Approach to Improving Safety
Setting of Care
A 64-year-old man was seen in the thoracic surgery clinic in June after being diagnosed with a right lower lobe lung cancer. The attending surgeon saw the patient along with his fellow, who was completing his 1-year fellowship. By that point in the year, the attending had supervised the fellow's operative and postoperative care of nearly 100 patients, and he trusted him completely. The patient was a good candidate for surgery, so the surgeon discussed the operative plan (a right lower lobe lobectomy) briefly with the fellow and had the procedure scheduled for a few weeks later.
The procedure was scheduled for the first week of July. However, by this time, the fellow who had seen the patient in clinic had graduated and left the institution. The procedure itself was uneventful and the patient was transferred to the intensive care unit (ICU) postoperatively. The initial postoperative orders were written by the new thoracic surgery fellow, who had just started his fellowship and was new to the organization. He wrote brief orders for postoperative care, assuming (as had been the case at the hospital where he did his residency) that the ICU team would write more comprehensive orders.
The patient was received in the ICU by a surgical intern, who was in her first rotation and had also graduated from medical school elsewhere. The patient's nurse noticed that there were no orders for venous thromboembolism (VTE) prophylaxis, despite the patient being at high risk for VTE. She brought this to the intern's attention. The intern assumed that VTE prophylaxis was contraindicated, since the fellow had not ordered VTE prophylaxis; she also recalled an incident during medical school where a surgery intern had been chastised for starting VTE prophylaxis inappropriately. Although the standard postoperative order set in the electronic medical record included a prompt for a VTE prophylaxis order, the intern found that she could easily skip this order and complete the rest of the order set without difficulty. Therefore, the patient was not prescribed VTE prophylaxis.
Two days later, the pharmacist on the ICU team was reviewing orders for the patient and realized that the patient was not receiving VTE prophylaxis. She brought this to the attention of the intern, who replied that she thought it was contraindicated so she had not ordered it. The pharmacist conferred with the ICU attending, who agreed that VTE prophylaxis should have been started postoperatively and made sure it was started that day. Fortunately, the patient experienced no adverse consequences as a result, but the pharmacist and ICU attending wondered what could have been done to limit the risk of such an event in the future.
by John Q. Young, MD, MPP
This patient's postoperative care occurred during the annual tsunami that disrupts teaching hospitals. Unlike most storms, this one is predictable. Teaching hospitals are among the few organizations (others being military units in combat or political administrations) that experience "cohort turnover"—the exit of a large number of workers coupled with a comparably sized entry of new workers at a single point in time. Cohort turnover is thought to lead to decreased productivity due to disruption in operations (1) and the loss of tacit knowledge held by the more experienced, departing workers.(2) Teaching hospitals experience cohort turnover among house staff when the most experienced trainees depart at the same time that a group of newbies enters. This affects more than 100,000 house staff in the United States (3) and 32,000 in Europe (4) and leads to an overnight voltage drop in organizational know-how. The consequence: the average experience of the teaching hospital's workforce abruptly declines, established teams are disrupted, and many of the remaining trainees are promoted and assume new roles in the care delivery process. This transition is popularly referred to as the "August killing season" in the United Kingdom or the "July phenomenon" or "July effect" in the United States.(5,6)
Over the past 25 years, at least 45 studies have been published on the July effect. Most studies have reported no difference in outcomes between July and the rest of the year. However, the majority of these studies are limited by significant methodological weaknesses, such as inappropriate comparators or inadequate adjustment for confounding factors (e.g., patient mix or the extent to which residents were involved in care). The most definitive study to date, a systematic review published in 2011, concluded that, among studies less subject to bias, mortality tends to increase during academic year-end changeovers.(7) Subsequent studies have been mixed. Several studies focused on surgical procedures—such as anterior cervical fusion (8), elective spine (9), and head and neck cancer (10) surgery—showed no effect while another study showed an increase in hospital-acquired complications for patients cared for in July.(11)
On the whole, it seems fair to conclude that outcomes likely worsen during the July transition for patients treated by clinical teams or services in which residents play a significant role. Such teams are more vulnerable to the deleterious effects of cohort turnover. It's important to note that the July effect entails a drop in the clinical experience of the physicians in the system as well as a drop in the systems literacy of those physicians. Though the overnight drop in medical knowledge is surely important, it's clear that the infusion of physicians unfamiliar with the local clinical microsystems is also critical. In fact, one of the most sophisticated studies found that undesirable events were committed as frequently by fifth-year trainees who were new to the hospital as by interns.(4) In this case, both the new thoracic surgery fellow and the new surgical intern in the ICU made significant errors, ones that did not seem to stem from inadequate knowledge but rather from not knowing the local practice protocol.
Future research needs to identify what specific features of a clinical care process are associated with the July effect. Possible drivers may include the type or severity of illness, the kind of treatment or procedure, role of the trainee in the care delivery process, and the team or supervision structure. Although the existing evidence base is problematic, it does suggest a number of reasonable approaches. First, we should continue efforts to design supervision frameworks on graded responsibility—in which autonomy increases with competency.(12,13) While this principle has been the bedrock of American graduate medical education for decades, we have yet to develop robust programs of assessment that utilize evidence-based instruments to generate valid judgments. Only with such judgments can we ensure that individual residents are entrusted with a level of responsibility appropriate for their skill level.(14) Furthermore, these assessments must address knowledge in context, including systems literacy, i.e., to what extent the trainee can be entrusted to perform this activity in this particular context. The ACGME requires direct (i.e., attending present) supervision until interns demonstrate competency to operate under indirect supervision. The mechanisms used by teaching hospitals to determine when an intern has met that threshold are highly variable and often not explicit. In this case, the intern was not directly supervised—interns were allowed to write orders without direct observation or review from an attending. In what context had the intern been determined ready to practice under indirect supervision? Moreover, not all interns enter with the same skill set. Before an intern is allowed to participate in clinical care, we need to develop baseline assessments that determine when an intern is ready to participate.(15) Clearly, an assessment system that legitimately judges competence would have identified that both the fellow and intern were not yet ready to practice in their roles, at least not without direct supervision.
Second, linking level of supervision with competence as determined by a valid process opens up the possibility of creating a more effective orientation of trainees to new roles and systems. Optimally, this sort of training would begin prior to assuming the new role (potentially using simulation or extended observation of clinical systems) and continue through the changeover. Finally, developing changeover systems that are informed by human factors principles (such as avoiding cognitive overload and fatigue) may have benefit. For example, hospitals may choose to reduce the initial trainee workload (e.g., through lower admission caps or panel sizes and use of physician extenders), enhance supervision, and increase use of multidisciplinary teams.(16,17) In this case, the pharmacist did detect the error, but only after 2 days had passed. Perhaps all medication orders written by a trainee new to a system should be directly evaluated by a pharmacist until the trainee is deemed competent with the system. An alternate approach would be to pursue throughout health system staggered schedule starts for trainees to reduce disruption by distributing over the time the turnover, although this would introduce significant logistical challenges.
Effective design of interventions such as those discussed above will require better information about causes and magnitudes of harms in various clinical settings. Addressing patient safety hazards during turnover is an opportunity for collaboration between residency programs, health system engineers, and medical center leaders. Until efficient mitigation strategies are developed, minimizing the impact of changeovers will likely require considerable resources in the form of increased supervision and extended orientations and assessments.(18)
- Teaching hospitals experience an annual significant disruption to their clinical teams and an overnight decline in experience. This occurs when the most experienced trainees are replaced by newbies at the beginning of the academic year.
- Taken as a whole, higher quality studies tend to report increased mortality and morbidity during this transition, popularly referred to as the "July effect."
- The drop in experience affects multiple dimensions, including clinical knowledge and local systems literacy. Both must be addressed.
- Future research needs to identify what specific features of a clinical care process are associated with the July effect. Possible drivers may include the type or severity of illness, treatment or procedure, role of the trainee in the care delivery process, team structure, supervision structure, and role clarity.
- Interventions include more robust and meaningful linkage of competence with level of supervision, orientation programs that do not permit entry into patient care until competence has been established, increased utilization of oversight from various members of the interprofessional team during the transition, and the implementation of staggered start schedules.
John Q. Young, MD, MPP
Professor and Vice Chair for Education
Director, Residency Training
Department of Psychiatry
Hofstra Northwell School of Medicine at The Zucker Hillside Hospital
Glen Oaks, NY
1. Krueger AB, Mas A. Strikes, scabs, and tread separations: labor strife and the production of defective Bridgestone/Firestone tires. J Polit Econ. 2004;112:253-289. [Available at]
2. Polanyi M. The Tacit Dimension. Garden City, NY: Doubleday; 1966.
3. Barzansky B, Etzel SI. Medical schools in the United States, 2008?2009. JAMA. 2009;302:1349-1355. [go to PubMed]
4. Haller G, Myles PS, Taff? P, Perneger TV, Wu CL. Rate of undesirable events at beginning of academic year: retrospective cohort study. BMJ. 2009;339:b3974. [go to PubMed]
5. Aylin P, Majeed FA. The killing season—fact or fiction? BMJ. 1994;309:1690. [go to PubMed]
6. Blumberg MS. It's not OK to get sick in July. JAMA. 1990;264:573. [go to PubMed]
7. Young JQ, Ranji SR, Wachter RM, Lee CM, Niehaus B, Auerbach AD. "July effect": impact of the academic year-end changeover on patient outcomes: a systematic review. Ann Intern Med. 2011;155:309-315. [go to PubMed]
8. Nandyala SV, Marquez-Lara A, Fineberg SJ, Singh K. Perioperative characteristics and outcomes of patients undergoing anterior cervical fusion in July: analysis of the "July effect." Spine (Phila Pa 1976). 2014;39:612-617. [go to PubMed]
9. Bohl DD, Fu MC, Gruskay JA, Basques BA, Golinvaux NS, Grauer JN. "July effect" in elective spine surgery: analysis of the American College of Surgeons National Surgical Quality Improvement Program database. Spine (Phila Pa 1976). 2014;39:603-611. [go to PubMed]
10. Hennessey PT, Francis HW, Gourin CG. Is there a "July effect" for head and neck cancer surgery? Laryngoscope. 2013;123:1889-1895. [go to PubMed]
11. Wen T, Attenello FJ, Wu B, Ng A, Cen SY, Mack WJ. The July effect: an analysis of never events in the nationwide inpatient sample. J Hosp Med. 2015;10:432-438. [go to PubMed]
12. Sterkenburg A, Barach P, Kalkman C, Gielen M, ten Cate O. When do supervising physicians decide to entrust residents with unsupervised tasks? Acad Med. 2010;85:1408-1417. [go to PubMed]
13. ten Cate O, Snell L, Carraccio C. Medical competence: the interplay between individual ability and the health care environment. Med Teach. 2010;32:669-675. [go to PubMed]
14. Green ML, Aagaard EM, Caverzagie KJ, et al. Charting the road to competence: developmental milestones for internal medicine residency training. J Grad Med Educ. 2009;1:5-20. [go to PubMed]
15. Cohen ER, Barsuk JH, Moazed F, et al. Making July safer: simulation-based mastery learning during intern boot camp. Acad Med. 2013;88:233-239. [go to PubMed]
16. Claridge JA, Schulman AM, Sawyer RG, Ghezel-Ayagh A, Young JS. The "July phenomenon" and the care of the severely injured patient: fact or fiction? Surgery. 2001;130:346-353. [go to PubMed]
17. Highstead RG, Johnson LC, Street JH III, Trankiem CT, Kennedy SO, Sava JA. July—as good a time as any to be injured. J Trauma. 2009;67:1087-1090. [go to PubMed]
18. Nuckols TK, Bhattacharya J, Wolman DM, Ulmer C, Escarce JJ. Cost implications of reduced work hours and workloads for resident physicians. N Engl J Med. 2009;360:2202-2215. [go to PubMed]