Explore the vast range of titles available.

In this tutorial, we provide a broad introduction to the topic of interaction between the effects of exposures. We discuss interaction on both additive and multiplicative scales using risks, and we discuss their relation to statistical models (e.g. linear, log-linear, and logistic models). We discuss and evaluate arguments that have been made for using additive or multiplicative scales to assess interaction. We further discuss approaches to presenting interaction analyses, different mechanistic forms of interaction, when interaction is robust to unmeasured confounding, interaction for continuous outcomes, qualitative or “crossover” interactions, methods for attributing effects to interactions, case-only estimators of interaction, and power and sample size calculations for additive and multiplicative interaction.

Given the emergence of the SARS-CoV-2 Omicron BA.1 and BA.2 variants and the roll-out of booster COVID-19 vaccination, evidence is needed on protection conferred by primary vaccination, booster vaccination and previous SARS-CoV-2 infection by variant. We employed a test-negative design on S-gene target failure data from community PCR testing in the Netherlands from 22 November 2021 to 31 March 2022 (n = 671,763). Previous infection, primary vaccination or both protected well against Delta infection. Protection against Omicron BA.1 infection was much lower compared to Delta. Protection was similar against Omicron BA.1 compared to BA.2 infection after previous infection, primary and booster vaccination. Higher protection was observed against all variants in individuals with both vaccination and previous infection compared with either one. Protection against all variants decreased over time since last vaccination or infection. We found that primary vaccination with current COVID-19 vaccines and previous SARS-CoV-2 infections offered low protection against Omicron BA.1 and BA.2 infection. Booster vaccination considerably increased protection against Omicron infection, but decreased rapidly after vaccination. The protection of COVID-19 vaccines against emerging variants needs to be monitored. Here, the authors use community testing data from the Netherlands and find that protection against infection by Omicron subvariants BA.1 and 2 is low and that booster vaccines considerably but temporarily increase protection.

An increasing proportion of the population has acquired immunity through COVID-19 vaccination and previous SARS-CoV-2 infection, i.e., hybrid immunity, possibly affecting the risk of new infection. We aim to estimate the protective effect of previous infections and vaccinations on SARS-CoV-2 Omicron infection, using data from 43,257 adult participants in a prospective community-based cohort study in the Netherlands, collected between 10 January 2022 and 1 September 2022. Our results show that, for participants with 2, 3 or 4 prior immunizing events (vaccination or previous infection), hybrid immunity is more protective against infection with SARS-CoV-2 Omicron than vaccine-induced immunity, up to at least 30 weeks after the last immunizing event. Differences in risk of infection are partly explained by differences in anti-Spike RBD (S) antibody concentration, which is associated with risk of infection in a dose-response manner. Among participants with hybrid immunity, with one previous pre-Omicron infection, we do not observe a relevant difference in risk of Omicron infection by sequence of vaccination(s) and infection. Additional immunizing events increase the protection against infection, but not above the level of the first weeks after the previous event. The relative protection against Omicron SARS-CoV-2 infection conferred by vaccination and previous infection are not fully understood. Here, the authors use data from a prospective cohort study in the Netherlands and show that hybrid immunity (vaccination plus previous infection) conferred strongest protection.

Measures of interaction on an additive scale (relative excess risk due to interaction [RERI], attributable proportion [AP], synergy index [S]), were developed for risk factors rather than preventive factors. It has been suggested that preventive factors should be recoded to risk factors before calculating these measures. We aimed to show that these measures are problematic with preventive factors prior to recoding, and to clarify the recoding method to be used to circumvent these problems. Recoding of preventive factors should be done such that the stratum with the lowest risk becomes the reference category when both factors are considered jointly (rather than one at a time). We used data from a case-control study on the interaction between ACE inhibitors and the ACE gene on incident diabetes. Use of ACE inhibitors was a preventive factor and DD ACE genotype was a risk factor. Before recoding, the RERI, AP and S showed inconsistent results (RERI = 0.26 [95% CI: -0.30; 0.82], AP = 0.30 [95% CI: -0.28; 0.88], S = 0.35 [95% CI: 0.02; 7.38]), with the first two measures suggesting positive interaction and the third negative interaction. After recoding the use of ACE inhibitors, they showed consistent results (RERI = -0.37 [95% CI: -1.23; 0.49], AP = -0.29 [95% CI: -0.98; 0.40], S = 0.43 [95% CI: 0.07; 2.60]), all indicating negative interaction. Preventive factors should not be used to calculate measures of interaction on an additive scale without recoding.

Most studies on risk factors for a SARS-CoV-2 infection were conducted in the pre-vaccination era with many non-pharmaceutical prevention measures in place. We investigated risk factors for symptomatic SARS-CoV-2 infections in vaccinated persons in a period with a varying degree of prevention measures. In a test-negative case control study among vaccinated adults attending community COVID-19 testing locations between June 1st 2021 till February 28th 2022, we compared symptomatic cases with symptomatic controls (to study risk factors specific for SARS-CoV-2) and with asymptomatic controls (to study risk factors that could apply to respiratory infections in general). We examined potential risk factors including household composition and mitigation behaviour by logistic regression, adjusting for age, sex, and week of testing. Risk factors for a positive SARS-CoV-2 test when symptomatic cases were compared to symptomatic controls were: having a household size of more than 4 (adjusted odds ratio: 1.47; 95% CI 1.14-1.92), being a healthcare worker (1.27;1.18-1.47), and visiting busy locations outside (1.49;1.19-1.87). When symptomatic cases were compared to asymptomatic controls, a household size of more than 4 members (1.71;1.25-2.33), living with children aged 0-12 (1.59;1.12-2.26), visiting busy locations outside (1.64;1.24-2.17) were independent risk factors for a positive SARS-CoV-2 test. Risk factors for separate periods and waves differed from the study period as a whole. This study was conducted in a period with a varying degree of prevention measures. Among vaccinated individuals, we identified several SARS-CoV-2 specific risk factors and SARS-CoV-2 risk factors that could be more general for respiratory infections. For SARS-CoV-2 transmission more attention could be given to visiting busy outdoor locations, having a household size that consists of more than 4 persons, being a healthcare worker, and living with children aged 0-12. Risk factors varied with different phases in the pandemic, emphasizing the importance of repeated assessment of risk factors.

Immunity induced by vaccination and infection, referred to as hybrid immunity, provides better protection against SARS-CoV-2 infections compared to immunity induced by vaccinations alone. To assess the development of hybrid immunity we investigated the induction of Nucleoprotein-specific antibodies in PCR-confirmed infections by Delta or Omicron in vaccinated individuals (n = 520). Eighty-two percent of the participants with a breakthrough infection reached N-seropositivity. N-seropositivity was accompanied by Spike S1 antibody boosting, and independent of vaccination status or virus variant. Following the infection relatively more antibodies to the infecting virus variant were detected. In conclusion, these data show that hybrid immunity through breakthrough infections is hallmarked by Nucleoprotein antibodies and broadening of the Spike antibody repertoire. Exposure to future SARS-CoV-2 variants may therefore continue to maintain and broaden vaccine-induced population immunity.

The 7-valent pneumococcal conjugate vaccine (PCV7) was introduced in The Netherlands in 2006 and was replaced by PHiD-CV10 in 2011. Data on carriage prevalence of S. pneumoniae serotypes in children and invasive pneumococcal disease (IPD) in children and older adults were collected to examine the impact of PCVs on carriage and IPD in The Netherlands. Pneumococcal carriage prevalence was determined by conventional culture of nasopharyngeal swabs in 24-month-old children in 2015/2016. Data were compared to similar carriage studies in 2005 (pre-PCV7 introduction), 2009, 2010/2011 and 2012/2013. Invasive pneumococcal disease isolates from hospitalized children <5 years and adults >65 years (2004-2016) were obtained by sentinel surveillance. All isolates were serotyped by Quellung. Serotype invasive disease potential was calculated using carriage and nationwide IPD data in children. The overall pneumococcal carriage rate was 48% in 2015/2016, lower than in 2010/2011 (64%) and pre-vaccination in 2005 (66%). Carriage of the previously dominant non-vaccine serotypes 19A and 11A has declined since 2010/2011, from 14.2% to 4.6% and 4.2% to 2.7%, respectively, whereas carriage of serotypes 6C and 23B has increased (4.2% to 6.7% and 3.9% to 7.3%), making serotypes 6C and 23B the most prevalent carriage serotypes. IPD incidence declined in children (20/100,000 cases in 2004/2006 to 6/100,000 cases in 2015/2016) as well as in older adults (63/100,000 cases to 51/100,000 cases). Serotypes 6C, 23B and 11A have high carriage prevalence in children, but show low invasive disease potential. Serotype 8 is the main causative agent for IPD in older adults (11.3%). In conclusion, 10 years after the introduction of pneumococcal vaccination in children in The Netherlands shifts in carriage and disease serotypes are still ongoing. Surveillance of both carriage and IPD is important to assess PCV impact and to predict necessary future vaccination strategies in both children and older adults.

Background New 15- and 20-valent pneumococcal vaccines (PCV15, PCV20) are available for both children and adults, while PCV21 for adults is in development. However, their cost-effectiveness for older adults, taking into account indirect protection and serotype replacement from a switch to PCV15 and PCV20 in childhood vaccination, remains unexamined. Methods We used a static model for the Netherlands to assess the cost-effectiveness of different strategies with 23-valent pneumococcal polysaccharide vaccine (PPV23), PCV15, PCV20, and PCV21 for a 65-year-old cohort from a societal perspective, over a 15-year time horizon. Childhood vaccination was varied from PCV10 to PCV13, PCV15, and PCV20. Indirect protection was assumed to reduce the incidence of vaccine serotypes in older adults by 80% (except for serotype 3, no effect), completely offset by an increase in non-vaccine serotype incidence due to serotype replacement. Results Indirect effects from childhood vaccination reduced the cost-effectiveness of vaccination of older adults, depending on the serotype overlap between the vaccines. With PCV10, PCV13, or PCV15 in children, PCV20 was more effective and less costly for older adults than PPV23 and PCV15. PCV20 costs approximately €10,000 per quality-adjusted life year (QALY) gained compared to no pneumococcal vaccination, which falls below the conventional Dutch €20,000/QALY gained threshold. However, with PCV20 in children, PCV20 was no longer considered cost-effective for older adults, costing €22,550/QALY gained. As indirect effects progressed over time, the cost-effectiveness of PCV20 for older adults further diminished for newly vaccinated cohorts. PPV23 was more cost-effective than PCV20 for cohorts vaccinated 3 years after the switch to PCV20 in children. PCV21 offered the most QALY gains, and its cost-effectiveness was minimally affected by indirect effects due to its coverage of 11 different serotypes compared to PCV20. Conclusions For long-term cost-effectiveness in the Netherlands, the pneumococcal vaccine for older adults should either include invasive serotypes not covered by childhood vaccination or become more affordable than its current pricing for individual use.

Background The impact of human papillomavirus (HPV) vaccination programs depends on the degree of indirect protection against new infections achieved among unvaccinated women. We estimated the indirect effect of bivalent HPV vaccination by comparing the HPV-type incidence in unvaccinated female participants between a cohort offered vaccination in 2009/2010 and a cohort of similar-aged women offered vaccination in 2014. Methods We compared the incidence rates of HPV types in the HAVANA cohort (follow-up from 2010/2011 until 2015/2016) with those from the HAVANA-2 cohort (2017–2022) using two regression approaches to estimate the indirect effect of HPV vaccination. First, we calculated the incidence ratio (IRR) for a vaccine or cross-protective type in HAVANA-2 versus HAVANA by Poisson regression and compared it to the IRR for a non-cross-protective type. The indirect vaccine effect is defined as 1-ratio of the IRRs. Second, we performed Cox regression with infection by vaccine or cross-protective type as the endpoint and calculated the hazard ratio (HR) for HAVANA-2 versus HAVANA after adjusting for time-varying sociodemographic variables. The indirect effect is defined as 1-HR. Results We included 661 unvaccinated participants in HAVANA and 927 in HAVANA-2. We observed a significant reduction in incident HPV16 infections of 70.9% (95% CI 48.3–83.7%) with Poisson regression and of 73.1% (95% CI 53.3–84.5%) with Cox regression. For HPV45, significant decreases of 67.3% (95% CI 8.8–88.3%) and 69.8% (95% CI 15.2–89.3%) were observed. For HPV18, HPV31, and HPV33, the indirect effect was not statistically significant. Conclusions Large indirect effects of the bivalent HPV vaccination program were observed for HPV16 and HPV45 infections.