EVIDENCE LIBRARY
Long-Acting Benzodiazepines in a Swedish Hospitalization Study: The Combined Risk Flag Lost Significance
A case-control study found that a potentially inappropriate medication category was not independently associated with hospitalization in its full model.
A case-control study found that a potentially inappropriate medication category was not independently associated with hospitalization in its full model.
The category included several drugs
The study formed 2,860 matched risk sets among older adults. Its potentially inappropriate medication definition included long-acting benzodiazepines, tramadol, propiomazine, and anticholinergic medicines. The univariate association with hospitalization disappeared after accounting for medication burden and chronic conditions, which remained associated.
Do not turn the combined result into a drug-specific conclusion
The null finding concerns the defined category, not a precise estimate for each benzodiazepine. Different screening definitions could produce different results. Claims-based availability, residual confounding, and the observational design constrain causation. The study supports examining illness and total treatment burden alongside a prescribing flag. It does not prove that long-acting benzodiazepines never contribute to hospitalization or that every flagged prescription was clinically appropriate.
Source paper
Original analysis of the supplied journal article, identified as PMC6773336.