Copyright ©ERS Journals Ltd 2007 The correlation between asthma control and health status: the GOAL study1 University of Cape Town, Cape Town, South Africa., 2 Arnaud de Villeneuve Hospital, Montpellier, France., 3 GlaxoSmithKline R&D, Greenford, and 5 Imperial College, London, UK. 4 University of Wisconsin Medical School, Madison, WI, USA. 6 University of Southern Denmark, Kolding Hospital, Kolding, Denmark. CORRESPONDENCE: E. D. Bateman, UCT Lung Institute, P.O. Box 34560, Groote Schuur 7937, Cape Town, South Africa. Fax: 27 214066902. E-mail: ebateman{at}uctgsh1.uct.ac.za Keywords: Asthma, control, fluticasone propionate, health status, quality of life, salmeterol
Received: November 3, 2005
The present study examined the association between guideline-derived asthma control and health-related quality of life, assessed using the Asthma Quality of Life Questionnaire (AQLQ), in patients with uncontrolled asthma whose treatment was directed towards achieving the highest possible level of control. The present randomised, double-blind, parallel-group study compared the efficacy of fluticasone propionate (FP) and salmeterol/fluticasone propionate combination (SFC) in achieving two composite, guideline-derived measures of control: total control (TC) and well-controlled (WC) asthma. Not achieving these levels was classed as not well-controlled (NWC). Doses were augmented until patients achieved TC or reached the maximum dose. This dose was maintained for the remainder of the study. AQLQ was assessed at baseline and at each clinic visit. AQLQ scores improved throughout the study, reaching near-maximal levels in patients achieving TC and WC, and 52-week mean scores in the three control groups were statistically significantly different. Clinically meaningful improvements (mean change from baseline) were: TC group (SFC 1.9, FP 1.8), WC (SFC 1.5, FP 1.5) and NWC (SFC 1.0, FP 0.9). In conclusion, the treatment aimed at controlling asthma improves the health-related quality of life to levels approaching normal. The difference in Asthma Quality of Life Questionnaire scores between total control and well-controlled confirms that patients distinguish even between these high levels of control. Guidelines for the management of asthma issued by the Global Initiative for Asthma (GINA)/National Institutes of Health (NIH) state that the therapeutic aim should be to achieve overall asthma control in order to minimise the impact of asthma on the individual patient 1, 2. However, it is also increasingly recognised that asthma patients have low expectations of their therapy, leading to an acceptance of a lower level of asthma control than might be achievable 3, 4. Assessment of health-related quality of life (HRQoL) alongside conventional clinical monitoring is increasingly proposed as a means of aligning patient expectations with the clinician's therapeutic goals 5. However, a number of studies 68 have demonstrated poor correlation between conventional clinical indices and the outcomes of the Asthma Quality of Life Questionnaire (AQLQ). Conversely, studies that have used more comprehensive measures of overall asthma control have found that achieving asthma control translates into significant improvements in AQLQ score 9, 10. Additionally, a study by Katz et al. 11 found that perceived control of asthma was strongly associated with improvements in both asthma-specific and generic health status outcomes. Based on retrospective analyses of the results of efficacy trials in asthma, the present authors have previously suggested that in contrast to conventional end-points of clinical trials, such as forced expiratory volume in one second (FEV1), the use of a composite measure incorporating a range of clinically relevant end-points provides a more complete view of the overall level of asthma control for the individual patient 12 and is likely to correlate with patient perception of control or freedom from disease 13. The Gaining Optimal Asthma controL (GOAL) study was designed to prospectively investigate whether, and in what proportion of patients, asthma control evaluated according to a rigorous composite measure derived from the GINA/NIH guidelines can be achieved. A further aim of the GOAL study was to compare the efficacy of individualised increasing doses of the two recommended controller therapies, fluticasone propionate (FP) alone or in combination with the long-acting ß2-agonist salmeterol, in achieving asthma control. The primary efficacy results from the GOAL study have been published in detail elsewhere 14. Some of the results of the AQLQ analysis have previously been presented in abstract form 15. The present analysis of the results examines the extent to which patients distinguish between the different levels of clinical asthma control achieved during the study using this diseasespecific health status measure.
Study design Full details of the GOAL study design and patient population have been reported elsewhere 14 and are summarised here. GOAL was a 52-week, randomised, double-blind, multicentre, stratified, parallel-group step-up study designed to compare the level of asthma control achieved in adults and adolescents with salmeterol/fluticasone propionate combination (SFC; Seretide®/Advair®; GlaxoSmithKline, Middlesex, UK) via Diskus® (Accuhaler®) dry powder inhaler and FP (Flixotide®/Flovent®; GlaxoSmithKline) alone, also via Diskus® (Accuhaler®).
Following a 4-week run-in, eligible patients were allocated to one of the following three strata based on their dose of inhaled corticosteroid (ICS) during the previous 6 months. Stratum 1: no ICS; stratum 2: Rigorous composite measures derived from the treatment goals of the GINA/NIH guidelines 1, 2 were used to define asthma control: TC or well-controlled (WC; the full criteria are reported elsewhere 14). When neither measure was achieved, the outcome was defined as not well-controlled (NWC).
Patient population AQLQ data was available from patients in 16 of the 44 countries involved in the study. Only those countries for which a validated translation in the local language was available were eligible for inclusion. All participants gave written informed consent prior to inclusion. The study was approved by the local research ethics committees.
Assessment of quality of life
Statistical analysis
The changes from baseline in AQLQ scores for each domain and the overall AQLQ score were plotted over the 1 yr treatment period for each stratum. A Chi-squared test was used to analyse the association of treatment with change from baseline in AQLQ score (
The changes from baseline in overall AQLQ scores at week 52 were categorised into An ANOVA model was fitted to AQLQ scores at 52 weeks, with the sole predictor variable being control status at the end of phase II (52 weeks). The p-values for all pair-wise differences in control status were calculated. To adjust for all the multiple comparisons, the Bonferroni correction method was used, which increases the p-value to account for the increased risk of incorrectly rejecting the null hypothesis. Unlike predictor variables traditionally used in ANOVA models, control status is not randomised and was not measured at any time before the AQLQ measurement.
The absolute AQLQ scores at 52 weeks were categorised and a two-sided Fisher's Exact test was used to analyse the association of treatment with AQLQ score (<6 versus
Patient demographics The total ITT population for the GOAL study comprised 3,416 patients. The baseline demographics, clinical characteristics and primary efficacy results of the overall GOAL population, including AQLQ scores achieved in each stratum, have been described elsewhere 14. A total of 1,994 patients (SFC n = 1,001; FP n = 993) in the ITT population completed the AQLQ at least once during the study. The demographics of the AQLQ population were comparable to those of the overall study population (table 1
Improvements in quality of life A significantly higher number of patients treated with SFC compared with FP in each stratum achieved either WC or TC status in each phase of the study (p 0.039) including at study end (52 weeks; table 2 0.5 (80% with SFC and 75% with FP; p<0.01). A total of 16 and 18% of patients achieved improvements 0.5<1.0 with SFC and FP, respectively; with 19 and 17% achieving improvements of 1.0<1.5, and 45 and 39% achieving improvements 1.5. A nonclinically meaningful improvement (>0<0.5) was achieved by 12 and 14% of SFC and FP patients, whilst 8 and 11% achieved no change or deterioration in quality of life indicated by a negative change in AQLQ score. However, the proportions of patients experiencing these different levels of change in AQLQ were similar in the three individual strata (stratum 13; table 3
In addition, there was a significant association between treatment and the proportion of patients with week-52 AQLQ scores of 6 versus <6. More patients in the SFC group achieved an AQLQ score 6 (minimal or no impairment) compared with those receiving FP (p<0.001). Across all strata, the proportions were 61 versus 52% for SFC versus FP, respectively (fig. 1 6 were 63 versus 62% (NS; stratum 1), 64 versus 53% (p<0.005; stratum 2) and 57 versus 45% (p<0.005; stratum 3) for SFC versus FP, respectively.
Relationship between level of asthma control and quality of life Mean values for the final score were significantly higher in patients achieving TC than in those with WC asthma (p<0.001), and between those with WC and NWC asthma (p<0.001, table 4 0.5 unit change were higher in patients with TC (SFC 89% and FP 85%) and WC status (SFC 85% and FP 84%) compared with those NWC (SFC 67% and FP 65%; fig. 2 1.0 (SFC 50% and FP 47%), with 31% in each treatment group achieving changes in AQLQ score 1.5 (fig. 2
Profile of improvements in AQLQ: all strata The largest improvement in overall score and in the scores for each domain was observed during the first 4 weeks of the treatment period. However, scores continued to improve throughout the study period, with highest values for each treatment being observed at 52 weeks (fig. 3
The GOAL study is the first prospective study to evaluate the concept of achieving complete clinical control, defined in GOAL as TC, based on the goals of treatment described in international treatment guidelines 1, 2. The AQLQ is a disease-specific, self-administered quality-of-life tool that is available in 36 languages and has been shown to be valid, reliable and reproducible for evaluating the impact of treatment regimens on the quality of life of asthma patients 1618. The unique design of the GOAL study permits evaluation of the relationship between asthma control and health status as measured using this AQLQ questionnaire. TC is associated with achievement of near maximal levels of HRQoL. The final values for the AQLQ for patients achieving lesser levels of clinical control (i.e. WC and NWC status) were lower but still statistically significant, and exceeded the minimal clinically significant difference in a large majority of patients. Furthermore, there was a statistically significant difference in mean total AQLQ score at 52 weeks between TC and WC patients, confirming that patients (assessed using this instrument) distinguish even between these high levels of control in spite of the probable "ceiling" effect as large proportions of subjects in both categories scored the maximum score of 7. The difference between patients designated controlled and NWC by the definitions used in the present study has recently been used by Juniper et al. 19 to define cut-points for the Asthma Control Questionnaire (ACQ) for distinguishing "well-controlled" and "not well-controlled" asthma. Although in their analysis the definition of TC was not used, but grouped under well controlled, a cut-point of 1.5 was associated with a probability of having well-controlled asthma of only 66%. A score of 0.75 (the ACQ score is inverse to the level of control) increased the likelihood of control to 85%, suggesting that the highest levels of control can be distinguished by control measures.
A further important conclusion of the current study is that, even when the desired levels of control were not achieved, a great majority of patients benefited from the treatment approach, with most achieving clinically significant improvements in AQLQ. Indeed, many patients not achieving TC or WC status reached high scores on the AQLQ, regardless of baseline values and treatment received. At the end of the 52-week randomised period, virtually all patients from all strata had achieved at least moderate improvements in HRQoL, as defined by an increase in AQLQ score of Throughout the present study, the AQLQ score values for patients treated with SFC were higher than for those treated with FP, except for the suggestion of a ceiling effect as values approached maximal levels. Comparisons between studies of different design should be performed with care. However, the magnitude of the increase and end-of-study values in the present study were high, comparing favourably with values in the Formoterol And Corticosteroids Establishing Therapy (FACET) study 6, which was of similar duration. This is to be expected since the individualised treatment was increased in GOAL with the purpose of achieving the best possible level of control, whereas the FACET study employed only a single fixed dose of treatment and a "step-down" study design. The GOAL results confirm the findings of the earlier retrospective analysis by Bateman et al. 13, that guideline-derived control is associated with attainment of near-normal AQLQ scores. The main improvements were seen during the initial dose-titrating phase of the study, particularly in the first 4 weeks of treatment. However, further improvement in AQLQ score was observed throughout the remainder of the 52-week study, beyond the point at which patients received no further dose increase in the controller treatment. The plateauing of the values towards the end of the study may reflect the absence of further dose increases and the fact that no further benefit or a ceiling effect 6 was being achieved (the limits of efficacy) as more and more patients approached maximum scores. By contrast, in the FACET study an initial large increase in AQLQ score was followed by a gradual decline over the remainder of the 1-yr study period, suggesting gradual loss of control 6. In the GOAL study, all three strata showed similar improvements in each of the four AQLQ domains. The exception was the greater improvement in the symptoms domain in stratum 1. It is reasonable to assume that the greatest impact of achieving control, as per the composite measure employed in the GOAL study, might have been in the symptoms domain because the parameters within the composite measure tend to be symptom-based. However, it is important to note that, in all strata, comparable improvement occurred in all AQLQ domains, even those not represented in the composite measure. This supports the view that the composite measure of control used in the GOAL study provides a simple measure that reflects a patient-reported outcome, such as the AQLQ. The absence of a placebo group is a potential limitation in the design of the GOAL study, with respect to HRQoL, which may restrict its validity in a wider patient population. For ethical reasons, it was not acceptable to include a placebo arm in a study of patients with uncontrolled asthma, of whom the majority in strata 2 and 3 had severe asthma. It seems improbable that spontaneous improvements could account for the high AQLQ scores at the end of the study. Other potential limitations are that no record was made of overall patient satisfaction with treatment and treatment approach, due to the current lack of validated and approved satisfaction instruments. In addition, AQLQ measurements were dependent on patient recall of the 2 weeks prior to the clinic visit. Since improving HRQoL is a slow process and changes may be subtle, there is a risk of perceived lack of progress and under-reporting of improvements, especially if patients have low expectations of their asthma treatments to begin with 3, 4. However, the clear, consistent and biologically plausible trends and correlations suggest that the results are reliable. Quality-of-life instruments, such as the Asthma Quality of Life Questionnaire, reflect patients' real experiences and perceptions of living with asthma. Despite "control" being described as the goal of asthma treatment, current surveys confirm that the majority of patients do not achieve control and are consequently condemned to an impaired quality of life 3, 4. The strong correlation between Asthma Quality of Life Questionnaire scores and guideline-derived asthma control seen in the Gaining Optimal Asthma controL study supports the case for attempting to achieve and maintain asthma control at a higher level than at present. It also confirms that patients are able to distinguish between and appreciate the benefits of this approach. The Gaining Optimal Asthma controL study confirms that impaired quality of life is an unnecessary hardship and can be avoided by aiming for total control (a composite measure derived from guideline goals) through individualised treatment escalated, where necessary, in accordance with accepted treatment steps. It further confirms that, with sustained dosing, gains are maintained and further improvements may occur. Since the Gaining Optimal Asthma controL study protocol made no provision for stepping down treatment in patients achieving control, further studies are required to examine whether it is possible to maintain the high levels of quality of life achieved in Gaining Optimal Asthma controL when controller treatment is reduced. Nevertheless, these results confirm that near-normal health-related quality of life can be achieved when treatment aims for total control of asthma, and that results with salmeterol/fluticasone propionate combination are superior to fluticasone propionate alone. This should serve to increase the expectations of patients and their caregivers regarding what can be achieved for all people with asthma.
The authors would like to thank H. Boushey, A. Woolcock and R. Pauwels for their valuable contributions to the study. Study management, editorial and statistical support were provided by M. Gibbs, L. Frith, N. Gul, J. Lim and D. Gor. Goal Investigators: M. de Vega, Salta; G Garcia, Buenos Aires; R Gene, Buenos Aires; R Rojas, Tucuman; E Schiavi, Buenos Aires; A Yanez, Buenos Aires, (all Argentina). A Cukier, Sao Paulo; CC Fritscher, Porto Alegre; ALG Fernandes, Sao Paulo; J R Lapa e Silva, Rio de Janeiro; JC Oliveira, Juiz de For a; E Pizzuchini, Florianspolis, (all Brazil). M Barros, Valparaiso; C Castillo, Temuco; L Vicherat, Santiago, (all Chile). S Gonzalez-Diaz, Nuevo-Leon; D Hernandez-Colin, Guadalajara, (all Mexico). F Montealegre, Ponce; R Reyes, San Juan, (all Puerto Rico). S Cala, North Gosford; M Chia, Toorak Gardens; N Freezer, Clayton; D Freiberg, Bankstown; K Karrasch, Kippa Ring; M McKeirnan, Carina Heights; R Scicchitano, Adelaide; C Steinfort, Geelong; T Underwood, Leura; EH Walters, Melbourne, (all Australia). B He, Beijing; SG Huang, Shanghai; S Niu, Shanghai; GS Qian, Chonqing; WZ Yao, Beijing; XRX Yi Shi, Nanjing; NA Zhong, Guangzhou; X Zhou, Shanghai; M Ip, Hong Kong; L Yam, Hong Kong; W Cho Yu, Hong Kong, (all China). YJ Cho, BW Choi, DC Choi, CS Hong, YY Kim, HB Moon, C Park, HJ Yoon, (all Seoul, South Korea); HS Park, Suwon, South Korea. R Harun, Kuala Lumpar; HB Liew, Sabah; R Mutalif, Kuala Terrengganu, (all Malasyia). C Burgess, Wellington; F Cullen, Hamilton; S Holt, Wellington; T Taylor, Dunedin; I Town, Christchurch; M Watson, Hamilton, (all New Zealand). M Banares, Quezon City; T de Guia, Quezon City; L Rivera, Makati; A Santos, Cebu City; J Wi, Dagupan City, (all Phillipines). A Goh and CT Yeo, Singapore. MA Gaffar, Durban; ED Bateman, Cape Town; J Killian, Pretoria; U Lalloo, Durban; G Naude, Verwoerdburg; J O'Brien, Cape Town; PG Williams, Johannesburg, (all South Africa). JY Hsu, Taichung City; YF Lai, Kaohsiun Hsien; TCY Tsao, Tao-Yuan Shein; JH Wang, Taipei; JH Wen, Changhua, (all Taiwan). W Boonsawat, Khonkaen; C Direkwattanachai, Bangkok; A Jatakanon, Bangkok; K Ruxrungtham, Bangkok; S Suntornlohanakul, Songkla, (all Thailand). W Janistyn, Wr. Neustadt; G Kaik, Vienna; N Vetter, Vienna; J Wuertz, Linz, (all Austria). J Behets, Paal-Beringen; JB Martinot, Namur; J Mortelmans, Ham; E Pieters, Tessenderlo; P Reynders, Tessenderlo, (all Belgium). AS Markovic, N Tudoric, E Verona, (all Zagreb, Croatia). P Cap, Praha; M Kasl, Tabor; V Petru, Prague; P Pohunek, Prague, (all Czech Republic). V Backer, Koebenhavn; F Egede, Svenborg; H Harving, Aalborg; JI Jensen, Varde; Peter Lange, Hvidovre; L Laursen, Hellerup; T Sorensen, Vaerloese, (all Denmark). R Jogi, Tartu; T Kipper, Tallin; E Losenko, Kohtal-Jdrve; L Raudla, Tallin; R Sepper, Tallinn, (all Estonia). J Jaakkola, Pietersaari; J Kotaniemi, Lahti; M Nieminen, Kangasala; IJ Strander, Vaasa, (all Finland). R Amar, Rouen; M Anton, Nantes; T Arfi, Nice; J Auger, Ronorantin; R Berman; L Bernabeu, Chauny; C Bertin, Montbeliard; F Blonde-Vincent, Lagny sur Marne; C Camus, Le Chesnay; Y Caubet, Lille; C Douillet, Montereau sur Yonne; JP Dumur, Aix en Provence; L Fouquert, Grasse; P Godard, Montpellier; P Greillier, Ollioules; N Grunchec, Cholet; JM Houssel, Oyonnax; P Hyvernat, Lyon; JY Jasnot, Nantes; JB Jouet, Reims; H Kafe, Saint Quentin; R Klink, Laon; MP Lafourcade, Angouleme; B Lebeaupin, Reze; B Lebrozec, Libourne; F Malaquin, Albi; F Marmouz, Pontoise; JPMathieu, Bayonne; B Mellin, Dunkerque; D Ortolan, Villejuif; F Payot, Lyon; H Pegliasco, Marseille; B Pigearias, Nice; D Proisy, Soissans; A PrudHomme, Tarbes; J Robert, Decines; F Saint Martin, Villebond sur Yvette; JC Severac, Beziers; C Traube, Villepinte; P Veve, Orange; F Wessel, Nantes, (all France). C Baumgarten, Berlin; A Benedix, Berlin; EM Bonninghoff, Beckum; R Braun, Unterschneidheim; HG Dammann, Hamburg; J Eller, Berlin; U Gehling, Schwetzingen; Jo Grosskopf, Wallerfing; T Jung, Deggingen; HG Kirchberg, Coburg; B Kroemer, Kaufbeuren; W Lieske, Deidesheim; A Linnhoff, Berlin; HH Ponitz, Berlin; G Preiss, Erlangen; R Schnorr, Berlin; HH Weber, Hannover; W Weede, Muenchen; W Westphal, Hamburg; H Witt, Lichtensfel; W Zachgo, Geesthacht, (all Germany). J Bibakis, Heraklion; P Christaki, Thessaloniki; N Galanis, Thessaloniki; NG Papageorgiou, Athens; C Gratziou, Athens; M Kakoura, Thessaloniki; N Siafakas, Heraklion, Crete, (all Greece). I Herjavecz, Budapest; P Magyar, Budapest; J Strausz, Tvrvkbalint, (all Hungary). D Gislason, Gardabaer; M Jonsson, Reykjavik, (all Iceland). D Ansalone, Salerno; A Bianchi, Roma; G D'Amato, Napoli; RD Negro, Verona; A Foresi, Milano, Sesto S. Giovanni; G Idotta, Padova; PL Paggiaro, Pisa; M Polverino, Napoli; A Potena, Ferrara; M Vignola, Palermo, (all Italy). A Krams, Riga; V Lozovskis, Riga; V Sillins, Riga, (all Latvia). A Bagdonas, Kaunas; R Dubakiene, Vilnius; G Norvaisas, Klaipeda, (all Lithuania). EJFM ten Berge, Hengelo; SJM Gans, Harderwijk; AJ van Harreveld, Apeldoorn; WJ den Ouden, Rotterdam; P Overberg, Hoorn; RCFM Vreuls, Enschede; J Westbroek, Heerenveen, (all The Netherlands). K Furuseth, Jessheain; T Oien, Trondheim; N Ringdal, Molde; K Skaug, Haegesund; A Sundset, Sandvika; I Vika, Karmsund, (all Norway). A Emeryk, Lublin; Z Sierglejko, Bialystok; A Slomka, Jelenia, (all Poland). J Almeida, Porto; AHR de Almeida, Vila Nova de Gaia; F Dias, Lisboa; R Pato, Cominbra, (all Portugal). M Radoi, Brasov; J R Radu, Bucharest; A Rascu, Bucharest, (all Romania). LA Goryachkina, NI Ivanovna, V E Nonikov, I Sidoreko, AN Tsoi, (all Moscow, Russia). M Hrubisko, Bratislava; M Lickova, Trebisov; P Pruzinez, Bratislava, (all Slovakia). S Kajba, Celje; R Kopriva, Litija; J Sorli, Golnik, (all Slovenia). JLA Sala, Madrid; T Bracamonte, Madrid; T Carrillo, Las Palmas; GG Hernandez, Madrid; PM Escribano, Malaga; F Ripolles, Sagunto; J Sanchis, Barcelona; V Sobradillo, Barakaldo, (all Spain). R Helleday, Umea; CG Lofdahl, Lund; B Olsson, Stockholm; EO Kulling, Stockholm; B Tilling, Atvidaberg; N Segurs, Boras; I Vinge, Lindingo; A Uusijarvi, Stockholm, (all Sweden). WRC Aitchison, Bridge of Weir; S Boyce, Glenrothes, JR Courtney, Bangor; SN Fearns, Doncaster; PJ Griffiths, Manchester; SK Holgate, Barry; JAV Jones, Cardiff; AG Lane, Cardiff; DJ Manchester, Manchester; JB McBride, Cookstown; M McCaughey, Randalstown; C McKinnon, Glasgow; PE Shearer, Ayr; AA Woodcock, Manchester, (all UK). E Amer, Mississauga; JS Anthony, Scarborough; WW Arkinstall, Kelowna; J Bouchard, La Malbaie; AS Cheema, Mississauga; RL Cowie, Calgary; GT Ford, Calgary; JR Herbert, Ste Foy; PK Keith, Hamilton; JHS Li, Moncton; MR Maleki-Yazdi, Toronto; BR Ramjattan, St Johns; P Renzi, Montreal; R Somani, Langley; RJ Welch, Peterborough; WH Yang, Ottawa, (all Canada). N Allegar, Hillsborough; L Bielory, Newark; RE Broker, Simpsonville; WW Busse, Madison; M Byrd, Roswell; EJ Campbell, Salt Lake City; NG Campbell, Houston; NT Chao, Trumbull; RG DeGarmo, Greer; MB DiDea, Orlando; D Dvorin, Mt. Laurel; SP Galant, Orange; KM Gehi, Anniston; LS Greos, Englewood; T Isakov, Lyndhurst; KW Jacobson, Eugene; EM Kerwin, Medford; WB Koppel, Baltimore; PE Korenblat, St Louis; V Krisciunas, Portland; BQ Lanier, Fort Worth; SD Miller, North Dartmouth; JD Mumper, Mechanicsville; RA Nathan, Colorado Springs; HS Nelson, Denver; JL Newman, Vista; MJ Noonan, Portland; JL Pinnas, Tucson; B Pogue, Boise; GD Raphael, Bethesda; MJ Round, Kissimme; NT Rupp, Charleston; G Salazar, San Antonio; EJ Schenkel, Easton; WD Sinclair, Missoula; M Strek, Chicago; RZ Surowitz, Jupiter; DR Webb, Kirkland; JM Weinberger, Kissimmee, (all USA).
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||