in reducing the number of these infections with evidence-based interventions, however, 23,000 such infections occurred in inpatient wards in 2009, at an extraordinary cost to the health care system and with an unacceptable risk of serious harm to patients (CDC, 2011). Such evidence-based interventions exist for many aspects of patient safety, yet few are used widely in patient care.

The Evidence Basis for Care

Another area for improvement is ensuring that clinical evidence guides patient care. For example, Americans receive only about half of the preventive, acute, and chronic care recommended by current research and evidence-based guidelines (McGlynn et al., 2003). Patients with diabetes, for instance, receive the recommended preventive care only 21 percent of the time (AHRQ, 2011b).

Estimates vary on the proportion of clinical decisions that are based on evidence, with some studies suggesting only 10-20 percent (Darst et al., 2010; IOM, 1985). The need for evidence also is reflected in clinical guidelines. A study of guidelines for the 10 most common types of cancer found that only 6 percent of the guidelines’ recommendations were based on a high level of evidence with uniform consensus (Poonacha and Go, 2011). An examination of 51 guidelines for treating lung cancer, for example, found that less than a third of the recommendations were evidence based (Harpole et al., 2003; IOM, 2009a). Another study found that fewer than half of the guidelines for treatment of infectious diseases are based on clinical trials (Lee and Vielemeyer, 2011).

Even when evidence-based guidelines are available, they are not always followed. For example, a recent analysis of implantable cardioverter-defibrillator (ICD) implants found that 22 percent were implanted in circumstances counter to the recommendations of professional society guidelines (Al-Khatib et al., 2011). While ICDs can be life-saving for many patients, they can be uncomfortable, inconvenient, and even life-threatening when implanted inappropriately.

This failure to deliver evidence-based care to patients results in suboptimal health outcomes. For example, consistently providing preventive services and interventions according to the best clinical evidence could prevent or postpone the majority of deaths from heart disease in the adult population (Kottke et al., 2009). The limited evidence supporting care delivery also contributes to widespread variations in clinical practice. For example, one study found that deliveries of normal-weight babies by caesarean section accounted for 7 percent of all births in some regions and almost 30 percent in others (Baicker et al., 2006).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement