Lithium and the Woozle effect.

OBJECTIVES

Lithium is the oldest and most-studied treatment for bipolar depression. Many treatment guidelines place lithium as a top treatment recommendation for bipolar depression; yet some guidelines do not recommend lithium at all. These discrepancies were explored by examining the following errors: the Woozle effect (evidence by citation), reference inflation (overreporting the findings of cited studies) and belief perseverance (maintaining a belief despite new contradictory evidence) as possible causes for these discrepancies.

METHODS

Various search engines were used to find treatment guidelines for bipolar depression. The references cited in these guidelines were examined and analyzed in-depth.

RESULTS

Ten guidelines recommend lithium as a first-line treatment for bipolar depression. Five did not recommend lithium at all for the treatment of bipolar depression. These discrepancies are remarkable. The references cited in the treatment guidelines were examined and do not favor lithium as a treatment for bipolar depression. The guidelines that favored lithium for the treatment of bipolar depression suffered numerous Woozle effects, reference inflation and belief perseverance are prevalent in guidelines that recommend lithium as a first-line treatment. All three errors are principally slippery slopes. These errors do not involve a deliberate attempt to mislead and may reflect failures of the peer review process.

CONCLUSIONS

These errors may be common, as demonstrated in the case of lithium, interfering with our understanding and practice of evidence-based medicine. Both authors and journals need to guard against these types of errors in order to achieve sound evidence-based medical practices.