There is an interesting public spat taking place over clinical guidelines for depression screening. A special report recently published by the BMJ titled, “Is the United States Preventive Services Task Force still a voice of caution?,” questions whether the Task Force’s recommendation to screen all children aged 12 years and older and all adults for depression
is too aggressive and could lead to overtesting and overtreatment—never a good thing. The USPSTF recommendation contrasts with those of Britain and Canada which recommend against routine screening.
The Journal’s Associate Editor Jeanne Lenzer writes in the report that some task force recommendations rely on questionable research methodologies. According to a BMJ press release on the report, Brett Thombs, professor of psychiatry at McGill University said, “In the absence of any trial evidence that screening would benefit patients, there is real concern that these recommendations may lead to more harm than good.”
Kirsten Bibbins-Domingo, current USPSTF chair and professor at the University of California, San Francisco, countered,“We evaluate the available evidence around preventive services by assessing a variety of valid trial designs and rigorously examining studies for potential bias.”
Here is the BMJ criticism of review methodology:
- Unpublished data. Lenzer notes that USPSTF recommendations have been based on reviews of evidence that have not always included unpublished data. Even though many systematic reviews besides USPSTF also fail to include unpublished data from regulators and manufacturers, Lenzer points to research by Erick Turner, a psychiatrist and former FDA reviewer, highlighting the dangers of omitting unpublished data. Albert Siu, immediate past chair of the task force, defended its reliance on published data, saying peer review “can address many sources of potential bias and methodological limitations.” Others, it is noted, reject this logic.
- Potential conflicts of interest. Lenzer points to USPSTF’s outsourcing of evidence reviews to evidence-based practice centers (EPCs), which she says “raises questions of whether financial conflicts could affect task force recommendations.” According to the press release, “although The BMJ found that both USPSTF members and the individual EPC researchers selected to work on reviews were almost entirely free of financial conflicts, several EPCs receive industry funding, raising questions of potential institutional conflicts of interest.“
- The need to acknowledge uncertainty. Some experts contacted by the BMJ said that “before making population-based screening and prevention recommendations, independent researchers should analyze more forms of raw data such as clinical study reports and patient-level data.“”When such rigorous analyses are not possible,” they continued, “it is important to acknowledge the resulting uncertainty.“
Finally, according to the press report, “some experts contacted by The BMJ said that the USPSTF’s advice for depression screening will lead to inappropriate treatment.” Allen Frances, Professor Emeritus at Duke University, former Chair of its Department of Psychiatry, and a well-known critic of overdiagnosis, was quoted as saying that “current services for severely mentally ill people were already strained to bursting point” and“we don’t need to create an army of mislabeled healthy people.”
Professor of evidence-based medicine at the University of Oxford, Carl Heneghan, concludes:
“What we need are fewer recommendations and more high-quality evidence to base decisions on. Currently, we seem to be seeing the exact opposite.”
The bottom line
When conflicts such as this arise, they make it clear what we need are rigorously designed, conflict-free, long-term studies to determine whether the screening guidelines themselves actually produce the results we hope for. When we screen aggressively as the USPSTF recommends, exactly how many people are diagnosed and treated for depression when they do not actually have it. What are the consequences of such overdiagnosis and overtreatment? When we fail to screen at all, how many people suffer severe adverse outcomes such as suicide? Could these people have been diagnosed using other types of interventions, such as those being suggested by devotees of digital health?