Limitations
As a cross-sectional online survey study, several major limitations
could have biased the findings. First, the survey invitation was
initially distributed to interventional cardiologists attending the
national cardiology conference and then circulated via several social
medias. The non-random recruitment method inherently poses a higher risk
of selection bias than a random recruitment method. However, it is
expected that the selection bias would be minimal as the distribution of
surveyed physicians reveals great regional variance among participants.
Second, the absence of a mandatory request for physicians who read the
invitation to complete this survey could associate the study with
response bias. Third, responses to questions that require participants
to recall memory or estimate (e.g., ICP workload, sick leave days due to
pain in the last five years) could generate information bias. Lastly,
the comorbidities information collected from the survey could be
substantially under-reported as the participants were not required to
provide medical evidence for their comorbidities.
However, the reported pain-related
OHHs should be reliable as pain is a sensory symptom.