Systematic cross-checking between physicians can reduce medical errors in the emergency department, according to results from the CHARMED randomized trial.
In contrast to other hospital settings, ED patients are commonly managed by a sole physician, and the need to make quick decisions with incomplete information contributes to an increased risk of medical errors and subsequent adverse events.

In an earlier pilot study, Dr. Yonathan Freund, from Hopital Pitie-Salpetriere in Paris, and colleagues found that involvement of more than one physician in the care of the patient was associated with a reduced risk of adverse events and medical error in the ED.

In the current study, online April 23, 2018, in JAMA Internal Medicine, they investigated whether systematic cross-checking between emergency physicians could reduce medical errors. The crossover trial included a random sample of 14 adult patients per day during two 10-day periods, separated by a one-month washout period, in six emergency departments.

During the intervention periods, a clinical research technician sought all emergency physicians working in the ED and asked them to meet one-on-one with a peer to discuss all his or her current patients three times a day, during the day shift, with one physician presenting a brief description of each medical case and soliciting feedback and comments from the other physician (and vice versa).

There were 54 adverse events among 840 patients (6.4 percent) during the cross-check intervention compared with 90 adverse events among 840 patients (10.7percent) during the control period, a relative reduction of 40 percent (95% confidence interval, 12–59%).

Errors in sepsis management accounted for more than 40 percent of the preventable serious adverse events.

Most of the reductions occurred in adverse events defined as near misses, where there was a reduction of 47 percent.

The rate of patients with a serious adverse event in the seven days after ED visit did not differ between the cross-checking and control periods.

The median duration of the cross-checking sessions was nine minutes, during which a median number of seven patients were cross-checked.

In a per-protocol analysis, there was no significant reduction in the rate of adverse events for patients that were actually crossed-checked than for patients who were not, but cross-check patients were sicker than per-protocol control patients, which may have been a confounding factor.

Dr. Laura E. Walker, from Mayo Clinic’s department of emergency medicine and health sciences research in Rochester, Minnesota, told Reuters Health by email, “There may be benefit from running through your patients with another provider—especially those with difficult diagnoses to pinpoint, such as sepsis.”