FOCUS - June 24, 2005

It’s an adage passed along in the medical community, especially in academic health centers, when the annual turnover of interns and residents occurs each summer. Each June and July, newly minted physicians, only a few weeks out of medical school, become interns. Termed the “July effect” or “July phenomenon,” care at teaching hospitals is, according to lore, chaotic and disorganized as the cohorts of incoming interns fill the ranks of first-year residents. But demonstrating that such a phenomenon exists has proven challenging. Recently, two Harvard Business School professors set out to do just that, with some telling results.

In a recent working paper from the National Bureau of Economic Research, Robert Huckman and Jason Barro set out to test the effects of the annual turnover on resource utilization, as measured by hospital length of stay, and quality, as measured by risk-adjusted mortality rates, using patient data from 700 hospitals each year over the 1993–2001 period. Huckman and Barro (now a consultant at Bain & Company) report significant negative effects on both measures, which initially increase with the ratio of residents to hospital beds. Hospitals with the greatest resident intensities, however, such as Brigham and Women’s Hospital and Massachusetts General Hospital, are less affected in terms of mortality than hospitals with medium resident intensities. Nevertheless, the average length of stay for the average major teaching hospital increases by two percent and the average mortality increases by four percent following the annual change of personnel. Across the country, this translates into 1,500 to 2,750 additional deaths each year.

“The July effect is more than medical knowledge, it's about a coming of age for physicians, their persona, and learning to be comfortable in their role as physicians.”

“We found some initial evidence supporting the view that hospitals with the greatest teach ing intensity experience a smaller July phenomenon because they provide higher levels of supervision to new members of their house staffs,” explained Huckman. It is also possible, he said, that hospitals with the highest teaching intensity also attract highly skilled interns who are capable of learning their new tasks rapidly, thereby minimizing the July effect. “In our discussions with physicians at teaching hospitals,” recounts Huckman, “we have heard that attendings and senior members of house staffs are sensitive to the higher need for supervision around the July changeover. In fact, many of them noted that their institutions have explicit policies against vacations by attendings during the end of June and beginning of July.”

Managing Change
Joel Katz, program director of the Internal Medicine Residency at BWH, was not surprised to see the evidence on increased resource utilization, as indicated by length of stay. When interns come to the wards, he said, they are often not as skilled at the process of discharging patients, qualifying them for Medicaid, and other administrative tasks, which may extend their patients’ stay. “In a teaching hospital, with all the inevitable change around July, my goal is not to avoid change, because it is necessary, but to manage it,” said Katz. Part of this management begins in late May, when the current interns attend special sessions to learn about their impending managerial and leadership responsibilities as residents.

Further, the most experienced and seasoned faculty act as attendings during the summer months to provide additional supervision and education to incoming interns. Katz said that the layers of supervision increase during this time, and noon conferences are dedicated to critical concepts for new physicians such as respiratory failure and neurological catastrophes. Hasan Bazari, director of the medical residency program at MGH, said that the focus of teaching changes during the year, following the interns’ growth. “The July effect is more than medical knowledge,” he said, “it’s about a coming of age for physicians, their persona, and learning to be comfortable in their role as physicians.”

At MGH, the hypervigilance of house staff during the annual turnover helps ease the transition, and there are efforts to improve communication
and coordination. Current interns and junior residents, who will assume junior and senior positions when the new interns arrive, attend a teaching retreat where they learn how to be effective leaders, and incoming interns attend workshops on how to work effectively with medical students. Night float is a job for junior residents during the first half of each year and for interns during the second half, providing interns with an opportunity to assume greater responsibility gradually.

In addition to high-level trainees and increased supervision, Katz adds integration of information technology to the factors allowing hospitals like BWH to mitigate some of the challenges of the annual cohort turnover. He explains that the use of computerized physician order entry and general IT infrastructure provide critical support to physicians in training and minimize the chance that they will commit errors.

Learning Curves
Ori Preis (HMS ’04), a medicine intern this past year at BWH, remembers bureaucratic snags during July, for example, the paperwork, called a page 1, required to transfer a patient between care facilities such as from hospital to rehab or a visiting nurses association (VNA). “I had the pleasure of spending four hours working on a page 1 after a patient had already been discharged because I did not know that VNAs require a page 1, nor did I know what a page 1 was,” he recounts. Preis said that the low number of errors he recalls is “more a testament to good supervision of the intern than to the skill of the intern.”

The July phenomenon provided Huckman and Barro with a natural experiment for testing the impact of cohort turnover, but their interest was also personal. “With relatives who are physicians, Jason and I have heard the cautionary statements, made in jest, about staying away from teaching hospitals in July,” said Huckman. “We thought it was time to put this assertion to a large-scale empirical test. At a broader level, we are both more interested in how organizational factors influence the performance of health care professionals, and we felt that this setting allowed us to make observations that are relevant to the fields of both health care policy and organizational management.”

—Erica Seiguer is a sixth-year MD–PhD
student studying economics in Harvard’s
Doctoral Program in Health Policy.

The opinions expressed in this column are not necessarily those of Harvard Medical School, its affiliated institutions, or Harvard University.

Related Website
Working paper from the National Bureau of Economic Research

top