Patients and disease characteristics at baseline

The patients in our cohort used oxymetazoline or xylometazoline in different forms, which are commercially available over the counter in Israel. However, other decongestants may also cause rhinitis medicamentosa.^{Reference Patel, Levi and Brook6}

The baseline characteristics of decongestant use before surgical or conservative intervention did not differ significantly (Table 2). Other studies presented a wide-ranging duration of decongestant use among rhinitis medicamentosa patients.^{Reference Caffier, Frieler, Scherer, Sedlmaier and Göktas2,Reference Zucker, Barton and McCoul3,Reference Patel, Levi and Brook6,Reference De Corso, Mastrapasqua, Tricarico, Settimi, Di Cesare and Mele7}

The pathophysiological mechanisms of rhinitis medicamentosa are not yet completely understood, and so there is no consensus on treatment protocol. Histological, pharmacokinetic and pathophysiological studies are necessary to understand the mechanisms of rhinitis medicamentosa, which would inform investigators with regard to targeted treatments.^{Reference Zucker, Barton and McCoul3}

Histological changes consistent with rhinitis medicamentosa include nasociliary loss, squamous cell metaplasia, epithelial oedema and goblet cell hyperplasia.^{Reference Li, Wang, Tsai and Wu8}

Patients in both treatment groups did not differ significantly in terms of reported allergies and positive allergy test rates (Table 1). To the best of our knowledge, allergic rhinitis is not a risk factor for rhinitis medicamentosa. Nasal decongestants have been used worldwide to relieve nasal congestion such as allergic and non-allergic rhinitis.^{Reference Caffier, Frieler, Scherer, Sedlmaier and Göktas2} According to US Food and Drug Administration recommendations, oxymetazoline hydrochloride should not be used for longer than 7 days, in order to avoid rhinitis medicamentosa.^{Reference Neighbors, Salvador, Zhu, Camacho and Tsai9} A systematic review found no reports of tachyphylaxis or rebound congestion after treatment with a combination of intranasal corticosteroid spray and oxymetazoline hydrochloride for allergic rhinitis during 14–28 days.^{Reference Neighbors, Salvador, Zhu, Camacho and Tsai9} Nasal decongestants are recommended together with intranasal corticosteroids for the management of acute rhinosinusitis, without an increased risk for developing rebound congestion after 10–30 days.^{Reference Mortuaire, de Gabory, François, Massé, Bloch and Brion10}

No significant differences were found between the surgical and conservative treatment groups in terms of smoking status (Table 1). Prolonged tobacco smoking has shown to cause mucociliary dysfunction and contribute to the development of rhinitis.^{Reference Wallace, Dykewicz, Bernstein, Blessing-Moore, Cox and Khan4} In one study, smoking correlated positively with decongestant abuse.^{Reference De Corso, Mastrapasqua, Tricarico, Settimi, Di Cesare and Mele7} Li et al. reported similar baseline characteristics between patients who ceased nasal decongestants and those who continued use, following nasal surgery. However, they found a higher rate of asthma among patients in whom surgical therapy failed.^{Reference Li, Misra, Harvey and Kalish11}

There is a strong addictive component to rhinitis medicamentosa, including psychological dependence on nasal decongestants and withdrawal symptoms. A higher risk of opioid abuse was seen in rhinitis medicamentosa patients.^{Reference Patel, Levi and Brook6}

Patients’ reported outcomes

The primary outcome was withdrawal from decongestant use, with changes in QoL after intervention (vs before intervention) being a secondary outcome.^{Reference Shapira Galitz, Halperin, Bavnik and Warman5} The high baseline SNOT-22 score of our cohort (mean score of 35.83 ± 22.2; Table 2) suggests that this condition significantly affects patients' QoL. In comparison, one study of rhinitis medicamentosa patients revealed a mean baseline SNOT-22 score of 40.3, and other studies of chronic rhinosinusitis patients reported a mean SNOT-22 score of 42.0, while a healthy control group had a mean score of 9.3.^{Reference Li, Misra, Harvey and Kalish11,Reference Hopkins, Gillett, Slack, Lund and Browne12}

One may argue that the Nasal Obstruction Symptom Evaluation (‘NOSE’) scale is more accurate than the SNOT-22 scale, which was originally developed for evaluating chronic sinusitis.^{Reference Hopkins, Gillett, Slack, Lund and Browne12,Reference Stewart, Witsell, Smith, Weaver, Yueh and Hannley13} We believe that the Nasal Obstruction Symptom Evaluation scale lacks some important questions that are relevant for rhinitis medicamentosa patients, such as waking up at night, waking up tired, and feeling frustrated, restless or embarrassed. We found that these parameters were significantly affected in rhinitis medicamentosa patients. Other studies used different symptom scores, such as nasal congestion visual analogue scale scores (of 1–10), Rhinitis Questionnaire Symptom Scores, Nasal Symptoms Scores, and the ‘need’ for topical decongestants.^{Reference Zucker, Barton and McCoul3,Reference Li, Misra, Harvey and Kalish11}

Outcomes following medical therapy

We did not use a uniform protocol for medical therapy. The most frequent regimen was a combination of intranasal corticosteroid and saline rinses, in parallel to weaning off from decongestant use (63.8 per cent of patients; Table 2). Similarly, in a survey of Canadian otolaryngologists, weaning off decongestants and intranasal corticosteroid use was the most common regimen for treating rhinitis medicamentosa (61 per cent of responders).^{Reference Fowler, Chin and Massoud14}

The conservative treatment group showed a significant decrease in the frequency of decongestant use after medical therapy, in comparison to the surgical treatment group (Table 2). The surgical treatment group was non-responsive to medical therapy, which was the reason they had been referred for surgery. This may be an important prognostic factor for rhinitis medicamentosa patients that could be an indication for inferior turbinate surgery. A possible explanation is the significantly shorter treatment duration relative to the conservative treatment group (mean of 1.6 vs 7.8 months; Table 2). In addition, some patients referred for surgery stopped using topical medications earlier than recommended (two months).

The ideal duration of medical therapy for rhinitis medicamentosa has not yet been determined, because the level of evidence is weak, and no randomised controlled trials have been completed.^{Reference Vaidyanathan, Williamson, Clearie, Khan and Lipworth15}

A systematic review included studies with topical medical therapy durations of two weeks to two months.^{Reference Zucker, Barton and McCoul3} A recent study presented lower rates of intranasal corticosteroid use by rhinitis medicamentosa patients before surgery.^{Reference Li, Misra, Harvey and Kalish11}

A randomised, placebo-controlled study of healthy subjects showed that nasal congestion resolved after the administration of fluticasone 200 μg twice daily for 3 days, despite continued use of a nasal decongestant.^{Reference Vaidyanathan, Williamson, Clearie, Khan and Lipworth15}

Another study evaluated the effect of fluticasone propionate (200 μg daily) versus placebo for 14 days on nasal congestion, nasal resistance, peak inspiratory flow and acoustic rhinomanometry after stopping nasal decongestants. Symptomatic improvement was observed on day 4 with corticosteroids and on day 7 with the placebo.^{Reference Hallén, Enerdal and Graf16}

The SNOT-22 scores in the conservative treatment group were not significantly reduced from baseline after medical therapy (Table 2, Figure 1). One possible reason is that some of the patients’ co-morbidities, such as obstructive sleep apnoea syndrome, might have affected SNOT-22 scores as well.

Outcomes following inferior turbinate surgery

Twenty-six patients, refractory to medical therapy, underwent endoscopic bilateral medial flap inferior turbinoplasty. We did not offer other inferior turbinate reduction techniques, such as Coblation® or electrocautery.

The frequency of decongestant use following surgery was reduced significantly, and their mean use was significantly lower than in the conservative treatment group following medical therapy (Table 2).

Similarly, SNOT-22 score reductions were more significant following surgery as compared to after medical therapy (Table 2, Figure 1), reflecting the greater effect of surgery and a better QoL for those patients. The mean SNOT-22 post-operative score of 10.3 ± 20.8 was similar to that of the healthy control group with a mean score of 9.3, according to European Position Paper on Rhinosinusitis and Nasal Polyps 2020 reports.^{Reference Fokkens, Lund, Hopkins, Hellings, Kern and Reitsma17}

Li et al. examined the effect of nasal surgery on rhinitis medicamentosa patients, and showed similar findings for surgical intervention in terms of decongestant use (91 per cent cessation rate) and SNOT-22 scores (change = −27.1), before and after surgery. However, that study did not compare conservative treatment to surgical intervention, and included other types of nasal surgery, such as septoplasty and limited functional endoscopic sinus surgery to address middle turbinate contact points.^{Reference Li, Misra, Harvey and Kalish11}

Co-morbidities such as ischaemic heart disease, atrial fibrillation and hypertension might have affected the otolaryngologist's decision of whether to refer the patient for inferior turbinate surgery; however, this is debatable. Turbinate surgery is considered safe, with a low rate of major complications.^{Reference Menezes, Guimarães, Breda, Vieira and Dias18} Several techniques of turbinate surgery performed under local or general anaesthesia exist.^{Reference Bergmark and Gray19}

Menezes et al. showed that septum and turbinate surgery performed under general anaesthesia are safe in the ambulatory setting, with a low rate of unexpected hospital revisits. No associations were found between prolonged hospital stay and co-morbidities or American Society of Anesthesiologists physical status classification.^{Reference Menezes, Guimarães, Breda, Vieira and Dias18}

Other surgical techniques for inferior turbinate surgery have been shown to be safe for co-morbid and older patients when performed under local anaesthesia. Age and the presence of a pulmonary co-morbidity did not significantly influence surgical outcomes.^{Reference Resende, Carmo, Mocellin, Pasinato and Mocellin20}

Our rate of minor complications post-operatively, mainly crusting, was somewhat higher than reported in other studies (Table 3).^{Reference Orabi, Sen, Timms and Morar21,Reference Barham, Thornton, Knisely, Marcells, Harvey and Sacks22} A possible explanation is the relatively short interval between the surgery and the follow-up examination (mean of 3.2 weeks).

Barham et al. found that long-term outcomes of endoscopic medial flap inferior turbinoplasty were superior to those of submucosal electrocautery and submucosal powered turbinate reduction for the indication of nasal obstruction. Post-operative complications included pain requiring analgesia (14 per cent), and bleeding requiring nasal packing or surgical intervention (4 per cent).^{Reference Barham, Thornton, Knisely, Marcells, Harvey and Sacks22}

To the best of our knowledge, this is the first study to compare conservative treatment with surgical intervention for rhinitis medicamentosa. In addition, it is the first to examine the effect of endoscopic medial flap inferior turbinoplasty solely for the indication of rhinitis medicamentosa.

The limitations of our study include its retrospective character, the low number of patients and the heterogenicity of treatment groups. In addition, there was no identical treatment protocol for rhinitis medicamentosa, and there was a lack of objective measures such as acoustic rhinomanometry or peak nasal inspiratory flow.

Other medical and surgical interventions have shown promising results for rhinitis medicamentosa. One animal study found that xylitol is as effective as a nasal steroid when evaluating histopathological responses to treatment in rhinitis medicamentosa.^{Reference Cam, Sari, Midi and Gergin23} Future clinical studies should examine its efficacy in treating rhinitis medicamentosa. Another study showed the efficacy of nebulised hyaluronic acid spray in reducing the use of topical decongestant in rhinitis medicamentosa patients.^{Reference Casale, Vella, Moffa, Sabatino, Rinaldi and Grimaldi24}

Caffier et al. found that out-patient diode laser treatment for inferior turbinate reduction was highly effective, safe and well tolerated by patients, and 88 per cent of patients stopped decongestant use within six months post-operatively.^{Reference Caffier, Frieler, Scherer, Sedlmaier and Göktas2}

Orabi et al. reported excellent symptomatic improvement and a low complication rate following potassium titanyl phosphate laser inferior turbinectomy for patients with allergic rhinitis and non-allergic rhinitis refractory to medical treatment.^{Reference Orabi, Sen, Timms and Morar21}