Increasing direct supervision of physician residents doesn't lower medical errors and can make trainees less autonomous, according to a new study published Monday.
Researchers found relatively no change in the medical error rate among residents at Boston's Massachusetts General Hospital who underwent increased supervision by attending physicians and those who were supervised the standard amount, according to the analysis published in JAMA Internal Medicine.
Standardized supervision of physician residents normally consists of attending physicians physically participating in hospital rounds with their residents for newly-admitted patients, but allow a team of residents and medical students to round on their own for established patients. Such training practices allow residents to come up with their own treatment plans for established patients.
Under increased supervision, which has become more popular in residency programs over the last 20 years, attending physicians are present throughout rounds.
Both increased supervised and standard supervised residents confer with attending physicians on treatment plans, and the supervisor reviews residents' medical decision-making.
Study lead author Dr. Kathleen Finn, an internal medicine specialist at Massachusetts General Hospital, estimated faculty participate in rounds at the majority of residency training programs. She said the results of the study provides an opportunity for programs to reevaluate whether increasing their supervision across the board leads to better patient outcomes.
Study co-author Dr. Christiana Iyasere, a hospitalist at MGH, said having the constant presence of a supervisor could impact a trainee's ability to learn how to work more independently, which could have negative results when those doctors eventually begin to practice on their own.
"Medicine training is an apprenticeship model, wherein you observe and get graduated levels of autonomy over time," Iyasere said. "But if we just constantly observe residents all the time it is very difficult to graduate them to progressive levels of independence so that at the end of residency they are independent, confident leaders that can practice without supervision."
For the study, 22 attending physicians participated in two-week periods of increased supervision for a total of 188 internal medicine residents from September 2015 and June 2016.
The overall medical error rate was 107.6 events per 1,000 patient-days in the standard-supervision group compared to 91.1 events per 1000 patient-days in the increased-supervision group. Although that represents a 15% reduction in the increased-supervision group, the researchers determined it was not statistically significant and fell well below their target of a 40% decline in errors.
The study also found no significant difference between the two study groups as far as their impact on the length of stay; patients seen by both standard and increased supervision residents had a median stay of six days. Other outcomes, including transfers to the intensive care unit as well as the number of deaths, and discharges were all similar.
Researchers also found residents in the increased supervision group talked less compared to their standard supervision peers. Residents surveyed reported feeling "less efficient", and felt less able to make independent decisions when under increased supervision. Residents receiving more standard oversight felt it was the right amount of supervision and that they received more feedback on their decision making.
"It seems consistent with our values that putting in an attending there all the time will make us better," Iyasere said. "But when we study these things, we find that the answer is much more complex or nuanced. Than we initially anticipated."
In a corresponding editorial to the study, co-author Dr. Brian Neubauer, an internal medicine specialist at Walter Reed National Military Medical Center, in Bethesda, Maryland, wrote the results of the study emphasized the importance for educators to strike the right balance between providing as much time to trainees as they need but only as "much as they need".