the originally desiredequally spaced setting (l = 0). Figure 2(a) shows that the minimum ASN attained by theO’Brien-Fleming design occurs at values of l between−0.1 and 0.1, while the minimum ASN for the Pocock design occurs atapproximately l = −0.06. In addition, Figure 2(b) shows that the maximal sample size for theO’Brien-Fleming design is fairly robust to the timing of analyses. It isclear that the ASN and maximal sample size for the Pocock design is more sensitiveto shifts in the analysis timing when compared to the O’Brien-Flemingdesign. This is because the Pocock is far less conservative at early analyses whencompared to the O’Brien-Fleming design.

Bottom Line: When formalized, monitoring is typically implemented using group sequential methods.This results in a missing data problem wherein a surrogate measure of response may provide useful information for interim decisions and future monitoring strategies.We illustrate the proposed methodology in the case of binary endpoints under various missingness mechanisms including missing completely at random assessments and when missingness depends on the IRC's measurement.

View Article: PubMed Central - HTML - PubMed

Affiliation: Harvard School of Public Health, Center for Biostatistics in AIDS Research, Boston, USA.

ABSTRACT

Due to ethical and logistical concerns it is common for data monitoring committees to periodically monitor accruing clinical trial data to assess the safety, and possibly efficacy, of a new experimental treatment. When formalized, monitoring is typically implemented using group sequential methods. In some cases regulatory agencies have required that primary trial analyses should be based solely on the judgment of an independent review committee (IRC). The IRC assessments can produce difficulties for trial monitoring given the time lag typically associated with receiving assessments from the IRC. This results in a missing data problem wherein a surrogate measure of response may provide useful information for interim decisions and future monitoring strategies. In this paper, we present statistical tools that are helpful for monitoring a group sequential clinical trial with missing IRC data. We illustrate the proposed methodology in the case of binary endpoints under various missingness mechanisms including missing completely at random assessments and when missingness depends on the IRC's measurement.

No MeSH data available.