University of Alberta
Influenza vaccine effectiveness – Patients with diabetes
Colquhoun, 1997 (ALL AGES)
- Estimate vaccine effectiveness in reducing hospital admissions for diabetic patients during periods when vaccine and wild strains were antigenically close.
- Case-control study.
- Data source
o Leicestershire Diabetes Register, which was specific for IDDM for the first influenza season studied (this bias should be similar in both cases and controls).
o Influenza epidemics = 1989-90, and 1993.
o Cases = admitted to hospital with ICD-9 codes for respiratory conditions and acute sequalae of diabetes. Exclusions of non-influenza etiologies (e.g.: adenovirus) where determined.
o Two controls per case were selected randomly from the Diabetes Register.
- Data collection
o GP notes of cases and controls
o Age, sex, duration of diabetes, other chronic medical disorders, number of previous GP consultations, influenza vaccination during the 3 years before the epidemic.
- 80 cases identified, but only 37 available for study.
- 166 controls selected, but only 77 available for study.
- Non-participation due mainly to lack of GP consent.
- Cases and controls similar WRT demographics co-morbidities, and HC utilization.
- Modal age band was 45-64, modal duration of diabetes = = 65 years benefit and should be targeted for vaccination, efforts should be renewed to ensure vaccination among those with high-risk conditions.
Heymann, 2004 (ELDERLY ONLY)
- Estimate the effectiveness of influenza vaccination of community-dwelling, diabetic, elderly individuals.
- Retrospective cohort study.
- Data sources
o Databases of Maccabi Healthcare Services.
o Patients at least 65 years of age on October 1, 2000.
o Reference group was low-risk = absence of heart disease, lung disease, diabetes or endocrine disorders, renal disease, stroke or dementia, vasculitis, rheumatologic disease, and cancer.
- Vaccination status available from the database.
- Outcome = any hospitalization or death.
- Influenza season: October through February of 2000-2001.
- Negative control: June through September of 2000.
- 15556 (48.8% vaccinated) patients with diabetes, 69097 (42.0% vaccinated) patients in the reference group.
- Results
o Vaccine protective, NS for differences between diabetic and reference group
o Diabetic patients: 8.29% vs 9.44%, OR = 0.87 [0.77, 0.97], VE = 13%.
o Reference patients: 6.99% vs 9.08%, OR = 0.77 [0.75, 0.80], VE = 23%.
o No effect in either group during the summer control period. Diabetic patients had 10% more events during this period than reference patients.
o Vaccination significantly associated with reduction in hospitalization rates in all age and sex categories examined, except a few diabetes substrata. Likewise mortality rate of older women with diabetes not significantly improved.
- Weaknesses: PPV may have been more common in reference patients. Diabetic influenza signal may have been cancelled out by generally higher event rates in diabetic patients. Cause-of-hospitalization or death not available. Types of diabetes mixed. No adjustment for comorbidity.
- This study supports the use of annual influenza vaccination in elderly patients, including those with diabetes.
Looijmans-Van Den Akker, 2006 (RESULTS PROVIDED FOR ELDERLY AND NON-ELDERLY)
- Determine the effectiveness of influenza vaccination in reducing occurrence of hospitalization and death from any cause in adults with diabetes during an influenza epidemic.
- Nested case control study – prospective data collection.
o Cases were hospitalizations identified from a variety of ICD-codes, including MI, CHF, stroke, and diabetes dysregulation, all-cause deaths.
o Controls were randomly selected in a roughly 1:4 ratio.
- Data source
o Primary care-based Prevention of Influenza, Surveillance and Management (PRISMA) study cohort, including 90 general practices from 1999 to 2002.
o Prospective recording of data in routine medical practice.
o 1999-2000 influenza season included.
o 75235 patients under observation during the 1999-2000 A(H3N1) epidemic were considered. 9238 patients with diabetes were eligible.
- 192 cases and 1561 controls were included. Vaccinated control subjects were older, more likely to have chronic heart or lung disease, took more medication in the previous year.
- Results
o VE adjusted for age, sex, health care insurance, presence of heart or lung disease, or other high-risk disease, and number of meds and GP visits in the previous year.
▪ Hospitalization VE = 54% [26, 71]
▪ Death from any case VE = 58% [13. 80]
▪ Hospitalization or death VE = 56% [36, 70]
o Adjusted VEs were higher for patients aged 18-64, which may explain much lower findings in Heymann compared with Colquhoun and the present study.
▪ Hospitalization VE = 70% [39, 85]
▪ Death VE = 24% [-706, 93]
▪ Hospitalization or death VE = 72% [46, 85]
o Adjusted VEs, patients aged >= 65 – non-significant prevention of hospitalizations.
▪ Hospitalization VE = 14% [-88, 60]
▪ Death VE = 56% [4, 80]
▪ Hospitalization or death VE = 39% [-5, 65]
- NNTs – Calculated by Darren – using Table 3 – crude incidence rates of total complications in the source population of 9238 patients with diabetes.
o Aged >= 65 years – 22.3 / 1000 – 29.9 / 1000 = AR = -7.6 / 1000. NNT = 132.
o Aged 18-64 – 14.0/1000 – 28.3/1000 = AR = -14.3/1000. NNT = 70.
- Adjustment in adults aged 18-64 appeared to increase VE estimates.
- IDDM mixed with NIDDM. Specificity of outcomes for influenza unclear. No non-epidemic control period.
- Most hospitalizations due to diabetes dysregulation.
- Patients with type 2 diabetes, like other high-risk individuals, benefit from annual influenza vaccination regardless of age.
Influenza vaccine effectiveness – Primary observational studies showing healthy user bias
Jackson LA, Nelson JC, Benson P, Neuzil KM, Reid RJ, Psaty BM, Heckbert SR, Larson EB, Weiss NS. Functional status is a confounder of the association of influenza vaccine and risk of all cause mortality in seniors. International Journal of Epidemiology, 35(2): 2006
- Objective: Explore the hypothesis that functional status and disease severity are confounders of the association of influenza vaccination and risk of death.
- Nested case-control study.
- Data source
o Group Health Cooperative databases. Cohort = subjects aged >= 65 on January 1, 1998.
o Community-dwelling.
o Cases = all subjects who died in Jan-Mar 1998.
o Controls – 2 or 3 controls per case, matched by age and sex, randomly selected – risk set sampling.
- The time period chosen was an influenza season of low vaccine matching – so the correct VE should be quite low.
- Comorbidity classification
o Group 1 – ICD-9-CM groupings
o Group 2 – Sub-groupings of group 1 variables for severity. E.g.: No cancer, non-serious cancer, serious or metastatic cancer; no diabetes, diabetes without complications, diabetes with complications.
o Group 3 – Functional status indicators – MRR – dementia, non-home residence, independent ambulation, assistance needed for bathing.
- 252 cases, 576 controls.
- Results
o Unadjusted OR = 0.59 [0.41, 0.83]
o Adjustment for Group 1 – OR = 0.45 [0.30, 0.68]
▪ Comorbid disease associated with both mortality and vaccination.
o However, within comorbidity severity sub-strata - Group 2 variables
▪ More severe disease associated with higher risk of death
▪ E.g.: severe cancer – 11-fold increased risk of death – no difference in vaccination vs no cancer
▪ Less severe strata were associated with increased vaccination, but only small or no differences in the risk of death.
▪ E.g.: Non-serious cancer – 3 times higher risk of vaccination – no difference in risk of death.
o Adjustment for Group 2 – OR = 0.51 [0.33, 0.78]
1. Residual confounding among those with illness – those assigned a chronic renal disease code – proportion with creatinine value >= 2.4 mg/dl – 85% vs 37%, cases vs controls.
2. Residual confounding among those without illness – presence of functional impairments – 73% vs 21%.
▪ Cases had more severe disease, less functional capacity, not captured by diagnostic codes.
o Adjustment for Group 3 – functional status – OR = 0.71 [0.47, 1.06]
o Restriction to create more homogeneous subgroups.
▪ ORs much closer to the null
▪ E.g.: Subjects who met the criteria for at least one of the group 1 variables – adjusted for functional status – OR = 0.83 [0.48, 1.41].
▪ Subjects with a serum creatinine value recorded – OR = 0.93 [0.53, 1.64].
▪ Subjects with an optometry department visit – OR = 0.86 [0.35, 2.12].
- Comorbidity classifications that do not distinguish severity jointly classify as diseased both persons with codes associated with an increased likelihood of vaccination and persons with codes associated with an increased risk of death.
- The combined group that appears to be at increased risk of death also appears more likely to be vaccinated – individually, though the opposite may be true, that those at most increased risk of death do not receive vaccine.
- An ecologic fallacy is committed.
- Residual confounding is also likely to be present in illness severity and functional status.
- Results suggest that functional status limitations identified by chart review are important confounders of the association of influenza vaccination and risk of death in seniors.
- Diagnostic codes incompletely address, if not worsen, confounding.
o Absence of diagnostic code not sensitive for absence of illness. Misclassification is differential.
o Heterogeneity in disease status among persons classified by broad groupings of diagnosis codes.
- Restriction produced estimates much closer to the null.
- Results suggest that further research is needed on methods to reduce selection bias in observational studies of influenza vaccine effectiveness.
Jackson LA, Jackson ML, Nelson JC, Neuzil KM, Weiss NS. Evidence of bias in estimates of influenza vaccine effectiveness in seniors. International Journal of Epidemiology, 35(2): 2006.
- Hypothesis – the magnitude of the underlying differences that predispose to death and hospitalization may diminish over time – depletion of the susceptible – groups become more similar.
- Objective – Evaluate the possible influence of bias. Evaluate ability of covariates defined by diagnosis codes and indicators of medical utilization to remove selection bias.
- Retrospective cohort study
- Data source
o Group Health Cooperative
o Subjects aged >= 65 years each September 1, entered every September 1.
o Cohort entry September 1995 to September 2002.
o Follow-up to August 31, 2003.
- Outcomes
o All cause mortality, P&I hospitalization
o Secondary – hospitalization – stroke, ischemic heart disease, CHF, and injury or trauma.
- Covariates – TVC for vaccination status, ICD diagnoses as secondary analyses.
- Time periods
o Influenza season – first and last weeks with >= 50 influenza positive isolates.
o Pre-season, post-influenza – to May 31, summer – to August 31.
- 75527 subjects – 338264 person-years of observation – 8 year study period.
- Vaccination coverage – 68 to 74%.
- Results
o RR – increased progressively in pre-influenza, influenza, and post-influenza time periods.
o Mortality, age-sex adjusted.
▪ Before influenza –RR = 0.39 [0.33, 0.47]
▪ Influenza – RR = 0.56 [0.52, 0.61]
▪ Post influenza – RR = 0.74 [0.67, 0.80]
o P&I hospitalization, age-sex adjusted.
▪ Before influenza – RR = 0.72 [0.59, 0.89]
▪ Influenza – RR = 0.82 [0.75, 0.89]
▪ After influenza – RR = 0.95 [0.85, 1.07]
o Injury or trauma hospitalization, age-sex adjusted
▪ Before influenza – RR = 0.67 [0.55, 0.82]
▪ During influenza – RR = 0.88 [0.79, 0.96]
▪ After influenza – RR = 0.85 [0.77, 0.94]
o Adjustment for disease covariates brought estimates away from the null, but did not materially change results.
o Similar trends were observed within influenza season.
o Robust to analyses restricted to individual study years.
- Influenza season estimates were consistent with those of Nichol et al for the 1999/2000 influenza season in three large HMOs.
- However, greatest reductions in risk occurred during the before influenza period. This study is novel because most previous studies examined a post-influenza control period.
- The pre-influenza RR could not have been due to vaccination.
- The magnitude of the bias was sufficient to account entirely for the associations observed during influenza season.
- Interpretation of a before influenza period is needed to appropriately interpret the relative risk observed in influenza and post-influenza periods.
- Finding does not mean that there is no effect of vaccination. Assuming that vaccination reduces fatal influenza infections by 58%, and that such deaths account for 10% of all influenza season deaths, the RR for all cause mortality expected in the absence of bias is 0.94. It simply means that all-cause mortality may be too non-specific an outcome to observe. The true effect of vaccination on many events may be negligible, but they are easily affected by differences in underlying health.
Jackson ML, Nelson JC, Weiss NS, Neuzil KM, Barlow W, Jackson LA. Influenza vaccination and risk of community-acquired pneumonia in immunocompetent elderly people: a population-based nested case-control study. Lancet, 372(9636): 2008.
- Estimate the effectiveness of influenza vaccine in preventing CAP in the elderly.
- Population-based, nested case-control study.
- Data source
o Group Health, 2000 – 2002 – 3 influenza season with good matching.
o Elderly subjects as of September 1.
o Community-dwelling, immunocompetent.
o Cases = validated episode of CAP, with radiographic evidence.
o Controls = 2 per case, risk set sampling.
- Covariates – comorbidities, severity indicators, functional status, utilization of health services, smoking, prescription data.
- Influenza season defined as before.
- Analysis
o Conditional logistic regression.
o Covariates added to achieve OR = 1.0 during pre-influenza periods – test adequacy of adjustment factors.
- 1173 cases, 2346 matched controls. Control vaccination rate = 78%.
- Results
o Pre-influenza
▪ Age and sex-adjusted OR = 0.60 [0.38, 0.95]
▪ Adjustment for age, sex, asthma, smoking history, antibiotics prescribed for lower respiratory tract conditions, abnormal FEV1, use of home oxygen, previous pneumonia episode, use of corticosteroids or bronchodilators, statins, or antipsychotic drugs, and any visit to an optometrist or ophthalmologist – OR = 1.01 [0.58, 1.76], p = 0.98.
o Influenza season
▪ Age and sex-adjusted – OR = 1.04 [0.88, 1.22]
▪ Fully adjusted – OR = 0.92 [0.77, 1.10]
o Restriction to hospital cases only
▪ Pre-season fully adjusted OR = 0.99
▪ Influenza season fully adjusted OR = 0.85 [0.62, 1.15].
▪ Peak season cases only – OR = 1.41 [0.35, 3.02]
o By conventional variable selection
▪ Pre-season OR = 0.87
▪ Influenza season OR = 1.07 [0.83, 1.38].
- Vaccination was not associated with a significant reduction in the risk of CAP in elderly individuals.
- Outcome makes a difference – outcome misclassification may render the analysis more susceptible to bias, since the majority of events, on which vaccination may have negligible effect, may nonetheless be easily affected by differences in health status.
- Cannot just restrict to inpatient admissions because these may be sicker and less likely to be vaccinated – over-estimate vaccine prevalence among controls.
- Ineffectiveness – because influenza only causes a small proportion of influenza, or because vaccine is inefficacious?
Eurich DT, Marrie TJ, Johnstone J, Majumdar SR. Mortality reduction with influenza vaccine in patients with pneumonia outside “flu” season: Pleiotropic benefits or residual confounding? American Journal of Respiratory and Critical Care Medicine, 178(5): 2008.
- Test the hypothesis that a mortality benefit of influenza vaccine is apparent during the influenza off-season because of healthy user bias.
- Reasons for suspecting previous observational estimates of vaccine effectiveness
o No trial data supporting such a mortality benefit.
o Off-season benefit observed by some studies
o No commensurate improvement in hospitalizations and mortality during a period of improving vaccination rates – in fact, both admission rates and mortality for CAP have been increasing.
o Studies better able to adjust for health and functional status or other measure of frailty demonstrate attenuated or abolished effects.
- Healthy user = a patient who, despite the presence of various coexisting conditions, is relatively healthier and has a predilection for better lifestyle behaviors, more health seeking and preventive activities, better adherence to medical advice and therapies, and greater likelihood of vaccination.
- Cohort study
- Data source
o Prospective CAP cohort in Capital Health.
o Patients aged 17+, hospitalized during influenza off-seasons in 2000-2002.
- Data collection
o Chart review
o Patient/proxy interview (functional status)
o Post codes for SES
o Lab measures
o PSI
- Outcomes
o Primary outcome = mortality.
o Secondary = ICU, ICU or mortality.
- Sample composed by 1:1 matching on propensity score including all variables indicative of a potential health user effect.
o PSI itself not included in PS – can’t affect vaccination decision.
o PPV – unable to include in PS – correlate with influenza vaccination.
o These variables were included in the fully adjusted model.
- 1813 patients, 352 vaccinated and 352 unvaccinated selected by propensity score matching. Median age 78 years, 29% with severe (PSI V) CAP. 12% mortality overall.
- Results
o Mortality
▪ Unadjusted 8% vs 15%, OR = 0.49 [0.30, 0.79], VE = 51%.
▪ Adjustment for age, sex, and then typical database adjustments increased the estimate of VE – OR = 0.45 [0.27, 0.76], VE = 55%.
▪ However, inclusion of laboratory and clinical data brought it back to OR = 0.52 [0.30, 0.90], VE = 48%.
▪ Full adjustment including functional status, advance directive, PPV, SES: OR = 0.81 [0.35, 1.85], VE = 19%, NS.
o ICU and ICU or death
▪ Sequential adjustment decreased estimate of VE, but all ORs were still significant.
▪ ICU – fully adjusted OR = 0.17 [0.04, 0.71]
▪ ICU or death – fully adjusted OR = 0.50 [0.25, 1.00].
o Sensitivity analyses
▪ Results robust to tightening up the off-season definition. No difference in effect before or after influenza season. Analysis based on entire cohort unchanged.
▪ Thought experiment: residual 19% VE could be abolished by postulating an unmeasured confounder in 10% of controls increasing risk of death 3-fold, or in 20% of controls increasing risk of death 2-fold. Some of the currently measured frailty variables are consistent with these parameters in the overall cohort – unmeasured confounders similar to these are plausible.
- Large mortality benefit in patients who received influenza vaccination before hospitalization for pneumonia observed, even though it is extremely unlikely that they had an influenza related illness.
- Results most consistent with residual and difficult-to-correct confounding.
- Previous studies – no off-season control, no frailty or healthy user variables.
- Results demonstrate empirically that the mortality benefits of influenza vaccination may have been largely overestimated – striking and highly significant associated attenuated and rendered NS after accounting for disease severity and functional status.
- Alternatively, vaccination may reduce severity of illness by augmenting innate immunity, or by preventing influenza, shifting the microbial spectrum toward less virulent pathogens.
Simonsen L, Viboud C, Taylor RJ, Miller MA, Jackson L. Influenza vaccination and mortality benefits: new insights, new opportunities. Vaccine, 27(45): 2009.
- Influenza and the sea of death
o Data from Thompson et al. estimate that 5% of the approximately 600000 senior deaths that occur in winter months is influenza-attributable mortality. This is the mortality that influenza vaccines can be expected to prevent.
o The notion that influenza vaccination may prevent 50% of all cause mortality is untenable, because most deaths do not involve influenza.
- Lack of evidence from clinical trials
o Single placebo-controlled clinical trial in Dutch seniors – VE is approximately 50% of laboratory-confirmed influenza infections among healthy people.
o No RCT evidence for mortality prevention.
- Evidence of selection bias in cohort studies
o Weird results, including increased benefit for older vs younger seniors in some studies.
o Jackson et al. – Cohort study – VE = 50% - but was at its highest in the period immediately after the vaccination season in Autumn – months before the influenza period – no evidence that vaccine prevented more deaths in the influenza period than in surrounding time periods.
o Jackson et al. – Nested case-control study – Routine database adjustments were counter-productive. Residual confounding demonstrated within strata of regular adjustment variables. Fuller adjustment led to attenuation of VE.
o Main source of bias is likely ta small subset of frail and terminally illseniors who are less likely to become vaccinated during the preceding autumn months because of their deteriorating health.
o Mis-measurement magnified by the use of non-specific endpoints – e.g.: all-cause mortality – vaccine likely to have negligible effect on most events – but health status differences likely to have a major impact.
- Modeling of this frail group
o Assume VE = 0%.
o Assume 5% of the population had a 20-fold higher risk of dying in any month starting in the Fall, and 0.5-fold smaller chance of being vaccinated.
o The VE benefits of Jackson et al. were reproduced remarkably well.
o These VE estimates, then, are the results of unrealized heterogeneity in the study population – depletion of the susceptible.
- Moving forward – research
o Laboratory-confirmed endpoints
o Better measurement of confounders by chart review – adjusting the VE in the pre-influenza period to 0%, NS.
o Use of pre-influenza season negative control
o Adjust VE by negative control VE
▪ Use RSV-confirmed hospital events
▪ Use a difference-in-differences approach
o RCTs – more likely to happen for new vaccination vs standard vaccinations, than for placebo controls.
- Moving forward – intervention
o Improve vaccine formulation – e.g.: adjuvanted vaccines, increased antigen dosing.
o More aggressive use of antivirals
o Interrupting transmission – e.g.: herd immunity, vaccination of household contacts
Fireman B, Lee J, Lewis N, Bembom O, Van Der Laan M, Baxter R. Influenza vaccination and mortality: Differentiating vaccine effects from bias. American Journal of Epidemiology, 170(5): 2009.
- Estimate the effect of flu shots on mortality using case-centered logistic regression and a differences-in-differences approach that obviates the need to adjust estimates using imperfect covariate measures of “nearness to death”.
- Data source – Kaiser Permanente in Northern California – 9 flu years from September 15, 1996 to September 14, 2005 – those >= 65 years at the start of the flu year.
- Covariates – age sex, DxCG insurance risk score predictive of costs, self-reproted heatlth status.
- Influenza season – 10% isolates positive. Every calendar day except Jan. 20-23 fell outside of flue season in at least 2 of 9 years studied.
- Outcome = all-cause mortality.
- Analysis
o Case centered logistic regression – focuses on cases (Deaths).
o Model of exposure odds among cases as a function of expected exposure odds in the absence of the case (estimated from those in the risk set age- sex- stratum at the calendar day of the case) modeled as an off-set term, with additional terms for a constant and for covariate adjustment.
o Risk set and case strata divided into those occurring during a flu season, and those occurring outside of a flu season.
o Expected - observed exposure odds in absence of influenza = estimate of selection bias. Difference graphed over calendar time to examine trends in selection bias. Exp(ln oddsobserved – ln oddsexpected) = VE due to selection bias only.
o The actual estimate of VE is taken as a difference of differences = the O-E (flu season) minus the O-E (off-season) for the same calendar days – implemented using explicit polynomial terms for the number of days since September 15.
- Results
o Predictors of vaccination – followed a curvilinear trend, with increasing, then decreasing vaccination rates across the continuum of most predictors
▪ Age – peak at approx. 80 years.
▪ Insurance risk score – peak in the 80th-90th percentile.
▪ Predicted probability of death – peak at 3.0-7.4%, but p(vaccination) fell below 50% in those with p(death) > 30%.
▪ Self-reported health status, on a E VG G F P, peaked at “good”.
o “VE” in the absence of influenza (O-E)
▪ O < E.
▪ “VE” descended from 0.80 to approximately 0.25 from October to September – behavior of selection-bias over time.
o Inclusion of flu season indicator in case-centered logistic regression, with smoothed estimation of calendar-time trajectories.
▪ True VE = 4.6% [0.7, 8.3].
▪ More effective in younger elderly (NS).
▪ More effective against mortality from cardiovascular and respiratory causes than against mortality from other causes (outcome specificity) (NS).
- A VE = 4.6%, given an excess influenza-attributable mortality of 7.8% in a population with over 60% vaccine coverage, suggests that vaccination produced a 47% reduction in the number of influenza-attributable deaths.
- NNT = 4000 per year.
- The curvilinear relationship between predictors of vaccination status and vaccination offer possible explanations for contrasting results in studies exploring potential healthy user bias – dichotomization or other simplification of data resulting in data loss.
o Patients with chronic diseases were more likely, on average, to get flu shots – however, most patients with these conditions had only a moderately elevated risk of death, often in the range where vaccine coverage was highest.
o In higher-risk patients, who drive mortality rates in the upcoming flu season, the propensity to obtain flu shots was lower.
o Frail patients and their providers may “give up” – but before the onset of frailty, those with chronic conditions have greater exposure to health care and therefore vaccination opportunities.
- Rising mortality rates (that, indeed, cross although remain parallel) over time within risk strata suggest that traditional covariate adjustment or exclusion of frail individuals does not capture that which determines “nearness to death”.
- So – implement a differences-in-differences approach.
o Large study cohort, long time
o Substantial year-to-year variations in flu season timing
o Assumed that real vaccine effectiveness is negligible each year until flu season arrives.
- Case-centered logistic regression – each record in the case-centered model summarizes an entire risk set in the corresponding Cox model. Risk set-sampling is applied. The OR is the dependent variable. Reduces computational burdens.
- The authors estimated a vaccine effectiveness of 4.6% against all-cause mortality, using a novel case-centered specification of logistic regression to differentiate vaccine effects from selection bias.
Influenza vaccine effectiveness – Primary observational studies with positive results
Hak E, Nordin J, Wei F, Mullooly J, Poblete S, Strikas R, Nichol KL. Influence of high-risk medical conditions on the effectiveness of influenza vaccination among elderly members of 3 large managed-care organizations. Clin Infect Dis, 35(4): 2002
- See Nichol et al. for methods (Nichol KL, Nordin J, Mullooly J, Lask R, Fillbrandt K, Iwane M. Influenza vaccination and reduction in hospitalizations for cardiac disease and stroke among the elderly. NEJM, 348(14): 2003).
- Classification of subject into those with combined pulmonary and cardiac disease, pulmonary disease, cardiac disease, diabetes and other endocrine disorders, immunosuppression, and being a healthy elderly individual.
- Outcome = composite of hospitalization for P&I or death from any cause.
- Data analysis – adjustment for age, sex, comorbidity, previous health care utilization, and previous hospitalization for P&I.
- 1997-98 – Poor match.
- Results
o Overall combined outcome – VE = 48% [42, 52] and 31% [26, 37].
o Effective in all sub-groups. Absolute benefits varied by subgroup.
▪ VE
• All – 48% [42, 52], 31% [26, 37]
• Healthy – 46% [34, 56], 42% [28, 52]
• Diabetes – 50% [37, 60], 21% [6, 34]
• Lung – 48%, 27%
• Heart – 49%, 30%
• Heart and lung disease – 47%, 28%.
• VE roughly similar in 1996-97, but more lower and variable in 1997-98, a year of poor vaccine match.
▪ Vaccine prevented AR – healthy elderly – 3.8/1000 and 3.5/1000.
▪ Vaccine prevented AR – high-risk elderly – 18.0/1000 and 8.5/1000.
▪ NNT – 26-56 high risk, 264 healthy; 50-150 high risk, 290 healthy.
- Both healthy and high-risk elderly people derive substantial benefits from vaccination, and age-based strategies have been more effective than risk condition-based vaccination strategies.
- However, elderly people with underlying medical conditions do have significantly higher rates of hospitalization and death. Though all persons aged >= 65 years benefit and should be targeted for vaccination, efforts should be renewed to ensure vaccination among those with high-risk conditions.
Nichol KL, Nordin JD, Nelson DB, Mullooly JP, Hak E. Effectiveness of influenza vaccine in the community-dwelling elderly. NEJM, 357(14): 2007.
- Objective – analyze the effectiveness of influenza vaccination among 18 cohorts of community-dwelling elderly members of HMOs during 10 seasons – provide a long-term view of effectiveness while addressing potential bias.
- Data source
o 18 cohorts from 3 US HMOs, over 10 seasons – HealthPartners (Minnesota, Eisconsin), Kaiser Permanente Northwest (Portland, Oregon, Vancouver), and Oxford Health Plans (New York City).
o Non-institutionalized elderly (>65 years old as of October 1)
- Retrospective cohort – administrative and clinical databases.
o Covariates = Age, sex, comorbidities, previous health care use (12 months), year, site, vaccination status.
o Outcomes = hospitalization for P&I, death from any cause.
o Influenza seasons – first and last influenza isolates, last influenza isolates + 2 weeks for delayed complications.
- Analysis
o Logistic regression, propensity score adjustment based on quintile stratification.
- Negative control – risk of hospitalization during noninfluenza period – two non-influenza periods – June through September 1999 and June through September 2000.
- Sensitivity analysis with hypothetical unmeasured confounder, modeled after impaired functional status in the elderly – 0.5 risk of vaccination, but 2-3-fold increased risk of outcome, prevalence varied.
- Results
o 713872 person-seasons.
o Vaccinated subjects – slightly older, higher prevalence rates of baseline medical conditions except dementia or stroke.
o P&I hospitalizations
▪ 0.6% vs 0.7%. (vaccinated vs un-vaccinated).
▪ OR = 0.73 [0.68, 0.77], VE = 27%
▪ No difference by vaccine match to circulating strains.
o Death
▪ 1.0% vs 1.6% (vaccinated vs un-vaccinated).
▪ OR = 0.52 [0.50, 0.55], VE = 48%
▪ Poor match – OR = 0.63 [0.57, 0.69]
▪ Good match – OR = 0.48 [0.46, 0.51]
o Negative control
▪ June-September 1999 – OR = 1.0 [0.78, 1.28]
▪ June-September 2000 – OR = 0.94 [0.74, 1.19]
o Sensitivity – Most extreme scenario – at a prevalence of 60%
▪ P&I hospitalization – VE = 7%
▪ Death – VE = 33%
- Large number of subjects, seasons, and sites, to account for year-to-year variability.
- National estimates of influenza-attributable mortality has not declined to the expected degree in light of increasing vaccination rates (Simonsen L, Reichert TA, Viboud C, Blackwelder WC, Taylor RJ, Miller MA. Arch Intern Med, 165: 2005) – but nation-level data do not include vaccination status or risk profile – confounded – wide range of geographic diversity – aging of the population – increase in chronic diseases – increased hospitalization for P&I in elderly – ecologic fallacy.
- Limitations – misclassification of vaccination status? Probably towards the null.
- Hospitalizations and deaths will be prevented if we succeed in increasing vaccination rates.
Kelly H, Carville K, Grant K, Jacoby P, Tran T, Barr I. Estimation of influenza vaccine effectiveness from routine surveillance data. PLoS ONE, 4(3): 2009.
- Use a case control design to estimate vaccine effectiveness during 5 influenza seasons between 2003 and 2007 in Victoria, Australia.
- Data source
o GP sentinel surveillance – reporting of ILI syndrome. Nose and throat swabs offered to patients presenting within 3 days of symptom onset, tested using RT-PCR. Positive isolates cultured.
o Influenza season – regular seasonal activity defined as >= 2.5 ILI cases per 1000 consultations, with epidemic influenza activity defined as >=35 ILI cases per 1000 consultations.
o Cases – those presenting with ILI and receiving a N-P swab from their physicians – laboratory-confirmation of disease.
o Controls – those presenting with ILI and receiving a N-P swab from their physicians – without a laboratory-confirmation of disease.
▪ Controls estimate the exposure distribution among those who would have attended the clinic if they had acquired influenza.
▪ Does sampling from those with ILI under- or over-estimate the source distribution of vaccination?
- Analysis
o VE = 1-OR.
o OR estimated by logistic regression for age adjustment.
- Results
o Higher than expected seasonal activity in 2003 and 2007.
o Swabs taken from approximately 40% of al ILI patients.
o Influenza detected in average of 36% of isolates.
o Vaccination status known for average 73% of patients with laboratory results. Unknown vaccination status appeared to be similarly distributed among cases and controls.
o Age- and year-adjusted VE
▪ Adjusted VE = 41% [19, 57] – however, test of homogeneity failed for age-group strata – note that, because exposure distribution varies within age strata, heavy confounding may be introduced by combining age strata without adjustment.
▪ Adults aged 20-49 – VE = 42% [13, 61]
▪ Adults aged 50-64 – VE = 57% [17 78]
▪ Adults aged >=65 – VE = 69% [8, 90]
▪ However, those aged >= 65 composed only 7% of presenting ILI.
o Unable to show any differences in VE in the years when the vaccine strains were well matched compared with years when the match was poorer. Non-significant tendency observed in some mis-matched years, but not all.
- No comorbidity data – unable to adjust for healthy users. Also, health status may explain differential VE across age groups.
- The Victoria sentinel surveillance network is able to provide estimates of influenza VE.
Skowronski DM, Masaro C, Kwindt TL, Mak A, Petric M, Li Y, Sebastian R, Chong M, Tam T, De Serres G. Estimating vaccine effectiveness against laboratory-confirmed influenza using a sentinel physician network: results from the 2005-2006 season of dual A and B vaccine mismatch in Canada. Vaccine, 25(15): 2007.
- Vaccine components selected as early as February for the influenza season that typically spans November to April in the northern hemisphere.
- From 1997-2007, 4 mis-matches in the H2N2 component
o 1997-1998 (emergence of A/Sydney/05/1997)
o 2003-04 (emergence of A/Fujian/411/2002)
o 2004-05 (emergence of A/California/7/2004)
o 2005-06 (emergence of A/Wisconsin/67/2005, WHO recommended vaccine strain was A/California/7/2004.)
- A/H1N1 was a minor contributor to influenza activity and was well-matched to vaccine over the last decade.
- Influenza B – two antigenically distinct lineages – B/Yamagata/16/88-like and B/Victoria/2/86-like viruses. B/Victoria re-appeared in 2001. The B component of the vaccine was mis-matched in 3 of 5 seasons since 2001.
- Objective: Evaluate degree of cross-protection offered by influenza vaccination during a period of vaccine mis-match to circulating influenza.
- Data sources
o BC sentinel surveillance network – Sentinel physicians collected respiratory specimens and questionnaires from patients with ILI – compensated $5.00 per specimen and questionnaire.
o Covariates – age, sex, chronic conditions, vaccination, timing of vaccination and medical visit WRT ILI onset.
- Case control study
o Patients presenting with ILI to a sentinel physician between November 1, 2005 and April 30, 2006, in whom a questionnaire and laboratory specimen was available. Cases = positive isolate. Controls = negative isolate.
o Case detection by PCR.
- Analysis – multi-variable – Cochran-Mantel-Haenszel method.
- Results
o 47% of ILI visits had a respiratory specimen collected = 442 visits included.
o 10% of ILI presented in the elderly.
o Vaccinated persons less often presented to a physician within 48 hours of ILI onset.
o Influenza diagnosed in 47% - of which 52% was influenza A and 48% was influenza B.
o VE, adjusted for age and chronic conditions.
▪ Influenza A or B – VE = 63% [15, 84], improved after sequential adjustment for age and then chronic conditions.
▪ Influenza A VE = 70% [9, 90], improved after sequential adjustment.
▪ Influenza B VE = 58% [-24, 87], fell after sequential adjustment.
▪ VE appeared much better among those without than among those with chronic conditions for influenza A (74% vs 5%, unadjusted), and overall (70% vs 18%, unadjusted).
- A sentinel surveillance system was used to derive VE estimates against laboratory-confirmed, medically attended influenza, instead of non-specific clinical outcomes.
- The vaccine effectiveness was lower than anticipated – this was among the first clues that the H3N2 strain in Canada, initially designated A/California/7/2004, was mis-matched to vaccine. The circulation of a new stain, A/Wisconsin/67/2005, which had been circulating in BC since the start of the season, was subsequently confirmed through gene sequencing, hemagluttinin inhibition, and retrospective re-testing of samples.
- Assumptions: Vaccinated and non-vaccinated persons …
o … have the same likelihood of influenza exposure.
o … present equally frequently to physician attention.
o … have the same probability of being tested by their physicians.
o … have equal viral diagnostic performance.
o … have similar susceptibility to laboratory confirmed influenza infection aside from vaccination status and other confounders for which adjustment was made
- Estimates mostly reflect the protection conferred to young healthy adults, which may mitigate the bias detected among the elderly.
Influenza vaccine efficiency / cost-effectiveness – Reviews and studies
Postma MJ, Jansema PJ, van Genugten MLL, Heijnen M-L A, Jager JC, de Jong-van den Berg LTW. Pharmacoeconomics of influenza vaccination for healthy working adults: reviewing the available evidence. Drugs, 62(7): 2002.
- Systematic review of available evidence on the pharmacoeconomics of influenza vaccination for healthy working adults.
- 11 included cost-benefit or cost-effectiveness studies.
- All studies – societal point of view – including indirect costs of production losses.
- Effectiveness data
o Observational – 5 studies
o Experimental – 2 studies (Nichol KL, Lind A, Margolis KL, et al.. NEJM, 333: 1995) (Bridges CB, Thompson WW, Meltzer ME, et al. JAMA, 284: 2000).
o Simulation – 4 studies
- Outcomes
o B/C ratio
▪ Direct benefits – e.g.: health care averted (medical), travel costs averted (non-medical).
▪ Indirect benefits – monetary valuations of vaccine-averted production losses.
▪ Direct costs – medical – purchase price, costs of administration.
▪ Indirect costs – e.g.: Treatment of vaccine adverse effects (medical), value of time investment to get vaccine (indirect), absence from work due to adverse vaccine effects (indirect).
o Vaccine effectiveness
▪ Not efficacy, i.e.: serologic.
▪ In economic evaluations of influenza vaccination, effectiveness is defined as the percentage reduction in the direct or indirect costs of influenza in healthy working adults.
▪ This is reflected by the excess relative risk percentage for a single outcome of costs are equal per episode for both vaccinated and unvaccinated individuals.
o Relative cost of the vaccine
▪ To correct for differences in relative prices and years, the per-person vaccination costs may be expressed as a percentage of daily labor costs, which reflect the most relevant financial determinant of vaccinating healthy working adults.
- Results – Components of benefit and cost
o 8 of 11 studies – included direct medical benefits. In all but one study, indirect benefits exceeded direct benefits – direct benefits reflect 0 to 52% of total benefits, with most studies at approximately 10%.
o No inclusion of care-giver time.
o No consideration of decreased productivity – presenteeism. Keech et al. suggest that productivity is reduced by 50% during the days surrounding the absence period due to influenza (Keech M, Scott AJ, Ryan PJJ. Occup Med, 48: 1998).
- Results – B/C ratios
o 3 studies – C > B, B/C < 1 – attributed to high relative costs of vaccination, low influenza incidence, or low baseline sickness-related absence days (the latter in Bridges et al.).
o 8 studies – B/C ratio approximately 2 or more (median B/C = 2.7), indicating net cost savings. Attributed to inclusion of indirect costs and benefits.
o Excluding production losses, only Nichol et al. remained cost saving, due to a relatively high attack rate and direct benefits. B/C ratio in this case ranged from 0 to 1.6, with a median value of 0.35.
- Results – effectiveness
o Effectiveness – preventing direct costs of influenza-related health care – 36% to 80%.
o Effectiveness – averting indirect costs of production losses – 33% to 80%.
- Results – costs of vaccination
o In proportion to labour costs, 2.6 to 18%.
- Results – robustness
o 3 of 11 studies with C > B were robust in sensitivity analyses presented by authors.
o Findings in the 8 other studies were also robust.
o Burckel et al. – B/C-ratio = 2.5 – Effectiveness = 80% for both direct costs and indirect costs – break even effectiveness in sensitivity analysis would be 32.5%.
- Most studies found that benefits exceeded costs – from a purely pharmacoeconomic perspective, any such program should be adopted.
- Exceptions
o Bridges et al. – may have been due to low sickness-related absence days per person with ILI, relative to other studies (see the Jefferson SR)
o Finnish study (Kumpulainen and Makela, 1997, cohort study, natural experiment) – influenza incidence low
o US study (Riddiough et al., simulation model, US OTA) – influenza incidence low.
o However, for both the Finnish and the US studies, results may be favorable in cost-effectiveness.
- Next to effectiveness, influenza incidence and number of sickness-related absence days of healthy workers appear to be key parameters.
- Do these findings apply to the elderly?
o In the elderly, the major benefit would be related to averted hospitalizations (direct medical benefits).
o For healthy working adults, cost-effectiveness depends crucially on the average salary and on sickness-related absenteeism in the absence of vaccination.
- Vaccination of healthy working adults in Western countries seems to be an intervention with favorable cost-effectiveness and cost-saving potentials if considered from the societal perspective.
- From a narrower perspective, almost all studies indicate that cost-saving potentials are lacking.
Nichol KL. The efficacy, effectiveness and cost-effectiveness of inactivated influenza virus vaccines. Vaccine, 21(16): 2003.
- Effective in the elderly
o RCT by Govaert, in the Netherolands – laboratory confirmed influenza illness – VE = 58%.
o Observational studies
o Meta-analysis – P&I hospitalizations VE = 33% [27, 38], all-cause mortality VE = 50% [45, 46].
o Observational study by Hak et al. – similar effectiveness in both healthy elderly and elderly with high-risk medical conditions.
o Meta-analysis by Gross et al. – effective among residents of LTCs (Gross PA, Hermogenes AW, Sacks HS, et al.. Ann Intern Med, 123: 1995).
▪ Acute respiratory illnesses VE = 56% [39 68]
▪ Pneumonia VE = 53% [35, 66]
▪ Hospitalizations VE = 48% [28, 65]
▪ All cuase mortality VE = 68% [56, 76]
- Economic benefits in the elderly
o See Table 2.
o 14 pharmacoecoomic studies – no RCT-based studies – all carried out before 2003 …
o Vaccination is cost-effective (4 of 13), with most (11 of 13) studies showing that vaccination is cost saving
o E.g.: US $73 per person vaccinated due to reductions in direct medical care costs (Nichol KL, Wuornema J, von Sternberg T. Arch Intern Med, 158: 1998).
o In 1 study, it was cost-effective for all elderly, and cost savings in high risk elderly. This study found positive ICERs as high as 6900 Euros per life-year gained for he low-risk elderly (Postma MJ, Bos JM, van Gennep M, et al.. Pharmacoeconomics, 16(Suppl 1): 1999).
o One study from Hong Kong found that vaccination was neither cost saving nor cost effective – CBA framework – C:B = HK$ 3.78 / 1 (Fitzner KA, Shortridge KF, McGhee SM, Hedley AJ. Health Policy, 56: 2001).
- Effective in healthy younger adults
o Lower risk for serious complications, but higher risk for work absenteeism, impaired work productivity, and interference with leisure time activities.
o Cochrane SR
▪ Serologically confirmed influenza – all seasons – VE = 65% [44, 79].
▪ Serologically confirmed influenza – match seaseons – VE = 72% [54, 83].
o Recent RCTs
▪ 80% to 90% VE against laboratory confirmed influenza during years with a good match.
▪ URTI/ILI, all causes – VE = 25-34%.
▪ URTI/ILI-associated work loss – VE = 32-53%.
▪ Health care visits – VE = 42-44%.
o US economic analyses based on RCTs – 3 US studies
▪ Cost-savings $47 / person (Nichol KL, Lind A, Margolis KL, et al.. NEJM, 333: 1995).
▪ Net costs (Bridges CB, Thopson WW, Meltzer ME, et al.. JAMA, 284: 2000).
▪ Break even cost = $43 / vaccine (Nichol KL, Mallon KP, Mendelman PM. Vaccine: 2003).
▪ At average $20/vaccine (USD) cost then savings most likely.
o US model based estimates
▪ E.g.: US OTA – no indirect cost savings in estimates – vaccination still cost effective - $278 / healthy year of life gained for persons aged 25-44 (lower limit of cost effectiveness, it appears).
- In addition to providing substantial health benefits, vaccination may also be associated with significant economic benefits, not only among the elderly, but also among healthy working adults and even children.
Nichol KL, Nordin J, Mullooly J. Influence of clinical outcome and outcome period definitions on estimates of absolute clinical and economic benefits of influenza vaccination in community dwelling elderly persons. Vaccine, 24(10): 2006.
- Objective – explore the cost-effectiveness of influenza vaccination using a model that incorporates additional cost savings for averted hospitalizations other than P&I, outside of peak influenza season, compared to a more conservative model.
- Vaccine Effectiveness
o Estimated from the 3-HMO cohort study, including effectiveness for reducing hospitalizations for P&I, heat disease, stroke, and all-cause mortality.
o Pooled 1998-99 and 1999-2000 data.
- Economic analysis
o Societal perspective
o CBA and CUA approaches taken.
o 3% discount.
o CBA
▪ Benefits = direct medical costs averted (hospitalizations prevented)
▪ Costs = direct costs of vaccination
▪ No indirect costs allowed.
▪ No value of a life lost estimated.
o CUA = net costs / number of life-years gained per person
▪ Life-years gained = death rate * mean life-years lost * effectiveness of vaccination for reducing deaths.
▪ No estimation for QoL.
o Primary model – hospitalizations for heart disease, stroke, all respiratory conditions, and all-cause deaths, throughout the influenza season.
o Secondary model – only P&I hospitalizations, and deaths occurring during the peak outbreak period.
o Data sources – Medicare (vaccination); HCUP and NIS of the AHRQ for hospitalizations.
o Probabilistic modeling with Monte Carlo simulation.
- Results
o Estimated mean net savings = $71 / person vaccinated, 90% likelihood interval [31, 117].
o Most important variables
▪ Hospitalization costs
▪ Hospitalization rate
▪ Vaccine effectiveness
o Vaccination – reduced 9 deaths per 1000 persons vaccinated.
o Estimated cost-effectiveness – direct medical care cost savings per year of life saved = $809 / life-year gained [331, 1450].
o Secondary analysis
▪ Net costs $3.50 [-11, 5]
▪ Cost-effectiveness = $-91 [-309, 126] per year of life saved.
▪ Impact of hospitalization and outcome period included affects results substantially.
- An economic analysis based only on pneumonia and influenza hospitalizations that occur during the peak of influenza season would underestimate the economic benefits of vaccination.
- Limitations
o Effectiveness estimates – potential bias – but estimates are similar to others.
o Cost savings – conservative – cost of hospitalizations only (ambulatory care omitted).
o Indirect cost savings omitted – 12.7% of those aged 65 years and older participated in the workforce in 2001.
- Influenza vaccination avoids hospitalizations and saves lives, at a lower cost than with no vaccination.
Influenza vaccine effectiveness – Systematic reviews
Jefferson T, Rivette D, Rivetti A, Rudin M, Di Pietrantonj C, Demicheli V. Efficacy and effectiveness of influenza vaccines in elderly people: a systematic review. Lancet, 366(9492): 2005.
- Systematic review, current to 2004 (Updated Cochrane version available).
- Included studies
o RCTs, cohort, case-control studies
o Patients aged >= 65 years – healthy elderly
o Exclusion – patients with particular chronic diseases
- Quality appraisal – Risk of bias used for RCTs, Newcastle-Ottawa Scales for NRS.
- Outcomes
o Laboratory-confirmed influenza (that which presents to medical attention only)
o ILI during influenza season
o Pneumonia
o Admission to hospital for influenza-associated illness or complications
o All-cause mortality
- Analysis
o Stratified by type of study, setting (community vs LTC), level of viral circulation, and vaccine matching.
o I2 for heterogeneity. Random effects model. Best-adjusted outcomes used.
- 64 studies (96 datasets) – 5 RTs, 49 cohort studies, 10 case-control studies.
- Most NRS at medium ROB, RCTs bimodally split almost evenly on either end.
- LTC – high viral circulation
o ILI – 16 datasets – 5963 observed – VE = 23% when matching was good.
o Influenza infection – 7 datasets – 1941 observed – NS
o Pneumonia – well matched – 8 datasets – 4482 observed – VE = 46%
o Hospital admission for P&I – well matched – 8 datasets – 2027 observed – VE = 45%
o Death from P&I – well matched – 16 datasets – 6127 observed – VE = 42%
o All-cause mortality – single study – 305 observed – well-matched – VE = 60%.
o NS when matching was poor or unknown. Heterogeneity high.
- LTC – low viral circulation
o Vaccines prevented ILI but not influenza. Prevention of hospital P&I admissions and P&I deaths, but few studies and high risk of bias in one heavily weighted study (Deguchi et al.).
- Community-dwelling elderly – 20 cohort studies (39 datasets) – over 3 million observations, mean follow-up of 5 months.
- Community-dwelling elderly – all elderly
o ILI, influenza, or pneumonia – Unable to stratify by vaccine matching – Inactivated influenza vaccine ineffective. Note trend to high VE (80%) for influenza – underpowered or badly designed studies.
o Hospital P&I admissions – 8 datasets – 779934 observed – VE = 26% if well-matched, also good VE = 28% if match poor or unknown.
o All-cause mortality – Unable to stratify by vaccine matching – 7 datasets – 404759 observed – VE = 42%.
- Community-dwelling elderly – high risk elderly
o 7 datasets, 6 studies.
o Only 1 outcome available – all cause mortality – 3 datasets – 68032 observed – VE = 61%.
- Community-dwelling elderly – healthy
o 7 datasets, 6 studies. Most outcomes driven by a single study.
o P&I hospital admission – 3 datasets – 101619 observed – VE = 50%.
o All-cause mortality NS, but RR = 0.65.
- Community-dwelling elderly –adjusted results – sex, age, smoking, comorbidities – 7 to 13 datasets, 667787 to 2608503 observed.
o All-cause mortality – VE = 47%
o P&I hospital admission – VE = 27%
o Respiratory disease hospital admission – VE = 22%.
o Cardiac disease hospital admissions – VE = 24%.
- Community-dwelling elderly – case-control studies –adjusted results
o P&I hospital admission – 5 datasets – 254830 observed = VE = 41%
o P&I death – 2 datasets – 251479 observed – VE = 26%
- RCTs – 5 studies – Just over 5000 observed.
o Heterogeneous in design and quality.
o ILI – 2 trials – 2047 observed – high viral circulation – VE = 43%
o Influenza – 3 trials – 2217 observed – high viral circulation – VE = 58%.
- Evidence of selection bias
o Counter-intuitive finding that vaccination is ineffective in preventing influenza, ILI, pneumonia, hospital admissions, or deaths from any respiratory disease, but are effective in preventing P&I hospitalizations and all-cause deaths (all estimates, community-dwelling elderly, all elderly).
o LTC findings suggest that vaccination may have a greater effect on influenza complications than ILI – control of disease through vaccination is a possibility.
o In the community, however, vaccine effectiveness is modest, irrespective of adjustment for systematic differences.
o Efforts should be concentrated on achieving high vaccination coverage in LTCs.
Jefferson T, Di Peitrantonj C, Rivette A, Bawazeer GA, AlAnsary LA, Ferroni E. Vaccines for preventing influenza in healthy adults. The Cochrane Library, 2010(7): 2010.
- Identify, retrieve, and assess all studies evaluating the efficacy, effectiveness, and harm of influenza vaccines in healthy adults.
- Studies
o Participants aged 16-65 years.
o Healthy
o RCTs, with observational evidence for harms.
- Outcomes
o Symptomatic influenza, ILI, hospital admissions for respiratory causes, complications, working days lost.
o Adverse events
o If trials offer a range of ILI case definitions or influenza season definitions, the more specific definitions were chosen.
- Quality appraisal – ROB, with Newcastle-Ottawa Scales for NRS.
- Data analysis
o Random effects meta-analysis.
o Stratified by type of vaccine.
- 57 datasets from 50 studies.
- Results – inactivated parenteral vaccines
o Effectiveness or efficacy in 18 datasets (12 studies), comprising 34573 participants. No formal estimation of RDs, provided here for clinical applicability.
o ILI – 10 datasets – 6984 patients – well matched – VE = 30% [17, 41]
▪ RD approximately 6% (24% vs 30%)
▪ NS when poor or unknown match.
o Influenza – 8 datasets – 11285 patients – well matched – VE = 73% [54, 84]
▪ RD approximately 3% (4% vs 1%).
▪ Poor or unknown match – 6 datasets – 10331 patients – VE = 44% [23, 59].
o Days ill – well matched – 3 datasets – 3670 patients – mean difference = -0.48 days [-0.62, -0.34]
▪ Exclusion of Nichol et al., 1999, due to reporting or the IR in that study.
▪ Only single study during poor or unknown match, NS.
o Working days lost – well-matched – 4 datasets – 4263 patients – VE = -0.21 days [-0.36, -0.05]
▪ Only a single study during poor or unknown match, p < 0.05 but VE = +0.09.
o Hospitalizations – well matched – 2 datasets – 2580 patients – NS but trend to VE = 63%.
o Pneumonia – well-matched – 1 dataset – 1402 patients – NS but trend to VE = 41%.
o Adverse effects – local tenderness (RR = 3), erythema (RR = 4), myalgia (RR = 1.5). No other systemic effects or induration or arm stiffness significantly associated.
o Serious and rare harms
▪ Oculo-respiratory syndrome – bilateral conjunctivitis, facial swelling, difficulty breathing, and chest discomfort. One RCT showed attributable risk of ORS = 2.9%, with increased risk of hoarseness and coughing as well.
▪ GBS – rapidly progressing symmetric paralysis, usually with spontaneous resolution – attributable risk – just below 1 case per 100000 vaccinations (Shoenberger 1979), or 1.6 extra cases per million vaccinations (Lasky 1998). RR = 1.8.
▪ Demyelinating diseases – 2 case-control studies – no evidence of association.
▪ Bell’s palsy – Mutsch 2004 reported OR = 84 in a case control study of intranasal inactivated virosomal influenza vaccine – withdrawn from market.
▪ Primary cardiac arrest – Siscovick 2000 – protective.
o Parenterally administered influenza vaccines appear significantly better than their comparators and can reduce the risk of developing influenza symptoms by around 4%, if the WHO recommendations are adhered to and the match is right. However, effect on clinical ILI, expressed as RD, is lower.
o No evidence that vaccines prevent viral transmission or complications (few datasets, NS, possibly under-powered).
o Inactivated influenza vaccine decreases the risks of symptomatic ILI and time off work, but their effects are minimal.
Warren-Gash C, Smeeth L, Hayward AC. Influenza as a trigger for acute myocardial infarction or death from cardiovascular disease: a systematic review. The Lancet Infectious Diseases, 9(10): 2009.
- Background
o Seasonal patterns of cardiovascular deaths similar to patters of influenza circulation.
o Clinical findings – frequent myocardial involvement (E.g.: Ison MG, Campbell V, Rembold C, Dent J, Hayden FG. Clin Infect Dis, 40: 2005) (E.g.: Geaves K, Oxford JS, Price CP, Clarke GH, Crake T. Arch Intern Med, 163: 2003).
o Markers of systemic inflammation.
o Cardiovascular disease – atherosclerosis = an inflammatory disease (Ross R. Atherosclerosis-an inflammatory disease. NEJM, 340: 1999).
o Influenza – extensive effects on inflammatory and coagulation pathways (Madjid M, Aboshady I, Awan I, Litovsky S, Casscells SW. Influenza and cardiovascular disease: is there a causal relationship? Tex heart Inst J, 31: 2004) – could lead to plaque destabilization.
- Objective: Examine the association between influenza infection and AMI or CVD death, including protection offered by influenza vaccines.
- Studies
o Include outcomes = MI or CVD death, where the latter could be more specific.
o Exposures classified as influenza, ILI, ARI, and influenza vaccination.
o Ecological, case-control, cohort, case-only studies (comparison to previous time period), and RCTs.
o Exclude – case reports, secondary reviews.
- Analysis
o Studies categorized by study design and exposure.
o RCTs meta-analysized – fixed and random-effects models.
o Study quality assessed using the STROBE statement criteria.
Warren-Gash C, Smeeth L, Hayward AC. Influenza as a trigger for acute myocardial infarction or death from cardiovascular disease: a systematic review. The Lancet Infectious Diseases, 9(10): 2009.
- 42 included studies (1932-2008)
- Results – Ecologic studies
o Association between timing of influenza circulation and CVD mortality or AMI incidence.
o 17 of 17 studies reported an increase in mortality due to CVD or incidence of AMI during times when influenza was circulating.
o Correlational studies – 8 studies
▪ Correlational studies – CVD deaths
• Outcome vs weekly or month rates of influenza circulation – 6 studies.
• Correlation coefficients rated from 0.61 to 0.77, and were 0.77 (age 45-64 years) to 0.98 (age over 75 years) in one study.
▪ Correlational studies – IHD deaths – 2 studies, “substantially” more deaths during influenza epidemic weeks.
▪ Correlational studies – AMI deaths – 2 studies – correlation coefficients were 0.38 and 0.5.
o Excess mortality studies
▪ Percentage of excess influenza deaths due to CVD – 9 studies – ranged from 18% (Collins SD, Public Health Rep, 47: 1932) to 66% (Mackenbach JP, Kunst AE, Looman CW. J Epidemiol Community Health, 46: 1992) – generally averaged around 35%-50%, but changing definitions of CVD make comparisons difficult.
o Limitations
▪ Exclusion of many studies that used overly broad definitions of CVD death (e.g.: respiratory and circulatory deaths).
▪ Ecologic fallacy – were those dying of CVD actually exposed to influenza?
▪ Confounder control limited.
▪ Induction period for delayed effects difficult to implement – is surveillance data lagged?
- Results – Observational studies – influenza infection or ARI
o ARI and AMI – 7 studies, 6 case-control
▪ Infections identified by self-reported symptoms or from family physician records.
▪ 5 of 7 studies significant associations, OR/RRs ranging from 2.1 [1.4, 3.2] (32) to 4.95 [4.43, 5.53] (37).
▪ 1 study – NS but fever associated with OR = 5.9.
▪ 1 study – NS for one visit for URTI but significant for 2 or more previous visits for URTI.
▪ Time period gradient explore in 3 studies. E.g.: Smeeth L, Thomas SL, Hall AJ, Hubbard R, Farrington P, Vallance P. NEJM, 351: 2004 – “self-controlled case series study of over 20000 patients with AMI” in the GPRD.
• IR = 4.95 for ARI within 1-3 days before AMI,
• IR = 3.2 for 4-7 days
• IR = 2.8 for 8-14 days
• IR = 1.4 for 15-28 days.
o ILI and AMI – 5 studies, 4 case-control
▪ Range of slightly positive but NS associations, to positive associations with OR = 3.8, among 4 case-control studies.
▪ Case-crossover study RR for AMI 1 day as opposed to 7 days after onset = 2.4 [1.7, 3.4].
o Influenza exposure and AMI – 5 studies, all case-control
▪ 1 of 5 studies – highly significant OR for IgG to influenza A (OR = 7.5) and B (OR = 27.3).
▪ 2 of 5 studies – NS
▪ 1 of 5 studies – No influenza antibodies detected in either group.
▪ 1 of 5 studies – pathologic study for influenza virus antigen on autopsy, AMI vs cancer deaths, no association, OR = 1.0 NS.
o Limitations
▪ Selection bias possible – hospitalized controls in 8 of 15 studies.
▪ Recall bias – 6 case-control studies relied on self-reported vaccination status or symptoms. Prospectively generated MR data (retrospectively collected) was used in 4 studies, and laboratory confirmation in 5 studies.
▪ Laboratory confirmation – serologic definition more specific than clinical ILI, but cross-sectional serology – difficult to distinguish vaccination from infection. 3 of 4 studies used paried acute and convalescent sera.
▪ Failure to control for potential confounders in 6 of 15 case control studies.
- Observational studies – influenza vaccination
o 8 studies
o Mixed results
o 3 of 8 – protective effect
o 1 of 8 – slightly protective effect NS
o 4 of 8 – no protective effect.
▪ However, 1 of 4 – underpowered
▪ Another 1 of 4 – short term effect aimed at detecting adverse effects of vaccination.
o Limitations
▪ Likely selection bias
▪ 2 of 3 cohort studies attempted to use off-season effect as a negative control.
- Interventional studies – RCT – 2 studies
o FLUVACS – 301 patients randomized – 1 year follow-up – previous Cochrane review – moderate ROB.
▪ CVD death HR = 0.34 [0.17, 0.71].
▪ But AMI HR = 0.99 [0.43, 2.32]
o FLUCAD – 658 patients randomized – median follow-up = 298 days – previous Cochrane review – low ROB.
▪ Coronary ischemic events – HR = 0.54 [0.29, 0.99]
▪ CVD death – HR = 1.06 [0.15, 7.56]
▪ MACE – HR = 0.54 [0.24, 1.21]
o Meta-analysis
▪ CVD death
• Heterogeneity, I2 = 61%, p = 0.08.
• Fixed effects model – RR = 0.39 [0.20, 0.77]
• Random effects model – RR = 0.51 [0.15, 1.76]
▪ AMI – NS either model – RR = 0.85 [0.44, 1.64]
o Limitations
▪ Small studies, few CVD events
▪ Methods for randomization and allocation concealment in FLUVACS unclear.
▪ Neither study examined influenza infection.
▪ Patients with established CVD disease.
o Despite the protective effect of influenza vaccination seen against death caused by CVD disease and some composite outcomes, overall data from these 2 RCTs were not enough to evaluate the effectiveness of influenza vaccine on CVD events (Keller T, Weeda VB, van Dongn CJ, Levi M. Influenza vaccines for preventing coronary heart disease. Cochrane Database Sys Rev, 3: 2008).
- Observational studies have tended to support the hypothesis that ARI, influenza in particular, can trigger AMI. More limited evidence for an adverse effect on CVD death. Two relatively small RCTs suggest that influenza vaccination reduces the risk of CVD death and some coronary ischemic events.
Influenza vaccine effectiveness – Randomized controlled trials
Govaert TM, Thijs CT, Masuerl N, Sprenger MJ, Dinant GJ, Knottnerus JA. The efficacy of influenza vaccination in elderly individuals. A randomized double-blind placebo-controlled trial. JAMA, 272(21): 1994.
- Estimate the efficacy of influenza vaccination in elderly individuals using clinical and serological outcome parameters.
- Single influenza season, 1991-1992. Patients selected from lists of 34 family physicians in 15 practices in the Netherlands. Patients aged 60 years or more, no high-risk indications (approx. 500 high risk patients were enrolled anyways, analyzed).
- 1838 eligible agreed to participate, LTFU approx. 3% < 10%.
- Patient-blinded placebo-randomized controlled trial. Randomization stratified by disease categories.
- Outcomes
o Serologic infection – S3 (6 months) and S2 (3 weeks) increase in titre >= 4-fold.
o Physician-reported ILI
o Patient-reported ILI – questionnaire – follow-up was 96%.
- Results
o Placebo group outcome rates
▪ Serology: 80/911 (9%)
▪ Family physician: 31/911 (3%)
▪ Sentinel station: 89/911 (10%)
▪ Self-report: 129/911 (14%)
▪ Clinical presentation with serologic confirmation: 38/911 (4.2%)
o RR
▪ Serology: 0.50 [0.35, 0.61], VE = 50%.
▪ Clinical presentation with serologic confirmation: 0.42 [0.23, 0.74], VE = 58%.
▪ Family physician: 0.53 [0.39, 0.73]
▪ Sentinel Stations: 0.69 [0.50, 0.87]
▪ Self-report: 0.83 [0.65, 1.05]
o Effectiveness similar for all sub-groups. However, sub-group confidence intervals quite wide (NS in most cases). As well, diabetes not reported separately from chronic heart and lung disease.
- Age >70 may be an effect modifier (potential decrease in effectiveness).
- The results of this study are consistent with VE = 50%.
Keitel WA, Cate TR, Couch RB, Huggins LL, Hess KR. Efficacy of repeated annual immunization with inactivated influenza virus vaccines over a five year period. Vaccine, 15(10): 1997.
- Assess the effects or repeated annual influenza vaccinations of adults.
- Volunteers
o 30-60 years old, healthy adults.
o Texas Medical Center in Houston.
o Those with no recent history of immunization and those who had received vaccine within the preceding 3 years.
o Continuous enrolment
o 1983-1987.
- Interventions
o Placebo
o First vaccination
o Multiple vaccination
- Continued vaccination in subsequent years.
- Re-randomization of placebo subjects every year, so that multiple vaccination group grew over time.
- Vaccine – WHO-compliant trivalent IM whole-virus vaccines.
- Patient and outcome assessment blinded.
- Endpoints
o Antibody response to vaccination – serology 1 month before and 1 month after intervention.
o ILI – prospectively reported, as well as retrospective reporting at the end of each season.
o Laboratory-confirmed influenza – culture, serology (1 month after ILI), serology (all patients after each flu season).
- 4233 subjects – LTFU – high – for those entering in 1983, only 62% remained at study completion. Overall follow-up rates higher, since more of those entering later completed.
- Results – Prospectively determined illness only
o Post-vaccination antibody titers similar for volunteers receiving different numbers of annual vaccinations except in a 1987-88 – small but significant reduction in titres with increasing numbers of prior vaccinations.
o Overall – illness
▪ Any illness – 30% vs 27% vs 28% (placebo, first vac, multivac)
▪ Flu-like illness – 21% vs 10% vs 11%
o Only twice was frequency of illness lower for a vaccine group
▪ Febrile illness (subset of flu-like illness) – 52.3% reduction – 1986-87 – multivac (p < 0.01)
▪ Flu-like illness – 48.7% reduction vs placebo – first vac (p < 0.05) – 1987-88.
o Sometimes illness lower in vaccination groups but NS.
o Influenza infection, laboratory confirmed – viral isolation and/or serology
▪ Placebo group – 13.8% to 23.2%, average 17.5%.
▪ 6.8% vs 6.6% vs 17.5%, first vac vs multivac vs placebo, average per year.
▪ Rates of infection significantly lower for each vaccine group than for the placebo group (p < 0.05) in 10/14 comparisons, borderline for an additional 2 comparisons.
▪ Isolation and infection rates higher or somewhat higher in the multivac than first vac groups.
▪ No significant association of influenza viral infection rates with increasing numbers of annual vaccinations on multivariate logistic regression during any epidemic period.
o Influenza virus infection-associated illness
▪ Average 30% of placebo volunteer reported a respiratory or systemic illness during each year’s flu season – average 31.6% of these were linked to laboratory-confirmed influenza.
▪ Placebo – 6.9% to 12.5%, average 9.2%.
▪ Reductions in rates for influenza vaccine groups vs placebo were significant during two seasons when vaccine matching was identical
• Any illness – VE = 50% firstvac, 44% multivac.
• Flu-like illness – VE = 48% firstvac, 50% multivac.
• Febrile illness – VE = 30% firstvac, 43% multivac.
▪ Multivac lower than placebo group more frequently than first vac – however most comparisons were NS, except in the two years of identical vaccine match (first vac significant for “any” during one year only, multivac significant for “any” during 1 year, “flu-like” during 2 years, and “febrile” during 1 year of the 2 identically matched seasons). Power issues may be engendered, especially for first vac during later years.
- Annual immunization was associated with reductions in virus shedding and infection, as well as infection-associated illness. Repeated annual vaccinations had a beneficial effect in years of identical vaccine match.
- No consistent differences in efficacy of primary vs sequential vaccinations were detected.
- Warning – methodological – vaccinations may prevent further significant rises in antibody titres leading to overestimates of VE based on serology alone.
- Findings are supportive of current recommendations for annual vaccination of adults.
Bridges CB, Thompson WW, Meltzer MI, Reeve GR, Talamonti WJ, Cox NJ, Lilac HA, Hall H, Klimov A, Fukuda K. Effectivenes and cost-benefit of influenza vaccination of healthy working adults: A randomized controlled trial. JAMA, 284(13): 2000.
- Double-blind randomized, placebo-controlled trial
o Random-numbers table
o Re-randomization in the second year of the study.
- Subjects – 18 to 64 years old, full-time employees of Ford Motor Co. in Michigan, healthy workers.
- Intervention – trivalent inactivated influenza vaccine vs sterile saline placebo.
- Data collection
o Follow-up surveys by E-mail for episodes of ILI
o End-of study questionnaire for episodes of ILI
o NP swab and viral culture for all cases of ILI – prospective notification of a study nurse required.
o Serology 3 weeks after injection and at season’s end – full serology only available for a fraction of all patients, selected randomly, due to costs.
- Outcomes
o Respiratory illness – ILI or URTI
o Associated physician visits
o Workdays
- Economic analysis
o Societal perspective and payer perspective
o Human capital approach – includes costs associated with work productivity.
o Cost of vaccination, including 15 minutes of nursing time and 30 minutes of time lost from work to attend vaccination clinic = $24.70 per person.
o Univariate sensitivity analysis – using the US average rate of $20.29 per hour, etc..
- 1184 participants randomized.
- Influenza season
o 1997-98 – antigenically distinct from the H3N2 vaccine component.
o 1998-99 – similar to the 1998-99 vaccine strains.
- Patient blinding verified by guessing assignment – findings consistent with chance, although just a bit off – authors say that they were not able to completely maintain blinding.
- Results
o Influenza illness (laboratory-confirmed, serology) (small numbers)
▪ 1997-98 – 2.2% vs 4.4% – VE = 50%, p = 0.33.
▪ 1998-99 – 1% vs 10% – VE = 86%, p = 0.001.
o Clinical illness – ILI
▪ 1997-98 – relative excess – illness 28% vs 24% = -18% (p = 0.25), days ill 2.4 vs 1.7 per person = -38% (p = 0.01), lost workdays 0.29 vs 0.20 per person = -45% (p 0.047). Results possibly skewed by single patient hospitalized for pneumonia reporting a 25 day LOS.
▪ 1998-99 – relative excess – illness 14% vs 22% = 34% (p < 0.001), days ill 1.02 vs 1.54 = 34% (p < 0.001), lost workdays 0.08 vs 0.12 per person = 32% (p < 0.001).
o Clinical illness – URTI
▪ Similarly, 1997-98 results show vaccine inefficacy – vaccine trended to be worse – but almost all findings NS except for lost workdays (0.31 vs 0.21 per person, VE = -46%, p = 0.02) and physician visit hours for patients.
▪ 1998-99 – relative excess – Positive benefits but mostly NS, except work days (barely NS, p = 0.07) 0.10 vs 0.12 per person, VE = 19%; proportion of patients with any lost work days 9% vs 12%, VE = 21%, p = 0.047.
o Economic analysis
▪ 1997-98 – societal perspective - $65.59 per person – net loss to society
▪ 1998-99 – societal perspective - $11.17 per person – net loss to society
▪ Payer perspective – not cost savings in any year.
▪ Doubling the base case rates of ILI resulted in a net societal benefit.
- Vaccination of healthy working adults provided no overall economic benefit in either year of the study.
- Vaccination had a clear health benefit when matched to circulating strains – efficacy against laboratory-confirmed influenza 86% with significant reductions in ILI, physician visits, and days lost from work.
- Some explanations for nonetheless finding that vaccination was not cost-saving
o Rates of influenza-associated severe illness and hospitalization are generally lower in healthy young adults than in elderly persons.
o Only a minority of the respiratory illnesses among adults were due to influenza.
o Approximately 1 in every 1- years there is poor antigenic match.
o Also, herd benefits and benefits to other parties not modeled.
Influenza vaccine effectiveness for cardiovascular outcomes – RCTs, and observational studies
Bainton D, Jones GR, Hole D. Influenza and ischaemic heart disease – a possible trigger for acute myocardial infarction? International Journal of Epidemiology, 7(3): 1978.
- Data source
o London – 21-year period; Six conurbanations – single epidemic from 1969-70.
o Death certification of arteriosclerotic of ischemic heart disease.
- Number of death certificates during weeks with influenza seasons compared to those without.
- Influenza season defined by deaths from influenza A
o London = 10 or more such deaths in persons aged >= 35 years constituted a flu week (20 or more after 1968 due to boundary changes).
o Criteria variable within conurbations
- Adjustment for temperature effect
o Regression – outcomes compared at a common temperature.
o Matching each flu week with a similar-temperature, or slightly colder control week
- Results – London
o Two of three broad multi-year periods – uncertain how each period defined – differences in heart disease deaths at a common temperature highly significant.
o One of three broad multi-year periods – deaths from heart disease in the flu weeks higher than in the control weeks.
o Cancer deaths – also higher during influenza weeks, but difference much smaller in absolute and proportional terms than that of heart disease.
o Matched analysis – similar results.
o Association appears largely confined to those aged 55 and over.
- Results – Six conurbanations
o Similar findings, no p-values reported, matched analysis only.
- Results show a temporal relationship between influenza and increased deaths from ischemic heart disease, allowing for differences in temperature.
Naghavi M, Barlas Z, Siadaty S, Naguib S, Madjid M, Casscells W. Association of influenza vaccination and reduced risk of recurrent myocardial infarction. Circulation, 102(25): 2000.
- Hypothesis – influenza vaccination may reduce the incidence of MI in patients with established coronary atherosclerosis.
- Case-control study
o University of Texas – Houston outpatient cardiology clinic.
o October 1, 1997 – March 31, 1998.
o Cases = Incident MI = 122
o Controls = Those who had not experienced new MI or exacerbation of their disease (hospital-based controls).
- Data collection – chart review and telephone interview. Exposure = current vaccination. Covariates included CVD RFs, multivitamin therapy, exercise, normal behaviors after URTI.
- 109 cases, 109 controls in analysis.
- Results
o Prevalence of previous (any time) vaccination 66% vs 79%, cases vs controls – proxy for health-maintaining behaviors.
o Unadjusted – Current vaccination – predictive of no subsequent MI vs new MI – OR = 21.9 [8.2, 54.3].
o Adjusted analysis – p < 0.1 univariate – forward stepwise – vaccination history (any time vaccination) (forced), HTN, hypercholesterolemia, current smoking – OR = 0.33 [0.13, 0.82], VE = 67% – associated despite including previous years vaccination as a proxy for health-maintaining behaviors.
- Unlikely that vaccination is associated with reduced risk simply because vaccinated patients are more health conscious – use of multivitamins, physical activity, and prior years vaccination NS in multiple logistic regression.
- Other potential confounders – seen as unlikely
o Medical therapy – uniform at the clinical site.
o Behavioral response to URTI may protect or predispose to MI – unclear.
o Can these account for VE = 67%?
- Nine potential mechanisms for influenza causing MI described.
- Vaccination against influenza may reduce the risk of recurrent MI.
Siscovick DS, Raghunathan ET, Lin D, Weinmann S, Arbogast P, Lemaitre RN, Psaty BM, Alexander R, Cobb LA. Influenza vaccination and the risk of primary cardiac arrest. American Journal of Epidemiology, 152(7): 2000.
- Objective – Is influenza vaccination during the previous year associated with a reduced risk of primary cardiac arrest?
- Case control study
o Cases – out-of-hospital PCA – paramedics in Washington – October 1988 – July 1994 – sudden pulse-less condition – exclusion of secondary cardiac arrest or non-cardiac causes.
o Controls – random digit dialing, matched 2:1 ratio to cases on age and sex.
- Covariates – demographics, physical parameter, comorbidities, smoking, physical activity, general health status, dietary intake of saturated and n-3 fatty acids, family history, education, employment.
- Validation of spouse reports – good, though reliability better for cases.
- Data collected from spouses for consistency with dead cases.
- Analysis – conditional logistic regression.
- Results
o 360 cases, 576 controls.
o Usually control-group correlations with vaccination status – older, comorbidities, former smokers – but less likely to be currently employed. More active, weighed less.
o Endpoint – PCA – OR – 0.51 [0.33, 0.79] – adjusted.
- For now, additional studies are needed to explore further whether the observed association of influenza vaccination with a reduced risk of PCA reflects selection or protection.
Lavallee P, Perchaud V, Gautier-Bertrand M, Grabli D, Amarenco P. Association between influenza vaccination and reduced risk of brain infarction. Stroke, 33(2): 2002.
- Case control study
o Cases – 60 years or older, consecutive patients, hospitalized for brain infarction during influenza epidemics from 1998-2000 (December 1998 – March 1999, January 2000 – March 2000).
o Controls – neighborhood controls – randomly identified – electoral rolls – 2 controls matched to cases on age, sex, and district of residency in Paris – Interviewed.
- Covariates – demographics, risk factors, CVD history, employment (proxy for SES).
- Analysis – conditional logistic regression.
- 90 patients, 180 controls
o Adjusted OR = 0.37 [0.19, 0.70], VE = 60% approximately (age, sex, diabetes, HTN, BMI, current smoking, cholesterol).
o Association stronger in those without previous history of CVD, and for those younger than 75 years (NS for those older than 75 years, but could be underpowered) (No test of interaction).
o Examination of effect within subgroups of other variables.
- Results provide indirect evidence that infection or inflammation plays a role in cardiovascular thrombotic events.
Jackson LA, Yu O, Heckbert SR, Psaty BM, Malais D, Barlow WE, Thompson WW. Influenza vaccination is not associated with a reduction in the risk of recurrent coronary events. American Journal of Epidemiology, 156(7): 2002.
- Objective – evaluate the possible role of influenza vaccination in preventing recurrent coronary events.
- Data source
o Group Health Cooperative, western Washington State.
o Members, aged >= 30 but = 50% stenosis of >=1 large coronary artery. Exclusions – severe CHF, other factors impeding follow-up.
o Primary PCI.
o Elective PCI.
o Stable angina (majority of patients)
- Intervention – single inactivated trivalent influenza vaccine IM with A/New Caledonia/20/99-like, A/Christchurch/28/03-like, and B/Jiangsu/10/03-like viruses.
o B-strain mismatched.
o Vaccine provided at discharge for patients identified under hospitalization for PCI, or during office visits to cardiologists.
- Endpoints
o Primary = CVD death
o Secondary = MACE (CVD death, MI, coronary revascularization, coronary ischemic event); coronary ischemic event = MACE or hospitalization for MI.
- Analysis – ITT
- 658 Caucasian patients aged 60 years +/- 10 years, with optimally treated CAD in a regional reference center – Follow-up median 298 days – LTFU = 0%.
- Results
o Adverse events – minor – 4% vs 3%.
o Self-reported ILI – 8% vs 13% (p = 0.042).
o Endpoint – primary – CVD death – 0.63% vs 0.76% – small cell sizes – HR = 1.06 [0.15, 7.56].
o Endpoint – secondary
▪ MACE – 3.00% vs 5.87%, HR = 0.54 [0.24, 1.21].
▪ Coronary ischemic event – 6.02% vs 9.97%, HR = 0.54 [0.29, 0.99].
o Multivariable analysis – vaccination emerged as a significant predictor of MACE.
- Results suggest a 50% reduction in coronary ischemic events, which is consistent with estimates from observational studies showing positive associations. No benefit for cardiovascular mortality was detected.
- Findings were not limited to the flu season, which ended April 30, 2005.
- Potential mechanisms
o Reduce the cardiovascular or inflammatory effects of any severe infection with fever, tachycardia, and dehydration.
o Prevent an influenza-specific inflammatory or otherwise detrimental effect on arteriosclerotic plaque stability.
o Exert an anti-inflammatory or otherwise plaque-stabilizing effect independent of influenza prophylaxis.
o Spurious effect.
- Influenza added to optimal and invasive medical care may improve the clinical course of CAD.
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related searches
- university of minnesota college of education
- university of minnesota school of social work
- wharton school of the university of pennsylvania
- cost of university of scranton
- university of minnesota school of education
- university of scranton cost of attendance
- university of south florida college of medicine
- university of minnesota masters of social work
- ecampus of university of phoenix
- university of minnesota college of continuing education
- university of illinois college of nursing
- university of north texas college of nursing