Patient Safety Strategies Often Lack Evidence Base

The overall benefit of this recommendation is now unclear, said Dr. Flanders, citing a 2005 meta-analysis of 22 clinical trials (BMJ 2005;331:313–21). In those trials, the use of perioperative β-blockers appeared to increase the risk of bradycardia and hypotension.

▸ Avoiding failure to rescue. Patients who experience cardiac arrest in the hospital often have preceding signs of clinical deterioration. Rapid response teams (RRTs) or medical emergency teams (METs) have been organized to react quickly and try to prevent deaths.

Two observational studies and one randomized study found a reduced risk of death associated with RRTs. But researchers in Australia found that METs had no substantial effect on the incidence of cardiac arrest, unplanned ICU admission, or unexpected death (Lancet 2005;365:2091–7).

▸ Preventing venous thromboembolism. Half of the estimated 2 million cases of venous thromboembolism (VTE) that occur each year develop in the hospital or within 30 days of discharge. Between 10% and 25% of inpatients may develop VTE, resulting in additional costs of up to $20,000 per episode.

Several meta-analyses have shown inconsistent evidence that pharmacologic prophylaxis can reduce the risk of VTE. “No study has ever shown a mortality benefit, including the meta-analyses that combined thousands of patients,” Dr. Flanders said, and studies have shown a roughly 50% increased risk of minor bleeding.

▸ Preventing methicillin-resistant Staphylococcus aureus (MRSA) infections. Prevention strategies aimed at limiting the spread of MRSA include hand hygiene, environmental decontamination, screening to identify colonized individuals, and the use of contact barriers to isolate colonized/infected patients.

Two large studies recently evaluated MRSA surveillance and decolonization. In an observational study involving three hospitals, polymerase chain reaction (PCR) was used to screen nasal specimens from more than 150,000 patients. Patients who tested positive for MRSA were put in contact isolation and decolonized with nasal mupirocin and chlorhexidine body washes. The baseline MRSA infection rate was about 9 per 10,000 patient days. The rate fell to about 7.5 per 10,000 after surveillance was implemented in the ICU only, and to about 4 per 10,000 after implementation of hospital-wide surveillance (Ann. Intern. Med. 2008;148:409–18).

But an interventional cohort study that involved almost 22,000 surgical patients failed to show a benefit. With use of a crossover design, PCR screening for MRSA on admission plus standard infection control measures was compared with standard infection control alone. There was no difference between the two approaches in the number of infections per 1,000 patient days (JAMA 2008;299:1149–57).