The case test-negative design for studies of the effectiveness of influenza vaccine

Background

A modification to the case–control study design has become popular to assess vaccine effectiveness (VE) against viral infections. Subjects with symptomatic illness seeking medical care are tested by a highly specific polymerase chain reaction (PCR) assay for the detection of the infection of interest. Cases are subjects testing positive for the virus; those testing negative represent the comparison group. Influenza and rotavirus VE studies using this design are often termed “test-negative case-control” studies, but this design has not been formally described or evaluated. We explicitly state several assumptions of the design and examine the conditions under which VE estimates derived with it are valid and unbiased.

Methods

We derived mathematical expressions for VE estimators obtained using this design and examined their statistical properties. We used simulation methods to test the validity of the estimators and illustrate their performance using an influenza VE study as an example.

Results

Because the marginal ratio of cases to non-cases is unknown during enrollment, this design is not a traditional case-control study; we suggest the name “case test-negative” design. Under sets of increasingly general assumptions, we found that the case test-negative design can provide unbiased VE estimates. However, differences in health care-seeking behavior among cases and non-cases by vaccine status, strong viral interference, or modification of the probability of symptomatic illness by vaccine status can bias VE estimates.

Conclusions

Vaccine effectiveness estimates derived from case test-negative studies are valid and unbiased under a wide range of assumptions. However, if vaccinated cases are less severely ill and seek care less frequently than unvaccinated cases, then an appropriate adjustment for illness severity is required to avoid bias in effectiveness estimates. Viral interference will lead to a non-trivial bias in the vaccine effectiveness estimate from case test-negative studies only when incidence of influenza is extremely high and duration of transient non-specific immunity is long.     

Estimates of 2012/13 influenza vaccine effectiveness using the case test-negative control design with different influenza negative control groups

Background

In recent years several reports of influenza vaccine effectiveness (VE) have been made early for public health decision. The majority of these studies use the case test-negative control design (TND), which has been showed to provide, under certain conditions, unbiased estimates of influenza VE. Nevertheless, discussions have been taken on the best influenza negative control group to use. The present study aims to contribute to the knowledge on this field by comparing influenza VE estimates using three test-negative controls: all influenza negative, non-influenza respiratory virus and pan-negative.

Methods

Incident ILI patients were prospectively selected and swabbed by a sample of general practitioners. Cases were ILI patients tested positive for influenza and controls ILI patients tested negative for influenza. The influenza negative control group was divided into non-influenza virus control group and pan-negative control group. Data were collected on vaccination status and confounding factors. Influenza VE was estimated as one minus the odds ratio of been vaccinated in cases versus controls adjusted for confounding effect by logistic regression.

Results

Confounder adjusted influenza VE against medically attended laboratory-confirmed influenza was 68.4% (95% CI: 20.7–87.4%) using all influenza negatives controls, 82.1% (95% CI: 47.6–93.9%) using non-influenza controls and 49.4% (95% CI: −44.7% to 82.3%) using pan-negative controls.

Conclusions

Influenza VE estimates differed according to the influenza negative control group used. These results are in accordance with the expected under the hypothesis of differential viral interference between influenza vaccinated and unvaccinated individuals. Given the wide importance of TND study further studies should be conducted in order to clarify the observed differences.     

The impact of selection bias on vaccine effectiveness estimates from test-negative studies

IntroductionEstimates of vaccine effectiveness (VE) from test-negative studies may be subject to selection bias. In the context of influenza VE, we used simulations to identify situations in which meaningful selection bias can occur. We also analyzed observational study data for evidence of selection bias.MethodsFor the simulation study, we defined a hypothetical population whose members are at risk for acute respiratory illness (ARI) due to influenza and other pathogens. An unmeasured “healthcare seeking proclivity” affects both probability of vaccination and probability of seeking care for an ARI. We varied the direction and magnitude of these effects and identified situations where meaningful bias occurred. For the observational study, we reanalyzed data from the United States Influenza VE Network, an ongoing test-negative study. We compared “bias-naïve” VE estimates to bias-adjusted estimates, which used data from the source populations to correct for sampling bias.ResultsIn the simulation study, an unmeasured care-seeking proclivity could create selection bias if persons with influenza ARI were more (or less) likely to seek care than persons with non-influenza ARI. However, selection bias was only meaningful when rates of care seeking between influenza ARI and non-influenza ARI were very different. In the observational study, the bias-naïve VE estimate of 55% (95% CI, 47-–62%) was trivially different from the bias-adjusted VE estimate of 57% (95% CI, 49-–63%).ConclusionsIn combination, these studies suggest that while selection bias is possible in test-negative VE studies, this bias in unlikely to be meaningful under conditions likely to be encountered in practice. Researchers and public health officials can continue to rely on VE estimates from test-negative studies.     

Influenza vaccine effectiveness against medically-attended influenza illness during the 2012–2013 season in Beijing,China

Background

Influenza vaccine coverage remains low in China, and there is limited information on the preventive value of local vaccination programs.

Methods

As part of influenza virological surveillance in Beijing, China during the 2012–2013 influenza season, we assessed the vaccine effectiveness (VE) of one or more doses of trivalent inactivated influenza vaccine (IIV3) in preventing medically-attended influenza-like-illness (ILI) associated with laboratory-confirmed influenza virus infection using a test-negative case–control design. Influenza vaccination was determined based on self-report by adult patients or the parents of child patients.

Results

Of 1998 patients with ILI, 695 (35%) tested positive for influenza viruses, including 292 (42%) A(H3N2), 398 (57%) A(H1N1)pdm09, and 5 (1%) not (sub)typed influenza viruses. The rate of influenza vaccination among all patients was 4% (71/1998). Among influenza positive patients, 2% (57/1303) were vaccinated compared to 4% (14/695) among influenza negative patients, resulting in VE for one or more doses of vaccine (adjusted for age, sex, week, and days since illness onset) against all circulating influenza viruses of 52% (95% CI = 12–74%). A significant adjusted VE for one or more doses of vaccine for all ages against A(H1N1)pdm09 of 59% (95% CI, 8–82%) was observed; however, the VE against A(H3N2) was 43% (95% CI, −30% to 75%). The point estimate of VE was 59% (95% CI, 19–79%) for those aged <60 years, but a negative VE point estimate without statistical significance was observed among those aged ≥60 years.

Conclusions

IIV3 conferred moderate protection against medically-attended influenza in Beijing, China during the 2012–2013 season, especially against the A(H1N1)pdm09 strain and among those aged <60 years old.     

Bias of influenza vaccine effectiveness estimates from test-negative studies conducted during an influenza pandemic

Test-negative (TN) studies have become the most widely used study design for the estimation of influenza vaccine effectiveness (VE) and are easily incorporated into existing influenza surveillance networks. We seek to determine the bias of TN-based VE estimates during a pandemic using a dynamic probability model. The model is used to evaluate and compare the bias of VE estimates under various sources of bias when vaccination occurs after the beginning of an outbreak, such as during a pandemic. The model includes two covariates (health status and health awareness), which may affect the probabilities of vaccination, developing influenza and non-influenza acute respiratory illness (ARI), and seeking medical care. Specifically, we evaluate the bias of VE estimates when (1) vaccination affects the probability of developing a non-influenza ARI; (2) vaccination affects the probability of seeking medical care; (3) a covariate (e.g. health status) is related to both the probabilities of vaccination and developing an ARI; and (4) a covariate (e.g. health awareness) is related to both the probabilities of vaccination and of seeking medical care. We considered two outcomes against which the vaccine is supposed to protect: symptomatic influenza and medically-attended influenza.When vaccination begins during an outbreak, we found that the effect of delayed onset of vaccination is unpredictable. VE estimates from TN studies were biased regardless of the source of bias present. However, if the core assumption of the TN design is satisfied, that is, if vaccination does not affect the probability of non-influenza ARI, then TN-based VE estimates against medically-attended influenza will only suffer from small (<0.05) to moderate bias (≥0.05 and <0.10). These results suggest that if sources of bias listed above are ruled out, TN studies are a valid study design for the estimation of VE during a pandemic.     

Seasonal influenza vaccine effectiveness against influenza in 2012–2013: A hospital-based case-control study in Lithuania

Background

Due to scarce information on seasonal influenza vaccine effectiveness (SIVE) against severe clinical influenza outcomes in risk populations, we conducted a case-control study to assess its effects against laboratory-confirmed influenza in hospitalized patients during the 2012–2013 influenza season.

Methods

We conducted a test-negative case-control study among ≥18 years old patients with influenza-like illness (ILI) hospitalized in two Lithuanian hospitals. Cases were influenza A(H1N1), A(H3) or influenza B positive by RT-PCR, and controls were influenza negative. Additional demographic and clinical data to assess the role of confounding were collected. SIVE and its confidence intervals (95% CI) were estimated by using multivariate logistic regression as (1 − OR) × 100%.

Results

The sample consisted of 185 subjects. Seasonal influenza vaccine uptake was 5%. Among 111 (60%) influenza positive cases, 24.3% were A(H1N1), 10.8% were A(H3) and 24.3% were influenza B cases. Unadjusted SIVE was 79% (95% CI −6% to 96%) and after the adjustment it increased to 86% (95% CI 19% to 97%).

Conclusions

Seasonal influenza vaccination in 2012–2013 was associated with reduced occurrence of laboratory-confirmed influenza, but due to low sample size the estimate of SIVE is imprecise. Given high prevalence of influenza in hospitalized ILI cases and low influenza vaccination coverage, there is a need to increase influenza vaccination rates.     

Effect of influenza vaccination on hospitalizations in persons aged 50 years and older

Objective

To estimate influenza vaccine effectiveness (VE) in preventing hospitalizations in persons over 50 years of age.

Design

We performed a retrospective, population based study, using a “difference-in-differences” approach to determine the association between hospitalization and prior vaccination. We examined this association when influenza was not circulating and compared it to the association found when influenza was circulating. VE was estimated from the difference in the association between hospitalization and prior vaccination, inside vs. outside influenza seasons.

Setting

Kaiser Permanente in Northern California.

Patients

Health plan members aged 50 years and older during the September 1997 to August 2008 study period, when there were about 68,000 pneumonia hospitalizations in 10 million person-years.

Results

Vaccination was associated with lower risk of hospitalization for pneumonia and influenza, even before flu season, presumably due to unmeasured confounders. When influenza arrived the hospitalization-vaccination association strengthened, yielding an adjusted VE estimate of 12.4% (95% CI: 1.6–22.0) in persons aged 50–64, and 8.5% (95% CI: 3.3–13.5) in those aged 65 years and older. There was no significant effect on hospitalizations for ischemic heart disease (IHD), congestive heart failure (CHF), cerebrovascular disease (CVD), or trauma.

Conclusions

Influenza vaccination has a modest but significant effect on prevention of hospitalization for pneumonia and influenza in persons 50 years of age and older.     

Estimates of influenza vaccine effectiveness in primary care in Scotland vary with clinical or laboratory endpoint and method—Experience across the 2010/11 season

Aim

This study examines estimation of seasonal influenza vaccine effectiveness (VE) for a cohort of patients attending general practice in Scotland in 2010/11. The study focuses on the variation in estimation of VE for both virological and clinical consultation outcomes and understanding the dependency on date of analysis during the season, methodological approach and the effect of use of a propensity score model.

Methods

For the clinical outcomes, three methodological approaches were considered; adjusted Poisson multi-level modelling splitting consultations in vaccinated individuals into those before and after vaccination, adjusted Cox proportional hazards modelling and finally the screening method. For the virological outcome, the test-negative case–control study design was employed.

Results

VE was highest for the most specific outcomes of ILI (Poisson end-of-season VE = 47% (95% CI: −69%, 83%); Cox VE = 34% (95% CI: −64%, 73.2%); Screening VE = 52.8% (95% CI: 3.8%, 76.8%)) and a virological diagnosis (VE = 54% (95% CI: −37%, 85%)). Using the Cox approach, adjusted for propensity score only gave VE = 46.5% (95% CI: −30.4%, 78.0%).

Conclusion

Our approach illustrated the ability to achieve relatively consistent estimates of seasonal influenza VE using both specific and less specific outcomes. Construction of a propensity score and use for bias adjustment increased the estimate of ILI VE estimated from the Cox model and made estimates more similar to the Poisson approach, which models differences in consultation behaviour of vaccinated individuals more inherently in its structure. VE estimation for the same data was found to vary by methodology which should be noted when comparing results from different studies and countries.     

On the bias of estimates of influenza vaccine effectiveness from test–negative studies

Estimates of the effectiveness of influenza vaccines are commonly obtained from a test-negative design (TND) study, where cases and controls are patients seeking care for an acute respiratory illness who test positive and negative, respectively, for influenza infection. Vaccine effectiveness (VE) estimates from TND studies are usually interpreted as vaccine effectiveness against medically-attended influenza (MAI). However, it is also important to estimate VE against any influenza illness (symptomatic influenza (SI)) as individuals with SI are still a public health burden even if they do not seek medical care. We present a numerical method to evaluate the bias of TND-based estimates of influenza VE with respect to MAI and SI. We consider two sources of bias: (a) confounding bias due to a (possibly unobserved) covariate that is associated with both vaccination and the probability of the outcome of interest and (b) bias resulting from the effect of vaccination on the probability of seeking care. Our results indicate that (a) VE estimates may suffer from substantial confounding bias when a confounder has a different effect on the probabilities of influenza and non-influenza ARI, and (b) when vaccination reduces the probability of seeking care against influenza ARI, then estimates of VE against MAI may be unbiased while estimates of VE against SI may be have a substantial positive bias.     

Influenza vaccination and risk of hospitalization among adults with laboratory confirmed influenza illness

Background

Influenza vaccine is moderately effective for preventing influenza illness. It is not known if vaccination reduces the risk of subsequent hospital admission among patients with vaccine failure and laboratory confirmed influenza illness.

Methods

Patients in a community cohort presenting with acute respiratory illness were prospectively enrolled and tested for influenza during 8 seasons to estimate seasonal vaccine effectiveness. Hospital admissions within 14 days after illness onset were identified for all participants aged ≥20 years with laboratory confirmed influenza. The association between vaccination and hospital admission was examined in a propensity score adjusted logistic regression model. The model was validated by examining the association between vaccination and hospital admission in participants without influenza.

Results

Influenza was identified in 1393 (28%) of 4996 participants. Sixty-two (6%) of 1020 with influenza A and 17 (5%) of 369 with influenza B were hospitalized. Vaccination was not associated with a reduced risk of hospital admission among all participants with influenza [adjusted odds ratio (aOR) = 1.08; 95% CI: 0.62, 1.88]; or among those with influenza A (aOR = 1.35; 95% CI: 0.71, 2.57) or influenza B (aOR = 0.67; 95% CI: 0.21, 2.15). Influenza vaccination was not associated with hospitalization after non-influenza respiratory illness (aOR = 1.14; 95% CI: 0.84, 1.54).

Conclusions

Influenza vaccination did not reduce the risk of subsequent hospital admission among patients with vaccine failure. These findings do not support the hypothesis that vaccination mitigates influenza illness severity.     

The effectiveness of influenza vaccination in preventing hospitalizations in children in Hong Kong, 2009–2013

Background

Influenza vaccination is widely recommended every year to protect individuals against influenza virus infection and illness. There are few published estimates of influenza vaccine effectiveness against hospitalization in children or from subtropical regions.

Methods

We conducted a test-negative year-round study between October 2009 and September 2013, recruiting children 6 months to 17 years of age admitted to two hospitals in Hong Kong with a febrile acute respiratory infection. Cases were tested for influenza A and B and conditional logistic regression was used to estimate vaccine effectiveness comparing influenza vaccination history of the trivalent influenza vaccine (TIV) among patients testing positive versus negative for influenza, adjusting for age and sex and matching by calendar week of recruitment.

Results

Overall vaccine effectiveness against hospitalization with laboratory-confirmed influenza A and B was estimated to be 61.7% (95% CI: 43.0%, 74.2%). The estimated vaccine effectiveness against A(H3N2) was 36.6% (95% CI: −25.5%, 67.9%) compared to 71.5% (95% CI: 39.4%, 86.6%) for A(H1N1)pdm09 and 68.8% (95% CI: 41.6%, 83.3%) for B.

Conclusions

Vaccine effectiveness against hospitalization in children varied from year to year, but was moderate to high overall even in an area with influenza activity throughout the year.     

Effectiveness of the influenza vaccination program for the elderly in Taiwan

Background and objectives

Since 1998, an influenza vaccination program has been implemented by the Taiwan government targeting people aged ≥65 years. However, the evidence of the effectiveness of this program in preventing influenza, which is based on the nation-wide database, is lacking. This study attempted to estimate the effectiveness of the influenza vaccination program in preventing influenza- and pneumonia-associated outpatient visits and hospitalization in the elderly.

Methods

Randomly sampled data of 1 million claims from the National Health Insurance Research Database compiled into seven consecutive cohorts were used to perform this analysis. Elderly claimants aged ≥65 years were included in each cohort. To decrease potential bias between vaccinated and unvaccinated subjects, the propensity score method was applied. Logistic regression and zero-inflated negative binominal regression were used to examine the effectiveness of vaccination in preventing influenza- and pneumonia-associated outpatient visits and hospitalization.

Results

A significant decrease in both the risk and frequency of hospitalization was observed in elderly people who received influenza vaccination compared with those who did not. No similar decrease was observed in the risk and frequency of outpatient visits for influenza and pneumonia.

Conclusion

Vaccination against influenza reduced hospitalization for influenza and pneumonia in elderly Taiwanese people. These results are meaningful for the promotion of vaccination policy. Annual influenza vaccination of the elderly should be encouraged.     

Acceptation de la vaccination antigrippale par les soignants au sein d’une institution gériatrique et médicosociale

Introduction

Seasonal influenza is a viral transmissible infectious disease causing increased morbidity or mortality in frail subjects, especially those living in institutions. Measures to prevent the impact of infectious diseases were proposed based on the use of influenza vaccination among health-care professionals. We wanted to evaluate the acceptance of our institutional vaccination procedure initiated in 2005 and possible improvement.

Methods

A questionnaire was sent in May 2007 to all health-care professionals (n = 730) to identify their current vaccine status in 2006 and their opinion concerning vaccination against influenza in 2007.

Results

Subsequently, 369 (50.2%) responses were obtained. Amongst those responding, 31.7% were vaccinated in 2006, 77.8% using the institutional procedure. Also, 221 (87.7%) nonvaccinated health-care professionals indicated their position concerning influenza vaccination: 37% of them would accept the vaccination in 2007 (on the condition that our current institutional procedure be improved), 63% of them would refuse influenza vaccines in spite of any improvement.

Conclusion

Our study emphasizes that the use of an adapted procedure for influenza vaccination among health-care professionals could improve vaccine coverage. It also emphasizes that a margin of those professionals are still reluctant to vaccination.     

Determinants of influenza immunization uptake in Canadian youths

Objective

To describe rate and determinants of influenza vaccination among Canadian youths.

Methods

We conducted an analysis of cross-sectional data from the Canadian Community Health Survey (CCHS) cycle 3.1 collected by Statistics Canada in 2005. This is a population-based survey collecting information pertaining to the Canadian population health status, health care utilization and health determinants. The CCHS 3.1 included 12,170 respondents age 12–17 years old who answered questions pertaining to influenza vaccination. We used multivariate logistic regression to estimate the odds of having received the influenza vaccination in the last 12 months, adjusting for potential confounders.

Results

Less than a quarter of Canadian youth reported receiving the influenza vaccination in the previous year. The most common reason for not getting the vaccination was “did not think it was necessary” (40.82%). Having chronic illness, and being an immigrant was significantly associated with a higher odds of receiving the influenza vaccination, while having an allergy and increasing frequency of alcohol drinking was associated with lower odds of receiving influenza vaccination. Smoking status acted as an effect modifier for many variables except for immigration status.

Conclusions

Influenza vaccination rate in Canadian youths is low. Judgement values on its necessity are a major factor in the decision to receive influenza vaccination. Strategies to involve youths in influenza vaccination programs and campaigns will be essential to achieve better national coverage.     

Intradermal influenza vaccination in immunocompromized patients is immunogenic and feasible

Background

Many strategies, including intradermal vaccination, have been tested to augment antibody responses upon vaccination. This strategy has not been evaluated in different groups of immunocompromized patients. We conducted a prospective, randomized study to compare the humoral response upon standard intramuscular influenza vaccination with the response upon reduced-dose intradermal vaccination in patients treated with anti-tumor necrosis factor (TNF)-alpha, human immunodeficiency virus (HIV)-infected patients, hematologic stem cell transplantation (HSCT) patients, and healthy controls.

Methods

In total 156 immunocompromized patients and 41 healthy controls were randomized to receive either 0.5 mL of the 2005/2006 trivalent influenza vaccine intramuscular or 0.1 mL intradermal. Humoral responses, determined by hemagglutination inhibition assay, were measured before and 28 days postvaccination. Geometric mean titers (GMTs) and protection rates (PRs) are reported as primary outcomes, adverse events as a secondary outcome.

Results

Reduced-dose intradermal vaccination leads to similar GMTs and PRs, within all tested groups, compared to the standard intramuscular vaccination. Healthy controls yielded significantly better GMTs and PRs than immunocompromized patients. Local skin reactions after intradermal vaccination occurred less frequent and were milder in immunocompromized patients than in healthy subjects and were predictive for a positive vaccination outcome for individual subjects.

Conclusions

Intradermal influenza vaccination is a feasible alternative for standard intramuscular vaccination in several groups of immunocompromized patients, including those treated with anti-TNF, HIV-infected patients and HSCT patients. The occurrence of a local skin reaction after intradermal vaccination is predictive of a response to at least one of the vaccine antigens.     

A(H1N1)v2009: a controlled observational prospective cohort study on vaccine safety in pregnancy

Background

A(H1N1)v2009 influenza vaccination of pregnant women was a challenge for health care providers, as little safety data were available.

Methods

We prospectively followed the pregnancies of women who were vaccinated at any time during pregnancy or ≤4 weeks prior to conception and compared these outcomes to a control cohort matched by the estimated date of birth. Primary endpoints: rate of spontaneous abortion and major malformations. Secondary endpoints: preeclampsia, gestational age at birth, and birth weight.

Results

Pregnancy outcome of 323 women immunized with adjuvanted or non-adjuvanted A(H1N1)v2009 influenza vaccines from 2009-09-28 to 2010-03-31 were compared to 1329 control subjects. The risk for spontaneous abortions (HR 0.89; 95% CI 0.36–2.19) and the rate of major malformations (all trimesters: OR 0.87; 95% CI 0.38–1.77; preconception and first trimester exposure: OR 0.79; 95% CI 0.13–2.64) did not vary between the two cohorts. Furthermore, there was no increase in preeclampsia, prematurity, and intrauterine growth retardation in the vaccinated cohort.

Conclusion

The results of our study do not indicate a risk for the pregnant woman and the developing embryo/fetus after H1N1 vaccination. We provide and apply methods novel in observational studies on pregnancy outcome, especially if a single dose exposure is investigated.