Association of Socioeconomic Status with the Use of Chronic Therapies and Healthcare Utilization in Children with Cystic Fibrosis

Abstract

Objective

To determine whether previously reported socioeconomic status (SES)-related disparities in cystic fibrosis (CF) health outcomes vary by the indicator used (median household income by zip code [MIZ], maternal educational attainment [MEA], and state insurance coverage [MA]), and whether these disparities can be explained by differences in medical treatment.

Study design

A cross-sectional analysis of data on patients age <18 years from the Epidemiologic Study of Cystic Fibrosis (ESCF).

Results

Disease severity showed a similar inverse correlation with all 3 SES measures. The number of stable clinic visits was unrelated to SES. Patients with MA had more sick outpatient visits and more courses of intravenous (IV) antibiotics for pulmonary exacerbations, and were more likely to be prescribed all chronic therapies. Low-MIZ patients had slightly fewer sick visits and more courses of IV antibiotics, and were more likely to receive oral nutrition supplements but less likely to receive macrolide prescriptions. Low-MEA patients were less likely to receive IV antibiotics at home, more likely to receive oral nutrition supplements, but less likely to receive macrolide prescriptions.

Conclusions

CF health outcomes are correlated with the SES spectrum, but these disparities are not explained by differential use of health services or prescription of chronic therapy. Future investigations should focus on the possible impact of environmental exposures and differences in disease self-management.

Socioeconomic status (SES) is a strong predictor of outcomes in patients with cystic fibrosis (CF).^1,2 Barriers to accessing quality care are an important cause of SES-related disparities in many patient populations,³ but previous analyses of CF Foundation Registry data have failed to show any SES-related difference in care at accredited CF care centers.² Until recently, however, the CF Registry collected only limited data on the use of specific therapies and practice patterns. Reports from the Epidemiologic Study of Cystic Fibrosis (ESCF) show large variations in the use of chronic medications of proven efficacy⁴ and in outpatient monitoring,⁵ and also demonstrate an association between these variations in care and outcomes.⁶ It is plausible that some of this variability in treatment and monitoring might be related to some aspect of SES and serve as a possible explanation for the disparities in outcomes.

There is no ideal measure of SES; different aspects impact health in different ways. Family income affects overall financial resources, which can affect the ability to access care and medications; measures of neighborhood income partly reflect individual income but have an independent effect related to features of the physical, social, or service environments, such as public transportation or distance from a CF center.⁷ Maternal education level affects disease self-management abilities. State insurance eligibility reflects financial resources, but this differs by state and is determined in part by healthcare expenditures, leading to the possible misinterpretation of the direction of causality in terms of conclusions regarding the link between SES and health.^8-13 Whether any specific indicator of SES most accurately predicts disease outcomes in CF is not clear; to date, no comparison of different variables associated with SES in the CF population has been reported.

We analyzed data from the ESCF to ascertain whether SES is associated with different patterns of routine outpatient monitoring or the likelihood of prescription of various chronic CF therapies. Given that a small minority of children with CF are uninsured,² we hypothesized that due to potential access barriers, such as cost and transportation,⁷ the most likely effect of SES would be in the number of outpatient visits, and any disparity in prescribed medications would not be seen for Medicaid recipients. In addition, we investigated whether any of 3 different indicators of SES for which ESCF tracked data (median household income by zip code, maternal education, or insurance status) might have a stronger association with disease severity or practice patterns compared with the others. We further hypothesized that state insurance status might have the strongest association with disease severity because of the higher likelihood that sicker patients are eligible for Medicaid, and that neighborhood income might have a weaker association because of the ecological nature of the variable.¹⁴

Methods

The design and implementation of the ESCF have been described previously.¹⁵ The ESCF is a large, multicenter, longitudinal, prospective observational study of the clinical course of patients with CF in the United States and Canada from 1994 through 2005. Data collected on each patient included demographics, use of chronic therapies, results of pulmonary function tests, results of respiratory tract cultures, information on growth and nutrition, and details of antibiotic, anti-inflammatory, and other treatments. ESCF II, the second phase of the ESCF, was conducted from 2003 to 2005 and included several additional variables, including those relevant to SES. Institutional review board approval was obtained at each individual study site.

The data collection form was completed by the study site personnel from medical records and identified certain prescribed therapies as chronic; there was no ascertainment of patient adherence. Data also were collected on treatment of pulmonary exacerbations, including site (home vs hospital) where intravenous (IV) antibiotics were administered.

Cohort Definition and SES Measures

This analysis included patients who were age <18 years at the time of enrollment in ESCF II, had 2 or more medical encounters at least 30 days apart recorded in the ESCF II database during 2004, and had no missing data on the outcome variables of interest and at least 1 SES variable. SES was evaluated using 3 indicators, defined as follows:

• Median income of zip code of residence (MIZ). ESCF data on zip code of residence were linked to United States Census reports from the year 2000 on median annual income for each zip code in the United States. These data were categorized into 4 groups: < $40 000, $40 000 to < $50 000, $50 000 to < $60 000, and $ $60 000.

• Maternal educational attainment (MEA), reported by participating centers in 3 categories: less than high school, high school diploma, or college diploma

• Eligibility for Medicaid or state health insurance cover age (MA), reported as a dichotomous yes/no variable, independent of whether or not the patient had any other form of health insurance.

Outcome Measures and Covariates

We evaluated the relationships between SES and disease severity, use of prescription or chronic medications, and healthcare utilization. Disease severity for patients age < 6 years was categorized using the first stable weight-for-age percentile (ie, < 5, 5 to < 10, 10 to < 25, 25 to < 50, $ 50) in 2003 or 2004. For older patients, the first stable percent predicted prebronchodilator forced expiratory volume in 1 second (FEV₁) value (ie, < 40, 40 to < 70, 70 to < 100, $ 100) in 2003 or 2004 was used to categorize disease severity. Values for percent predicted FEV₁ were calculated using the formula of Wang et al¹⁶ for males up to age 17 years and females up to age 15 years. The equations of Hankinson et al¹⁷ were used for all older patients. In terms of chronic medications, the reported use of oral and enteral nutritional supplements, vitamins, oral macrolide antibiotics, inhaled antibiotics, and dornase alfa in 2004 was evaluated. Health-care utilization was evaluated as the number of outpatient visits (distinguished by report as “stable” or “sick”) and number of pulmonary exacerbations requiring IV antibiotics (as diagnosed by treating physician and distinguished by a report as treated in the hospital or at home) in 2004. Age at enrollment and disease severity were considered potential confounders and included as covariates in the analysis of chronic medication use and healthcare utilization.

Statistical Analysis

Differences among patients in each SES category were evaluated by ײ, Mantel-Haenszel ײ, or t tests, as appropriate. Separate multivariate models were used to evaluate each SES indicator as a predictor of each outcome measure, controlling for age at enrollment in the ESCF II and for disease severity. Additional models containing either all 3 SES categories or 2 categories at a time provided no novel outcomes and thus are not reported here. Because disease severity was classified differently for children age < 6 and those age 6 to < 18, the model was established with a different set of severity effects for these 2 age groups. Logistic regression was used to model the likelihood of receiving chronic therapies and medications, and linear regression was used to model the number of healthcare visits.

Results

Of the 32 585 patients enrolled in the ESCF between 1994 and 2005, 17 722 were enrolled in the ESCF II. Of these patients, 11 447 were age < 18 years at the time of enrollment in the ESCF II. Approximately 83% of these patients (n = 9534) had at least 2 medical encounters in 2004 spanning a minimum of 30 days (mean number of encounters, 4.93 ± 2.65), and no missing SES or disease severity data. MIZ was known for 9134 patients, MEA was reported for 3731 patients, and MA status was known for 9168 patients. Figure 1 shows the overlap of subjects with SES data.

Figure 1.

Distribution of subjects for each SES cohort.

The cohorts were similar with respect to mean age, sex, race/ethnicity, percent predicted FEV₁, height-for-age and weight-for-age percentiles, body mass index (BMI), and disease severity (data not shown). The distributions of mother's education, median household income, and type of healthcare insurance also were similar among the cohorts. For the MIZ and MEA cohorts, there were no differences between the study population analyzed and those patients excluded from the analysis because of missing SES data (data not shown). However, compared with the group included in the MA cohort, those with missing insurance data tended to be older (mean age, 9.1 ± 5.0 vs 8.3 5.2 years; P = .007), to be non-Hispanic white (94.3% vs 85.3%; P < .001), to have higher BMI (mean BMI, 18.7 ± 3.4 vs 18.3 ± 3.3; P = .013), and to have higher percent predicted FEV₁ (mean, 94.9 ± 19.7 vs 91.6 ± 21.6; P = .014).

As assessed by Spearman correlation coefficients, the correlation among SES indicators was small to moderate¹⁸: 0.193 between MIZ and MEA, 0.260 between MA and MEA, and 0.343 between MA and MIZ (P < .001 for all).

Comparisons of disease severity, as measured by both weight-for-age percentile and percent predicted FEV₁, and all measures of SES exhibited a linear inverse correlation (ie, low SES associated with greater disease severity; all P < .0001) (Figure 2). There was no difference in the prevalence of positive airway cultures for Pseudomonas aeruginosa by any SES measure.

Figure 2.

Distribution of SES by disease severity, defined by weight-for-age percentile in children age < 6 years A, and percent predicted FEV₁ in children age 6 to 17 years B.

Likelihood of Chronic Medication Use

In models adjusting for age and disease severity, patients living in zip codes with the lowest median income were more likely to receive oral nutritional supplements and inhaled antibiotics but less likely to receive macrolides; there were no discernible trends in the prescription of other chronic therapies (Figure 3, A). Similarly, patients whose mothers had less than a high school education were more likely to be prescribed oral nutritional supplements but less likely to be prescribed vitamins and macrolides compared with patients whose mothers were high school graduates (Figure 3, B). There was no significant difference between high school and college graduates in the use of any chronic therapies, so the results from a model using a dichotomous MEA variable are shown. Patients with MA were about 30% to 50% more likely than patients without MA to be prescribed all chronic therapies and medications except macrolides (Figure 3, C).

Figure 3.

Adjusted odds ratios for the likelihood of being prescribed chronic therapy according to SES status, as defined by MIZ A, MEA B, and MA C. Logistic regression models include age at enrollment and severity as covariates.

Healthcare Utilization

In models adjusting for age and disease severity, no SES-related differences in the number of stable outpatient visits were found. Patients living in zip codes with the lowest median income had the fewest sick outpatient visits (Table). There was a trend across the category for lower-income patients to receive more IV antibiotics for pulmonary exacerbations. MEA was not associated with any difference in outpatient visits or treatment for IV exacerbations, but the children of mothers who did not graduate high school were less likely to be treated for IV exacerbations at home. There was no significant difference between high school and college graduates in terms of healthcare utilization, so the results from a model using a dichotomous MEA variable are shown. Patients with MA had more sick visits than those without MA, and they also had more pulmonary exacerbations treated with IV antibiotics.

Table.

Adjusted^* averages for healthcare utilization by SES cohort

Healthcare utilization (mean ± SE)	MIZ					MEA			MA
	<$40 000 (n = 1998)	$40 000-<$50 000 (n = 2585)	$50 000-<$60 000 (n = 1731)	≥$60 000 (n = 2820)	P †	< High school (n = 344)	≥ High school (n = 3387)	P ‡	Medicaid (n = 4073)	Other (n = 5095)	P ‡
Stable visits	3.56 ± 0.04	3.61 ± 0.04	3.67 ± 0.04	3.65 ± 0.03	.18	3.48 ± 0.10	3.66 ± 0.03	.078	3.59 ± 0.03	3.61 ± 0.03	.72
Sick visits	0.94 ± 0.04	1.05 ± 0.03	1.16 ± 0.04	1.07 ± 0.03	<.001	0.99 ± 0.10	1.14 ± 0.03	.15	1.22 ± 0.03	0.93 ± 0.02	<.001
Total IV exacerbations	0.47 ± 0.02	0.48 ± 0.02	0.45 ± 0.02	0.41 ± 0.02	.006	0.51 ± 0.05	0.46 ± 0.02	.43	0.56 ± 0.01	0.37 ± 0.01	<.001
Hospital IV exacerbations	0.44 ± 0.02	.44 ± 0.02	0.40 ± 0.02	0.35 ± 0.02	<.001	0.47 ± 0.05	0.41 ± 0.01	.17	0.51 ± 0.01	0.32 ± 0.01	<.001
Home IV exacerbations	0.19 ± 0.01	0.22 ± 0.01	0.24 ± 0.01	0.22 ± 0.01	.091	0.14 ± 0.03	0.24 ± 0.01	.003	0.23 ± 0.01	0.21 ± 0.01	.062

^*Linear regression models include age at enrollment and severity as covariates.

^†Calculated from F-test statistics.

^‡Calculated from t-test statistics.

Discussion

SES is a strong predictor of disease severity in CF; patients of low SES have increased mortality, worse lung function, and worse nutrition.^1,2 This is confirmed by our ESCF data, which also show a similar association whether SES is defined by median household income by zip code, by maternal education, or by Medicaid status. Our findings suggest the existence of some previously unappreciated SES-related differences in healthcare utilization patterns in patients attending CF care centers in the United States that vary, to some degree on the marker of SES used. Patients with low SES as indicated by MIZ) have fewer outpatient sick visits, but patients with MA have more visits, and MEA does not seem to influence the number of visits. The relationship between MIZ and number of sick visits might reflect reduced access for patients with this marker, particularly if they live at a distance from the CF center.¹⁹ Pulmonary exacerbations requiring IV antibiotics are diagnosed and treated more frequently in patients with lower SES, as indicated by MIZ or MA status but not by MEA. The reason for this is unclear; perhaps the CF providers sense that patients with fewer resources are likely to deteriorate faster and thus need more aggressive care. Among the SES measures, MEA primarily seems to influence whether a patient's pulmonary exacerbations are treated with IV medications at home or in the hospital, suggesting that providers are less likely to believe that low-MEA mothers can perform all of the necessary treatments on their own or that the mothers perceive barriers to home care.

Most chronic medications are used at similar or slightly greater rates in patients with lower SES. Patients with MA in particular are more likely to receive all medications evaluated except macrolides, perhaps reflecting the absence of financial barriers faced by patients who have private insurance and need to make copayments for medications. Oral nutritional supplements are more likely to be prescribed to low-MIZ and low-MEA patients. Low-MIZ patients receive more inhaled antibiotics, even though they are no more likely than others to have Pseudomonas airway infection when rates are adjusted for disease severity and age. The greater prescription of chronic therapies in this group might indicate that healthcare providers recognize that patients with low SES have worse disease outcomes and attempt to treat these patients more aggressively.

Children of more highly educated mothers and those living in wealthier zip codes are more likely to be prescribed macro-lides. This analysis evaluated practice in 2004, the year after publication of a major study documenting the efficacy of azithromycin in CF.²⁰ It took several years before the use of chronic macrolide therapy was consistently adopted at CF centers, and even now it is used nationally in fewer than 60% of patients for whom it would be appropriate.²¹ One possible explanation for this is that CF healthcare providers are prompted by wealthier or better-educated mothers to prescribe macrolides sooner than they might have otherwise. This is consistent with the theory that wealth or parental education influences disease outcomes in children through their impact on disease self-management.²² Similarly, the greater use of vitamins in children with more educated mothers also might be related to parental prompting.

Overall, we uncovered no compelling differences in health-care utilization or the prescription of chronic therapies that would explain the significant and striking disparities in health outcomes in low-SES groups. This contrasts with findings in other chronic diseases, such as asthma,³ for which access to quality care is an important contributor to SES-related disparities. Other mechanism are probably of greater importance in the CF population.²³ For example, disease self-management is likely an important determinant of long-term outcomes for chronic illnesses such as CF and appears to be associated with SES.^22,24 Health behaviors, such as cigarette smoking, also are clearly linked to SES;²⁵ secondhand smoke exposure is a likely contributing factor to SES disparities in CF.²⁶ Other potential risk factors include exposure to psychological stress and environmental pollutants.²⁷

No previous comparison of SES measures has been undertaken in patients with CF; in fact, very few comparison studies are available in the general pediatric literature.²⁸ SES is a broad sociological concept with different attributes and no single ideal measure.^8,29 Income is an intuitively obvious measure of SES and is a good measure of material resources, but it is potentially confounded by age, education, family size, regional cost of living, and whether the income is generated by one or more family members.⁸ We do not have direct information on family income in the ESCF database, but we used US Census data to link to median household income by zip code. This approach has the usual weaknesses and imprecision of an ecological measure,¹⁴ but any resulting misclassification would typically weaken statistical associations rather than cause spurious ones.³⁰ Furthermore, some specific neighborhood effects are probably relevant to an analysis of the impact of SES on health.³¹ In particular, in our population of primarily Caucasian patients with CF, low median household income is more likely to be associated with rural areas than inner-city urban areas, and rural populations are more likely to encounter barriers due to distance from a CF center.¹⁹ Educational attainment may best predict disease self-management skills. Furthermore, because the resources required to care for a sick child may drain a family of wealth and time for work, low education is more likely than low income to be the cause, rather than the result, of greater disease severity.²² Unfortunately, we have information on MEA for only one-third of our patients, making it difficult to ensure that our sample is not biased. State insurance (Medicaid) eligibility is closely tied to family income and is readily ascertained, making it a simple measure to use. But eligibility is determined differently by individual states, undermining its usefulness when comparing outcomes across centers. Of medical expenses may be considered when evaluating eligibility;³² thus, sicker patients are more likely to be eligible for MA at any given income level. The finding that MA was associated with greater likelihood of receiving many therapeutic interventions suggests that finances might play a role in prescribing patterns for low-SES patients, but this also may be because higher medical expenses increase the likelihood of being eligible for MA. This makes the finding that MA was no stronger a predictor of disease severity than the other 2 SES measures even more interesting.

The low correlation among SES measures is intuitively surprising (although it has been reported previously⁷) and reflects the fact that SES is a heterogeneous category with different components. This has presented significant methodological challenges in attempting to differentiate the relative contribution of SES to racial disparities³³ in CF along with other conditions; the contradictory reports on the relative impact of race on outcomes in CF stem from the use of different SES measures to control for that factor.^1,2 Attempts to combine components of SES into a single composite measure have met with limited acceptance;⁸ in our own analysis, models combining all measures provided essentially the same results as those containing only 1 or 2 measures.

Because this was a cross-sectional study, the direction and causality of the relationship between SES and disease severity could not be determined definitively. Nevertheless, our findings are consistent with those from previous longitudinal analyses. In addition, without performing a prospective cohort study, we cannot technically conclude that SES status preceded the differences (or similarities) in treatment, or that those treatments preceded the SES-related outcome disparities, but it is likely that both contribute to these findings.

In conclusion, we have confirmed previous observations regarding SES and disease severity in CF and have found that this relationship exists in association with all 3 of the SES markers that we studied. Although our data show mild differences in the use of outpatient and inpatient services, these are unlikely to have an important relationship with SES-associated disparities. The differences in the use of chronic therapies suggest that low-SES patients are treated equally or more intensively than higher-SES patients, so again this aspect of care is unlikely to explain the SES disparities. MA eligibility is clearly associated with an increased likelihood of receiving various prescribed treatments and appears to increase the use of the outpatient clinic for sick visits. □

Glossary

BMI

Body mass index

CF

Cystic fibrosis

ESCF

Epidemiologic Study of Cystic Fibrosis

FEV₁

Forced expiratory volume in 1 second

IV

Intravenous

MA

Medicaid or state insurance

MEA

Maternal educational attainment

MIZ

Median household income by zip code

SES

Socioeconomic status

Appendix 1

Appendix 1.

ESCF Investigators and NASAG Members

ESCF Investigators
Dana Brasfield	Alabama
Raymond Lyrene	Alabama
Lawrence Sindel	Alabama
Dion Roberts	Alaska
John Carroll	Arkansas
Robert Warren	Arkansas
Louay Nassri	Arkansas
Paula Anderson	Arkansas
Mark Brown	Arizona
Amy Silverthorn	Arizona
Peggy Radford	Arizona
Gerald Gong	Arizona
Gregory Legris	Arizona
Gerald Greene	California
Reddivalam Sudhakar	California
Arnold Platzker	California
Bruce Nickerson	California
Karen Hardy	California
Ivan Harwood	California
Gregory Shay	California
Bryon Quick	California
Allan Lieberthal	California
Richard Moss	California
Chris Landon	California
Yvonne Fanous	California
Jay Lieberman	California
Eugene Spiritus	California
Bradley Chipps	California
Ruth McDonald	California
Mark Pian	California
Gerd Cropp	California
Nancy Lewis	California
Dennis Nielson	California
Bertrand Shapiro	California
Jeff Wagener	Colorado
Frank Accurso	Colorado
Milene Saavedra	Colorado
Karen Daigle	Connecticut
Jacob Hen	Connecticut
Regina Palazzo	Connecticut
Kathryn Dodds	Delaware
Raj Padman	Delaware
John Goodill	Delaware
Glenna Winnie	District of Columbia
Lea Davies	District of Columbia
Tony Kriseman	Florida
Jorge Sallent	Florida
Joseph Chiaro	Florida
Martin Kubiet	Florida
Sue Goldfinger	Florida
Morton Schwartzman	Florida
Carlosenrique Diaz	Florida
Kevin Maupin	Florida
Eduardo Riff	Florida
David Geller	Florida
Floyd Livingston	Florida
Kunjana Mavunda	Florida
Jose Birriel Jr.	Florida
Luis Faverio	Florida
David Rosenberg	Florida
David Schaeffer	Florida
James Sherman	Florida
Mary Wagner	Florida
Michael Light	Florida
Bruce Schnapf	Florida
Gary Montgomery	Georgia
Kevin Kirchner	Georgia
Mark Weatherly	Georgia
Daniel Caplan	Georgia
Margaret Guill	Georgia
Valera Hudson	Georgia
Javeed Akhter	Illinois
Donald Davison	Illinois
Steven Boas	Illinois
Susanna McColley	Illinois
Youngran Chung	Illinois
Rennee Latner	Illinois
Gabriel Aljadeff	Illinois
Youngran Chan	Illinois
Jerome Kraut	Illinois
Arvey Stone	Illinois
John Lloyd Still	Illinois
Girish Sharma	Illinois
Lanie Eagleton	Illinois
Patricia Hopkins	Illinois
Umesh Chatrath	Illinois
Lucille Lester	Illinois
Young-Jee Kim	Illinois
Veena Anthony	Indiana
Howard Eigan	Indiana
Michelle Howenstine	Indiana
Pushpom James	Indiana
Edward Gergesha	Indiana
James Harris	Indiana
Robert Plant	Indiana
Veljko Zivkovich	Iowa
Angela Collins	Iowa
Edward Nassif	Iowa
Richard Ahrens	Iowa
Daniel Doornbos	Kansas
Joseph Kanarek	Kansas
Richard Leff	Kansas
Pamela Shaw	Kansas
Elanor Demoss	Kansas
Maria Riva	Kansas
Leonard Sullivan	Kansas
Michael Anstead	Kentucky
Jamshed Kanga	Kentucky
Nemr Eid	Kentucky
Ron Morton	Kentucky
Bettina Hilman	Louisana
Kim Jones	Louisana
Scott Davis	Louisana
Ralph Harder	Maine
Tom Lever	Maine
Anne Marie Cairns	Maine
Edgar Caldwell	Maine
Jonathan Zuckerman	Maine
Peter Mogayzel	Maryland
Beryl Rosenstein	Maryland
John McQuestion	Maryland
Donna Perry	Maryland
Samuel Rosenberg	Maryland
Robert Gerstle	Massachusetts
Andrew Colin	Massachusetts
Mary Ellen Wohl	Massachusetts
Allan Lapey	Massachusetts
William Yee	Massachusetts
Brian O'Sullivan	Massachusetts
Robert Zwerdling	Massachusetts
Ibrahim Abdulhamid	Michigan
Adrian O'Hagan	Michigan
John Schuen	Michigan
Lawrence Kurlandsky	Michigan
Richard Honicky	Michigan
Douglas Homnick	Michigan
John Marks	Michigan
Bohdan Pichurko	Michigan
Norma Maxvold	Michigan
Samya Nasr	Michigan
Richard Simon	Michigan
Wan Tsai	Michigan
Dana Kissner	Michigan
John Mc Namara	Minnesota
Nancy Henry	Minnesota
Stephen Marker	Minnesota
Michael Pryor	Minnesota
Warren Regelmann	Minnesota
Lynn Walker	Minnesota
Jim Woodward	Mississippi
Louis Mizell	Mississippi
Suzanne Miller	Mississippi
Daniel Rosenbluth	Missouri
Philip Black	Missouri
Michael McCubbin	Missouri
Alan Cohen	Missouri
Thomas Ferkol	Missouri
George Mallory	Missouri
Anthony Rejent	Missouri
Bruce Rubin	Missouri
Gavin Graff	Missouri
Peter Konig	Missouri
John Colombo	Nebraska
Peter Murphy	Nebraska
William Boyle H	New Hampshire
Worth Parker	New Hampshire
Chandler Patton	New Jersey
Robert Zanni	New Jersey
Arthur Atlas	New Jersey
Nelson Turcios	New Jersey
Lourdes Laraya-Cuasay	New Jersey
Dorothy Bisberg	New Jersey
Helen Aguila	New Jersey
Sarah Allen	New Mexico
David James	New Mexico
Elizabeth Perkett	New Mexico
Marsha Thompson	New Mexico
Sonia Budhecha	Nevada
Ruben Diaz	Nevada
Jonathan Rosen	New York
Robert Kaslovsky	New York
Ronald Percciacante	New York
Drucy Borowitz	New York
Joseph Cronin	New York
Colin McMahon	New York
Lynne Quittell	New York
Robert Giusti	New York
Rubin Cohen	New York
Joan DeCelie-Germana	New York
Jack Gorvoy	New York
Kalpan Patel	New York
Meyer Kattan	New York
Allen Dozor	New York
Emily DiMango	New York
Maria Berdella	New York
Ran Anbar	New York
Debra Ianuzzi	New York
James Sexton	New York
Catherine Tayag-Kier	New York
John McBride	New York
Clement Ren	New York
Karen Voter	New York
Mary Dimaio	New York
Gerald Georgitis	North Carolina
Joseph Marc Majure	North Carolina
Maria Martinez J	North Carolina
Clarke McIntosh	North Carolina
Margaret Leigh	North Carolina
Michael Schechter	North Carolina
Hugh Black	North Carolina
James Hughes	North Dakota
Anand Kantak	North Dakota
Robert Wilmott	Ohio
Gregory Omlor	Ohio
Robert Stone	Ohio
Karen McCoy	Ohio
James Acton	Ohio
Carl Doershuk	Ohio
Michael Konstan	Ohio
Robert Fink	Ohio
Michael Steffan	Ohio
Pierre Vauthy	Ohio
Patricia Joseph	Ohio
Santiago Reyes	Oklahoma
John Kramer	Oklahoma
James Royall	Oklahoma
Jay Eisenberg	Oregon
Michael Wall	Oregon
Stanley Fiel	Pennsylvania
Thomas Scanlin	Pennsylvania
Shroti Phadke	Pennsylvania
Glenna Winnie	Pennsylvania
Joel Weinberg	Pennsylvania
William Sexauer	Pennsylvania
Stephen Wolf	Pennsylvania
Douglas Holsclaw	Pennsylvania
Debra Klein W.	Pennsylvania
Stuart Warren	Pennsylvania
Robert Kinsey	Pennsylvania
Carlos Perez	Pennsylvania
Muttiah Ganeshanathan	Pennsylvania
James Shinnick	Pennsylvania
Howard Panitch	Pennsylvania
Laurie Varlotta	Pennsylvania
Cynthia Robinson	Pennsylvania
Jose Rodriguez Santana	Puerto Rico
Mary Ann Passero	Rhode Island
Jane Gwinn	South Carolina
Robert Baker C.	South Carolina
Michael Bowman	South Carolina
Patrick Flume	South Carolina
Daniel Brown	South Carolina
Roxanne Marville	South Carolina
James Wallace	South Dakota
Rodney Parry	South Dakota
Don Ellenburg	Tennessee
John Rogers	Tennessee
Ricky Mohon	Tennessee
Joel Ledbetter	Tennessee
Aram Hanissian	Tennessee
Robert Schoumacher	Tennessee
Preston Campbell	Tennessee
Christopher Harris	Tennessee
Bonnie Slovis	Tennessee
Dennis Stokes	Tennessee
Kathryn Hale	Texas
Marcia Katz	Texas
Dan Seilheimer	Texas
Marianne Sockrider	Texas
Allan Frank	Texas
James Daniel	Texas
James Cunningham	Texas
Iley Browning	Texas
John Bray	Texas
Amanda Dove J.	Texas
Fernando Mandujano	Texas
Larry Tremper	Texas
Martha Morse	Texas
Donna Willey-Courand	Texas
Steven Copenhaver	Texas
John Pohl	Texas
Bennie McWilliams	Texas
Marie Martine-Logvinoff	Texas
Marsh Wallace	Texas
Robert Klein	Texas
Rodolfo Amaro	Texas
Leslie Couch	Texas
Michael Brown	Texas
Claude Prestidge	Texas
Stephen Inscore	Texas
Andrew Lipton	Texas
Barbara Chatfield	Utah
Theodore Liou	Utah
Bruce Marshall	Utah
Karl Karlson	Virginia
Ignacio Ropoll	Virginia
Thomas Rubio	Virginia
Joel Schmidt	Virginia
David Thomas	Virginia
John Osborn	Virginia
Deborah Froh	Virginia
Benjamin Gaston	Virginia
Greg Elliott	Virginia
Thomas Lahiri	Vermont
Donald Swartz	Vermont
Laurie Whittaker	Vermont
Ronald Gibson	Washington
Bonnie Ramsey	Washington
Michael McCarthy	Washington
Lawrence Larson	Washington
David Ricker	Washington
Mark Robbins	Washington
Moira Aitken	Washington
Julia Emerson	Washington
Julie Biller	Wisconsin
Mark Splaingard	Wisconsin
Bradley Sullivan	Wisconsin
Paul Pritchard	Wisconsin
Stu Adair	Wisconsin
Peter Holzwarth	Wisconsin
Guillermo Dopico	Wisconsin
Keith Meyer	Wisconsin
Christopher Green	Wisconsin
Michael Rock	Wisconsin
Stephen Aronoff	West Virginia
Kathryn Moffett	West Virginia
NASAG Members
Wayne Morgan (chair)
Michael Konstan (co-chair)
Ted Liou
Susanna McColley
Ann McMullen
Alexandra Quittner
Warren Regelmann
Clement Ren
Michael Schechter
Jeff Wagener
Marlyn Woo

Appendix 2

Funding and Conflict of Interest Disclosures

M.S.S., S.A.M., M.W.K., and J.S.W. have received honoraria during the last 3 years to attend meetings as members of the North American Scientific Advisory Group for the Epidemiologic Study of Cystic Fibrosis (ESCF), and their respective institutions previously received grant support from Genentech, Inc, for participating in the study. S.A.M. is a member of a speakers’ bureau for Genentech and a consultant to Axcan Pharma and Novartis; T.H. is a consultant to Genentech; M.W.K. is a member of a speakers’ bureau for Genentech and has received grant support from and is a consultant to Axcan Pharma, Digestive Care Inc, Genentech, and Novartis; and J.S.W. is a member of a speakers bureau for Genentech and is a consultant to Genentech and Novartis, and previously was an employee of Genentech. S.S. is an employee of ICON Clinical Research, which was paid by Genentech to provide biostatistical and analytical services for ESCF. No honoraria, grants, or other form of payment were given to any of these authors to produce this manuscript. The authors were responsible for the study design, interpretation of data, and drafting and completion of the manuscript. The initial draft and all subsequent drafts of the manuscript were written by Dr Schechter in consultation with the coauthors. The decision to submit the manuscript was made by the authors and was approved by Genentech.