j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Racial disparity in early graft failure after infrainguinal bypass Shalini Selvarajah, MD, MPH,a James H. Black III, MD,b Adil H. Haider, MD, MPH,a,c and Christopher J. Abularrage, MDa,b,* a
Department of Surgery, Center for Surgical Trials and Outcomes Research (CSTOR), Johns Hopkins University School of Medicine, Baltimore, Maryland b Division of Vascular Surgery and Endovascular Therapy, Department of Surgery, Johns Hopkins Hospital, Baltimore, Maryland c Division of Trauma and Acute Care Surgery, Department of Surgery, Johns Hopkins Hospital, Baltimore, Maryland
article info
abstract
Article history:
Background: Racial disparities have been shown to be associated with increasing health-
Received 6 March 2014
care costs. We sought to identify racial disparities in 30-d graft failure rates after infrain-
Received in revised form
guinal bypass in an effort to define targets for improved health care among minorities.
8 April 2014
Methods: The 2005e2011 National Surgical Quality Improvement Program database was
Accepted 15 April 2014
queried for patients with peripheral arterial disease who underwent infrainguinal bypass
Available online xxx
as their primary procedure. A bivariate analysis was done to assess pre and intraoperative risk factors across race (whites, blacks, and Hispanics). Multivariate logistic regression was
Keywords:
performed to assess the independent association of race with 30-d graft failure.
NSQIP
Results: Of a total of 16,276 patients, 12,536 (77.0%) were whites, 2940 (18.1%) blacks, and 800
Infrainguinal bypass
(4.9%) Hispanics. Black patients were more likely to be younger, female, current smokers,
Racial disparity
and on dialysis (P < 0.001, all). In addition, whites were less likely to present with critical
Early graft failure
limb ischemia compared with blacks and Hispanics (44.2 versus 55.4 versus 52.8%, respec-
Complications
tively; P < 0.001). Similarly, fewer whites underwent femoral-tibial (31.4 vs. 34.7 vs. 38.6% respectively) or popliteal-tibial level bypasses (8.9 versus 13.4 versus 16.1%, respectively) than blacks and Hispanics (P < 0.001, all). There was no difference in the use of autogenous conduit across the groups (P ¼ 0.266). Proportionally more blacks than whites developed early graft failure (6.7 versus 4.5%; P < 0.001) but there was no difference comparing Hispanics to whites (6.0 versus 4.5%; P ¼ 0.057). On multivariable analysis, black race remained independently associated with early graft failure (adjusted odds ratio ¼ 1.26, 95% confidence interval 1.05e1.51; P ¼ 0.011). Conclusions: More blacks and Hispanics present with critical limb ischemia, requiring distal revascularization. Even when controlling for anatomic differences and degree of peripheral arterial disease, black race remained independently associated with early graft failure after infrainguinal bypass. These results identify a target for improved outcomes. ª 2014 Elsevier Inc. All rights reserved.
Presented as an oral presentation at the Academic Surgical Congress in San Diego, California on February 6, 2014. * Corresponding author. Division of Vascular Surgery and Endovascular Therapy, Department of Surgery, The Johns Hopkins Hospital, Halsted 671, 600 N. Wolfe Street, Baltimore, MD 21287. Tel.: þ1 410 955 5174; fax: þ1 410 614 2079. E-mail address: [email protected] (C.J. Abularrage). 0022-4804/$ e see front matter ª 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2014.04.029
2
1.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
Introduction
National efforts to improve postoperative outcomes are being made in an effort to control the rising costs of health care. The presence of racial disparities in health care outcomes is well established including in the realm of vascular surgery, and is responsible for many such costs [1e3]. An estimated 4.3% of the United States adult population (5 million Americans) aged >40 y have lower extremity peripheral arterial disease (PAD) [4]. PAD, an atherosclerotic disease known to be influenced by smoking, hypertension, diabetes, hypercholesterolemia, and advancing age, has been shown to disproportionately affect minority populations. Blacks and Hispanics are reported to have PAD at a rate 2e3 times greater than is seen in whites [5,6]. However, blacks and Hispanics are less likely to undergo revascularization procedures such as infrainguinal bypass surgery or balloon angioplasty for limb salvage, and are instead two to four times more likely to have amputations compared with whites [7]. Postoperative complication rates in minorities are also reported to be higher [2,3,8]. There is an intricate interplay of multiple factors that gives rise to such disparities. Most of prior studies have shown that minority populations are more likely to come from a lower socioeconomic background, be uninsured, and have reduced access to medical care, including preventative care [1,7]. These alone maybe reasons why minority populations tend to present at later stages of the disease process, often presenting with features of critical limb ischemia (CLI), which may partially explain why there are marked discrepancies in limb salvage and amputation rates across races [1,7]. Early graft failure after infrainguinal bypass surgery is less frequently described compared with 1- or 5-y graft patency. Robinson et al. [10] in 2009 reported that blacks were two times more likely than whites (11% versus 5%) to develop early graft failure in a single-center analysis involving 1646 patients. Although not significantly different, Hispanics (n ¼ 57) showed a tendency toward having greater early graft failure rates compared with whites (7% versus 5%). There is a need to obtain objective information to fill the gap of understanding if such disparities remain in a larger cohort of patients recruited from multiple centers around the United States. Primary graft patency rates continue to be a crucial target for improvement, as they can help minimize readmission rates, reoperations, disablement due to amputations, and consequently, rising health-care cost [11]. We sought to evaluate the presence of racial disparities in the occurrence of early graft failure after infrainguinal bypass surgery using a validated, nationwide, multicenter surgical database in an effort to define targets for improved health care in minorities [12].
2.
Methods
2.1.
Data source and inclusion criteria
We queried the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database
from 2005e2011. The ACS-NSQIP is a prospectively maintained database of surgical procedures with administrative and clinical information at the pre-, intra-, and post-operative period collected from non-Veterans Affairs (VA) hospitals. It was developed in line with the objective of the original VANSQIP database that was established in the early 1970s for the purpose of monitoring quality improvement efforts and improving surgical outcomes in VA facilities [13]. Surgical case reviewers from participating hospitals abstract information from clinical and administrative notes to populate >100 variables in the database for every surgical procedure that is systematically sampled. Detailed information on ACS-NSQIP participating facilities, sampling methodology, and program implementation can be obtained from the ACS website and previous literature [13,14]. The study population was identified using Current Procedural Terminology codes for infrainguinal bypass recorded in the principal procedure field (35556, 35566, 35570, 35571, 35583, 35585, 35587, 35656, 35666, and 35671). Patients with PAD identified using International Classification of Diseases9-Clinical Modification (ICD-9-CM) diagnosis codes of 440.20e440.24, 443.9, or a ACS-NSQIP definition of “peripheral vascular disease” were retained. Subsequently, patients with a primary postoperative diagnosis representing acute ischemia (444.0e444.9), or those who underwent a concomitant non-infrainguinal lower extremity procedure, reoperations, or revisions were excluded from the analysis. Because of the small sample size of other race categories, only observations with race reported as whites, blacks, and Hispanics were retained. A complete iteration of the inclusion and exclusion criteria used to define the final study population is in the Figure.
2.2.
Data collection
Demographic variables that were collected for this study included age, sex, and race. Race was categorized as whites for “white, not of Hispanic origin”, blacks for “black, not of Hispanic origin”, and Hispanic for “Hispanic, white; Hispanic, black; and Hispanic, color unknown”. Other variables that were analyzed include preoperative cigarette smoking within 1 y before surgery (active smoking), consumption of more than moderate amounts (>2 drinks/d) of alcohol within 2 wk before surgery, body mass index, and comorbid conditions (cardiac, pulmonary, renal, cerebrovascular, infection-related, and diabetes). Operative details such as CLI defined by ACS-NSQIP as the presence of ischemic ulceration and/or tissue loss related to peripheral vascular disease; bypass target level (femoralpopliteal, femoral-tibial, popliteal-tibial, and tibial-tibial); type of graft (vein versus prosthetic); emergent or elective surgery; preoperative thrombocytosis defined as platelet count of >400 103/mm3; American Society of Anesthesiologists (ASA) classification; presence of a resident during surgery; type of anesthesia; and operative time in minutes were also collected. All postoperative complications collected by ACS-NSQIP, comprising mortality, return to operating room, graft failure, septicemia, bleeding, wound, cardiac, respiratory, renal, and thromboembolic complications were reported in this study.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
3
Figure e Patient selection criteria. *BMI [ body mass index; SIRS [ systemic inflammatory response syndrome; TIA [ transient ischemic attack.
2.3.
Statistical analysis
Univariate analysis for all demographic characteristics, preoperative risk factors, operative details, and postoperative complications was performed for the whole study population then, stratified by race. Counts and proportions were presented for categorical variables, and, chi-square and Fisher exact tests were subsequently performed. For each continuous variable (i.e.,: age, body mass index, and operative time), we first assessed the distribution of the data, and tested for normality using histograms and q-q plots. As operative time had a skewed distribution, we presented median and interquartile ranges (IQR), and performed a test of association using the KruskaleWallis test. For other normally distributed continuous variables, we presented means and standard deviations then, a one-way analysis of variance was performed for comparison of means. We subsequently fitted unadjusted and adjusted logistic regression models using a threshold P value of 0.100 for sequential addition of covariates into the model to assess for the impact of race on early graft failure. Interaction terms between race and CLI, level of bypass, type of graft, active smoking, and alcohol consumption were tested for and were found to be not statistically significant hence a parsimonious model without interaction terms was used. The final model chosen was tested for goodness of fit, and had a
HosmereLemeshow P value of >0.050. Variance inflation factors and a condition index were calculated for the final model. We present odds ratios (OR) and their corresponding 95% confidence intervals (CI) and P values. Level of statistical significance was set at P ¼ 0.050. White race was the reference race category. All data analyses were performed using Stata Statistical Software: Release 12 (College Station, TX) [15].
3.
Results
A total of 16,276 patients undergoing infrainguinal bypass surgery were identified, with 12,536 (77.0%) whites, 2940 (18.1%) blacks, and 800 (4.9%) Hispanics. Infrainguinal bypass was predominantly performed in males (n ¼ 10,412; 64.0%) and the mean age ( standard error) at time of surgery was 68.0 0.09 y. CLI was present in 46.6% (n ¼ 7587) of procedures that were performed and most of the revascularizations were at the femoral-popliteal level (n ¼ 9340; 57.4%) and had used autogenous conduits (n ¼ 11,063; 68.0%). Table 1 describes basic demography and patient characteristics of the study population stratified by race. Black patients were more likely to be younger (65.1 0.21 y versus 68.6 0.10 y), female (44.9% versus 33.7%), and be current smokers (47.1% versus 39.7%) compared with whites
4
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
Table 1 e Description of the study population stratified by race. Infrainguinal bypass (N ¼ 16,276) Whites n ¼ 12,536 (%) Blacks n ¼ 2940 (%) Hispanics n ¼ 800 (%) Total n ¼ 16,276 (%) Age (y)* Male sex BMI (kg/m2)* Active smoker >Moderate alcohol intake CLI Level of bypass Femoral-popliteal Femoral-tibial Popliteal-tibial Tibial-tibial Type of graft Vein Prosthetic Concurrent adjunct procedure(s)
68.58 0.10 8314 (66.3) 27.62 0.06 4973 (39.7) 761 (6.1) 5537 (44.2) 7462 3938 1110 26
(59.5) (31.4) (8.9) (0.2)
65.09 0.21 1621 (55.1) 27.29 0.12 1385 (47.1) 120 (4.1) 1628 (55.4) 1520 1020 395 5
(51.7) (34.7) (13.4) (0.2)
68.74 0.40 477 (59.6) 27.32 0.22 198 (24.8) 23 (2.9) 422 (52.8) 358 309 129 4
(44.8) (38.6) (16.1) (0.5)
67.96 0.09 10,412 (64.0) 27.54 0.05 6556 (40.3) 904 (5.6) 7587 (46.6) 9340 5267 1634 35
P
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Racial disparity in early graft failure after infrainguinal bypass Shalini Selvarajah, MD, MPH,a James H. Black III, MD,b Adil H. Haider, MD, MPH,a,c and Christopher J. Abularrage, MDa,b,* a
Department of Surgery, Center for Surgical Trials and Outcomes Research (CSTOR), Johns Hopkins University School of Medicine, Baltimore, Maryland b Division of Vascular Surgery and Endovascular Therapy, Department of Surgery, Johns Hopkins Hospital, Baltimore, Maryland c Division of Trauma and Acute Care Surgery, Department of Surgery, Johns Hopkins Hospital, Baltimore, Maryland
article info
abstract
Article history:
Background: Racial disparities have been shown to be associated with increasing health-
Received 6 March 2014
care costs. We sought to identify racial disparities in 30-d graft failure rates after infrain-
Received in revised form
guinal bypass in an effort to define targets for improved health care among minorities.
8 April 2014
Methods: The 2005e2011 National Surgical Quality Improvement Program database was
Accepted 15 April 2014
queried for patients with peripheral arterial disease who underwent infrainguinal bypass
Available online xxx
as their primary procedure. A bivariate analysis was done to assess pre and intraoperative risk factors across race (whites, blacks, and Hispanics). Multivariate logistic regression was
Keywords:
performed to assess the independent association of race with 30-d graft failure.
NSQIP
Results: Of a total of 16,276 patients, 12,536 (77.0%) were whites, 2940 (18.1%) blacks, and 800
Infrainguinal bypass
(4.9%) Hispanics. Black patients were more likely to be younger, female, current smokers,
Racial disparity
and on dialysis (P < 0.001, all). In addition, whites were less likely to present with critical
Early graft failure
limb ischemia compared with blacks and Hispanics (44.2 versus 55.4 versus 52.8%, respec-
Complications
tively; P < 0.001). Similarly, fewer whites underwent femoral-tibial (31.4 vs. 34.7 vs. 38.6% respectively) or popliteal-tibial level bypasses (8.9 versus 13.4 versus 16.1%, respectively) than blacks and Hispanics (P < 0.001, all). There was no difference in the use of autogenous conduit across the groups (P ¼ 0.266). Proportionally more blacks than whites developed early graft failure (6.7 versus 4.5%; P < 0.001) but there was no difference comparing Hispanics to whites (6.0 versus 4.5%; P ¼ 0.057). On multivariable analysis, black race remained independently associated with early graft failure (adjusted odds ratio ¼ 1.26, 95% confidence interval 1.05e1.51; P ¼ 0.011). Conclusions: More blacks and Hispanics present with critical limb ischemia, requiring distal revascularization. Even when controlling for anatomic differences and degree of peripheral arterial disease, black race remained independently associated with early graft failure after infrainguinal bypass. These results identify a target for improved outcomes. ª 2014 Elsevier Inc. All rights reserved.
Presented as an oral presentation at the Academic Surgical Congress in San Diego, California on February 6, 2014. * Corresponding author. Division of Vascular Surgery and Endovascular Therapy, Department of Surgery, The Johns Hopkins Hospital, Halsted 671, 600 N. Wolfe Street, Baltimore, MD 21287. Tel.: þ1 410 955 5174; fax: þ1 410 614 2079. E-mail address: [email protected] (C.J. Abularrage). 0022-4804/$ e see front matter ª 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2014.04.029
2
1.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
Introduction
National efforts to improve postoperative outcomes are being made in an effort to control the rising costs of health care. The presence of racial disparities in health care outcomes is well established including in the realm of vascular surgery, and is responsible for many such costs [1e3]. An estimated 4.3% of the United States adult population (5 million Americans) aged >40 y have lower extremity peripheral arterial disease (PAD) [4]. PAD, an atherosclerotic disease known to be influenced by smoking, hypertension, diabetes, hypercholesterolemia, and advancing age, has been shown to disproportionately affect minority populations. Blacks and Hispanics are reported to have PAD at a rate 2e3 times greater than is seen in whites [5,6]. However, blacks and Hispanics are less likely to undergo revascularization procedures such as infrainguinal bypass surgery or balloon angioplasty for limb salvage, and are instead two to four times more likely to have amputations compared with whites [7]. Postoperative complication rates in minorities are also reported to be higher [2,3,8]. There is an intricate interplay of multiple factors that gives rise to such disparities. Most of prior studies have shown that minority populations are more likely to come from a lower socioeconomic background, be uninsured, and have reduced access to medical care, including preventative care [1,7]. These alone maybe reasons why minority populations tend to present at later stages of the disease process, often presenting with features of critical limb ischemia (CLI), which may partially explain why there are marked discrepancies in limb salvage and amputation rates across races [1,7]. Early graft failure after infrainguinal bypass surgery is less frequently described compared with 1- or 5-y graft patency. Robinson et al. [10] in 2009 reported that blacks were two times more likely than whites (11% versus 5%) to develop early graft failure in a single-center analysis involving 1646 patients. Although not significantly different, Hispanics (n ¼ 57) showed a tendency toward having greater early graft failure rates compared with whites (7% versus 5%). There is a need to obtain objective information to fill the gap of understanding if such disparities remain in a larger cohort of patients recruited from multiple centers around the United States. Primary graft patency rates continue to be a crucial target for improvement, as they can help minimize readmission rates, reoperations, disablement due to amputations, and consequently, rising health-care cost [11]. We sought to evaluate the presence of racial disparities in the occurrence of early graft failure after infrainguinal bypass surgery using a validated, nationwide, multicenter surgical database in an effort to define targets for improved health care in minorities [12].
2.
Methods
2.1.
Data source and inclusion criteria
We queried the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database
from 2005e2011. The ACS-NSQIP is a prospectively maintained database of surgical procedures with administrative and clinical information at the pre-, intra-, and post-operative period collected from non-Veterans Affairs (VA) hospitals. It was developed in line with the objective of the original VANSQIP database that was established in the early 1970s for the purpose of monitoring quality improvement efforts and improving surgical outcomes in VA facilities [13]. Surgical case reviewers from participating hospitals abstract information from clinical and administrative notes to populate >100 variables in the database for every surgical procedure that is systematically sampled. Detailed information on ACS-NSQIP participating facilities, sampling methodology, and program implementation can be obtained from the ACS website and previous literature [13,14]. The study population was identified using Current Procedural Terminology codes for infrainguinal bypass recorded in the principal procedure field (35556, 35566, 35570, 35571, 35583, 35585, 35587, 35656, 35666, and 35671). Patients with PAD identified using International Classification of Diseases9-Clinical Modification (ICD-9-CM) diagnosis codes of 440.20e440.24, 443.9, or a ACS-NSQIP definition of “peripheral vascular disease” were retained. Subsequently, patients with a primary postoperative diagnosis representing acute ischemia (444.0e444.9), or those who underwent a concomitant non-infrainguinal lower extremity procedure, reoperations, or revisions were excluded from the analysis. Because of the small sample size of other race categories, only observations with race reported as whites, blacks, and Hispanics were retained. A complete iteration of the inclusion and exclusion criteria used to define the final study population is in the Figure.
2.2.
Data collection
Demographic variables that were collected for this study included age, sex, and race. Race was categorized as whites for “white, not of Hispanic origin”, blacks for “black, not of Hispanic origin”, and Hispanic for “Hispanic, white; Hispanic, black; and Hispanic, color unknown”. Other variables that were analyzed include preoperative cigarette smoking within 1 y before surgery (active smoking), consumption of more than moderate amounts (>2 drinks/d) of alcohol within 2 wk before surgery, body mass index, and comorbid conditions (cardiac, pulmonary, renal, cerebrovascular, infection-related, and diabetes). Operative details such as CLI defined by ACS-NSQIP as the presence of ischemic ulceration and/or tissue loss related to peripheral vascular disease; bypass target level (femoralpopliteal, femoral-tibial, popliteal-tibial, and tibial-tibial); type of graft (vein versus prosthetic); emergent or elective surgery; preoperative thrombocytosis defined as platelet count of >400 103/mm3; American Society of Anesthesiologists (ASA) classification; presence of a resident during surgery; type of anesthesia; and operative time in minutes were also collected. All postoperative complications collected by ACS-NSQIP, comprising mortality, return to operating room, graft failure, septicemia, bleeding, wound, cardiac, respiratory, renal, and thromboembolic complications were reported in this study.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
3
Figure e Patient selection criteria. *BMI [ body mass index; SIRS [ systemic inflammatory response syndrome; TIA [ transient ischemic attack.
2.3.
Statistical analysis
Univariate analysis for all demographic characteristics, preoperative risk factors, operative details, and postoperative complications was performed for the whole study population then, stratified by race. Counts and proportions were presented for categorical variables, and, chi-square and Fisher exact tests were subsequently performed. For each continuous variable (i.e.,: age, body mass index, and operative time), we first assessed the distribution of the data, and tested for normality using histograms and q-q plots. As operative time had a skewed distribution, we presented median and interquartile ranges (IQR), and performed a test of association using the KruskaleWallis test. For other normally distributed continuous variables, we presented means and standard deviations then, a one-way analysis of variance was performed for comparison of means. We subsequently fitted unadjusted and adjusted logistic regression models using a threshold P value of 0.100 for sequential addition of covariates into the model to assess for the impact of race on early graft failure. Interaction terms between race and CLI, level of bypass, type of graft, active smoking, and alcohol consumption were tested for and were found to be not statistically significant hence a parsimonious model without interaction terms was used. The final model chosen was tested for goodness of fit, and had a
HosmereLemeshow P value of >0.050. Variance inflation factors and a condition index were calculated for the final model. We present odds ratios (OR) and their corresponding 95% confidence intervals (CI) and P values. Level of statistical significance was set at P ¼ 0.050. White race was the reference race category. All data analyses were performed using Stata Statistical Software: Release 12 (College Station, TX) [15].
3.
Results
A total of 16,276 patients undergoing infrainguinal bypass surgery were identified, with 12,536 (77.0%) whites, 2940 (18.1%) blacks, and 800 (4.9%) Hispanics. Infrainguinal bypass was predominantly performed in males (n ¼ 10,412; 64.0%) and the mean age ( standard error) at time of surgery was 68.0 0.09 y. CLI was present in 46.6% (n ¼ 7587) of procedures that were performed and most of the revascularizations were at the femoral-popliteal level (n ¼ 9340; 57.4%) and had used autogenous conduits (n ¼ 11,063; 68.0%). Table 1 describes basic demography and patient characteristics of the study population stratified by race. Black patients were more likely to be younger (65.1 0.21 y versus 68.6 0.10 y), female (44.9% versus 33.7%), and be current smokers (47.1% versus 39.7%) compared with whites
4
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 4 ) 1 e9
Table 1 e Description of the study population stratified by race. Infrainguinal bypass (N ¼ 16,276) Whites n ¼ 12,536 (%) Blacks n ¼ 2940 (%) Hispanics n ¼ 800 (%) Total n ¼ 16,276 (%) Age (y)* Male sex BMI (kg/m2)* Active smoker >Moderate alcohol intake CLI Level of bypass Femoral-popliteal Femoral-tibial Popliteal-tibial Tibial-tibial Type of graft Vein Prosthetic Concurrent adjunct procedure(s)
68.58 0.10 8314 (66.3) 27.62 0.06 4973 (39.7) 761 (6.1) 5537 (44.2) 7462 3938 1110 26
(59.5) (31.4) (8.9) (0.2)
65.09 0.21 1621 (55.1) 27.29 0.12 1385 (47.1) 120 (4.1) 1628 (55.4) 1520 1020 395 5
(51.7) (34.7) (13.4) (0.2)
68.74 0.40 477 (59.6) 27.32 0.22 198 (24.8) 23 (2.9) 422 (52.8) 358 309 129 4
(44.8) (38.6) (16.1) (0.5)
67.96 0.09 10,412 (64.0) 27.54 0.05 6556 (40.3) 904 (5.6) 7587 (46.6) 9340 5267 1634 35
P
