Phenobarbital and diazepam performed alike in alcohol withdrawal, while cannabis for insomnia stays unproven

The edition in brief

In 24,643 US inpatients with alcohol withdrawal, starting phenobarbital rather than diazepam on admission made no clear difference to length of stay (4.1 vs 4.3 days, difference -0.21, 95% CI -0.42 to 0.00) or invasive ventilation (3.1% vs 2.7%); self-directed discharge was less frequent with phenobarbital (9.5% vs 12.6%). The design was a target trial emulation using hospital billing data. A Cochrane review of nine trials (537 participants) found that medical cannabis may have little to no clinically meaningful effect on sleep quality, total sleep time or daytime functioning against placebo, with very low-certainty evidence, and may increase adverse events (OR 11.25, 95% CI 4.46 to 28.35; low certainty). A systematic review of five observational studies linked shingles to a transient rise in vascular risk, most consistently for stroke, peaking about 3 to 4 weeks after onset and larger after ophthalmic zoster. A retrospective cohort of 1,835 patients with heart failure and advanced kidney disease associated calcium channel blocker use with lower mortality (IPTW-adjusted HR 0.85, 95% CI 0.73 to 0.99), but this is observational and does not justify treatment changes. A single-site service report described a hospitalist-led rehabilitation unit that shortened referral-to-discharge time from 6.89 to 3.35 days over two years. The pearl covers thiamine in alcohol withdrawal.