Data that empower: The success and promise of CF patient registries

Global disparities in cystic fibrosis outcomes prior to CFTR modulators: A CF registries cohort study in South Africa and Canada

BACKGROUND

Outcomes of cystic fibrosis (CF) differ between low-middle income and high-income countries, but comparative data are lacking. We compared South African (SA) and Canadian CF outcomes to explore what disparities existed prior to access of CFTR modulators in Canada.

METHODS

A cross-sectional study of SA and Canadian CF registries data for period 1 January to 31 December 2018. CF registry data were harmonised between countries to compare lung function and nutrition outcomes. Poor nutrition was defined as BMIz-score < -1 in children and < 18.5 kg/m2 in adults. Standardised mean difference (SMD) >10 was considered significant.

RESULTS

After excluding Canadians on CFTR modulators and lung transplant recipients, data on 4049 Canadian and 446 SA people was analysed. Compared to Canada, people in SA were younger (median age 15.8 years vs. 24.1 years: SMD 52) with fewer males (47.8% vs 54.2%; SMD 12.5) and White (70.9% vs. 93.3%; SMD 61.3). Class I-III CFTR mutation frequency was similar in SA (n = 384, 86.1%) and Canada (n = 3426, 84.9%). After adjusting for age, gender, diagnosis age, genotype, P.aeruginosa infection and pulmonary treatments, FEV1pp was 8.9% lower (95% CI 6.3% to 11.4%) and poor nutrition 1.7-fold more common (OR 1.70; 95% CI 1.19-2.41) in SA compared to Canada.

CONCLUSION

Lung function and nutrition was significantly lower in SA compared to Canada. Global disparities in CF outcomes between high and low-middle income countries are likely to widen as CFTR modulators are rapidly scaled up in only high-income countries.

Establishing a Cystic Fibrosis Learning Network: Interventions to promote collaboration and data-driven improvement at scale

Introduction

A learning health network is a type of learning health system in which stakeholders use network organization to improve health and health care. Building on existing resources in the cystic fibrosis (CF) community, the Cystic Fibrosis Learning Network (CFLN) was designed to improve medical outcomes and quality of life through an intentional focus on achieving reliable evidence-based chronic care delivery and creating a system for data-driven collaborative learning.

Methods

We describe the development and growth of the CFLN considering six domains of a Network Maturity Grid: system leadership; governance and policy management; quality improvement (QI); engagement and community building; data and analytics; and research. We illustrate the impact of the CFLN experience on chronic care processes and indicators of collaborative infrastructure.

Results

The CFLN represents 36 accredited care centers in the CF Foundation Care Center Network caring for over 6300 patients. Of 6779 patient clinical care visits/quarter, 77% are entered into the CF Foundation Patient Registry within 30 days, providing timely means to track outcomes. Collaborative visit planning is occurring in 93% of clinical care visits to share agenda setting with patients and families. Almost all CFLN teams (94%, n = 34) have a patient/family partner (PFP), and 74% of PFPs indicate they are actively participating, taking ownership of, or leading QI initiatives with the interdisciplinary care team. In 2022, 97% of centers reported completing 1-13 improvement cycles per month, and 82% contributed to monthly QI progress reports to share learning.

Conclusion

The CFLN is a maturing, collaborative infrastructure. CFLN centers practice at an advanced level of coproduction. The CFLN fosters interdisciplinary and PFP leadership and the performance of consistent data-driven improvement cycles. CFLN centers are positioned to respond to rapid changes in evidence-based care and advance the practice of QI and implementation science on a broader scale.

Global burden of nontuberculous mycobacteria in the cystic fibrosis population: a systematic review and meta-analysis

Background

People living with cystic fibrosis have an increased risk of lung infection with nontuberculous mycobacteria (NTM), the prevalence of which is reportedly increasing. We conducted a systematic review of the literature to estimate the burden (prevalence and incidence) of NTM in the cystic fibrosis population.

Methods

Electronic databases, registries and grey literature sources were searched for cohort and cross-sectional studies reporting epidemiological measures (incidence and prevalence) of NTM infection or NTM pulmonary disease in cystic fibrosis. The last search was conducted in September 2021; we included reports published since database creation and registry reports published since 2010. The methodological quality of studies was appraised with the Joanna Briggs Institute tool. A random effects meta-analysis was conducted to summarise the prevalence of NTM infection, and the remaining results are presented in a narrative synthesis.

Results

This review included 95 studies. All 95 studies reported on NTM infection, and 14 of these also reported on NTM pulmonary disease. The pooled estimate for the point prevalence of NTM infection was 7.9% (95% CI 5.1-12.0%). In meta-regression, sample size and geographical location of the study modified the estimate. Longitudinal analysis of registry reports showed an increasing trend in NTM infection prevalence between 2010 and 2019.

Conclusions

The overall prevalence of NTM infection in cystic fibrosis is 7.9% and is increasing over time based on international registry reports. Future studies should report screening frequency, microbial identification methods and incidence rates of progression from NTM infection to pulmonary disease.

Gut Dysbiosis in Children with Cystic Fibrosis: Development, Features and the Role of Gut-Lung Axis on Disease Progression

Cystic fibrosis (CF) is the most common autosomal recessive disease among Caucasians. Over the last 20 years, culture-independent analysis, including next-generation sequencing, has paired with culture-based microbiology, offering deeper insight into CF lung and gut microbiota. The aim of this review is to analyse the features of gut microbiota in patients with CF and its possible role in the progression of the disease, establishing the basis for a potential role in microbe-based therapies. The literature analysis showed that the gut environment in CF patients has unique features due to the characteristics of the disease, such as decreased bicarbonate secretion, increased luminal viscosity, and an acidic small intestinal environment, which, due to the treatment, includes regular antibiotic use or a high-energy and fat-dense diet. As a result, the gut microbial composition appears altered, with reduced richness and diversity. Moreover, the population of pro-inflammatory bacteria is higher, while immunomodulatory genera, such as Bacteroides and Bifidobacterium, are scarcer. The imbalanced gut microbial population has a potential role in the development of systemic inflammation and may influence clinical outcomes, such as respiratory exacerbations, spirometry results, and overall growth. Although a better understanding of the pathophysiology behind the gut-lung axis is needed, these findings support the rationale for considering gut microbiota manipulation as a possible intervention to regulate the severity and progression of the disease.

Sharing is caring: a call for a new era of rare disease research and development

Scientific advances in the understanding of the genetics and mechanisms of many rare diseases with previously unknown etiologies are inspiring optimism in the patient, clinical, and research communities and there is hope that disease-specific treatments are on the way. However, the rare disease community has reached a critical point in which its increasingly fragmented structure and operating models are threatening its ability to harness the full potential of advancing genomic and computational technologies. Changes are therefore needed to overcome these issues plaguing many rare diseases while also supporting economically viable therapy development. In "Data silos are undermining drug development and failing rare disease patients (Orphanet Journal of Rare Disease, Apr 2021)," we outlined many of the broad issues underpinning the increasingly fragmented and siloed nature of the rare disease space, as well as how the issues encountered by this community are representative of biomedical research more generally. Here, we propose several initiatives for key stakeholders - including regulators, private and public foundations, and research institutions - to reorient the rare disease ecosystem and its incentives in a way that we believe would cultivate and accelerate innovation. Specifically, we propose supporting non-proprietary patient registries, greater data standardization, global regulatory harmonization, and new business models that encourage data sharing and research collaboration as the default mode. Leadership needs to be integrated across sectors to drive meaningful change between patients, industry, sponsors, and academic medical centers. To transform the research and development landscape and unlock its vast healthcare, economic, and scientific potential for rare disease patients, a new model is ultimately the goal for all.

Incidence of thrombotic microangiopathies in Quebec: insight from a laboratory centralizing ADAMTS-13 testing

BACKGROUND

Thrombotic microangiopathies (TMA) are serious medical conditions requiring a prompt diagnosis to adapt treatment. The determination of ADAMTS-13 activity enables discriminating thrombotic thrombocytopenic purpura (TTP) from other forms of TMA. The purpose of this study was to provide an estimate of the incidence of TTP and TMA in the Canadian Quebec province using data collected from a laboratory centralizing ADAMTS-13 testing for the whole province.

RESULTS

From 2012 to 2019, 846 patients were evaluated for plasma ADAMTS-13 activity due to a suspicion of TMA. TTP was identified in 147 patients. Of these, 118 patients with a median age of 51.5 years and a male-female ratio of 1:1.4 had their first episode of TTP during the study period. The number of ADAMTS-13 tests performed and the number of patients with suspected TMA increased annually by 19% and 21% respectively. While the incidence of non-TTP TMA increased annually, that for TTP remained unchanged. This averaged 10.2 (95% CI 5.9-14.4) per million persons per year for suspected non-TTP TMA and 1.8 (95% CI 1.3-2.4) for confirmed TTP. The incidence rate of TMA other than TTP was higher in the age group 70-79 years (21.8; 95% CI 5.4-38.1) for females and in the age group 80-89 years (24.4; 95% CI 7.2-41.7) for males compared to other age groups. The incidence rate of TTP was higher in the age group 40-49 years (4.0; 95% CI 2.0-5.9) for women and in the age group 60-69 years (3.4; 95% CI 1.1-5.6) for men compared to other age groups.

CONCLUSION

The analysis of centralized data measuring ADAMTS-13 activity allowed us to adequately establish the incidence rate and demographic characteristics of TMA, particularly TTP, in Quebec. TTP incidence remained stable while suspected non-TTP TMA steadily increased from 2012 to 2019.

Trends in cystic fibrosis survival over 40 years in South Africa: An observational cohort study

INTRODUCTION

Temporal trends in cystic fibrosis (CF) survival from low-middle-income settings is poorly reported. We describe changes in CF survival after diagnosis over 40 years from a South African (SA) CF center.

METHODS

An observational cohort study of people diagnosed with CF from 1974 to 2019. Changes in age-specific mortality rates from 2000 (vs. before 2000) were estimated using multivariable Poisson regression. Data were stratified by current age < or ≥10 years and models controlled for diagnosis age, sex, ethnicity, genotype, and Pseudomonas aeruginosa (PA) infection. A second analysis explored the association of mortality with weight and forced expiratory volume in 1 s reported as z-scores (FEV1z-scores) at age 5-8 years.

RESULTS

A total of 288 people (52% male; 57% Caucasian; 44% p.Phe508del homozygous) were included (median diagnosis age 0.5 years: Q1, Q3: 0.2, 2.5); 100 (35%) died and 30 (10%) lost to follow-up. Among age >10 years, age-specific mortality from 2000 was significantly lower (adjusted hazard ratio [aHR]: 0.14; 95% confidence interval [CI]: 0.06, 0.29; p < 0.001), but not among age <10 years (aHR: 0.67; 95% CI: 0.28, 1.64; p = 0.383). In children <10 years, Caucasian ethnicity was associated with lower mortality (aHR 0.17; 95% CI: 0.05, 0.63), and longer times since first PA infection with higher mortality (aHR: 1.31; 95% CI: 1.01, 1.68). Mortality was sevenfold higher if FEV1z was <-2.0 at age 5-8 years (aHR: 7.64; 95% CI: 2.58, 22.59).

CONCLUSION

Overall, CF survival has significantly improved in SA from 2000 in people older than 10 years. However, increased risk of mortality persists in young non-Caucasian children, and with FEV1z <-2.0 at age 5-8 years.

A model for building a national, patient-driven database to track contraceptive use in women with rare diseases

Data on the safety and effectiveness of contraception among women with rare diseases are critical and sorely lacking. To fill this gap, we propose a national, patient-driven database that tracks contraceptive safety and effectiveness among women with rare diseases. We built a pilot database focusing on women with cystic fibrosis in 3 phases: (1) database design input from patients and experts, (2) merging of contraceptive survey data with relevant clinical outcomes from the Cystic Fibrosis Foundation Patient Registry (CFFPR), and (3) forming a data guide to facilitate accessible output data. We successfully linked 62 contraceptive survey variables with 362 relevant clinical outcome variables for 150 patients. This pilot represents a breakthrough in linking contraceptive data to disease-specific outcomes and informs how to build a national, patient-driven contraceptive database for women with rare diseases.

Demographic characteristics, clinical and laboratory features, and the distribution of pathogenic variants in the CFTR gene in the Cypriot cystic fibrosis (CF) population demonstrate the utility of a national CF patient registry

BACKGROUND

Specialized clinical care for cystic fibrosis (CF) in Cyprus, a small island country, has been implemented since the 1990s. However, only recently, a national CF patient registry has been established for the systematic recording of patients' data. In this study, we aim to present data on the epidemiological, genotypic and phenotypic features of CF patients in the country from the most recent data collection in 2019, with particular emphasis on notable rare or unique cases.

RESULTS

Overall, data from 52 patients are presented, 5 of whom have deceased and 13 have been lost to follow-up in previous years. The mean age at diagnosis was 7.2 ± 12.3 years, and the mean age of 34 alive patients by the end of 2019 was 22.6 ± 13.2 years. Patients most commonly presented at diagnosis with acute or persistent respiratory symptoms (46.2%), failure to thrive or malnutrition (40.4%), and dehydration or electrolyte imbalance (32.7%). Sweat chloride levels were diagnostic (above 60 mmol/L) in 81.8% of examined patients. The most common identified mutation was p.Phe508del (F508del) (45.2%), followed by p.Leu346Pro (L346P) (6.7%), a mutation detected solely in individuals of Cypriot descent. The mean BMI and FEV1 z-scores were 0.2 ± 1.3 and - 2.1 ± 1.7 across all age groups, respectively, whereas chronic Pseudomonas aeruginosa colonization was noted in 26.9% of patients. The majority of patients (74.5%) were eligible to receive at least one of the available CFTR modulator therapies. In 25% of patients we recovered rare or unique genotypic profiles, including the endemic p.Leu346Pro (L346P), the rare CFTR-dup2, the co-segregated c.4200_4201delTG/c.489 + 3A > G, and the polymorphism p.Ser877Ala.

CONCLUSIONS

CF patient registries are particularly important in small or isolated populations, such as in Cyprus, with rare or unique disease cases. Their operation is necessary for the optimization of clinical care provided to CF patients, enabling their majority to benefit from evolving advances in precision medicine.

Cystic fibrosis in South Africa: spectrum of disease and determinants of outcome

Introduction

Little is known about cystic fibrosis (CF) in low- to middle-income settings. This study aimed to describe the spectrum and outcomes of CF in South Africa (SA) from the recently established SA CF registry (SACFR).

Methods

Demographic, diagnosis and clinical data were extracted from the SACFR. Cross-sectional univariable and multivariable regression analysis of best forced expiratory volume in 1 s (FEV₁; age≥6 years) and nutrition (all ages) in 2018 was conducted to investigate factors associated with severe lung disease (SLD; FEV₁ ≤3.0 z-score) and undernutrition.

Results

By December 2018, ancestry of 447 individuals included in the SACFR was Caucasian (315; 70%), mixed (87; 19%) and black African (41; 9%). Median diagnosis age was 7.6 months (IQR 2.7-37.1). Genotype was p.Phe508del homozygous (220; 49%); p.Phe508del heterozygous (144; 32%) and neither p.Phe508del or unknown Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) variant in 83 (19%); the second most frequent CFTR variant was 3120+1G>A, common in black Africans. Median age of patients in 2018 was 14.7 years (IQR 7.4-24.4). SLD was independently associated with chronic methicillin-resistant Staphylococcus aureus (MRSA) (adjusted odds ratio( aOR) 16.75; 95% CI 1.74-161.50), undernutrition (aOR 5.20; 95% CI 2.23-12.13) and age (aOR 2.23 per 10 years; 95% CI 1.50-3.31). Undernutrition was associated in univariable analysis with low weight at diagnosis, non-Caucasian ancestry, chronic P. aeruginosa infection and lower socioeconomic status.

Conclusion

Interventions targeting MRSA infection and nutrition are needed to improve CF outcomes in SA. Most people with CF in SA are eligible for highly effective CFTR modulator therapy.

Neurodevelopmental and psychosocial interventions for individuals with CHD: a research agenda and recommendations from the Cardiac Neurodevelopmental Outcome Collaborative

In 2018, the Neurodevelopmental and Psychosocial Interventions Working Group of the Cardiac Neurodevelopmental Outcome Collaborative convened through support from an R13 grant from the National Heart, Lung, and Blood Institute to survey the state of neurodevelopmental and psychosocial intervention research in CHD and to propose a slate of critical questions and investigations required to improve outcomes for this growing population of survivors and their families. Prior research, although limited, suggests that individualised developmental care interventions delivered early in life are beneficial for improving a range of outcomes including feeding, motor and cognitive development, and physiological regulation. Interventions to address self-regulatory, cognitive, and social-emotional challenges have shown promise in other medical populations, yet their applicability and effectiveness for use in individuals with CHD have not been examined. To move this field of research forward, we must strive to better understand the impact of neurodevelopmental and psychosocial intervention within the CHD population including adapting existing interventions for individuals with CHD. We must examine the ways in which dedicated cardiac neurodevelopmental follow-up programmes bolster resilience and support children and families through the myriad transitions inherent to the experience of living with CHD. And, we must ensure that interventions are person-/family-centred, inclusive of individuals from diverse cultural backgrounds as well as those with genetic/medical comorbidities, and proactive in their efforts to include individuals who are at highest risk but who may be traditionally less likely to participate in intervention trials.

Lessons learned from hemolytic uremic syndrome registries: recommendations for implementation

BACKGROUND

Hemolytic uremic syndrome (HUS) is a rare condition which diagnosed with the triad of thrombocytopenia, microangiopathic hemolytic anemia, and acute renal injury. There is a high requirement for research to discover treatments. HUS registries can be used as an important information infrastructure. In this study, we identified and compared the different features of HUS registries to present a guide for the development and implementation of HUS registries.

RESULTS

The purposes of registries were classified as clinical (9 registries), research (7 registries), and epidemiological (5 registries), and only 3 registries pursued all three types of purposes. The data set included demographic data, medical and family history, para-clinical and diagnostic measures, treatment and pharmacological data, complications, and outcomes. The assessment strategies of data quality included monthly evaluation and data audit, the participation of physicians to collect data, editing and correcting data errors, increasing the rate of data completion, following guidelines and data quality training, using specific data quality indicators, and real-time evaluation of data at the time of data entry. 8 registries include atypical HUS patients, and 7 registries include all patients regardless of age. Only two registries focused on children. 4 registries apply prospective and 4 applied both prospective, and retrospective data collection. Finally, specialized hospitals were the main data source for these registries.

CONCLUSION

Based on the findings, we suggested a learning framework for developing and implementing an HUS registry. This framework includes lessons learned and suggestions for HUS registry purposes, minimum data set, data quality assurance, data collection methods, inclusion and exclusion criteria as well as data sources. This framework can help researchers develop HUS registries.

Albertans for Health Research Network: Form, Fit, and Function

The patient engagement (PE) platform staff of Alberta Strategy for Patient Oriented Research Support for People and Patient Oriented Research Trials Unit developed a patient-powered PE network called Albertans for Health Research Network (AB4HR); an enhanced tool to better connect patient partners and researchers online. AB4HR was developed in response to an identified need—a user-friendly online forum for both patient partners and researchers to access, so that they can better work together, as partners, in health research. We codesigned AB4HR and identified ways to improve the form, fit, and function of an existing registry through discussion groups with patient partners and researchers. We found 3 main themes derived from the perspective shared by patient partners and researchers. Patient partners and researchers agreed that the existing registry provides a forum to connect with one another in an easy, low-barrier way. However, there were opportunities for improvement with AB4HR, including possibilities for greater interaction between patient partners and researchers to promote more collaborative partnerships.

Establishing Patient Registries for Rare Diseases: Rationale and Challenges

Globally, an estimated 350 million people are affected by a rare disease diagnosis. Knowledge limitations persist for the majority of rare conditions due to systemic and structural challenges in healthcare and research. Disease-specific patient populations are often small and geographically dispersed; funding support for research is restricted; and diagnostic delays are common due to disease complexities, limited medical training for practitioners, and evolving foundational knowledge related to disease characterization. Patient registries can be effective, convenient, and cost-efficient tools to support documentation of the natural history of a disease, centering patients as research partners in the process while uniting rare communities around a common initiative. Current global trends towards innovative and patient-centered healthcare are enabling patient registries to increasingly emerge as valuable tools for use within rare disease research and drug development. This article describes the value of and rationale for establishing rare disease patient registries and the considerations and challenges that stakeholders, such as researchers, industry, health care providers, and patient community organizations, may encounter.

Cystic fibrosis in Turkey: First data from the national registry

BACKGROUND

Cystic fibrosis (CF) care has been implemented in Turkey for a long time; however, there had been no patient registry. For this purpose, the Turkish National CF Registry was established. We present the first results of registry using data collected in 2017.

METHODS

The data were collected using a data-entry software system, which was accessed from the internet. Demographic and annually recorded data consisted of 15 and 79 variables, respectively.

RESULTS

There were 1170 patients registered from 23 centers; the estimated coverage rate was 30%. The median age at diagnosis was 1.7 years (median current age: 7.3 years); 51 (4.6%) patients were aged over 18 years. Among 293 patients who were under 3 years of age, 240 patients (81.9%) were diagnosed through newborn screening. Meconium ileus was detected in 65 (5.5%) patients. Genotyping was performed in 978 (87.4%) patients and 246 (25.2%) patients' mutations were unidentified. The most common mutation was deltaF508 with an allelic frequency of 28%, followed by N1303K (4.9%). The median FEV1% predicted was 86. Chronic colonization with Pseudomonas aeruginosa was seen in 245 patients. The most common complication was pseudo-Bartter syndrome in 120 patients. The median age of death was 13.5 years in a total of 15 patients.

CONCLUSIONS

Low coverage rate, lack of genotyping, unidentified mutations, and missing data of lung functions are some of our greatest challenges. Including data of all centers and reducing missing data will provide more accurate data and help to improve the CF care in Turkey in the future.

Gut microbiota signatures in cystic fibrosis: Loss of host CFTR function drives the microbiota enterophenotype

Background

Cystic fibrosis (CF) is a disorder affecting the respiratory, digestive, reproductive systems and sweat glands. This lethal hereditary disease has known or suspected links to the dysbiosis gut microbiota. High-throughput meta-omics-based approaches may assist in unveiling this complex network of symbiosis modifications.

Objectives

The aim of this study was to provide a predictive and functional model of the gut microbiota enterophenotype of pediatric patients affected by CF under clinical stability.

Methods

Thirty-one fecal samples were collected from CF patients and healthy children (HC) (age range, 1–6 years) and analysed using targeted-metagenomics and metabolomics to characterize the ecology and metabolism of CF-linked gut microbiota. The multidimensional data were low fused and processed by chemometric classification analysis.

Results

The fused metagenomics and metabolomics based gut microbiota profile was characterized by a high abundance of Propionibacterium, Staphylococcus and Clostridiaceae, including Clostridium difficile, and a low abundance of Eggerthella, Eubacterium, Ruminococcus, Dorea, Faecalibacterium prausnitzii, and Lachnospiraceae, associated with overexpression of 4-aminobutyrate (GABA), choline, ethanol, propylbutyrate, and pyridine and low levels of sarcosine, 4-methylphenol, uracil, glucose, acetate, phenol, benzaldehyde, and methylacetate. The CF gut microbiota pattern revealed an enterophenotype intrinsically linked to disease, regardless of age, and with dysbiosis uninduced by reduced pancreatic function and only partially related to oral antibiotic administration or lung colonization/infection.

Conclusions

All together, the results obtained suggest that the gut microbiota enterophenotypes of CF, together with endogenous and bacterial CF biomarkers, are direct expression of functional alterations at the intestinal level. Hence, it’s possible to infer that CFTR impairment causes the gut ecosystem imbalance.This new understanding of CF host-gut microbiota interactions may be helpful to rationalize novel clinical interventions to improve the affected children’s nutritional status and intestinal function.

The prevalence, clinical status and genotype of cystic fibrosis patients living in Cuba using national registry data

BACKGROUND

We aimed to establish a national cystic fibrosis (CF) registry for Cuba, a developing country.

METHODS

Regional centres that deliver care for all CF patients provided information for a national database.

FINDINGS

The prevalence of CF in Cuba is 26.3 cases per 1,000,000 population. The median age at diagnosis is 2 years, and the median age of the total population was 15 years. Of those aged 16 years or older, the prevalence of Pseudomonas aeruginosa infection was 46%, the prevalence of Staphylococcus aureus infection was 36%, and 80% of individuals were receiving oral azithromycin. The commonest gene mutation was F508del which was observed in 50% of patients.

INTERPRETATION

These data demonstrate that it is possible to establish a national CF registry in a developing country such as Cuba. This provides baseline data to permit evaluation of health care delivery enable the spread of good clinical practice nationally.

Evaluating the impact of 2006 Australasian Clinical Practice Guidelines for nutrition in children with cystic fibrosis in Australia

OBJECTIVES

To determine the association between the implementation of the 2006 Australasian Clinical Practice Guidelines for Nutrition in Cystic Fibrosis (CF) and the nutritional status of children participating in the Australian Cystic Fibrosis Data Registry (ACFDR).

METHODS

This research consisted of a quantitative study using ACFDR data and a survey of clinicians and dietitians treating children with CF. Two independent cohorts of children (2-5 years and 6-11 years) were selected from ACFDR between 1998 and 2014 (N = 2304). Generalised estimating equation model was used to assess weight, height and body mass index (BMI) z-scores for each patient before and after the implementation of the nutrition guidelines. A nationwide online survey was sent to 48 clinicians to explore the enablers and barriers to implementation of the guidelines.

RESULTS

Data analysis showed significant increase (p < 0.05) in mean weight, height and BMI z-scores ranging from 0.06 to 0.18 after implementation of the guidelines in both cohorts of children. Nineteen (39%) clinicians participated in the survey. The majority of the respondents adopted the recommendations into their practice and used the guidelines as part of their professional development. Structural barriers included a lack of adequate staff resources and clinic space for consultations, inappropriate staff classification, high staff turnover and lack of mentoring support.

CONCLUSION

In children participating in the ACFDR, nutritional status improved after the implementation of the 2006 guidelines. Survey results revealed enablers and barriers to guideline implementation and will inform implementation strategies for the revised Australasian nutrition guidelines for CF, released in 2017.

Prognostication and Risk Factors for Cystic Fibrosis via Automated Machine Learning

Accurate prediction of survival for cystic fibrosis (CF) patients is instrumental in establishing the optimal timing for referring patients with terminal respiratory failure for lung transplantation (LT). Current practice considers referring patients for LT evaluation once the forced expiratory volume (FEV1) drops below 30% of its predicted nominal value. While FEV1 is indeed a strong predictor of CF-related mortality, we hypothesized that the survival behavior of CF patients exhibits a lot more heterogeneity. To this end, we developed an algorithmic framework, which we call AutoPrognosis, that leverages the power of machine learning to automate the process of constructing clinical prognostic models, and used it to build a prognostic model for CF using data from a contemporary cohort that involved 99% of the CF population in the UK. AutoPrognosis uses Bayesian optimization techniques to automate the process of configuring ensembles of machine learning pipelines, which involve imputation, feature processing, classification and calibration algorithms. Because it is automated, it can be used by clinical researchers to build prognostic models without the need for in-depth knowledge of machine learning. Our experiments revealed that the accuracy of the model learned by AutoPrognosis is superior to that of existing guidelines and other competing models.