Kaatje Bollaerts

Simple and multiple P‐splines regression with shape constraints

In many research areas, especially within social and behavioural sciences, the relationship between predictor and criterion variables is often assumed to have a particular shape, such as monotone, single-peaked or U-shaped. Such assumptions can be transformed into (local or global) constraints on the sign of the nth-order derivative of the functional form. To check for such assumptions, we present a non-parametric regression method, P-splines regression, with additional asymmetric discrete penalties enforcing the constraints. We show that the corresponding loss function is convex and present a Newton-Raphson algorithm to optimize. Constrained P-splines are illustrated with an application on monotonicity-constrained regression with both one and two predictor variables, using data from research on the cognitive development of children.

A systematic review and meta-analysis on the safety of newly adjuvanted vaccines among children

INTRODUCTION New adjuvants such as the AS- or the MF59-adjuvants improve vaccine efficacy and facilitate dose-sparing. Their use in influenza and malaria vaccines has resulted in a large body of evidence on their clinical safety in children. METHODS We carried out a systematic search for safety data from published clinical trials on newly adjuvanted vaccines in children ≤10 years of age. Serious adverse events (SAEs), solicited AEs, unsolicited AEs and AEs of special interest were evaluated for four new adjuvants: the immuno-stimulants containing adjuvant systems AS01 and AS02, and the squalene containing oil-in-water emulsions AS03 and MF59. Relative risks (RR) were calculated, comparing children receiving newly adjuvanted vaccines to children receiving other vaccines with a variety of antigens, both adjuvanted and unadjuvanted. RESULTS Twenty-nine trials were included in the meta-analysis, encompassing 25,056 children who received at least one dose of the newly adjuvanted vaccines. SAEs did not occur more frequently in adjuvanted groups (RR 0.85, 95%CI 0.75-0.96). Our meta-analyses showed higher reactogenicity following administration of newly adjuvanted vaccines, however, no consistent pattern of solicited AEs was observed across adjuvant systems. Pain was the most prevalent AE, but often mild and of short duration. No increased risks were found for unsolicited AEs, febrile convulsions, potential immune mediated diseases and new onset of chronic diseases. CONCLUSIONS Our meta-analysis did not show any safety concerns in clinical trials of the newly adjuvanted vaccines in children ≤10 years of age. An unexplained increase of meningitis in one Phase III AS01-adjuvanted malaria trial and the link between narcolepsy and the AS03-adjuvanted pandemic vaccine illustrate that continued safety monitoring is warranted.

Pandemrix™ and narcolepsy: A critical appraisal of the observational studies

A link between Pandemrix™ (AS03-adjuvanted H1N1 pandemic influenza vaccine, GSK Vaccines, Belgium) and narcolepsy was first suspected in 2010 in Sweden and Finland following a number of reports in children and adolescents. Initial scepticism about the reported association faded as additional countries reported similar findings, leading several regulatory authorities to restrict the use of Pandemrix™. The authors acknowledge that currently available data suggest an increased risk of narcolepsy following vaccination with Pandemrix™; however, from an epidemiologists perspective, significant methodological limitations of the studies have not been fully addressed and raise questions about the reported risk estimates. We review the most important biases and confounders that potentially occurred in 12 European studies of the observed association between Pandemrix™ and narcolepsy, and call for further analyses and debate.

Estimating the population prevalence and force of infection directly from antibody titres

The use of threshold values in order to diagnose individual subjects as being ‘susceptible’ or ‘infected or recovered/immune’ for a specific infection is virtually always prone to false positive, false negative or inconclusive classifications. Such misclassifications might lead to biased estimates for epidemiological parameters, such as the prevalence and the force of infection. In this article, we propose to estimate these epidemiological parameters directly from antibody titres, using an underlying mixture model. The method is applied to estimate the Salmonella serological prevalence in pigs and the age-dependent force of infection using serological data on the Varicella-Zoster virus (VZV) in humans. The threshold and direct method are compared through a simulation study.

Effectiveness of pneumococcal vaccines in preventing pneumonia in adults, a systematic review and meta-analyses of observational studies

Introduction S. pneumoniae can cause a wide spectrum of diseases, including invasive pneumococcal disease and pneumonia. Two types of pneumococcal vaccines are indicated for use in adults: 23-valent pneumococcal polysaccharide vaccines (PPV23) and a 13-valent pneumococcal conjugate vaccine (PCV13). Objective To systematically review the literature assessing pneumococcal vaccine effectiveness (VE) against community-acquired pneumonia (CAP) in adults among the general population, the immunocompromised and subjects with underlying risk factors in real-world settings. Methods We searched for peer-reviewed observational studies published between 1980 and 2015 in Pubmed, SciELO or LILACS, with pneumococcal VE estimates against CAP, pneumococcal CAP or nonbacteremic pneumococcal CAP. Meta-analyses and meta-regression for VE against CAP requiring hospitalization in the general population was performed. Results 1159 unique articles were retrieved of which 33 were included. No studies evaluating PCV13 effectiveness were found. Wide ranges in PPV23 effectiveness estimates for any-CAP were observed among adults ≥65 years (-143% to 60%). The meta-analyzed VE estimate for any-CAP requiring hospitalization in the general population was 10.2% (95%CI: -12.6; 33.0). The meta-regression indicates that VE against any-CAP requiring hospitalization is significantly lower in studies with a maximum time since vaccination ≥60 months vs. <60 months and in countries with the pediatric PCV vaccine available on the private market. However, these results should be interpreted cautiously due to the high influence of two studies. The VE estimates for pneumococcal CAP hospitalization ranged from 32% (95%CI: -18; 61) to 51% (95%CI: 16; 71) in the general population. Conclusions Wide ranges in PPV23 effectiveness estimates for any-CAP were observed, likely due to a great diversity of study populations, circulation of S. pneumoniae serotypes, coverage of pediatric pneumococcal vaccination, case definition and time since vaccination. Despite some evidence for short-term protection, effectiveness of PPV23 against CAP was not consistent in the general population, the immunocompromised and subjects with underlying risk factors.

Application of Probabilistic Multiple-Bias Analyses to a Cohort- and a Case-Control Study on the Association between Pandemrix™ and Narcolepsy.

Background An increase in narcolepsy cases was observed in Finland and Sweden towards the end of the 2009 H1N1 influenza pandemic. Preliminary observational studies suggested a temporal link with the pandemic influenza vaccine Pandemrix™, leading to a number of additional studies across Europe. Given the public health urgency, these studies used readily available retrospective data from various sources. The potential for bias in such settings was generally acknowledged. Although generally advocated by key opinion leaders and international health authorities, no systematic quantitative assessment of the potential joint impact of biases was undertaken in any of these studies. Methods We applied bias-level multiple-bias analyses to two of the published narcolepsy studies: a pediatric cohort study from Finland and a case-control study from France. In particular, we developed Monte Carlo simulation models to evaluate a potential cascade of biases, including confounding by age, by indication and by natural H1N1 infection, selection bias, disease- and exposure misclassification. All bias parameters were evidence-based to the extent possible. Results Given the assumptions used for confounding, selection bias and misclassification, the Finnish rate ratio of 13.78 (95% CI: 5.72–28.11) reduced to a median value of 6.06 (2.5th- 97.5th percentile: 2.49–15.1) and the French odds ratio of 5.43 (95% CI: 2.6–10.08) to 1.85 (2.5th—97.5th percentile: 0.85–4.08). Conclusion We illustrate multiple-bias analyses using two studies on the Pandemrix™-narcolepsy association and advocate their use to better understand the robustness of study findings. Based on our multiple-bias models, the observed Pandemrix™-narcolepsy association consistently persists in the Finnish study. For the French study, the results of our multiple-bias models were inconclusive.

Thyroid cancer incidence around the Belgian nuclear sites: Surrogate exposure modelling

BACKGROUND In a recent ecological study among residents living around Belgian nuclear sites (the NUCABEL study), significant increased incidences of thyroid cancer were observed around the two nuclear facilities with industrial and research activities (Mol-Dessel and Fleurus), prompting further research. METHODS The data from the NUCABEL study were reanalysed to test the hypothesis of a gradient in cancer incidence with increasing levels of exposure from these sites using three measures of surrogate exposure, being (i) residential proximity, (ii) prevailing wind directions and (iii) simulated dispersion of radioactive discharges. Single-site focussed hypothesis tests were complemented with Generalized Additive Models to estimate the exposure-response relationships. RESULTS For Mol-Dessel, the results of the focussed hypothesis tests were far from significant. For Fleurus, the p-values were much closer to significance with p=0.05 for Bithells Linear Risk Score using radioactive discharge estimates as surrogate. CONCLUSIONS The re-analyses refute an association with the nuclear facilities for the site of Mol-Dessel. For the site of Fleurus, one of Europes major production sites of radio-iodines, the results were less conclusive and further research suggests itself.

Methodology for computing the burden of disease of adverse events following immunization.

Composite disease burden measures such as disability‐adjusted life‐years (DALY) have been widely used to quantify the population‐level health impact of disease or injury, but application has been limited for the estimation of the burden of adverse events following immunization. Our objective was to assess the feasibility of adapting the DALY approach for estimating adverse event burden.

Background rates of disease in Latin American children from a rotavirus vaccine study

ABSTRACT Background: Knowledge of background rates of adverse events is crucial to assess vaccine safety concerns. We used data from a rotavirus vaccine study (Ruiz-Palacios et al., NEJM, 2006) including 63,225 infants from 11 Latin American countries to investigate reporting rates of serious adverse events (SAEs) among these infants, and describe rates by country, gender, age, and season. Methods: For this randomized, double-blind, placebo-controlled, phase 3 trial, investigators from Argentina, Brazil, Chile, Colombia, Dominican Republic, Honduras, Mexico, Nicaragua, Panama, Peru, and Venezuela recruited 6-to-13-week-old healthy infants. The infants received 2 oral doses of vaccine or placebo. The study population was followed 100 d for the assessment of adverse events. SAEs were captured by an active surveillance system. Results: Strong differences in event rates could be observed between countries (min. 48.1/10,000 person-years in Dominican Republic/Peru; max. 296.2/10,000 person-years in Brazil) and between genders: gastroenteritis, pneumonia, bronchiolitis and bronchitis occurred significantly more frequently in males. In addition, infections and infestations, and most disorders, including immune system and cardiac disorders, were more frequent at earlier ages. Finally, looking at seasonality we noted higher rates of SAEs in the second half of the year in all countries except Mexico. Discussion: Significant differences in reporting rates of SAEs between countries, gender and calendar months illustrate the importance of knowing the local epidemiology when interpreting SAEs. Data from clinical trials can be used to better understand background rates of diseases that may be perceived as potential adverse events following immunization.

A systematic review and meta-analysis on the safety of newly adjuvanted vaccines among older adults

INTRODUCTION New adjuvants have been developed to improve the efficacy of vaccines and for dose-sparing capacity and may overcome immuno senescence in the elderly. We reviewed the safety of newly-adjuvanted vaccines in older adults. METHODS We searched Medline for clinical trials (CTs) including new adjuvant systems (AS01, AS02, AS03, or MF59), used in older adults, published between 01/1995 and 09/2017. Safety outcomes were: serious adverse events (SAEs); solicited local and general AEs (reactogenicity); unsolicited AEs; and potentially immune-mediated diseases (pIMDs). Standard random effects meta-analyses were conducted by type of safety event and adjuvant type, reporting Relative Risks (RR) with 95% confidence intervals (95% CI). RESULTS We identified 1040 publications, from which we selected 7, 7, and 12 CTs on AS01/AS02, AS03 and MF59, respectively. 47,602 study participants received newly-adjuvanted vaccine and 44,521 control vaccine, or placebo. Rates of SAEs (RR = 0.99, 95% CI = 0.96-1.02), deaths (RR = 0.99, 95% CI = 0.92-1.06) and pIMDs (RR = 0.94, 95% CI = 0.79-1.1) were comparable in newly-adjuvanted and control groups. Vaccine-related SAEs occurred in <1% of the subjects in both groups. The reactogenicity of AS01/AS02 and AS03 adjuvanted vaccines was higher compared to control vaccines, whereas MF59-adjuvanted vaccines resulted only in more pain. Grade 3 reactogenicity was reported infrequently, with fatigue (RR = 2.48, 95% CI = 1.69-3.64), headache (RR = 2.94, 95% CI = 1.24-6.95), and myalgia (RR = 2.68, 95% CI = 1.86-3.80) occurring more frequently in newly-adjuvanted groups. Unsolicited AEs occurred slightly more frequently in newly-adjuvanted groups (RR = 1.04, 95% CI = 1.00-1.08). CONCLUSIONS Our review suggests that, within the clinical trial setting, the use of new adjuvants in older adults has not led to any safety concerns, with no increase in SAEs or fatalities. Higher rates for solicited AEs were observed, especially for AS01/AS02 and AS03 adjuvanted vaccines, but AEs were mostly mild and transient. Further evidence will need to come from the use of new adjuvants in the real-world setting, where larger numbers can be studied to potentially detect rare reactions.