Surg Endosc (2014) 28:1712–1719 DOI 10.1007/s00464-013-3380-y
and Other Interventional Techniques
Elective antireflux surgery in the US: an analysis of national trends in utilization and inpatient outcomes from 2005 to 2010 Luke M. Funk • Aliyah Kanji • W. Scott Melvin Kyle A. Perry
•
Received: 13 August 2013 / Accepted: 9 December 2013 / Published online: 1 January 2014 Ó Springer Science+Business Media New York 2013
Abstract Background Previous research suggested that antireflux surgery reached its peak volume in the US more than a decade ago. Factors such as changes in population demographics and improvements in surgical outcomes may have reversed this decline. We sought to examine national trends in the management of antireflux surgery patients and identify patient and hospital characteristics associated with postoperative complications. Methods We analyzed data from the Nationwide Inpatient Sample to identify adults with gastroesophageal reflux disease or esophagitis who underwent elective antireflux surgery between 2005 and 2010. Patient and hospital characteristics were analyzed. A multivariate logistic regression model was used to identify characteristics associated with an increased risk of postoperative complications following laparoscopic antireflux surgery. Results The volume of elective antireflux surgery remained relatively stable between 2005 (n = 15,819) and 2010 (n = 18,780). The percentage of patients older than 64 years of age increased from 21.1 % in 2005 to 30.9 % in 2010 (p \ 0.01), while the percentage with a Charlson score over 2 more than doubled (1.2–2.7 %; p \ 0.01). Inpatient complication rates (6.3 vs. 6.6 %; p = 0.21) and mortality (0.08 vs. 0.21; p = 0.72) were unchanged. On multivariate analysis, patients older than 79 years were three times as likely to develop a complication (odds ratio [OR] 3.1; 95 % CI 2.1–4.5) as were patients with a Charlson score over 2 (OR 3.1; 95 % CI 2.2–4.3).
L. M. Funk (&) A. Kanji W. Scott Melvin K. A. Perry Department of Surgery, Center for Minimally Invasive Surgery, The Ohio State University Medical Center, Columbus, OH, USA e-mail: [email protected]
123
Conclusions Today’s antireflux surgery patient population is a higher-risk cohort, but complication rates have remained stable and inpatient mortality has declined more than 50 % over the past decade. Given these findings, additional research is needed to understand why antireflux surgery is underutilized, with a decline of more than twothirds since its peak in 1999. Keywords Minimally invasive surgery Gastroesophageal reflux disease Laparoscopic antireflux surgery
With the widespread application of laparoscopic surgery during the 1990s, the volume of antireflux surgery for the treatment of gastroesophageal reflux disease (GERD) increased dramatically. The initial population-based study of antireflux surgery in the US found that the total number of antireflux operations performed annually nearly quadrupled between 1990 and 1997, when nearly 24,000 operations (12 cases/100,000 adults) were performed [1]. This corresponded with a massive increase in the utilization of laparoscopy, from 1 % of antireflux procedures in 1990 to 64 % in 1997. Subsequent analyses found that antireflux surgery volume peaked in 1999 at 16 cases per 100,000 adults, declined 30 % over the next 4 years to 11 per 100,000 adults, and continued to gradually decline until at least 2006 [2, 3]. The factors driving this decline in surgical volume included publication of a prominent randomized controlled trial that questioned the benefit of surgery compared with medical therapy, concerns regarding the side effects of antireflux surgery, and the availability of overthe-counter proton pump inhibitors (PPIs) [4, 5]. Current national antireflux surgery volumes are unknown, and changes in both the prevalence of GERD
Surg Endosc (2014) 28:1712–1719
and perception of the utility of long-term PPI use may have impacted physicians’ treatment strategies over the past decade. While the prevalence of GERD has risen significantly due to multiple factors, including the obesity epidemic and an aging population, concerns regarding complications of long-term PPI use, such as osteoporosis and enteric infections, have increased [6–8]. Meanwhile, surgical outcomes have continued to improve with the evolution of minimally invasive techniques [9]. Several randomized controlled trials published within the past decade have also demonstrated durable reflux control following laparoscopic surgery for refractory GERD [10–13]. In light of these changes in GERD epidemiology and treatment, the primary aim of this study was to characterize national trends in the utilization of antireflux surgery in the US between 2005 and 2010. We also sought to assess trends in the outcomes of patients undergoing elective laparoscopic antireflux surgery and identify predictors of postoperative complications.
Materials and methods Data sources We used the Nationwide Inpatient Sample (NIS) from 2005 through 2010 to generate our patient cohort. Managed by the Healthcare Cost and Utilization Project (HCUP) and sponsored by the Agency for Healthcare Research and Quality, the NIS contains all-payer data on inpatient hospitalizations from participating states (representing [95 % of the US population). Data comprising approximately 8 million hospitalizations from 1,000 hospitals are included annually to create a 20 % stratified probability sample of hospitals in the US. Sampling probabilities are based on the number of hospitals in each stratum. Strata are generated based on five hospital characteristics: ownership/control, number of hospital beds, teaching status, urban versus rural location, and region. Post-stratification weights provided by HCUP are used to calculate national estimates for all nonfederal hospitals in the US [14]. National census data were used to calculate rates of surgery per 100,000 people [15].
1713 Table 1 Inclusion and exclusion diagnoses and procedures with associated ICD-9 codes ICD-9 procedure codes Open creation of esophagogastric sphincteric competence (inclusion) Laparoscopic creation of esophagogastric sphincteric competence (inclusion) ICD-9-CM diagnosis codes
44.66 44.67
Malignant neoplasm of the esophagus (exclusion)
150.0–150.9
Achalasia and cardiospasm (exclusion)
530.0
Reflux esophagitis (inclusion)
530.11
Esophageal reflux (inclusion)
530.81
Diaphragmatic hernia with gangrene (exclusion)
551.3
ICD-9-CM International Classification of Diseases, 9th Revision, Clinical Modification
Study variables Patient demographic variables included age, gender, race, comorbid status, payment type (private, medicare, medicaid, or other), and income level (low, lower-middle, uppermiddle, or high). Comorbid status was characterized using Charlson scores for each patient based on ICD-9-CM diagnosis codes. The Charlson score represents a weighted composite measure of overall comorbid status that has been validated for use with administrative data [16]. Hospital characteristics included location (urban or rural), teaching status, size (small, medium, or large), and hospital antireflux procedure volume (high or low, based on the median number of fundoplications performed annually). The median number of laparoscopic operations performed annually at the hospital level was 12 (2005, 2007), 13 (2006, 2008), 15 (2009), and 17 (2010). The primary outcome measure was the occurence of a postoperative complication following laparoscopic antireflux surgery. This was assessed as a composite variable that included reoperation, stroke, myocardial infarction, respiratory failure, small-bowel obstruction, hemodialysis, splenic injury, bleeding, blood transfusion, deep venous thrombosis, wound complications, percutaneous drain placement, or shock. Secondary outcomes included inpatient mortality and length of inpatient stay.
Study population Statistical analyses We used International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure and diagnosis codes to identify all hospitalizations during which an adult patient underwent an elective antireflux surgery. All patients had a diagnosis of either GERD or reflux esophagitis. Patients with a diagnosis of esophageal cancer, achalasia or hiatal hernia with obstruction/gangrene were excluded from the analysis (Table 1).
Analyses of patient demographics, hospital characteristics, and surgical outcomes were performed for all patients who underwent laparoscopic antireflux surgery. We used the Rao–Scott v2 test to compare patient demographics, hospital characteristics and outcomes by year, and accounted for clustering at the hospital level in these analyses [17]. Multivariate logistic regression modeling was used to
123
1714
Surg Endosc (2014) 28:1712–1719
Table 2 Annual case volume estimates of elective fundoplication for patients with GERD/esophagitis in the US between 2005 and 2010 Total
2005
2006
2007
2008
2009
2010
Open and laparoscopic cases
96,702
15,819
15,325
15,026
14,684
17,068
18,780
Total cases per 100,000 people
5.3
5.3
5.1
5.0
4.8
5.6
6.1
Laparoscopic cases
78,233
12,557
12,405
12,371
12,121
13,548
15,231
Percent performed laparoscopically
81.6
79.4
80.9
82.3
82.5
79.4
81.1
p Value
0.54
GERD gastroesophageal reflux disease
Fig. 1 Annual number of antireflux operations performed in the US, 2005–2010
characterize the relationship between patient and hospital factors and postoperative complications. Characteristics with a p value of \0.05 on univariate analysis were selected for inclusion in the regression model along with hospital volume, which was included based on the a priori hypothesis of a correlation between hospital volume and surgical outcome. Race was not included in the model due to the high proportion of missing data (33 % of patients). Generalized estimating equations were used to account for clustering of patients by hospital [18]. All p values were two-sided, and p values \0.05 were considered statistically significant. Data were analyzed using SAS version 9.3 (SAS Institute, Cary, NC, USA).
esophageal cancer, achalasia, or obstruction/gangrene. The annual national volume of elective fundoplications decreased from 15,819 (5.3 cases per 100,000 people) in 2005 to a nadir of 14,684 (4.8 cases per 100,000 people) in 2008. In 2010, a total of 18,780 elective fundoplications (6.1 cases per 100,000 people) were performed (Fig. 1). Of the 96,702 elective fundoplications, 81.6 % underwent a laparoscopic fundoplication and 19.4 % had an open procedure (p \ 0.01). There was no statistically significant difference in the proportion of cases performed laparoscopically during the 6-year period (p = 0.54). The frequency of laparoscopy compared with an open approach was similar in urban (80.5 %) and rural (83.4 %; p = 0.11) hospitals and teaching (80.7 %) versus non-teaching (81.1 %; p = 0.84) hospitals.
Results
Patient characteristics
National volume estimates
The proportion of patients older than 65 years of age who underwent a laparoscopic antireflux operation increased from 21.1 % in 2005 to 30.9 % in 2010 (p \ 0.01), while the percentage with a Charlson score [2 increased from 1.2 % in 2005 to 2.7 % in 2010 (p \ 0.01; Table 3). The proportion of patients with private insurance decreased between 2005 (62.7 %) and 2010 (52.8 %; p = 0.02), while the percentage of Medicare patients increased during
Annual case volume estimates of elective fundoplication for patients with reflux in the US between 2005 and 2010 are summarized in Table 2. Overall, 103,702 patients with a diagnosis of GERD or reflux esophagitis underwent an elective fundoplication between 2005 and 2010; 7,000 were excluded from this analysis due to a diagnosis of
123
Surg Endosc (2014) 28:1712–1719
1715
Lowermiddle
30.4
30.4
24.8
31.2
28.2
29.4
after laparoscopic antireflux surgery according to the year (Table 4). Length of stay varied from a mean of 2.3 days in 2005 and 2007 to 2.5 days in 2008 and 2009 (p = 0.42). Complication rates ranged from 6.3 % in 2005 to 8.1 % in 2006 (p = 0.21), while mortality rates ranged from 0.08 % in 2005 and 2008 to 0.21 % in 2010 (p = 0.72). On univariate analysis, the complication rate increased with age and comorbidity status (Table 5). Patients older than 80 years of age had a 15.8 % complication rate compared with 4.6 % in those under 34 years of age (p \ 0.01). Patients with a Charlson score of 0 developed complications 5.2 % of the time, whereas those with a score greater than 2 had a 17.3 % complication rate (p \ 0.01). Young (18–50 years of age), healthy (Charlson score = 0) patients had the lowest complication rates (3.7 %), while elderly patients with significant comorbidities ([75 years of age, Charlson score [2) developed complications 24.6 % of the time (Fig. 2). Payer status was also associated with statistically significant differences in complication rates. Medicare patients experienced the highest complication rates (9.8 %) and privately insured patients experienced the lowest (5.5 %; p \ 0.01). No hospital characteristics, including procedure volume, were associated with statistically significant differences in complication rates.
Uppermiddle
25.2
27.5
26.2
26.1
27.2
26.8
Multivariate analysis
High
24.2
22.5
23.2
21.6
23.5
22.7
Table 3 Demographic data for patients undergoing elective laparoscopic antireflux surgery in the US between 2005 and 2010 2005
2006 2007
2008
2009
2010
p Value
Age (years, %) 18–34
13.5
11.7
11.0
9.4
9.0
9.1
35–49
30.2
28.8
27.1
25.0
24.5
22.7
50–64
35.2
35.4
35.4
36.9
36.9
37.2
65–79
19.5
21.2
23.6
24.6
26.1
27.0
1.6
2.9
2.8
4.0
3.5
3.9
64.6 12.5
65.1 10.2
66.6 16.2
68.5 12.8
67.5 14.2
68.5 14.7
0.02 0.03
and Other Interventional Techniques
Elective antireflux surgery in the US: an analysis of national trends in utilization and inpatient outcomes from 2005 to 2010 Luke M. Funk • Aliyah Kanji • W. Scott Melvin Kyle A. Perry
•
Received: 13 August 2013 / Accepted: 9 December 2013 / Published online: 1 January 2014 Ó Springer Science+Business Media New York 2013
Abstract Background Previous research suggested that antireflux surgery reached its peak volume in the US more than a decade ago. Factors such as changes in population demographics and improvements in surgical outcomes may have reversed this decline. We sought to examine national trends in the management of antireflux surgery patients and identify patient and hospital characteristics associated with postoperative complications. Methods We analyzed data from the Nationwide Inpatient Sample to identify adults with gastroesophageal reflux disease or esophagitis who underwent elective antireflux surgery between 2005 and 2010. Patient and hospital characteristics were analyzed. A multivariate logistic regression model was used to identify characteristics associated with an increased risk of postoperative complications following laparoscopic antireflux surgery. Results The volume of elective antireflux surgery remained relatively stable between 2005 (n = 15,819) and 2010 (n = 18,780). The percentage of patients older than 64 years of age increased from 21.1 % in 2005 to 30.9 % in 2010 (p \ 0.01), while the percentage with a Charlson score over 2 more than doubled (1.2–2.7 %; p \ 0.01). Inpatient complication rates (6.3 vs. 6.6 %; p = 0.21) and mortality (0.08 vs. 0.21; p = 0.72) were unchanged. On multivariate analysis, patients older than 79 years were three times as likely to develop a complication (odds ratio [OR] 3.1; 95 % CI 2.1–4.5) as were patients with a Charlson score over 2 (OR 3.1; 95 % CI 2.2–4.3).
L. M. Funk (&) A. Kanji W. Scott Melvin K. A. Perry Department of Surgery, Center for Minimally Invasive Surgery, The Ohio State University Medical Center, Columbus, OH, USA e-mail: [email protected]
123
Conclusions Today’s antireflux surgery patient population is a higher-risk cohort, but complication rates have remained stable and inpatient mortality has declined more than 50 % over the past decade. Given these findings, additional research is needed to understand why antireflux surgery is underutilized, with a decline of more than twothirds since its peak in 1999. Keywords Minimally invasive surgery Gastroesophageal reflux disease Laparoscopic antireflux surgery
With the widespread application of laparoscopic surgery during the 1990s, the volume of antireflux surgery for the treatment of gastroesophageal reflux disease (GERD) increased dramatically. The initial population-based study of antireflux surgery in the US found that the total number of antireflux operations performed annually nearly quadrupled between 1990 and 1997, when nearly 24,000 operations (12 cases/100,000 adults) were performed [1]. This corresponded with a massive increase in the utilization of laparoscopy, from 1 % of antireflux procedures in 1990 to 64 % in 1997. Subsequent analyses found that antireflux surgery volume peaked in 1999 at 16 cases per 100,000 adults, declined 30 % over the next 4 years to 11 per 100,000 adults, and continued to gradually decline until at least 2006 [2, 3]. The factors driving this decline in surgical volume included publication of a prominent randomized controlled trial that questioned the benefit of surgery compared with medical therapy, concerns regarding the side effects of antireflux surgery, and the availability of overthe-counter proton pump inhibitors (PPIs) [4, 5]. Current national antireflux surgery volumes are unknown, and changes in both the prevalence of GERD
Surg Endosc (2014) 28:1712–1719
and perception of the utility of long-term PPI use may have impacted physicians’ treatment strategies over the past decade. While the prevalence of GERD has risen significantly due to multiple factors, including the obesity epidemic and an aging population, concerns regarding complications of long-term PPI use, such as osteoporosis and enteric infections, have increased [6–8]. Meanwhile, surgical outcomes have continued to improve with the evolution of minimally invasive techniques [9]. Several randomized controlled trials published within the past decade have also demonstrated durable reflux control following laparoscopic surgery for refractory GERD [10–13]. In light of these changes in GERD epidemiology and treatment, the primary aim of this study was to characterize national trends in the utilization of antireflux surgery in the US between 2005 and 2010. We also sought to assess trends in the outcomes of patients undergoing elective laparoscopic antireflux surgery and identify predictors of postoperative complications.
Materials and methods Data sources We used the Nationwide Inpatient Sample (NIS) from 2005 through 2010 to generate our patient cohort. Managed by the Healthcare Cost and Utilization Project (HCUP) and sponsored by the Agency for Healthcare Research and Quality, the NIS contains all-payer data on inpatient hospitalizations from participating states (representing [95 % of the US population). Data comprising approximately 8 million hospitalizations from 1,000 hospitals are included annually to create a 20 % stratified probability sample of hospitals in the US. Sampling probabilities are based on the number of hospitals in each stratum. Strata are generated based on five hospital characteristics: ownership/control, number of hospital beds, teaching status, urban versus rural location, and region. Post-stratification weights provided by HCUP are used to calculate national estimates for all nonfederal hospitals in the US [14]. National census data were used to calculate rates of surgery per 100,000 people [15].
1713 Table 1 Inclusion and exclusion diagnoses and procedures with associated ICD-9 codes ICD-9 procedure codes Open creation of esophagogastric sphincteric competence (inclusion) Laparoscopic creation of esophagogastric sphincteric competence (inclusion) ICD-9-CM diagnosis codes
44.66 44.67
Malignant neoplasm of the esophagus (exclusion)
150.0–150.9
Achalasia and cardiospasm (exclusion)
530.0
Reflux esophagitis (inclusion)
530.11
Esophageal reflux (inclusion)
530.81
Diaphragmatic hernia with gangrene (exclusion)
551.3
ICD-9-CM International Classification of Diseases, 9th Revision, Clinical Modification
Study variables Patient demographic variables included age, gender, race, comorbid status, payment type (private, medicare, medicaid, or other), and income level (low, lower-middle, uppermiddle, or high). Comorbid status was characterized using Charlson scores for each patient based on ICD-9-CM diagnosis codes. The Charlson score represents a weighted composite measure of overall comorbid status that has been validated for use with administrative data [16]. Hospital characteristics included location (urban or rural), teaching status, size (small, medium, or large), and hospital antireflux procedure volume (high or low, based on the median number of fundoplications performed annually). The median number of laparoscopic operations performed annually at the hospital level was 12 (2005, 2007), 13 (2006, 2008), 15 (2009), and 17 (2010). The primary outcome measure was the occurence of a postoperative complication following laparoscopic antireflux surgery. This was assessed as a composite variable that included reoperation, stroke, myocardial infarction, respiratory failure, small-bowel obstruction, hemodialysis, splenic injury, bleeding, blood transfusion, deep venous thrombosis, wound complications, percutaneous drain placement, or shock. Secondary outcomes included inpatient mortality and length of inpatient stay.
Study population Statistical analyses We used International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure and diagnosis codes to identify all hospitalizations during which an adult patient underwent an elective antireflux surgery. All patients had a diagnosis of either GERD or reflux esophagitis. Patients with a diagnosis of esophageal cancer, achalasia or hiatal hernia with obstruction/gangrene were excluded from the analysis (Table 1).
Analyses of patient demographics, hospital characteristics, and surgical outcomes were performed for all patients who underwent laparoscopic antireflux surgery. We used the Rao–Scott v2 test to compare patient demographics, hospital characteristics and outcomes by year, and accounted for clustering at the hospital level in these analyses [17]. Multivariate logistic regression modeling was used to
123
1714
Surg Endosc (2014) 28:1712–1719
Table 2 Annual case volume estimates of elective fundoplication for patients with GERD/esophagitis in the US between 2005 and 2010 Total
2005
2006
2007
2008
2009
2010
Open and laparoscopic cases
96,702
15,819
15,325
15,026
14,684
17,068
18,780
Total cases per 100,000 people
5.3
5.3
5.1
5.0
4.8
5.6
6.1
Laparoscopic cases
78,233
12,557
12,405
12,371
12,121
13,548
15,231
Percent performed laparoscopically
81.6
79.4
80.9
82.3
82.5
79.4
81.1
p Value
0.54
GERD gastroesophageal reflux disease
Fig. 1 Annual number of antireflux operations performed in the US, 2005–2010
characterize the relationship between patient and hospital factors and postoperative complications. Characteristics with a p value of \0.05 on univariate analysis were selected for inclusion in the regression model along with hospital volume, which was included based on the a priori hypothesis of a correlation between hospital volume and surgical outcome. Race was not included in the model due to the high proportion of missing data (33 % of patients). Generalized estimating equations were used to account for clustering of patients by hospital [18]. All p values were two-sided, and p values \0.05 were considered statistically significant. Data were analyzed using SAS version 9.3 (SAS Institute, Cary, NC, USA).
esophageal cancer, achalasia, or obstruction/gangrene. The annual national volume of elective fundoplications decreased from 15,819 (5.3 cases per 100,000 people) in 2005 to a nadir of 14,684 (4.8 cases per 100,000 people) in 2008. In 2010, a total of 18,780 elective fundoplications (6.1 cases per 100,000 people) were performed (Fig. 1). Of the 96,702 elective fundoplications, 81.6 % underwent a laparoscopic fundoplication and 19.4 % had an open procedure (p \ 0.01). There was no statistically significant difference in the proportion of cases performed laparoscopically during the 6-year period (p = 0.54). The frequency of laparoscopy compared with an open approach was similar in urban (80.5 %) and rural (83.4 %; p = 0.11) hospitals and teaching (80.7 %) versus non-teaching (81.1 %; p = 0.84) hospitals.
Results
Patient characteristics
National volume estimates
The proportion of patients older than 65 years of age who underwent a laparoscopic antireflux operation increased from 21.1 % in 2005 to 30.9 % in 2010 (p \ 0.01), while the percentage with a Charlson score [2 increased from 1.2 % in 2005 to 2.7 % in 2010 (p \ 0.01; Table 3). The proportion of patients with private insurance decreased between 2005 (62.7 %) and 2010 (52.8 %; p = 0.02), while the percentage of Medicare patients increased during
Annual case volume estimates of elective fundoplication for patients with reflux in the US between 2005 and 2010 are summarized in Table 2. Overall, 103,702 patients with a diagnosis of GERD or reflux esophagitis underwent an elective fundoplication between 2005 and 2010; 7,000 were excluded from this analysis due to a diagnosis of
123
Surg Endosc (2014) 28:1712–1719
1715
Lowermiddle
30.4
30.4
24.8
31.2
28.2
29.4
after laparoscopic antireflux surgery according to the year (Table 4). Length of stay varied from a mean of 2.3 days in 2005 and 2007 to 2.5 days in 2008 and 2009 (p = 0.42). Complication rates ranged from 6.3 % in 2005 to 8.1 % in 2006 (p = 0.21), while mortality rates ranged from 0.08 % in 2005 and 2008 to 0.21 % in 2010 (p = 0.72). On univariate analysis, the complication rate increased with age and comorbidity status (Table 5). Patients older than 80 years of age had a 15.8 % complication rate compared with 4.6 % in those under 34 years of age (p \ 0.01). Patients with a Charlson score of 0 developed complications 5.2 % of the time, whereas those with a score greater than 2 had a 17.3 % complication rate (p \ 0.01). Young (18–50 years of age), healthy (Charlson score = 0) patients had the lowest complication rates (3.7 %), while elderly patients with significant comorbidities ([75 years of age, Charlson score [2) developed complications 24.6 % of the time (Fig. 2). Payer status was also associated with statistically significant differences in complication rates. Medicare patients experienced the highest complication rates (9.8 %) and privately insured patients experienced the lowest (5.5 %; p \ 0.01). No hospital characteristics, including procedure volume, were associated with statistically significant differences in complication rates.
Uppermiddle
25.2
27.5
26.2
26.1
27.2
26.8
Multivariate analysis
High
24.2
22.5
23.2
21.6
23.5
22.7
Table 3 Demographic data for patients undergoing elective laparoscopic antireflux surgery in the US between 2005 and 2010 2005
2006 2007
2008
2009
2010
p Value
Age (years, %) 18–34
13.5
11.7
11.0
9.4
9.0
9.1
35–49
30.2
28.8
27.1
25.0
24.5
22.7
50–64
35.2
35.4
35.4
36.9
36.9
37.2
65–79
19.5
21.2
23.6
24.6
26.1
27.0
1.6
2.9
2.8
4.0
3.5
3.9
64.6 12.5
65.1 10.2
66.6 16.2
68.5 12.8
67.5 14.2
68.5 14.7
0.02 0.03
