- Email: [email protected]
The Surgical Hospitalist: A New Model for Emergency Surgical Care
The Surgical Hospitalist: A New Model for Emergency Surgical Care John Maa, MD, FACS, Jonathan T Carter, MD, Jessica E Gosnell, MD, Robert Wachter, MD, Hobart W Harris, MD, MPH, FACS Quality of acute surgical care in the US is threatened by a shortage of surgeons performing emergency procedures because of rising costs of uncompensated care, liability concerns, declining reimbursement, and lifestyle considerations. In July 2005, we restructured the general surgery service at our medical center into a hospitalist model to improve patient access to surgical care. STUDY DESIGN: We hypothesized that a surgical hospitalist program could improve timeliness of care, emergency department (ED) efficiency and physician satisfaction, resident supervision, continuity of care, and revenue generation. We reviewed our program after 1 year, including patient demographics, diagnosis, and time to consult. RESULTS: Three surgical hospitalists cared for 853 patients during 1 year. Patients ranged from 17 to 100 years of age and presented with abdominal pain (66%), infection (18%), malignancy (6%), hernia (4%), and trauma (3%). Fifty-seven percent of consults originated from the ED; 8% came from other surgeons. Mean time to consult was 20 minutes. A survey of ED physicians reported shorter ED length of stay, better patient satisfaction, improved professionalism and resident supervision, and better overall quality of care. Average waiting time for patients with acute appendicitis to undergo operation was reduced from 16 ⫾ 10 hours to 8 ⫾ 4 hours (p ⬍ 0.05). Forty-two percent of consults resulted in an operative procedure, and revenue increased as the number of billable consults rose by 190%. CONCLUSIONS: The surgical hospitalist model provides a cost-effective way for general surgeons to provide timely and high-quality emergency surgical care and enhance patient and referring provider satisfaction. (J Am Coll Surg 2007;205:704–711. © 2007 by the American College of Surgeons) BACKGROUND:
Lack of access to acute surgical care in the US has reached crisis proportions and is now the topic of a major national debate.1 A 2005 survey conducted by the American College of Emergency Physicians revealed that nearly 75% of emergency department (ED) medical directors believed they had inadequate surgical coverage,2 and general surgeons were among the key specialists in short supply.3 The cause of the surgeon shortage is multifactorial and includes declining reimbursement, escalating malpractice premiums, a trend toward practice in ambulatory surgery centers, a shrinking supply of surgeons in the face of a growing patient population, an aging surgeon population, and an unwillingness to disrupt elective surgical practice.4-6 The
recent Institute of Medicine report “Hospital-Based Emergency Care: At the Breaking Point” reported that hospitals in several states across America have closed their EDs because of inadequate surgical coverage, a trend with devastating consequences for critically injured or uninsured patients.7 Even within “general surgery,” the growing popularity of specialization (often achieved through additional fellowship training) has added to the problem. As a result, many surgeons have become less comfortable dealing with general surgical conditions outside their areas of expertise, especially when these conditions present emergently.8-10 Additionally, because surgical acute care in academic medical centers was historically dependent on housestaff, the Accreditation Council for Graduate Medical Education’s 80-hour workweek, introduced in 2001, has limited even more the availability of acute surgical consultation services.11 Given the critical shortage of surgeons available and willing to take emergency call, it is clear that new models of care are needed to address growing public demand.
Competing Interests Declared: None. Received January 8, 2007; Revised March 29, 2007; Accepted May 9, 2007. From the Departments of Surgery (Maa, Carter, Gosnell, Harris) and Medicine (Wachter), University of California-San Francisco, San Francisco, CA. Correspondence address: John Maa, MD, FACS, Department of General Surgery, University of California-San Francisco, 1600 Divisadero, Rm C-322D, Box 1674, San Francisco, CA 94103. email: [email protected]
© 2007 by the American College of Surgeons Published by Elsevier Inc.
704
ISSN 1072-7515/07/$32.00 doi:10.1016/j.jamcollsurg.2007.05.008
Vol. 205, No. 5, November 2007
In 2005, we reorganized our acute general surgical service at the University of California-San Francisco (UCSF) Medical Center into a hospitalist model in an effort to address these challenges. As originally conceived, hospitalists are physicians who dedicate the majority of their time to the care of hospitalized inpatients of primary care providers, with the intent of handing the care back to the primary care physician after the patient is discharged from the hospital. This model of care was introduced on the medicine services at our hospital in the early 1990s, in response to pressures of the managed care movement, and resulted in improved quality, reduced length of stay, and improved patient safety.12 By focusing on continuity and comprehensiveness of care, medical hospitalists have made substantial progress in both quality and efficiency improvement efforts for inpatients. The model has become well accepted nationally,13 with ⬎ 15,000 hospitalists in the US today. Surveys of hospitalists show that the vast majority are general internists (74%) or general pediatricians (6%); few, if any, “surgical hospitalists” have been reported.14 Instead, discussions about the role of hospitalists on surgical services tend to center on medical co-management of surgical patients,15,16 quite possibly a salutary trend, but one that fails to address the needs of surgical inpatients for timely, highquality, and cost-effective surgical diagnosis and treatment decisions. We hypothesized that creation of a surgical hospitalist program could realize many of the same benefits achieved by medical hospitalists: improved timeliness of care, improved satisfaction of referring ED providers, increased consult volume and revenue, better resident supervision and continuity of care, and shortened waiting times, using appendectomy as an index case. In this article, we describe the surgical hospitalist model and report our initial experience.
METHODS Structure of the surgical hospitalist service
Before July 2005, emergency general surgical care at UCSF was provided by a diverse faculty of gastrointestinal, hepatobiliary, colorectal, oncologic, endocrine, breast, and bariatric surgeons spread across 2 campuses separated by 3 miles, each surgeon taking call on a 24-hour basis. This system of care was problematic for several reasons. First, daytime consultations disrupted the elective procedures and clinics of on-call surgeons. Patients in the ED or acute care ward might wait hours until the on-call surgeon was available to evaluate them. Second, the diversity of emergency surgical conditions left many surgeons uncomfortable caring for diseases outside their usual practice. Third, the 24-hour structure of the call schedule and separate campuses disrupted continuity of care, particularly for
Maa et al
The UCSF Surgical Hospitalist Program
705
those patients treated by surgeons at the remote site who would need subsequent care in the ED at the main hospital. Surgical housestaff provided the only continuity and were constrained even more by the 80-hour Accreditation Council for Graduate Medical Education workweek. Finally, there was little economic incentive for taking call; the only benefit to a surgeon or the department was revenue generated from the minority of consultations that resulted in a surgical procedure. In July 2005, we restructured our general surgery call system into a surgical hospitalist model to provide coverage at the main campus, which houses the ED and accounts for 95% of the urgent consult requests. Three full-time boardcertified general surgeons staffed the service on a rotating weekly basis, dedicating 100% of their time to ED and inpatient consults. A per-diem fee was historically paid by the hospital to the department for ED coverage, and this compensation continued unchanged to the hospitalist team. These surgeons had minimal elective procedures or clinics scheduled during their on-call service weeks. Surgical hospitalist attendings rounded daily, providing continuity of care and supervision of housestaff. A senior resident, a midlevel resident, a nurse practitioner, and medical students were assigned to the hospitalist service to gain experience in emergency surgical disease. This team also cared for the elective practice of the bariatric and endocrine surgery services, with a volume of approximately 400 inpatients per year. A goal of timely surgical consultation (ie, within 30 minutes during weekdays and 45 minutes on weeknights and weekends) was implemented as a primary measure of performance. A backup surgeon (usually one of the other hospitalist surgeons) was always available. For those patients possibly requiring complex surgical intervention (advanced hepatobiliary, foregut, thoracic, or endocrine procedures), the surgical hospitalist would consult a senior experienced surgeon for additional care after the patient had been initially stabilized and admitted. In addition to the direct clinical care, hospitalists also managed the majority of surgical transfer requests from referring hospitals throughout Northern California. Once discharged, patients were followed as outpatients in a surgical hospitalist clinic, staffed by off-call surgical hospitalists. The key elements of the surgical hospitalist model are listed in Table 1. Patient, disease, and financial characteristics
We prospectively recorded the age, gender, race, chief complaint, diagnosis, medical comorbidities, service requesting consultation, and time to consult (defined as the number of minutes from initial phone consultation to bedside evaluation) for each patient evaluated by surgical hospitalists. We reviewed the percentage of consultations resulting in an operative procedure, the type of procedure performed, and
706
Maa et al
The UCSF Surgical Hospitalist Program
Table 1. Key Elements of the Surgical Hospitalist Model On-call period lasts continuously for 1 week, not 24 hours, in order to improve continuity of care. During the on-call period, no elective clinics or procedures are scheduled that might disrupt or conflict with acute surgical care. A resident or attending should evaluate the patient within 30 minutes of consultation during business hours and within 45 minutes off hours. If the resident is unavailable, then the on-call attending will be contacted directly to see the patient independently. Patients requiring special expertise are initially assessed by the team and then reassigned (triaged) to a higher level of expert care as indicated. After the on-call period, the care of inpatients and consults is handed off to the next on-call surgeon in a group-practice model. Patient safety is enhanced though increased resident supervision and improved signouts. Revenue stems primarily from per-diem payment from the hospital, procedural fees, and attending documentation of nonoperative care.
the percentage of consultations requiring additional surgical expertise outside our practice (such as advanced esophageal, colorectal, or hepatic resection). Financial data, including year-to-year billing charges, collections, and payor mix, were provided by the Division of Finance of the Department of General Surgery. Emergency department physician satisfaction and patient outcomes
Because a primary goal of the surgical hospitalist program was to provide timely and efficient consultations to ED providers, we conducted an anonymous survey of ED providers 6 months after the start of the program. Nine questions were structured on a 5-point Likert scale (1 ⫽ improved dramatically, 5 ⫽ worsened dramatically) and included the following, “since July 1, 2005, how have the following changed: 1) the timeliness of general surgery evaluations and recommendations, 2) the professionalism of general surgeons toward ED staff, and 3) supervision of surgical residents by surgery attendings.” We also reviewed the records of all patients presenting to the ED with acute appendicitis and tracked the time from ED triage to appendectomy as another measure of our program’s performance. Patients undergoing appendectomy during the 6 months before the start of the study served as controls. Data collection and statistical analysis
All data were stored on a password-protected Health Insurance Portability and Accountability Act⫺compliant server. This study was approved by the UCSF Committee on Human Research. Student’s t-test was used to compare appendectomy wait times. Statistical significance was defined as p ⬍ 0.05.
J Am Coll Surg
RESULTS Patient and disease characteristics
Three surgical hospitalists cared for 853 patients during a 1-year period, averaging 2.3 new consults per day. Patient and disease characteristics are listed in Table 2. Patients averaged 53 years of age (range 17 to 100), and the maleto-female ratio was nearly equal (51:49). Patients were racially diverse (63% Caucasian, 17% Asian, 11% African American, and 9% Hispanic). The majority presented with acute abdominal pain (66%), and soft-tissue infections (18%), malignancy (6%), and hernia (4%) were other major sources of referrals. Because our center is a Level II trauma center, whereby most trauma patients are regionalized to San Francisco General Hospital, trauma accounted Table 2. Characteristics of 853 Patients Evaluated by Surgical Hospitalists at University of California-San Francisco from July 1, 2005, to June 30, 2006 Age (y), mean (range) 53 (17⫺100) Gender, n (%) Male 432 (51) Female 421 (49) Ethnicity, n (%) African American 97 (11) Asian/Pacific Islander 144 (17) Caucasian 540 (63) Hispanic/Latino 72 (9) Reason for consultation, n (%) Acute appendicitis 118 (14) Bowel obstruction or ileus 87 (10) Biliary tract disease 91 (10) Colorectal or anal disease 66 (8) Mesenteric ischemia 33 (4) Pancreatitis (excluding biliary) 33 (4) Other acute abdomen 138 (16) Hernia 33 (4) Superficial or deep space infection 157 (18) Trauma 22 (3) Malignancy 47 (6) Other 28 (3) Source of consultation, n (%) Emergency department 487 (57) General medicine or pediatrics 167 (20) Medical specialist, including critical care 61 (7) Obstetrics/gynecology 18 (2) Surgical specialist 67 (8) Transfer from another institution 53 (6) Wait time before surgical evaluation (min), mean (range) 20 (1⫺270) Patients evaluated, n (%) Within 30 min of consult 679 (80) Within 45 min of consult 727 (85)
Vol. 205, No. 5, November 2007
Maa et al
Table 3. Case Log of Surgical Procedures Performed Appendectomy Incision and drainage of abscess Major abdominal (intestinal resection, lysis of adhesions) Cholecystectomy Complex liver/spleen/pancreas procedures Hernia repair Soft tissue subcutaneous procedures Other Total cases
n
%
103 70
29 19
69 39
19 11
24 15 14 25 359
7 4 4 7
for only 3% of consultations. The majority of consultations originated from the ED (57%), although general medicine (20%) and medical specialists (7%), including critical care physicians, accounted for large numbers. Interestingly, other surgeons were responsible for 8% of the total consult volume; these were generally from nongeneral surgical specialists, such as neurosurgeons or urologists. Transfer patients from other institutions accounted for 6% of consults. Surgical procedures performed and triage to senior expert surgeons
Forty-two percent of consultations (n ⫽ 359) resulted in an operative procedure during the index admission (Table 3). The most common procedures performed included appendectomy (29%), incision and drainage of abscess (19%), exploratory laparotomy for intestinal resection or lysis of adhesions (19%), cholecystectomy (11%), and complex liver/spleen/pancreas procedures (7%). We involved other
The UCSF Surgical Hospitalist Program
707
Table 4. Patients Triaged and Referred to Senior General Surgery Specialist
Service
Advanced laparoscopic Bariatric Breast Colorectal Endocrine Foregut Hepatobiliary Thoracic Totals
Needed consult from a senior specialist
Needed operation by that specialist
7 1 5 8 1 7 6 2 37
7 0 2 5 1 6 6 2 29
surgeons with advanced expertise in 37 of 853 patients (4.3%). Surgical intervention was necessary in 29 of these 37 patients (Table 4), and the remaining 92% (330 of 359) of all procedures were performed by surgical hospitalists without involving other surgeons. Time to consultation and ED provider satisfaction
Time to consult, defined as the time from telephone consultation to bedside evaluation, averaged 16 minutes. Eighty percent of consults were seen within 30 minutes, and 85% were seen within 45 minutes (Table 2). The confidential survey of ED providers 6 months after our program began had a response rate of 76% (13 of 17). Results of the survey indicated that two-thirds of ED providers were aware of the surgical hospitalist program’s creation and intent (Fig. 1). All providers who participated in the survey believed that the surgical hospitalist program had
Figure 1. Survey of emergency department providers after 6 months of University of CaliforniaSan Francisco surgical hospitalist care.
708
Maa et al
The UCSF Surgical Hospitalist Program
improved timeliness of care, ED length of stay, supervision of housestaff, patient satisfaction, and professionalism of the surgical staff. Only 1 provider believed overall quality of care had decreased; the remainder believed the quality was either the same (38%) or better (46%). When we compared the wait time (from ED triage to skin incision) for patients undergoing appendectomy in a 6-month period before and after the start of our program, we found that it decreased 50%, from 16 ⫾ 10 hours to 8 ⫾ 4 hours (p ⬍ 0.05). The rate of wound infection and postoperative intraabdominal abscess were both ⬍ 4%. Revenue generation and payor mix
In the first year of the surgical hospitalist program, we saw a 190% increase in requested consults from the ED and inpatient wards. This led to a 415% increase in year-overyear billable consult revenue, with a minimal impact on the individual consult revenue for the remainder of the division of general surgery (which will be addressed in future studies). The greatest increase was observed in the area of subsequent care and followup (“subsequent hospital care,” by CPT code), where a 24-fold increase in revenue was achieved through improved documentation and billing. Fifty-one percent of patients treated were insured under capitated care plans, 44% had Medicare/Medicaid, and 4% were uninsured. Nonprocedural in-hospital care accounted for 20% of total revenue for the program. Financial startup costs for the program were minimal. Primary cost for the program was the 1.5 full-time equivalent dedicated salary support for the hospitalist surgeons. With the support from the hospital per diem and the increased revenue (particularly from collections for improved documentation), the program was self-sustaining.
DISCUSSION In this study, we explored whether reorganizing an acute surgical consultation system into a hospitalist model could improve the quality and timeliness of hospital-based emergency surgical care. We found that a surgical hospitalist delivery system led to higher satisfaction among emergency room referring physicians, shorter patient waiting times (using appendectomy as an index case), and a substantial increase in the number of consults and reimbursement. The service’s success depended on the experience and availability of a core group of surgeons dedicated to assessment, stabilization, and appropriate triage of complex patients. Surgical departments at teaching hospitals nationally are confronting new obstacles to fulfilling the three traditional academic missions of education, research, and the provision of high-quality patient care. General surgery has become highly fragmented in university hospitals, as surgeons have become
J Am Coll Surg
increasingly subspecialized. The combination of increased disease acuity and limits on housestaff duty hours has resulted in a shortage of faculty interested in taking call. The care of these complex patients is time intensive and associated with prolonged lengths of stay, and often requires specialized procedures performed in the middle of night or on weekends.There are few promotional or financial incentives for taking call or for the ongoing care of these patients. These challenges have prompted many to question traditional models of general surgery and trauma call. Although the surgery hospitalist model is probably most relevant to academic medical centers, it might also prove applicable to community-based group practices and other nontrauma settings seeking solutions to the emerging national crisis in patient access to emergency surgical care. Preliminary data suggest that perhaps the greatest beneficiaries of the surgical hospitalist model were the surgeons themselves. Under the new system, most general surgeons at our center were relieved of emergency call and could focus on elective patient care, research, and teaching. Likewise, the 3 surgical hospitalists found their 1 week of continuous call, free from elective clinics and procedures, preferable to the combination of a traditional 24-hour call schedule with an elective practice. The triage system provided a way for junior faculty (who currently represent the majority of the hospitalist surgeons) to become more proficient in the care of complex and challenging patients, with the support of senior experienced surgeons to provide assistance as necessary. An added benefit of this model was the opportunity for the hospitalists to take active roles in medical center quality-improvement projects. A critical element for hospitalist surgeon satisfaction was research time that was free from on-call and inpatient activities when they were not on service. During the 2 weeks off service, these surgeons could focus on their academic interests with only a minimal responsibility for the ongoing care of inpatients or other clinical interruptions. The first of these weeks included only moderate elective cases and clinics. The goal of the second was to be free of all clinical duties with protected time for research. Although other medical centers, such as Vanderbilt, have combined emergency surgical care with trauma care, or into “acute care surgery” programs, in an effort to improve quality of care and reduce costs,17-19 we believe our system of care is the first true surgical hospitalist model for general surgical care in the US. The focus of our program is on improving access and the processes of care for general surgical patients, without an attempt to extend into the domains of neurosurgical or orthopaedic procedures. The European model of “acute care” surgery practice, which originated in Germany, is a trauma care model that emphasizes performing procedures in other trauma subspecialties, with a lesser emphasis on emergency
Vol. 205, No. 5, November 2007
Maa et al
The UCSF Surgical Hospitalist Program
709
general surgery.20 There is no earlier literature describing a similar system of care for general surgical care in the US. The success of the surgical hospitalist program depends on several key principles.
some cases, the revenue received from a moderately complex consult, followed for several days without an operation, could exceed the revenue generated from lesscomplex consults that required an operation.
Timeliness
Team-based group practice
One of our most important findings was that the time between an emergency consult being called and the patient being seen by a surgical hospitalist averaged 20 minutes. Often this would require that attending surgeons evaluate patients directly, before the housestaff could see the patients themselves. This was a distinct change from the previous model, where housestaff would often see patients without the attending being directly involved, or would delay “staffing” the consult with the on-call attending until after a lengthy evaluation had been completed in the ED. This increased timeliness of care served as the foundation of other hospital-based quality improvement efforts, such as one aimed at decreasing patient waiting times for appendectomy and reducing infectious complications, as has been observed in other medical centers with acute care surgical programs.21 One of the advantages of a dedicated hospitalist model is to improve relationships with the operating room, and thereby negotiate more flexible early AM operating times (such as 6:00 AM) for appendectomy rather than having to perform these procedures in the middle of the night.
Surgical hospitalists freely transferred the care of inpatients from one to another at the end of the on-call week, sharing a financial bottom line. Off-call surgical hospitalists were freed from most clinical activities for protected research time. This required a willingness to share in the care of patients and to adopt a team approach to perioperative care, which represents a major departure from the general surgical tradition of a solo practitioner philosophy. Even after signing out, the primary surgeon was still updated about the patient’s care plans after the team rounded each AM and had the opportunity to participate in the patient’s care. An important area of additional work will be to maximize continuity of care during longer hospitalizations and readmissions of a single patient.
Triage
Some cases required specialized expertise, frequently from another member of the surgery faculty. Surgical hospitalists performed the initial stabilization and assessment (often on weekends or late in the PM), after which patient needs were matched to the appropriate surgical expert. Interestingly, ⬍ 5% of patients required a second surgical specialist, as surgical hospitalists were able to provide care to the vast majority of patients presenting from the ED and wards at a tertiary care medical center. Documentation
Surgical hospitalists increased the complexity and frequency of documentation by attending surgeons for billing purposes, particularly of nonoperative care. Historically, surgeons have not focused on revenue generation from the delivery of care that does not result in an operation. In this study, 58% of consultations did not result in a procedure. One of the primary aims of the hospitalist program was to “recapture” this lost revenue, which eventually yielded approximately 20% of overall revenue to support the service. Using the medicine hospitalist billing model, surgical hospitalist attendings documented the initial consultation and daily progress notes, resulting in an overall 415% increase in revenue generation after creation of the program. In
Resident supervision and surgical education
An important benefit was increased resident supervision on both the wards and in the operating room, as noted in the survey of ED physicians. An important overall goal of the program was to strengthen the commitment of the Department of Surgery to surgical education, and address the challenges in fragmentation and continuity of resident education after the introduction of the 80-hour workweek. Although concerns have been raised about housestaff learning under such a model, our impression was that housestaff valued the real-time contact with faculty surgeons and the role modeling of timely and professional care. In preliminary housestaff evaluations, 42% of the respondents cited teaching as the greatest strength of the hospitalist service, similar to the experience with medical hospitalist programs.22 Overall resident satisfaction is similar to other surgical services, though the residents did need extra time to adapt to the structure of the hospitalist service and its increased role of nurse practitioners. Future work will focus on finding the proper balance between resident supervision and autonomy that preserves both resident education and patient safety. There are a number of limitations of this study. First, this work is retrospective and descriptive without a randomized comparison group. We could not compare the response times to consultation request before implementation of the program, and the group of surgeons who performed appendectomy in the prehospitalist era did not know that this end point would be studied in the future. Second, it describes the experience of a single tertiary academic medical center, where trauma represents a low percentage of patients and the patient population has a relatively favorable payor mix, limiting the generalizability of results. Although
710
Maa et al
The UCSF Surgical Hospitalist Program
the revenue generation from consultations and operation increased substantially, the level of additional hospital or departmental support of the program necessary beyond professional fees cannot be determined from our study and can vary from institution to institution. Our study period was for 1 year; it is possible that the 3 surgical hospitalists were highly motivated to perform and that such results might not be sustainable longterm. Emerging data from the medical hospitalist field indicate that full-time and intense clinical work might be a formula for burnout.23 Calibrating the amount of time on call versus off (and adding in the possibilities of creative work and academic advancement) will be important for surgical hospitalist programs to be sustainable.24 Recognizing this, and to maximize patient safety, our program always required a surgeon to be available on backup call, and expanded in 2006 by adding 3 new surgeons. Currently, none of the hospitalist surgeons is critical care board⫺certified, but we plan to expand the intensive care presence in the future. Future research will assess the effect of surgical hospitalists on average length of hospital stay for several complex conditions and determine whether morbidity and mortality are reduced (which will require adequate sample sizes and a possible multicenter study). Patient satisfaction, and the satisfaction of the other nonhospitalist surgical faculty members, will be assessed through survey instruments. It will be important to determine whether an increased value of the quality of care that is delivered can offset the costs to sustain the program. Although our initial experience showed that substantial resources can be generated by the hospitalist service, an important unanswered question will be to determine the level of external support appropriate to sustain the surgical hospitalist program. Length of stay and costs can actually rise if increasingly complex and ill patients are referred to UCSF from the community. The medical hospitalist field has been nationally supported by medical centers, which have recognized the value added by the hospitalists and the need for institutional support to make up for shortfalls in clinical revenue generation that occur because of time spent coordinating care, often for an unfavorable payor mix. A similar dialogue will need to take place about the support of surgical hospitalist programs. As important as support is for the physicians, medical centers will not reap the full benefits of surgical hospitalists until they deploy resources like weekend social workers, PM CT scan technicians, or a dedicated emergency operating room suite. Finally, the question of how surgical hospitalists will interface with surgical subspecialists and with medical hospitalists remains unanswered at present. We view the surgical hospitalist program as a stepping stone toward
J Am Coll Surg
hospital-based patient safety initiatives, such as perioperative wound infection prevention, deep venous thrombosis prophylaxis, and myocardial infarction prevention. We envision strong and productive collaborations with our medical hospitalist colleagues to develop clinical investigations that will transform the surgical delivery system, with an emphasis on error reduction, improved signouts, and evidence-based operations. Preliminary evidence suggests that increasing expertise in the delivery of care in an emergency setting can lead to improvements in the care of elective patients as well. Longer-range efforts will seek to identify coordinated efforts between medicine and surgery hospitalist programs to maximize hospital efficiencies and to address the national challenges of ED boarding and diversion and inadequate access to emergency surgical care.25 The surgical hospitalist model is a new paradigm for the organization of surgical departments in academic medical centers that has potential applicability to a wide variety of practice settings. It proposes a new strategy to addressing the crisis in access to emergency surgical care in the US. Its creation recognizes and responds to several important trends: the evidence that on-site availability is critical to provision of high-quality, cost-effective care; the need in academic centers to provide quality and prompt care in the face of fewer available housestaff hours; and the demonstrated value of on-site availability for urgent and emergent care. Although patients with general surgical emergencies have come to be regarded as the orphans of academic surgery, the time has arrived to offer these patients better outcomes by enhancing the structure and processes of emergency surgical care in teaching hospitals. Author Contributions
Study conception and design: Maa, Gosnell, Harris Acquisition of data: Maa, Carter, Gosnell, Harris Analysis and interpretation of data: Maa, Carter, Wachter, Harris Drafting of manuscript: Maa, Carter, Wachter, Harris Critical revision: Maa, Carter, Gosnell, Wachter, Harris Acknowledgment: We wish to thank Erik Nakakura, Guilherme Campos, Robin Andersen, Danielle Kreiger, Joan O’Mahoney, Danielle Blanc, and Nancy Ascher for their support of the UCSF surgical hospitalist program. We also thank Pamela Derish in the UCSF Department of Surgery for editorial advice, and the Writing Seminar members at the UCSF Institute of Health Policy Studies for their review and assistance with the article preparation.
Vol. 205, No. 5, November 2007
REFERENCES 1. A growing crisis in patient access to emergency surgical care. Report of the American College of Surgeons Division of Advocacy and Health Policy. October 2006. Available at: http://www.facs.org/ ahp/emergcarecrisis.pdf. Accessed: October 30, 2006. 2. On-call specialist coverage in US emergency departments, American College of Emergency Physicians Survey of Emergency Department Directors, April 2006. Available at: http://www.acep.org/NR/ rdonlyres/. Accessed: October 30, 2006. 3. The Lewin Group analysis of AHA ED Hospital Capacity Survey, 2002. Available at: http://www.aha.org/ahapolicyforum/resources/ EDdiversionsurvey0404/html. Accessed: October 30, 2006. 4. Salsberg E, Grover A. Physician workforce shortages: implications and issues for academic health centers and policymakers. Acad Med 2006;81:782–787. 5. Richardson JD. Workforce and lifestyle issues in general surgery training and practice. Arch Surg 2002;137:515–520. 6. US General Accounting Office. Medical malpractice: implications of rising premiums on access to healthcare. Washington, DC: US Government Printing Office; 2003 (GAO-03-836). 7. Institute of Medicine. Hospital-based emergency care: at the breaking point. Washington, DC: National Academy Press; 2006. 8. Garner JP, Prytherch D. Subspecialization in general surgery: the problem of providing a safe emergency general surgical service. Colorectal Dis 2006;8:273–277. 9. Dawson EJ, Paterson-Brown S. Emergency general surgery and the implications for specialisation. Surgeon 2004;2:165–170. 10. Lankester BJ, Britton DC, Holbrook AG, et al. Emergency surgery: atavistic refuge of the general surgeon? J R Soc Med 2001; 94:180–182. 11. Debas HT, Bass BL, Brennan MF, et al. American Surgical Association blue ribbon committee report on surgical education: 2004. Ann Surg 2005;241:1–8. 12. Wachter RM, Goldman L. The emerging role of hospitalists in the American health care system. N Engl J Med 1996;335:514–517.
Maa et al
The UCSF Surgical Hospitalist Program
711
13. Wachter RM, Goldman L. The hospitalist movement 5 years later. JAMA 2002;287:487–494. 14. Hoff TH, Whitcomb WF, Williams K, et al. Characteristics and work experiences of hospitalists in the United States. Arch Intern Med 2001;161:851–858. 15. Merli GJ. The hospitalist joins the surgical team. Ann Intern Med 2004;141:67–69. 16. Huddleston JM, Hall Long K, Naessens JM, et al. Medical and surgical comanagement after elective hip and knee arthoplasty. A randomized, controlled trial. Ann Intern Med 2004;141:28–38. 17. Austin MT, Diaz JJ, Feurer ID, et al. Creating an emergency general surgery service enhances the productivity of trauma surgeons, general surgeons and the hospital. J Trauma 2005;58:906–910. 18. Pryor JP, Reilly PM, Schwab W, et al. Integrating emergency general surgery with a trauma service: impact on the care of injured patients. J Trauma 2004;57:467–473. 19. Scherer LA, Battistella FD. Trauma and emergency surgery: an evolutionary direction for trauma surgeons. J Trauma 2004;56:7–12. 20. Moore EE, Maier RV, Hoyt DB, et al. Acute care surgery: Eraritjaritjaka. J Am Coll Surg 2006;202:698–701. 21. Earley AS, Pryor JP, Kim PK, et al. An acute care surgery model improves outcomes in patients with appendicitis. Ann Surg 2006;244:498–504. 22. Hauer KE, Wachter RM. Implications of the hospitalist model for medical students’ education. Acad Med 2001;76:324–330. 23. Hoff TH, Whitcomb WF, Williams K, et al. Characteristics and work experiences of hospitalists in the United States. Arch Intern Med 2001;161:851–8. 24. Wachter RM, Goldman L. Implications of the hospitalist movement for academic departments of medicine: lessons from the UCSF experience. Am J Med 1999;106:127–133. 25. Olshaker JS, Rathlev NK. Emergency department overcrowding and ambulance diversion: the impact and potential solutions of extended boarding of admitted patients in the emergency department. J Emerg Med 2006;30:351–356.
Lack of access to acute surgical care in the US has reached crisis proportions and is now the topic of a major national debate.1 A 2005 survey conducted by the American College of Emergency Physicians revealed that nearly 75% of emergency department (ED) medical directors believed they had inadequate surgical coverage,2 and general surgeons were among the key specialists in short supply.3 The cause of the surgeon shortage is multifactorial and includes declining reimbursement, escalating malpractice premiums, a trend toward practice in ambulatory surgery centers, a shrinking supply of surgeons in the face of a growing patient population, an aging surgeon population, and an unwillingness to disrupt elective surgical practice.4-6 The
recent Institute of Medicine report “Hospital-Based Emergency Care: At the Breaking Point” reported that hospitals in several states across America have closed their EDs because of inadequate surgical coverage, a trend with devastating consequences for critically injured or uninsured patients.7 Even within “general surgery,” the growing popularity of specialization (often achieved through additional fellowship training) has added to the problem. As a result, many surgeons have become less comfortable dealing with general surgical conditions outside their areas of expertise, especially when these conditions present emergently.8-10 Additionally, because surgical acute care in academic medical centers was historically dependent on housestaff, the Accreditation Council for Graduate Medical Education’s 80-hour workweek, introduced in 2001, has limited even more the availability of acute surgical consultation services.11 Given the critical shortage of surgeons available and willing to take emergency call, it is clear that new models of care are needed to address growing public demand.
Competing Interests Declared: None. Received January 8, 2007; Revised March 29, 2007; Accepted May 9, 2007. From the Departments of Surgery (Maa, Carter, Gosnell, Harris) and Medicine (Wachter), University of California-San Francisco, San Francisco, CA. Correspondence address: John Maa, MD, FACS, Department of General Surgery, University of California-San Francisco, 1600 Divisadero, Rm C-322D, Box 1674, San Francisco, CA 94103. email: [email protected]
© 2007 by the American College of Surgeons Published by Elsevier Inc.
704
ISSN 1072-7515/07/$32.00 doi:10.1016/j.jamcollsurg.2007.05.008
Vol. 205, No. 5, November 2007
In 2005, we reorganized our acute general surgical service at the University of California-San Francisco (UCSF) Medical Center into a hospitalist model in an effort to address these challenges. As originally conceived, hospitalists are physicians who dedicate the majority of their time to the care of hospitalized inpatients of primary care providers, with the intent of handing the care back to the primary care physician after the patient is discharged from the hospital. This model of care was introduced on the medicine services at our hospital in the early 1990s, in response to pressures of the managed care movement, and resulted in improved quality, reduced length of stay, and improved patient safety.12 By focusing on continuity and comprehensiveness of care, medical hospitalists have made substantial progress in both quality and efficiency improvement efforts for inpatients. The model has become well accepted nationally,13 with ⬎ 15,000 hospitalists in the US today. Surveys of hospitalists show that the vast majority are general internists (74%) or general pediatricians (6%); few, if any, “surgical hospitalists” have been reported.14 Instead, discussions about the role of hospitalists on surgical services tend to center on medical co-management of surgical patients,15,16 quite possibly a salutary trend, but one that fails to address the needs of surgical inpatients for timely, highquality, and cost-effective surgical diagnosis and treatment decisions. We hypothesized that creation of a surgical hospitalist program could realize many of the same benefits achieved by medical hospitalists: improved timeliness of care, improved satisfaction of referring ED providers, increased consult volume and revenue, better resident supervision and continuity of care, and shortened waiting times, using appendectomy as an index case. In this article, we describe the surgical hospitalist model and report our initial experience.
METHODS Structure of the surgical hospitalist service
Before July 2005, emergency general surgical care at UCSF was provided by a diverse faculty of gastrointestinal, hepatobiliary, colorectal, oncologic, endocrine, breast, and bariatric surgeons spread across 2 campuses separated by 3 miles, each surgeon taking call on a 24-hour basis. This system of care was problematic for several reasons. First, daytime consultations disrupted the elective procedures and clinics of on-call surgeons. Patients in the ED or acute care ward might wait hours until the on-call surgeon was available to evaluate them. Second, the diversity of emergency surgical conditions left many surgeons uncomfortable caring for diseases outside their usual practice. Third, the 24-hour structure of the call schedule and separate campuses disrupted continuity of care, particularly for
Maa et al
The UCSF Surgical Hospitalist Program
705
those patients treated by surgeons at the remote site who would need subsequent care in the ED at the main hospital. Surgical housestaff provided the only continuity and were constrained even more by the 80-hour Accreditation Council for Graduate Medical Education workweek. Finally, there was little economic incentive for taking call; the only benefit to a surgeon or the department was revenue generated from the minority of consultations that resulted in a surgical procedure. In July 2005, we restructured our general surgery call system into a surgical hospitalist model to provide coverage at the main campus, which houses the ED and accounts for 95% of the urgent consult requests. Three full-time boardcertified general surgeons staffed the service on a rotating weekly basis, dedicating 100% of their time to ED and inpatient consults. A per-diem fee was historically paid by the hospital to the department for ED coverage, and this compensation continued unchanged to the hospitalist team. These surgeons had minimal elective procedures or clinics scheduled during their on-call service weeks. Surgical hospitalist attendings rounded daily, providing continuity of care and supervision of housestaff. A senior resident, a midlevel resident, a nurse practitioner, and medical students were assigned to the hospitalist service to gain experience in emergency surgical disease. This team also cared for the elective practice of the bariatric and endocrine surgery services, with a volume of approximately 400 inpatients per year. A goal of timely surgical consultation (ie, within 30 minutes during weekdays and 45 minutes on weeknights and weekends) was implemented as a primary measure of performance. A backup surgeon (usually one of the other hospitalist surgeons) was always available. For those patients possibly requiring complex surgical intervention (advanced hepatobiliary, foregut, thoracic, or endocrine procedures), the surgical hospitalist would consult a senior experienced surgeon for additional care after the patient had been initially stabilized and admitted. In addition to the direct clinical care, hospitalists also managed the majority of surgical transfer requests from referring hospitals throughout Northern California. Once discharged, patients were followed as outpatients in a surgical hospitalist clinic, staffed by off-call surgical hospitalists. The key elements of the surgical hospitalist model are listed in Table 1. Patient, disease, and financial characteristics
We prospectively recorded the age, gender, race, chief complaint, diagnosis, medical comorbidities, service requesting consultation, and time to consult (defined as the number of minutes from initial phone consultation to bedside evaluation) for each patient evaluated by surgical hospitalists. We reviewed the percentage of consultations resulting in an operative procedure, the type of procedure performed, and
706
Maa et al
The UCSF Surgical Hospitalist Program
Table 1. Key Elements of the Surgical Hospitalist Model On-call period lasts continuously for 1 week, not 24 hours, in order to improve continuity of care. During the on-call period, no elective clinics or procedures are scheduled that might disrupt or conflict with acute surgical care. A resident or attending should evaluate the patient within 30 minutes of consultation during business hours and within 45 minutes off hours. If the resident is unavailable, then the on-call attending will be contacted directly to see the patient independently. Patients requiring special expertise are initially assessed by the team and then reassigned (triaged) to a higher level of expert care as indicated. After the on-call period, the care of inpatients and consults is handed off to the next on-call surgeon in a group-practice model. Patient safety is enhanced though increased resident supervision and improved signouts. Revenue stems primarily from per-diem payment from the hospital, procedural fees, and attending documentation of nonoperative care.
the percentage of consultations requiring additional surgical expertise outside our practice (such as advanced esophageal, colorectal, or hepatic resection). Financial data, including year-to-year billing charges, collections, and payor mix, were provided by the Division of Finance of the Department of General Surgery. Emergency department physician satisfaction and patient outcomes
Because a primary goal of the surgical hospitalist program was to provide timely and efficient consultations to ED providers, we conducted an anonymous survey of ED providers 6 months after the start of the program. Nine questions were structured on a 5-point Likert scale (1 ⫽ improved dramatically, 5 ⫽ worsened dramatically) and included the following, “since July 1, 2005, how have the following changed: 1) the timeliness of general surgery evaluations and recommendations, 2) the professionalism of general surgeons toward ED staff, and 3) supervision of surgical residents by surgery attendings.” We also reviewed the records of all patients presenting to the ED with acute appendicitis and tracked the time from ED triage to appendectomy as another measure of our program’s performance. Patients undergoing appendectomy during the 6 months before the start of the study served as controls. Data collection and statistical analysis
All data were stored on a password-protected Health Insurance Portability and Accountability Act⫺compliant server. This study was approved by the UCSF Committee on Human Research. Student’s t-test was used to compare appendectomy wait times. Statistical significance was defined as p ⬍ 0.05.
J Am Coll Surg
RESULTS Patient and disease characteristics
Three surgical hospitalists cared for 853 patients during a 1-year period, averaging 2.3 new consults per day. Patient and disease characteristics are listed in Table 2. Patients averaged 53 years of age (range 17 to 100), and the maleto-female ratio was nearly equal (51:49). Patients were racially diverse (63% Caucasian, 17% Asian, 11% African American, and 9% Hispanic). The majority presented with acute abdominal pain (66%), and soft-tissue infections (18%), malignancy (6%), and hernia (4%) were other major sources of referrals. Because our center is a Level II trauma center, whereby most trauma patients are regionalized to San Francisco General Hospital, trauma accounted Table 2. Characteristics of 853 Patients Evaluated by Surgical Hospitalists at University of California-San Francisco from July 1, 2005, to June 30, 2006 Age (y), mean (range) 53 (17⫺100) Gender, n (%) Male 432 (51) Female 421 (49) Ethnicity, n (%) African American 97 (11) Asian/Pacific Islander 144 (17) Caucasian 540 (63) Hispanic/Latino 72 (9) Reason for consultation, n (%) Acute appendicitis 118 (14) Bowel obstruction or ileus 87 (10) Biliary tract disease 91 (10) Colorectal or anal disease 66 (8) Mesenteric ischemia 33 (4) Pancreatitis (excluding biliary) 33 (4) Other acute abdomen 138 (16) Hernia 33 (4) Superficial or deep space infection 157 (18) Trauma 22 (3) Malignancy 47 (6) Other 28 (3) Source of consultation, n (%) Emergency department 487 (57) General medicine or pediatrics 167 (20) Medical specialist, including critical care 61 (7) Obstetrics/gynecology 18 (2) Surgical specialist 67 (8) Transfer from another institution 53 (6) Wait time before surgical evaluation (min), mean (range) 20 (1⫺270) Patients evaluated, n (%) Within 30 min of consult 679 (80) Within 45 min of consult 727 (85)
Vol. 205, No. 5, November 2007
Maa et al
Table 3. Case Log of Surgical Procedures Performed Appendectomy Incision and drainage of abscess Major abdominal (intestinal resection, lysis of adhesions) Cholecystectomy Complex liver/spleen/pancreas procedures Hernia repair Soft tissue subcutaneous procedures Other Total cases
n
%
103 70
29 19
69 39
19 11
24 15 14 25 359
7 4 4 7
for only 3% of consultations. The majority of consultations originated from the ED (57%), although general medicine (20%) and medical specialists (7%), including critical care physicians, accounted for large numbers. Interestingly, other surgeons were responsible for 8% of the total consult volume; these were generally from nongeneral surgical specialists, such as neurosurgeons or urologists. Transfer patients from other institutions accounted for 6% of consults. Surgical procedures performed and triage to senior expert surgeons
Forty-two percent of consultations (n ⫽ 359) resulted in an operative procedure during the index admission (Table 3). The most common procedures performed included appendectomy (29%), incision and drainage of abscess (19%), exploratory laparotomy for intestinal resection or lysis of adhesions (19%), cholecystectomy (11%), and complex liver/spleen/pancreas procedures (7%). We involved other
The UCSF Surgical Hospitalist Program
707
Table 4. Patients Triaged and Referred to Senior General Surgery Specialist
Service
Advanced laparoscopic Bariatric Breast Colorectal Endocrine Foregut Hepatobiliary Thoracic Totals
Needed consult from a senior specialist
Needed operation by that specialist
7 1 5 8 1 7 6 2 37
7 0 2 5 1 6 6 2 29
surgeons with advanced expertise in 37 of 853 patients (4.3%). Surgical intervention was necessary in 29 of these 37 patients (Table 4), and the remaining 92% (330 of 359) of all procedures were performed by surgical hospitalists without involving other surgeons. Time to consultation and ED provider satisfaction
Time to consult, defined as the time from telephone consultation to bedside evaluation, averaged 16 minutes. Eighty percent of consults were seen within 30 minutes, and 85% were seen within 45 minutes (Table 2). The confidential survey of ED providers 6 months after our program began had a response rate of 76% (13 of 17). Results of the survey indicated that two-thirds of ED providers were aware of the surgical hospitalist program’s creation and intent (Fig. 1). All providers who participated in the survey believed that the surgical hospitalist program had
Figure 1. Survey of emergency department providers after 6 months of University of CaliforniaSan Francisco surgical hospitalist care.
708
Maa et al
The UCSF Surgical Hospitalist Program
improved timeliness of care, ED length of stay, supervision of housestaff, patient satisfaction, and professionalism of the surgical staff. Only 1 provider believed overall quality of care had decreased; the remainder believed the quality was either the same (38%) or better (46%). When we compared the wait time (from ED triage to skin incision) for patients undergoing appendectomy in a 6-month period before and after the start of our program, we found that it decreased 50%, from 16 ⫾ 10 hours to 8 ⫾ 4 hours (p ⬍ 0.05). The rate of wound infection and postoperative intraabdominal abscess were both ⬍ 4%. Revenue generation and payor mix
In the first year of the surgical hospitalist program, we saw a 190% increase in requested consults from the ED and inpatient wards. This led to a 415% increase in year-overyear billable consult revenue, with a minimal impact on the individual consult revenue for the remainder of the division of general surgery (which will be addressed in future studies). The greatest increase was observed in the area of subsequent care and followup (“subsequent hospital care,” by CPT code), where a 24-fold increase in revenue was achieved through improved documentation and billing. Fifty-one percent of patients treated were insured under capitated care plans, 44% had Medicare/Medicaid, and 4% were uninsured. Nonprocedural in-hospital care accounted for 20% of total revenue for the program. Financial startup costs for the program were minimal. Primary cost for the program was the 1.5 full-time equivalent dedicated salary support for the hospitalist surgeons. With the support from the hospital per diem and the increased revenue (particularly from collections for improved documentation), the program was self-sustaining.
DISCUSSION In this study, we explored whether reorganizing an acute surgical consultation system into a hospitalist model could improve the quality and timeliness of hospital-based emergency surgical care. We found that a surgical hospitalist delivery system led to higher satisfaction among emergency room referring physicians, shorter patient waiting times (using appendectomy as an index case), and a substantial increase in the number of consults and reimbursement. The service’s success depended on the experience and availability of a core group of surgeons dedicated to assessment, stabilization, and appropriate triage of complex patients. Surgical departments at teaching hospitals nationally are confronting new obstacles to fulfilling the three traditional academic missions of education, research, and the provision of high-quality patient care. General surgery has become highly fragmented in university hospitals, as surgeons have become
J Am Coll Surg
increasingly subspecialized. The combination of increased disease acuity and limits on housestaff duty hours has resulted in a shortage of faculty interested in taking call. The care of these complex patients is time intensive and associated with prolonged lengths of stay, and often requires specialized procedures performed in the middle of night or on weekends.There are few promotional or financial incentives for taking call or for the ongoing care of these patients. These challenges have prompted many to question traditional models of general surgery and trauma call. Although the surgery hospitalist model is probably most relevant to academic medical centers, it might also prove applicable to community-based group practices and other nontrauma settings seeking solutions to the emerging national crisis in patient access to emergency surgical care. Preliminary data suggest that perhaps the greatest beneficiaries of the surgical hospitalist model were the surgeons themselves. Under the new system, most general surgeons at our center were relieved of emergency call and could focus on elective patient care, research, and teaching. Likewise, the 3 surgical hospitalists found their 1 week of continuous call, free from elective clinics and procedures, preferable to the combination of a traditional 24-hour call schedule with an elective practice. The triage system provided a way for junior faculty (who currently represent the majority of the hospitalist surgeons) to become more proficient in the care of complex and challenging patients, with the support of senior experienced surgeons to provide assistance as necessary. An added benefit of this model was the opportunity for the hospitalists to take active roles in medical center quality-improvement projects. A critical element for hospitalist surgeon satisfaction was research time that was free from on-call and inpatient activities when they were not on service. During the 2 weeks off service, these surgeons could focus on their academic interests with only a minimal responsibility for the ongoing care of inpatients or other clinical interruptions. The first of these weeks included only moderate elective cases and clinics. The goal of the second was to be free of all clinical duties with protected time for research. Although other medical centers, such as Vanderbilt, have combined emergency surgical care with trauma care, or into “acute care surgery” programs, in an effort to improve quality of care and reduce costs,17-19 we believe our system of care is the first true surgical hospitalist model for general surgical care in the US. The focus of our program is on improving access and the processes of care for general surgical patients, without an attempt to extend into the domains of neurosurgical or orthopaedic procedures. The European model of “acute care” surgery practice, which originated in Germany, is a trauma care model that emphasizes performing procedures in other trauma subspecialties, with a lesser emphasis on emergency
Vol. 205, No. 5, November 2007
Maa et al
The UCSF Surgical Hospitalist Program
709
general surgery.20 There is no earlier literature describing a similar system of care for general surgical care in the US. The success of the surgical hospitalist program depends on several key principles.
some cases, the revenue received from a moderately complex consult, followed for several days without an operation, could exceed the revenue generated from lesscomplex consults that required an operation.
Timeliness
Team-based group practice
One of our most important findings was that the time between an emergency consult being called and the patient being seen by a surgical hospitalist averaged 20 minutes. Often this would require that attending surgeons evaluate patients directly, before the housestaff could see the patients themselves. This was a distinct change from the previous model, where housestaff would often see patients without the attending being directly involved, or would delay “staffing” the consult with the on-call attending until after a lengthy evaluation had been completed in the ED. This increased timeliness of care served as the foundation of other hospital-based quality improvement efforts, such as one aimed at decreasing patient waiting times for appendectomy and reducing infectious complications, as has been observed in other medical centers with acute care surgical programs.21 One of the advantages of a dedicated hospitalist model is to improve relationships with the operating room, and thereby negotiate more flexible early AM operating times (such as 6:00 AM) for appendectomy rather than having to perform these procedures in the middle of the night.
Surgical hospitalists freely transferred the care of inpatients from one to another at the end of the on-call week, sharing a financial bottom line. Off-call surgical hospitalists were freed from most clinical activities for protected research time. This required a willingness to share in the care of patients and to adopt a team approach to perioperative care, which represents a major departure from the general surgical tradition of a solo practitioner philosophy. Even after signing out, the primary surgeon was still updated about the patient’s care plans after the team rounded each AM and had the opportunity to participate in the patient’s care. An important area of additional work will be to maximize continuity of care during longer hospitalizations and readmissions of a single patient.
Triage
Some cases required specialized expertise, frequently from another member of the surgery faculty. Surgical hospitalists performed the initial stabilization and assessment (often on weekends or late in the PM), after which patient needs were matched to the appropriate surgical expert. Interestingly, ⬍ 5% of patients required a second surgical specialist, as surgical hospitalists were able to provide care to the vast majority of patients presenting from the ED and wards at a tertiary care medical center. Documentation
Surgical hospitalists increased the complexity and frequency of documentation by attending surgeons for billing purposes, particularly of nonoperative care. Historically, surgeons have not focused on revenue generation from the delivery of care that does not result in an operation. In this study, 58% of consultations did not result in a procedure. One of the primary aims of the hospitalist program was to “recapture” this lost revenue, which eventually yielded approximately 20% of overall revenue to support the service. Using the medicine hospitalist billing model, surgical hospitalist attendings documented the initial consultation and daily progress notes, resulting in an overall 415% increase in revenue generation after creation of the program. In
Resident supervision and surgical education
An important benefit was increased resident supervision on both the wards and in the operating room, as noted in the survey of ED physicians. An important overall goal of the program was to strengthen the commitment of the Department of Surgery to surgical education, and address the challenges in fragmentation and continuity of resident education after the introduction of the 80-hour workweek. Although concerns have been raised about housestaff learning under such a model, our impression was that housestaff valued the real-time contact with faculty surgeons and the role modeling of timely and professional care. In preliminary housestaff evaluations, 42% of the respondents cited teaching as the greatest strength of the hospitalist service, similar to the experience with medical hospitalist programs.22 Overall resident satisfaction is similar to other surgical services, though the residents did need extra time to adapt to the structure of the hospitalist service and its increased role of nurse practitioners. Future work will focus on finding the proper balance between resident supervision and autonomy that preserves both resident education and patient safety. There are a number of limitations of this study. First, this work is retrospective and descriptive without a randomized comparison group. We could not compare the response times to consultation request before implementation of the program, and the group of surgeons who performed appendectomy in the prehospitalist era did not know that this end point would be studied in the future. Second, it describes the experience of a single tertiary academic medical center, where trauma represents a low percentage of patients and the patient population has a relatively favorable payor mix, limiting the generalizability of results. Although
710
Maa et al
The UCSF Surgical Hospitalist Program
the revenue generation from consultations and operation increased substantially, the level of additional hospital or departmental support of the program necessary beyond professional fees cannot be determined from our study and can vary from institution to institution. Our study period was for 1 year; it is possible that the 3 surgical hospitalists were highly motivated to perform and that such results might not be sustainable longterm. Emerging data from the medical hospitalist field indicate that full-time and intense clinical work might be a formula for burnout.23 Calibrating the amount of time on call versus off (and adding in the possibilities of creative work and academic advancement) will be important for surgical hospitalist programs to be sustainable.24 Recognizing this, and to maximize patient safety, our program always required a surgeon to be available on backup call, and expanded in 2006 by adding 3 new surgeons. Currently, none of the hospitalist surgeons is critical care board⫺certified, but we plan to expand the intensive care presence in the future. Future research will assess the effect of surgical hospitalists on average length of hospital stay for several complex conditions and determine whether morbidity and mortality are reduced (which will require adequate sample sizes and a possible multicenter study). Patient satisfaction, and the satisfaction of the other nonhospitalist surgical faculty members, will be assessed through survey instruments. It will be important to determine whether an increased value of the quality of care that is delivered can offset the costs to sustain the program. Although our initial experience showed that substantial resources can be generated by the hospitalist service, an important unanswered question will be to determine the level of external support appropriate to sustain the surgical hospitalist program. Length of stay and costs can actually rise if increasingly complex and ill patients are referred to UCSF from the community. The medical hospitalist field has been nationally supported by medical centers, which have recognized the value added by the hospitalists and the need for institutional support to make up for shortfalls in clinical revenue generation that occur because of time spent coordinating care, often for an unfavorable payor mix. A similar dialogue will need to take place about the support of surgical hospitalist programs. As important as support is for the physicians, medical centers will not reap the full benefits of surgical hospitalists until they deploy resources like weekend social workers, PM CT scan technicians, or a dedicated emergency operating room suite. Finally, the question of how surgical hospitalists will interface with surgical subspecialists and with medical hospitalists remains unanswered at present. We view the surgical hospitalist program as a stepping stone toward
J Am Coll Surg
hospital-based patient safety initiatives, such as perioperative wound infection prevention, deep venous thrombosis prophylaxis, and myocardial infarction prevention. We envision strong and productive collaborations with our medical hospitalist colleagues to develop clinical investigations that will transform the surgical delivery system, with an emphasis on error reduction, improved signouts, and evidence-based operations. Preliminary evidence suggests that increasing expertise in the delivery of care in an emergency setting can lead to improvements in the care of elective patients as well. Longer-range efforts will seek to identify coordinated efforts between medicine and surgery hospitalist programs to maximize hospital efficiencies and to address the national challenges of ED boarding and diversion and inadequate access to emergency surgical care.25 The surgical hospitalist model is a new paradigm for the organization of surgical departments in academic medical centers that has potential applicability to a wide variety of practice settings. It proposes a new strategy to addressing the crisis in access to emergency surgical care in the US. Its creation recognizes and responds to several important trends: the evidence that on-site availability is critical to provision of high-quality, cost-effective care; the need in academic centers to provide quality and prompt care in the face of fewer available housestaff hours; and the demonstrated value of on-site availability for urgent and emergent care. Although patients with general surgical emergencies have come to be regarded as the orphans of academic surgery, the time has arrived to offer these patients better outcomes by enhancing the structure and processes of emergency surgical care in teaching hospitals. Author Contributions
Study conception and design: Maa, Gosnell, Harris Acquisition of data: Maa, Carter, Gosnell, Harris Analysis and interpretation of data: Maa, Carter, Wachter, Harris Drafting of manuscript: Maa, Carter, Wachter, Harris Critical revision: Maa, Carter, Gosnell, Wachter, Harris Acknowledgment: We wish to thank Erik Nakakura, Guilherme Campos, Robin Andersen, Danielle Kreiger, Joan O’Mahoney, Danielle Blanc, and Nancy Ascher for their support of the UCSF surgical hospitalist program. We also thank Pamela Derish in the UCSF Department of Surgery for editorial advice, and the Writing Seminar members at the UCSF Institute of Health Policy Studies for their review and assistance with the article preparation.
Vol. 205, No. 5, November 2007
REFERENCES 1. A growing crisis in patient access to emergency surgical care. Report of the American College of Surgeons Division of Advocacy and Health Policy. October 2006. Available at: http://www.facs.org/ ahp/emergcarecrisis.pdf. Accessed: October 30, 2006. 2. On-call specialist coverage in US emergency departments, American College of Emergency Physicians Survey of Emergency Department Directors, April 2006. Available at: http://www.acep.org/NR/ rdonlyres/. Accessed: October 30, 2006. 3. The Lewin Group analysis of AHA ED Hospital Capacity Survey, 2002. Available at: http://www.aha.org/ahapolicyforum/resources/ EDdiversionsurvey0404/html. Accessed: October 30, 2006. 4. Salsberg E, Grover A. Physician workforce shortages: implications and issues for academic health centers and policymakers. Acad Med 2006;81:782–787. 5. Richardson JD. Workforce and lifestyle issues in general surgery training and practice. Arch Surg 2002;137:515–520. 6. US General Accounting Office. Medical malpractice: implications of rising premiums on access to healthcare. Washington, DC: US Government Printing Office; 2003 (GAO-03-836). 7. Institute of Medicine. Hospital-based emergency care: at the breaking point. Washington, DC: National Academy Press; 2006. 8. Garner JP, Prytherch D. Subspecialization in general surgery: the problem of providing a safe emergency general surgical service. Colorectal Dis 2006;8:273–277. 9. Dawson EJ, Paterson-Brown S. Emergency general surgery and the implications for specialisation. Surgeon 2004;2:165–170. 10. Lankester BJ, Britton DC, Holbrook AG, et al. Emergency surgery: atavistic refuge of the general surgeon? J R Soc Med 2001; 94:180–182. 11. Debas HT, Bass BL, Brennan MF, et al. American Surgical Association blue ribbon committee report on surgical education: 2004. Ann Surg 2005;241:1–8. 12. Wachter RM, Goldman L. The emerging role of hospitalists in the American health care system. N Engl J Med 1996;335:514–517.
Maa et al
The UCSF Surgical Hospitalist Program
711
13. Wachter RM, Goldman L. The hospitalist movement 5 years later. JAMA 2002;287:487–494. 14. Hoff TH, Whitcomb WF, Williams K, et al. Characteristics and work experiences of hospitalists in the United States. Arch Intern Med 2001;161:851–858. 15. Merli GJ. The hospitalist joins the surgical team. Ann Intern Med 2004;141:67–69. 16. Huddleston JM, Hall Long K, Naessens JM, et al. Medical and surgical comanagement after elective hip and knee arthoplasty. A randomized, controlled trial. Ann Intern Med 2004;141:28–38. 17. Austin MT, Diaz JJ, Feurer ID, et al. Creating an emergency general surgery service enhances the productivity of trauma surgeons, general surgeons and the hospital. J Trauma 2005;58:906–910. 18. Pryor JP, Reilly PM, Schwab W, et al. Integrating emergency general surgery with a trauma service: impact on the care of injured patients. J Trauma 2004;57:467–473. 19. Scherer LA, Battistella FD. Trauma and emergency surgery: an evolutionary direction for trauma surgeons. J Trauma 2004;56:7–12. 20. Moore EE, Maier RV, Hoyt DB, et al. Acute care surgery: Eraritjaritjaka. J Am Coll Surg 2006;202:698–701. 21. Earley AS, Pryor JP, Kim PK, et al. An acute care surgery model improves outcomes in patients with appendicitis. Ann Surg 2006;244:498–504. 22. Hauer KE, Wachter RM. Implications of the hospitalist model for medical students’ education. Acad Med 2001;76:324–330. 23. Hoff TH, Whitcomb WF, Williams K, et al. Characteristics and work experiences of hospitalists in the United States. Arch Intern Med 2001;161:851–8. 24. Wachter RM, Goldman L. Implications of the hospitalist movement for academic departments of medicine: lessons from the UCSF experience. Am J Med 1999;106:127–133. 25. Olshaker JS, Rathlev NK. Emergency department overcrowding and ambulance diversion: the impact and potential solutions of extended boarding of admitted patients in the emergency department. J Emerg Med 2006;30:351–356.