Placebo may be unnecessary in some clinical trials gauging subjective outcomes

Sham and no-treatment controls had similar patient-reported outcomes.

Sham treatment may be unnecessary in all clinical trials when patient-reported outcomes weigh heavily on results and interviewers are masked to treatment, a study suggested.

Randomization of patients between sham and no-treatment control groups in clinical trials of new interventions for conditions for which no effective treatment is available would provide better data than their retrospective analysis, the study authors said.

Sham treatment showed no additional subjective outcomes compared with no-treatment controls, they said.

“The most noteworthy result of this investigation was our findings,” Barbara S. Hawkins, PhD, the corresponding author, said in an email interview with Ocular Surgery News. “Frankly, I did not expect to see so much similarity in 2-year patient-reported outcomes between the two groups of matched controls.”

Data were culled from the MARINA trial for Lucentis (ranibizumab, Genentech) for neovascular age-related macular degeneration and the Submacular Surgery Trials (SST) for subfoveal choroidal neovascularization secondary to AMD.

The findings underscored the importance of interviewers being masked to treatment in clinical trials or to information such as exposure status in case-control or cohort studies, regardless of the interview’s purpose, Dr. Hawkins said.

“When patient perceptions of function or other aspects of their health are the subject of the interview, masked administration of interviews is very important,” she said. “A masked interviewer tends to be more consistent in the way questions are asked. In the case of MARINA and SST, interviewers were trained not to prompt or suggest responses to patients in addition to being masked to treatment assignment.”

The retrospective study was published in Archives of Ophthalmology.

Matched control groups

Investigators compared 2-year changes in National Eye Institute Visual Function Questionnaire (NEI-VFQ) and Short Form Health Survey (SF-36) scores between sham and no-treatment controls from the MARINA trial and SST.

The MARINA trial included a control group that received sham injection; patients were masked to treatment. The SST included a no-treatment control group in which patients were not masked.

The NEI-VFQ and SF-36 were administered in both trials.

The analysis included 62 sham controls and 62 no-treatment controls matched according to seven criteria: NEI-VFQ scores, best corrected visual acuity in the better eye, presence or absence of CNV in the non-study eye, SF-36 physical component summary scores, SF-36 mental component summary scores, age, and gender.

Data were available for 42 matched pairs interviewed at 2 years of the 56 matched pairs with at least one follow-up interview.

Inclusion in the SST required that the subfoveal neovascular lesion contain some classic CNV. Inclusion in the MARINA trial did not require the presence of classic CNV. However, only control patients who had some classic CNV were included in the retrospective study.

Results and observations

Results showed no significant differences between the sham treatment and no-treatment control groups’ changes in overall NEI-VFQ scores or subscores at 2 years. Changes on NEI-VFQ driving and mental health subscales showed the largest differences between control groups but were not statisically significant (P = .06 and P = .08, respectively), the authors reported.

“Most of the subscale differences in 2-year changes in NEI-VFQ scores were small, consistent with the overall difference,” Dr. Hawkins said. “The largest difference, using both methods of analysis, was for the driving subscale, a very important subscale for patients. The no-treatment controls reported more difficulty with driving with both methods of analysis of 2-year outcomes.”

Limitations of the study included its small size. Further study with control patients randomized to sham vs. no-treatment is warranted, Dr. Hawkins said.

“I hope that future designers of randomized trials of ophthalmologic conditions for which there is no known effective treatment will randomize the control arm patients to sham treatment vs. no treatment to provide higher quality data for comparison of outcomes, whether clinical or patient-reported. Perhaps our findings will stimulate such research,” she said.

In addition, patient-reported outcomes and clinical outcomes should be appropriately weighted to determine meaningful relationships between the two in clinical trials, Dr. Hawkins said.

“It surely is important to know whether the patient perceives any benefit from treatments deemed effective based on clinical outcomes,” she said. “This is an area that deserves more investigation, perhaps using utility analyses or other methods.” – by Matt Hasson

Reference:

• Hawkins BS, Bressler NM, Reynolds SM. Patient-reported outcomes among sham vs. no-treatment controls from randomized trials. Arch Ophthalmol. 2011;129(2):200-205.

• Barbara S. Hawkins, PhD, can be reached at bhawkins@jhmi.edu.
• Disclosure: Dr. Hawkins has no relevant financial disclosures. Fellow author Neil M. Bressler, MD, is a principal investigator of grants to the Johns Hopkins University School of Medicine that are sponsored by Genentech. Under School of Medicine policy, support for costs of research administered by the institution does not constitute a financial conflict of interest.