Presenter: Benjamin Movsas, MD Presenter's Affiliation: Henry Ford Hospital, Detroit, MI
Missing data are a significant problem in clinical trials, particularly for QOL, which cannot be obtained retrospectively. Unlike traditional endpoints, such as survival, QOL data can never be obtained retrospectively if it is not provided by the patient at the appropriate time. This limits researchers' ability to accurately perform QOL statistical analyses and negatively impacts the clinical relevance of this effort.
RTOG 0415 was a randomized study of radiation fractionation in prostate cancer with a patient reported QOL component. Initially, using paper forms, the baseline QOL compliance was >90%, but the 6 month (mo) compliance rate was only 52%.
Typically, QOL forms are filled out on paper forms. To provide a more convenient method of filling out QOL information, RTOG began using VisionTree Software to enable patients on this pilot study to fill out their QOL forms from any computer with Internet access.
The most common reported reason for noncompliance in RTOG 0415 in reporting QOL was "institutional error" (>30%). An institutional error might be something as simple as the staff neglecting to provide patients with the QOL instruments. The VTOC software automatically gives patients the opportunity to fill out QOL forms. It also allows for real-time monitoring of the QOL completion rates, thereby providing a method for intervention via real-time reminders to both patients and research associates to reduce missing QOL data.
This electronic technology allowed patients to fill out QOL forms on-line and to receive e-mail reminders. E-mail reminders were sent to research associates to remind them that a QOL time point window is about to close so that a patient can be contacted to fill out QOL information on time, before it becomes "missing."
The QOL instruments were the Expanded Prostate Index Composite (EPIC), EuroQol (EQ5D), Hopkins Symptom Checklist (HSCL25), & Sexual Adjustment Questionnaire (SAQ).
The purpose of this prospective companion study was to test the ability of an electronic web-based technology within the cooperative group mechanism to improve QOL data collection compared to using paper forms.
The goal was to determine if the 6 month QOL completion rate could be improved from the prior 52% level via paper forms (in the 20 institutions) to 75% using an online system.
Materials and Methods
RTOG 0828 was an IRB-approved companion study of RTOG 0415. 20 selected institutions were involved in the recruitment of patients.
49 patients (Mean age was 63 years old) enrolled onto RTOG 0828 between September 2008 and December 2009. To be eligible, patients had to consent to the RTOG 0415 QOL component, be English-literate, and have an email address.
This system used VisionTree Optimal Care (VTOC), a HIPAA-secure web-based technology, encrypted for security, that allows email reminders.
For 85% power and significance level 0.025, the one-sided z test yielded a minimum accrual of 40 pts with a primary endpoint of 6 mo EPIC compliance rate.
The QOL compliance rates (for all instruments) at baseline ranged from 96-98%. Compared to the 52% paper form 6 mo completion rate, the EPIC QOL web-based completion rate at 6 mo was 90% (p<0.001) & similarly for EQ5D, HSCL25, & SAQ, the 6 mo rates were 96%, 86%, & 88%, respectively (all p<0.001).
Reasons for noncompliance were patient refusal (2-4%), patient could not be contacted (2%), institutional error (2%) or other (4%).
A survey of research associates (RAs) suggested that this system saved the research associates an average of 10 minutes per QOL form.
This prospective study demonstrated a dramatic increase in the 6 month QOL compliance rates from ~50% to ~90% (p<0.001) by switching from paper forms to a web-based technology.
To our knowledge, this is the first NCI cooperative group "proof of principle" study of a web-based technology to reduce missing QOL data.
This novel approach almost eliminated institutional error as a cause of missing data by using real-time email reminders. Institutional error declined from 30% to 2%.
RTOG plans to next test this electronic system in head & neck cancer patients across all RTOG sites.
This is a very interesting and thought-provoking prospective trial, sponsored by the RTOG. The authors reported the results of a pilot study that was remarkably effective in decreasing non-compliance in QOL reporting by patients.
As radiation modalities improve, QOL continues to become more important in the design of prospective trials. Because QOL data are time sensitive and cannot be retrospectively gathered, accurate and timely reporting is critical.
Surprisingly, with the effectively of this pilot study, the mean age of patients was 63 years old. It is hyothesized that randomized trials tends to attract healthier patients with a higher level of education. It will be interesting to see how applicable this technology will be when applied to larger, more diverse patient groups.
As more patients become computer-savvy, this technology may be used as a novel strategy for real-time QOL collection to enhance the "quality" of QOL studies.
Jan 11, 2011 - Quality-of-life measurements are frequently taken into account in trials for breast cancer, but an updated review and analysis of the literature suggests the reporting methodology has room for improvement; the research has been published online Jan. 7 in the Journal of the National Cancer Institute.