Vaccine effectiveness in pandemic influenza - primary care reporting (VIPER): an observational study to assess the effectiveness of the pandemic influenza A (H1N1) v vaccine

Notes

Permissions

Chapter 1 Introduction

In the last century, there were three pandemics (global epidemics) of influenza (1918–19, 1957–58, 1968–69), with these resulting in considerable morbidity and mortality; the numbers of deaths in these pandemics have been estimated at 20–40 million, 1 million and 1 million, respectively. The lack of herd immunity to the novel influenza viruses implicated (i.e. H1N1, H2N2 and H3N2) is believed to have been a key factor contributing to these very high numbers of deaths. 1 The influenza A subtype: H1N1 virus, which emerged in Mexico in March 2009, was subsequently declared a pandemic by the World Health Organization in June 2009. 2

In the light of data that incentivised immunisation programmes delivered in primary health-care settings being shown to be acceptable (as evidenced by high uptake rates) and effective in reducing morbidity and mortality,3 and evidence that seasonal influenza vaccination has been shown to reduce the risk of hospitalisation and death from pneumonia or influenza by 27% and 48% respectively,4 the Chief Medical Officer (CMO) for England and the Department of Health instituted a targeted vaccination programme. 5 This was mirrored in Scotland by the CMO (Scotland) and Scottish Government. Production of influenza A (H1N1)v vaccinations began soon after outbreaks in the USA and Europe, with two vaccines being adopted for the UK national immunisation programme: Pandemrix (GlaxoSmithKline), which requires one dose, and Celvapan (Baxter Healthcare), which requires two doses at least 3 weeks apart. The vaccination process in the UK began on 22 October 2009 and was initially offered to frontline health-care workers and pregnant women; those with underlying health conditions that may predispose (and in particular people with respiratory disease) (Table 1) who were at increased risk of serious illness or death from influenza-like illness were also targeted in this first phase. In December 2009, phase II of the immunisation programme targeted children aged between 6 months and 5 years to receive the vaccination.

Observational studies can be used to estimate the effectiveness of health-care interventions in situations where it is unethical and/or not feasible to mount more rigorous experimental studies, as is the case in the context of the 2009 HIN1 pandemic. 6 Building on related pilot work,3 we sought to determine influenza A (H1N1)v vaccine uptake and effectiveness for 2009 in the Scottish population using a sentinel surveillance network of general practices, the Practice Team Information (PTI) network.

Chapter 2 Methods

Overview of methods

The impact of the Scottish 2009 pandemic H1N1 vaccination programme was evaluated using a retrospective cohort design to study vaccination effectiveness (VE). This was achieved by ascertaining the uptake of the influenza A (H1N1)v vaccine by the relevant at-risk populations, i.e. patients with relevant comorbidities and pregnant women, and assessing the reduction in the expected incidence of influenza-related serious morbidity. Figure 1 gives an overview of the study design and the data sources used.

FIGURE 1.

Flow diagram for the VIPER study. HPS, Health Protection Scotland; ISD, Information Services Division; PCCIU, Primary Care Clinical Informatics Unit; PTI, Practice Team Information (network).

Chapter 3 Results

We recruited all 41 PTI practices with a combined list size of 246,368 patients. As only one practice software system, gpass, was used for the study, the general practitioner (GP) practices in this study are not representative of the spatial distribution of the population of Scotland. There was an under-representation of practices from the north-east of Scotland, in particular Orkney, Grampian and Tayside areas (where practices tend to use other GP software systems). There was also a preponderance of practices in west and central Scotland, which have higher levels of socioeconomic deprivation (Table 2).

Overall, 24.2% individuals (n = 59,721) were deemed to be in the at-risk category on the basis of existing illnesses; 4.3% (n = 2203) of 15- to 44-year-old women (n = 51,404) were found to be pregnant and 0.8% (n = 1314) of people of working age (18–64 years inclusive, n = 159,873) registered with the practices worked for the NHS (Table 2).

Chapter 4 Discussion

During the immediate period after the introduction of the influenza A (H1N1)v vaccination, more than one-third of patients registered in primary care with PTI practices and deemed to have an at-risk comorbidity were vaccinated for influenza A (H1N1)v. However, men and younger patients (outwith the 0–4 years age group) were less likely to be vaccinated. Our interim results suggest that during the study period, the vaccine seems to have been particularly effective for people with an at-risk comorbidity. This is reassuring as they were also more likely to be tested positive for having influenza A (H1N1)v (when compared with those not at risk) and to be at much higher rates of hospitalisation for severe complication of influenza. Of the patients tested using swabs after their vaccination, only one tested positive for influenza A (H1N1)v.

Our findings indicate that our estimated influenza A (H1N1)v VE is at least comparable to the seasonal influenza vaccine in preventing hospitalisation admission for: influenza and/or pneumonia (27%),16 influenza-like illness (27%) in all patients,4 acute respiratory disease and cardiovascular disease (97%) in high-risk patients,17 and medically attended acute respiratory illness in children (18% in those aged 18 months to 18 years). 18 As expected, overall uptake rates of vaccine reported in this study were similar to those reported by HPS in similar practices. 19 However, rates of vaccine uptake for health-care workers in occupational settings were lower than those reported by HPS, which relied on occupational health services reporting (as opposed to information captured by the practices). This under-reporting probably reflects a partial success in practices recording of occupational status on patients records, and therefore any inference from the occupational data should be treated with caution. Our findings that females and, outwith the youngest age group (0–4 years), older patients were more likely to be vaccinated is similar to other studies looking at uptake of vaccines by different groups. 20 This is likely to be due to greater levels of perceived susceptibility to, and perceived severity of, influenza A (H1N1)v and a greater belief in the effectiveness of recommended behaviours to protect against the disease. There is also evidence that greater levels of state anxiety and greater trust in authorities are associated with increased vaccine uptake. Our finding that the vaccine seemed to be particularly effective in patients with an at-risk comorbidity endorses the targeted approach advised by the Joint Committee on Vaccination and Immunisation and adopted by the CMOs for England, Northern Ireland, Scotland and Wales and their respective Government Health Departments/Directorates. 5

There are benefits as well as drawbacks to the evaluation of an influenza vaccine campaign based on just a single postimplementation season. Retrospective ascertainment of vaccination status is less dependable, of course, than prospective clarification, but the use of GP records is more reliable than self-reporting methods,21 as is the electronic recording of uptake rates in the sample PTI population. Also, the relatively small size of the Scottish population makes it feasible to centrally collate almost all cases of H1N1 disease, allowing for completeness of reporting. Observational studies can be used to assess the effects of health-care interventions without influencing the care that is provided or the patients who receive it;6 therefore, when used in the assessment of vaccination programmes they have high external validity and broad generalisability. However, non-randomised studies, such as the current evaluation, are limited by the extent to which there may be dissimilarities between vaccinated and non-vaccinated persons, in both their likelihood of receiving vaccination and in their subsequent care and follow-up. Our findings that at-risk patients who received the vaccine were more likely to be admitted for a trauma-associated emergency (and were possibly more likely to be frail, thus leading to negative confounding) than those who did not may mean that there is some underestimate of VE. A further assessment of the possible impact of any bias caused by the preferential receipt of vaccine by relatively healthy individuals will need to be checked, however, outside the influenza season. 22 The results from the single outcome of emergency admission for cardiovascular-related illness or the use of the composite outcome: influenza, pneumococcal disease, COPD and cardiovascular disease should also be treated with some caution as it may have led to a healthy vaccine effect. For instance, within the high-risk group those at lower risk of heart disease or those who do not smoke (and hence are less likely to have COPD) may be more likely to be vaccinated. However, it is of note that other authors have considered the role of influenza in the generation of cardiac-related conditions and debated its contribution to the excess mortality observed each winter during the annual influenza season. 23–25 The lack of a recorded vaccination amongst health-care workers (which was probably carried out in occupational settings) may have biased the estimation of VE towards zero. The fact that PTI is based on a small sample of practices in Scotland means that the data collected may be subject to fluctuations as a result of any factors that have an impact locally, such as changes to the way that PTI practices manage their services. 9 However, apart from a reduction in precision, it is unlikely that the small sample size, or other associated factors, will have a substantial impact on the overall estimates of VE. A convenience sampling approach was used to collect data on patient H1N1 virological status rather than a more systematic sampling approach. This was adopted in recognition that the implementation of a surveillance programme could not affect routine clinical practice. Conventionally, the seasonal influenza vaccine is thought to require 14 days to establish a protective effect; however, there is evidence from ongoing studies (involving HPS and the Health Protection Agency, Colindale), that 7 days is probably sufficient for the influenza A (H1N1)v vaccine. A sensitivity analysis using the 14-day cut-off period was carried out with similar estimates of vaccine effect being found (results not presented).

Chapter 5 Implications for practice

Policy-makers and clinicians should be encouraged that the VE estimates obtained are comparable to those found for seasonal influenza, which should strengthen the evidence base for health-care practitioners involved in distributing vaccine in other countries.

Influenza A (H1N1)v immunisation in primary health-care settings is both effective and widely acceptable as evidenced by high uptake rates, and should continue to be the mainstay of disease prevention for at-risk patients.

Chapter 6 Research recommendations

Study time constraints resulting from the unexpected ‘slow-burn’ nature of the second wave of influenza A (H1N1)v and the later than expected roll-out of the vaccine meant that this analysis was limited to studying only the short-term effectiveness of the influenza A (H1N1)v vaccine. A further analysis encompassing the whole influenza season is required to cover more days of vaccination exposure, which will increase precision (with resulting narrower confidence intervals). Also, for pregnant women and under-5-year-olds, a further study using a greater time period of exposure is required to calculate and present meaningful results. We were also unable, in this study, to estimate whether the vaccine was effective in reducing mortality (as only mortality data for the prevaccination period were available). A future study that repeats this data linkage and allows the calculation of longer-term VE (in reducing both morbidity and mortality) should be undertaken later in 2010.

Chapter 7 Conclusions

Evidence from swabs submitted from patients in the cohort presenting with influenza-like illness in general practice suggests that the introduction of influenza A (H1N1)v vaccine in Scotland during 2009 was associated with a high degree of protection against influenza A (H1N1)v. In addition, receipt of influenza A (H1N1)v vaccine was associated with a reduction in both admission for cardiac-related conditions and for the combined category of influenza, pneumonia, COPD and cardiac conditions. Policy-makers ought to be encouraged that the VE estimates obtained are comparable to those found for seasonal influenza, and this possibly reflects the suitability of primary care as a means of delivery. Additionally, as the first large-scale demonstration of effectiveness in a UK population, the results should strengthen the evidence base for health-care practitioners involved in distributing vaccine in other countries. Influenza A (H1N1)v immunisation in primary health-care settings is both effective and widely acceptable, as evidenced by high uptake rates, and should continue to be the mainstay of disease prevention for at-risk patients. Whether the reduced incidence of severe complications of influenza or death will persist will be apparent only when data from later in 2010 are analysed.

Acknowledgements

The authors would like to thank Dr Martin Donaghy, Dr Oliver Blatchford, Arlene Reynolds, Heather Murdoch, Eisin McDonald, Alison Potts, Ruth Thomson, Cathy Hunter, Lynne Otterson and Stuart Yule at HPS, the staff at all of the general practices and virus laboratories that contributed data to the study, and also Katie Wilde, Fiona Chaloner, Dr Margaret Watson, Carol Mackie, Alasdair Coutts, Alastair Soutar and Cristina de Cario at PCCIU, and Carole Morris and Bradley Kirby at ISD. We would also like to thank the Independent Steering Committee, which comprised Dr Neil Kelly and Dr Robert Milne.

References

1. Nicholls H. Pandemic influenza: the inside story. PLoS Biol 2006;4.
2. Chan M. World now at start of 2009 influenza pandemic. Geneva: World Health Organization; 2009.
3. Mooney JD, Weir A, McMenamin J, Ritchie LD, Macfarlane TV, Simpson CR, et al. The impact of pneumococcal vaccination in Scotland for those aged 65 and over during winter 2003/2004. BMC Infect Dis 2008;8.
4. Nichol KL, Nordin JD, Nelson DB, Mullooly JP, Hak E. Effectiveness of influenza vaccine in the community-dwelling elderly. N Engl J Med 2007;14:1373-81.
5. Donaldson LJ, Rutter PD, Ellis BM, Greaves FEC, Mytton OT, Pebody RG, et al. Mortality from pandemic A/H1N1 2009 influenza in England: public health surveillance study. BMJ 2009;339.
6. Smith S, Sinclair D, Raine R, Reeves B, Smith S, Sinclair D, et al. Health care evaluation. London: Open University Press; 2005.
7. Whitelaw FG, Nevin SL, Milne RM, Taylor RJ, Taylor MW, Watt AH. Completeness and accuracy of morbidity and repeat prescribing records held on general practice computers in Scotland. Br J Gen Pract 1996;46:181-6.
8. Ekins-Daukes S, Helms PJ, Taylor MW, Simpson CR, McLay JS. Utility of routinely acquired primary care data for epidemiology and pharmacoepidemiology. Brit J Clin Pharmacol 2005;59:684-90.
9. Information Services Division . General Practice – Practice Team Information 2008. www.isdscotland.org/pti (accessed November 2009).
10. General Register Office for Scotland . GRO Statistics Library (population Estimates) n.d. www.gro-scotland.gov.uk/statistics/publications-and-data/populationestimates/index.html (accessed July 2005).
11. Harley K, Jones C. Quality of Scottish Morbidity Record (SMR) data. Health Bull 1996;54:410-17.
12. Carstairs V, Morris R. Deprivation and health in Scotland. Aberdeen: Aberdeen University Press; 1991.
13. Fleming DM, Elliot AJ. Estimating the risk population in relation to influenza vaccination policy. Vaccine 2006;24:4378-85.
14. Graham PL, Mengersen K, Morton AP. Confidence limits for the ratio of two rates based on likelihood scores: non-iterative method. Stat Med 2003;22:2071-83.
15. Newcombe R. Confidence Intervals for Proportions and Related Quantities n.d. www.cardiff.ac.uk/medic/aboutus/departments/primarycareandpublichealth/ourresearch/resources/index.html (accessed January 2010).
16. Jefferson T, Rivetti D, Rivetti A, Rudin M, Di Pietrantonj C, Demicheli V. Efficacy and effectiveness of influenza vaccines in elderly people: a systematic review. Lancet 2005;366:1165-74.
17. Hak E, Buskens E, Nichol KL, Verheij TJM. Do recommended high-risk adults (18–64 years) benefit from a first influenza vaccination?. Vaccine 2006;24:2799-802.
18. Holloran ME, Longini IM, Gaglani MJ, Piedra PA, Chu H, Herschler GB, et al. Estimating efficacy of trivalent, cold-adapted, influenza virus vaccine (CAIV-T) against influenza A (H1N1) and B using surveillance cultures. Am J Epidemiol 2003;158:305-11.
19. Health Protection Scotland . Weekly Influenza Situation Report 2009. www.documents.hps.scot.nhs.uk/respiratory/swine-influenza/situation-reports/weekly-influenza-sitrep-2009–12–24.pdf (accessed January 2010).
20. Bish A, Michie S. Demographic and attitudinal determinants of protective behaviours during a pandemic: a review. Br J Health Psych 2010. 10.1348/135910710X485826.
21. Mangtani P, Shah A, Roberts JA. Validation of influenza and pneumococcal vaccine status in adults based on self-report. Epidemiol Infect 2007;135:139-43.
22. Jackson LA, Jackson ML, Nelson JC, Neuzil KM, Weiss NS. Evidence of bias in estimates of influenza vaccine effectiveness in seniors. Int J Epidemiol 2006;35:337-44.
23. Madjid M, Aboshady I, Awan I, Litovsky S, Ward Casscells S. Influenza and cardiovascular disease: is there a causal relationship?. Tex Heart Inst J 2004;31:4-13.
24. Mamas MA, Fraser D, Neyses L. Cardiovascular manifestations associated with influenza virus infection. Int J Cardiol 2008;130:304-9.
25. Warren-Gash C, Smeeth L, Hayward A. Cardiovascular manifestations associated with influenza virus infection. Lancet Infect Dis 2009;9:601-10.

List of abbreviations

CHI

Community Health Index

CI

confidence interval

CMO

Chief Medical Officer

COPD

chronic obstructive pulmonary disease

CRH

cardiovascular-related hospitalisation

GP

general practitioner

GROS

General Register Office for Scotland

HPS

Health Protection Scotland

ISD

Information Services Division

OR

odds ratio

PCCIU

Primary Care Clinical Informatics Unit

PTI

Practice Team Information

RR

rate ratio

SMR

Scottish Morbidity Record

VE

vaccination effectiveness