WTA 2014 PLENARY PAPER
Attribution: Whose complication is it? Jason Murry, MD, Greg Hambright, MD, Nimesh Patel, MD, Peter Rappa, MD, Michael Truitt, MD, and Ernest Dunn, MD, Dallas, Texas
To improve quality, programs such as accountable care organizations need to determine the part of the health care system most ‘‘responsible’’ for a complication. This is referred to as attribution. This provides a framework to compare physicians for patients and third-party payers. Traditionally, the attribution of complications has been to the admitting physician. This may misidentify the physician ‘‘responsible’’ for the complication. This is especially difficult in trauma patients who have multiple providers. We hypothesized that the current mechanism for attributing complications in trauma patients is inadequate and will need to be modernized. METHODS: All trauma admissions during a 12-month period were reviewed. Patients with single-system trauma were excluded. We reviewed our trauma database for mechanism of injury, complications, and readmissions. The trauma director and the medical director of our accountable care organizations reviewed all complications and attributed them to the appropriate health care provider. These were compared with the hospital decisions using the traditional definition. RESULTS: The trauma service had 1,526 admissions. After exclusions, 1,019 patients were reviewed. One hundred twenty-five complications occurred in 73 patients. Using the traditional definition, the acute care surgery service was assigned all 125 complications. Using the trauma director and medical director method, the neurosurgical attending accounted for 36% (45 of 125) of complications. The acute care surgery attending was responsible for 34% (43 of 125) of complications, and orthopedic surgery was identified as the causative factor in 22% (27 of 125). The remaining 8% (10 of 125) were attributed to various other services. Seven patients had unexpected readmissions. Most (6 of 7) of these were related to orthopedics. CONCLUSION: Hospital complications are now being assigned to individual surgeons. Which physician is responsible for each complication will be a controversial matter. Without a critical review process with physician input, up to two thirds of complications could be attributed incorrectly. The attribution process needs to be refined. (J Trauma Acute Care Surg. 2014;77: 974Y977. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level IV. KEY WORDS: Attribution; complications; accountable care organizations. BACKGROUND:
A
ttribution is the process of assigning a complication to a treating physician in the health care system. At our trauma center, the policy has been to attribute all complications for trauma patients to the admitting physician. This policy assumes that the admitting physician delivers the majority of care. As such, he or she is likely responsible for procedures as well as complications during a hospitalization. The team approach to the delivery of health care has brought about new reimbursement models. Accountable care organizations purportedly seek to collectively save on costs through quality and benchmarking.1 These programs have been developed in parallel with a focus on increased transparency with the public in regard to outcomes.2 A key aspect in the evolution of health care reporting is the reporting of individual physician data. For this to occur, there must be a mechanism for the attribution of complications to a specific physician.3 The model of patient care is changing in the United States. The traditional practice of being admitted to the hospital by one’s primary care provider has been supplanted by the Submitted: January 15, 2014, Revised: March 17, 2014, Accepted: March 20, 2014, Published online: July 21, 2014. From the Department of Surgery, Methodist Dallas Medical Center, Dallas, Texas. This study was presented at 44th Annual Meeting of the Western Trauma Association, March 2Y7, 2014, in Steamboat Springs, Colorado. Address for reprints: Jason Murry, MD, Medical Education Department, Methodist Dallas Medical Center, 1441 N. Beckley, Dallas, TX 75204; email: [email protected]. DOI: 10.1097/TA.0000000000000344
974
hospitalist model. The surgical community is seeing a similar trend.4 Acute care surgeons (ACSs) are hospital-based surgeons who typically cover trauma and emergent surgical needs. Appropriate consultations are made to other teams, when needed, with the general surgeon serving as the captain of the ship. The trauma service is the quintessential example of a team approach to health care delivery. Multiple services are typically involved in the care of a single trauma patient. This makes the attribution of complications especially difficult. It is our hypothesis that a significant number of complications have been attributed to the ACS that more appropriately belong to other physicians. As a result, we created a physician-led process to determine attribution. We then compared our attribution results with those of the existing hospital policy and critically evaluated discrepancies.
PATIENTS AND METHODS After institutional review board approval, all patients admitted to the trauma service during a 12-month period were identified. Patients with single-system trauma who were admitted and had care provided solely by the admitting service were excluded. The remaining patients were evaluated for any complication experienced during the index admission or a subsequent admission that could be related to their trauma. For those patients with complications, we evaluated the following: mechanism of injury, age, race, sex, Injury Severity Score J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Murry et al.
(ISS), and number of readmissions. Consulting services for each patient were also noted. The trauma director and the medical director (TD/MD) of our accountable care organizations reviewed all complications and attributed them to the ‘‘appropriate’’ health care provider. The results of this were then compared with the attribution data report based on the existing hospital policy.
RESULTS There were 1,526 trauma admissions. After exclusions, 1,019 patients were reviewed. These patients were admitted to the acute care surgery service with subspecialty input when appropriate. Complications occurred in 7.2% (73 of 1,019) of the cohort and accounted for 125 total complications (1.7 complications per patient). Mechanism of injury in these patients was overwhelmingly blunt. Falls accounted for 41% (30 of 73), and motor vehicle collisions accounted for 43% (32 of 73) of the patients. The remaining injuries included gunshot wounds (4 of 73), assault (1 of 73), pedestrianYmotor vehicle collisions (5 of 73), and a stab wound (1 of 73). The mean age was 48.7 years, and the ISS was 21. The mean length of stay for all patients was 16.4 days, with a mean intensive care unit length of stay of 8.6 days. There were a total of seven unexpected readmissions during the study period. The cohort was composed of approximately equal numbers of white (34%), African American (33%), Hispanic (30%), and Asian (3%) patients. Most of the patients were male, 63% (46 of 73). With the use of the traditional definition, the acute care surgery service was assigned all 125 complications. Using the TD/MD method, the neurosurgical attending accounted for 36% (45 of 125) of complications. The acute care surgery attending was deemed responsible for 34% (43 of 125) of complications, and orthopedic surgery was identified as the causative factor in 22% (27 of 125). The remaining 8% (10 of 125) were attributed to various other services. Urinary tract infections were the most frequent complication, with 22 identified. The neurosurgery service was attributed 41% (9 of 22), while 32% (7 of 22) went to the acute care surgery service, and 18% (4 of 22) went to orthopedics. Sacral decubitus ulcers were the next most common, with 15 noted during the study period. Neurosurgery was identified as accountable in 53% (8 of 15); and orthopedics, in 27% (4 of 15). The acute care surgery service was attributed 13% (2 of 15), and plastic surgery was identified as the attributable party in 7% (1 of 15). There were a total of 17 wound infections. Wound infections were attributed to neurosurgery in 41% (7 of 17), orthopedics in 35% (6 of 17), plastics in 18% (3 of 17), and acute care surgery in 6% (1 of 17). Seven patients had unexpected readmissions. Most (5 of 7) of these were related to orthopedics. Table 1 demonstrates a breakdown of all complications as attributed by the TD/MD.
DISCUSSION Physicians work together as a team in the delivery of health care. Despite providing appropriate care, it is inevitable that complications are going to occur. Traditionally, these have been discussed in a morbidity and mortality conference, with a
goal to identify opportunities for improvement. While these conferences still occur, they are provincial and do not necessarily add to the collective consciousness of health care. As a result, lessons learned in one hospital are not necessarily shared with others. Changes in reimbursement have focused on several hospital-acquired conditions that shift costs to the hospital and away from government payers. This has motivated hospitals to be more proactive in preventing complications since they may decrease reimbursement. Hospitals, therefore, are increasingly interested in identifying complications and the physicians who are responsible for a disproportionate number of them. Surgeons have been leaders in performance improvement for many years, and the attribution of complications is not a new concept.5 What has changed is the potential effect on physicians in the future when it comes to professional viability, reimbursement, and public reporting. Two thirds of all complications in our trauma cohort were potentially misattributed to the ACS under the existing model. This can have deleterious effects on a practicing physician for several reasons. For a surgeon who needs to relocate, there may be difficulty obtaining privileges as a high rate of complications could be seen as an unacceptable risk to a new hospital. In addition, hospitals may be in a position to restrict a physician’s privileges if certain procedures are deemed too high risk. This places the hospital, not a physician governance, in a position to severely limit a physician’s scope of practice. Similar to the issue of economic credentialing in the past, this could lead to complication-based credentialing in the future. In addition, this may negatively affect the already dire issue of emergency call coverage. Emergency operations typically carry a higher complication rate than do elective cases.6 Therefore, physicians may avoid call responsibilities to preserve a desirable complication profile. This could lead to a decrease in the quantity and, potentially, the quality of the workforce caring for the injured/severely ill patient. For referral-based practices, there is a risk that patients seeking surgical care will look at public reporting and avoid the surgeons who cover trauma call as their rate of complications may be artificially elevated. This risk exists not only for acute care surgery but also for each of the surgical specialties that are involved in the care of these complex patients. This could exacerbate the subspecialist call shortage as well, particularly in disciplines such as neurosurgery or orthopedics. Patient access to excellent surgical care could suffer, as a result, in both the elective and emergent settings. In addition, insurance companies could impose an outcomes requirement to qualify as an in-network physician. This could significantly affect access to care and patient referral patterns. The loss of insurance contracts from an abnormally high rate of misattributed complications would be devastating to a practice. The conversation regarding who is responsible for a complication needs to be dynamic. There are many areas where the line will be clearly delineated. Several examples were encountered in this review. A trauma patient who develops a wound and hardware infection after an open reduction internal fixation should be attributed to the orthopedist and not the ACS, while a subsequent anastomotic leak from a small bowel resection is clearly the responsibility of the ACS. In contrast, a
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
975
J Trauma Acute Care Surg Volume 77, Number 6
Murry et al.
TABLE 1. Complications per Treating Service Responsible Complication AKI ARDS CPR Decubitus ulcer Deep surgical site infection Drug/EtOH withdrawal DVT Extremity compartment syndrome Graft/flap failure MI Surgical space infection Pneumonia PE Stroke/CVA Superficial SSI Unplanned intubation UTI CLABSI Osteomyelitis Unplanned return to OR Unplanned return to ICU Severe sepsis Pneumothorax* Totals
Acute Care Surgery
Neurosurgery
Orthopedics
Plastics
Urology
Total
6 3 1 2 0 1 2 0 0 0 2 8 2 0 0 2 7 3 0 0 0 3 1 43
2 2 2 8 1 3 2 0 0 0 0 6 0 2 5 1 9 0 0 0 1 1 0 45
1 0 0 4 0 0 1 1 1 1 0 1 0 0 3 2 4 0 2 4 1 1 0 27
0 0 0 1 0 0 0 0 0 1 2 0 0 0 0 1 2 0 0 0 0 1 0 8
0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 2
9 5 3 15 1 4 5 1 1 2 5 15 2 2 8 6 22 3 2 5 2 6 1 125
*Not listed as reportable complication for the Committee on Trauma. AKI, Acute Kidney Injury; ARDS, adult respiratory distress syndrome; CLABSI, Central Line Associated Bloodstream Infection; CVA, Cerebrovascular Accident; DVT, Deep Vein Thrombosis; EtOH, Alcohol; ICU, intensive care unit; MI, myocardial Infarction; OR, Operating Room; PE, Pulmonary Embolus; SSI, Surgical Site Infection; UTI, Urinary Tract Infection.
patient with a spinal fracture and paraplegia who develops a decubitus ulcer is not so clear cut. Is this a nursing complication? Or should it be ascribed to the service whose injury most closely contributed to the complication? Such a system could unjustly punish a service for having complications that are not preventable or readily treatable. It could also penalize specific service lines who have patients who are more prone to hospital complications. What about physician extenders? Are they responsible for their own complications, or do they fall under the purview of their supervising physician? The attribution of complications in health care raises many questions that are beyond the scope of this article. The lessons learned from trauma performance improvement may be used as a model for the attribution of complications in general. There are times when an individual physician is not responsible for a complication. Perhaps, the system should include a mechanism to assign complications to a hospital deficiency. If a patient requires medication, equipment, or expertise to avoid a complication but a hospital does not provide the needed resource, there should be accountability at the system level. Patients should also be expected to take an active role in their own well-being. To attribute a pulmonary embolism to a physician when the patient did not take his or her prescribed 976
anticoagulant is misguided. In addition, modifiable risk factors that place a patient at increased risk for complications should not fall at the feet of the physicians who care for them. Patients are discharged with instructions to abstain from tobacco or illicit drugs. If they ignore this directive, it may have an adverse effect on their outcome.7 The physician’s ability to effect change in the life of his or her patients can extend only so far. The attribution of complications will need to take these and other situations into account. Discussions with all services that care for these patients to determine the ultimate outcome will be necessary. There will always be some degree of disagreement, and an adjudication process will be necessary. To our knowledge, few hospitals have addressed this issue. Clearly, this should be a physicianled process and should be formally defined within the medical staff bylaws. The mechanisms by which this will be executed are not entirely clear and may vary from hospital to hospital. Ideally, there would be a universal process that codifies the data to allow comparison both regionally and nationally. There are several limitations to this study. We evaluated only trauma patients, but this will affect the entire spectrum of providers/patients. Trauma surgery provides a unique framework from which to evaluate this issue. Multiple services are involved, and most issues are acute conditions that are readily treated. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Murry et al.
Defining complications can, in and of itself, be a difficult task. Before the adjudication of responsibility for complications, we must define what constitutes a complication. Finally, the process we used to determine attribution may be flawed.
CONCLUSION The attribution of complications is not new. The policy of assigning all trauma complications to the admitting ACS resulted in misattribution 66% of the time. The fallout of incorrectly assigning complications is significant. This could have severe unintended consequences to the health care system. A physician-led, multidisciplinary solution is necessary and should be applied uniformly across the country. AUTHORSHIP M.T. and E.D. designed this study. J.M. and G.H. conducted the literature search. J.M. performed data collection. J.M., G.H., N.P., and P.R. contributed to data analysis. J.M., N.P., M.T., and E.D. interpreted the data. J.M., G.H., and P.R. wrote the manuscript, which P.R., M.T., and E.D. critically revised.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Berwick DM. Making good on ACO’s promiseVthe final rule for Medicare shared savings program. N Engl J Med. 2011;365(19):1753Y1756. 2. Harrington R, Coffin J, Chauhan B. Understanding how the Physician Quality Reporting System affects primary care physicians. J Med Pract Manage. 2013;28(4):248Y250. 3. Weissler M. ACS perspective: public reporting of surgical outcomes. ACS Surg News. 2013;9(11):2. 4. Charles AG, Ortiz-Pujols S, Ricketts T, Fraher E, Neuwahl S, Cairns B, Sheldon GF. The employed surgeon: a changing professional paradigm. JAMA Surg. 2013;148(4):323Y328. 5. Orlander J, Barber T, Fincke B. The morbidity and mortality conference: the delicate nature of learning from error. Acad Med. 2002;77(10):1001Y1006. 6. Becher RD, Hoth JJ, Miller PR, et al. A critical assessment of outcomes in emergency versus nonemergency general surgery using the American College of Surgeons National Surgical Quality Improvement Program database. Am Surg. 2011;77(7):951Y959. 7. Sørensen LT. Wound healing and infection in surgery: the pathophysiological impact of smoking, smoking cessation, and nicotine replacement therapy: a systematic review. Ann Surg. 2012;255(6):1069Y1079.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
977
Attribution: Whose complication is it? Jason Murry, MD, Greg Hambright, MD, Nimesh Patel, MD, Peter Rappa, MD, Michael Truitt, MD, and Ernest Dunn, MD, Dallas, Texas
To improve quality, programs such as accountable care organizations need to determine the part of the health care system most ‘‘responsible’’ for a complication. This is referred to as attribution. This provides a framework to compare physicians for patients and third-party payers. Traditionally, the attribution of complications has been to the admitting physician. This may misidentify the physician ‘‘responsible’’ for the complication. This is especially difficult in trauma patients who have multiple providers. We hypothesized that the current mechanism for attributing complications in trauma patients is inadequate and will need to be modernized. METHODS: All trauma admissions during a 12-month period were reviewed. Patients with single-system trauma were excluded. We reviewed our trauma database for mechanism of injury, complications, and readmissions. The trauma director and the medical director of our accountable care organizations reviewed all complications and attributed them to the appropriate health care provider. These were compared with the hospital decisions using the traditional definition. RESULTS: The trauma service had 1,526 admissions. After exclusions, 1,019 patients were reviewed. One hundred twenty-five complications occurred in 73 patients. Using the traditional definition, the acute care surgery service was assigned all 125 complications. Using the trauma director and medical director method, the neurosurgical attending accounted for 36% (45 of 125) of complications. The acute care surgery attending was responsible for 34% (43 of 125) of complications, and orthopedic surgery was identified as the causative factor in 22% (27 of 125). The remaining 8% (10 of 125) were attributed to various other services. Seven patients had unexpected readmissions. Most (6 of 7) of these were related to orthopedics. CONCLUSION: Hospital complications are now being assigned to individual surgeons. Which physician is responsible for each complication will be a controversial matter. Without a critical review process with physician input, up to two thirds of complications could be attributed incorrectly. The attribution process needs to be refined. (J Trauma Acute Care Surg. 2014;77: 974Y977. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level IV. KEY WORDS: Attribution; complications; accountable care organizations. BACKGROUND:
A
ttribution is the process of assigning a complication to a treating physician in the health care system. At our trauma center, the policy has been to attribute all complications for trauma patients to the admitting physician. This policy assumes that the admitting physician delivers the majority of care. As such, he or she is likely responsible for procedures as well as complications during a hospitalization. The team approach to the delivery of health care has brought about new reimbursement models. Accountable care organizations purportedly seek to collectively save on costs through quality and benchmarking.1 These programs have been developed in parallel with a focus on increased transparency with the public in regard to outcomes.2 A key aspect in the evolution of health care reporting is the reporting of individual physician data. For this to occur, there must be a mechanism for the attribution of complications to a specific physician.3 The model of patient care is changing in the United States. The traditional practice of being admitted to the hospital by one’s primary care provider has been supplanted by the Submitted: January 15, 2014, Revised: March 17, 2014, Accepted: March 20, 2014, Published online: July 21, 2014. From the Department of Surgery, Methodist Dallas Medical Center, Dallas, Texas. This study was presented at 44th Annual Meeting of the Western Trauma Association, March 2Y7, 2014, in Steamboat Springs, Colorado. Address for reprints: Jason Murry, MD, Medical Education Department, Methodist Dallas Medical Center, 1441 N. Beckley, Dallas, TX 75204; email: [email protected]. DOI: 10.1097/TA.0000000000000344
974
hospitalist model. The surgical community is seeing a similar trend.4 Acute care surgeons (ACSs) are hospital-based surgeons who typically cover trauma and emergent surgical needs. Appropriate consultations are made to other teams, when needed, with the general surgeon serving as the captain of the ship. The trauma service is the quintessential example of a team approach to health care delivery. Multiple services are typically involved in the care of a single trauma patient. This makes the attribution of complications especially difficult. It is our hypothesis that a significant number of complications have been attributed to the ACS that more appropriately belong to other physicians. As a result, we created a physician-led process to determine attribution. We then compared our attribution results with those of the existing hospital policy and critically evaluated discrepancies.
PATIENTS AND METHODS After institutional review board approval, all patients admitted to the trauma service during a 12-month period were identified. Patients with single-system trauma who were admitted and had care provided solely by the admitting service were excluded. The remaining patients were evaluated for any complication experienced during the index admission or a subsequent admission that could be related to their trauma. For those patients with complications, we evaluated the following: mechanism of injury, age, race, sex, Injury Severity Score J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Murry et al.
(ISS), and number of readmissions. Consulting services for each patient were also noted. The trauma director and the medical director (TD/MD) of our accountable care organizations reviewed all complications and attributed them to the ‘‘appropriate’’ health care provider. The results of this were then compared with the attribution data report based on the existing hospital policy.
RESULTS There were 1,526 trauma admissions. After exclusions, 1,019 patients were reviewed. These patients were admitted to the acute care surgery service with subspecialty input when appropriate. Complications occurred in 7.2% (73 of 1,019) of the cohort and accounted for 125 total complications (1.7 complications per patient). Mechanism of injury in these patients was overwhelmingly blunt. Falls accounted for 41% (30 of 73), and motor vehicle collisions accounted for 43% (32 of 73) of the patients. The remaining injuries included gunshot wounds (4 of 73), assault (1 of 73), pedestrianYmotor vehicle collisions (5 of 73), and a stab wound (1 of 73). The mean age was 48.7 years, and the ISS was 21. The mean length of stay for all patients was 16.4 days, with a mean intensive care unit length of stay of 8.6 days. There were a total of seven unexpected readmissions during the study period. The cohort was composed of approximately equal numbers of white (34%), African American (33%), Hispanic (30%), and Asian (3%) patients. Most of the patients were male, 63% (46 of 73). With the use of the traditional definition, the acute care surgery service was assigned all 125 complications. Using the TD/MD method, the neurosurgical attending accounted for 36% (45 of 125) of complications. The acute care surgery attending was deemed responsible for 34% (43 of 125) of complications, and orthopedic surgery was identified as the causative factor in 22% (27 of 125). The remaining 8% (10 of 125) were attributed to various other services. Urinary tract infections were the most frequent complication, with 22 identified. The neurosurgery service was attributed 41% (9 of 22), while 32% (7 of 22) went to the acute care surgery service, and 18% (4 of 22) went to orthopedics. Sacral decubitus ulcers were the next most common, with 15 noted during the study period. Neurosurgery was identified as accountable in 53% (8 of 15); and orthopedics, in 27% (4 of 15). The acute care surgery service was attributed 13% (2 of 15), and plastic surgery was identified as the attributable party in 7% (1 of 15). There were a total of 17 wound infections. Wound infections were attributed to neurosurgery in 41% (7 of 17), orthopedics in 35% (6 of 17), plastics in 18% (3 of 17), and acute care surgery in 6% (1 of 17). Seven patients had unexpected readmissions. Most (5 of 7) of these were related to orthopedics. Table 1 demonstrates a breakdown of all complications as attributed by the TD/MD.
DISCUSSION Physicians work together as a team in the delivery of health care. Despite providing appropriate care, it is inevitable that complications are going to occur. Traditionally, these have been discussed in a morbidity and mortality conference, with a
goal to identify opportunities for improvement. While these conferences still occur, they are provincial and do not necessarily add to the collective consciousness of health care. As a result, lessons learned in one hospital are not necessarily shared with others. Changes in reimbursement have focused on several hospital-acquired conditions that shift costs to the hospital and away from government payers. This has motivated hospitals to be more proactive in preventing complications since they may decrease reimbursement. Hospitals, therefore, are increasingly interested in identifying complications and the physicians who are responsible for a disproportionate number of them. Surgeons have been leaders in performance improvement for many years, and the attribution of complications is not a new concept.5 What has changed is the potential effect on physicians in the future when it comes to professional viability, reimbursement, and public reporting. Two thirds of all complications in our trauma cohort were potentially misattributed to the ACS under the existing model. This can have deleterious effects on a practicing physician for several reasons. For a surgeon who needs to relocate, there may be difficulty obtaining privileges as a high rate of complications could be seen as an unacceptable risk to a new hospital. In addition, hospitals may be in a position to restrict a physician’s privileges if certain procedures are deemed too high risk. This places the hospital, not a physician governance, in a position to severely limit a physician’s scope of practice. Similar to the issue of economic credentialing in the past, this could lead to complication-based credentialing in the future. In addition, this may negatively affect the already dire issue of emergency call coverage. Emergency operations typically carry a higher complication rate than do elective cases.6 Therefore, physicians may avoid call responsibilities to preserve a desirable complication profile. This could lead to a decrease in the quantity and, potentially, the quality of the workforce caring for the injured/severely ill patient. For referral-based practices, there is a risk that patients seeking surgical care will look at public reporting and avoid the surgeons who cover trauma call as their rate of complications may be artificially elevated. This risk exists not only for acute care surgery but also for each of the surgical specialties that are involved in the care of these complex patients. This could exacerbate the subspecialist call shortage as well, particularly in disciplines such as neurosurgery or orthopedics. Patient access to excellent surgical care could suffer, as a result, in both the elective and emergent settings. In addition, insurance companies could impose an outcomes requirement to qualify as an in-network physician. This could significantly affect access to care and patient referral patterns. The loss of insurance contracts from an abnormally high rate of misattributed complications would be devastating to a practice. The conversation regarding who is responsible for a complication needs to be dynamic. There are many areas where the line will be clearly delineated. Several examples were encountered in this review. A trauma patient who develops a wound and hardware infection after an open reduction internal fixation should be attributed to the orthopedist and not the ACS, while a subsequent anastomotic leak from a small bowel resection is clearly the responsibility of the ACS. In contrast, a
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
975
J Trauma Acute Care Surg Volume 77, Number 6
Murry et al.
TABLE 1. Complications per Treating Service Responsible Complication AKI ARDS CPR Decubitus ulcer Deep surgical site infection Drug/EtOH withdrawal DVT Extremity compartment syndrome Graft/flap failure MI Surgical space infection Pneumonia PE Stroke/CVA Superficial SSI Unplanned intubation UTI CLABSI Osteomyelitis Unplanned return to OR Unplanned return to ICU Severe sepsis Pneumothorax* Totals
Acute Care Surgery
Neurosurgery
Orthopedics
Plastics
Urology
Total
6 3 1 2 0 1 2 0 0 0 2 8 2 0 0 2 7 3 0 0 0 3 1 43
2 2 2 8 1 3 2 0 0 0 0 6 0 2 5 1 9 0 0 0 1 1 0 45
1 0 0 4 0 0 1 1 1 1 0 1 0 0 3 2 4 0 2 4 1 1 0 27
0 0 0 1 0 0 0 0 0 1 2 0 0 0 0 1 2 0 0 0 0 1 0 8
0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 2
9 5 3 15 1 4 5 1 1 2 5 15 2 2 8 6 22 3 2 5 2 6 1 125
*Not listed as reportable complication for the Committee on Trauma. AKI, Acute Kidney Injury; ARDS, adult respiratory distress syndrome; CLABSI, Central Line Associated Bloodstream Infection; CVA, Cerebrovascular Accident; DVT, Deep Vein Thrombosis; EtOH, Alcohol; ICU, intensive care unit; MI, myocardial Infarction; OR, Operating Room; PE, Pulmonary Embolus; SSI, Surgical Site Infection; UTI, Urinary Tract Infection.
patient with a spinal fracture and paraplegia who develops a decubitus ulcer is not so clear cut. Is this a nursing complication? Or should it be ascribed to the service whose injury most closely contributed to the complication? Such a system could unjustly punish a service for having complications that are not preventable or readily treatable. It could also penalize specific service lines who have patients who are more prone to hospital complications. What about physician extenders? Are they responsible for their own complications, or do they fall under the purview of their supervising physician? The attribution of complications in health care raises many questions that are beyond the scope of this article. The lessons learned from trauma performance improvement may be used as a model for the attribution of complications in general. There are times when an individual physician is not responsible for a complication. Perhaps, the system should include a mechanism to assign complications to a hospital deficiency. If a patient requires medication, equipment, or expertise to avoid a complication but a hospital does not provide the needed resource, there should be accountability at the system level. Patients should also be expected to take an active role in their own well-being. To attribute a pulmonary embolism to a physician when the patient did not take his or her prescribed 976
anticoagulant is misguided. In addition, modifiable risk factors that place a patient at increased risk for complications should not fall at the feet of the physicians who care for them. Patients are discharged with instructions to abstain from tobacco or illicit drugs. If they ignore this directive, it may have an adverse effect on their outcome.7 The physician’s ability to effect change in the life of his or her patients can extend only so far. The attribution of complications will need to take these and other situations into account. Discussions with all services that care for these patients to determine the ultimate outcome will be necessary. There will always be some degree of disagreement, and an adjudication process will be necessary. To our knowledge, few hospitals have addressed this issue. Clearly, this should be a physicianled process and should be formally defined within the medical staff bylaws. The mechanisms by which this will be executed are not entirely clear and may vary from hospital to hospital. Ideally, there would be a universal process that codifies the data to allow comparison both regionally and nationally. There are several limitations to this study. We evaluated only trauma patients, but this will affect the entire spectrum of providers/patients. Trauma surgery provides a unique framework from which to evaluate this issue. Multiple services are involved, and most issues are acute conditions that are readily treated. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Murry et al.
Defining complications can, in and of itself, be a difficult task. Before the adjudication of responsibility for complications, we must define what constitutes a complication. Finally, the process we used to determine attribution may be flawed.
CONCLUSION The attribution of complications is not new. The policy of assigning all trauma complications to the admitting ACS resulted in misattribution 66% of the time. The fallout of incorrectly assigning complications is significant. This could have severe unintended consequences to the health care system. A physician-led, multidisciplinary solution is necessary and should be applied uniformly across the country. AUTHORSHIP M.T. and E.D. designed this study. J.M. and G.H. conducted the literature search. J.M. performed data collection. J.M., G.H., N.P., and P.R. contributed to data analysis. J.M., N.P., M.T., and E.D. interpreted the data. J.M., G.H., and P.R. wrote the manuscript, which P.R., M.T., and E.D. critically revised.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Berwick DM. Making good on ACO’s promiseVthe final rule for Medicare shared savings program. N Engl J Med. 2011;365(19):1753Y1756. 2. Harrington R, Coffin J, Chauhan B. Understanding how the Physician Quality Reporting System affects primary care physicians. J Med Pract Manage. 2013;28(4):248Y250. 3. Weissler M. ACS perspective: public reporting of surgical outcomes. ACS Surg News. 2013;9(11):2. 4. Charles AG, Ortiz-Pujols S, Ricketts T, Fraher E, Neuwahl S, Cairns B, Sheldon GF. The employed surgeon: a changing professional paradigm. JAMA Surg. 2013;148(4):323Y328. 5. Orlander J, Barber T, Fincke B. The morbidity and mortality conference: the delicate nature of learning from error. Acad Med. 2002;77(10):1001Y1006. 6. Becher RD, Hoth JJ, Miller PR, et al. A critical assessment of outcomes in emergency versus nonemergency general surgery using the American College of Surgeons National Surgical Quality Improvement Program database. Am Surg. 2011;77(7):951Y959. 7. Sørensen LT. Wound healing and infection in surgery: the pathophysiological impact of smoking, smoking cessation, and nicotine replacement therapy: a systematic review. Ann Surg. 2012;255(6):1069Y1079.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
977
