A 77-year-old man with a history of chronic obstructive pulmonary disease (COPD) was admitted with a left hip fracture to the orthopedic surgery service, which has internal medicine hospitalists comanage its patients. The surgical repair went smoothly. On postoperative day 2, the patient was doing well when seen by the comanaging hospitalist. Later that day, the patient's oxygen requirement increased and the patient noted that he was feeling somewhat more short of breath compared to his baseline. The nurse notified the orthopedic surgery resident of the change in clinical status. A chest x-ray, ordered by orthopedics, showed new bilateral basilar consolidations. The orthopedic resident did not communicate these findings to the hospitalist, nor did he start antibiotics. The orthopedic resident assumed that the hospitalist was keeping up-to-date on developments and would initiate the appropriate treatment, while the hospitalist assumed that he would be contacted with any change in clinical status.
When the hospitalist next saw the patient (postoperative day 3), the patient was even more hypoxic. A computed tomography (CT) angiogram was done, which was negative for pulmonary embolism but showed much more extensive consolidations of his bilateral lung fields. He was started on broad-spectrum antibiotics; however, the patient's respiratory status continued to decline. He was ultimately transferred to the intensive care unit (ICU), intubated, and later died of hypoxic respiratory failure and sepsis (presumably from his pneumonia). It was believed that the delay in diagnosis of pneumonia and initiation of antibiotics may have contributed to the patient's downhill course.
The orthopedic surgery resident and hospitalist in this case jointly cared for the patient under an arrangement described as "comanagement." Compared to traditional consultation, comanagement gives the nonsurgical provider (often a hospitalist) increased authority, responsibility, and accountability for the care of surgical patients.(1) Rather than simply offering advice to the referring surgeon on a specific question, comanaging physicians have a broad scope of practice that often includes identifying patients and problems to manage, writing orders, and coordinating care with other providers. Comanagement is growing in popularity, and may improve care. However, as illustrated by this case, it can also pose patient safety risks.
Comanagement has grown in parallel with the hospitalist movement. The chances that a surgical patient would be comanaged grew by more than 11% per year between 2001 and 2006, and by 2006 more than 40% of patients admitted for 1 of 15 common surgical procedures was comanaged, usually by a hospitalist.(2) Like the 77-year-old patient with a hip fracture and history of COPD in this case, comanaged patients are more likely to be admitted emergently and are typically older and have greater medical comorbidity. Comanagement is more common in orthopedic and general surgery patients. Comanagement services often include criteria for selecting patients for comanagement based on the number or severity of medical problems.(3,4) With surgery increasingly being performed on patients once deemed too old or too sick, comanagement is likely to continue to grow in popularity. Another factor driving growth is the diminished availability of surgeons to attend to patients on the wards, due to greater surgical caseloads, lifestyle considerations, and resident duty hour restrictions.
The potential benefits of medical comanagement of surgical patients include improvements in provider satisfaction, clinical outcomes, and efficiency of care. Of these three goals, improved provider satisfaction has been easiest to demonstrate. Both surgeons and nurses greatly preferred comanagement to traditional medical consultation at tertiary care hospitals.(3,4) Providers felt that comanagement improved patient care and led to improved recognition of perioperative medical needs. Despite this perception, few data support the premise that comanagement leads to better clinical outcomes. In a randomized trial comparing comanagement to traditional consultation in patients undergoing arthroplasty, minor complications (e.g., urinary tract infections) were reduced, while major complications (e.g., myocardial infarctions) and mortality were unaffected.(4) Comanagement also failed to improve mortality and 30-day readmission rates in patients undergoing neurosurgery or hip fracture repair.(3,5) Efficiency of care, as measured by direct costs and length of stay, seems more likely to improve under comanagement. Hospital costs decreased when comanagement was applied to neurosurgical patients, and length of stay decreased in comanaged patients undergoing elective and emergency arthroplasty and spine surgery.(3-7)
The popularity and possible benefits of comanagement must be weighed against its potential risks to patient safety. The primary danger arises from fragmentation of care delivery. Responsibilities once held by a single physician are now divided between several providers. Without frequent communication and a shared understanding of the division of labor between partners, care can easily be omitted, duplicated, or in conflict. In this case, the surgical resident and hospitalist had differing assumptions about their responsibilities, each incorrectly believing that their colleague would monitor the patient's status. The harm from this error could have been averted if the physicians had a plan in place to communicate with each other, but, unfortunately, no such protocol was available.
Another risk arising from shared responsibility is disengagement of the surgeon.(8) The presence of a hospitalist on the surgical wards frees up the surgeon to spend more time in the operating room or clinic. This can untether surgeons from postoperative care responsibilities, leaving hospitalists to evaluate and manage surgical problems that are beyond their experience or training. This situation is more common in non-teaching hospitals, where surgical patients are often admitted to the hospitalist rather than the surgeon.
A key strategy for improving patient safety with comanagement is the creation of a carefully considered agreement between the surgeons and medicine physicians. To minimize the harm from fragmentation of care, the comanagement agreement should delineate each provider's roles and responsibilities. Some questions to address in the agreement include:
- Who is responsible for identifying patients to be comanaged?
- What problems and complications will the medicine physician, as opposed to the surgeon, take responsibility for managing?
- Who is responsible for routine patient safety measures, such as deep venous thrombosis prophylaxis?
- What are the limitations on the comanaging physician's order-writing privileges?
- What role will each provider play in ensuring a safe discharge?
For example, at UCSF, a written comanagement agreement between the Department of Neurologic Surgery and the Division of Hospital Medicine describes each service's role. In addition to evaluating patients at the request of the neurosurgeons, the hospitalist screens all patients admitted to Neurosurgery and automatically rounds on those who have significant end-organ medical disease. The neurosurgeon has primary responsibility for managing all surgical care and neurologic complications (including increased intracranial pressure). The hospitalist has primary responsibility for managing chronic medical issues, such as diabetes, and medical complications such as postoperative atrial fibrillation or pneumonia. Because of the risk of central nervous system bleeding, anticoagulant use is a shared responsibility—hospitalists always contact neurosurgery physicians before starting antiplatelet or anticoagulant agents.
The comanagement agreement should also describe expectations for communication between providers. Identifying a reliable method of contacting each other, and setting forth an expectation that any concerns will be promptly addressed, can reduce the risk of provider disengagement. The agreement should also establish an expectation of how and when routine communication between providers will occur. The surgeon and medicine physician should determine whom nurses would call when they have specific questions or concerns, and this protocol should be explained to the hospital staff. Finally, the agreement should indicate how disagreements about patient care will be resolved.
Comanagement arrangements create an opportunity to improve care for patients through the increased availability and the complementary knowledge base of the comanaging physician. To optimize its potential, systems need to be developed to ensure that role expectations are clear and that lines of communication are always open.
- Comanagement is an increasingly common practice that gives medicine physicians, especially hospitalists, enhanced responsibilities and authority in the care of surgical patients.
- Studies find that comanagement can improve provider satisfaction and efficiency of care, but its impact on clinical outcomes is less certain.
- Potential safety risks of comanagement arise from fragmentation of care and inadequate communication between providers, as well as possibility of the surgeon's disengagement.
- These risks can be reduced by creating a comanagement agreement that describes each provider's roles and responsibilities and establishes protocols for communication.
Hugo Q. Cheng, MD Clinical Professor of Medicine
Division of Hospital Medicine
Department of Medicine
University of California, San Francisco
1. Hospitalist co-management with surgeons and specialists. Society of Hospital Medicine. [Available at]
2. Sharma G, Kuo YF, Freeman J, Zhang DD, Goodwin JS. Comanagement of hospitalized surgical patients by medicine physicians in the United States. Arch Intern Med. 2010;170:363-368. [go to PubMed]
3. Auerbach AD, Wachter RM, Cheng HQ, et al. Comanagement of surgical patients between neurosurgeons and hospitalists. Arch Intern Med. 2010;170:2004-2010. [go to PubMed]
4. Huddleston JM, Long KH, Naessens JM, et al. Medical and surgical comanagement after elective hip and knee arthroplasty: a randomized, controlled trial. Ann Intern Med. 2004;141:28-38. [go to PubMed]
5. Phy MP, Vanness DJ, Melton LJ III, et al. Effects of a hospitalist model on elderly patients with hip fracture. Arch Intern Med. 2005;165:796-801. [go to PubMed]
6. Pinzur MS, Gurza E, Kristopaitis T, et al. Hospitalist–orthopedic co-management of high-risk patients undergoing lower extremity reconstruction surgery. Orthopedics. 2009;32:495. [go to PubMed]
7. Simon TD, Eilert R, Dickinson LM, Kempe A, Benefield E, Berman S. Pediatric hospitalist comanagement of spinal fusion surgery patients. J Hosp Med. 2007;2:23-30. [go to PubMed]
8. Siegal EM. Just because you can, doesn't mean that you should: a call for the rational application of hospitalist comanagement. J Hosp Med. 2008;3:398-402. [go to PubMed]