BREAST Outcomes Article
Lessons Learned from the American College of Surgeons National Surgical Quality Improvement Program Database: Has Centralized Data Collection Improved Immediate Breast Reconstruction Outcomes and Safety? Frederick Wang, M.D. Peter F. Koltz, M.D. Hani Sbitany, M.D. San Francisco, Calif.; and Rochester, N.Y.
Background: The American College of Surgeons National Surgical Quality Improvement Program database was implemented to longitudinally track surgical 30-day surgical outcomes and complications. The authors analyze the programreported outcomes for immediate breast reconstruction from 2007 to 2011, to assess whether longitudinal data collection has improved national outcomes and to highlight areas in need of continued improvement. Methods: The authors reviewed the database from 2007 to 2011 and identified encounters for immediate breast reconstruction using Current Procedural Terminology codes for prosthetic and autologous reconstruction. Demographics and comorbidities were tabulated for all patients. Postoperative complications analyzed included surgical-site infection, wound dehiscence, implant or flap loss, pulmonary embolism, and respiratory infections. Results: A total of 15,978 patients underwent mastectomy and immediate reconstruction. Fewer smokers underwent immediate reconstruction over time (p = 0.126), whereas more obese patients (p = 0.001) and American Society of Anesthesiologists class 3 and 4 patients (p < 0.001) underwent surgery. An overall increase in superficial surgical-site infection was noted, from 1.7 percent to 2.3 percent (p = 0.214). Wound dehiscence (p = 0.036) increased over time, whereas implant loss (p = 0.015) and flap loss (p = 0.012) decreased over time. Mean operative times increased over the analyzed years, as did all complications for prosthetic and autologous reconstruction. Conclusions: The American College of Surgeons National Surgical Quality Improvement Program data set has shown an increase in complications for immediate breast reconstruction over time, because of a longitudinally higher number of comorbid patients and longer operative times. This knowledge allows plastic surgeons the unique opportunity to improve patient selection criteria and efficiency. (Plast. Reconstr. Surg. 134: 859, 2014.) CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, III.
C
urrently in the United States, breast cancer remains the most prevalent cancer diagnosis in women, regardless of race and ethnicity. In 2010, the incidence of new breast cancer diagnosis was 206,966 women.1 In 2012, a total From the Division of Plastic and Reconstructive Surgery, University of California, San Francisco; and the Division of Plastic and Reconstructive Surgery, University of Rochester. Received for publication January 21, 2014; accepted April 7, 2014. Copyright © 2014 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0000000000000615
of 91,655 breast reconstruction procedures were performed in the United States, and the majority of these, 64,575 cases (70 percent of the total), were two-stage expander/implant-based breast reconstructions.2 Given the high volume of reconstructive breast surgery being performed and the clearly defined psychosocial and aesthetic Disclosure: The authors have no financial disclosures to report related to the content of this study. No funding was provided for data collection or writing of this article.
www.PRSJournal.com
859
Plastic and Reconstructive Surgery • November 2014 benefits that breast reconstruction affords to women undergoing mastectomy, reconstructive surgeons must continue to improve techniques and find ways to further minimize morbidity.3–7 The landscape of techniques for mastectomy and breast reconstruction continues to change. The advent of new procedures, such as nipple-sparing mastectomy and total skin-sparing mastectomy, and the increased use of acellular dermal matrices in implant-based breast reconstruction have given the reconstructive surgeon a variety of tools with which to maximize patient satisfaction with aesthetic outcomes.8–10 However, resultant morbidity and outcomes are still being delineated. With changing practices over time, the need to track outcomes is further magnified in the effort to offer our patients the best product. The American College of Surgeons National Surgical Quality Improvement Program database represents a unique data collection program to measure and improve the delivery of quality surgical care.11 With the combined 30-day risk-adjusted surgical outcomes at over 200 hospitals in the United States offered by the database, we have the ability to assess national trends in outcomes for operations such as immediate breast reconstruction.12 Furthermore, the database allows us to track improvement of outcomes over time, which was the intended purpose of this database collection. In this study, we evaluate national trends in immediate breast reconstruction practices and report short-term outcomes and complications. Furthermore, we aim to assess the effect that American College of Surgeons National Surgical Quality Improvement Program data collection has had on improving the national outcomes in immediate breast reconstruction. In this way, we can track the effect of this database on clinical practices in our specialty and make recommendations for future directions of outcomes-based improvement.
PATIENTS AND METHODS American College of Surgeons National Surgical Quality Improvement Program Database We reviewed the 2007 to 2011 American College of Surgeons National Surgical Quality Improvement Program databases and identified encounters for simultaneous mastectomy with breast reconstruction.11 The database was populated by trained research staff at each institution using systematic sampling of operative procedures performed at participating institutions.
860
Each data set contains 240 variables for each case, including patient demographics, baseline comorbidities, preoperative risk factors, intraoperative variables, and 30-day postoperative complications. Patients who were discharged before 30 days were contacted by letter or telephone survey to complete the full 30-day follow-up period.13 We obtained and accessed this database on August 22, 2013. Patient Selection We used 2012 Current Procedural Terminology codes to select cases from the 2006 to 2011 program databases. Implant-based reconstructions included immediate implant (19340), delayed implant (19342), and tissue expander placement (19357). Autologous reconstructions included the latissimus dorsi flap with or without tissue expander or implant (19361), pedicled transverse rectus abdominis myocutaneous (TRAM) flap (19367, 19368, and 19369), and free flap (19364). Immediate reconstructions were defined as cases performed concurrent with mastectomy during the initial hospitalization. Current Procedural Terminology codes for mastectomy included simple mastectomy (19303), subcutaneous mastectomy (19304), radical mastectomy (19305 and 19306), and modified radical mastectomy (19307). Codes for the use of acellular dermal matrix (15170, 15171, 15330, 15331, and 15350) were used to identify patients who had acellular dermal matrix–assisted reconstructions. Patient encounters with a code with both mastectomy and reconstruction were included as immediate reconstructions for the purposes of this study. We further narrowed the cohort down to only clean cases, defined as wound class 1 cases, as an attempt to increase the specificity of our case selection. Classification of Variables We analyzed the 240 variables that were collected in standard fashion in the program’s database and focused on several comorbidities, including obesity (body mass index > 30), diabetes, smoking, chemotherapy within 30 days, and radiotherapy within 90 days. The primary outcome variables for postoperative complications included postoperative superficial surgical-site infection, deep surgical-site infection, organ/ space surgical-site infection, wound dehiscence, return to the operating room, loss of flap or implant-based reconstruction, pulmonary embolism, and postoperative pneumonia.
Volume 134, Number 5 • Improving Breast Reconstruction Safety Statistical Analysis Differences with respect to continuous variables of age, body mass index, albumin, and operative times were analyzed using the t test and reported as mean values. All categorical variables were analyzed using the Pearson chi-square test. Postoperative complications were tracked as dichotomous variables and reported as whole values and percentage of cases in the cohort. A logistic regression model was used to examine the interaction of autologous reconstruction with the three main predictors of smoking, diabetes, and obesity. Risk ratios were obtained for outcomes data and hypotheses were analyzed with the Pearson chi-square test. All data were analyzed using Stata version 13 (StataCorp LP, College Station, Texas). All p values were two-tailed, and values of p < 0.05 were considered to be significant.
RESULTS During the study period, there were 15,978 clean cases of patients undergoing mastectomy with concurrent breast reconstruction using either implant-based reconstruction (i.e., implant or tissue expander placement) or autologous tissue reconstruction (i.e., latissimus flap, TRAM flap, or free flap). The number of cases increased every year over this period, likely because of increasing annual enrollment of hospitals contributing to the database. Overall patient demographics and preoperative comorbidities were summarized by year (Table 1). The racial makeup of patients was mostly Caucasian (88.3 percent), followed by African American (7.9 percent), Asian (3.0 percent), other (0.4 percent), Pacific Islander (0.3 percent), and American Indian (0.2 percent). More nonCaucasians contributed to the cohort over time (Table 1). The mean age of the patients was 51.1 ± 10.4 years, and this remained relatively stable throughout the study period. There was a trend toward a decrease in the percentage of smokers undergoing immediate breast reconstruction from 2007 (13.8 percent) to 2011 (11.8 percent). The mean body mass index of patients was 26.9 ± 6.6 kg/m2, and 4276 patients (26.8 percent) were classified as being obese with a body mass index greater than 30. There were 180 type 1 diabetics (1.1 percent) and 576 type 2 diabetics (3.6 percent) in our patient population. There were 672 patients (4.2 percent) who had undergone chemotherapy within 30 days and 60 patients (0.4 percent) who had undergone radiotherapy within 90 days. The American Society of Anesthesiologists
Classification of Physical Status was also evaluated, and we noted a trend toward increasing percentage of more complex class 3 and 4 patients and a decrease in the percentage of lower class 1 and 2 patients (Table 1). There were 12,373 cases of implant-based reconstruction (77.4 percent) and 2876 (18.0 percent) cases of autologous reconstruction (Table 2). The majority of cases involved tissue expander placement, and there was an increasing trend in these cases from 2007 (64.1 percent) to 2011 (69.3 percent). With regard to autologous reconstruction, there is a decreasing trend in the use of the pedicled TRAM flap from 2007 (12.8 percent) to 2011 (6.2 percent), with an increasing trend in free flap reconstruction from 2007 (3.2 percent) to 2011 (6.4 percent). Intraoperative characteristics of the study cohort are summarized in Table 3. The mean operative time was 205 ± 84 minutes for cases with implant-based reconstruction and 387 ± 197 minutes for autologous reconstruction (p < 0.005, t test). There was a trend of increasing operative times from 2007 to 2011 for both implant-based and autologous reconstructions. The overall incidences of postoperative complications within 30 days for this cohort are summarized in Table 4. Complications analyzed included superficial surgical-site infection (1.9 percent), deep surgical-site infection (1.2 percent), organ/ space surgical-site infection (0.7 percent), wound dehiscence (0.64 percent), unplanned return to the operating room (7.66 percent), loss of either flaps (2.8 percent) or implants (0.8 percent), pulmonary embolism (0.25 percent), and postoperative pneumonia (0.12 percent). There was a trend toward an increase in superficial surgicalsite infections from 2007 (1.7 percent) to 2011 (2.3 percent). There was also a trend toward an increase in wound dehiscence from 2007 (0.3 percent) to 2011 (0.9 percent). The overall rate of implant loss within 30 days was low (0.5 to 1.1 percent per year), and the incidence of flap loss decreased between 2007 (3.9 percent) and 2011 (1.1 percent). The overall incidence of pulmonary embolisms was low, but there was a trend toward a decrease in pulmonary embolisms recorded between 2007 (0.52 percent) and 2011 (0.21 percent). We evaluated postoperative complications with regard to several risk factors, including operative time, type of reconstruction, smoking status, diabetes, obesity, chemotherapy within 30 days, radiotherapy within 90 days, and use of acellular dermal matrix. Increasing mean operative time
861
Plastic and Reconstructive Surgery • November 2014 Table 1. Patient Demographics and Comorbidities for Patients Undergoing Mastectomy and Immediate Breast Reconstruction in the 2007 to 2011 American College of Surgeons National Surgical Quality Improvement Program Data Set per Year* Characteristic No. of cases Race (%) White Black American Indian Pacific Islander Asian Other Mean age ± SD, yr Mean BMI ± SD, kg/m2 No. of obese patients (BMI > 30) Smoking history within 1 yr Alcohol consumption > 2 drinks/day Diabetes Type 1 Type 2 Dyspnea Moderate At rest COPD CHF MI history within 6 mo Prior cardiac stent placement Prior cardiac surgery Hypertension PVD with operative intervention Preoperative HD Prior TIA Prior CVA with defect Prior CVA without defect Disseminated cancer Recent weight loss Mean albumin, mg/dl Bleeding disorder Steroid use Preoperative chemotherapy within 30 days Preoperative irradiation within 90 days ASA physical classification 1 2 3 4
2007 (%)
2008 (%)
2009 (%)
2010 (%)
2011 (%)
1931
2654
3490
3680
4223
1564 (90.1) 115 (6.6) 0 (0) 0 (0) 0 (0) 57 (3.3) 50.7 ± 10.1 26.5 ± 6.2 465 (24.1) 267 (13.8) 20 (1.0)
2119 (89.8) 160 (6.8) 2 (0.1) 3 (0.1) 77 (3.3) 0 (0) 50.7 ± 10.4 26.8 ± 6.4 707 (26.6) 360 (13.6) 30 (1.1)
2806 (89.4) 237 (7.6) 5 (0.2) 12 (0.4) 78 (2.5) 0 (0) 51.1 ± 10.3 26.9 ± 6.8 918 (26.3) 479 (13.7) 43 (1.3)
2879 (87.4) 276 (8.4) 10 (0.3) 10 (0.3) 120 (3.6) 0 (0) 51.5 ± 10.6 27.4 ± 7.2 1072 (29.1) 461 (12.5) 44 (1.2)
15 (0.8) 62 (3.2)
35 (1.3) 81 (3.1)
45 (1.3) 135 (3.9)
41 (1.1) 150 (4.1)
85 (4.4) 3 (0.16) 13 (0.67) 2 (0.10) 0 (0) 15 (0.78) 10 (0.52) 455 (23.6) 3 (0.16) 3 (0.16) 14 (0.73) 2 (0.10) 7 (0.36) 11 (0.57) 6 (0.31) 4.2 ± 0.4 12 (0.62) 22 (1.14) 86 (4.45) 15 (0.78)
108 (4.1) 2 (0.08) 18 (0.68) 1 (0.04) 1 (0.04) 15 (0.57) 10 (0.38) 613 (23.1) 6 (0.23) 3 (0.11) 10 (0.38) 5 (0.19) 9 (0.34) 17 (0.64) 12 (0.45) 4.2 ± 0.4 18 (0.68) 25 (0.94) 102 (3.84) 8 (0.30)
110 (3.2) 1 (0.03) 24 (0.69) 1 (0.03) 0 (0) 16 (0.46) 12 (0.34) 831 (23.8) 3 (0.09) 1 (0.03) 14 (0.40) 13 (0.37) 12 (0.34) 30 (0.86) 14 (0.40) 4.2 ± 0.4 26 (0.74) 23 (0.66) 119 (3.41) 9 (0.26)
134 (3.6) 3 (0.08) 37 (1.01) 0 (0) 0 (0) 26 (0.71) 19 (0.52) 904 (24.6) 4 (0.11) 4 (0.11) 21 (0.57) 9 (0.24) 17 (0.46) 28 (0.76) 13 (0.35) 4.2 ± 0.4 19 (0.52) 30 (0.82) 172 (4.67) 15 (0.41)
207 (10.7) 1412 (73.1) 309 (16.0) 1 (0.05)
289 (10.9) 1896 (71.4) 461 (17.4) 6 (0.23)
368 (10.5) 2441 (70.0) 669 (19.2) 10 (0.29)
329 (8.9) 2575 (70.0) 762 (20.7) 10 (0.27)
p
3328 (86.35)
Lessons Learned from the American College of Surgeons National Surgical Quality Improvement Program Database: Has Centralized Data Collection Improved Immediate Breast Reconstruction Outcomes and Safety? Frederick Wang, M.D. Peter F. Koltz, M.D. Hani Sbitany, M.D. San Francisco, Calif.; and Rochester, N.Y.
Background: The American College of Surgeons National Surgical Quality Improvement Program database was implemented to longitudinally track surgical 30-day surgical outcomes and complications. The authors analyze the programreported outcomes for immediate breast reconstruction from 2007 to 2011, to assess whether longitudinal data collection has improved national outcomes and to highlight areas in need of continued improvement. Methods: The authors reviewed the database from 2007 to 2011 and identified encounters for immediate breast reconstruction using Current Procedural Terminology codes for prosthetic and autologous reconstruction. Demographics and comorbidities were tabulated for all patients. Postoperative complications analyzed included surgical-site infection, wound dehiscence, implant or flap loss, pulmonary embolism, and respiratory infections. Results: A total of 15,978 patients underwent mastectomy and immediate reconstruction. Fewer smokers underwent immediate reconstruction over time (p = 0.126), whereas more obese patients (p = 0.001) and American Society of Anesthesiologists class 3 and 4 patients (p < 0.001) underwent surgery. An overall increase in superficial surgical-site infection was noted, from 1.7 percent to 2.3 percent (p = 0.214). Wound dehiscence (p = 0.036) increased over time, whereas implant loss (p = 0.015) and flap loss (p = 0.012) decreased over time. Mean operative times increased over the analyzed years, as did all complications for prosthetic and autologous reconstruction. Conclusions: The American College of Surgeons National Surgical Quality Improvement Program data set has shown an increase in complications for immediate breast reconstruction over time, because of a longitudinally higher number of comorbid patients and longer operative times. This knowledge allows plastic surgeons the unique opportunity to improve patient selection criteria and efficiency. (Plast. Reconstr. Surg. 134: 859, 2014.) CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, III.
C
urrently in the United States, breast cancer remains the most prevalent cancer diagnosis in women, regardless of race and ethnicity. In 2010, the incidence of new breast cancer diagnosis was 206,966 women.1 In 2012, a total From the Division of Plastic and Reconstructive Surgery, University of California, San Francisco; and the Division of Plastic and Reconstructive Surgery, University of Rochester. Received for publication January 21, 2014; accepted April 7, 2014. Copyright © 2014 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0000000000000615
of 91,655 breast reconstruction procedures were performed in the United States, and the majority of these, 64,575 cases (70 percent of the total), were two-stage expander/implant-based breast reconstructions.2 Given the high volume of reconstructive breast surgery being performed and the clearly defined psychosocial and aesthetic Disclosure: The authors have no financial disclosures to report related to the content of this study. No funding was provided for data collection or writing of this article.
www.PRSJournal.com
859
Plastic and Reconstructive Surgery • November 2014 benefits that breast reconstruction affords to women undergoing mastectomy, reconstructive surgeons must continue to improve techniques and find ways to further minimize morbidity.3–7 The landscape of techniques for mastectomy and breast reconstruction continues to change. The advent of new procedures, such as nipple-sparing mastectomy and total skin-sparing mastectomy, and the increased use of acellular dermal matrices in implant-based breast reconstruction have given the reconstructive surgeon a variety of tools with which to maximize patient satisfaction with aesthetic outcomes.8–10 However, resultant morbidity and outcomes are still being delineated. With changing practices over time, the need to track outcomes is further magnified in the effort to offer our patients the best product. The American College of Surgeons National Surgical Quality Improvement Program database represents a unique data collection program to measure and improve the delivery of quality surgical care.11 With the combined 30-day risk-adjusted surgical outcomes at over 200 hospitals in the United States offered by the database, we have the ability to assess national trends in outcomes for operations such as immediate breast reconstruction.12 Furthermore, the database allows us to track improvement of outcomes over time, which was the intended purpose of this database collection. In this study, we evaluate national trends in immediate breast reconstruction practices and report short-term outcomes and complications. Furthermore, we aim to assess the effect that American College of Surgeons National Surgical Quality Improvement Program data collection has had on improving the national outcomes in immediate breast reconstruction. In this way, we can track the effect of this database on clinical practices in our specialty and make recommendations for future directions of outcomes-based improvement.
PATIENTS AND METHODS American College of Surgeons National Surgical Quality Improvement Program Database We reviewed the 2007 to 2011 American College of Surgeons National Surgical Quality Improvement Program databases and identified encounters for simultaneous mastectomy with breast reconstruction.11 The database was populated by trained research staff at each institution using systematic sampling of operative procedures performed at participating institutions.
860
Each data set contains 240 variables for each case, including patient demographics, baseline comorbidities, preoperative risk factors, intraoperative variables, and 30-day postoperative complications. Patients who were discharged before 30 days were contacted by letter or telephone survey to complete the full 30-day follow-up period.13 We obtained and accessed this database on August 22, 2013. Patient Selection We used 2012 Current Procedural Terminology codes to select cases from the 2006 to 2011 program databases. Implant-based reconstructions included immediate implant (19340), delayed implant (19342), and tissue expander placement (19357). Autologous reconstructions included the latissimus dorsi flap with or without tissue expander or implant (19361), pedicled transverse rectus abdominis myocutaneous (TRAM) flap (19367, 19368, and 19369), and free flap (19364). Immediate reconstructions were defined as cases performed concurrent with mastectomy during the initial hospitalization. Current Procedural Terminology codes for mastectomy included simple mastectomy (19303), subcutaneous mastectomy (19304), radical mastectomy (19305 and 19306), and modified radical mastectomy (19307). Codes for the use of acellular dermal matrix (15170, 15171, 15330, 15331, and 15350) were used to identify patients who had acellular dermal matrix–assisted reconstructions. Patient encounters with a code with both mastectomy and reconstruction were included as immediate reconstructions for the purposes of this study. We further narrowed the cohort down to only clean cases, defined as wound class 1 cases, as an attempt to increase the specificity of our case selection. Classification of Variables We analyzed the 240 variables that were collected in standard fashion in the program’s database and focused on several comorbidities, including obesity (body mass index > 30), diabetes, smoking, chemotherapy within 30 days, and radiotherapy within 90 days. The primary outcome variables for postoperative complications included postoperative superficial surgical-site infection, deep surgical-site infection, organ/ space surgical-site infection, wound dehiscence, return to the operating room, loss of flap or implant-based reconstruction, pulmonary embolism, and postoperative pneumonia.
Volume 134, Number 5 • Improving Breast Reconstruction Safety Statistical Analysis Differences with respect to continuous variables of age, body mass index, albumin, and operative times were analyzed using the t test and reported as mean values. All categorical variables were analyzed using the Pearson chi-square test. Postoperative complications were tracked as dichotomous variables and reported as whole values and percentage of cases in the cohort. A logistic regression model was used to examine the interaction of autologous reconstruction with the three main predictors of smoking, diabetes, and obesity. Risk ratios were obtained for outcomes data and hypotheses were analyzed with the Pearson chi-square test. All data were analyzed using Stata version 13 (StataCorp LP, College Station, Texas). All p values were two-tailed, and values of p < 0.05 were considered to be significant.
RESULTS During the study period, there were 15,978 clean cases of patients undergoing mastectomy with concurrent breast reconstruction using either implant-based reconstruction (i.e., implant or tissue expander placement) or autologous tissue reconstruction (i.e., latissimus flap, TRAM flap, or free flap). The number of cases increased every year over this period, likely because of increasing annual enrollment of hospitals contributing to the database. Overall patient demographics and preoperative comorbidities were summarized by year (Table 1). The racial makeup of patients was mostly Caucasian (88.3 percent), followed by African American (7.9 percent), Asian (3.0 percent), other (0.4 percent), Pacific Islander (0.3 percent), and American Indian (0.2 percent). More nonCaucasians contributed to the cohort over time (Table 1). The mean age of the patients was 51.1 ± 10.4 years, and this remained relatively stable throughout the study period. There was a trend toward a decrease in the percentage of smokers undergoing immediate breast reconstruction from 2007 (13.8 percent) to 2011 (11.8 percent). The mean body mass index of patients was 26.9 ± 6.6 kg/m2, and 4276 patients (26.8 percent) were classified as being obese with a body mass index greater than 30. There were 180 type 1 diabetics (1.1 percent) and 576 type 2 diabetics (3.6 percent) in our patient population. There were 672 patients (4.2 percent) who had undergone chemotherapy within 30 days and 60 patients (0.4 percent) who had undergone radiotherapy within 90 days. The American Society of Anesthesiologists
Classification of Physical Status was also evaluated, and we noted a trend toward increasing percentage of more complex class 3 and 4 patients and a decrease in the percentage of lower class 1 and 2 patients (Table 1). There were 12,373 cases of implant-based reconstruction (77.4 percent) and 2876 (18.0 percent) cases of autologous reconstruction (Table 2). The majority of cases involved tissue expander placement, and there was an increasing trend in these cases from 2007 (64.1 percent) to 2011 (69.3 percent). With regard to autologous reconstruction, there is a decreasing trend in the use of the pedicled TRAM flap from 2007 (12.8 percent) to 2011 (6.2 percent), with an increasing trend in free flap reconstruction from 2007 (3.2 percent) to 2011 (6.4 percent). Intraoperative characteristics of the study cohort are summarized in Table 3. The mean operative time was 205 ± 84 minutes for cases with implant-based reconstruction and 387 ± 197 minutes for autologous reconstruction (p < 0.005, t test). There was a trend of increasing operative times from 2007 to 2011 for both implant-based and autologous reconstructions. The overall incidences of postoperative complications within 30 days for this cohort are summarized in Table 4. Complications analyzed included superficial surgical-site infection (1.9 percent), deep surgical-site infection (1.2 percent), organ/ space surgical-site infection (0.7 percent), wound dehiscence (0.64 percent), unplanned return to the operating room (7.66 percent), loss of either flaps (2.8 percent) or implants (0.8 percent), pulmonary embolism (0.25 percent), and postoperative pneumonia (0.12 percent). There was a trend toward an increase in superficial surgicalsite infections from 2007 (1.7 percent) to 2011 (2.3 percent). There was also a trend toward an increase in wound dehiscence from 2007 (0.3 percent) to 2011 (0.9 percent). The overall rate of implant loss within 30 days was low (0.5 to 1.1 percent per year), and the incidence of flap loss decreased between 2007 (3.9 percent) and 2011 (1.1 percent). The overall incidence of pulmonary embolisms was low, but there was a trend toward a decrease in pulmonary embolisms recorded between 2007 (0.52 percent) and 2011 (0.21 percent). We evaluated postoperative complications with regard to several risk factors, including operative time, type of reconstruction, smoking status, diabetes, obesity, chemotherapy within 30 days, radiotherapy within 90 days, and use of acellular dermal matrix. Increasing mean operative time
861
Plastic and Reconstructive Surgery • November 2014 Table 1. Patient Demographics and Comorbidities for Patients Undergoing Mastectomy and Immediate Breast Reconstruction in the 2007 to 2011 American College of Surgeons National Surgical Quality Improvement Program Data Set per Year* Characteristic No. of cases Race (%) White Black American Indian Pacific Islander Asian Other Mean age ± SD, yr Mean BMI ± SD, kg/m2 No. of obese patients (BMI > 30) Smoking history within 1 yr Alcohol consumption > 2 drinks/day Diabetes Type 1 Type 2 Dyspnea Moderate At rest COPD CHF MI history within 6 mo Prior cardiac stent placement Prior cardiac surgery Hypertension PVD with operative intervention Preoperative HD Prior TIA Prior CVA with defect Prior CVA without defect Disseminated cancer Recent weight loss Mean albumin, mg/dl Bleeding disorder Steroid use Preoperative chemotherapy within 30 days Preoperative irradiation within 90 days ASA physical classification 1 2 3 4
2007 (%)
2008 (%)
2009 (%)
2010 (%)
2011 (%)
1931
2654
3490
3680
4223
1564 (90.1) 115 (6.6) 0 (0) 0 (0) 0 (0) 57 (3.3) 50.7 ± 10.1 26.5 ± 6.2 465 (24.1) 267 (13.8) 20 (1.0)
2119 (89.8) 160 (6.8) 2 (0.1) 3 (0.1) 77 (3.3) 0 (0) 50.7 ± 10.4 26.8 ± 6.4 707 (26.6) 360 (13.6) 30 (1.1)
2806 (89.4) 237 (7.6) 5 (0.2) 12 (0.4) 78 (2.5) 0 (0) 51.1 ± 10.3 26.9 ± 6.8 918 (26.3) 479 (13.7) 43 (1.3)
2879 (87.4) 276 (8.4) 10 (0.3) 10 (0.3) 120 (3.6) 0 (0) 51.5 ± 10.6 27.4 ± 7.2 1072 (29.1) 461 (12.5) 44 (1.2)
15 (0.8) 62 (3.2)
35 (1.3) 81 (3.1)
45 (1.3) 135 (3.9)
41 (1.1) 150 (4.1)
85 (4.4) 3 (0.16) 13 (0.67) 2 (0.10) 0 (0) 15 (0.78) 10 (0.52) 455 (23.6) 3 (0.16) 3 (0.16) 14 (0.73) 2 (0.10) 7 (0.36) 11 (0.57) 6 (0.31) 4.2 ± 0.4 12 (0.62) 22 (1.14) 86 (4.45) 15 (0.78)
108 (4.1) 2 (0.08) 18 (0.68) 1 (0.04) 1 (0.04) 15 (0.57) 10 (0.38) 613 (23.1) 6 (0.23) 3 (0.11) 10 (0.38) 5 (0.19) 9 (0.34) 17 (0.64) 12 (0.45) 4.2 ± 0.4 18 (0.68) 25 (0.94) 102 (3.84) 8 (0.30)
110 (3.2) 1 (0.03) 24 (0.69) 1 (0.03) 0 (0) 16 (0.46) 12 (0.34) 831 (23.8) 3 (0.09) 1 (0.03) 14 (0.40) 13 (0.37) 12 (0.34) 30 (0.86) 14 (0.40) 4.2 ± 0.4 26 (0.74) 23 (0.66) 119 (3.41) 9 (0.26)
134 (3.6) 3 (0.08) 37 (1.01) 0 (0) 0 (0) 26 (0.71) 19 (0.52) 904 (24.6) 4 (0.11) 4 (0.11) 21 (0.57) 9 (0.24) 17 (0.46) 28 (0.76) 13 (0.35) 4.2 ± 0.4 19 (0.52) 30 (0.82) 172 (4.67) 15 (0.41)
207 (10.7) 1412 (73.1) 309 (16.0) 1 (0.05)
289 (10.9) 1896 (71.4) 461 (17.4) 6 (0.23)
368 (10.5) 2441 (70.0) 669 (19.2) 10 (0.29)
329 (8.9) 2575 (70.0) 762 (20.7) 10 (0.27)
p
3328 (86.35)
