In the aftermath of SARS, many issues surrounding health care worker behaviour and professionalism have been discussed [10–14], however to the best of our knowledge, only a few studies have addressed these issues in relation to the anticipated influenza pandemic . Our results suggest that most HCWs (83.3%) working in the two hospitals surveyed saw it as their professional obligation to treat sick patients and would continue working despite the potential risks. These results are in corroboration with previous studies from Hong Kong (84%) . However, they differ from other research in Australia (67%) , and from the United States (50%) . The differences may relate to non standard survey questions or a factor of time, with knowledge and intentions changing as exposure to information about PI increases. Suggested reasons for this include the fact that HCWs consider it unethical to abandon their professional responsibilities in order to protect themselves or their families [14, 19].
When participants were asked whether they consider their ward/department to be sufficiently prepared for an influenza pandemic, only 24.8% responded in the affirmative. This high lack of confidence in the department's preparation may actually stem from a real lack of preparation by the hospital or department, or may just be a result of HCWs being unaware of any planning which has been conducted. Further studies would need to be conducted to ascertain which of these two options the most likely reason is.
While many health care workers will willingly attend work during an infectious diseases emergency, history provides many stories of physicians who have avoided responsibility for treating patients . The appearance of an exotic, highly virulent disease, challenges HCWs to question their interpretation of the duty of care, in particular, its limits. This challenge was apparent both in the HIV/AIDS epidemics of the 1980s, where fear about contact with infected patients among some clinicians challenged their responsibilities to these patients and secondly in the 2003 SARs outbreaks [21–24]. There were several reports during the SARs outbreak that HCWs in Hong Kong and Toronto either avoided the physical examination of sick patients or refused to work altogether on the grounds that they presented too great a risk. In China, at the height of the SARs epidemic, at least one hospital had difficulty maintaining services due to absenteeism some of which was driven by fear of getting sick . A recent survey assessing the willingness and ability of HCWs to report to duty during catastrophic disaster in New York City, found that although more than 80% were willing and/or able to report to work for mass casualty or environmental disaster, only 57% to 68% would be willing to report to work during a severe acute respiratory syndrome (SARS) or smallpox outbreak .
Fears for personal safety and family responsibilities are commonly the main issues underlying possibly absenteeism during a pandemic – in our study the rate of absenteeism doubled in the scenario of a family member being infected. Our finding was also echoed in the New York survey, where fears and concern for the family and for themselves were the most frequently stated reasons for not being willing to report to work . Whilst in a second study, the authors reported that almost one third of the HCWs they surveyed either strongly agreed or agreed that it was professionally acceptable for HCWs to abandon their workplace during a pandemic in order to protect themselves and their families .
It is interesting to note that willingness to work varied considerably according to the individual's knowledge and their job classification. Medical and nursing staff was significantly more likely to report to work, whereas ancillary staff was unsure of their intentions during this period. A recent study by Ehrenstein et al, reported a similar finding in that the rate of administrators not willing to accept personal risk was approximately twice as high as the rate of other staff . This difference may correlate with a perception of the importance of one's role in the hospital during the response. Whilst it's important to encourage all categories of staff members to turn up for work – it must be done in an appropriate manner. Inappropriate working behaviours were identified when participants were asked to nominate their potential response in a given situation. For example, when asked whether they would attend work during a staff shortage, if they had symptoms consistent with pandemic influenza, 24% of medical staff and 26% of ancillary staff said yes. In a second scenario, participants were asked whether they would attend work if a close family member was diagnosed. This time over half of the participants stated that they would. Whilst this may not be detrimental in light of antiviral and vaccine availability, in the early stages of a pandemic, this behaviour may be linked to the spread of the disease.
Quarantine is a key public health measure in pandemic influenza plans. It is also one of the oldest methods of controlling communicable disease outbreaks. Australia's pandemic influenza plans, amongst others, emphasize the use of quarantine measures (home quarantine for up to one week) in combination with social distancing measures and antiviral medications. Our study defined quarantine as involving: (i) being forbidden to use public transport, (ii) being allowed only to travel directly from home to work and back, (iii) being isolated from other family members in their home (as was enforced in Canada during the 2003 SARS outbreak) . We found that most HCWs would comply with such quarantine measures (and adhere to antiviral medications), however, a large proportion of those surveyed stated that they would be very unhappy about it, which could ultimately affect compliance with the measures. Helping people to understand the reason for various protocols might increase their belief in their effectiveness and thus, their compliance. The use of education was previously examined in a post SARS article which examined a cohort of persons quarantined during the 2003 SARS outbreak in Canada . The authors found that compliance could have been improved by providing enhanced education and support. The authors also felt that this could have also reduced the psychological distress in the quarantined adults.
In a study by Shiao et al, factors related to nurse's consideration of leaving their jobs during the SARS outbreak in Taiwan were examined . The authors found that the main predictors were short tenure, increased work stress, perceived risk of fatality from SARS, and affected social relationships. On the contrary, belief in the effectiveness of personal protective equipment (PPE) was not an important predictive factor for nurse's consideration of leaving their job. This was because most nurses surveyed believed that protective measures at work were generally effective. While we cannot ascertain from our study whether there is a link between PPE effectiveness and work attendance, we can make postulations about the confidence the staff have in the different protective measures. For example, when respondents were asked what will protect you from getting pandemic influenza, the most common response was washing hands, wearing masks and vaccination. However, when it came to antiviral use, more respondents stated that "eating well" would offer better personal protection then the anti-viral drugs oseltamivir ("Tamiflu") or zanamvir ("Relenza") against pandemic influenza.
There are a number of limitations to this study that need to be discussed. It is important to recognize that the generalisability of our results may have been affected by the limitations inherent to any voluntary questionnaire-based cross sectional study. Given the design of the study, we relied on a convenience sample of 1200 staff members from the total number (8000) of hospital staff in both hospitals. From this number, only 894 actually returned the survey. As responses were voluntary, there may have been responder bias in the sample. We were unable to compare the demographics of the respondents versus non-respondents to examine how representative the sample actually was. This information was not available to the study authors. We can only rely on the fact that we obtained a 13.5% sample from the two hospitals. This sample is considerable large for such a study. Another limitation relates to the online response rate. The online invitation to complete the survey electronically was seen by an unknown number of staff at the paediatric hospital leading to a lack of denominator data about the 185 completing it. Nevertheless, the paper-based survey was comprehensive, yielding a large sample and acceptable response rate.
There may also be limitations with generalisability since we included only two hospitals in a single Australian city. It is also unknown as to whether responses given to the hypothetical situations posed in a questionnaire accurately reflect real-world responses of the respondents in the event of an actual influenza pandemic.
Despite these issues, the large sample size of our study, broad spectrum of HCWs represented, and representative age/gender demographics provide a general indication of what responses to a pandemic may occur and provide information on differences between health care worker groups. In addition, much research focusing on behavioural intentions indicates the potential for these to be reasonable in predicting actual behaviour .