End of 2022/23 Season Influenza Vaccine Effectiveness in Primary Care in Great Britain.

BACKGROUND: The 2022/23 influenza season in the United Kingdom saw the return of influenza to prepandemic levels following two seasons with low influenza activity. The early season was dominated by A(H3N2), with cocirculation of A(H1N1), reaching a peak late December 2022, while influenza B circulated at low levels during the latter part of the season. From September to March 2022/23, influenza vaccines were offered, free of charge, to all aged 2-13 (and 14-15 in Scotland and Wales), adults up to 49 years of age with clinical risk conditions and adults aged 50 and above across the mainland United Kingdom. METHODS: End-of-season adjusted vaccine effectiveness (VE) estimates against sentinel primary-care attendance for influenza-like illness, where influenza infection was laboratory confirmed, were calculated using the test negative design, adjusting for potential confounders. METHODS: Results In the mainland United Kingdom, end-of-season VE against all laboratory-confirmed influenza for all those > 65 years of age, most of whom received adjuvanted quadrivalent vaccines, was 30% (95% CI: -6% to 54%). VE for those aged 18-64, who largely received cell-based vaccines, was 47% (95% CI: 37%-56%). Overall VE for 2-17 year olds, predominantly receiving live attenuated vaccines, was 66% (95% CI: 53%-76%). CONCLUSION: The paper provides evidence of moderate influenza VE in 2022/23.

Influenza vaccination during the 2021/22 season: A data-linkage test-negative case-control study of effectiveness against influenza requiring emergency care in England and serological analysis of primary care patients.

We present England 2021/22 end-of-season adjusted vaccine effectiveness (aVE) against laboratory confirmed influenza related emergency care use in children aged 1-17 and in adults aged 50+, and serological findings in vaccinated vs unvaccinated adults by hemagglutination inhibition assay. Influenza vaccination has been routinely offered to all children aged 2-10 years and adults aged 65 years + in England. In 2021/22, the offer was extended to children to age 15 years, and adults aged 50-64 years. Influenza activity rose during the latter half of the 2021/22 season, while remaining comparatively low due to COVID-19 pandemic control measures. Influenza A(H3N2) strains predominated. A test negative design was used to estimate aVE by vaccine type. Cases and controls were identified within a sentinel laboratory surveillance system. Vaccine histories were obtained from the National Immunisation Management Service (NIMS), an influenza and COVID-19 vaccine registry. These were linked to emergency department presentations (excluding accidents) with respiratory swabbing ≤ 14 days before or ≤ 7 days after presentation. Amongst adults, 423 positive and 32,917 negative samples were eligible for inclusion, and 145 positive and 6,438 negative samples among children. Those admitted to hospital were further identified. In serology against the circulating A(H3N2) A/Bangladesh/4005/2020-like strain, 61 % of current season adult vaccinees had titres ≥ 1:40 compared to 17 % of those unvaccinated in 2020/21 or 2021/22 (p < 0.001). We found good protection from influenza vaccination against influenza requiring emergency care in children (72.7 % [95 % CI 52.7, 84.3 %]) and modest effectiveness in adults (26.1 % [95 % CI 4.5, 42.8 %]). Adult VE was higher for A(H1N1) (81 % [95 % CI 50, 93 %]) than A(H3N2) (33 % [95 % CI 6, 53 %]). Consistent protection was observable across preschool, primary and secondary school aged children. Imperfect test specificity combined with very low prevalence may have biased estimates towards null. With limited influenza circulation, the study could not determine differences by vaccine types.

Erratum to "Risk of cardiac arrhythmia and cardiac arrest after primary and booster COVID-19 vaccination in England: A self-controlled case series analysis" [Vaccine: X 15 (2023) 100418].

[This corrects the article DOI: 10.1016/j.jvacx.2023.100418.].

Risk of cardiac arrhythmia and cardiac arrest after primary and booster COVID-19 vaccination in England: A self-controlled case series analysis.

Background: Various cardiac arrhythmias have been reported after COVID-19 infection and vaccination. We assessed the risk after primary immunisation with the ChAdOx1 adenovirus vectored vaccine, and primary and booster immunisation with an mRNA vaccine in 40 million vaccinated adults with 121 million doses (33.9% ChAdOx1 and 66.1% mRNA) in England. Methods: Hospital admissions for a cardiac arrhythmia and emergency care attendance for a cardiac arrest in individuals aged 18 years and older on the 31st March 2021 were linked to the national COVID-19 immunisation register. The incidence of events 1-14 and 15-28 days after vaccination relative to a post-vaccination control period was estimated using the self-controlled case series method modified for fatal events. Outcomes were stratified by arrhythmia type, vaccine type, age group and dose number (up to five). Elevated relative incidence (RI) estimates with p < 0.001 were considered strong evidence of an association. Findings: There was an increased risk of admission for arrhythmia events that were largely palpitations without myocarditis within 14 days of a second priming dose of an mRNA vaccine in 18-49 year olds with an RI of 1.66 (95 % confidence interval 1.47,1.86) for BNT162b2 and 3.75 (2.52,5.57) for mRNA-1273 (p < 0.001) and also after a first booster dose, 1.34 (1.17,1.53) and 1.75 (1.43,2.15) respectively (p < 0.001). No other cardiac arrhythmia, including cardiac arrest, showed an elevated incidence within 28 days of vaccination for any dose, age group or vaccine type. In contrast the risk of a cardiac arrhythmia of all types, including a cardiac arrest, was consistently elevated in those testing positive for SARS-CoV-2 infection. Interpretation: Our study provides reassuring evidence of the safety of the ChAdOx1 and mRNA COVID-19 vaccines with respect to serious cardiac arrhythmias and of the favourable risk benefit of mRNA booster vaccination.

COVID-19 vaccine effectiveness against hospitalisation and death of people in clinical risk groups during the Delta variant period: English primary care network cohort study.

BACKGROUND: COVID-19 vaccines have been shown to be highly effective against hospitalisation and death following COVID-19 infection. COVID-19 vaccine effectiveness estimates against severe endpoints among individuals with clinical conditions that place them at increased risk of critical disease are limited. METHODS: We used English primary care medical record data from the Oxford-Royal College of General Practitioners Research and Surveillance Centre sentinel network (N > 18 million). Data were linked to the National Immunisation Management Service database, Second Generation Surveillance System for virology test data, Hospital Episode Statistics, and death registry data. We estimated adjusted vaccine effectiveness (aVE) against COVID-19 infection followed by hospitalisation and death among individuals in specific clinical risk groups using a cohort design during the delta-dominant period. We also report mortality statistics and results from our antibody surveillance in this population. FINDINGS: aVE against severe endpoints was high, 14-69d following a third dose aVE was 96.4% (95.1%-97.4%) and 97.9% (97.2%-98.4%) for clinically vulnerable people given a Vaxzevria and Comirnaty primary course respectively. Lower aVE was observed in the immunosuppressed group: 88.6% (79.1%-93.8%) and 91.9% (85.9%-95.4%) for Vaxzevria and Comirnaty respectively. Antibody levels were significantly lower among the immunosuppressed group than those not in this risk group across all vaccination types and doses. The standardised case fatality rate within 28 days of a positive test was 3.9/1000 in people not in risk groups, compared to 12.8/1000 in clinical risk groups. Waning aVE with time since 2nd dose was also demonstrated, for example, Comirnaty aVE against hospitalisation reduced from 96.0% (95.1-96.7%) 14-69days post-dose 2-82.9% (81.4-84.2%) 182days+ post-dose 2. INTERPRETATION: In all clinical risk groups high levels of vaccine effectiveness against severe endpoints were seen. Reduced vaccine effectiveness was noted among the immunosuppressed group.

Risk of myocarditis and pericarditis after a COVID-19 mRNA vaccine booster and after COVID-19 in those with and without prior SARS-CoV-2 infection: A self-controlled case series analysis in England.

BACKGROUND: An increased risk of myocarditis or pericarditis after priming with mRNA Coronavirus Disease 2019 (COVID-19) vaccines has been shown but information on the risk post-booster is limited. With the now high prevalence of prior Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection, we assessed the effect of prior infection on the vaccine risk and the risk from COVID-19 reinfection. METHODS AND FINDINGS: We conducted a self-controlled case series analysis of hospital admissions for myocarditis or pericarditis in England between 22 February 2021 and 6 February 2022 in the 50 million individuals eligible to receive the adenovirus-vectored vaccine (ChAdOx1-S) for priming or an mRNA vaccine (BNT162b2 or mRNA-1273) for priming or boosting. Myocarditis and pericarditis admissions were extracted from the Secondary Uses Service (SUS) database in England and vaccination histories from the National Immunisation Management System (NIMS); prior infections were obtained from the UK Health Security Agency's Second-Generation Surveillance Systems. The relative incidence (RI) of admission within 0 to 6 and 7 to 14 days of vaccination compared with periods outside these risk windows stratified by age, dose, and prior SARS-CoV-2 infection for individuals aged 12 to 101 years was estimated. The RI within 27 days of an infection was assessed in the same model. There were 2,284 admissions for myocarditis and 1,651 for pericarditis in the study period. Elevated RIs were only observed in 16- to 39-year-olds 0 to 6 days postvaccination, mainly in males for myocarditis. Both mRNA vaccines showed elevated RIs after first, second, and third doses with the highest RIs after a second dose 5.34 (95% confidence interval (CI) [3.81, 7.48]; p < 0.001) for BNT162b2 and 56.48 (95% CI [33.95, 93.97]; p < 0.001) for mRNA-1273 compared with 4.38 (95% CI [2.59, 7.38]; p < 0.001) and 7.88 (95% CI [4.02, 15.44]; p < 0.001), respectively, after a third dose. For ChAdOx1-S, an elevated RI was only observed after a first dose, RI 5.23 (95% CI [2.48, 11.01]; p < 0.001). An elevated risk of admission for pericarditis was only observed 0 to 6 days after a second dose of mRNA-1273 vaccine in 16 to 39 year olds, RI 4.84 (95% CI [1.62, 14.01]; p = 0.004). RIs were lower in those with a prior SARS-CoV-2 infection than in those without, 2.47 (95% CI [1.32,4.63]; p = 0.005) versus 4.45 (95% [3.12, 6.34]; p = 0.001) after a second BNT162b2 dose, and 19.07 (95% CI [8.62, 42.19]; p < 0.001) versus 37.2 (95% CI [22.18, 62.38]; p < 0.001) for mRNA-1273 (myocarditis and pericarditis outcomes combined). RIs 1 to 27 days postinfection were elevated in all ages and were marginally lower for breakthrough infections, 2.33 (95% CI [1.96, 2.76]; p < 0.001) compared with 3.32 (95% CI [2.54, 4.33]; p < 0.001) in vaccine-naïve individuals respectively. CONCLUSIONS: We observed an increased risk of myocarditis within the first week after priming and booster doses of mRNA vaccines, predominantly in males under 40 years with the highest risks after a second dose. The risk difference between the second and the third doses was particularly marked for the mRNA-1273 vaccine that contains half the amount of mRNA when used for boosting than priming. The lower risk in those with prior SARS-CoV-2 infection, and lack of an enhanced effect post-booster, does not suggest a spike-directed immune mechanism. Research to understand the mechanism of vaccine-associated myocarditis and to document the risk with bivalent mRNA vaccines is warranted.

Real-world data on immune responses following heterologous prime-boost COVID-19 vaccination schedule with Pfizer and AstraZeneca vaccines in England.

BACKGROUND: There are limited data on immune responses to heterologous COVID-19 immunisation schedules, especially following an extended ≥12-week interval between doses. METHODS: SARS-CoV-2 infection-naïve and previously-infected adults receiving ChAd-BNT (ChAdOx1 nCoV-19, AstraZeneca followed by BNT162b2, Pfizer-BioNTech) or BNT-ChAd as part of the UK national immunisation programme provided blood samples at 30 days and 12 weeks after their second dose. Geometric mean concentrations (GMC) of anti-SARS-CoV-2 spike (S-antibody) and nucleoprotein (N-antibody) IgG antibodies and geometric mean ratios (GMR) were compared with a contemporaneous cohort receiving homologous ChAd-ChAd or BNT-BNT. RESULTS: During March-October 2021, 75,827 individuals were identified as having received heterologous vaccination, 9,489 invited to participate, 1,836 responded (19.3%) and 656 were eligible. In previously-uninfected adults, S-antibody GMC at 30 days post-second dose were lowest for ChAd-ChAd (862 [95% CI, 694 - 1069]) and significantly higher for ChAd-BNT (6233 [5522-7035]; GMR 6.29; [5.04-7.85]; p<0.001), BNT-ChAd (4776 [4066-5610]; GMR 4.55 [3.56-5.81]; p<0.001) and BNT-BNT (5377 [4596-6289]; GMR 5.66 [4.49-7.15]; p<0.001). By 12 weeks after dose two, S-antibody GMC had declined in all groups and remained significantly lower for ChAd-ChAd compared to ChAd-BNT (GMR 5.12 [3.79-6.92]; p<0.001), BNT-ChAd (GMR 4.1 [2.96-5.69]; p<0.001) and BNT-BNT (GMR 6.06 [4.32-8.50]; p<0.001). Previously infected adults had higher S-antibody GMC compared to infection-naïve adults at all time-points and with all vaccine schedules. CONCLUSIONS: These real-world findings demonstrate heterologous schedules with adenoviral-vector and mRNA vaccines are highly immunogenic and may be recommended after a serious adverse reaction to one vaccine product, or to increase programmatic flexibility where vaccine supplies are constrained.

Pfizer-BioNTech and Oxford AstraZeneca COVID-19 vaccine effectiveness and immune response amongst individuals in clinical risk groups.

Background COVID-19 vaccines approved in the UK are highly effective in general population cohorts, however, data on effectiveness amongst individuals with clinical conditions that place them at increased risk of severe disease are limited. Methods We used GP electronic health record data, sentinel virology swabbing and antibody testing within a cohort of 712 general practices across England to estimate vaccine antibody response and vaccine effectiveness against medically attended COVID-19 amongst individuals in clinical risk groups using cohort and test-negative case control designs. Findings There was no reduction in S-antibody positivity in most clinical risk groups, however reduced S-antibody positivity and response was significant in the immunosuppressed group. Reduced vaccine effectiveness against clinical disease was also noted in the immunosuppressed group; after a second dose, effectiveness was moderate (Pfizer: 59.6%, 95%CI 18.0-80.1%; AstraZeneca 60.0%, 95%CI -63.6-90.2%). Interpretation In most clinical risk groups, immune response to primary vaccination was maintained and high levels of vaccine effectiveness were seen. Reduced antibody response and vaccine effectiveness were seen after 1 dose of vaccine amongst a broad immunosuppressed group, and second dose vaccine effectiveness was moderate. These findings support maximising coverage in immunosuppressed individuals and the policy of prioritisation of this group for third doses.

Impact of COVID-19 vaccination program on seroprevalence in blood donors in England, 2021.

The COVID-19 vaccination programme commenced in England on 8th December 2020 primarily based on age; by 7th March 2021 approximately 93% of the English population aged 70+ years had received at least 1 dose of either the Pfizer BioNTech or AstraZeneca vaccines. Using a nucleoprotein assay that detects antibodies following natural infection only and a spike assay that detects infection and vaccine-induced responses, we aim to describe the impact of vaccination on SARS-CoV-2 antibody prevalence in English blood donors.

Serological surveillance of SARS-CoV-2: Six-month trends and antibody response in a cohort of public health workers.

BACKGROUND: Antibody waning after SARS-CoV-2 infection may result in reduction in long-term immunity following natural infection and vaccination, and is therefore a major public health issue. We undertook prospective serosurveillance in a large cohort of healthy adults from the start of the epidemic in England. METHODS: Clinical and non-clinical healthcare workers were recruited across three English regions and tested monthly from March to November 2020 for SARS-CoV-2 spike (S) protein and nucleoprotein (N) antibodies using five different immunoassays. In positive individuals, antibody responses and long-term trends were modelled using mixed effects regression. FINDINGS: In total, 2246 individuals attended 12,247 visits and 264 were seropositive in ≥ 2 assays. Most seroconversions occurred between March and April 2020. The assays showed > 85% agreement for ever-positivity, although this changed markedly over time. Antibodies were detected earlier with Abbott (N) but declined rapidly thereafter. With the EuroImmun (S) and receptor-binding domain (RBD) assays, responses increased for 4 weeks then fell until week 12-16 before stabilising. For Roche (N), responses increased until 8 weeks, stabilised, then declined, but most remained above the positive threshold. For Roche (S), responses continued to climb over the full 24 weeks, with no sero-reversions. Predicted proportions sero-reverting after 52 weeks were 100% for Abbott, 59% (95% credible interval 50-68%) Euroimmun, 41% (30-52%) RBD, 10% (8-14%) Roche (N) < 2% Roche (S). INTERPRETATION: Trends in SARS-CoV-2 antibodies following infection are highly dependent on the assay used. Ongoing serosurveillance using multiple assays is critical for monitoring the course and long-term progression of SARS-CoV-2 antibodies.

Control yourself: ISPE-endorsed guidance in the application of self-controlled study designs in pharmacoepidemiology.

PURPOSE: Consensus is needed on conceptual foundations, terminology and relationships among the various self-controlled "trigger" study designs that control for time-invariant confounding factors and target the association between transient exposures (potential triggers) and abrupt outcomes. The International Society for Pharmacoepidemiology (ISPE) funded a working group of ISPE members to develop guidance material for the application and reporting of self-controlled study designs, similar to Standards of Reporting Observational Epidemiology (STROBE). This first paper focuses on navigation between the types of self-controlled designs to permit a foundational understanding with guiding principles. METHODS: We leveraged a systematic review of applications of these designs, that we term Self-controlled Crossover Observational PharmacoEpidemiologic (SCOPE) studies. Starting from first principles and using case examples, we reviewed outcome-anchored (case-crossover [CCO], case-time control [CTC], case-case-time control [CCTC]) and exposure-anchored (self-controlled case-series [SCCS]) study designs. RESULTS: Key methodological features related to exposure, outcome and time-related concerns were clarified, and a common language and worksheet to facilitate the design of SCOPE studies is introduced. CONCLUSIONS: Consensus on conceptual foundations, terminology and relationships among SCOPE designs will facilitate understanding and critical appraisal of published studies, as well as help in the design, analysis and review of new SCOPE studies. This manuscript is endorsed by ISPE.

Assessing the impact of preventive mass vaccination campaigns on yellow fever outbreaks in Africa: A population-level self-controlled case series study.

BACKGROUND: The Eliminate Yellow fever Epidemics (EYE) strategy was launched in 2017 in response to the resurgence of yellow fever in Africa and the Americas. The strategy relies on several vaccination activities, including preventive mass vaccination campaigns (PMVCs). However, to what extent PMVCs are associated with a decreased risk of outbreak has not yet been quantified. METHODS AND FINDINGS: We used the self-controlled case series (SCCS) method to assess the association between the occurrence of yellow fever outbreaks and the implementation of PMVCs at the province level in the African endemic region. As all time-invariant confounders are implicitly controlled for in the SCCS method, this method is an alternative to classical cohort or case-control study designs when the risk of residual confounding is high, in particular confounding by indication. The locations and dates of outbreaks were identified from international epidemiological records, and information on PMVCs was provided by coordinators of vaccination activities and international funders. The study sample consisted of provinces that were both affected by an outbreak and targeted for a PMVC between 2005 and 2018. We compared the incidence of outbreaks before and after the implementation of a PMVC. The sensitivity of our estimates to a range of assumptions was explored, and the results of the SCCS method were compared to those obtained through a retrospective cohort study design. We further derived the number of yellow fever outbreaks that have been prevented by PMVCs. The study sample consisted of 33 provinces from 11 African countries. Among these, the first outbreak occurred during the pre-PMVC period in 26 (79%) provinces, and during the post-PMVC period in 7 (21%) provinces. At the province level, the post-PMVC period was associated with an 86% reduction (95% CI 66% to 94%, p < 0.001) in the risk of outbreak as compared to the pre-PMVC period. This negative association between exposure to PMVCs and outbreak was robustly observed across a range of sensitivity analyses, especially when using quantitative estimates of vaccination coverage as an alternative exposure measure, or when varying the observation period. In contrast, the results of the cohort-style analyses were highly sensitive to the choice of covariates included in the model. Based on the SCCS results, we estimated that PMVCs were associated with a 34% (95% CI 22% to 45%) reduction in the number of outbreaks in Africa from 2005 to 2018. A limitation of our study is the fact that it does not account for potential time-varying confounders, such as changing environmental drivers of yellow fever and possibly improved disease surveillance. CONCLUSIONS: In this study, we provide new empirical evidence of the high preventive impact of PMVCs on yellow fever outbreaks. This study illustrates that the SCCS method can be advantageously applied at the population level in order to evaluate a public health intervention.

Self-controlled case series studies: Just how rare does a rare non-recurrent outcome need to be?

The self-controlled case series method assumes that adverse outcomes arise according to a non-homogeneous Poisson process. This implies that it is applicable to independent recurrent outcomes. However, the self-controlled case series method may also be applied to unique, non-recurrent outcomes or first outcomes only, in the limit where these become rare. We investigate this rare outcome assumption when the self-controlled case series method is applied to non-recurrent outcomes. We study this requirement analytically and by simulation, and quantify what is meant by 'rare' in this context. In simulations we also apply the self-controlled risk interval design, a special case of the self-controlled case series design. To illustrate, we extract data on the incidence rate of some recurrent and non-recurrent outcomes within a defined study population to check whether outcomes are sufficiently rare for the rare outcome assumption to hold when applying the self-controlled case series method to first or unique outcomes. The main findings are that the relative bias should be no more than 5% when the cumulative incidence over total time observed is less than 0.1 per individual. Inclusion of age (or calendar time) effects will further reduce bias. Designs that begin observation with exposure maximise bias, whereas little or no bias will be apparent when there is no time trend in the distribution of exposures, or when exposure is central within time observed.

Investigating the assumptions of the self-controlled case series method.

We describe some simple techniques for investigating 2 key assumptions of the self-controlled case series (SCCS) method, namely, that events do not influence subsequent exposures and that events do not influence the length of observation periods. For each assumption, we propose some simple tests based on the standard SCCS model, along with associated graphical displays. The methods also enable the user to investigate the robustness of the results obtained using the standard SCCS model to failure of assumptions. The proposed methods are investigated by simulations and applied to data on measles, mumps and rubella vaccine, and antipsychotics.

Thromboembolic and neurologic sequelae of discontinuation of an antihyperlipidemic drug during ongoing warfarin therapy.

Warfarin and antihyperlipidemics are commonly co-prescribed. Some antihyperlipidemics may inhibit warfarin deactivation via the hepatic cytochrome P450 system. Therefore, antihyperlipidemic discontinuation has been hypothesized to result in underanticoagulation, as warfarin metabolism is no longer inhibited. We quantified the risk of venous thromboembolism (VTE) and ischemic stroke (IS) due to statin and fibrate discontinuation in warfarin users, in which warfarin was initially dose-titrated during ongoing antihyperlipidemic therapy. Using 1999-2011 United States Medicaid claims among 69 million beneficiaries, we conducted a set of bidirectional self-controlled case series studies-one for each antihyperlipidemic. Outcomes were hospital admissions for VTE/IS. The risk segment was a maximum of 90 days immediately following antihyperlipidemic discontinuation, the exposure of interest. Time-varying confounders were included in conditional Poisson models. We identified 629 study eligible-persons with at least one outcome. Adjusted incidence rate ratios (IRRs) for all antihyperlipidemics studied were consistent with the null, and ranged from 0.21 (0.02, 2.82) for rosuvastatin to 2.16 (0.06, 75.0) for gemfibrozil. Despite using an underlying dataset of millions of persons, we had little precision in estimating IRRs for VTE/IS among warfarin-treated persons discontinuing individual antihyperlipidemics. Further research should investigate whether discontinuation of gemfibrozil in warfarin users results in serious underanticoagulation.

Relative risk of hemorrhage during pregnancy in patients with brain arteriovenous malformations.

Background It is unclear whether the risk of bleeding from brain arteriovenous malformations is higher during pregnancy, delivery, or puerperium. We compared occurrence of brain arteriovenous malformation hemorrhage in women during this period with occurrence of hemorrhage outside this period during their fertile years. Methods We included all women with ruptured brain arteriovenous malformations (16-41 years) from a retrospective database of patients with brain arteriovenous malformations in four Dutch university hospitals (n = 95) and from the population-based Scottish Audit of Intracranial Vascular Malformations (n = 44). We estimated the relative rate of brain arteriovenous malformation rupture (before any treatment) during exposed time (pregnancy, delivery, puerperium) versus non-exposed time during fertile years, using the case-crossover design as primary analysis, and the self-controlled case-series design as secondary analysis. Results In 17 of 95 Dutch women and in 3 of 44 Scottish women, hemorrhages occurred while pregnant; none occurred during delivery or puerperium. In Dutch women, the relative rate of brain arteriovenous malformation rupture during pregnancy, delivery, or puerperium was 6.8 (95% confidence interval 3.6-13) according to the case-crossover method and 7.1 (95% confidence interval 3.4-13) using the self-controlled case-series method. In Scottish women, the relative rate was 1.3 (95% confidence interval 0.39-4.1) using the case-crossover method and 1.7 (95% confidence interval 0.0-4.4) according to the self-controlled case-series method. Because of limited overlap of confidence intervals, we refrained from pooling the cohorts. Conclusions Case-crossover and self-controlled case series analyses reveal an increase in relative rate of brain arteriovenous malformation rupture during pregnancy in the Dutch cohort but not in the Scottish cohort. Since point estimates varied between both cohorts and numbers are relatively small, the clinical implications of our findings are uncertain.

Spline-based self-controlled case series method.

The self-controlled case series (SCCS) method is an alternative to study designs such as cohort and case control methods and is used to investigate potential associations between the timing of vaccine or other drug exposures and adverse events. It requires information only on cases, individuals who have experienced the adverse event at least once, and automatically controls all fixed confounding variables that could modify the true association between exposure and adverse event. Time-varying confounders such as age, on the other hand, are not automatically controlled and must be allowed for explicitly. The original SCCS method used step functions to represent risk periods (windows of exposed time) and age effects. Hence, exposure risk periods and/or age groups have to be prespecified a priori, but a poor choice of group boundaries may lead to biased estimates. In this paper, we propose a nonparametric SCCS method in which both age and exposure effects are represented by spline functions at the same time. To avoid a numerical integration of the product of these two spline functions in the likelihood function of the SCCS method, we defined the first, second, and third integrals of I-splines based on the definition of integrals of M-splines. Simulation studies showed that the new method performs well. This new method is applied to data on pediatric vaccines. Copyright © 2017 John Wiley & Sons, Ltd.

Flexible modelling of vaccine effect in self-controlled case series models.

The self-controlled case series (SCCS) method, commonly used to investigate the safety of vaccines, requires information on cases only and automatically controls all age-independent multiplicative confounders, while allowing for an age-dependent baseline incidence. Currently, the SCCS method represents the time-varying exposures using step functions with pre-determined cut points. A less prescriptive approach may be beneficial when the shape of the relative risk function associated with exposure is not known a priori, especially when exposure effects can be long-lasting. We therefore propose to model exposure effects using flexible smooth functions. Specifically, we used a linear combination of cubic M-splines which, in addition to giving plausible shapes, avoids the integral in the log-likelihood function of the SCCS model. The methods, though developed specifically for vaccines, are applicable more widely. Simulations showed that the new approach generally performs better than the step function method. We applied the new method to two data sets, on febrile convulsion and exposure to MMR vaccine, and on fractures and thiazolidinedione use.

Self-controlled case series method with smooth age effect.

The self-controlled case series method, commonly used to investigate potential associations between vaccines and adverse events, requires information on cases only and automatically controls all age-independent multiplicative confounders while allowing for an age-dependent baseline incidence. In the parametric version of the method, we modelled the age-specific relative incidence by using a piecewise constant function, whereas in the semiparametric version, we left it unspecified. However, mis-specification of age groups in the parametric version can lead to biassed estimates of exposure effect, and the semiparametric approach runs into computational problems when the number of cases in the study is moderately large. We, thus, propose to use a penalized likelihood approach where the age effect is modelled using splines. We use a linear combination of cubic M-splines to approximate the age-specific relative incidence and integrated splines for the cumulative relative incidence. We conducted a simulation study to evaluate the performance of the new approach and its efficiency relative to the parametric and semiparametric approaches. Results show that the new approach performs equivalently to the existing methods when the sample size is small and works well for large data sets. We applied the new spline-based approach to data on febrile convulsions and paediatric vaccines. Co