Published online January 2, 2007
PEDIATRICS Vol. 119 No. 1 January 2007, pp. 19-28 (doi:10.1542/peds.2006-1498)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sims, E. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sims, E. J.
Related Collections
Right arrow Respiratory Tract

ARTICLE

Cystic Fibrosis Diagnosed After 2 Months of Age Leads to Worse Outcomes and Requires More Therapy

Erika J. Sims, PhDa,b, Allan Clark, PhDb, Jonathan McCormick, MDc, Gita Mehta, MPhila, Gary Connett, FRCPCHd, Anil Mehta, FRCP(Edin)a on behalf of the United Kingdom Cystic Fibrosis Database Steering Committee

a United Kingdom Cystic Fibrosis Database, Division of Maternal and Child Health Sciences, Ninewells Hospital and Medical School, University of Dundee, Dundee, Scotland, United Kingdom
b Department of Population Health, School of Medicine, Health Policy and Practice, University of East Anglia, Norwich, United Kingdom
c Respiratory Unit, Royal Hospital for Sick Children, Yorkhill National Health Service Trust, Glasgow, United Kingdom
d Department of Paediatrics, Southampton University Hospitals National Health Service Trust, Southampton, Hampshire, United Kingdom


   ABSTRACT
TOP
 ABSTRACT
PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
OBJECTIVE. Newborn screening for cystic fibrosis remains controversial because improved pulmonary function has not been established. Studies to date have not accounted for differences in treatments delivered to clinically diagnosed children and newborn-screened controls. Here, we compare outcomes and treatment of patients clinically diagnosed within the newborn-screening reporting window (early-clinically diagnosed), those presenting after this period (late-clinically diagnosed), and patients diagnosed by newborn screening.

PATIENTS AND METHODS. In a cross-sectional analysis of cohorts retrospectively ascertained, patients who were homozygous {Delta}F508 with cystic fibrosis, attending specialist cystic fibrosis centers, and 1 to 10 years of age between 2000 and 2002 were identified from the United Kingdom Cystic Fibrosis Database and stratified into newborn-screened, early-clinically diagnosed, or late-clinically diagnosed cohorts. Two analyses were performed: (1) after restricting to the most recent year of data collection, early-clinically diagnosed and late-clinically diagnosed cohorts were matched to newborn-screened patients by patient age and year of data collection (133 patients per cohort were identified); and (2) for all years of data collection, annual sets of data for early-clinically diagnosed and late-clinically diagnosed patients were matched to newborn-screened patients by patient age and year of data collection (291 data sets per cohort were identified). Median height and weight z scores, proportion of patients with height and weight <10th percentile, prevalence of chronic Pseudomonas aeruginosa infection, Shwachman-Kulczyki morbidity scores, percent predicted forced expiratory volume in 1 second, and numbers of long-term therapies were compared.

RESULTS. In both analyses, newborn screening was associated with higher height z score, higher Shwachman-Kulczyki score, lower likelihood of height <10th percentile, and fewer long-term therapies compared with late-clinically diagnosed patients. No other differences were found.

CONCLUSIONS. Newborn screening was associated with improved growth, reduced morbidity, and reduced therapy, yet generated equivalent pulmonary outcome compared with late clinical diagnosis, suggesting that newborn screening may slow cystic fibrosis lung disease progression.

Key Words: cystic fibrosis • newborn screening • clinical diagnosis

Abbreviations: CF—cystic fibrosis • CD—clinical diagnosis • MI—meconium ileus • NBS—newborn screening • UKCFD—United Kingdom Cystic Fibrosis Database • YDC—most recent year of data collection • SK—Shwachman-Kulczyki morbidity • NEBS—nebulized therapies • FEV1—forced expiratory volume in 1 second • RR—relative risk • CI—confidence interval • rhDNase—recombinant human deoxyribonuclease • IRT—immunoreactive trypsinogen

Cystic fibrosis (CF) is the most common life-threatening autosomal recessive, progressive disease in white individuals, affecting ~1 in 2500 births.1 Clinical diagnosis (CD) can be difficult unless neonatal bowel obstruction (meconium ileus [MI]) occurs, typically in ~13% of presentations.2 In the remainder of patients, CF masquerades as persistent lower respiratory tract infections, failure to thrive, or diarrheal states resulting in either a delayed or misdiagnosis.3–5 Many countries and States are, therefore, deliberating newborn screening (NBS) for CF, which can facilitate the early diagnosis (≥2 months after birth) of >90% of true-positives. Both prospective and cross-sectional studies comparing NBS and CD cohorts show less stunting of growth.6–8 However, better pulmonary outcomes are shown by some9–13 but not others,8,14,15 and the issue remains controversial.

A limitation of current outcome data lies in the comparison of NBS patients with all CD patients irrespective of their age at CD. Some studies have partially addressed this issue by excluding MI presentations6,7,9–12,14,15 because they do not benefit from NBS, primarily because they present with symptoms before completion of the processing of a NBS result and a confirmatory sweat test (6–8 weeks). Consequently, NBS is predicted to only benefit those infants who would otherwise present after this processing period, pragmatically designed here as beyond the first 2 months of life (late CD).

Using the United Kingdom Cystic Fibrosis Database (UKCFD), a well-validated disease register,8,16–18 we tested the hypothesis that diagnosis within 2 months of birth by NBS or if symptomatic, by early CD, is associated with improved clinical outcome compared with late CD. In addition, we tested whether outcomes were confounded by intercohort differences in treatment intensity. As CF clinical-phenotype and age at diagnosis vary considerably and are partially dependent on CF transmembrane conductance regulator genotype,19,20 we restricted our analysis to patients who were homozygous {Delta}F508 (50% of most populations of North European descent) and would be diagnosed by most NBS programs. This restriction had the additional advantage that >95% of these patients present clinically in early childhood, thus reducing ascertainment bias.


   PATIENTS AND METHODS
TOP
 ABSTRACT
 PATIENTS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Cohort
UKCFD data were collected, verified, and error checked as described recently8,17,21 and also at www.cystic-fibrosis.org.uk. All procedures were compliant with multicenter research ethics protocols and United Kingdom legislation on patient confidentiality. Patients between 1 and 10 years of age (between 2000 and 2002) were eligible for study from the 7294 UKCFD-registered CF clinic population. As described previously, NBS for CF in the United Kingdom is limited to a small number of regional CF centers (by 2002, ~12% of patients were diagnosed by NBS).8,16 Furthermore, the geographical area in which screening is available is not always geographically congruous with the referral population. Thus, some centers will be treating both patients diagnosed by NBS, as well as those diagnosed on the basis of symptom presentation. Therefore, the United Kingdom CF population offers the opportunity to compare outcomes between cohorts of patients diagnosed by NBS and CD.

As summarized in Fig 1, 3 cohorts were created: NBS (diagnosed by blood-spot screening followed by a confirmatory sweat test within 2 months of birth; 162 patients), early-CD (diagnosed by clinical presentation within 2 months of birth, including those diagnosed by MI; 404 patients), and late-CD (diagnosed by clinical presentation any time within 2 months of birth; 542 patients). Because patients diagnosed by early-CD would always present with symptoms within 2 months of birth irrespective of whether a NBS program was in existence or not, those patients diagnosed by both NBS and early-CD were included in the early-CD cohort. Patients for whom an age at diagnosis or mode of clinical presentation were not available, or for whom the diagnosis of CF was based on a significant family history (with or without clinical presentation) were excluded, because these patients would have been diagnosed early irrespective of NBS (59 patients). In the United Kingdom, antenatal or prenatal screening for CF is primarily ultrasound-based. Where evidence of echogenic bowel is found, an amniocentesis may be performed to confirm the diagnosis. With the exception of 1 early-CD patient and 1 late-CD patient, all eligible study patients were receiving pancreatic enzyme–replacement therapy as expected for {Delta}F508 homozygotes.


Figure 1
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 1 Identification of primary and secondary analysis NBS, early-CD, and late-CD study cohorts. Homozygous {Delta}F508 UKCFD patients aged 1 to 10 years were stratified into NBS, early-CD, and late-CD cohorts. At each level, patient numbers are indicated. The final number of matched data sets in the primary and secondary analyses are shown (left and right branches, respectively).

 

Figure 3
View larger version (24K):
[in this window]
[in a new window]

  		FIGURE 3 Increasing P aeruginosa infection status, not mode of diagnosis, determines long-term therapy demands (primary analysis). Shown in the median (interquartile range) number of long-term therapies for early-CD, late-CD, and NBS cohorts stratified according to no, intermittent, or chronic P aeruginosa infection. The black bars indicates the median value, and the black dots indicate the 95% CIs. a Significant (P < .01) difference between NBS and late-CD cohorts. For bars 1, 5, and 6, the median numbers of long-term therapies are 1, 2, and 2, respectively.

 
Two analyses were performed. In the primary analysis, a cross-sectional analysis of cohorts retrospectively ascertained for the years 2000 to 2002 was performed. After restriction to the most recent year of data collection (YDC), NBS patients were matched on a 1:1:1 basis, by patient age and YDC, using an {alpha}-numeric patient-identifier-based randomization process, to the equivalent early-CD and late-CD data sets. A total of 133 patients per diagnostic cohort were identified (Fig 1, primary analysis), demographic data for which are detailed in Table 1. This analysis facilitated direct comparison of the cohorts; however, restriction of the available data to the most recent data set per patient generated a preponderance of patients 10 years of age (31 [~ 23%] of 133). Given the association of increasing disease severity with increasing age, this may have skewed our analysis toward an older and, therefore, more severely affected population. To overcome this potential confounding variable, we performed a secondary analysis. In this secondary analysis, all 3 consecutive years of data for each patient were included (Fig 1, secondary analysis), and patients were matched as described for the primary analysis. Up to 3 annual sets of data were included in this matching process for patients with clinical data recorded at any time between 2000 and 2002 (153 NBS, 356 early-CD, and 481 late-CD patients). After excluding unmatched cases, 291 unique 3-way matches were identified, which generated data from 144 NBS, 169 early-CD, and 171 late-CD patients (Table 1). In contrast to the primary analysis, in the secondary analysis, matched data sets were evenly distributed across all ages.


View this table:
[in this window]
[in a new window]

  		TABLE 1 General Characteristics of Parent and Matched (Analyses 1 and 2) Study Cohorts

 
Clinical outcomes for early-CD and late-CD patients were compared against outcomes for the NBS cohort (a comparison of early-CD versus late-CD was not performed). In both primary and secondary analyses, it was not possible to match cohorts by treatment center because of insufficient patient numbers. As detailed in Table 1, no differences were found in the age in 2002 or age at diagnosis of CF, or in the proportion of patients presenting with MI, other gastrointestinal disorders, or respiratory symptoms between the parent populations and matched study cohorts.

Outcome Measures
Median outcomes were compared for each diagnostic cohort: height and weight z scores, Shwachman-Kulczyki morbidity (SK) scores, and percent predicted forced expiratory volume in 1 second (FEV1; in patients ≥6 years of age only). Where 2 or more records of an outcome measure were available with the same YDC, the mean of the values was used, as described previously.8,21 The median total number of long-term therapies (prescribed for >3 months; Table 2) and the proportion of patients receiving only low intensity therapy (as defined in Table 2), ≥3 long-term therapies, ≥2 nebulized therapies (NEBS), no intravenous antibiotics, and ≥2 courses of intravenous antibiotics were also compared. To determine whether treatment mirrored disease severity, Pseudomonas aeruginosa infection was used as a surrogate marker of severity. The UKCFD defines P aeruginosa infection as either intermittent (1 or 2 positive cultures in 12 months) or chronic (≥3 positive cultures in 12 months). NBS, early-CD, and late-CD cohorts were stratified according to P aeruginosa infection status, and long-term therapy requirements were determined. Outcome measures evaluated in the secondary analysis were height and weight z scores, SK score, FEV1, and the total number of long-term therapies. Because of the potential for confounding by repeated measures, direct intercohort comparisons were not permitted for this secondary analysis; however, the use of multiple data per patient facilitated age-matched multiple cross-sectional comparisons similar to those performed prospectively by Farrell et al.6,7,14


View this table:
[in this window]
[in a new window]

  		TABLE 2 Long-term (Administered for ≥3 Months) Therapies

 
Statistics
UKCFD analyses used Microsoft Access and Excel (Microsoft Corporation, Redmond, WA), with graphs created in SigmaPlot (SigmaPlot 4.01 for Windows; SPSS Inc, Chicago, IL). Statistical analyses were performed by using MINITAB 13.1 (Minitab Inc, State College, PA) and Stata SE 8.2 (Stata Corp, College Station, TX). In the primary analysis, data were not normally distributed; the Mann-Whitney test was used to determine statistically significant differences between medians. For the comparison of proportions of each cohort, the cohort sizes were considered to be large enough for the Central Limit Theorem to apply.22 Given the multiple comparisons, a P value of <.01 was considered to represent a significant difference. Indeed, using a Bonferroni correction for multiple comparisons, the actual P value should be .017, but we decided to reduce this to be more conservative. In the secondary analysis, height and weight z score, SK score, and FEV1 were compared by using a random effects analysis of variance model, which allowed for between-child variation in the number of years of contributing data and included a cohort effect, a time trend, and an interaction. Long-term therapies were compared by using a Poisson random-effects analysis of variance model and results were presented as relative risks (RRs).


   RESULTS
TOP
 ABSTRACT
 PATIENTS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Primary Analysis
NBS patients had significantly higher median height z scores and higher SK scores (ie, were healthier), and fewer patients were below the 10th percentile for height compared with late-CD patients (Table 3). No differences were found in FEV1. No differences were found in the clinical outcomes between NBS versus early CD or between early-CD versus late-CD cohorts.


View this table:
[in this window]
[in a new window]

  		TABLE 3 Difference (99% CI) in Clinical and Therapeutic Outcomes for NBS, Early-CD, and Late-CD

 
Patients in the NBS cohort received significantly fewer long-term therapies than late-CD patients. Furthermore, fewer NBS patients received ≥3 long-term therapies, ≥2 NEBS, or intravenous antibiotics. Indeed, a higher cumulative proportion of patients in the early-CD and late-CD cohorts received NEBS and recombinant human deoxyribonuclease (rhDNase) specifically, compared with those in the NBS cohort (Fig 2). Overall, no differences in long-term therapy outcomes for early-CD versus late-CD cohorts were found.


Figure 2
View larger version (19K):
[in this window]
[in a new window]

  		FIGURE 2 Higher use of NEBS and rhDNase in early-CD and late-CD cohorts compared with NBS (primary analysis). Shown is the cumulative age-related percentage of patients in early-CD, late-CD, and NBS cohorts receiving any long-term (>3 months) nebulized therapy (top; defined as an antibiotic [see Table 2] or rhDNase) or rhDNase (bottom). Note the difference in the y-axis scale.

 
Chronic P aeruginosa infection is a marker for a poor prognosis. Overall, no differences were found in the prevalence of chronic or intermittent P aeruginosa infection between the cohorts, but significantly fewer 1- to 5-year-olds who were NBS were chronically infected with P aeruginosa than in the late-CD cohort (3% vs 14%; difference of –11%; 95% confidence interval [CI]: –1.4 to –20.5). Patients who were NBS and free from P aeruginosa infection (Fig 3) received fewer long-term therapies compared with matched late-CD but not early-CD patients. No difference in long-term therapies was found between intermittent or chronically P aeruginosa-infected cohorts. For all cohorts, a higher number of long-term therapies was associated with P aeruginosa infection status in the rank order: no P aeruginosa < intermittent P aeruginosa infection < chronic P aeruginosa infection. No differences were found in the proportion of patients with P aeruginosa infection (intermittent or chronic) or in respect to treatment requirements for patients with and without P aeruginosa infection between early-CD and late-CD cohorts (data not shown).

Secondary Analysis
Late-CD but not early-CD patients had significantly lower height z scores compared with NBS patients, which persisted until 7.5 years of age (average difference: 0.87 SD score [SDS]; 95% CI: –0.48 to –1.25; P = .0001 and –0.12 SDS; 95% CI: –0.27 to 0.51; P = .555, respectively; Fig 4). Compared with the NBS cohort, only the height z score for the late-CD cohort increased significantly with time (average: 0.09 SDS per year; 95% CI: 0.03 to 0.15). Both late-CD and early-CD cohorts had lower SK scores compared with the NBS cohort, although the difference between the NBS and early-CD cohorts did not reach significance (average difference of –6.36; 95% CI: –10.73 to –2.00; P = .004 and –3.47; 95% CI: –7.72 to 0.78; P = .110, respectively; Fig 3). Irrespective of mode of diagnosis, SK score was found to decrease significantly with time (–0.97 per year; 95% CI: –1.41 to –0.53; P < .001). However, no difference was found in the temporal trend for SK score between NBS and late-CD or NBS and early-CD cohorts (0.49 per year; –0.19–1.15 and 0.14 per year; –0.52–0.80, respectively). For weight z score and percent FEV1, no differences were found between NBS and late-CD cohorts or NBS and early-CD cohorts.


Figure 4
View larger version (28K):
[in this window]
[in a new window]

  		FIGURE 4 NBS associates with improved clinical outcomes and fewer long-term therapies compared with late-CD but not early-CD (secondary analysis). Shown is a comparison of mean (SEM) of height z score (top), SK score (second-to-bottom), and long-term therapy requirements (bottom) for NBS versus late-CD (left column) cohorts and NBS versus early-CD (right column) cohorts. Numbers indicate the number of patients. Also shown is the proportion of each cohort according to age below the 10th percentile for height (second-to-top). The total numbers of patients in each age group are shown in the top figures.

 
Compared with the NBS cohort, a trend toward a higher long-term therapy requirement was observed for the late-CD (average RR: 1.30; 95% CI: 0.99 to 1.70; P = .063), but not early-CD (average RR: 1.10; 95% CI: 0.83 to 1.45; P = .505) cohorts, although the difference bordered on significance (Fig 4). As expected for a chronic disease, irrespective of the mode of diagnosis, increasing age was significantly associated with a 6% increase per year in long-term therapy requirements (average RR: 1.06; 95% CI: 1.03 to 1.10; P < .001).


   DISCUSSION
TOP
 ABSTRACT
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
According to the UKCFD, in the absence of a NBS program, patients with CF who are homozygous for the common {Delta}F508 mutation most often present after 2 months of age, with over 95% diagnosed by 6 years of age (E.J.S., unpublished data, 2005). Our data demonstrate that the benefits of NBS associate with early, presymptomatic diagnosis (within 2 months of birth) compared with a CD after the typical 2-month NBS reporting window. Compared with their late diagnosed, genetically matched peers, NBS diagnosed patients were taller, had reduced morbidity (ie, higher SK score), and received less treatment but nevertheless had similar lung function. Importantly, we demonstrate that although receiving fewer therapies, a similar pulmonary outcome can be achieved in genetically similar patients with CF diagnosed by NBS as compared with patients who were CD. Indeed, as clinicians respond to signs and symptoms of increasing disease severity, our data here, and elsewhere for patients with CF of mixed genotypes,8,16 suggest that treatment burden is a reasonable and justifiable surrogate marker for increasing disease severity (ie, lower SK score) rather than poor pulmonary function, which may manifest later. An alternative explanation for our results could be that clinicians may treat late-CD patients more aggressively than NBS on the basis that they had been diagnosed later. However, the higher morbidity and worse height z score of the late-CD cohort, and comparable treatment in response to intermittent or chronic P aeruginosa infection to that of NBS or early-CD patients, would suggest that irrespective of mode of diagnosis, patients were treated proportionately against signs and symptoms rather than on a preventative basis. Indeed, Cystic Fibrosis Trust (United Kingdom) guidelines advocate treatment appropriate to disease severity, with the exception of prophylactic anti-Staphylococcus aureus therapy (oral flucloxacillin). As might be expected, we found no therapeutic or nutritional difference between NBS and early-CD cohorts, suggesting that NBS benefits most those patients who would otherwise have been missed or would have been subject to delayed diagnoses (eg, the late-CD cohort), the precise rationale for a screening program.

No previous comparisons of NBS and early-CD cohorts have been reported, to our knowledge, although 2 comparisons of NBS and MI-diagnosed cohorts have been described.23,24 Both report comparable clinical outcomes to those shown here, although in contrast to our study, the above studies showed significantly lower FEV1 for cohorts diagnosed with MI compared with NBS. However, restriction of our early-CD cohort to MI-diagnosed patients (86 of 133) did not alter our results (data not shown). Li et al24 reported that the reduced FEV1 for their MI-diagnosed cohort only become apparent by the age of 8 to 10 years, and the younger age of our cohorts could explain this discrepancy. Alternatively, improvements in surgical care or more aggressive treatment of P aeruginosa infection in the United Kingdom25 compared with that used in Australia23 and the United States24 could be explanatory factors. Another remote possibility is that, although all of our patients were homozygous for the {Delta}F508 genotype, the early-CD cohort may represent a subset manifesting a more severe phenotype in a spectrum of {Delta}F508 severity.

One single center study compared early-CD and late-CD patients.26 In agreement, we found no difference in the clinical outcomes of matched early-CD and late-CD cohorts. However, unlike that longitudinal study, we found only minimal differences in treatment requirements. Because their patients were recruited over a 17-year period, during which new pancreatic therapies and clinical, dietary, and antimicrobial6,14 protocols were introduced, it is possible that this study could have been confounded by a cohort effect.27

Differences in the age distribution of the cohorts in the primary and secondary analyses may account for the significant excess in long-term therapies found in the late-CD cohort compared with NBS cohort in the primary analysis, but that is only observed as a trend in the secondary analysis. The persistence of lower treatment requirements and improved morbidity after NBS with increasing age could explain why pregnant females with CF have highly significantly better fertility outcomes if they were diagnosed by NBS,28 suggesting that the stress of pregnancy is better tolerated in screened cohorts. Thus, NBS might benefit the next generation of families affected by NBS.

Our analyses may be subject to several biases. First, the use of patient matching could have introduced selection bias. However, as patients were matched randomly and no differences in the age at diagnosis between our parent population and the derived cohorts were found, this is highly unlikely. Also, although restriction of the study cohorts to those homozygous for the {Delta}F508 genotype maximized the comparability of the cohorts in terms of potential disease severity, it reduced the likelihood of inclusion of patients with atypical CF phenotypes associated with genotypes other than homozygous {Delta}F508 that may be diagnosed via NBS (depending on screening protocol), thereby potentially biasing the study against NBS. Second, the stratification of patients into NBS, early-CD, and late-CD cohorts could have introduced ascertainment bias. Furthermore, some homozygous {Delta}F508 patients with a very mild CF disease phenotype may be represented in the NBS screening group but not in the late-CD cohort because they had not been diagnosed within the time frame of the study, thus biasing the study in favor of NBS. However, because only 64 (1.68%) of 3797 patients homozygous for the {Delta}F508 genotype were diagnosed after the 10 years of age, it is unlikely that this would have a significant impact on our results. Also, CD patients with very severe disease that would have otherwise been included in the late-CD cohort could have died before the study was performed, potentially biasing the study against NBS. However, again, this is unlikely as a recent review of deaths attributable to CF revealed that only 3 patients under the age of 10 died in 2002, of which 2 were early- CD (both by MI) and 1 was late-CD (A.M., unpublished data, 2005). Indeed, a recent systematic review examining the potential impact of NBS on child survival concluded that up to 10 years old, NBS associates with a reduction in mortality risk of between 2 and 10 per 100 children with CF (without MI) compared with CD children with CF.29 Third, the use of data spanning a 3-year time period whose aggregation produces a study population spanning up to 13 years could have introduced cohort bias.27 However, because modern nutritional care (especially acid-resistant pancreatic enzyme–replacement therapy) has been available to all patients since the late 1980s, it is unlikely that improved nutrition would have introduced a large cohort effect. Furthermore, because all patients in the study were receiving CF care at specialist CF centers or care provided in association with a Specialist CF Centre under a shared care framework, it is likely that the cohorts would have received comparable care and antimicrobial therapy according to National guidelines,25,30 although some clinician preferences may exist. Indeed, between 1990 and 2002, no substantial changes in treatment practices or guidelines25,30 for CF have been made. In addition, given the direct comparability of our NBS outcome data in the United Kingdom with that in a prospective US study,8 and similarity with outcome data from other National Registries reported in the literature,6,7,9,13,15 it is unlikely that our results could be explained by an unknown bias or potential pitfalls of database analyses as discussed in detail elsewhere.31,32 As NBS programs exist in the United Kingdom in only a minority of specialist CF centers, it could be argued that the differences we have reported here may be attributable to differences in clinical practice within the small number of CF centers performing NBS and not mode of diagnosis. However, this is highly unlikely, because we have previously demonstrated a comparable magnitude of benefit in terms of height, SK score, and number of long-term therapies for NBS compared with CD patients when we restricted our analysis to only those centers that received 15% to 85% of referrals from NBS, as compared with all UK CF specialist centers.16 Finally, although we report significant differences in the clinical and treatment outcomes between patients who would present clinically with CF before the age of 10, we are unable to determine if a diagnosis by NBS would have benefit over CD in those patients with mild CF disease and that would otherwise present with symptoms later in life.

Up to and including 2002, a reactive rather than proactive approach was used in the treatment of CF in the United Kingdom.25 Indeed, with the exception of long-term oral flucloxacillin, the use of long-term therapies including the mucolytic, rhDNase (because of funding limitations) and nebulized antibiotics (including tobramycin) was primarily focused on attenuating rather than preventing increasing disease severity. It should also be noted that up to and including 2002, the nebulized antibiotic, tobramycin, was primarily used in the United Kingdom as a second-line treatment for chronic P aeruginosa infection after resistance or intolerance to colistin or gentamicin had been exhibited, whereas azithromycin, a macrolide with immunomodulatory properties was only used in a very small number of centers. The lower treatment requirements associated with NBS that we demonstrate here, would suggest that considerable cost savings could be implied in favor of NBS. However, the demonstration that rhDNase is potentially most beneficial in patients with mild CF disease rather than those with moderate-to-severe disease has led to its increasing earlier proactive use in younger, milder, patients.33 Given the high costs associated with the long-term rhDNase and TOBI, the increasing use of these therapies may absorb some of these potential cost-savings. On the other hand, if the use of rhDNase in NBS diagnosed patients with very mild CF disease additional maintains the good health of these children, fewer hospital visits could lead to substantial reductions in hospital and indirect costs.

Crucially, policy makers have widely used the reasoning that CF screening is of no benefit because lung function is no better for patients with CF whether they have been screened at birth or not. However, although our data confirms no difference in lung function between pediatric patients who are diagnosed by NBS or late-CD, children diagnosed by NBS received significantly fewer long-term therapies and had improved growth and morbidity compared with their late-CD controls. Furthermore, because the rate of decline of SK score (1% per annum) was consistent across all cohorts and no temporal change was observed, the lower SK score for the late-CD cohort could be attributable to increased morbidity (possibly related to untreated airway inflammation, malnutrition, or pulmonary infection) in the period before diagnosis. Because patients who received fewer long-term therapies were found to have improved morbidity, it is not unreasonable to suggest that these patients are likely to incur lower health care costs. Therefore, early, presymptomatic diagnosis by NBS may afford potential long-term cost-savings to the drugs budget and health care budget over late clinical presentation, which could offset part or all of the cost of running a NBS program, although the increasing use of prophylactic therapies (eg, rhDNase) in patients with mild symptoms may offset some of these potential savings.

Shwachman et al34 created the current definition of early diagnosis of CF of within 3 months of birth in the 1970s. However, since this landmark article, significant improvements in laboratory based, rapid throughput technology for immunoreactive trypsinogen (IRT) assays and CF transmembrane conductance regulator DNA mutation screening means that it is now feasible for a positive diagnosis of CF using an IRT/IRT or IRT/DNA (or a combination) screening program to be made within 2 months of birth. Indeed, as indicated in Fig 5, excluding those patients diagnosed by the presentation of symptoms within 2 months of birth (eg, the early-CD) group, increasing the window of diagnosis from 2 months to 3 months or beyond increases the proportion of patients with poor, long-term growth (eg, height below the 10th percentile). On the basis of our results, we, therefore, propose that the definition of early diagnosis of CF is redefined to a diagnosis made within 2 months of birth.


Figure 5
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 5 Increasing window of early diagnosis associates with worse growth outcome. Shown is a sensitivity analysis of the percentage of patients below the 10th percentile for height (A), weight (B), and BMI (C) in early-diagnosed and late-diagnosed cohorts with increasing window of early diagnosis. Patients diagnosed by presentation of clinical symptoms within 2 months of age are excluded from this analysis. Results shown at 2 months of age correspond to results shown in Table 3. As the definition of early diagnosis increases to the right, late-diagnosed patients move into the earlydiagnosed cohort, increasing the size of the early-diagnosed cohort as shown in the table at the bottom of the figure.

 

   CONCLUSIONS
TOP
 ABSTRACT
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
In the absence of a NBS program, patients with CF diagnosed by symptom presentation after the age of 2 months manifest worse clinical outcomes despite receiving higher levels of long-term therapy for at least the first 10 years. NBS for CF provides an opportunity to maximize the clinical potential of patients whose survival may otherwise be limited to early adulthood. Although NBS will not be of benefit for patients presenting with symptoms within the 2 months required to obtain and confirm an NBS result (early-CD), we have demonstrated that these patients have comparable clinical outcomes to symptom-free patients diagnosed by NBS, although they required treatment of a higher intensity. In assessing the merits of implementing a NBS program, we suggest that differences in treatment intensity should be recognized as an important confounding outcome measure.


   ACKNOWLEDGMENTS
 
We are grateful for the financial support of the CF Trust and the National Services Division of the National Health Service (Scotland). Neither funding body played a role in the design or conduct of this study, analysis or interpretation of the data, or preparation, review, or approval of the manuscript. Both funding bodies provided funds to assist the collection and management of data for this project.

Data were supplied by the directors of specialist CF centers in the United Kingdom. We thank M. Fraser and S. Krawczyk (UKCFD) for expert data validation and the directors and data managers at specialist CF centers and CF clinics throughout the United Kingdom who contributed data to the UKCFD. We also thank Dr Anne Thomson (Oxford) for critical review.


   FOOTNOTES
 
Accepted Sep 25, 2006.

Address correspondence to Erika J. Sims, PhD, School of Medicine, Health Policy and Practice, University of East Anglia, University Plain, Norwich NR4 7TJ, United Kingdom. E-mail: e.sims{at}uea.ac.uk

Financial Disclosure: Dr Sims and Ms G. Mehta were funded by a grant awarded by the Cystic Fibrosis Trust at the time this study was performed. Dr A. Mehta was the director of the United Kingdom Cystic Fibrosis Database and received grants from the Cystic Fibrosis Trust. Drs Clark, McCormick, and Connett have indicated they have no financial relationships relevant to this article to disclose.


   REFERENCES
TOP
 ABSTRACT
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 

  1. Davis PB, Drumm M, Konstan MW. Cystic fibrosis. Am J Respir Crit Care Med. 1996;154 :1229 –1256[ISI][Medline]
  2. McCormick J, Green MW, Mehta G, Culross F, Mehta A. Demographics of the UK cystic fibrosis population: implications for neonatal screening. Eur J Hum Genet. 2002;10 :583 –590[CrossRef][ISI][Medline]
  3. Rosenstein BJ, Langbaum TS. Diagnostic delay in Duchennes muscular dystrophy and cystic fibrosis. JAMA. 1982;247 :3313[CrossRef][ISI][Medline]
  4. Wilcken B, Towns SJ, Mellis CM. Diagnostic delay in cystic fibrosis: lessons from newborn screening. Arch Dis Child. 1983;58 :863 –866[Abstract]
  5. Farrell PM, Mischler EH. Newborn screening for cystic fibrosis: the Cystic Fibrosis Neonatal Screening Study Group. Adv Pediatr. 1992;39 :35 –70[Medline]
  6. Farrell PM, Kosorok MR, Laxova A, et al. Nutritional benefits of neonatal screening for cystic fibrosis. Wisconsin Cystic Fibrosis Neonatal Screening Study Group. N Engl J Med. 1997;337 :963 –969[Abstract/Free Full Text]
  7. Farrell PM, Kosorok MR, Rock MJ, et al. Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Pediatrics. 2001;107 :1 –13[Abstract/Free Full Text]
  8. Sims EJ, McCormick J, Mehta G, Mehta A. Neonatal screening for cystic fibrosis is beneficial even in the context of modern treatment. J Pediatr. 2005;147 (3 suppl):S42–S46
  9. Waters DL, Wilcken B, Irwing L, et al. Clinical outcomes of newborn screening for cystic fibrosis. Arch Dis Child Fetal Neonatal Ed. 1999;80 :F1 –F7[Abstract/Free Full Text]
  10. Dankert-Roelse JE, te Meerman GJ. Long term prognosis of patients with cystic fibrosis in relation to early detection by neonatal screening and treatment in a cystic fibrosis centre. Thorax. 1995;50 :712 –718[Abstract]
  11. Merelle ME, Schouten JP, Gerritsen J, Dankert-Roelse JE. Influence of neonatal screening and centralized treatment on long-term clinical outcome and survival of CF patients. Eur Respir J. 2001;18 :306 –315[Abstract/Free Full Text]
  12. Mastella G, Zanolla L, Castellani C, et al. Neonatal screening for cystic fibrosis: long-term clinical balance. Pancreatology. 2001;1 :531 –537[CrossRef][ISI][Medline]
  13. McKay KO, Waters DL, Gaskin KJ. The influence of newborn screening for cystic fibrosis on pulmonary outcomes in new South Wales. J Pediatr. 2005;147 (3 suppl):S47–S50
  14. Farrell PM, Li Z, Kosorok MR, et al. Bronchopulmonary disease in children with cystic fibrosis after early or delayed diagnosis. Am J Respir Crit Care Med. 2003;168 :1100 –1108[Abstract/Free Full Text]
  15. Siret D, Bretaudeau G, Branger B, et al. Comparing the clinical evolution of cystic fibrosis screened neonatally to that of cystic fibrosis diagnosed from clinical symptoms: a 10-year retrospective study in a French region (Brittany). Pediatr Pulmonol. 2003;35 :342 –349[CrossRef][ISI][Medline]
  16. Sims EJ, McCormick J, Mehta G, Mehta A. Newborn screening for cystic fibrosis associates with reduced treatment intensity. J Pediatr. 2005;147 :306 –311[CrossRef][ISI][Medline]
  17. Sims EJ, Green MW, Mehta A. Decreased lung function in female but not male subjects with established cystic fibrosis-related diabetes. Diabetes Care. 2005;28 :1581 –1587[Abstract/Free Full Text]
  18. McCormick J, Ogston SA, Sims EJ, Mehta A. Asians with cystic fibrosis in the UK have worse disease outcomes than clinic matched white homozygous delta F508 controls. J Cyst Fibros. 2005;4 :53 –58[CrossRef][Medline]
  19. Cystic Fibrosis Genotype-Phenotype Consortium. Correlation between genotype and phenotype in patients with cystic fibrosis. N Engl J Med. 1993;329 :1308 –1313[Abstract/Free Full Text]
  20. Zielenski J. Genotype and phenotype in cystic fibrosis. Respiration. 2000;67 :117 –133[CrossRef][ISI][Medline]
  21. Mehta G, Sims EJ, Culross F, McCormick J, Mehta A. Potential benefits of the UK Cystic Fibrosis Database. J R Soc Med. 2004;97 (suppl 44):60–71
  22. Bland M. An Introduction to Medical Statistics. Oxford, United Kingdom: Oxford University Press; 1990
  23. Evans AK, Fitzgerald DA, McKay KO. The impact of meconium ileus on the clinical course of children with cystic fibrosis. Eur Respir J. 2001;18 :784 –789[Abstract/Free Full Text]
  24. Li Z, Lai HJ, Kosorok MR, Laxova A, Rock MJ, Splaingard ML, Farrell PM. Longitudinal pulmonary status of cystic fibrosis children with meconium ileus. Pediatr Pulmonol. 2004;38 :277 –284[CrossRef][ISI][Medline]
  25. Littlewood JM, Bevan A, Connett G, et al. Antibiotic Treatment for Cystic Fibrosis: Report of the UK Cystic Fibrosis Trust Antibiotic Group. 2nd ed. Bromley, Kent, United Kingdom: Cystic Fibrosis Trust; 2002
  26. Connett GJ, Yeatman SL. Delayed diagnosis in cystic fibrosis patients homozygous for the delta F508 mutation results in increased treatment to maintain health in childhood [abstract]. Pediatr Pulmonol. 2002;24S :A407
  27. Dodge JA, Morison S, Lewis PA, et al. Incidence, population, and survival of cystic fibrosis in the UK, 1968–95. Arch Dis Child. 1997;77 :493 –496[Abstract/Free Full Text]
  28. Boyd JM, Mehta A, Murphy DJ. Fertility and pregnancy outcomes in men and women with cystic fibrosis in the United Kingdom. Hum Reprod. 2004;19 :2238 –2243[Abstract/Free Full Text]
  29. Grosse SD, Rosenfeld M, Devine OJ, Lai HJ, Farrell PM. Potential impact of newborn screening for cystic fibrosis on child survival: a systematic review and analysis. J Pediatr. 2006;149 :362 –366[CrossRef][ISI][Medline]
  30. Standards for the Clinical Care of Children and Adults With Cystic Fibrosis in the UK 2001. Bromley, Kent, United Kingdom: Cystic Fibrosis Trust; 2001
  31. Black N. Developing high quality clinical databases. BMJ. 1997;315 :381 –382[Free Full Text]
  32. Black N. High-quality clinical databases: breaking down barriers. Lancet. 1999;353 :1205 –1206[CrossRef][ISI][Medline]
  33. Quan JM, Tiddens HA, Sy JP, et al. A two-year randomized, placebo-controlled trial of dornase alfa in young patients with cystic fibrosis with mild lung function abnormalities. J Pediatr. 2001;139 :813 –820[CrossRef][ISI][Medline]
  34. Shwachman H, Redmond A, Khaw KT. Studies in cystic fibrosis: report of 130 patients diagnosed under 3 months of age over a 20-year period. Pediatrics. 1970;46 :335 –343[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics



This article has been cited by other articles:


Home page
 Arch. Dis. Child.Home page
I. M Balfour-Lynn
Newborn screening for cystic fibrosis: evidence for benefit
Arch. Dis. Child., January 1, 2008; 93(1): 7 - 10.
[Full Text] [PDF]

Home page
 Proc Am Thorac SocHome page
L. B. Ware
Clinical Year in Review III: Asthma, Lung Transplantation, Cystic Fibrosis, Acute Respiratory Distress Syndrome
Proceedings of the ATS, September 15, 2007; 4(6): 489 - 493.
[Full Text] [PDF]

Home page
 PediatricsHome page
P. M. Farrell
The Meaning of "Early" Diagnosis in a New Era of Cystic Fibrosis Care
Pediatrics, January 1, 2007; 119(1): 156 - 157.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow P3Rs: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when P3Rs are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sims, E. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sims, E. J.
Related Collections
Right arrow Respiratory Tract