Thirty-three studies (n=11,147) were included in this review: 3 randomised controlled trials (RCTs; n=240), 3 controlled clinical trials (CCTs; n=2,699) and 27 case series reports (n=8,208).
Symptomatic improvement (5 RCTs/CCTs).
Three trials reported greater symptomatic improvement for FESS techniques, with the improvement ranging from 78 to 85% compared with 43 to 72% for comparative techniques. Two of these studies displayed statistically significant differences. A further trial reported no difference between the techniques (82%). Another trial reported that radical nasalisation showed greater improvement than conventional ethmoidectomy (41% versus 8%); the differences were statistically significant.
Symptomatic improvement (case series).
For participants with nasal polyps, symptomatic improvement ranged from 37 to 99% (n=10). In the mixed patient group (with and without nasal polyps), overall symptomatic improvement ranged from 40 to 98% (n=17).
Recurrence of polyps or disease (2 RCTs/CCTs).
In the one trial that reported disease recurrence, the rates were not significantly different between FESS and Caldwell-Luc techniques (8% versus 14%). A different trial reported no significant differences in polyp recurrence between endoscopic ethmoidectomy and polypectomy (28% versus 35%).
Recurrence of polyps or disease (case series).
Amongst patients with polyps, recurrence ranged from 8 to 66% (n=13). Amongst those with mixed disease, recurrence ranged from 4 to 33% (n=6).
Residual disease.
RCTs/CCTs: no comparative study reported on this outcome.
Case series: residual disease ranged from 12 to 75% of sides that were operated on (n=4).
Revision surgery.
In the RCTs/CCTs, neither of the two trials reporting revision surgery as an outcome detected statistically significant differences in rates between the groups. In the case series, revision ranged from 6 to 42% amongst patients with polyps (n=4), and from 3 to 9% amongst those with mixed disease (n=9).
Complications (3 RCTs/CCTs).
One trial comparing FESS with Caldwell-Luc reported no major complications in either group. Another trial reported a rate of total complications of 1.4% for FESS techniques compared with 0.8% for conventional procedures. A trial comparing functional ethmoidectomy with radical nasalisation reported total complications of 7.7% for the latter technique, but none for the former.
Complications (case series).
The total complications ranged from 0.3 to 22.4% (n=23).
Validity.
The authors highlighted the low generalisability of results to the UK setting, the non-independent assessment of outcomes, and the variable follow-up of all the included studies, all of which threaten external validity.
In addition, the internal validity of the included RCTs was reduced by limited study power (n=2), inadequate randomisation (n=3), baseline differences between the groups (n=2), variation in the intervention of interest (n=3) and loss to follow-up (n=3). Similarly, the CCTs were subject to non-random allocation (n=3), baseline differences between the groups (n=3), variation in the intervention of interest (n=3) and loss to follow-up (n=2). Finally, the authors pointed out the main threats to the internal validity of the case series studies: the lack of a control group, susceptibility to selection bias, variation in the intervention applied and loss to follow-up (n=15).