In caring for their patients, physicians strive to uphold the fundamental principle of medicine: primum non nocere – first do no harm. However, previous studies have reported that more than 80% of physicians come to work when they are ill.1^–4 Numerous infections can be transmitted nosocomially, with some of the most common being the respiratory tract infections (RTIs).5 6

To explore this phenomenon, we created and sent three versions of an online survey to third year medical students, internal medicine and surgical residents, and staff physicians from the University of Toronto between June and August 2006. The questionnaire explored the frequency of working with an RTI and the factors that influenced this behaviour.

The response rates for medical students, residents and staff physicians were 149/202 (73.8%), 317/650 (48.9%) and 202/350 (57.7%), respectively.

The vast majority of respondents were ill for 1–2 days or more. Linear regression showed that when compared with residents, staff physicians reported an average of 0.9 fewer days with an RTI (p = 0.001) and students an average of 1.3 more days (p = 0.008). The Kruskal–Wallis test (p<0.001) strongly suggested that the distribution of days with an RTI was different between the three groups.

All three groups reported a high incidence of working when ill despite acknowledging that working through an RTI posed a risk to others. When ill with an RTI, staff physicians were the most likely to work more than 80% of the time, compared to residents and medical students (60% vs 51% vs 48%, p = 0.05) and considered the risk of transmitting that infection to others to be the lowest (p<0.001).

Medical students, residents and staff physicians reported the same reasons for staying home when ill with an RTI:

1. feeling too unwell to perform clinical duties

2. concern over infecting others

3. having obvious symptoms of an RTI.

The reasons for coming to one’s clinical duties with an RTI varied markedly between the three groups. Medical students cited extrinsic factors that were dependent on the opinions and impressions of others. They also reported that a “doctor’s note” was often required by the university to explain their absence and was difficult for them to obtain. This was in contrast to staff physicians who were more influenced by intrinsic factors such as concern over the delivery of patient care. Surgeons were especially concerned about the patient impact of rescheduling procedures. Residents reported a mix of the above factors.

When asked if they believed in the legitimacy of a colleague’s sick day, medical students and residents were significantly less likely than staff physicians to agree (45% vs 49% vs 79%, p<0.001). Residents were also more likely than medical students or staff physicians to feel annoyed when a colleague was absent because of illness (22% vs 11% vs 6%, p<0.001).

Surgical residents and staff were more likely than their internal medicine counterparts to come to their clinical duties more than 80% of the time when ill with an RTI (77.5% vs 49.8%, p<0.001). Surgeons also reported a higher threshold for staying at home than internists (8.48 vs 7.55 (on a scale of 1–10 where 5 = visibly ill), p<0.001) and were less willing to cover a colleague’s workload rather than having them come to work when sick (74.7% vs 88.70%, p<0.001). Finally, surgical residents and staff were less likely to be pleased that a sick colleague stayed at home (75.90% vs 90%, p<0.001).

While most respiratory illnesses are mild, there is evidence to indicate that in susceptible populations they are associated with significant morbidity.7^–11 Additionally, there is a risk of impaired judgment, decreased level of functioning and reduced productivity when working while ill.12 13

From the results of our study, we suggest that during the first few days of an RTI, healthcare workers should try to either stay at home or take steps to minimise the risk of infecting other (wear a mask, perform frequent hand washing, avoid susceptible patients). Healthcare workers should also be educated about the sickness policies in their respective departments. Medical students should not be required to produce documentation when taking a sick day. Trainees who are not critical to the functioning of a service should be encouraged by their colleagues to go home when symptomatic from an RTI. Finally, mechanisms (coverage, patient switching) need to be developed so that residents or staff involved in performing procedures can remain home when too ill to carry out their duties.

Our results confirm the high level of physician “presenteeism” seen anecdotally and in previous studies. Further research is required to explore and address this issue, not only for common RTIs but to ensure that proper infection control practices are in place to prevent healthcare workers from acting as vectors – as happened in SARS – should another pandemic respiratory illness occur.