Limitations
Several factors must be understood when interpreting the results of this
study. While the largest study included 168 patients the majority had
less than 15 subjects. Additionally, all but one of these studies were
retrospective reviews. There was also heterogeneity among the studies
with some measuring a more obese and medically complex patients,
multilevel airway surgery, and PSG results as secondary outcomes. This
was accounted for in our statistical process described above. Within
each study, demographic data and multi-level surgical targets were not
always clearly linked to PSG data, presenting a source of confounding
bias and potential selection bias. Finally, postoperative PSG data was
collected on average 2.1 months postoperatively which may not be enough
time to permit healing leading to possible under or over-estimation of
OSA.2