Monitoring quality of care between healthcare providers

Transcription

1 Brief Report Variations in Acute Hospital Stroke Care and Factors Influencing Adherence to Quality Indicators in 6 European Audits Silke Wiedmann, PhD; Steffi Hillmann, MPH; Sònia Abilleira, MD, PhD; Martin Dennis, MD; Peter Hermanek, MD, PhD; Maciej Niewada, MD, PhD; Bo Norrving, MD, PhD; Kjell Asplund, MD; Anthony G. Rudd, MD; Charles D.A. Wolfe, MD; Peter U. Heuschmann, MD, MPH; on behalf of the European Implementation Score Collaboration Downloaded from by guest on September 18, 2016 Background and Purpose We compared compliance with standards of acute stroke care between 6 European audits and identified factors associated with delivery of appropriate care. Methods Data were derived from stroke audits in Germany, Poland, Scotland, Catalonia, Sweden, and England/Wales/ Northern-Ireland participating within the European Implementation Score (EIS) collaboration. Associations between demographic and clinical characteristics with adherence to predefined quality indicators were investigated by hierarchical logistic regression analyses. Results In 2007/2008 data from patients with stroke were documented. Substantial variations in adherence to quality indicators were found; older age was associated with a lower probability of receiving thrombolytic therapy, anticoagulant therapy, or stroke unit treatment and a higher probability of being tested for dysphagia. Women were less likely to receive anticoagulant or antiplatelet therapy or stroke unit treatment. No major weekend effect was found. Conclusions Detected variations in performance of acute stroke services were found. Differences in adherence to quality indicators might indicate population subgroups with specific needs for improving care delivery. (Stroke. 2015;46: DOI: /STROKEAHA ) Key Words: health services research quality indicators, health care quality of healthcare stroke care Monitoring quality of care between healthcare providers might reveal current service needs and identify gaps in delivery of appropriate care. Valid comparisons between regions and healthcare systems might clarify factors driving implementation of research evidence into practice. We analyzed data from European acute stroke care audits to compare quality of care by predefined quality indicators (QIs) and to identify patient and clinical characteristics being associated with appropriate care delivery. Methods Data from national (German Stroke Register Study Group [ADSR], Germany; the Scottish Stroke Care Audit [SSCA], Scotland; the National Stroke Register in Sweden [Riks-Stroke], Sweden; the National Sentinel Audit of Stroke [NSSA], England/Wales/Northern- Ireland; and the Hospital Stroke Registry of National Program for Prevention and Treatment of Cardiovascular Diseases [POLKARD], Poland) and regional (Catalan Stroke Audit [CSA], Catalonia, Spain) audits cooperating within an EU FP7 project (European Implementation Score [EIS] Collaboration) were included with details described previously (Table I in the online-only Data Supplement). 1 Variables that were documented in a comparable way in 5 of the 6 audits were used for the analyses (definitions in Appendix in the online-only Data Supplement). QIs of acute hospital stroke care were defined a priori by a European consensus group within the European Implementation Score project. Compliance with QIs (definitions in Appendix in the online-only Data Supplement) was estimated for measures that could be calculated in 5 of the 6 audits including: Received September 19, 2014; final revision received November 12, 2014; accepted November 18, From the Institute of Clinical Epidemiology and Biometry (S.W., S.H., P.U.H.), Comprehensive Heart Failure Centre (S.W., P.U.H.), University of Würzburg, Würzburg, Germany; Stroke Programme/Agency for Health Quality and Assessment of Catalonia, Barcelona, Spain (S.A.); Department of Clinical Neurosciences, Western General Hospital, Edinburgh, United Kingdom (M.D.); Bavarian Permanent Working Party for Quality Assurance, Munich, Germany (P.H.); Department of Experimental and Clinical Pharmacology, Medical University of Warsaw, Warsaw, Poland (M.N.); Department of Clinical Sciences, Neurology, Lund University, Lund, Sweden (B.N.); Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden (K.A.); and Division of Health and Social Care Research (A.G.R.) and National Institute for Health Research Biomedical Research Centre Guy s & St Thomas NHS Foundation Trust (C.D.A.W.), King s College London, London, United Kingdom. Guest Editor for this article was Eric E. Smith, MD, MPH. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Silke Wiedmann, PhD, Institute of Clinical Epidemiology and Biometry, University of Würzburg, Würzburg, Germany. silke. wiedmann@uni-wuerzburg.de 2014 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 2 Stroke February 2015 Downloaded from by guest on September 18, 2016 stroke unit treatment; thrombolysis; dysphagia screening; antiplatelet therapy; and anticoagulation in atrial fibrillation. The effect of factors associated with delivery of appropriate care was estimated by hierarchical logistic regression analyses. For estimating odds ratios and resulting 95% confidence intervals, distinct multilevel models were built, taking into account clustering of patients within regions/countries. Cross-level interactions between first- level (patients) and second-level (country/region) variables were assessed. For estimating variations of adherence by center, standardized ratios were calculated by comparing the observed proportion of an indicator within a center with the proportion expected when factor-specific event rates in the respective audit derived from multiple logistic regression models were applied to the center-specific population. Results Audit and patient s characteristics are presented in Table I in the online-only Data Supplement; data collection details are presented in Table II in the online-only Data Supplement. Adherence to QIs varied substantially across audits (Table 1). Figure I in the online-only Data Supplement shows the observed adherence to QIs by the expected adherence based on the center-specific case-mix. Patient characteristics influenced delivery of appropriate care in univariable (Table III in the online-only Data Supplement) and multivariable analyses (Table 2). No significant crosslevel interactions were found for QIs between patient characteristics and region/country, including missing variables did not change associations substantially (data not shown). Discussion Substantial variations in adherence to QIs were identified between 6 audits in Europe, especially for thrombolysis, dysphagia screening, and anticoagulant therapy. Patient characteristics, such as age, sex, or stroke subtype, were identified that were associated with delivery of appropriate care. A consistent proportion of 3 quarters or more of patients were treated on a stroke unit, a higher proportion compared with reports from Canada. 2 Different stroke unit definitions between healthcare systems might limit comparability of findings. Substantial variations in tissue-type plasminogen activator -rates were found between audits ranging from 1.3% to 9.1%. The higher proportion of tissue-type plasminogen activator in previous studies might be because of differences in study populations (eg, considering exclusion criteria for tissue-type plasminogen activator). 3 The proportions of patients with ischemic stroke with atrial fibrillation and anticoagulant therapy at discharge ranged between 23.7% and 57.3% in our study. Substantial higher rates were observed in previous studies from the United States. 4 However, in contrast to our data, in the later study patients were excluded if a contraindication was documented. 4 Homogeneous patterns between proportions of appropriate healthcare delivery observed and proportions expected within centers were documented for antiplatelet and anticoagulant therapy, indicating that these measures might have already been translated successfully into clinical practice. Similar to our data, previous studies have shown that older patients receive less often anticoagulation if they experience atrial fibrillation 5,6 and more often a swallow screening. 7 This might be because of the fact that older age is associated with more severe strokes being associated with higher risk of complications. In accordance with our findings, other studies have found sex differences in diagnosis and treatment. 8 The study was designed retrospectively; therefore, no common definitions and methods were used for data collection. Potential selection biases might have contributed to our findings as in some audits participation was voluntary. No uniform criteria were applied for ensuring completeness of case ascertainment. We cannot exclude that some of our findings were caused by potential confounders. No outcome data, such as in-hospital mortality, could be calculated because of limited data availability. Data were derived from 2007 to 2008, and current treatment patterns within regions/audits might be substantially different. Differences in adherence to QI might indicate population subgroups with specific needs for improving care delivery. Acknowledgments German Stroke Register Study Group (Misselwitz, Seidel, Bruder, Berger, Hoffmann, Matthis, Janssen, and Burmeister); Hospital Stroke Registry of National Program for Prevention and Treatment of Cardiovascular Diseases (Członkowska, Ryglewicz, Sarzyńska- Długosz, and Skowrońska), Scottish Stroke Care Audit (Murphy, Dodds, McLeod, Langhorne, and Barber); Catalan Stroke Audit (Gallofré); Riks-Stroke (Jonsson); National Sentinel Audit of Stroke (Hoffmann); and Quality Register of Flemish Hospital Network of the KU Leuven, Flanders (Thijs). Table 1. Acute Treatment, Management, and Prevention Procedures During Hospitalization, 2007 to 2008* ADSR POLKARD SSCA CSA Riks-Stroke NSSA Brain imaging (CT or MRI), % Stroke unit treatment, % Swallowing test done, % Thrombolysis, % Antiplatelet therapy, % Anticoagulant therapy, % Length of stay, mean (SD), d 10.7 (8.4) 12.6 (10.2) 26.5 (34.0) 10.8 (10.5) 16.7 (17.9) 23.4 (27.8) ADSR indicates German Stroke Register Study Group; CSA, Catalan Stroke Audit; NSSA, National Sentinel Audit of Stroke; POLKARD, Hospital Stroke Registry of National Program for Prevention and Treatment of Cardiovascular Diseases; and SSCA, Scottish Stroke Care Audit. *Analysis restricted to patients with ischemic stroke, intracranial hemorrhage, and stroke not known whether ischemic or hemorrhagic and without missing values. Variable not documented or not documented in a comparable way.

3 Wiedmann et al Quality of Stroke Care in 6 European Audits 3 Table 2. Multivariate Analyses: Patients Receiving Appropriate Care by Patient and Clinical Characteristics, 2007 to 2008* Downloaded from by guest on September 18, 2016 Thrombolysis Dysphagia Screening Anticoagulation Antiplatelet Therapy Stroke Unit OR (95% CI) P Value OR (95% CI) P Value OR (95% CI) P Value OR (95% CI) P Value OR (95% CI) P Value Age group, y < < < < < < ( ) ( ) 1.12 ( ) 1.18 ( ) 85 n.a ( ) Sources of Funding This study was supported by EU FP7 (European Implementation Score Collaboration [EIS]; No ). Disclosures Dr Heuschmann received research support from the European Union (EIS; No ). The other authors report no conflicts. References 1. Wiedmann S, Norrving B, Nowe T, Abilleira S, Asplund K, Dennis M, et al. Variations in quality indicators of acute stroke care in 6 European countries: the European Implementation Score (EIS) Collaboration. Stroke. 2012;43: doi: /STROKEAHA Kapral MK, Laupacis A, Phillips SJ, Silver FL, Hill MD, Fang J, et al. Stroke care delivery in institutions participating in the Registry of the Canadian Stroke Network. Stroke. 2004;35: Reeves MJ, Grau-Sepulveda MV, Fonarow GC, Olson DM, Smith EE, Schwamm LH. Are quality improvements in the get with the guidelines: stroke program related to better care or better data documentation? Circ Cardiovasc Qual Outcomes. 2011;4: doi: / CIRCOUTCOMES ( ) 0.48 ( ) 0.18 ( ) 1.19 ( ) 1.17 ( ) 1.11 ( ) 0.97 ( ) 0.91 ( ) 0.73 ( ) Sex < Women 1.02 ( ) 0.98 ( ) 0.82 ( ) 0.92 ( ) 0.95 ( ) Day of admission 0.08 < Weekend 1.04 ( ) 1.04 ( ) 0.97 ( ) 0.98 ( ) 0.96 ( ) Year of admission < < < ( ) 1.52 ( ) 1.10 ( ) 1.07 ( ) 1.03 ( ) Atrial fibrillation < < < < Yes 1.41 ( ) 1.23 ( ) n.a ( ) 1.10 ( ) Stroke subtype < < Ischemic stroke n.a. 1 n.a. n.a. 1 Intracranial hemorrhage Unclassified/ unknown n.a ( ) n.a ( ) n.a. n.a ( ) n.a. n.a ( ) CI indicates confidence; n.a., not applicable; and OR, odds ratio. *Hierarchical multilevel model. Ischemic stroke patients aged years. In patients with ischemic stroke, intracranial hemorrhage, and stroke not known whether ischemic or hemorrhagic. Anytime during stay or at discharge or recommended (not available in all audits) at discharge in patients with ischemic stroke and atrial fibrillation alive at discharge. Anytime during stay or at discharge or recommended at discharge in patients with ischemic stroke alive at discharge without anticoagulant therapy. 4. Lewis WR, Fonarow GC, Grau-Sepulveda MV, Smith EE, Bhatt DL, Hernandez AF, et al. Improvement in use of anticoagulation therapy in patients with ischemic stroke: results from Get With The Guidelines-Stroke. Am Heart J. 2011;162: e2. doi: /j. ahj Fonarow GC, Reeves MJ, Zhao X, Olson DM, Smith EE, Saver JL, et al; Get With the Guidelines-Stroke Steering Committee and Investigators. Age-related differences in characteristics, performance measures, treatment trends, and outcomes in patients with ischemic stroke. Circulation. 2010;121: doi: / CIRCULATIONAHA Ferrari J, Knoflach M, Seyfang L, Lang W; Austrian Stroke Unit Registry Collaborators. Differences in process management and in-hospital delays in treatment with iv thrombolysis. PLoS One. 2013;8:e doi: /journal.pone Saposnik G, Black SE, Hakim A, Fang J, Tu JV, Kapral MK; Investigators of the Registry of the Canadian Stroke Network (RCSN); Stroke Outcomes Research Canada (SORCan) Working Group. Age disparities in stroke quality of care and delivery of health services. Stroke. 2009;40: doi: /STROKEAHA Smith MA, Lisabeth LD, Brown DL, Morgenstern LB. Gender comparisons of diagnostic evaluation for ischemic stroke patients. Neurology. 2005;65: doi: /01.wnl f.

4 Downloaded from by guest on September 18, 2016 Variations in Acute Hospital Stroke Care and Factors Influencing Adherence to Quality Indicators in 6 European Audits Silke Wiedmann, Steffi Hillmann, Sònia Abilleira, Martin Dennis, Peter Hermanek, Maciej Niewada, Bo Norrving, Kjell Asplund, Anthony G. Rudd, Charles D.A. Wolfe and Peter U. Heuschmann on behalf of the European Implementation Score Collaboration Stroke. published online December 30, 2014; Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 2014 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: Data Supplement (unedited) at: Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Stroke is online at:

5 SUPPLEMENTAL MATERIAL Tables: 3 Figures: 5 Word count: 1595

6 Table I: Audit and patient characteristics, * Audit characteristics No of patients, n (%) ADSR POLKARD SSCA CSA Riks-Stroke NSSA Total 222,327 30,060 14,362 1,718 49,286 11, ,232 (39.2) 24,935 (82.9) 6,870 (47.8) 1,718 (100.0) 24,492 (49.7) # ,095 (60.8) 5,213 (17.1) 7,492 (52.2) # 24,794 (50.3) 11,369 (100.0) No of centers per year, n (%) Total No of patients per centre per year, median (IQR) (72-436) 181 ( ) 339 ( ) # 262 ( ) # (72-482) 39 (19-62) 382 ( ) # 283 ( ) # No of centers providing tpa, per year, n (%) Total 381 (63.6) 51 (41.1) 20 (45.5) 77 (97.5) 72 (33.3) (65.2) 48 (39.0) 20 (45.5) 72 (92.3) (63.0) 17 (16.8) 0 74 (93.7) 72 (33.3)

7 Patient characteristics* Age, median (IQR), y 75 (66-82) 73 (63-80) 76 (66-83) 77 (68-84) 78 (69-85) 78 (69-85) Age group, y < Female Sex, % Dependency pre stroke, % Stroke subtype, % Ischemic stroke Intracerebral haemorrhage Unknown/ undefined Interval onset admission <3 hours, % Admission at the weekend, % Atrial fibrillation, % *Patients with IS, ICH or UND only; patients with missing center allocation or with missing/ default year of admission were not considered; no of centers refer to participating hospital, trusts or departments; analyses restricted to patients without missing values in the respective variable; # for these audits, a predefined number of patients was included based on the center volume, thus, no median number of patients was calculated; no complete year of documentation; variable not documented in a comparable way

8 Table II: Data collection across participating audits; Audit Country Participation Participating hospitals Data collection Documentation period German Stroke Register Study Group [ADSR] Germany Voluntary; mandatory in some regions and for SU Selected acute care hospitals cooperating within regional stroke register in different regions across Germany Continuously, consecutive patients (IS, ICH, SAH, TIA, UND) 2007 (8 registers)-2008 (9 registers) POLKARD Hospital Stroke Registry [POLKARD] Poland Voluntary Selected acute care hospitals across Poland Continuously, consecutive patients (IS, ICH, SAH, UND) (not whole calendar years) Scottish Stroke Care Audit [SSCA] Scotland, UK Mandatory All acute hospital trusts, Scotland* Continuously (IS, ICH, TIA, UND) Catalan Stroke Audit [CSA] Autonomous Community of Catalonia, Spain Mandatory All publicly financed acute hospitals in Catalonia Predefined number of consecutive patients per hospital (IS, ICH, UND ) 2007 National Stroke Register in Sweden [RIKS-Stroke] Sweden Voluntary All acute hospitals in Sweden Continuously (IS, ICH, UND) National Sentinel Audit of Stroke [NSSA] England/ Wales/ N-Ireland, UK Voluntary All acute hospitals in England, Wales, N. Ireland Predefined number of consecutive patients (IS, ICH, UND) 2008 *Due to technical reasons, some hospitals were combined based on the data extraction process in Scotland; one NHS board was not included in the analysis due to differences in the data collection system

9 Table III: Univariate analysis: patients receiving appropriate care by patient and clinical characteristics range between audits, * Thrombolysis # Dysphagia Anticoagulation if Antiplatelet Stroke Unit Screening AF therapy Audit range, % Audit range, % Audit range, % Audit range, % Audit range, % Age group, y < (75-80) $ n. a Sex Male Female Day of admission Weekday Weekend Year of admission Pre-stroke risk factors No Atrial fibrillation Atrial fibrillation Stroke subtype Ischemic stroke Intracranial haemorrhage n. a n. a. n. a Unclassified/unknown n. a n. a. n. a *Analysis restricted to patients with ischemic stroke, intracerebral haemorrhage or undefined stroke and without missing values in the respective variables; audits not documenting a respective variable in a comparable way were excluded; n.a. not applicable as quality indicator did include the specific subgroup; $ oldest age group restricted to 80 years; # ischemic stroke patients age y; patients with ischemic stroke, intracerebral haemorrhage or undefined stroke; anytime during stay or at discharge or recommended (not available in all audits) at discharge in patients with ischemic stroke and atrial fibrillation alive at discharge; anytime during stay or at discharge or recommended at discharge in patients with ischemic stroke alive at discharge and without anticoagulant therapy.

10 Figure I: observed/ expected ratio by center for patients being treated at a SU O/E SU care

11 Figure II: observed/ expected ratio by center for patients receiving thrombolysis O/E Thrombolysis

12 Figure III: observed/ expected ratio by center for IS patients with AF receiving anticoagulant therapy O/E Anticoagulation in AF ADSR POLKARD SSCA CSA RIKS-Stroke NSSA

13 Figure IV: observed/ expected ratio by center for patients being tested for swallowing disorders O/E Dysphagia Screening

14 Figure V (online only): observed/ expected ratio by center for patients being receiving antiplatelet therapy O/E Antiplatelet therapy ADSR POLKARD SSCA CSA RIKS-Stroke NSSA

15 Figure legend I to V For estimating variations of adherence to predefined QI by center, standardized ratios on the center level were calculated by comparing the observed proportion of an indicator within a center with the proportion expected when factor-specific event rates in the respective audit derived from multiple logistic regression models were applied to the center-specific population.

16 METHODS Data definitions Demographics: age; sex; dependency pre-stroke (dependent, independent); stroke subtype (ischemic stroke [IS], intracerebral haemorrhage [ICH], undefined [UND]); admission: time interval between onset and admission ( 3h, 3h/missing); day of admission (weekend; weekday); comorbidities/risk factors: atrial fibrillation [AF] (No/Yes); process of care: brain imaging (No/Yes); intravenous treatment with tissue-plasminogen-activator [rt-pa] (No/Yes); Stroke Unit [SU] treatment (No/Yes); testing for swallowing disorders (NO/YES or unassessable if documented); antiplatelet therapy during hospital stay (No/Yes); anticoagulant therapy during hospital stay or recommended at discharge (No/Yes); length of hospital stay (days). Quality Indicators The following QI and definitions were used: SU treatment (numerator: number of stroke patients treated on a SU; denominator: all patients with IS, ICH and UND; available in the following audits: POLKARD, SSCA, CSA, Riks-Stroke, NSSA); thrombolysis (numerator: number of stroke patients treated with intravenous thrombolysis; denominator: all patients with IS, aged 18 to 80 years; available in the following audits: ADSR, POLKARD, SSCA, CSA, Riks-Stroke, NSSA); dysphagia screening (numerator: number of stroke patients screened for swallowing disorders and patients that are unassessable (if documented); denominator: all patients with IS, ICH and UND; available in the following audits: SSCA, CSA, Riks-Stroke, NSSA); antiplatelet therapy (numerator: number of patients treated with antiplatelet therapy anytime during stay or recommended at discharge; denominator: all patients with IS alive and without anticoagulants at discharge; available in the following audits: ADSR, POLKARD, SSCA, CSA, Riks-Stroke, NSSA); anticoagulation if AF (numerator: number of patients treated with anticoagulants at discharge or recommended at discharge; denominator: all patients with IS and AF alive at discharge; available in the following audits: ADSR, POLKARD (without recommendation at discharge), SSCA, CSA, Riks-Stroke, NSSA). Statistical analyses Variations in acute stroke care between the participating audits were assessed by calculating adherence to predefined quality indicators with definitions described above. The impact of factors associated with delivery of appropriate care within the audit data sets was estimated by hierarchical logistic regression analyses. Age group, sex, day of admission, comorbidities (atrial fibrillation), stroke subtype, and time from onset to admission (for the indicator thrombolytic therapy) were investigated as patient and clinical characteristics potentially influencing quality of care. For estimating odds ratios (OR) and resulting 95% confidence intervals (CI) for these factors, distinct multilevel models were built for each of the defined QI, taking into account clustering of patients within audits. This approach allows the simultaneous examination of the effects of the individual patient and clinical characteristics (first-level) and the region/country (second-level). A mixed model with a random region/country effect was applied modeling the random region effect as a G-side effect using the SAS procedure GLIMMIX. At estimation the variance matrix is blocked by region/country, which means the marginal variance matrix is assumed block-diagonal and patients from the same region/country form the blocks. We investigated whether strengths and direction of patient-level factors depends on the second level factor (region/country). For this we added terms for random cross-level interactions between first-level (age, sex, weekday of admission, atrial fibrillation, stroke subtype and time from onset to hospitalization (if applicable)) and second-level (region/country) variables. With this we used statistical tests for

17 significant differences between the effects of a first-level factor between regions/countries. For estimating variations of adherence to predefined QI at the center level within the audits, we also calculated standardized ratios on the center level (trust, hospital or department) for the defined QI. For this purpose, an indirect standardization method was used comparing the observed proportion in a respective center with the proportion expected when factor-specific event rates in the respective country or region of the center applied to the specific center population. The respective expected proportions were derived from multiple logistic regression models adjusted for age, sex, weekday of admission, stroke subtype (if applicable), time from onset to admission (for thrombolysis), and atrial fibrillation. The sum and the mean of the predicted outcome probabilities was calculated for each hospital within a region/country providing the denominator for the observed to expected ratio. Analyses for estimating the observed/ expected ratio were restricted to hospitals with at least 10 complete cases per hospital for the QI thrombolysis, dysphagia screening, SU treatment and antiplatelet therapy or at least 5 complete cases per hospital for the quality indicator anticoagulation. Graphical presentation of O/E ratios was restricted to a ratio of four for thrombolytic therapy. All analyses were restricted to patients without missing values. The number of patients with missing values ranged for example from 0% for age, sex, or AF in at least 3 of the audits up to 38.9% for dependency pre-stroke in ADSR. Sensitivity analyses were performed to assess the potential impact of missingness, such as full data set analysis with missing values as own category, showing no substantial variations to complete case analyses. All tests were twotailed and statistical significance was determined at an alpha level of Statistical analyses were performed using the SAS 9.2 Software Package.