RESEARCH
Plastic Surgery Mortality An 11-Year, Single-Institution Experience Steven M. Levine, MD, Stelios C. Wilson, MD, Sammy Sinno, MD, Jason Rothwax, BA, Alexander Dillon, BS, and Pierre B. Saadeh, MD, FACS Background: Systematic reporting of mortality data is lacking in many surgical fields including plastic surgery. Current plastic surgery literature is largely limited to adverse events associated with specific procedures. Without mortality data, it is unclear how the recent growth of patient safety initiatives can rationally impact outcomes. Methods: We evaluated 11 years of patient outcome data collected prospectively and updated monthly by our department. Paper records were entered into a Health Insurance Portability and Accountability Act-compliant digital database capable of prospectively maintaining future data. Data were reviewed for 5 surgical services in 4 different hospitals that comprise our department's activity. Results: Between 2000 and 2011, a total of 60,834 cases were performed. In this time, a total of 829 (1.4%) negative outcome reports were identified. Of these, a total of 25 (0.04%) cases had an outcome of death (24) or brain death (1). Deaths were either directly or indirectly associated with cardiopulmonary causes, multisystem organ failure, sepsis, massive bleeding, CVA, saddle embolism, or unknown causes. Conclusions: This study is the largest reported series of cases performed by a single academic plastic surgery service to report overall mortality data.
the United States.2 This study reported specific death rates for certain “high-risk” procedures including cardiovascular procedures, cancer procedures, above-the-knee amputations, and exploratory laparotomies.2 Still, the systematic reporting of mortality data is lacking in many surgical fields. A review of the plastic surgery literature demonstrates minimal information is available regarding total patient mortality rates and causes of death. Current plastic surgery literature is largely limited to adverse events associated with specific types of procedures. Notably, some institutions have sought to increase the transparency of mortality rates of craniofacial surgery,3,4 sternal wound repair,5,6 head and neck reconstruction,7,8 burn reconstruction,9 pressure ulcer management,10 abdominoplasty,11 cleft lip and cleft palate repair,12 and office-based plastic surgery.13 Our study is unique in the fact that it includes all deaths that were either directly or indirectly related to any operative procedure performed by a single academic institution's plastic surgery department. The goal of this study was to increase transparency of patient death at our institution and continue our ongoing efforts to decrease patient mortality across the field of plastic surgery.
Key Words: plastic surgery, death, mortality, morbidity
METHODS
(Ann Plast Surg 2016;76: 556–558)
“To rectify past blunders is impossible, but we might profit by the experience of them.” —George Washington
S
ystematic reporting of surgical mortality data is lacking in many surgical fields including plastic surgery. Despite the widespread presence of intradepartmental conferences and discourse, little attention has been paid within the literature to analyze departmental-wide mortality. Without these data, it is unclear how the recent growth of patient safety initiatives has impacted, or will impact, patient outcomes. There have been some gains in transparency in the field of general surgery. Specifically, a study conducted in the Department of Surgery at the Massachusetts General Hospital that reviewed morbidity and mortality conference data reported a complication rate of 6.4% between July 1, 2002, and June 30, 2003. This study also reported a total of 53 deaths (0.9% of their total 5905 cases) during the 1-year period and also included data about procedures, diagnosis, and other pertinent information regarding the death of their patients.1 In addition, a different study using the National Inpatient Sample from 2006 found an overall mortality rate of 1.32% within the first 30 days after any surgery in
Received June 25, 2014, and accepted for publication, after revision, September 13, 2014. From the Institute of Reconstructive Plastic Surgery, New York University Langone Medical Center, New York, NY. Conflicts of interest and sources of funding: none declared. Reprints: Pierre B. Saadeh, MD, FACS, Institute of Reconstructive Plastic Surgery, New York University, 305 East 33rd St, Lower Level, New York, NY 10016. E-mail: [email protected]. Steven M. Levine and Stelios C. Wilson contributed equally to this work. Copyright © 2014 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/16/7605–0556 DOI: 10.1097/SAP.0000000000000369
556
www.annalsplasticsurgery.com
We evaluated 11 years of patient outcome data, collected prospectively and updated monthly by our department—the Institute of Reconstructive Plastic Surgery at the New York University Langone Medical Center. Complications and deaths are identified and reported by residents and faculty so that teachable points can be discussed. Once a case has been identified, the Chief Resident is responsible for compiling the information in a preformatted document and submitting it to the department. Residents and attending physicians are required to report all possible complications and deaths. Paper records from 2000 to 2011 were entered into a Health Insurance Portability and Accountability Act-compliant digital database capable of prospectively maintaining future data. Data were reviewed for 5 surgical services (2 private, 2 public, and 1 VA) located in 4 different hospitals that comprise our department's activity. Occurrence rates were calculated using the total number of submissions as the numerator and the total number of cases performed by our department as the denominator. Mortality rate was calculated in a similar fashion. Postoperative time to death was defined as the number of days between the date of death and the date of a patient's last operative procedure.
RESULTS Between 2000 and 2011, a total of 60,834 were performed. In this time, a total of 829 negative outcome reports were identified (1.4%). Of these, a total of 25 (0.04%) cases had an end outcome of death (24) or brain death (1). Deaths occurred in every year but 2002. The mean age of patients in this study was 56.3 years (0.75– 93 years); 44% (11) were men and 56% (14) were women. Two patients in the study had undergone recent trauma. Of the 25 deaths, all but 2 of the patients were managed by multiple surgical departments. Plastic surgery operative goals included sternal wound closure (6), mandibular reconstruction (5), cranial vault remodeling (3), hand/foot amputation (2), lower extremity wound Annals of Plastic Surgery • Volume 76, Number 5, May 2016
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Annals of Plastic Surgery • Volume 76, Number 5, May 2016
closure (2), panniculectomy (2), rhytidectomy (1), excision of carcinoma (1), sacral wound closure (1), glossal reconstruction (1), and abdominal closure (1; Table 1). Five cases used microsurgical free flaps and an additional 6 cases used local flaps. Seven patients died while still in the OR, 17 patients died in the hospital, and a single patient died at home. Advanced cardiovascular life support was called for 10 patients and 10 patients were made do not resuscitate/ do not intubate at some point before the time of death. Deaths occurred an average of 10.2 days (0–47 days) after the patients' final operative procedures. Deaths were either directly or indirectly associated with cardiopulmonary causes (6), multisystem organ failure (7), sepsis (3), massive bleeding (3), cerebral vascular accident (CVA) (3), saddle embolism (1), or unknown causes (2). Regarding the 2 patients where the causes of death were unknown, no autopsies were performed at the request of the patients' families.
DISCUSSION Our data demonstrate that deaths occurred in individuals with varying ages, preoperative diagnoses, and intraoperative and postoperative complications. However, more than half of patients who died in our study had undergone one of the following 3 types of procedures: sternal wound closure, mandibular reconstruction, and cranial vault remodeling. A total of 6 (24%) patients in our study died after sternal wound debridement and coverage. Plastic surgeons often care for patients with sternal wounds to assist with complications of wound healing and infection. Current treatment modalities for the management of sternal wound infection include surgical debridement, vacuum-assisted closure therapy, flap coverage, and sternal plating.14 Our management included sternal debridement and attempted wound closure for all 6 patients as well as 3 local pectoral flaps and 1 rectus abdominis flap. Four of these patients had a coronary artery bypass graft (CABG) before their presentation. Complications are common in this patient population. Patients undergoing a median sternotomy for CABG have been found to have the highest rate of sternal wound infection when compared to other
Mortality in Plastic Surgery
sternal surgeries.14 Further, infection is a major cause of morbidity and mortality for these patients. Specifically, the mortality rate of superficial sternal wound infections ranges in the literature between 0.5% and 9%,15,16 whereas the mortality rate of deep sternal wound infection ranges in the literature between 10% and 47%.5,6 Of note, there were 2 patients with sternal wounds that did not have CABG procedures before presentation. One patient had a sternal wound from a thymectomy (and adjuvant radiation therapy) performed in a different country and ultimately died of massive bleeding from her aorta. The final patient had a sternal wound after an aortic arch repair but experienced a CVA before the wound closure procedure. The family eventually made this patient do not resuscitate and the patient died. Although our institution experienced 6 deaths related to sternal wound complications between 2004 and 2006, we have yet to experience a death associated with a sternal wound in over 5 years. A total of 5 (20%) patients in our study died after mandibular reconstruction. Four of these patients required a fibula osteocutaneous free flap (3 for squamous cell carcinoma and 1 for multiple myeloma). The average age of these 4 patients was nearly 63 years (range, 60–67 years). Two of the 4 patients had a documented history of alcohol dependence. A history of alcoholism not only increases the risk of head and neck cancer but is also associated with a high rate of postoperative mortality.8,17 Specifically, a study by Kao et al8 found that that mortality rates can range from 4.8% to 30% depending on the severity of liver disease for patients undergoing surgical resection of head and neck cancer followed by microsurgical free tissue transfer. Although we did find that one of the microsurgical flaps failed, it was later found that the failure was due to a blowout of the free flap's arterial supply secondary to an infection. All other flaps were viable until the patients died. One of the aforementioned 4 patients died from sepsis, whereas the other 3 died secondary to multisystem organ failure or cardiac arrest. These data support the findings of Jones et al7 who published a 1% mortality rate for their 100 patient consecutive case series of head and neck microsurgical reconstruction. The authors noted that medical complications, not microsurgical complications, comprise most morbidity, mortality, and overall cost of health care related to these procedures.7 Of note, one additional patient died after mandibular reconstruction. After
TABLE 1. Plastic Surgery Mortality During a Consecutive 11-Year Period at our Institution Plastic Surgery Operative Goals
No. Cases
Age (range), y
Oncologic (n)
Trauma (n)
Single Team (n)/Multi-team (n)
Location of Death
Sternal wound closure
6
71.2 (56–77)
N
N
0/6
Hospital (4) OR (2)
Mandibular reconstruction
5
55–0 (24–67)
Y (4)
Y (1)
1/4
Cranial vault remodeling
3
N
N
0/3
Hospital (3) OR (2) OR (2) Home (1)
Hand/foot amputation
2
66.0 (52–80)
N
N
0/2
Hospital (2)
Lower extremity wound closure
2
55.0 (35–75)
N
Y(1)
0/2
Hospital (2)
Panniculectomy
2
63.5 (53–74)
N
N
0/2
Hospital (2)
Rhytidectomy Excision of carcinoma Sacral wound closure Glossal reconstruction Abdominal reconstruction
1 1 1 1 1
N Y N Y N
N N N N N
1/0 1/0 0/1 0/1 0/1
OR Hospital Hospital Hospital Hospital
9.2 (0.75–27)
54.0 93.0 65.0 41.0 56.0
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
Cause of Death
Multisystem organ failure (3) Intraoperative bleed (2) CVA (1) Multisystem organ failure (3) Sepsis (2) Intraoperative bleed (1) Unknown cerebral event (1) Unknown (1) Cardiac arrest (1) Multisystem organ failure (1) CVA (1) Cardiopulmonary failure (1) Saddle embolism (1) Unknown (1) Cardiac arrest Cardiac arrest Cardiac arrest Respiratory arrest Sepsis
www.annalsplasticsurgery.com
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
557
Annals of Plastic Surgery • Volume 76, Number 5, May 2016
Levine et al
experiencing a gunshot wound to the face, this patient's mandible was repaired via open reduction and internal fixation by our plastic surgery service, but the patient ultimately died from sepsis 1 week later. Given our experience, our department implemented guidelines for vigilant monitoring of infection in patients undergoing mandibular reconstruction. Specifically, at the first sign of local infection, rise in body temperature, rise in white count, or increased secretions, patients are pancultured and empirically treated with broad-spectrum antibiotics against anaerobic bacteria while awaiting culture speciation and sensitivity. The third most common type of procedure resulting in death was cranial vault remodeling. Two patients underwent cranial vault remodeling at 9 months of age for Apert syndrome and Pfeiffer syndrome; both of which are associated with the premature fusion of cranial suture lines (craniosynostosis).18 If left untreated, one third of these patients will go on to develop clinically significant intracranial hypertension including developmental delay.19,20 A study by Czerwinski et al4 reviewed 18 years of craniofacial surgery and found that, of the 8101 major procedures, there were 8 perioperative deaths (
Plastic Surgery Mortality An 11-Year, Single-Institution Experience Steven M. Levine, MD, Stelios C. Wilson, MD, Sammy Sinno, MD, Jason Rothwax, BA, Alexander Dillon, BS, and Pierre B. Saadeh, MD, FACS Background: Systematic reporting of mortality data is lacking in many surgical fields including plastic surgery. Current plastic surgery literature is largely limited to adverse events associated with specific procedures. Without mortality data, it is unclear how the recent growth of patient safety initiatives can rationally impact outcomes. Methods: We evaluated 11 years of patient outcome data collected prospectively and updated monthly by our department. Paper records were entered into a Health Insurance Portability and Accountability Act-compliant digital database capable of prospectively maintaining future data. Data were reviewed for 5 surgical services in 4 different hospitals that comprise our department's activity. Results: Between 2000 and 2011, a total of 60,834 cases were performed. In this time, a total of 829 (1.4%) negative outcome reports were identified. Of these, a total of 25 (0.04%) cases had an outcome of death (24) or brain death (1). Deaths were either directly or indirectly associated with cardiopulmonary causes, multisystem organ failure, sepsis, massive bleeding, CVA, saddle embolism, or unknown causes. Conclusions: This study is the largest reported series of cases performed by a single academic plastic surgery service to report overall mortality data.
the United States.2 This study reported specific death rates for certain “high-risk” procedures including cardiovascular procedures, cancer procedures, above-the-knee amputations, and exploratory laparotomies.2 Still, the systematic reporting of mortality data is lacking in many surgical fields. A review of the plastic surgery literature demonstrates minimal information is available regarding total patient mortality rates and causes of death. Current plastic surgery literature is largely limited to adverse events associated with specific types of procedures. Notably, some institutions have sought to increase the transparency of mortality rates of craniofacial surgery,3,4 sternal wound repair,5,6 head and neck reconstruction,7,8 burn reconstruction,9 pressure ulcer management,10 abdominoplasty,11 cleft lip and cleft palate repair,12 and office-based plastic surgery.13 Our study is unique in the fact that it includes all deaths that were either directly or indirectly related to any operative procedure performed by a single academic institution's plastic surgery department. The goal of this study was to increase transparency of patient death at our institution and continue our ongoing efforts to decrease patient mortality across the field of plastic surgery.
Key Words: plastic surgery, death, mortality, morbidity
METHODS
(Ann Plast Surg 2016;76: 556–558)
“To rectify past blunders is impossible, but we might profit by the experience of them.” —George Washington
S
ystematic reporting of surgical mortality data is lacking in many surgical fields including plastic surgery. Despite the widespread presence of intradepartmental conferences and discourse, little attention has been paid within the literature to analyze departmental-wide mortality. Without these data, it is unclear how the recent growth of patient safety initiatives has impacted, or will impact, patient outcomes. There have been some gains in transparency in the field of general surgery. Specifically, a study conducted in the Department of Surgery at the Massachusetts General Hospital that reviewed morbidity and mortality conference data reported a complication rate of 6.4% between July 1, 2002, and June 30, 2003. This study also reported a total of 53 deaths (0.9% of their total 5905 cases) during the 1-year period and also included data about procedures, diagnosis, and other pertinent information regarding the death of their patients.1 In addition, a different study using the National Inpatient Sample from 2006 found an overall mortality rate of 1.32% within the first 30 days after any surgery in
Received June 25, 2014, and accepted for publication, after revision, September 13, 2014. From the Institute of Reconstructive Plastic Surgery, New York University Langone Medical Center, New York, NY. Conflicts of interest and sources of funding: none declared. Reprints: Pierre B. Saadeh, MD, FACS, Institute of Reconstructive Plastic Surgery, New York University, 305 East 33rd St, Lower Level, New York, NY 10016. E-mail: [email protected]. Steven M. Levine and Stelios C. Wilson contributed equally to this work. Copyright © 2014 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/16/7605–0556 DOI: 10.1097/SAP.0000000000000369
556
www.annalsplasticsurgery.com
We evaluated 11 years of patient outcome data, collected prospectively and updated monthly by our department—the Institute of Reconstructive Plastic Surgery at the New York University Langone Medical Center. Complications and deaths are identified and reported by residents and faculty so that teachable points can be discussed. Once a case has been identified, the Chief Resident is responsible for compiling the information in a preformatted document and submitting it to the department. Residents and attending physicians are required to report all possible complications and deaths. Paper records from 2000 to 2011 were entered into a Health Insurance Portability and Accountability Act-compliant digital database capable of prospectively maintaining future data. Data were reviewed for 5 surgical services (2 private, 2 public, and 1 VA) located in 4 different hospitals that comprise our department's activity. Occurrence rates were calculated using the total number of submissions as the numerator and the total number of cases performed by our department as the denominator. Mortality rate was calculated in a similar fashion. Postoperative time to death was defined as the number of days between the date of death and the date of a patient's last operative procedure.
RESULTS Between 2000 and 2011, a total of 60,834 were performed. In this time, a total of 829 negative outcome reports were identified (1.4%). Of these, a total of 25 (0.04%) cases had an end outcome of death (24) or brain death (1). Deaths occurred in every year but 2002. The mean age of patients in this study was 56.3 years (0.75– 93 years); 44% (11) were men and 56% (14) were women. Two patients in the study had undergone recent trauma. Of the 25 deaths, all but 2 of the patients were managed by multiple surgical departments. Plastic surgery operative goals included sternal wound closure (6), mandibular reconstruction (5), cranial vault remodeling (3), hand/foot amputation (2), lower extremity wound Annals of Plastic Surgery • Volume 76, Number 5, May 2016
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
Annals of Plastic Surgery • Volume 76, Number 5, May 2016
closure (2), panniculectomy (2), rhytidectomy (1), excision of carcinoma (1), sacral wound closure (1), glossal reconstruction (1), and abdominal closure (1; Table 1). Five cases used microsurgical free flaps and an additional 6 cases used local flaps. Seven patients died while still in the OR, 17 patients died in the hospital, and a single patient died at home. Advanced cardiovascular life support was called for 10 patients and 10 patients were made do not resuscitate/ do not intubate at some point before the time of death. Deaths occurred an average of 10.2 days (0–47 days) after the patients' final operative procedures. Deaths were either directly or indirectly associated with cardiopulmonary causes (6), multisystem organ failure (7), sepsis (3), massive bleeding (3), cerebral vascular accident (CVA) (3), saddle embolism (1), or unknown causes (2). Regarding the 2 patients where the causes of death were unknown, no autopsies were performed at the request of the patients' families.
DISCUSSION Our data demonstrate that deaths occurred in individuals with varying ages, preoperative diagnoses, and intraoperative and postoperative complications. However, more than half of patients who died in our study had undergone one of the following 3 types of procedures: sternal wound closure, mandibular reconstruction, and cranial vault remodeling. A total of 6 (24%) patients in our study died after sternal wound debridement and coverage. Plastic surgeons often care for patients with sternal wounds to assist with complications of wound healing and infection. Current treatment modalities for the management of sternal wound infection include surgical debridement, vacuum-assisted closure therapy, flap coverage, and sternal plating.14 Our management included sternal debridement and attempted wound closure for all 6 patients as well as 3 local pectoral flaps and 1 rectus abdominis flap. Four of these patients had a coronary artery bypass graft (CABG) before their presentation. Complications are common in this patient population. Patients undergoing a median sternotomy for CABG have been found to have the highest rate of sternal wound infection when compared to other
Mortality in Plastic Surgery
sternal surgeries.14 Further, infection is a major cause of morbidity and mortality for these patients. Specifically, the mortality rate of superficial sternal wound infections ranges in the literature between 0.5% and 9%,15,16 whereas the mortality rate of deep sternal wound infection ranges in the literature between 10% and 47%.5,6 Of note, there were 2 patients with sternal wounds that did not have CABG procedures before presentation. One patient had a sternal wound from a thymectomy (and adjuvant radiation therapy) performed in a different country and ultimately died of massive bleeding from her aorta. The final patient had a sternal wound after an aortic arch repair but experienced a CVA before the wound closure procedure. The family eventually made this patient do not resuscitate and the patient died. Although our institution experienced 6 deaths related to sternal wound complications between 2004 and 2006, we have yet to experience a death associated with a sternal wound in over 5 years. A total of 5 (20%) patients in our study died after mandibular reconstruction. Four of these patients required a fibula osteocutaneous free flap (3 for squamous cell carcinoma and 1 for multiple myeloma). The average age of these 4 patients was nearly 63 years (range, 60–67 years). Two of the 4 patients had a documented history of alcohol dependence. A history of alcoholism not only increases the risk of head and neck cancer but is also associated with a high rate of postoperative mortality.8,17 Specifically, a study by Kao et al8 found that that mortality rates can range from 4.8% to 30% depending on the severity of liver disease for patients undergoing surgical resection of head and neck cancer followed by microsurgical free tissue transfer. Although we did find that one of the microsurgical flaps failed, it was later found that the failure was due to a blowout of the free flap's arterial supply secondary to an infection. All other flaps were viable until the patients died. One of the aforementioned 4 patients died from sepsis, whereas the other 3 died secondary to multisystem organ failure or cardiac arrest. These data support the findings of Jones et al7 who published a 1% mortality rate for their 100 patient consecutive case series of head and neck microsurgical reconstruction. The authors noted that medical complications, not microsurgical complications, comprise most morbidity, mortality, and overall cost of health care related to these procedures.7 Of note, one additional patient died after mandibular reconstruction. After
TABLE 1. Plastic Surgery Mortality During a Consecutive 11-Year Period at our Institution Plastic Surgery Operative Goals
No. Cases
Age (range), y
Oncologic (n)
Trauma (n)
Single Team (n)/Multi-team (n)
Location of Death
Sternal wound closure
6
71.2 (56–77)
N
N
0/6
Hospital (4) OR (2)
Mandibular reconstruction
5
55–0 (24–67)
Y (4)
Y (1)
1/4
Cranial vault remodeling
3
N
N
0/3
Hospital (3) OR (2) OR (2) Home (1)
Hand/foot amputation
2
66.0 (52–80)
N
N
0/2
Hospital (2)
Lower extremity wound closure
2
55.0 (35–75)
N
Y(1)
0/2
Hospital (2)
Panniculectomy
2
63.5 (53–74)
N
N
0/2
Hospital (2)
Rhytidectomy Excision of carcinoma Sacral wound closure Glossal reconstruction Abdominal reconstruction
1 1 1 1 1
N Y N Y N
N N N N N
1/0 1/0 0/1 0/1 0/1
OR Hospital Hospital Hospital Hospital
9.2 (0.75–27)
54.0 93.0 65.0 41.0 56.0
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
Cause of Death
Multisystem organ failure (3) Intraoperative bleed (2) CVA (1) Multisystem organ failure (3) Sepsis (2) Intraoperative bleed (1) Unknown cerebral event (1) Unknown (1) Cardiac arrest (1) Multisystem organ failure (1) CVA (1) Cardiopulmonary failure (1) Saddle embolism (1) Unknown (1) Cardiac arrest Cardiac arrest Cardiac arrest Respiratory arrest Sepsis
www.annalsplasticsurgery.com
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
557
Annals of Plastic Surgery • Volume 76, Number 5, May 2016
Levine et al
experiencing a gunshot wound to the face, this patient's mandible was repaired via open reduction and internal fixation by our plastic surgery service, but the patient ultimately died from sepsis 1 week later. Given our experience, our department implemented guidelines for vigilant monitoring of infection in patients undergoing mandibular reconstruction. Specifically, at the first sign of local infection, rise in body temperature, rise in white count, or increased secretions, patients are pancultured and empirically treated with broad-spectrum antibiotics against anaerobic bacteria while awaiting culture speciation and sensitivity. The third most common type of procedure resulting in death was cranial vault remodeling. Two patients underwent cranial vault remodeling at 9 months of age for Apert syndrome and Pfeiffer syndrome; both of which are associated with the premature fusion of cranial suture lines (craniosynostosis).18 If left untreated, one third of these patients will go on to develop clinically significant intracranial hypertension including developmental delay.19,20 A study by Czerwinski et al4 reviewed 18 years of craniofacial surgery and found that, of the 8101 major procedures, there were 8 perioperative deaths (
