1
|
Tsai YT, Fulcher IR, Li T, Sukums F, Hedt-Gauthier B. Predicting facility-based delivery in Zanzibar: The vulnerability of machine learning algorithms to adversarial attacks. Heliyon 2023; 9:e16244. [PMID: 37234636 PMCID: PMC10205516 DOI: 10.1016/j.heliyon.2023.e16244] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/04/2022] [Revised: 05/01/2023] [Accepted: 05/10/2023] [Indexed: 05/28/2023] Open
Abstract
Background Community health worker (CHW)-led maternal health programs have contributed to increased facility-based deliveries and decreased maternal mortality in sub-Saharan Africa. The recent adoption of mobile devices in these programs provides an opportunity for real-time implementation of machine learning predictive models to identify women most at risk for home-based delivery. However, it is possible that falsified data could be entered into the model to get a specific prediction result - known as an "adversarial attack". The goal of this paper is to evaluate the algorithm's vulnerability to adversarial attacks. Methods The dataset used in this research is from the Uzazi Salama ("Safer Deliveries") program, which operated between 2016 and 2019 in Zanzibar. We used LASSO regularized logistic regression to develop the prediction model. We used "One-At-a-Time (OAT)" adversarial attacks across four different types of input variables: binary - access to electricity at home, categorical - previous delivery location, ordinal - educational level, and continuous - gestational age. We evaluated the percent of predicted classifications that change due to these adversarial attacks. Results Manipulating input variables affected prediction results. The variable with the greatest vulnerability was previous delivery location, with 55.65% of predicted classifications changing when applying adversarial attacks from previously delivered at a facility to previously delivered at home, and 37.63% of predicted classifications changing when applying adversarial attacks from previously delivered at home to previously delivered at a facility. Conclusion This paper investigates the vulnerability of an algorithm to predict facility-based delivery when facing adversarial attacks. By understanding the effect of adversarial attacks, programs can implement data monitoring strategies to assess for and deter these manipulations. Ensuring fidelity in algorithm deployment secures that CHWs target those women who are actually at high risk of delivering at home.
Collapse
Affiliation(s)
- Yi-Ting Tsai
- Department of Biostatistics, Harvard Chan School of Public Health, Boston, USA
| | - Isabel R. Fulcher
- Department of Global Health and Social Medicine, Harvard Medical School, Boston, USA
- Harvard Data Science Initiative, Harvard University, Cambridge, USA
| | - Tracey Li
- D-tree International, Zanzibar, Tanzania
| | - Felix Sukums
- Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania
| | - Bethany Hedt-Gauthier
- Department of Biostatistics, Harvard Chan School of Public Health, Boston, USA
- Department of Global Health and Social Medicine, Harvard Medical School, Boston, USA
| |
Collapse
|
2
|
Wilkes C, Bava M, Graham HR, Duke T. What are the risk factors for death among children with pneumonia in low- and middle-income countries? A systematic review. J Glob Health 2023; 13:05003. [PMID: 36825608 PMCID: PMC9951126 DOI: 10.7189/jogh.13.05003] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/25/2023] Open
Abstract
Background Knowledge of the risk factors for and causes of treatment failure and mortality in childhood pneumonia is important for prevention, diagnosis, and treatment at an individual and population level. This review aimed to identify the most important risk factors for mortality among children aged under ten years with pneumonia. Methods We systematically searched MEDLINE, EMBASE, and PubMed for observational and interventional studies reporting risk factors for mortality in children (aged two months to nine years) in low- and middle-income countries (LMICs). We screened articles according to specified inclusion and exclusion criteria, assessed risk of bias using the EPHPP framework, and extracted data on demographic, clinical, and laboratory risk factors for death. We synthesized data descriptively and using Forest plots and did not attempt meta-analysis due to the heterogeneity in study design, definitions, and populations. Findings We included 143 studies in this review. Hypoxaemia (low blood oxygen level), decreased conscious state, severe acute malnutrition, and the presence of an underlying chronic condition were the risk factors most strongly and consistently associated with increased mortality in children with pneumonia. Additional important clinical factors that were associated with mortality in the majority of studies included particular clinical signs (cyanosis, pallor, tachypnoea, chest indrawing, convulsions, diarrhoea), chronic comorbidities (anaemia, HIV infection, congenital heart disease, heart failure), as well as other non-severe forms of malnutrition. Important demographic factors associated with mortality in the majority of studies included age <12 months and inadequate immunisation. Important laboratory and investigation findings associated with mortality in the majority of studies included: confirmed Pneumocystis jirovecii pneumonia (PJP), consolidation on chest x-ray, pleural effusion on chest x-ray, and leukopenia. Several other demographic, clinical and laboratory findings were associated with mortality less consistently or in a small numbers of studies. Conclusions Risk assessment for children with pneumonia should include routine evaluation for hypoxaemia (pulse oximetry), decreased conscious state (e.g. AVPU), malnutrition (severe, moderate, and stunting), and the presence of an underlying chronic condition as these are strongly and consistently associated with increased mortality. Other potentially useful risk factors include the presence of pallor or anaemia, chest indrawing, young age (<12 months), inadequate immunisation, and leukopenia.
Collapse
Affiliation(s)
- Chris Wilkes
- Murdoch Children’s Research Institution, Royal Children’s Hospital, Parkville, Victoria, Australia
| | - Mohamed Bava
- Murdoch Children’s Research Institution, Royal Children’s Hospital, Parkville, Victoria, Australia
| | - Hamish R Graham
- Murdoch Children’s Research Institution, Royal Children’s Hospital, Parkville, Victoria, Australia,Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Parkville, Victoria, Australia
| | - Trevor Duke
- Murdoch Children’s Research Institution, Royal Children’s Hospital, Parkville, Victoria, Australia,Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Parkville, Victoria, Australia
| | | |
Collapse
|
3
|
Li X, Zhang X, Lin X, Cai L, Wang Y, Chang Z. Classification and Prognosis Analysis of Pancreatic Cancer Based on DNA Methylation Profile and Clinical Information. Genes (Basel) 2022; 13:genes13101913. [PMID: 36292798 PMCID: PMC9601656 DOI: 10.3390/genes13101913] [Citation(s) in RCA: 5] [Impact Index Per Article: 2.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/02/2022] [Revised: 10/15/2022] [Accepted: 10/18/2022] [Indexed: 11/04/2022] Open
Abstract
Pancreatic adenocarcinoma (PAAD) has a poor prognosis with high individual variation in the treatment response among patients; however, there is no standard molecular typing method for PAAD prognosis in clinical practice. We analyzed DNA methylation data from The Cancer Genome Atlas database, which identified 1235 differentially methylated DNA genes between PAAD and adjacent tissue samples. Among these, 78 methylation markers independently affecting PAAD prognosis were identified after adjusting for significant clinical factors. Based on these genes, two subtypes of PAAD were identified through consistent clustering. Fourteen specifically methylated genes were further identified to be associated with survival. Further analyses of the transcriptome data identified 301 differentially expressed cancer driver genes between the two PAAD subtypes and the degree of immune cell infiltration differed significantly between the subtypes. The 14 specific genes characterizing the unique methylation patterns of the subtypes were used to construct a Bayesian network-based prognostic prediction model for typing that showed good predictive value (area under the curve value of 0.937). This study provides new insight into the heterogeneity of pancreatic tumors from an epigenetic perspective, offering new strategies and targets for personalized treatment plan evaluation and precision medicine for patients with PAAD.
Collapse
Affiliation(s)
- Xin Li
- Harbin Medical University Cancer Hospital, Harbin 150081, China
| | - Xuan Zhang
- Harbin Medical University Cancer Hospital, Harbin 150081, China
| | - Xiangyu Lin
- Harbin Institute of Technology, School of Life Science and Technology, Harbin 150001, China
| | - Liting Cai
- The First Affiliated Hospital of Baotou Medical College Cancer Center, Baotou 014016, China
| | - Yan Wang
- Harbin Medical University Cancer Hospital, Harbin 150081, China
- Correspondence: (Y.W.); (Z.C.)
| | - Zhiqiang Chang
- College of Bioinformatics Science and Technology, Harbin Medical University, Harbin 150081, China
- Correspondence: (Y.W.); (Z.C.)
| |
Collapse
|
4
|
Fredriksson A, Fulcher IR, Russell AL, Li T, Tsai YT, Seif SS, Mpembeni RN, Hedt-Gauthier B. Machine learning for maternal health: Predicting delivery location in a community health worker program in Zanzibar. Front Digit Health 2022; 4:855236. [PMID: 36060544 PMCID: PMC9428344 DOI: 10.3389/fdgth.2022.855236] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/14/2022] [Accepted: 07/25/2022] [Indexed: 11/24/2022] Open
Abstract
Background Maternal and neonatal health outcomes in low- and middle-income countries (LMICs) have improved over the last two decades. However, many pregnant women still deliver at home, which increases the health risks for both the mother and the child. Community health worker programs have been broadly employed in LMICs to connect women to antenatal care and delivery locations. More recently, employment of digital tools in maternal health programs have resulted in better care delivery and served as a routine mode of data collection. Despite the availability of rich, patient-level data within these digital tools, there has been limited utilization of this type of data to inform program delivery in LMICs. Methods We use program data from 38,787 women enrolled in Safer Deliveries, a community health worker program in Zanzibar, to build a generalizable prediction model that accurately predicts whether a newly enrolled pregnant woman will deliver in a health facility. We use information collected during the enrollment visit, including demographic data, health characteristics and current pregnancy information. We apply four machine learning methods: logistic regression, LASSO regularized logistic regression, random forest and an artificial neural network; and three sampling techniques to address the imbalanced data: undersampling of facility deliveries, oversampling of home deliveries and addition of synthetic home deliveries using SMOTE. Results Our models correctly predicted the delivery location for 68%–77% of the women in the test set, with slightly higher accuracy when predicting facility delivery versus home delivery. A random forest model with a balanced training set created using undersampling of existing facility deliveries accurately identified 74.4% of women delivering at home. Conclusions This model can provide a “real-time” prediction of the delivery location for new maternal health program enrollees and may enable early provision of extra support for individuals at risk of not delivering in a health facility, which has potential to improve health outcomes for both mothers and their newborns. The framework presented here is applicable in other contexts and the selection of input features can easily be adapted to match data availability and other outcomes, both within and beyond maternal health.
Collapse
Affiliation(s)
- Alma Fredriksson
- Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, MA, United States
- Correspondence: Alma Fredriksson
| | - Isabel R. Fulcher
- Department of Global Health and Social Medicine, Harvard Medical School, Boston, MA, United States
- Harvard Data Science Initiative, Cambridge, MA, United States
| | | | - Tracey Li
- D-tree International, Dar es Salaam, Tanzania
| | - Yi-Ting Tsai
- Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, MA, United States
| | | | - Rose N. Mpembeni
- Department of Epidemiology and Biostatistics, Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania
| | - Bethany Hedt-Gauthier
- Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, MA, United States
- Department of Global Health and Social Medicine, Harvard Medical School, Boston, MA, United States
| |
Collapse
|
5
|
Haneef R, Tijhuis M, Thiébaut R, Májek O, Pristaš I, Tolenan H, Gallay A. Methodological guidelines to estimate population-based health indicators using linked data and/or machine learning techniques. Arch Public Health 2022; 80:9. [PMID: 34983651 PMCID: PMC8725299 DOI: 10.1186/s13690-021-00770-6] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/27/2021] [Accepted: 12/17/2021] [Indexed: 12/23/2022] Open
Abstract
BACKGROUND The capacity to use data linkage and artificial intelligence to estimate and predict health indicators varies across European countries. However, the estimation of health indicators from linked administrative data is challenging due to several reasons such as variability in data sources and data collection methods resulting in reduced interoperability at various levels and timeliness, availability of a large number of variables, lack of skills and capacity to link and analyze big data. The main objective of this study is to develop the methodological guidelines calculating population-based health indicators to guide European countries using linked data and/or machine learning (ML) techniques with new methods. METHOD We have performed the following step-wise approach systematically to develop the methodological guidelines: i. Scientific literature review, ii. Identification of inspiring examples from European countries, and iii. Developing the checklist of guidelines contents. RESULTS We have developed the methodological guidelines, which provide a systematic approach for studies using linked data and/or ML-techniques to produce population-based health indicators. These guidelines include a detailed checklist of the following items: rationale and objective of the study (i.e., research question), study design, linked data sources, study population/sample size, study outcomes, data preparation, data analysis (i.e., statistical techniques, sensitivity analysis and potential issues during data analysis) and study limitations. CONCLUSIONS This is the first study to develop the methodological guidelines for studies focused on population health using linked data and/or machine learning techniques. These guidelines would support researchers to adopt and develop a systematic approach for high-quality research methods. There is a need for high-quality research methodologies using more linked data and ML-techniques to develop a structured cross-disciplinary approach for improving the population health information and thereby the population health.
Collapse
Affiliation(s)
- Romana Haneef
- Department of Non-Communicable Diseases and Injuries, Santé Publique France, Saint-Maurice, France.
| | - Mariken Tijhuis
- National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands
| | - Rodolphe Thiébaut
- Bordeaux University, Bordeaux School of Public Health, Bordeaux, France.,INSERM / INRIA SISTM team, Bordeaux Population health, Bordeaux, France.,Medical Information Department, Bordeaux University Hospital, Bordeaux, France
| | - Ondřej Májek
- Institute of Health Information and Statistics of the Czech Republic, Prague, Czech Republic.,Institute of Biostatistics and Analyses, Faculty of Medicine, Masaryk University, Brno, Czech Republic
| | - Ivan Pristaš
- National Institute of public health, division of health informatics and biostatistics, Zagreb, Croatia
| | - Hanna Tolenan
- Finnish Institute for Health and Welfare (THL), Helsinki, Finland
| | - Anne Gallay
- Department of Non-Communicable Diseases and Injuries, Santé Publique France, Saint-Maurice, France
| |
Collapse
|
6
|
Heightened Local T h17 Cell Inflammation Is Associated with Severe Community-Acquired Pneumonia in Children under the Age of 1 Year. Mediators Inflamm 2021; 2021:9955168. [PMID: 34602860 PMCID: PMC8482031 DOI: 10.1155/2021/9955168] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/08/2021] [Revised: 07/29/2021] [Accepted: 08/05/2021] [Indexed: 12/31/2022] Open
Abstract
Severe community-acquired pneumonia (sCAP) early in life is a leading cause of morbidity, mortality, and irreversible sequelae. Herein, we report the clinical, etiological, and immunological characteristics of 62 children age < 1 year. We measured 27 cytokines in plasma and bronchoalveolar lavage (BAL) from 62 children age < 1 year who were diagnosed with CAP, and then, we analyzed correlations among disease severity, clinical parameters, and etiology. Of the entire cohort, three cytokines associated with interleukin-17- (IL-17-) producing helper T cells (Th17 cells), IL-1β, IL-6, and IL-17, were significantly elevated in sCAP patients with high fold changes (FCs); in BAL, these cytokines were intercorrelated and associated with blood neutrophil counts, Hb levels, and mixed bacterial-viral infections. BAL IL-1β (area under the curve (AUC) 0.820), BAL IL-17 (AUC 0.779), and plasma IL-6 (AUC 0.778) had remarkable predictive power for sCAP. Our findings revealed that increased local Th17 cell immunity played a critical role in the development of sCAP in children age < 1 year. Th17 cell-related cytokines could serve as local and systemic inflammatory indicators of sCAP in this age group.
Collapse
|
7
|
Lenahan JL, Nkwopara E, Phiri M, Mvalo T, Couasnon MT, Turner K, Ndamala C, McCollum ED, May S, Ginsburg AS. Repeat assessment of examination signs among children in Malawi with fast-breathing pneumonia. ERJ Open Res 2020; 6:00275-2019. [PMID: 32494572 PMCID: PMC7248340 DOI: 10.1183/23120541.00275-2019] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/03/2019] [Accepted: 03/04/2020] [Indexed: 01/04/2023] Open
Abstract
Background As part of a randomised controlled trial of treatment with placebo versus 3 days of amoxicillin for nonsevere fast-breathing pneumonia among Malawian children aged 2–59 months, a subset of children was hospitalised for observation. We sought to characterise the progression of fast-breathing pneumonia among children undergoing repeat assessments to better understand which children do and do not deteriorate. Methods Vital signs and physical examination findings, including respiratory rate, arterial oxygen saturation measured by pulse oximetry (SpO2), chest indrawing and temperature were assessed every 3 h for the duration of hospitalisation. Children were assessed for treatment failure during study visits on days 1, 2, 3 and 4. Results Hospital monitoring data from 436 children were included. While no children had SpO2 90–93% at baseline, 7.4% (16 of 215) of children receiving amoxicillin and 9.5% (21 of 221) receiving placebo developed SpO2 90–93% during monitoring. Similarly, no children had chest indrawing at enrolment, but 6.6% (14 of 215) in the amoxicillin group and 7.2% (16 of 221) in the placebo group went on to develop chest indrawing during hospitalisation. Conclusion Repeat monitoring of children with fast-breathing pneumonia identified vital and physical examination signs not present at baseline, including SpO2 90–93% and chest indrawing. This information may support providers and policymakers in developing guidance for care of children with nonsevere pneumonia. This study characterised the progression of fast-breathing pneumonia among children in Malawi. Repeat monitoring of children identified vital and physical exam signs not present at baseline, including oxygen saturation of 90–93% and chest indrawing.http://bit.ly/2vUlckS
Collapse
Affiliation(s)
- Jennifer L Lenahan
- International Programs, Save the Children Federation Inc., Westport, CT, USA
| | - Evangelyn Nkwopara
- International Programs, Save the Children Federation Inc., Westport, CT, USA
| | - Melda Phiri
- Dept of Pediatrics, University of North Carolina Project, Lilongwe Medical Relief Fund Trust, Lilongwe, Malawi
| | - Tisungane Mvalo
- Dept of Pediatrics, University of North Carolina Project, Lilongwe Medical Relief Fund Trust, Lilongwe, Malawi
| | - Mari T Couasnon
- International Programs, Save the Children Federation Inc., Westport, CT, USA
| | - Kali Turner
- International Programs, Save the Children Federation Inc., Westport, CT, USA
| | - Chifundo Ndamala
- Dept of Pediatrics, University of North Carolina Project, Lilongwe Medical Relief Fund Trust, Lilongwe, Malawi
| | - Eric D McCollum
- Dept of Pediatrics, Eudowood Division of Pediatric Respiratory Sciences, Johns Hopkins School of Medicine, Baltimore, MD, USA
| | - Susanne May
- Dept of Biostatistics, University of Washington, Seattle, WA, USA
| | - Amy Sarah Ginsburg
- International Programs, Save the Children Federation Inc., Westport, CT, USA
| |
Collapse
|
8
|
Aurilio RB, Sant'Anna CC, March MDFBP. CLINICAL PROFILE OF CHILDREN WITH AND WITHOUT COMORBIDITIES HOSPITALIZED WITH COMMUNITY-ACQUIRED PNEUMONIA. REVISTA PAULISTA DE PEDIATRIA 2020; 38:e2018333. [PMID: 32401948 PMCID: PMC7212558 DOI: 10.1590/1984-0462/2020/38/2018333] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Subscribe] [Scholar Register] [Received: 10/30/2018] [Accepted: 04/20/2019] [Indexed: 11/22/2022]
Abstract
Objective: To describe the clinical profile of children and adolescents hospitalized
with community-acquired pneumonia (CAP). They were divided into two groups:
those with and those without comorbidities. Methods: An observational, cross-sectional, descriptive study with prospective data
collection, was carried out in a cohort of patients aged zero to 11, who
were hospitalized with a clinical and radiological diagnosis of
community-acquired pneumonia, from January 2010 to January 2012. As an
exploratory study, the two groups were compared through logistic regression
for possible risk factors associated with community-acquired pneumonia.
Relative risk (RR) was used with a 95% confidence interval (95%CI). The
process of selection for independent variables was stepwise forward, with a
significance level of 5%. Results: There were 121 cases of community-acquired pneumonia evaluated, and 47.9%
had comorbidities. In the bivariate analysis, patients with comorbidities
demonstrated higher chances for: age >60 months (p=0.005), malnutrition
(p=0.002), previous use of antibiotics (p=0.008) and previous
hospitalization for community-acquired pneumonia in the last 24 months
(p=0.004). In the multivariate analysis, these variables were independent
predictors of community-acquired pneumonia in patients with the
comorbidities: age >60 months (p=0.002; RR=5.39; 95%CI 1.89-15.40);
malnutrition (p=0.008; RR=1.75; 95%CI 1.75-44.60); previous use of
antibiotics (p=0.0013; RR=3.03; 95%CI 1.27-7.20); and previous
hospitalization for community-acquired pneumonia (p=0.035; RR=2.91; 95%CI
1.08-7.90). Conclusions: Most patients with community-acquired pneumonia and comorbidities were aged
>60 months, were malnourished, had used antibiotics and had been
hospitalized for community-acquired pneumonia. Comorbidities were associated
with a higher chance of malnutrition and hospitalizations for
community-acquired pneumonia in an older age group, compared to children
without comorbidities. Knowledge of this clinical profile may contribute to
better assist pediatric patients with community-acquired pneumonia
hospitalized in referral centers.
Collapse
|
9
|
Gachau S, Quartagno M, Njagi EN, Owuor N, English M, Ayieko P. Handling missing data in modelling quality of clinician-prescribed routine care: Sensitivity analysis of departure from missing at random assumption. Stat Methods Med Res 2020; 29:3076-3092. [PMID: 32390503 DOI: 10.1177/0962280220918279] [Citation(s) in RCA: 6] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/20/2023]
Abstract
Missing information is a major drawback in analyzing data collected in many routine health care settings. Multiple imputation assuming a missing at random mechanism is a popular method to handle missing data. The missing at random assumption cannot be confirmed from the observed data alone, hence the need for sensitivity analysis to assess robustness of inference. However, sensitivity analysis is rarely conducted and reported in practice. We analyzed routine paediatric data collected during a cluster randomized trial conducted in Kenyan hospitals. We imputed missing patient and clinician-level variables assuming the missing at random mechanism. We also imputed missing clinician-level variables assuming a missing not at random mechanism. We incorporated opinions from 15 clinical experts in the form of prior distributions and shift parameters in the delta adjustment method. An interaction between trial intervention arm and follow-up time, hospital, clinician and patient-level factors were included in a proportional odds random-effects analysis model. We performed these analyses using R functions derived from the jomo package. Parameter estimates from multiple imputation under the missing at random mechanism were similar to multiple imputation estimates assuming the missing not at random mechanism. Our inferences were insensitive to departures from the missing at random assumption using either the prior distributions or shift parameters sensitivity analysis approach.
Collapse
Affiliation(s)
- Susan Gachau
- Health Services Unit, Kenya Medical Research Institute-Wellcome Trust Research Programme, Nairobi, Kenya
- School of Mathematics, University of Nairobi, Nairobi, Kenya
| | - Matteo Quartagno
- Institute of Clinical Trials and Methodology, University College London, London, UK
| | - Edmund Njeru Njagi
- Department of Non-Communicable Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK
| | - Nelson Owuor
- School of Mathematics, University of Nairobi, Nairobi, Kenya
| | - Mike English
- Health Services Unit, Kenya Medical Research Institute-Wellcome Trust Research Programme, Nairobi, Kenya
- Nuffield Department of Medicine, University of Oxford, Oxford, UK
| | - Philip Ayieko
- Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK
- Mwanza Intervention Trials Unit, Mwanza, Tanzania
| |
Collapse
|
10
|
Brown N, Rizvi A, Kerai S, Nisar MI, Rahman N, Baloch B, Jehan F. Recurrence of WHO-defined fast breathing pneumonia among infants, its occurrence and predictors in Pakistan: a nested case-control analysis. BMJ Open 2020; 10:e035277. [PMID: 31915178 PMCID: PMC6955570 DOI: 10.1136/bmjopen-2019-035277] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 02/03/2023] Open
Abstract
OBJECTIVES Studies in low-income and middle-income countries have shown an adverse association between environmental exposures including poverty. There is little literature from South Asia. We aimed to test the associations between housing, indoor air pollution and children's respiratory health and recurrent fast breathing pneumonia in a poor urban setting in Pakistan. SETTING Primary health centres in a periurban slum in Karachi, Pakistan. METHODS Nested matched case-control study within a non-inferiority randomised controlled trial of fast breathing pneumonia (Randomised Trial of Amoxicillin vs Placebo for Pneumonia (RETAPP)) in periurban slums of Karachi, Pakistan. Cases were children aged 2-60 months enrolled in RETAPP with fast breathing pneumonia who presented again with fast breathing between 8 weeks and 12 months after full recovery. Controls, selected in a 2:1 ratio, were age-matched participants who did not represent. Multivariable conditional logistic regression analysis was undertaken to explore associations with potentially modifiable environmental predictors including housing type, indoor air quality, exposure to tobacco smoke, outdoor pollution, household crowding, water and sanitation quality, nutritional status, immunisation completeness, breast feeding and airways hyperactivity. RESULTS Fast breathing recurred in 151 (3.7%) of children out of the total (4003) enrolled in the trial. Poor-quality housing of either katcha or mixed type strongly predicted recurrence with adjusted matched ORs 2.43 (95% CI 1.02 to 5.80) and 2.44 (1.11 to 5.38), respectively. Poor air quality, cooking fuel, inadequate ventilation, nutritional status, water, sanitation and hygiene (WASH) index, wheeze at first presentation and group of initial trial assignment were not independently predictive of recurrence. CONCLUSION Poor-quality housing independently predicted recurrence of fast breathing pneumonia. TRIAL REGISTRATION NUMBER NCT02372461.
Collapse
Affiliation(s)
- Nick Brown
- International Maternal and Child Health, Department of Women's and Children's Health, Uppsala University, Academiska Sjukhuset, Uppsala, 75185, Sweden
- Department of Child Health, Aga Khan University Hospital, National Stadium Rd, Karachi, Sindh, 74800, Pakistan
| | - Arjumand Rizvi
- Department of Child Health, Aga Khan University Hospital, National Stadium Rd, Karachi, Sindh, 74800, Pakistan
| | - Salima Kerai
- University of British Columbia, Vancouver, British Columbia, Canada
| | - Muhammad Imran Nisar
- Department of Child Health, Aga Khan University Hospital, National Stadium Rd, Karachi, Sindh, 74800, Pakistan
| | - Najeeb Rahman
- Department of Child Health, Aga Khan University Hospital, National Stadium Rd, Karachi, Sindh, 74800, Pakistan
| | - Benazir Baloch
- Department of Child Health, Aga Khan University Hospital, National Stadium Rd, Karachi, Sindh, 74800, Pakistan
| | - Fyezah Jehan
- Department of Child Health, Aga Khan University Hospital, National Stadium Rd, Karachi, Sindh, 74800, Pakistan
| |
Collapse
|
11
|
Ginsburg AS, May S, Nkwopara E, Ambler G, McCollum ED, Mvalo T, Phiri A, Lufesi N. Clinical Outcomes of Pneumonia and Other Comorbidities in Children Aged 2-59 Months in Lilongwe, Malawi: Protocol for the Prospective Observational Study "Innovative Treatments in Pneumonia". JMIR Res Protoc 2019; 8:e13377. [PMID: 31359870 PMCID: PMC6690162 DOI: 10.2196/13377] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/11/2019] [Revised: 04/12/2019] [Accepted: 04/30/2019] [Indexed: 01/29/2023] Open
Abstract
Background Pneumonia is the leading infectious cause of death worldwide among children below 5 years of age. Clinical trials are conducted to determine optimal treatment; however, these trials often exclude children with comorbidities and severe illness. Conclusions Given the paucity of data from Africa, African-based research is necessary to establish optimal management of childhood pneumonia in malaria-endemic settings in the region. An expanded evidence base that includes children with pneumonia and other comorbidities, who are at high risk for mortality or have other complications and are therefore typically excluded from childhood pneumonia clinical trials, can contribute to future iterations of the World Health Organization Integrated Management of Childhood Illness guidelines. Methods The study enrolled 1000 children with pneumonia presenting to the outpatient departments of Kamuzu Central or Bwaila District Hospitals in Lilongwe, Malawi, who were excluded from concurrent randomized controlled clinical trials investigating fast breathing and chest indrawing pneumonia and who met the inclusion criteria for this prospective observational study. Each child received standard care for their illnesses per Malawian guidelines and hospital protocol and was prospectively followed up with scheduled study visits on days 1, 2 (if hospitalized), 6, 14 (in person), and 30 (by phone). Our primary objectives are to describe the clinical outcomes of children who meet the inclusion criteria for this study and to investigate whether the percentages of children cured at day 14 among those with either fast breathing or chest indrawing pneumonia and comorbidities such as severe malaria, anemia, severe acute malnutrition, or HIV are lower than those in children without these comorbidities in the standard care groups in concurrent clinical trials. This study was approved by the Western Institutional Review Board, Malawi College of Medicine Research and Ethics Committee, and the Malawi Pharmacy, Medicines and Poisons Board. Objective This prospective observational study aimed to assess the clinical outcomes of children aged 2-59 months with both pneumonia and other comorbidities in a malaria-endemic region of Malawi. Results The Innovative Treatments in Pneumonia project was funded by the Bill and Melinda Gates Foundation (OPP1105080) in April 2014. Enrollment in this study began in 2016, and the primary results are expected in 2019. International Registered Report Identifier (IRRID) DERR1-10.2196/13377
Collapse
Affiliation(s)
| | - Susanne May
- Department of Biostatistics, University of Washington, Seattle, WA, United States
| | | | | | - Eric D McCollum
- Eudowood Division of Pediatric Respiratory Sciences, John Hopkins School of Medicine, Baltimore, MD, United States
| | - Tisungane Mvalo
- University of North Carolina Project: Lilongwe, Central Region, Lilongwe, Malawi
| | - Ajib Phiri
- Department of Paediatrics and Child Health, College of Medicine, University of Malawi, Blantyre, Malawi
| | - Norman Lufesi
- Ministry of Health, Republic of Malawi, Lilongwe, Malawi
| |
Collapse
|
12
|
Ardura-Garcia C, Kuehni CE. Reducing childhood respiratory morbidity and mortality in low and middle income countries: a current challenge. Eur Respir J 2019; 54:54/1/1900987. [PMID: 31296784 DOI: 10.1183/13993003.00987-2019] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/16/2019] [Accepted: 05/16/2019] [Indexed: 01/01/2023]
Affiliation(s)
| | - Claudia E Kuehni
- Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland .,Paediatric Respiratory Medicine, Children's University Hospital of Bern, University of Bern, Bern, Switzerland
| |
Collapse
|
13
|
Dunstan J, Aguirre M, Bastías M, Nau C, Glass TA, Tobar F. Predicting nationwide obesity from food sales using machine learning. Health Informatics J 2019; 26:652-663. [PMID: 31106648 DOI: 10.1177/1460458219845959] [Citation(s) in RCA: 14] [Impact Index Per Article: 2.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/16/2022]
Abstract
The obesity epidemic progresses everywhere across the globe, and implementing frequent nationwide surveys to measure the percentage of obese population is costly. Conversely, country-level food sales information can be accessed inexpensively through different suppliers on a regular basis. This study applies a methodology to predict obesity prevalence at the country-level based on national sales of a small subset of food and beverage categories. Three machine learning algorithms for nonlinear regression were implemented using purchase and obesity prevalence data from 79 countries: support vector machines, random forests and extreme gradient boosting. The proposed method was validated in terms of both the absolute prediction error and the proportion of countries for which the obesity prevalence was predicted satisfactorily. We found that the most-relevant food category to predict obesity is baked goods and flours, followed by cheese and carbonated drinks.
Collapse
|
14
|
Agweyu A, Lilford RJ, English M. Appropriateness of clinical severity classification of new WHO childhood pneumonia guidance: a multi-hospital, retrospective, cohort study. Lancet Glob Health 2018; 6:e74-e83. [PMID: 29241618 PMCID: PMC5732316 DOI: 10.1016/s2214-109x(17)30448-5] [Citation(s) in RCA: 44] [Impact Index Per Article: 7.3] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/22/2017] [Revised: 09/23/2017] [Accepted: 11/02/2017] [Indexed: 12/13/2022]
Abstract
BACKGROUND Management of pneumonia in many low-income and middle-income countries is based on WHO guidelines that classify children according to clinical signs that define thresholds of risk. We aimed to establish whether some children categorised as eligible for outpatient treatment might have a risk of death warranting their treatment in hospital. METHODS We did a retrospective cohort study of children aged 2-59 months admitted to one of 14 hospitals in Kenya with pneumonia between March 1, 2014, and Feb 29, 2016, before revised WHO pneumonia guidelines were adopted in the country. We modelled associations with inpatient mortality using logistic regression and calculated absolute risks of mortality for presenting clinical features among children who would, as part of revised WHO pneumonia guidelines, be eligible for outpatient treatment (non-severe pneumonia). FINDINGS We assessed 16 162 children who were admitted to hospital in this period. 832 (5%) of 16 031 children died. Among groups defined according to new WHO guidelines, 321 (3%) of 11 788 patients with non-severe pneumonia died compared with 488 (14%) of 3434 patients with severe pneumonia. Three characteristics were strongly associated with death of children retrospectively classified as having non-severe pneumonia: severe pallor (adjusted risk ratio 5·9, 95% CI 5·1-6·8), mild to moderate pallor (3·4, 3·0-3·8), and weight-for-age Z score (WAZ) less than -3 SD (3·8, 3·4-4·3). Additional factors that were independently associated with death were: WAZ less than -2 to -3 SD, age younger than 12 months, lower chest wall indrawing, respiratory rate of 70 breaths per min or more, female sex, admission to hospital in a malaria endemic region, moderate dehydration, and an axillary temperature of 39°C or more. INTERPRETATION In settings of high mortality, WAZ less than -3 SD or any degree of pallor among children with non-severe pneumonia was associated with a clinically important risk of death. Our data suggest that admission to hospital should not be denied to children with these signs and we urge clinicians to consider these risk factors in addition to WHO criteria in their decision making. FUNDING Wellcome Trust.
Collapse
Affiliation(s)
- Ambrose Agweyu
- Health Services Unit, KEMRI-Wellcome Trust Research Programme, Nairobi, Kenya; Nuffield Department of Medicine, University of Oxford, Oxford, UK.
| | - Richard J Lilford
- Department of Health Sciences, Warwick Medical School, University of Warwick, Coventry, UK
| | - Mike English
- Health Services Unit, KEMRI-Wellcome Trust Research Programme, Nairobi, Kenya; Nuffield Department of Medicine, University of Oxford, Oxford, UK
| |
Collapse
|
15
|
Tuti T, Agweyu A, Mwaniki P, Peek N, English M. Correction to: An exploration of mortality risk factors in non-severe pneumonia in children using clinical data from Kenya. BMC Med 2017; 15:212. [PMID: 29207988 PMCID: PMC5715507 DOI: 10.1186/s12916-017-0980-8] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 11/28/2017] [Accepted: 11/28/2017] [Indexed: 11/14/2022] Open
Abstract
The original article contains an omission in the Acknowledgements sub-section of the Declarations.
Collapse
Affiliation(s)
- Timothy Tuti
- KEMRI - Wellcome Trust Research Programme, Nairobi, Kenya.
| | - Ambrose Agweyu
- KEMRI - Wellcome Trust Research Programme, Nairobi, Kenya
| | - Paul Mwaniki
- KEMRI - Wellcome Trust Research Programme, Nairobi, Kenya
| | - Niels Peek
- Centre for Health Informatics, Division of Informatics, Imaging & Data Sciences, Faculty of Biology, Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK.,NIHR Greater Manchester Primary Care Patient Safety Translational Research Centre, Manchester, UK
| | - Mike English
- KEMRI - Wellcome Trust Research Programme, Nairobi, Kenya.,Nuffield Department of Medicine, Oxford University, Oxford, UK
| | | |
Collapse
|