ORIGINAL ARTICLE
A randomized controlled trial of a middle meatal silastic stent for reducing adhesions and middle turbinate lateralization following endoscopic sinus surgery Chun L. Chan, BRadMedImag (Hons), BMBS1 , Behrad Elmiyeh, MBBS, FRCS (ORL-HNS)1 , Charmaine Woods, BBiotech (Hons), PhD1 , Shahid Ullah, BSc, MSc, PhD2 , Indunil Gunawardena, MBBS, FRACS1 , A. Simon Carney, FRCS, FRACS1 and Eng H. Ooi, MBBS, FRACS, PhD1
Background: Endoscopic sinus surgery (ESS) is indicated for patients with recalcitrant chronic rhinosinusitis symptoms. However, surgical revision can be required because of adhesion formation and middle turbinate lateralization. We investigate the efficacy of a middle meatal silastic stent in reducing these complications aer ESS. Methods: Thirty-six patients were randomized to receive a silastic stent in the middle meatus unilaterally aer ESS. The surgeon was blinded to the side receiving the stent until completion of the ESS. The contralateral side, with no stent, was the control side. Patients completed a 7day, postoperative, visual analog scale symptom diary and were blinded to the stent side until its removal at the first postoperative visit. Patients were followed up aer 2, 8, and 24 weeks. Endoscopic video of the sinus cavities were recorded at all visits and 2 blinded, independent ear/nose/throat (ENT) surgeons assessed the videos using a modified Lund-Kennedy scoring system. Results: Thirty-five of 36 patients completed 6-months’ follow-up. Middle turbinate lateralization was observed in
E
ndoscopic sinus surgery (ESS) is indicated for chronic rhinosinusitis (CRS) refractive to medical therapy.
1 Department
of Otolaryngology–Head and Neck Surgery, Flinders Medical Centre, Flinders University, Adelaide, Australia; 2 Flinders Centre for Epidemiology and Biostatistics, School of Medicine, Faculty of Medicine, Nursing and Health Sciences, Flinders University, Adelaide, Australia
Correspondence to: Eng H. Ooi, MBBS, FRACS, PhD, Department of Otolaryngology–Head and Neck Surgery, Flinders Medical Centre, Bedford Park, SA 5042, Australia; e-mail: [email protected] Potential conflict of interest: None provided. Received: 13 October 2014; Revised: 18 December 2014; Accepted: 23 December 2014 DOI: 10.1002/alr.21490 View this article online at wileyonlinelibrary.com.
517
International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
13 sides without a stent vs 1 side with a stent. There was a significant reduction in adhesions at weeks 2 and 8 (p < 0.001) and crusting (p < 0.01) in the stent side compared to control. Video scores at 6 months aer surgery for stent and control sides remained unchanged from the 8-week visit. There was no difference between sides for symptom scores, edema, or nasal discharge. Conclusion: Middle meatal silastic stents are well tolerated by patients and effective in reducing middle turbinate lateralization, adhesions, and crusting postoperatively in ESS. C 2015 ARS-AAOA, LLC.
Key Words: surgical procedure; endoscopic; intranasal surgery; rhinosinusitis; middle turbinates; tissue adhesions How to Cite this Article: Chan CL, Elmiyeh B, Woods C, et al. A randomized controlled trial of a middle meatal silastic stent for reducing adhesions and middle turbinate lateralization following endoscopic sinus surgery. Int Forum Allergy Rhinol. 2015;5:517–523.
However, revision rates due to surgical failure are a significant problem despite advances in operative and postoperative management, with revision rates reported between 10% and 66% in the literature.1 Postoperative adhesion rates have been reported in up to 27% of patients despite sinus irrigation and postoperative debridement being performed.2–4 Middle turbinate lateralization and formation of adhesions postoperatively have been implicated as the most common cause for suboptimal ESS outcomes requiring revision surgery.5 There have been numerous techniques reported in the literature attempting to reduce adhesion formation and middle turbinate lateralization. These include placement of absorbable and/or nonabsorbable spacers/stents, yet there is no ideal acceptable material.6 There is evidence that the use of nasal packs or topical agents can adversely affect
Chan et al.
wound healing and potentially result in excessive adhesion formation.7 Two recent double-blinded randomized controlled trials have found a sinus silastic stent to be effective in preventing adhesions; however, some patients found these stents to have an adverse effect on postoperative patient comfort.8, 9 In our department, silastic stents are selectively used in ESS as a spacer between the middle turbinate and lateral nasal wall after ESS. The technique that we employ was first described by Gall and Witterick2 in their cohort study of 500 patients, during which they found that their stent was tolerated by the majority of their patients; they reported adhesion formation in 4.2% of their cases. However, since that report was published, no randomized control trial had been performed looking at the effectiveness of the technique. The aim of this study was to perform a double-blinded, randomized controlled trial to evaluate the efficacy of a silastic stent in preventing adhesions and middle turbinate lateralization and the extent to which patients tolerated the presence of the silastic stent following ESS.
Patients and methods The Southern Adelaide Clinical Human Research Ethics Committee’s approval was obtained. All recruited patients provided written informed consent prior to surgery.
Study design We designed a double-blinded randomized controlled trial; based on our unit’s observed adhesion rate of approximately 20%, it was determined that 34 patients were required for a study to have an 80% chance of detecting a significant difference of at least 25% on a visual analog scale (VAS) in a 2-sided t test. Inclusion criteria were adult patients (ࣙ18 years of age) who had CRS ± polyps who had failed medical therapy and underwent ESS with bilateral equivalent dissections (at least middle meatal antrostomies and ethmoidectomies). Patients were excluded if they had unilateral disease, previous sinus surgeries, or any underlying medical condition that may have interfered with wound healing, such as diabetes or autoimmune diseases. Patients were also excluded from final analysis if they required middle turbinate excision during surgery and/or required nasal packing postoperatively. Computer-generated randomization (http://www. randomization.com) was employed preoperatively to determine the side of the nose that would receive the silastic stent, which is referred to as the “stent side.” The other side of the nasal cavity without a silastic stent is referred to as the “control side.” Preoperative computed tomography (CT) scans were reviewed and Lund-Mackay10 scores were calculated. All patients underwent ESS by a single surgeon (B.E.) with the technique described in the next paragraph. If sphenoidotomy was indicated, it was performed via a transethmoidal route to avoid lateralization of the middle
turbinate. Upon completion of surgery, the theatre nurse opened a sealed envelope containing the randomization of the stent side to inform the operating surgeon to place the middle meatal stent prior to extubation. Stents were fashioned from medical grade silastic, using a commercially available silicone sheet (5 × 5 cm, thickness of 1.02 mm [0.04 inches]; Medtronic, Minneapolis, MN) ensuring standard thickness and type of stent between patients. The stent was held together at each end with a silk suture, creating a slightly “sprung-open” stent to be placed in the middle meatus (Fig. 1). The stent was positioned to rest on the superior surface of the inferior turbinate to achieve self-anchoring, thus negating the need for anchoring or holding sutures. No other techniques to prevent middle turbinate lateralization were performed. A “drainage hole” was cut out of the stent superiorly to allow frontal sinus drainage. At the completion of the operation, the surgeon ensured that the septum was midline with both middle turbinates medialized. Patients were blinded to the stent side until the first postoperative visit when stents were removed in the clinic under topical anesthesia.
Postoperative care Standard postoperative treatment included irrigation with R sinus care device (ENT Technologies, Hawthorn the FLO East, VIC, Australia) 4 times a day. Antibiotics are not prescribed routinely for our patients, unless indicated by the presence of pus at the time of surgery. Patients with nasal polyposis were commenced on topical nasal steroid spray (Mometasone furoate 100 μg daily in each nostril; Nasonex; Merck, Kirkland, QC, Canada) after stent removal.
Evaluation of nasal discomfort Patients were asked to fill in a daily VAS diary for the first 7 days postoperation. The VAS was scaled from 0 to 10, with 0 indicating none and 10 indicating maximal levels of a symptom as judged by the patient, with each side assessed independently of the other side. The symptoms assessed included: discomfort/pain, bleeding, nasal blockage, and nasal secretion. Patients were also encouraged to document any further comments/complaints on these questionnaires.
Evaluation of sinuses Patients were reviewed in the outpatient setting at 2 and 8 weeks postoperatively and up to 6 months after surgery. The stent was removed at the 2-week review to allow for mucosalization as described by Lee and Lee.11 At each follow-up, recorded video sinoscopy was performed. Two experienced rhinologists (E.H.O. and A.S.C.), who were blinded as to which side had received the stent, scored the endoscopy videos independently using the modified LundKennedy scoring system.12 Middle turbinate lateralization was assessed as “present” if a rigid 4-mm endoscope was unable to be passed into the middle meatus because of lateralization of the turbinate and “absent” if the scope was able to pass into the middle meatus. International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
518
Silastic stent use in endoscopic sinus surgery
FIGURE 1. Silastic stent preparation and placement. A 5 × 5 cm silastic sheet with a thickness of 1.02 mm (0.04 inches) (Medtronic, Minneapolis, MN) was folded in half and the ends were held together with a silk suture (A, B). A small perforation was created to allow for drainage of secretions/blood (B). This configuration resulted in a slightly “sprung-open” stent, which was then placed in the middle meatus (C).
Statistical analysis All statistical analyzes were performed by a biostatistician from the Flinders University Epidemiology and Biostatistics Centre using STATA software, version 12.0 (StataCorp, College Station, TX), R version 2.15.1 (R development Core Team, Vienna, Austria), R version 2.15.1 (R development Core Team, Vienna, Austria). An interrater reliability of video scores was calculated using a weighted kappa (kw2) statistic with quadratic weights for ordinal assessments.13 Kappa values were graded as slight agreement (0.01 to 0.20), fair agreement (0.21 to 0.40), moderate agreement (0.41 to 0.60), substantial agreement (0.61 to 0.80), and almost perfect agreement (0.81 to 0.99). The 95% confidence intervals (CIs) of weighted kappa were based on the empirical sampling distribution generated by the computer-intensive bias corrected bootstrapping resampling method.14 Because 2 researchers assessed the video scores, we choose 1 researcher depending on the high agreement of interrater reliability. A simple random sampling strategy was applied to choose the researcher. The video scores from the selected researcher were used for further analysis. Mann-Whitney U test was used to identify LundMackay and baseline VAS scores between stent and nonstent patients. Chi-square analysis was used to compare the effect of stent vs no stent at 2 and 8 weeks postoperatively. A multilevel mixed effect logistic regression model using the “xtmelogit” command was applied in STATA. A maximum likelihood estimation procedure was used to compare the significant differences between the stent and non-stent groups. The model was used to determine the group effects (adjusted odds ratio [OR] of 2 groups at weeks 2 and 8). Because the Lund-Mackay and baseline VAS scores were not significant between the groups, the model was not adjusted for potential confounders. All analysis was performed with 2-tailed tests and the level of significance was set at p < 0.05. Where appropriate, 95% CIs were reported along with p values.
Results Forty-one patients recruited during the period (August 2012 until August 2013) satisfied the inclusion criteria and underwent ESS. Three patients required unilateral partial
519
International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
TABLE 1. Interrater reliability of postoperative sinoscopy
videos Video scorea Agreement (%) w-kappa
95% CI
Judgmentb
Edema
93.1
0.55
0.51–0.57 Moderate agreement
Discharge
85.2
0.19
0.11–0.31 Slight agreement
Adhesion
88.1
0.62
0.54–0.71 Substantial agreement
Crusting
90.3
0.56
0.45–0.70 Moderate agreement
Videos were scored for the characteristics listed using the following scale: 1 = normal; 2 = mild; and 3 = severe. Definition of w-kappa values: ࣘ0 = no agreement; 0.01–0.2 = slight agreement; 0.21–0.4 = fair agreement; 0.4–0.6 = moderate agreement; 0.61–0.8 = substantial agreement; and 0.81–1.0 = almost perfect agreement. CI = confidence interval. a
b
resection of a middle turbinate and 2 patients experienced postoperative hemorrhage that required packing; these 5 patients were thus excluded from the final analysis. The remaining 36 patients completed follow-up at 2 and 8 weeks. There were 35 patients with 6-month clinic video endoscopy data following ESS because 1 patient failed to attend the 6-month postoperative visit. The patient was contacted by phone and declined to return for follow-up because of being symptom free; of the remaining 35 patients, 20 (52.6%) were male and the median age was 39.5 years (interquartile range [IQR] = 28.0 to 39.5 years). Concurrent septoplasty was performed in 78.9% (30/38) of patients. Interrater reliability results (Table 1) indicated that 3 of the 4 video scores (edema, crusting, and scarring) reached at least moderate agreement. Substantial agreement was found for scarring, whereas there was moderate agreement for edema and crusting and slight agreement between observers when rating discharge. No statistically significant differences were found between the stent and non-stent sides for Lund-Mackay scores (median: 8 for non-stent vs 7 with stent, p = 0.79) and baseline (postoperative day 0) VAS scores (Mann-Whitney U test). Video endoscopy assessment showed that the stent reduced the number of lateralized middle turbinates postoperatively, with only 1 of 36 (2.8%) middle turbinate lateralized in the stent side compared with 13 of 36 (36.1%) in the control non-stented sides. Analysis of the 7-day patient diary scores showed natural resolution of
Chan et al.
FIGURE 2. VAS assessing patient symptoms (pain, secretion, bleeding, and obstruction) in the first 7 days after ESS. No statistically significant difference was found between patient symptoms when comparing stent and non-stent sides. Data are presented as median ± IQR. ESS = endoscopic sinus surgery; IQR = interquartile range; VAS = visual analog scale.
symptoms over the week following ESS and both sides were indistinguishable regarding facial pain, bleeding, obstruction, and secretions (p > 0.05) (Fig. 2). When analyzed, the sinoscopy scores showed no difference between sides for edema (p = 0.41 and 0.52, weeks 2 and 8, respectively) or nasal discharge (p = 0.72 and 0.17, respectively). However, the stent side demonstrated a statistically significant decrease in adhesion formation at both 2 weeks (8% vs 64%; OR = 0.01; 95% CI, 0.002 to 0.10; p < 0.001) and 8 weeks (8% vs 61%; OR = 0.02; 95% CI, 0.003 to 0.12; p < 0.001) postoperatively. Similarly crusting at postoperative week 2 was reduced on the stent side (27% vs 58%; OR = 0.13; 95% CI, 0.03 to 0.53; p < 0.01), but in contrast, this benefit resolved by the 8-week follow-up (p = 0.62) (Table 2, Fig. 3). There was no difference in the number of presentations to clinics between the stent and non-stent sides. Thirty-five patients completed their 6-month follow-up with no change in the sinoscopy scores for patients from their 8-week assessment (1 patient failed to attend their 6-month follow-up). At the 6-month follow-up visit a single surgeon’s endoscopic scores were found to have no change in the individual patient’s endoscopic scores from their 8-week scores (EO scores) in regard to adhesions, crusting. and middle turbinate lateralization. There were 3 episodes of postoperative hemorrhage; 2 patients requiring nasal packing were subsequently excluded from analysis; the third patient resolved spontaneously with conservative measures. In addition, 2 patients were recorded as presenting to their family physicians, presenting with symptoms consistent with sinusitis; both were
prescribed oral antibiotics and 1 also received oral steroids. No patients presented to the emergency department or required an unscheduled outpatient clinic review. At the time of submission 1 of the patients has required revision ESS. This was because of adhesions and middle turbinate lateralization that formed on the non-stented side.
Discussion The incidence of adhesion formation following ESS has been reported in up to 27% of cases2 and may be a significant factor predisposing to recurrent disease and revision surgery.5, 15 Adhesions, which form between the middle turbinate and the lateral nasal wall, can lead to lateralization of the middle turbinate, obstruction of the sinuses, and inability for topical medications to reach the sinuses effectively. Nasal packing has been employed postoperatively in an attempt to reduce the incidence of reobstruction and to promote hemostasis.16, 17 The desired characteristics of nasal packing are to achieve hemostasis, to promote wound healing, and to be well tolerated by the patient; ideally, the packing is absorbable. However, a recent review concludes that currently there is no dressing available to fulfill these requirements,17 with some surgeons choosing to avoid packing altogether.6 Similarly, Zhao et al.18 addressed the issue of adhesion formation following ESS using absorbable vs nonabsorbable dressings in their systematic review and meta-analysis. They found that there was no significant difference in adhesion rates between the 2 groups when
International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
520
Silastic stent use in endoscopic sinus surgery
TABLE 2. A multilevel mixed effect logistic regression model of patient sinoscopy scores at postoperative week 2 and week 8 Week 2
a
Scores
Non-stent (n = 36)
Edema
Week 8
Stent (n = 36)
p
Stent (n = 36)a
p
Reference
1.70 (0.54–5.37)
0.37
Reference
0.60 (0.14–2.48)
0.48
Discharge
Reference
1.20 (0.45–3.20)
0.36
Reference
0.42 (0.13–1.38)
0.13
Adhesions
Reference
0.01 (0.002–0.10)
A randomized controlled trial of a middle meatal silastic stent for reducing adhesions and middle turbinate lateralization following endoscopic sinus surgery Chun L. Chan, BRadMedImag (Hons), BMBS1 , Behrad Elmiyeh, MBBS, FRCS (ORL-HNS)1 , Charmaine Woods, BBiotech (Hons), PhD1 , Shahid Ullah, BSc, MSc, PhD2 , Indunil Gunawardena, MBBS, FRACS1 , A. Simon Carney, FRCS, FRACS1 and Eng H. Ooi, MBBS, FRACS, PhD1
Background: Endoscopic sinus surgery (ESS) is indicated for patients with recalcitrant chronic rhinosinusitis symptoms. However, surgical revision can be required because of adhesion formation and middle turbinate lateralization. We investigate the efficacy of a middle meatal silastic stent in reducing these complications aer ESS. Methods: Thirty-six patients were randomized to receive a silastic stent in the middle meatus unilaterally aer ESS. The surgeon was blinded to the side receiving the stent until completion of the ESS. The contralateral side, with no stent, was the control side. Patients completed a 7day, postoperative, visual analog scale symptom diary and were blinded to the stent side until its removal at the first postoperative visit. Patients were followed up aer 2, 8, and 24 weeks. Endoscopic video of the sinus cavities were recorded at all visits and 2 blinded, independent ear/nose/throat (ENT) surgeons assessed the videos using a modified Lund-Kennedy scoring system. Results: Thirty-five of 36 patients completed 6-months’ follow-up. Middle turbinate lateralization was observed in
E
ndoscopic sinus surgery (ESS) is indicated for chronic rhinosinusitis (CRS) refractive to medical therapy.
1 Department
of Otolaryngology–Head and Neck Surgery, Flinders Medical Centre, Flinders University, Adelaide, Australia; 2 Flinders Centre for Epidemiology and Biostatistics, School of Medicine, Faculty of Medicine, Nursing and Health Sciences, Flinders University, Adelaide, Australia
Correspondence to: Eng H. Ooi, MBBS, FRACS, PhD, Department of Otolaryngology–Head and Neck Surgery, Flinders Medical Centre, Bedford Park, SA 5042, Australia; e-mail: [email protected] Potential conflict of interest: None provided. Received: 13 October 2014; Revised: 18 December 2014; Accepted: 23 December 2014 DOI: 10.1002/alr.21490 View this article online at wileyonlinelibrary.com.
517
International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
13 sides without a stent vs 1 side with a stent. There was a significant reduction in adhesions at weeks 2 and 8 (p < 0.001) and crusting (p < 0.01) in the stent side compared to control. Video scores at 6 months aer surgery for stent and control sides remained unchanged from the 8-week visit. There was no difference between sides for symptom scores, edema, or nasal discharge. Conclusion: Middle meatal silastic stents are well tolerated by patients and effective in reducing middle turbinate lateralization, adhesions, and crusting postoperatively in ESS. C 2015 ARS-AAOA, LLC.
Key Words: surgical procedure; endoscopic; intranasal surgery; rhinosinusitis; middle turbinates; tissue adhesions How to Cite this Article: Chan CL, Elmiyeh B, Woods C, et al. A randomized controlled trial of a middle meatal silastic stent for reducing adhesions and middle turbinate lateralization following endoscopic sinus surgery. Int Forum Allergy Rhinol. 2015;5:517–523.
However, revision rates due to surgical failure are a significant problem despite advances in operative and postoperative management, with revision rates reported between 10% and 66% in the literature.1 Postoperative adhesion rates have been reported in up to 27% of patients despite sinus irrigation and postoperative debridement being performed.2–4 Middle turbinate lateralization and formation of adhesions postoperatively have been implicated as the most common cause for suboptimal ESS outcomes requiring revision surgery.5 There have been numerous techniques reported in the literature attempting to reduce adhesion formation and middle turbinate lateralization. These include placement of absorbable and/or nonabsorbable spacers/stents, yet there is no ideal acceptable material.6 There is evidence that the use of nasal packs or topical agents can adversely affect
Chan et al.
wound healing and potentially result in excessive adhesion formation.7 Two recent double-blinded randomized controlled trials have found a sinus silastic stent to be effective in preventing adhesions; however, some patients found these stents to have an adverse effect on postoperative patient comfort.8, 9 In our department, silastic stents are selectively used in ESS as a spacer between the middle turbinate and lateral nasal wall after ESS. The technique that we employ was first described by Gall and Witterick2 in their cohort study of 500 patients, during which they found that their stent was tolerated by the majority of their patients; they reported adhesion formation in 4.2% of their cases. However, since that report was published, no randomized control trial had been performed looking at the effectiveness of the technique. The aim of this study was to perform a double-blinded, randomized controlled trial to evaluate the efficacy of a silastic stent in preventing adhesions and middle turbinate lateralization and the extent to which patients tolerated the presence of the silastic stent following ESS.
Patients and methods The Southern Adelaide Clinical Human Research Ethics Committee’s approval was obtained. All recruited patients provided written informed consent prior to surgery.
Study design We designed a double-blinded randomized controlled trial; based on our unit’s observed adhesion rate of approximately 20%, it was determined that 34 patients were required for a study to have an 80% chance of detecting a significant difference of at least 25% on a visual analog scale (VAS) in a 2-sided t test. Inclusion criteria were adult patients (ࣙ18 years of age) who had CRS ± polyps who had failed medical therapy and underwent ESS with bilateral equivalent dissections (at least middle meatal antrostomies and ethmoidectomies). Patients were excluded if they had unilateral disease, previous sinus surgeries, or any underlying medical condition that may have interfered with wound healing, such as diabetes or autoimmune diseases. Patients were also excluded from final analysis if they required middle turbinate excision during surgery and/or required nasal packing postoperatively. Computer-generated randomization (http://www. randomization.com) was employed preoperatively to determine the side of the nose that would receive the silastic stent, which is referred to as the “stent side.” The other side of the nasal cavity without a silastic stent is referred to as the “control side.” Preoperative computed tomography (CT) scans were reviewed and Lund-Mackay10 scores were calculated. All patients underwent ESS by a single surgeon (B.E.) with the technique described in the next paragraph. If sphenoidotomy was indicated, it was performed via a transethmoidal route to avoid lateralization of the middle
turbinate. Upon completion of surgery, the theatre nurse opened a sealed envelope containing the randomization of the stent side to inform the operating surgeon to place the middle meatal stent prior to extubation. Stents were fashioned from medical grade silastic, using a commercially available silicone sheet (5 × 5 cm, thickness of 1.02 mm [0.04 inches]; Medtronic, Minneapolis, MN) ensuring standard thickness and type of stent between patients. The stent was held together at each end with a silk suture, creating a slightly “sprung-open” stent to be placed in the middle meatus (Fig. 1). The stent was positioned to rest on the superior surface of the inferior turbinate to achieve self-anchoring, thus negating the need for anchoring or holding sutures. No other techniques to prevent middle turbinate lateralization were performed. A “drainage hole” was cut out of the stent superiorly to allow frontal sinus drainage. At the completion of the operation, the surgeon ensured that the septum was midline with both middle turbinates medialized. Patients were blinded to the stent side until the first postoperative visit when stents were removed in the clinic under topical anesthesia.
Postoperative care Standard postoperative treatment included irrigation with R sinus care device (ENT Technologies, Hawthorn the FLO East, VIC, Australia) 4 times a day. Antibiotics are not prescribed routinely for our patients, unless indicated by the presence of pus at the time of surgery. Patients with nasal polyposis were commenced on topical nasal steroid spray (Mometasone furoate 100 μg daily in each nostril; Nasonex; Merck, Kirkland, QC, Canada) after stent removal.
Evaluation of nasal discomfort Patients were asked to fill in a daily VAS diary for the first 7 days postoperation. The VAS was scaled from 0 to 10, with 0 indicating none and 10 indicating maximal levels of a symptom as judged by the patient, with each side assessed independently of the other side. The symptoms assessed included: discomfort/pain, bleeding, nasal blockage, and nasal secretion. Patients were also encouraged to document any further comments/complaints on these questionnaires.
Evaluation of sinuses Patients were reviewed in the outpatient setting at 2 and 8 weeks postoperatively and up to 6 months after surgery. The stent was removed at the 2-week review to allow for mucosalization as described by Lee and Lee.11 At each follow-up, recorded video sinoscopy was performed. Two experienced rhinologists (E.H.O. and A.S.C.), who were blinded as to which side had received the stent, scored the endoscopy videos independently using the modified LundKennedy scoring system.12 Middle turbinate lateralization was assessed as “present” if a rigid 4-mm endoscope was unable to be passed into the middle meatus because of lateralization of the turbinate and “absent” if the scope was able to pass into the middle meatus. International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
518
Silastic stent use in endoscopic sinus surgery
FIGURE 1. Silastic stent preparation and placement. A 5 × 5 cm silastic sheet with a thickness of 1.02 mm (0.04 inches) (Medtronic, Minneapolis, MN) was folded in half and the ends were held together with a silk suture (A, B). A small perforation was created to allow for drainage of secretions/blood (B). This configuration resulted in a slightly “sprung-open” stent, which was then placed in the middle meatus (C).
Statistical analysis All statistical analyzes were performed by a biostatistician from the Flinders University Epidemiology and Biostatistics Centre using STATA software, version 12.0 (StataCorp, College Station, TX), R version 2.15.1 (R development Core Team, Vienna, Austria), R version 2.15.1 (R development Core Team, Vienna, Austria). An interrater reliability of video scores was calculated using a weighted kappa (kw2) statistic with quadratic weights for ordinal assessments.13 Kappa values were graded as slight agreement (0.01 to 0.20), fair agreement (0.21 to 0.40), moderate agreement (0.41 to 0.60), substantial agreement (0.61 to 0.80), and almost perfect agreement (0.81 to 0.99). The 95% confidence intervals (CIs) of weighted kappa were based on the empirical sampling distribution generated by the computer-intensive bias corrected bootstrapping resampling method.14 Because 2 researchers assessed the video scores, we choose 1 researcher depending on the high agreement of interrater reliability. A simple random sampling strategy was applied to choose the researcher. The video scores from the selected researcher were used for further analysis. Mann-Whitney U test was used to identify LundMackay and baseline VAS scores between stent and nonstent patients. Chi-square analysis was used to compare the effect of stent vs no stent at 2 and 8 weeks postoperatively. A multilevel mixed effect logistic regression model using the “xtmelogit” command was applied in STATA. A maximum likelihood estimation procedure was used to compare the significant differences between the stent and non-stent groups. The model was used to determine the group effects (adjusted odds ratio [OR] of 2 groups at weeks 2 and 8). Because the Lund-Mackay and baseline VAS scores were not significant between the groups, the model was not adjusted for potential confounders. All analysis was performed with 2-tailed tests and the level of significance was set at p < 0.05. Where appropriate, 95% CIs were reported along with p values.
Results Forty-one patients recruited during the period (August 2012 until August 2013) satisfied the inclusion criteria and underwent ESS. Three patients required unilateral partial
519
International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
TABLE 1. Interrater reliability of postoperative sinoscopy
videos Video scorea Agreement (%) w-kappa
95% CI
Judgmentb
Edema
93.1
0.55
0.51–0.57 Moderate agreement
Discharge
85.2
0.19
0.11–0.31 Slight agreement
Adhesion
88.1
0.62
0.54–0.71 Substantial agreement
Crusting
90.3
0.56
0.45–0.70 Moderate agreement
Videos were scored for the characteristics listed using the following scale: 1 = normal; 2 = mild; and 3 = severe. Definition of w-kappa values: ࣘ0 = no agreement; 0.01–0.2 = slight agreement; 0.21–0.4 = fair agreement; 0.4–0.6 = moderate agreement; 0.61–0.8 = substantial agreement; and 0.81–1.0 = almost perfect agreement. CI = confidence interval. a
b
resection of a middle turbinate and 2 patients experienced postoperative hemorrhage that required packing; these 5 patients were thus excluded from the final analysis. The remaining 36 patients completed follow-up at 2 and 8 weeks. There were 35 patients with 6-month clinic video endoscopy data following ESS because 1 patient failed to attend the 6-month postoperative visit. The patient was contacted by phone and declined to return for follow-up because of being symptom free; of the remaining 35 patients, 20 (52.6%) were male and the median age was 39.5 years (interquartile range [IQR] = 28.0 to 39.5 years). Concurrent septoplasty was performed in 78.9% (30/38) of patients. Interrater reliability results (Table 1) indicated that 3 of the 4 video scores (edema, crusting, and scarring) reached at least moderate agreement. Substantial agreement was found for scarring, whereas there was moderate agreement for edema and crusting and slight agreement between observers when rating discharge. No statistically significant differences were found between the stent and non-stent sides for Lund-Mackay scores (median: 8 for non-stent vs 7 with stent, p = 0.79) and baseline (postoperative day 0) VAS scores (Mann-Whitney U test). Video endoscopy assessment showed that the stent reduced the number of lateralized middle turbinates postoperatively, with only 1 of 36 (2.8%) middle turbinate lateralized in the stent side compared with 13 of 36 (36.1%) in the control non-stented sides. Analysis of the 7-day patient diary scores showed natural resolution of
Chan et al.
FIGURE 2. VAS assessing patient symptoms (pain, secretion, bleeding, and obstruction) in the first 7 days after ESS. No statistically significant difference was found between patient symptoms when comparing stent and non-stent sides. Data are presented as median ± IQR. ESS = endoscopic sinus surgery; IQR = interquartile range; VAS = visual analog scale.
symptoms over the week following ESS and both sides were indistinguishable regarding facial pain, bleeding, obstruction, and secretions (p > 0.05) (Fig. 2). When analyzed, the sinoscopy scores showed no difference between sides for edema (p = 0.41 and 0.52, weeks 2 and 8, respectively) or nasal discharge (p = 0.72 and 0.17, respectively). However, the stent side demonstrated a statistically significant decrease in adhesion formation at both 2 weeks (8% vs 64%; OR = 0.01; 95% CI, 0.002 to 0.10; p < 0.001) and 8 weeks (8% vs 61%; OR = 0.02; 95% CI, 0.003 to 0.12; p < 0.001) postoperatively. Similarly crusting at postoperative week 2 was reduced on the stent side (27% vs 58%; OR = 0.13; 95% CI, 0.03 to 0.53; p < 0.01), but in contrast, this benefit resolved by the 8-week follow-up (p = 0.62) (Table 2, Fig. 3). There was no difference in the number of presentations to clinics between the stent and non-stent sides. Thirty-five patients completed their 6-month follow-up with no change in the sinoscopy scores for patients from their 8-week assessment (1 patient failed to attend their 6-month follow-up). At the 6-month follow-up visit a single surgeon’s endoscopic scores were found to have no change in the individual patient’s endoscopic scores from their 8-week scores (EO scores) in regard to adhesions, crusting. and middle turbinate lateralization. There were 3 episodes of postoperative hemorrhage; 2 patients requiring nasal packing were subsequently excluded from analysis; the third patient resolved spontaneously with conservative measures. In addition, 2 patients were recorded as presenting to their family physicians, presenting with symptoms consistent with sinusitis; both were
prescribed oral antibiotics and 1 also received oral steroids. No patients presented to the emergency department or required an unscheduled outpatient clinic review. At the time of submission 1 of the patients has required revision ESS. This was because of adhesions and middle turbinate lateralization that formed on the non-stented side.
Discussion The incidence of adhesion formation following ESS has been reported in up to 27% of cases2 and may be a significant factor predisposing to recurrent disease and revision surgery.5, 15 Adhesions, which form between the middle turbinate and the lateral nasal wall, can lead to lateralization of the middle turbinate, obstruction of the sinuses, and inability for topical medications to reach the sinuses effectively. Nasal packing has been employed postoperatively in an attempt to reduce the incidence of reobstruction and to promote hemostasis.16, 17 The desired characteristics of nasal packing are to achieve hemostasis, to promote wound healing, and to be well tolerated by the patient; ideally, the packing is absorbable. However, a recent review concludes that currently there is no dressing available to fulfill these requirements,17 with some surgeons choosing to avoid packing altogether.6 Similarly, Zhao et al.18 addressed the issue of adhesion formation following ESS using absorbable vs nonabsorbable dressings in their systematic review and meta-analysis. They found that there was no significant difference in adhesion rates between the 2 groups when
International Forum of Allergy & Rhinology, Vol. 5, No. 6, June 2015
520
Silastic stent use in endoscopic sinus surgery
TABLE 2. A multilevel mixed effect logistic regression model of patient sinoscopy scores at postoperative week 2 and week 8 Week 2
a
Scores
Non-stent (n = 36)
Edema
Week 8
Stent (n = 36)
p
Stent (n = 36)a
p
Reference
1.70 (0.54–5.37)
0.37
Reference
0.60 (0.14–2.48)
0.48
Discharge
Reference
1.20 (0.45–3.20)
0.36
Reference
0.42 (0.13–1.38)
0.13
Adhesions
Reference
0.01 (0.002–0.10)
