This systematic review aimed to compare the presence of clinical features versus the polysomnography results in predicting major postoperative respiratory complications in children undergoing adenotonsillectomy for OSA. The study was prepared according to the PRISMA checklist. Overall 22 studies met the criteria, which included available clinical and PSG data and additionally specified presentation of respiratory complications. Ten of these studies had sample sizes over 100. Three of the studies were case controls, with the remainder being case series. The case mix was slightly skewed, given that all the studies included a high risk population. Twelve studies included crude data which could be pooled for individual analysis.
This showed that only 8.9% had a moderate or severe PSG as the sole predictive factor of a postoperative complication. The remaining patients had a readily identifiable clinical risk factor. Following statistical analysis, the group estimated that 192 preoperative PSG’s would need to be performed in unselected patients to identify a child with no other clinical findings that would be expected to have a major respiratory complication.
Overall this group found that major postoperative respiratory complication rate was low, even amongst high risk patients, and that the yield of using PSG for stratifying postoperative risk did not appear to be justified.