Pandemic influenza vaccination: Will the health care system take its own medicine?

Authors


Dr Craig D. Ferguson, School of Medicine & Dentistry, James Cook University, Townsville, Queensland, 4811, Australia. Email: craig.ferguson@jcu.edu.au

Abstract

Objective: To identify factors associated with acceptance of the pandemic Influenza A H1N1 California 7/09 vaccine (pH1N1 vaccine) in rural hospital staff.

Design: Cross-sectional group administered survey.

Setting: A 58-bed rural Queensland hospital, over a 10-day period starting five days after commencement of the pH1N1 vaccination program.

Participants: Hospital staff and students.

Main outcome measures: Intention to receive pandemic influenza vaccination, seasonal influenza vaccination.

Results: The survey was completed by 252 staff and students (full response rate 76.1%, cooperation rate 98.4%). The majority of respondents had already decided whether or not to receive the pH1N1 vaccine: 111 (44.0%) intended to be vaccinated, 63 (25.0%) intended not to be vaccinated and 78 (31.0%) were undecided. Of 125 staff who had received the seasonal influenza vaccine, 86 (68.8%) intended to receive the pandemic vaccine. Respondents were less likely to state acceptance of the pH1N1 vaccine if they identified concern about vaccine adverse reactions (odds ratio (OR) 0.29; 95% confidence interval (CI), 0.11–0.78), vaccine safety (OR 0.20; 95% CI, 0.09–0.46), vaccine efficacy (OR 0.20; 95% CI, 0.04–0.87) or perceived limitations of vaccine trials (OR 0.08; 95% CI, 0.01–0.99) as primary reasons discouraging pandemic influenza vaccination.

Conclusions: There is considerable concern among rural hospital staff regarding pH1N1 vaccine adverse reactions, safety, efficacy and the adequacy of clinical trials on its use. Education specifically addressing these concerns might increase pandemic vaccine acceptability.

Introduction

The Australian Government has responded to the first global influenza pandemic in 41 years with a free, population-wide vaccination program.1 A novel influenza A (H1N1) virus emerged in Mexico in March 2009 and following rapid global spread was recognised as a pandemic influenza strain (pH1N1) by the World Health Organization on 11 June 2009.2 The pandemic Influenza A H1N1 California 7/09 vaccine (pH1N1 vaccine) (CSL Biotherapies, Parkville, Australia) has been produced3 and approved by the Therapeutic Goods Administration on 18 September 2009.4 As of 26 February 2010, 191 deaths in Australia had been associated with pH1N1, with a disproportionately high mortality rate in Australian Indigenous people.5

A combination of hospital and community control measures, vaccination and therapeutic antiviral treatment is optimal to reduce the burden of pandemic influenza.6 However, there are concerns about the acceptability of pandemic influenza vaccination among health care workers, which has implications for the health of the employee, their patients, their families and supply of the health care workforce.7,8

Acceptance of seasonal influenza vaccine among hospital staff in Australia is low9,10 despite guidelines recommending vaccination to reduce staff illness, absenteeism, virus transmission and patient morbidity and mortality.7,11 An online survey of staff in the Gold Coast Health Service District to determine expected pH1N1 vaccine uptake was conducted in the fortnight before vaccine availability (Dr Brian Bell, Gold Coast Health Service District, Queensland Health, pers. comm., 30 October 2009). The survey found 48.1% of respondents planned to be vaccinated, 28.0% were undecided and 23.9% planned not to be vaccinated; however, less than 14% of eligible staff completed the survey. The attitudes of rural hospital staff towards pandemic and seasonal influenza vaccines have not been explored.

Staff attitudes towards influenza vaccination are the most significant barrier to increasing vaccine acceptance.12 Education that targets identified objections are likely to be more effective than generic educational programs.12 Therefore, the primary aim of this study was to assess the association of factors likely to motivate or discourage pH1N1 vaccination with the stated intention of rural hospital staff to receive the pH1N1 vaccine.

What is already known on this subject:

  • • Hospital staff vaccination is required for an optimal response to pandemic influenza.
  • • Acceptance of seasonal influenza vaccination by the Australian health care workforce is low.
  • • Staff attitudes and beliefs are significant barriers to the uptake of seasonal influenza vaccination; however, these have not been explored in rural hospital staff, or in relation to pandemic influenza vaccines.

What this study adds:

  • • This study identifies staff attitudes correlated with intention to receive the pandemic H1N1 and seasonal influenza vaccine in a rural hospital (Rural, Remote Metropolitan Areas classification 5).
  • • These results are useful for developing targeted education addressing key concerns to improve staff acceptance of influenza vaccination, which is likely to be more effective than generic education.

Methods

Source population

We surveyed staff and students working in a rural (Rural, Remote Metropolitan Areas classification 5) Queensland hospital between 5 and 15 October 2009. Data collection began five days after pH1N1 vaccines became available to staff. Staff and students were eligible to receive pandemic and seasonal influenza vaccination free of charge. Number of staff and students in the hospital during data collection was calculated from payroll data and student registrations.

Survey instrument

The study design was an opportunistic cross-sectional survey administered in groups of one to three volunteers who completed the survey independently. A single page survey was adapted from an instrument developed by Clark et al., which explored knowledge, attitudes and beliefs regarding the seasonal influenza vaccine among registered nurses.13 Respondents gave written informed consent before completing the survey. Demographic data collected included age (years), sex, Indigenous status, staff designation and hospital department. Survey items consisted of closed-ended questions including receipt of seasonal influenza vaccine in 2009, receipt of the pandemic influenza vaccine, intention to receive the pandemic influenza vaccine, interaction with hospital inpatients and factors previously identified by Clark et al. as variables associated with acceptance of the seasonal influenza vaccine.13 A medical condition with an indication for influenza vaccination was defined as a condition predisposing to severe influenza including pregnancy.11 All respondents were instructed to select the primary reasons that motivated and discouraged them to receive pH1N1 vaccine irrespective of whether they did or did not intend to receive the vaccine. A free text response was solicited for any unlisted motivating and discouraging factors. CF and TF (authors) regularly attended hospital departments at shift changes during the data collection phase to invite participation from staff, which included administrative, allied health, operational staff, doctors, nurses and students. Non-shift staff were surveyed opportunistically. The study was approved by the Cairns & Hinterland Health Service District Human Research Ethics Committee (HREC/09/QCH/119 – 631 LR).

Statistical analyses

Solicited survey items were coded and entered into a spreadsheet. Unsolicited factors cited by three or more participants were also coded. Data were imported into SPSS 17.0 (SPSS Inc., Chicago, IL, USA) for statistical analysis. Staff characteristics were compared by χ2- and Mann–Whitney U-tests. Age, sex and designation of staff who responded were compared with all hospital staff. We compared the characteristics of staff with a stated intention to receive the pH1N1 vaccine with those who were either undecided or did not intend to be vaccinated. Factors significantly different between the groups in univariate analysis were entered into a logistic regression to determine the relationship between motivating and discouraging factors and intention to receive the pH1N1 vaccine. Participant age was scaled to aid comparison of odds ratios (ORs). Alpha of 0.05 was set for statistical tests.

Results

An estimated 331 staff and students worked in the hospital during data collection. The survey was completed by 252 of 256 staff and students invited to participate (full response rate 76.1%, cooperation rate 98.4%). There was no significant difference of age, sex or designation between respondents and all hospital staff (data not shown). Approximately half of respondents had received the seasonal influenza vaccine in 2009 (n = 125, 49.6%). Uptake of seasonal influenza vaccine by administrative, allied health, operational staff, doctors, nurses and students was 64.3%, 60.0%, 34.6%, 54.5%, 49.6% and 25.0%, respectively. Most respondents had decided whether or not to receive the pH1N1 vaccine. 111 (44.0%) intended to be vaccinated, 63 (25.0%) intended not to be vaccinated and 78 (31.0%) were undecided. A definite intention to receive the pH1N1 vaccine was recorded for 68.8% of staff who had received the seasonal influenza vaccine compared with 19.7% of staff who did not receive the seasonal influenza vaccine (P < 0.001).

Table 1 shows the characteristics of participants grouped by intention to receive the pH1N1 vaccine. Those intending to be vaccinated were older, more likely to have received the seasonal influenza vaccine and have a medical condition for which influenza vaccination is advised. Primary motivators that were cited significantly more frequently by participants intending to receive the pH1N1 vaccine included protecting self or others from infection, and ease of access to the vaccine. Factors less likely to be cited by participants intending to receive the pH1N1 vaccine included concern about vaccine safety or adverse effects, vaccine efficacy, limitations of vaccine trials, fear of needles or associated pain, low chance of contracting influenza, history of contracting influenza in 2009 and limited contact with high-risk patients.

Table 1. Participant characteristics compared by intention to receive pH1N1 vaccine
 DefinitelyUndecided or definitely notP value
n = 111 (44.0%)n = 141 (56.0%)
  1. Nominal variables n (%); †Mann–Whitney U-test, other characteristics compared by χ2-test;n = 105 and§n = 134 valid responses;¶unsolicited responses cited by three or more respondents.

Age, median [interquartile range]48 [42–55]43.5 [35–52]§0.003
Men24 (21.6)21 (14.9)0.223
Interact with inpatients88 (79.3)118 (83.7)0.462
Received seasonal influenza vaccine86 (77.5)39 (27.7)<0.001
Medical condition with indication for vaccine25 (22.5)12 (8.5)0.003
Staff designation   
 Administrative8 (7.2)20 (14.2)0.122
 Allied health18 (16.2)12 (8.5)0.093
 Doctors19 (17.1)14 (9.9)0.136
 Nurses44 (39.6)71 (50.4)0.117
 Operational staff12 (10.8)14 (9.9)0.984
 Students10 (9.0)10 (7.1)0.746
Primary factors influencing decision to receive the H1N1 vaccine   
 Protect myself from influenza92 (82.9)44 (31.2)<0.001
 Protect patients from influenza56 (50.5)30 (21.3)<0.001
 Protect family from influenza52 (46.8)42 (29.8)0.008
 Previous severe influenza infection10 (9.0)5 (3.5)0.121
 Member of target group for vaccination27 (24.3)11 (7.8)0.001
 Ease of access to vaccination30 (27.0)16 (11.3)0.002
 Workplace expectation30 (27.0)24 (17.0)0.077
 Severity of H1N1 influenza strain18 (16.2)16 (11.3)0.349
 Risk of adverse reactions14 (12.6)58 (41.1)<0.001
 Concern about vaccine safety18 (16.2)86 (61.0)<0.001
 Vaccine not effective enough4 (3.6)21 (14.9)0.006
 Limitations or insufficiency of vaccine trials1 (0.9)14 (9.9)0.006
 Fear of needles/pain0 (0)8 (5.7)0.029
 Contraindicated0 (0)5 (3.5)0.121
 Low chance of contracting influenza2 (1.8)20 (14.2)0.001
 Limited contact with high-risk patients0 (0)12 (8.5)0.004
 Already contracted influenza this year2 (1.8)22 (15.6)<0.001
 Too busy6 (5.4)8 (5.7)1.000

Age (OR per 10 years 0.62; 95% confidence interval (CI), 0.44–0.90), but not presence of a medical condition with an indication for influenza vaccination, was associated with intention to receive the pH1N1 vaccine in the multivariate logistic regression analysis (Table 2). Protecting self (OR 4.72; 95% CI, 1.96–11.40) or patients (OR 2.61; 95% CI, 1.00–6.81) from influenza infection or being a member of a target group for vaccination (OR 3.98; 95% CI, 1.15–13.81) were primary motivators associated with intention to receive the pH1N1 vaccine. Discouraging factors independently correlated with reduced likelihood of intention to receive the pH1N1 vaccine included concern regarding risk of adverse reactions, vaccine safety, vaccine efficacy, limitations of vaccine trials and history of influenza infection in 2009 (Table 2).

Table 2. Independent correlates of survey responses with intention to receive pH1N1 vaccine and receipt of seasonal influenza vaccine (n = 239)
 Intention to receive pH1N1 vaccineReceipt of seasonal influenza vaccine
Odds ratio95% CIOdds ratio95% CI
LowerUpperLowerUpper
  • *

    P < 0.05.

  • Respondents with no specified age were excluded from the analysis (n = 13).

  • ‡No upper bound.

Age (per 10 years)0.62*0.440.900.72*0.550.94
Medical condition with indication for vaccine0.830.242.860.25*0.090.73
Protect myself from influenza4.72*1.9611.403.38*1.627.06
Protect patients from influenza2.61*1.006.811.540.733.23
Protect family from influenza1.010.412.480.810.401.65
Member of target group for vaccination3.98*1.1513.811.190.453.18
Ease of access to vaccination1.310.463.741.160.492.77
Risk of adverse reactions0.29*0.110.780.41*0.200.84
Concern about vaccine safety0.20*0.090.460.970.491.95
Vaccine not effective enough0.20*0.040.870.630.211.90
Limitations or insufficiency of vaccine trials0.08*0.010.990.270.051.37
Fear of needles/pain0.000.001.270.246.60
Low chance of contracting influenza0.200.031.180.810.262.54
Limited contact with high-risk patients0.000.000.000.00
Already contracted influenza this year0.16*0.030.840.730.262.01

In the same logistic regression model, with receipt of seasonal influenza vaccine in 2009 as the outcome, age (OR per 10 years 0.72; 95% CI, 0.55–0.94) and a medical condition with an indication for vaccination (OR 0.25; 95% CI, 0.09–0.73) were both significantly associated with seasonal influenza vaccination (Table 2). Only two factors associated with intention to receive the pandemic influenza vaccine were associated with seasonal influenza vaccination: protecting self from influenza infection (OR 3.38; 95% CI, 1.62–7.06) and risk of adverse reactions (OR 0.41 95% 0.20–0.84).

Discussion

Rural hospital staff acceptance of pandemic and seasonal influenza vaccines is within the range reported in Australian studies of seasonal influenza.9,10 Stated intention to receive the pH1N1 vaccine is lower than acceptance of the seasonal influenza vaccine in this cohort. Participants who received seasonal influenza vaccination were more likely to accept pH1N1 vaccination. However, one in five staff intending to receive the pH1N1 vaccine had not been vaccinated against seasonal influenza. Furthermore, several survey items correlated with intention to receive the pH1N1 vaccine in the multivariate analysis were not factors in relation to seasonal influenza vaccination. Together these results suggest that respondents perceive the pH1N1 vaccine differently to the seasonal influenza vaccine.

A substantial proportion of staff had not decided whether to receive pandemic influenza vaccination. Although compulsory vaccination policies have been applied successfully for some diseases, such as hepatitis B, requiring staff to undertake medical interventions raises several ethical problems: first, it denies employees' right to autonomy regarding personal health care decisions; second, compulsion is generally reserved unless there is immediate and significant danger to others, which is difficult to demonstrate in relation to influenza vaccination; finally, the intended beneficiary of employment-related influenza vaccination is the patient and individuals are rarely compelled to undergo medical treatment primarily for the benefit of others.14 For these reasons, other means of improving compliance with vaccination guidelines are preferable.14

Appropriately targeted education might motivate staff to accept pH1N1 vaccination. Prioritising factors that correlate with pH1N1 vaccine acceptance, as compared with a generic education program, is more likely to improve compliance.12,15 From our findings, the positive points required in staff education include evidence that influenza vaccines reduce influenza infection in staff and patients7 and the established immunogenicity of the pH1N1 vaccine to benchmarks used to evaluate seasonal influenza vaccines.3 This information addresses three survey items independently correlated with intention to receive pandemic influenza vaccination.

Staff with a history of influenza-like illness (ILI) in the last year thought they might have immunity to pH1N1 and declined vaccination. However, without positive laboratory confirmation, staff with a history of ILI should be encouraged to re-consider vaccination.

Regarding vaccine safety, we interpreted risk of adverse reactions as referring to short-term vaccine effects and concern about vaccine safety referring to more serious and delayed-onset effects. We accept that these survey items are ambiguous and might not have been interpreted similarly by all respondents. Both items were negatively associated with a decision to receive the pH1N1 vaccine, which suggests that addressing concerns about safety and adverse events should be a high priority in staff education.

In relation to short-term adverse reactions, results from a pH1N1 vaccine trial demonstrated that incidence and severity of adverse events were comparable with the seasonal influenza vaccine. Local adverse events were common; however, most were rated as mild by the respondents. No serious adverse reactions were reported.3

Concerns regarding long-term adverse events are more difficult to address since useful influenza vaccines have an incompletely described safety profile.16 However, the historical safety profile of influenza vaccination is robust and the pH1N1 vaccine has been produced using proven techniques.3 A unique vaccine safety concern related to the pH1N1 strain is the association of Guillain–Barré syndrome with the influenza A(H1N1)/New Jersey 1976 vaccine.17 The attributable risk for Guillain–Barré syndrome within six weeks of receiving this vaccine was 8.8 cases per million vaccinations.17 Studies of influenza vaccines since have demonstrated low rates of serious long-term adverse events and evidence suggests that influenza vaccination is a protective factor for Guillain–Barré syndrome.18 Post-licensure surveillance for the pH1N1 vaccine will be required to assess this outcome in vaccinated and unvaccinated populations.3

Addressing staff concerns might also be critical for ensuring adequate uptake of future pandemic vaccines and the trivalent 2010 seasonal influenza vaccine, which includes the pH1N1 strain.19 Promoting vaccines to employees is often allocated full-time staff resources in tertiary hospitals; however, this is often not available in smaller centres. The findings of this study are therefore useful to direct the finite staff education resources available in rural hospitals. Promotional campaigns that specifically incorporate health care worker concerns should be used and their efficacy measured.

Author contributions

CF, study conception, research design, ethics application, survey administration, data analysis, data interpretation, manuscript draft and final edit of manuscript; TF, study conception, ethics application, survey administration, data analysis, data interpretation, manuscript draft and final edit of manuscript; JG, research design, data interpretation, manuscript draft and final edit of manuscript; WM, research design, ethics application, data interpretation, manuscript draft and final edit of manuscript.

Ancillary