鼻内皮质类固醇治疗非过敏性鼻炎
Abstract
研究背景
非过敏性鼻炎的定义是由除过敏原、微生物之外的其他刺激因素引起的非传染性炎症和鼻粘膜功能障碍。它很常见，患病率约为 10% 到 20%。患者会出现鼻塞、前鼻流涕、鼻后滴漏感和打喷嚏的症状。根据诱发症状的诱发因素，可将非过敏性鼻炎分为多个亚组，包括职业、香烟香雾、激素、药物、食物和年龄。在细胞分子水平上，也可以确定不同的致病机制。对于非过敏性鼻炎的患者来说，由于缺乏对疾病的了解和对潜在致病机制的认识，常常缺少有效的治疗方法。鼻内皮质类固醇是鼻炎或鼻窦炎症状患者（包括非过敏性鼻炎患者）最常用的药物类型之一。然而， 目前还不清楚鼻内皮质类固醇是否确实对这些患者有效。
研究目的
评价鼻内皮质类固醇对非过敏性鼻炎管理的效果。
检索策略
Cochrane耳鼻喉文献信息检索专员检索了Cochrane耳鼻喉注册库（the Cochrane ENT Register）、Cochrane 对照试验中心注册库（CENTRAL, 2019年，卷7）、PubMed、Ovid EMBASE、CINAHL、Web of Science、ClinicalTrials.gov、ICTRP和其他数据库来寻找发表及未发表的试验。检索日期为2019年7月1日。
标准/纳入排除标准
对成人和儿童（≥12岁）以任何方式和任何剂量给予鼻内皮质类固醇与（a）安慰剂/不干预或（b）其他积极治疗进行比较的随机对照试验（RCTs）。
数据收集与分析
我们使用了Cochrane推荐的标准方法学流程。主要结局是患者报告的疾病严重程度和
一种明显的不良反应——鼻出血。次要结局是（特定疾病的）健康相关生活质量，气流（呼吸道通畅程度）的客观测量和其他不良事件。我们使用GRADE来评价每个结局的证据质量。
主要结果
我们纳入了34项研究（4452名受试者）。然而，只有13项研究提供了鼻内皮质类固醇与安慰剂的比较相关资料（本研究主要的对比措施）。受试者主要为多年鼻炎症状且过敏测试阴性的患者。尽管少数研究仅纳入了特定临床表现的受试者，例如妊娠鼻炎，血管舒缩性鼻炎，药物性鼻炎或老年性鼻炎，但无法区分不同的表型（临床表现）和内型（内在机制）。大多数研究的实施环境是二级或三级卫生保健机构。没有研究报告的随访结果超过了3个月。在此次评价中，鼻内皮质类固醇剂量从每天50μg到2000μg不等。
鼻内皮质类固醇与安慰剂对比
13项研究（2045名受试者）为这一比较提供了资料。这些研究使用了不同的评分系统来评价患者报告的疾病严重程度，因此我们使用标准化平均差（standardised mean difference，SMD）进行分析。与安慰剂相比，在最多4周的时间里，鼻内皮质类固醇治疗可以改善患者报告的疾病严重程度（总鼻部症状评分）（SMD=‐0.74，95% CI [‐1.15, ‐0.33]; 4项研究；131受试者，I 2 =22%）（低质量证据）。然而，在4周到3个月之间，疗效证据非常不确定（SMD=‐0.24, 95% CI [‐0.67, 0.20]; 3项研究；85名受试者；I 2 =0%）（极低质量证据）。与安慰剂相比，在长达4周的时间里，鼻腔皮质类固醇激素治疗可以轻微改善患者报告的疾病严重程度（总鼻部症状得分相对于基线水平的变化）（SMD=‐0.15, 95% CI [‐0.25, ‐0.05]; 4个研究； 1465名受试者； I 2 =35％）（低质量性证据） 。
所有四项研究中鼻出血风险的结果表明，与安慰剂（每1000人中31人鼻出血）相比，鼻内皮质类固醇（每1000人中65人鼻出血）的风险可能更高（RR=2.10, 95% CI [1.24, 3.57]; 4项研究；1174名受试者；I 2 =0%）（中等质量证据）。绝对风险差 （RD） 为 0.04，额外的伤害结局的需治疗数（NNTH）为25（95% CI [16.7, 100]）。
只有1项研究报告了生活质量的数值数据。确实报告了鼻内皮质类固醇的生活质量得分更高（152.3对比145.6; SF‐12v2范围 0 至 800）。然而，这种情况在长期随访中消失了（148.4对比145.6）（低质量证据）。
只有2项研究提供了气流（气道流畅情况）的客观测量结局。因为所使用不同的结局指标测量方法不同，所以无法合并数据。两者均未发现鼻内皮质类固醇和安慰剂组之间的显著差异（鼻测量值SMD=‐0.46, 95% CI [‐1.06, 0.14]; 44名受试者；呼气峰值流速SMD=0.78, 95% CI [‐0.47, 2.03]; 11名受试者）（极低质量证据）。
与安慰剂相比，鼻内皮质类固醇导致的其他不良事件的风险可能差别较小或没有差别（RR=0.99, 95% CI [0.87, 1.12]; 3项研究；1130名受试者；I 2 = 0%）（中等质量证据） 。
鼻内皮质类固醇与其他治疗对比
只有1项或少数几项研究评价了其他比较（鼻内皮质类固醇与生理盐水冲洗，鼻内抗组胺药，辣椒素，色甘酸钠，异丙托溴铵，鼻内皮质类固醇联合鼻内抗组胺药，鼻内皮质类固醇联合鼻内抗组胺药比较；鼻内皮质类固醇联合盐水与单独使用盐水比较）。因此，尚不确定鼻内皮质类固醇与其他积极治疗之间，对所报告的任何结局是否存在差异。
作者结论
总体而言，本综述中大多数结局的证据质量为低或极低。与安慰剂相比，在三个月的随访报告中，尚不确定鼻内皮质类固醇是否能降低非过敏性鼻炎患者报告的疾病严重程度。然而，使用鼻内皮质类固醇可能有较高的不良反应（鼻出血）风险。很少有研究将鼻内皮质类固醇与其他治疗方式进行比较，因此很难下结论。
PICOs
Plain language summary
鼻内皮质类固醇治疗非过敏性鼻炎
综述问题
我们想了解鼻内皮质类固醇（将类固醇药物直接应用于鼻）是否能有效治疗非过敏性鼻炎。
研究背景
非过敏性鼻炎是一种非感染、非过敏反应导致的鼻腔慢性疾病。非过敏性鼻炎的症状会影响患者的生活质量，如鼻塞、流鼻涕和打喷嚏。非过敏性鼻炎患者根据病因的不同，可以分为不同的亚组。因为非过敏性鼻炎的根本病因尚未完全了解，所以对患者的治疗往往并不成功。
局部（鼻腔内）使用皮质类固醇可用于减轻炎症。它们是鼻和鼻窦的其他慢性疾病最常使用的处方药，如过敏性鼻炎和慢性鼻窦炎。鼻内皮质类固醇治疗可以使用喷雾剂或滴剂，并在不同的时间段使用。
研究特征
我们在此次系统综述中，纳入了34项随机对照试验（RCTs），共涉及4452名受试者。尽管最大型的研究共有983名受试者，但大多数研究相对为小型研究。所有受试者为患有非过敏性鼻炎的成人或青少年（12至18岁）。这些研究对鼻内皮质类固醇的各种类型，剂量和给药方法（例如喷雾剂，滴剂）都进行了研究。9项研究由药厂资助或存在商业资助。1项研究由政府资助。在几项研究中，药厂或商业资助者可能提供了药物，但具体的资助者不清楚。8项研究没有报告资助情况。
重要结局
鼻内皮质类固醇与安慰剂对照
与安慰剂相比，在三个月的观察报告中，尚不确定鼻内皮质类固醇是否能降低非过敏性鼻炎患者报告的疾病严重程度。与安慰剂相比，它们（鼻内皮质类固醇）可在4周的时间内改善患者报告的疾病严重程度，但是这一证据质量较低。鼻内皮质类固醇的治疗可能会增加鼻衄（鼻出血）的风险，但在其他不良反应风险方面没有差异。从本综述中无法得知鼻内皮质类固醇的不同浓度、给药方法或治疗方案之间（效果）是否存在差异。尚无评价使用鼻内皮质类固醇改善生活质量的高质量研究。
鼻内皮质类固醇与其他治疗对照
没有足够的证据表明鼻内皮质类固醇治疗是否比其他治疗方法更好、更差或与相同，如使用盐水冲洗，鼻内抗组胺药，辣椒素或异丙托溴铵治疗非过敏性鼻炎。
证据质量
总体而言，与安慰剂相比，大多数有关鼻内皮质类固醇治疗结局的证据为低质量（我们对估计的效应的信心较低）或极低质量（我们对估计的效应的信心非常低）。这是因为大多数研究的规模很小，而且使用不同的测量方法来测量同一结局指标。证据更新至2019年7月。
Authors' conclusions
Summary of findings
|
Intranasal corticosteroids compared to placebo for non‐allergic rhinitis |
|||||||
|
Patient or population:
adults and children > 12 with non‐allergic rhinitis
|
|||||||
|
Outcomes |
Anticipated absolute effects * (95% CI) |
Relative effect
|
№ of participants
|
Certainty of the evidence
|
Comments |
||
|
Risk with placebo |
Risk with intranasal corticosteroids |
||||||
|
Disease severity as measured by patient‐reported symptom score (total nasal symptom score) |
Follow‐up ≤ 4 weeks |
— |
SMD 0.74 lower
|
— |
131
|
⊕⊕⊝⊝
|
Intranasal corticosteroids may improve patient‐reported disease severity at a follow‐up of up to 4 weeks compared to placebo. The mean difference in disease severity score was 0.74 standard deviations lower (1.15 lower to 0.33 lower) with intranasal corticosteroids compared to placebo. This represents a medium effect size ( Cohen 1988 ). |
|
Follow‐up > 4 weeks |
— |
SMD 0.24 lower
|
— |
85
|
⊕⊝⊝⊝
|
It is uncertain whether intranasal corticosteroids improve patient‐reported disease severity with a follow‐up of more than 4 weeks compared to placebo, because the certainty of the evidence is very low. |
|
|
Change from baseline Follow‐up ≤ 4 weeks |
— |
SMD 0.15 lower (0.25 lower to 0.05 lower) |
— |
1465
|
⊕⊕⊝⊝
|
Intranasal corticosteroids may slightly improve patient‐reported disease severity change from baseline with a follow‐up of up to 4 weeks compared to placebo. The SMD of 0.15 represents a small effect size. There are two large studies ( Jacobs 2009 ; Webb 2002 ). Jacobs 2009 reports with a high degree of certainty a small improvement in favour of intranasal corticosteroids. Webb 2002 reports a less certain clinically relevant improvement in favour of intranasal corticosteroids. Jacobs 2009 has an adjusted SD value (presented SD are most likely SEM). |
|
|
Significant adverse event: epistaxis Follow‐up: 2 weeks to 33 days |
Study population |
RR 2.10 (1.24 to 3.57) |
1174
|
⊕⊕⊕⊝
|
There is probably a higher risk of epistaxis with intranasal steroids compared to placebo. |
||
|
(31 per 1000 |
65 per 1000
|
||||||
|
Disease‐specific health‐related quality of life Short Form 12 (SF‐12v2) (range 0 to 800) Follow‐up: 1 month to 3 months |
Just one study reported quality of life ( Lin 2017 ). Quality of life was better in the intranasal corticosteroids group versus the placebo group, however while this difference was significant at a follow‐up of 1 month (152.3 versus 145.6) it was barely noticeable at a follow‐up of 3 months (148.4 versus 145.6). |
49 (1 RCT) |
⊕⊕⊝⊝ low 5 |
There is not enough information (1 study) to conclude whether there is a difference. |
|||
|
Objective measurement of airflow: peak flow rate (expiratory) Follow‐up: 2 weeks to 4 weeks |
Just two studies reported objective measurement of airflow ( Malm 1981 ; Spector 1980 ). However, they used different outcome measurements to measure outflow: rhinomanometry and expiratory peak flow rate (PEFR). Neither found a significant difference between groups: rhinomanometry ( Malm 1981 ) SMD ‐0.46, 95% CI ‐1.06 to 0.14; 44 participants; PEFR ( Spector 1980 ) SMD 0.78, 95% CI ‐0.47 to 2.03; 11 participants. |
55
|
⊕⊝⊝⊝ very low 6 |
There is not enough information (2 studies with different methods of measurement) to conclude whether there is a difference. |
|||
|
Other adverse events Follow‐up: 1 month to 6 weeks |
Study population |
RR 0.99
|
1130
|
⊕⊕⊕⊝
|
Intranasal corticosteroids probably result in little or no difference in the risk of other adverse events compared to placebo. |
||
|
454 per 1000 |
450 per 1000
|
||||||
|
*
The risk in the intervention group
(and its 95% confidence interval) is based on the assumed risk in the comparison group and the
relative effect
of the intervention (and its 95% CI).
|
|||||||
|
GRADE Working Group grades of evidence
|
|||||||
|
1 Due to the small sample size we downgraded once for imprecision and once due to the risk of publication bias. The I 2 value in this pooled analysis was 22% so there was no reason to downgrade for heterogeneity. 2 We downgraded twice for serious imprecision due to the small sample size and because the confidence interval includes both meaningful benefit and harm. We downgraded once for risk of publication bias due to the small sample size. 3 We downgraded twice for serious inconsistency because the I 2 value was 96% when the original SD values presented in the Jacobs 2009 study were used and in that case the confidence interval for Jacobs 2009 does not overlap with the other studies. Jacobs 2009 has an unlikely SD value, which does not match the mean, n and P values. It is likely that the originally SD value presented should actually be a standard error of the mean (SEM), therefore we re‐calculated the SD values. 4 We downgraded once due to study limitations (risk of bias) because there were unclear blinding domains, which could have influenced the significant adverse events (epistaxis) outcome. The I 2 value in this pooled analysis is 0% so there is no reason to downgrade for heterogeneity. We judged that there were no other reasons to downgrade. 5 Due to the small sample size we downgraded once for imprecision and once due to the risk of publication bias. 6 We downgraded once for inconsistency (when the two studies are combined the I 2 value is 67%). The two studies used different methods for measuring objective airflow, which contributed to the heterogeneity. Due to the small sample size we downgraded once for imprecision and once due to the risk of publication bias. 7 We downgraded once due to study limitations (risk of bias) as there were unclear blinding domains, which could have influenced the adverse events outcome. The I 2 value in the pooled analysis is 0% so there is no reason to downgrade for heterogeneity. We judged that there were no other reasons to downgrade. |
|||||||
Background
Description of the condition
Chronic rhinitis (allergic and non‐allergic) affects up to 30% to 40% of the general population ( Bousquet 2008 ). Non‐allergic rhinitis is diagnosed when anatomic, infectious and allergic aetiologies are excluded and symptoms have been present for more than 12 weeks. The symptoms include nasal congestion, clear rhinorrhoea, sneezing and, less frequently, nasal itching. It is common, with an estimated prevalence of around 10% to 20% ( Bachert 2008 ). Most epidemiological studies report that 25% to 50% of chronic rhinitis patients can be categorised as having non‐allergic rhinitis ( Fokkens 2002 ), with a worldwide estimated prevalence of 200 to 400 million people ( Bousquet 2008a ). Most studies agree on a female predominance ( Knudsen 2009 ; Molgaard 2007 ). A recent study has shown that quality of life is significantly impaired in people with non‐allergic rhinitis and this impairment is equal to that in people with allergic rhinitis ( Segboer 2018 ). Around 60% of non‐allergic rhinitis patients develop non‐allergic asthma ( Hellings 2017 ).
Within non‐allergic rhinitis several phenotypes can be differentiated: environmental (occupational, smoking), hormonal (e.g. pregnancy), gustatory, age (rhinitis of the elderly), medication‐induced and inflammatory (non‐allergic rhinitis with eosinophilia syndrome (NARES) or local allergic rhinitis) ( Hellings 2017 ; Papadopoulos 2015 ). In local allergic rhinitis, patients have the clinical characteristics of allergic rhinitis and an allergen sensitisation but no systemic signs of atopy. NARES patients have high numbers of eosinophils in their nasal mucosa and can have micro‐polyposis, hyposmia and signs of bronchial hyper‐responsiveness in a limited way, comparable to patients with chronic rhinosinusitis. The prevalence rates of these different phenotypes are unknown.
Environmental (occupational (chemical) and smoking) rhinitis can be clearly linked to an affecting agent. In close to 60% of cases, occupational rhinitis can be associated with occupational asthma ( Ameille 2013 ). Smoking is considered a specific irritant of the nasal mucosa, which can cause non‐allergic rhinitis ( van Rijswijk 2005 ). Hormonal rhinitis can occur during the menstrual cycle and puberty, due to hypothyroidism or acromegaly, as well as during pregnancy, where it resolves postpartum. Gustatory rhinitis is accompanied by oversecretion of nasal mucus in response to irritating gustatory agents, usually spicy foods ( Waibel 2008 ). Rhinitis of the elderly (senile rhinitis) is encountered in the older generation and characterised by the presence of constant rhinorrhoea and a lack of other nasal complaints.
In the case of medication‐induced rhinitis (rhinitis medicamentosa), several medications have been implicated ( Varghese 2010 ). The most common is the misuse of topical sympathomimetics (e.g. oxy‐ or xylometazoline) for more than 10 days, resulting in dysregulation of the adrenergic receptors in the nasal mucosa and a relative increase of the parasympathetic drive, leading to significant rhinorrhoea and nasal obstruction. These symptoms cause the patients to continue using topical adrenergics, perpetuating a vicious cycle. Treatment is usually focused on cessation of the affecting agent, as well as support with intranasal corticosteroids.
In terms of the pathophysiological mechanisms, neurogenic, inflammatory and idiopathic endotypes can be distinguished. Two phenotypes clearly belong to the inflammatory endotype: local allergic rhinitis and non‐allergic rhinitis with eosinophilia syndrome (NARES). Within the neurogenic endotype, neurogenic dysbalance (for example, senile rhinitis) and neurogenic inflammation (for example, idiopathic rhinitis) can be differentiated.
Local allergic rhinitis is diagnosed when skin prick and serum specific IgE testing are negative, however a nasal allergen provocation test is positive ( Rondon 2012a ). A recent report attributed over a quarter of chronic rhinitis patients to local allergic rhinitis ( Rondon 2012b ). NARES is considered in the presence of rhinitis symptoms, no evidence of allergy and more than 20% eosinophilia on nasal smears ( Ellis 2007 ). Its pathophysiology is poorly understood, but is thought to involve a local, self‐perpetuating nasal inflammation with eosinophilia ( Groger 2012 ). Idiopathic rhinitis has for a long time remained a diagnosis of exclusion, when the other causes of rhinitis have been ruled out ( Burns 2012 ). Its suggested pathophysiology includes chronic inflammation of an antigenic or neurogenic nature ( van Rijswijk 2005 ).
In explaining non‐allergic rhinitis to patients, doctors have often referred to the concept of nasal hyperreactivity. For that reason, non‐allergic rhinitis ‐ or idiopathic rhinitis ‐ was also called vasomotor rhinitis in the past. However, recent literature shows us that nasal hyperreactivity is a common symptom in both allergic and non‐allergic rhinitis. The terminology of vasomotor rhinitis is therefore no longer used.
Treatment of non‐allergic rhinitis includes trigger avoidance, topical and systemic medications, and surgery. When rhinitis is caused by a known aetiologic factor, such as smoking or chemical exposure, the mainstay of treatment is trigger avoidance.
Several medications are widely utilised in the treatment of non‐allergic rhinitis, including oral and topical nasal antihistamines, intranasal and (rarely) systemic corticosteroids, and anticholinergics (ipratropium bromide). Other medical options include capsaicin, intranasal injection of botulinum toxin type A, intranasal saline rinse, local and systemic sympathomimetics and cromolyn sodium. The exact mechanisms of effect of these therapies in non‐allergic rhinitis remain largely unknown.
Some medications are particularly useful in specific types of non‐allergic rhinitis. Specifically, ipratropium bromide is mostly used in the treatment of rhinitis of the elderly, due to its alleviation of the main symptom, rhinorrhoea ( van Rijswijk 2005 ). Intranasal antihistamines are usually prescribed when sneezing is the main symptom ( Schroer 2012 ). Capsaicin (8‐methyl‐N‐vanillyl‐6‐nonenamide), the active component of chili peppers, appears to have a therapeutic effect in idiopathic rhinitis, based on several randomised controlled trials ( Ciabatti 2009 ; van Rijswijk 2003 ).
Surgical reduction can be considered to treat inferior turbinate hypertrophy, when it contributes to nasal obstruction and mucosal hypersecretion in chronic rhinitis ( Garzaro 2012 ). Vidian neurectomy, causing denervation of the autonomic supply of the nasal mucosa, can reduce the symptoms of rhinorrhoea and nasal obstruction ( Robinson 2006 ).
Description of the intervention
Topical (local) intranasal corticosteroids are administered as nasal sprays or drops. Intranasal corticosteroids act locally on the nasal mucosa, eliciting anti‐inflammatory and immunosuppressant effects, while mostly avoiding the systemic side effects of corticosteroids ( Bruni 2009 ; Emin 2011 ; Mizrachi 2012 ).
Currently available intranasal corticosteroid preparations include the earlier generation medications beclomethasone dipropionate, triamcinolone acetonide, flunisolide and budesonide, and the newer preparations fluticasone propionate, fluticasone furoate and mometasone furoate. They differ in their local potency, lipid solubility, bioavailability, and local and systemic side effects.
The local side effects of intranasal corticosteroids include epistaxis (5% to 10%), nasal irritation (5% to 10%, including dryness, burning and stinging), headache, nasal septal perforation (< 1%), candida infection of the nose and pharynx, and impaired wound healing after recent nasal surgery or trauma ( Merck 2012 ).
How the intervention might work
Corticosteroids have immunosuppressant and anti‐inflammatory effects, modifying and reducing inflammation through suppression of the synthesis of pro‐inflammatory cytokines and pro‐inflammatory enzymes, inhibiting lymphocyte proliferation and chemotaxis ( Mygind 2001 ).
The local pharmacology of intranasal corticosteroids is connected with absorption characteristics (lipid solubility), topical potency (receptor‐binding ability) and systemic bioavailability ( Benninger 2003 ). The delivery mechanism (sprays versus drops) can also influence local drug concentration and its subsequent metabolism.
In allergic rhinitis, optimal therapeutic efficacy can be achieved after daily use of intranasal corticosteroids for two weeks ( Bousquet 2008 ). However, it is unknown when optimal therapeutic efficacy in non‐allergic rhinitis can be achieved.
Intranasal corticosteroids are likely to work better for the inflammatory endotypes of non‐allergic rhinitis, i.e. NARES and local allergic rhinitis ( Mygind 2001 ).
Why it is important to do this review
Establishing the clinical effectiveness of intranasal corticosteroids in non‐allergic rhinitis could have important clinical implications. Several well‐conducted randomised controlled trials have evaluated intranasal corticosteroids for non‐allergic rhinitis. Most of these studies have small numbers of participants and variations in the included non‐allergic rhinitis phenotypes, as well as variations in the dosages and schedule of intranasal corticosteroid administration. However, there are no reported meta‐analyses on this topic.
This review aims to assess the evidence for the use of intranasal corticosteroids in non‐allergic rhinitis, to define the responsive subgroups and, specifically, to establish the most advantageous dosing and scheduling regimens.
Objectives
To assess the effects of intranasal corticosteroids in the management of non‐allergic rhinitis.
Methods
Criteria for considering studies for this review
Types of studies
We included studies with the following design characteristics:
-
randomised controlled trials (RCTs), including cluster‐randomised and cross‐over trials (cross‐over trials were only to be included if the data from the first phase were available); and
-
patients were followed up for at least two weeks.
We excluded studies with the following design characteristics:
-
randomised patients by side of nose (within‐patient controlled) because it is difficult to ensure that the effects of any interventions considered can be localised; or
-
peri‐operative studies.
Types of participants
Adults and children ≥ 12 years with all phenotypes of non‐allergic rhinitis. We consider patients 12 years of age and above to have the same phenotype as adults. We included studies in which participants with perennial rhinitis were enrolled when it was possible to extract data for those participants with non‐allergic rhinitis.
We excluded studies that included a majority of patients with:
-
allergic rhinitis;
-
infectious rhinitis;
-
acute or chronic rhinosinusitis;
-
auto‐immune rhinitis;
-
rhinitis related to anatomical abnormalities.
Types of interventions
Intervention
We included all intranasal corticosteroids in nasal spray and nasal drops form, at any dose and frequency, and for any duration.
First‐generation intranasal corticosteroids:
-
Beclomethasone dipropionate
-
Triamcinolone acetonide
-
Flunisolide
-
Budesonide
Second‐generation intranasal corticosteroids:
-
Fluticasone furoate
-
Fluticasone propionate
-
Mometasone furoate
-
Betamethasone sodium phosphate
-
Ciclesonide
If other interventions (for example, decongestants) were used, these must have been used equally in all treatment arms.
Comparisons
The comparators were placebo or no intervention or other active treatments.
The main comparison pair was:
-
intranasal corticosteroids versus placebo.
Other possible comparison pairs included:
-
intranasal corticosteroids versus saline irrigation;
-
intranasal corticosteroids versus intranasal antihistamine;
-
intranasal corticosteroids versus capsaicin;
-
intranasal corticosteroids versus sodium cromoglycate;
-
intranasal corticosteroids versus ipratropium.
Types of outcome measures
Primary outcomes
-
Disease severity as measured by patient‐reported symptom score (such as a total nasal symptom score (TNSS) or visual analogue scale (VAS)).
-
Significant adverse event: epistaxis.
Secondary outcomes
-
Disease‐specific health‐related quality of life (using disease‐specific health‐related quality of life questionnaires scores such as the Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ and the Mini Rhinoconjunctivitis Quality of Life Questionnaire (mini‐RQLQ)).
-
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow.
-
Other adverse events: for example, local irritation/discomfort.
Outcomes were measured at follow‐up time points of ≤ 4 weeks and > 4 weeks.
Search methods for identification of studies
The Cochrane ENT Information Specialist conducted systematic searches for randomised controlled trials and controlled clinical trials. There were no language, publication year or publication status restrictions. The date of the search was 1 July 2019.
Electronic searches
The Information Specialist searched:
-
the Cochrane ENT Register (searched via the Cochrane Register of Studies 1 July 2019);
-
the Cochrane Central Register of Controlled Trials (CENTRAL 2019, Issue 7) (searched via the Cochrane Register of Studies 1 July 2019);
-
PubMed (1946 to 1 July 2019);
-
Ovid EMBASE (1974 to 1 July 2019);
-
EBSCO CINAHL (1982 to 1 July 2019);
-
Ovid CAB Abstracts (1910 to 1 July 2019);
-
LILACS (Latin American and Caribbean Health Science Information database), lilacs.bvsalud.org (searched 1 July 2019);
-
Web of Science (1945 to 1 July 2019);
-
ClinicalTrials.gov (searched via the Cochrane Register of Studies and clinicaltrials.gov 1 July 2019);
-
World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP), www.who.int/ictrp (searched 1 July 2019).
The Information Specialist modelled subject strategies for databases on the search strategy designed for CENTRAL. Where appropriate, they were combined with subject strategy adaptations of the highly sensitive search strategy designed by Cochrane for identifying randomised controlled trials and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0, Box 6.4.b. ( Handbook 2011 ). Search strategies for major databases including CENTRAL are provided in Appendix 1 .
Searching other resources
We scanned the reference lists of identified publications for additional trials and contacted trial authors where necessary. In addition, the Information Specialist searched PubMed to retrieve existing systematic reviews relevant to this systematic review, so that we could scan their reference lists for additional trials. The Information Specialist also ran non‐systematic searches of Google Scholar to retrieve grey literature and other sources of potential trials.
We did not perform a separate search for adverse effects of intranasal steroids. We considered adverse effects described in the included studies only.
Data collection and analysis
Selection of studies
We merged the identified studies using the Covidence online reference management software. We removed any duplicate records of the same report.
Two authors (AG and CS, a rhinology fellow and a junior otorhinolaryngology trainee, respectively) independently examined the titles and abstracts of the studies and removed obviously irrelevant reports. We then retrieved the full texts of potentially relevant articles. We linked multiple reports of the same study together. The same two authors independently examined the full‐text reports for compliance of the studies with the eligibility criteria. We contacted the study authors, where appropriate, to clarify study eligibility. The two authors then independently made final decisions on study inclusion. Any disagreements on study inclusion were resolved by discussion. If necessary, disagreement was resolved by arbitration of a third author (KS). We noted the primary reason for exclusion.
Data extraction and management
Two authors (AG and CS) independently extracted the data with a predetermined data collection form ( Appendix 2 ). We piloted the form on a small number of studies to identify any discrepancies in coding. If there were multiple reports of the same study, each author collected data separately from each report and then we collated this into a single study report. Disagreements were again resolved by discussion, with arbitration by a third author (KS) if necessary.
For dichotomous outcomes, we extracted the numbers in each of the two outcome categories in each of the intervention groups, or odds ratio, or risk accompanied by measures of uncertainty (e.g. standard error, 95% confidence interval or an exact P value). For continuous outcomes, we extracted the mean value of the outcome measurements in each intervention group, respective standard deviation and number of participants. If the data were presented in another format, we made appropriate calculations and/or transformations according to the Cochrane Handbook for Systematic Reviews of Interventions ( Handbook 2011 ). We extracted ordinal outcomes and outcomes presented as counts in the form reported in the original studies.
Assessment of risk of bias in included studies
AG and CS undertook assessment of the risk of bias of the included studies independently, with the following taken into consideration, as guided by the Cochrane Handbook for Systematic Reviews of Interventions ( Handbook 2011 ):
-
sequence generation;
-
allocation concealment;
-
blinding;
-
incomplete outcome data;
-
selective outcome reporting; and
-
other sources of bias.
We used the Cochrane 'Risk of bias' tool in RevMan 5 ( RevMan 2014 ), which involved describing each of these domains as reported in the trial and then assigning a judgement about the adequacy of each entry: 'low', 'high' or 'unclear' risk of bias.
Measures of treatment effect
We calculated a weighted treatment effect across studies using RevMan 5 ( RevMan 2014 ). For dichotomous outcomes, we calculated risk ratios (RR) after appropriate conversions. For continuous outcomes, we calculated a mean difference (MD) or a standardised mean difference (SMD) as appropriate. We analysed longer ordinal scales (e.g. visual analogue scale (VAS) scores) as continuous data, using MD or SMD. As suggested in the Cochrane Handbook for Systematic Reviews of Interventions ( Handbook 2011 ), we used standard rules of thumb in the interpretation of SMD effect sizes (SMD, or Cohen's effect size of < 0.41 = small, 0.40 to 0.70 = moderate, > 0.70 = large effect size) ( Cohen 1988 ). We analysed short ordinal scales as dichotomous data (using RR), combining adjacent scores together whenever it was possible to find an appropriate cut‐off point. We treated more frequent count data as continuous. We expressed pooled treatment effects with their 95% confidence intervals (95% CI) for all types of data.
Unit of analysis issues
We determined appropriate units of analysis from the included studies and presented them in the results. We analysed cluster‐randomised trials based on the level of allocation, i.e. clusters of patients. Cross‐over trials were only included if the data from the first phase were available.
Dealing with missing data
We recorded all missing data on the data collection form and reported this in the 'Risk of bias' tables. Whenever possible, we contacted the original investigators to request missing data and information for our risk of bias assessments.
Imputing total symptom scores
We planned to adopt the strategy outlined by Chong et al to deal with missing total disease severity outcomes ( Chong 2016 ). Where a paper did not present information for the total disease severity in terms of patient‐reported symptom scores but did present data for the results of individual symptoms, we used the symptoms rhinorrhoea, blockage and sneezing to calculate a total symptom score. Where mean final values or changes from baseline were presented in the paper for the individual symptoms we summed these to calculate a 'total symptom score'. We calculated standard deviations for the total symptom score as if the symptoms were independent, random variables that were normally distributed. We acknowledge that there is likely to be a degree of correlation between the individual symptoms, however we used this process because the magnitude of correlation between the individual symptoms is not currently well understood (no evidence found). If the correlation is high, the summation of variables as discrete variables is likely to give a conservative estimate of the total variance of the summed final score. If the correlation is low, this method of calculation will underestimate the standard deviation of the total score. However, the average patient‐reported symptom scores have a correlation coefficient of about 0.5; if this is also applicable to non‐allergic rhinitis, the method used should have minimal impact ( Balk 2012 ). As this method of calculation does not take into account weighting of different symptoms (no evidence found), we downgraded all the disease severity outcomes for lack of use of validated scales whenever this occurred.
Assessment of heterogeneity
To assess the heterogeneity of effect size across pooled studies, we calculated the I 2 statistic in RevMan 5. We did not plan to perform a meta‐analysis if heterogeneity was considered substantial (50% to 90%) or considerable (75% to 100%).
Assessment of reporting biases
We had planned to use a funnel plot to detect reporting biases if there were at least 10 studies included in the meta‐analysis and to analyse the visual asymmetry of the plot. However, none of our meta‐analyses included more than 10 studies.
Data synthesis
We used RevMan 5 to perform a meta‐analysis using the random‐effects model if we did not consider the heterogeneity of the included studies to be substantial or considerable. We performed a meta‐analysis of studies that were sufficiently homogenous in terms of participants, treatments and outcome measures. When a meta‐analysis could not be performed due to the level of heterogeneity, we provided a narrative analysis. We analysed the data on an intention‐to‐treat basis using the generic inverse variance method. We made comparisons for all available outcomes between intranasal corticosteroids and no therapy, intranasal corticosteroids and placebo, intranasal corticosteroids and other topical or systemic medications, intranasal corticosteroids and two or more of the above therapies in combination, and between different intranasal corticosteroids regimens (dose, frequency or duration comparisons, if available).
Subgroup analysis and investigation of heterogeneity
We performed subgroup analysis to compare the effects of intranasal corticosteroids:
-
different types of intranasal corticosteroids (type A versus type B).
The following subgroup analyses were planned but not conducted due to insufficient data.
-
different types of non‐allergic rhinitis (e.g. rhinitis medicamentosa, pregnancy rhinitis);
-
different doses of intranasal corticosteroids (dose A versus dose B, e.g. 200 μg versus 400 μg/day budesonide);
-
different regimens of intranasal corticosteroids (regimen A versus regimen B, e.g. once a day versus twice a day);
-
different delivery devices for intranasal corticosteroids (device A versus device B).
Sensitivity analysis
We carried out sensitivity analyses on the basis of the methodological diversity of the included studies. We considered the following factors when repeating the analysis:
-
risk of bias: excluding studies with high risk of bias (defined as four out of seven domains deemed to have high risk).
GRADE and 'Summary of findings' table
Two authors (CS, AG) independently used the GRADE approach to rate the overall certainty of evidence for each outcome using the GDT tool ( https://gradepro.org/ ). The certainty of evidence reflects the extent to which we are confident that an estimate of effect is correct and we applied this in the interpretation of results. There are four possible ratings: 'high', 'moderate', 'low' and 'very low'. A rating of 'high' certainty evidence implies that we are confident in our estimate of effect and that further research is very unlikely to change our confidence in the estimate of effect. A rating of 'very low' certainty implies that any estimate of effect obtained is very uncertain.
The GRADE approach rates evidence from RCTs that do not have serious limitations as high certainty. However, several factors can lead to the downgrading of the evidence to moderate, low or very low. The degree of downgrading is determined by the seriousness of these factors:
-
study limitations (risk of bias);
-
inconsistency;
-
indirectness of evidence;
-
imprecision;
-
publication bias.
The 'Summary of findings' table presents only the outcomes for the main comparison, intranasal corticosteroids versus placebo.
Results
Description of studies
See Characteristics of included studies ; Characteristics of excluded studies .
Results of the search
The results of the search are presented in the study flow diagram in Figure 1 . The search retrieved 17,319 references. We identified no further references by screening the reference lists of studies. We screened and excluded duplicates and obviously irrelevant studies, leaving 6013 studies. After screening of the titles and abstracts of these references, we discarded 5858 studies, leaving 155 references to assess for eligibility. We assessed the full texts of these 155 references. We discarded 77 of these references after full‐text review.
Process for sifting search results and selecting studies for inclusion.
We formally excluded 43 studies with reasons recorded in the review (see Excluded studies ). The most common reason for exclusion was the lack of separate reporting of the non‐allergic rhinitis subpopulation in studies that initially enrolled patients with perennial rhinitis or mixed rhinitis.
We identified one ongoing study ( NCT04002349 ). This study is a randomised, open‐label clinical trial comparing both nasal saline, intranasal corticosteroid, intranasal antihistamine and combination therapy in non‐allergic rhinitis patients. The expected completion date of the study is 31 March 2020. See Characteristics of ongoing studies . No studies are awaiting assessment.
We included 34 studies in the systematic review. Out of these, we were able to include data from 18 studies in our analyses.
Included studies
We included 34 studies in this review ( Arikan 2006 ; Balle 1982 ; Behncke 2006 ; Boechat 2019 ; Blom 1997 ; Day 1990 ; Ellegård 2001 ; Guo 2015 ; Hallén 1997 ; Havas 2002 ; Hillas 1980 ; Incaudo 1980 ; Jacobs 2009 ; Jessen 1990 ; Kalpaklioglu 2010 ; Lin 2017 ; Löfkvist 1976 ; Lundblad 2001 ; Malm 1976 ; Malm 1981 ; Meltzer 1994 ; Miller 1969 ; O'Reilly 1991 ; Scadding 1995 ; Schulz 1978 ; Singh 2017 ; Song 2018 ; Spector 1980 ; Tantilipikorn 2010 ; Tarlo 1977 ; Turkeltaub 1982 ; Varricchio 2011 ; Warland 1982 ; Webb 2002 ). See Characteristics of included studies .
Design
Most of the included studies were randomised ( Arikan 2006 ; Balle 1982 ; Behncke 2006 ; Blom 1997 ; Boechat 2019 ; Day 1990 ; Ellegård 2001 ; Guo 2015 ; Hallén 1997 ; Hillas 1980 ; Jacobs 2009 ; Jessen 1990 ; Kalpaklioglu 2010 ; Lin 2017 ; Lundblad 2001 ; Malm 1981 ; Meltzer 1994 ; Scadding 1995 ; Schulz 1978 ; Singh 2017 ; Song 2018 ; Spector 1980 ; Tantilipikorn 2010 ; Tarlo 1977 ; Turkeltaub 1982 ; Varricchio 2011 ; Warland 1982 ; Webb 2002 ). Two studies were quasi‐randomised ( Havas 2002 ; Miller 1969 ). Randomisation was unclear in four studies ( Incaudo 1980 ; Löfkvist 1976 ; Malm 1976 ; O'Reilly 1991 ).
The majority of the studies used a parallel‐group design ( Arikan 2006 ; Behncke 2006 ; Blom 1997 ; Boechat 2019 ; Day 1990 ; Ellegård 2001 ; Guo 2015 ; Hallén 1997 ; Havas 2002 ; Incaudo 1980 ; Jacobs 2009 ; Kalpaklioglu 2010 ; Lin 2017 ; Lundblad 2001 ; Meltzer 1994 ; Scadding 1995 ; Schulz 1978 ; Singh 2017 ; Song 2018 ; Spector 1980 ; Tantilipikorn 2010 ; Turkeltaub 1982 ; Varricchio 2011 ; Webb 2002 ). Ten studies had cross‐over design ( Balle 1982 ; Hillas 1980 ; Jessen 1990 ; Löfkvist 1976 ; Malm 1976 ; Malm 1981 ; Miller 1969 ; O'Reilly 1991 ; Tarlo 1977 ; Warland 1982 ).
Funding sources were reported in 11 studies, of which 10 were industry‐sponsored ( Ellegård 2001 ; Hallén 1997 ; Hillas 1980 ; Jacobs 2009 ; Lin 2017 ; Lundblad 2001 ; Singh 2017 ; Spector 1980 ; Tantilipikorn 2010 ; Webb 2002 ), and one was government‐sponsored ( Song 2018 ). In another five studies, the industry provided drugs for the study, but no grant support ( Balle 1982 ; Day 1990 ; Havas 2002 ; Löfkvist 1976 ; Malm 1976 ). In five studies, the industry was involved and may have provided medication, but the funding role was unclear ( Incaudo 1980 ; Malm 1981 ; Scadding 1995 ; Schulz 1978 ; Turkeltaub 1982 ). Finally, funding was not reported in 12 studies ( Arikan 2006 ; Behncke 2006 ; Blom 1997 ; Guo 2015 ; Jessen 1990 ; Kalpaklioglu 2010 ; Meltzer 1994 ; Miller 1969 ; O'Reilly 1991 ; Tarlo 1977 ; Varricchio 2011 ; Warland 1982 ). For the other studies it was unclear whether there was any funding.
Conflicts of interest were not clearly reported. In 10 studies, at least one of the authors was an employee of a pharmaceutical company ( Day 1990 ; Ellegård 2001 ; Incaudo 1980 ; Jacobs 2009 ; Malm 1981 ; Scadding 1995 ; Schulz 1978 ; Tantilipikorn 2010 ; Turkeltaub 1982 ; Webb 2002 ). Other conflicts of interest were not reported.
Sample size
Samples sizes ranged from 15 ( Balle 1982 ) to 983 ( Webb 2002 ).
Setting
Most studies took place in secondary or tertiary referral hospital outpatient clinic departments. The countries involved were Australia, Belgium, Brazil, Canada, China, the Czech Republic, Denmark, Finland, France, Germany, Iceland, Ireland, Italy, the Netherlands, New Zealand, Norway, Romania, Thailand, Turkey, Sweden, Switzerland, the UK and the USA.
Participants
There were 4452 patients reported in 34 included studies. The exact number of randomised patients is difficult to assess, given that many studies included both allergic and non‐allergic rhinitis patients and the total number randomised was reported only for the combined population.
Overall, there were more females then males. In 17 studies where information was available 60% were female ( Blom 1997 ; Ellegård 2001 ; Hallén 1997 ; Havas 2002 ; Incaudo 1980 ; Jacobs 2009 ; Jessen 1990 ; Lin 2017 ; Lundblad 2001 ; Löfkvist 1976 ; Malm 1976 ; Malm 1981 ; Miller 1969 ; Singh 2017 ; Spector 1980 ; Tantilipikorn 2010 ; Varricchio 2011 ). In the study Song 2018 the proportion male/female was comparable. In the other studies, the exact proportions were not reported, or were reported for a combined allergic and non‐allergic rhinitis population; in most cases there were more females. Interestingly, Incaudo 1980 was comprised only of male patients. Conversely, Ellegård 2001 was a study of pregnancy rhinitis in females. Behncke 2006 studied rhinitis symptoms in geriatric patients.
Several studies reported mean patient age, which was between 29 and 49 years. Age range also varied by study, for example Boechat 2019 included elderly patients. The overall range was from 9 years ( Miller 1969 ) to 87 years ( Boechat 2019 ). Most patients were between 18 and 70 years of age.
Description of non‐allergic rhinitis in included patients
The majority of studies used a conventional description of rhinitis, by which patients with chronic perennial rhinitis had negative allergy testing. This included the description of non‐allergic rhinitis as vasomotor rhinitis ( Arikan 2006 ; Löfkvist 1976 ; Malm 1976 ; Miller 1969 ; Song 2018 ; Warland 1982 ) and non‐allergic, non‐infectious perennial rhinitis (NANIPER) ( Blom 1997 ). Only a few studies focused on specific subtypes of non‐allergic rhinitis: Hallén 1997 studied rhinitis medicamentosa, Jacobs 2009 investigated a weather and temperature‐sensitive subtype of vasomotor rhinitis, while Tantilipikorn 2010 focused on the irritant subtype due to air pollution, wind/temperature triggers and strong odours. Boechat 2019 focused on senile rhinitis patients (≥ 60 years), with both allergic and non‐allergic rhinitis. Webb 2002 subdivided the overall non‐allergic rhinitis population into NARES and non‐NARES subtypes. Finally, Ellegård 2001 specifically studied pregnancy rhinitis. Interestingly, the majority of studies purposefully excluded pregnant women from their populations. All patients had perennial symptoms. In most studies severity was rated as moderate or severe.
Interventions
Comparisons
Twenty‐five studies compared intranasal corticosteroids with placebo ( Arikan 2006 ; Balle 1982 ; Blom 1997 ; Day 1990 ; Ellegård 2001 ; Hallén 1997 ; Incaudo 1980 ; Jacobs 2009 ; Lin 2017 ; Lundblad 2001 ; Löfkvist 1976 ; Malm 1976 ; Malm 1981 ; Meltzer 1994 ; Miller 1969 ; O'Reilly 1991 ; Scadding 1995 ; Schulz 1978 ; Spector 1980 ; Tantilipikorn 2010 ; Tarlo 1977 ; Turkeltaub 1982 ; Varricchio 2011 ; Warland 1982 ; Webb 2002 ). In all but one of these studies, placebo was described as the inactive vehicle of the intervention medication, or its ingredients were not described. In Varricchio 2011 , isotonic saline solution was used as placebo.
Among these, three studies also compared different doses of intranasal corticosteroids in a multiple‐arm study ( Blom 1997 ; Scadding 1995 ; Webb 2002 ), one study compared different regimens of intranasal corticosteroids ( Blom 1997 ), and one study compared two different types of intranasal corticosteroids ( Scadding 1995 ).
Two studies compared azelastine combined with an intranasal corticosteroid to an intranasal corticosteroid alone ( Guo 2015 ; Song 2018 ). One study compared azelastine combined with fluticasone propionate to placebo ( Singh 2017 ). Another study compared intranasal corticosteroids with capsaicin ( Havas 2002 ).
One study compared intranasal corticosteroids with ipratropium ( Jessen 1990 ). Three studies compared intranasal corticosteroids with intranasal antihistamine ( Behncke 2006 ; Kalpaklioglu 2010 ; Song 2018 ). One study compared intranasal corticosteroids versus saline, versus no treatment and versus intranasal corticosteroids combined with saline ( Lin 2017 ). Another study compared intranasal corticosteroids with sodium cromoglycate ( Hillas 1980 ). One study compared intranasal corticosteroids with azelastine ( Kalpaklioglu 2010 ). Finally, one study compared intranasal corticosteroids with saline to saline alone ( Boechat 2019 ).
Types of steroids
Fluticasone propionate was the most commonly used intranasal corticosteroid and was the main intervention in 10 studies ( Arikan 2006 ; Behncke 2006 ; Blom 1997 ; Ellegård 2001 ; Guo 2015 ; Hallén 1997 ; Meltzer 1994 ; Scadding 1995 ; Singh 2017 ; Webb 2002 ). It was used in total daily doses (calculated as a sum of total dose for both nostrils) of 200 µg ( Arikan 2006 ; Blom 1997 ; Ellegård 2001 ; Hallén 1997 ; Scadding 1995 ; Singh 2017 ; Webb 2002 ) or 400 µg daily ( Blom 1997 ; Scadding 1995 ; Webb 2002 ). Singh 2017 and Guo 2015 used a combination of fluticasone propionate and azelastine. The length of treatment varied from two weeks to three months. Arikan 2006 used treatment for three months; Blom 1997 , Blom 1997 and Ellegård 2001 for eight weeks; Guo 2015 for six weeks; Hallén 1997 and Singh 2017 for two weeks; Scadding 1995 for 12 weeks; and Webb 2002 for four weeks.
Beclomethasone dipropionate was used in seven studies ( Jessen 1990 ; Hillas 1980 ; Löfkvist 1976 ; Malm 1976 ; O'Reilly 1991 ; Scadding 1995 ; Tarlo 1977 ). Daily doses varied from 200 µg to 800 µg per day. Hillas 1980 used 400 µg daily for four weeks. Jessen 1990 used 400 µg daily for two weeks. Löfkvist 1976 used 300 µg daily for four weeks. Malm 1976 used daily doses of 200 µg, 400 µg and 800 µg for two weeks. O'Reilly 1991 used 600 µg per day for 12 weeks. Finally, Scadding 1995 used 200 µg and 400 µg daily for 12 weeks.
Flunisolide nasal spray was used in six studies ( Incaudo 1980 ; Schulz 1978 ; Spector 1980 ; Turkeltaub 1982 ; Varricchio 2011 ; Warland 1982 ). The daily doses ranged from 200 µg to 2 mg per day. Incaudo 1980 used 200 µg per day for six weeks. Schulz 1978 used 300 µg for six weeks. Spector 1980 used 400 µg daily for four weeks. Turkeltaub 1982 used 300 µg daily for 12 weeks. Varricchio 2011 used 2 mg daily for eight weeks, which appears to be at least a five times higher dose compared to the other four studies.
Budesonide was used in five studies ( Balle 1982 ; Day 1990 ; Havas 2002 ; Malm 1981 ; Song 2018 ). The daily doses ranged from 200 µg to 800 µg daily. Balle 1982 used 200 µg and 500 µg daily for two weeks. Day 1990 used 400 µg daily for four weeks. Havas 2002 used a total daily dose of 512 µg for two weeks. Finally, Malm 1981 used 50 µg, 200 µg and 800 µg daily for two weeks.
Fluticasone furoate was used in two studies ( Jacobs 2009 ; Tantilipikorn 2010 ). Both studies used 100 µg once daily for four weeks.
Triamcinolone acetonide was used in Kalpaklioglu 2010 . A total daily dose of 220 µg was used for two weeks.
Mometasone furoate was used in Lundblad 2001 and Boechat 2019 . Lundblad 2001 used a total daily dose of 200 µg for six weeks. Boechat 2019 used a total daily dose of 200 µg for two weeks.
Finally, dexamethasone nasal spray was used in Miller 1969 . A total daily dose of 672 µg or 1008 µg was used (patients used two to three times per day) for one month.
Steroid dosage
Different doses of the same intranasal corticosteroids were used in six studies in addition to the placebo comparison ( Balle 1982 ; Blom 1997 ; Malm 1976 ; Malm 1981 ; Scadding 1995 ; Webb 2002 ). Balle 1982 used budesonide at daily doses of 200 µg and 400 µg in a cross‐over study design. Blom 1997 (parallel‐group study) used fluticasone propionate respectively 200 µg once daily and twice daily in different regimens: a) fluticasone propionate 200 µg once daily and placebo once daily for eight weeks; b) fluticasone propionate 200 µg once daily and placebo once daily for four weeks followed by fluticasone propionate 200 µg twice daily for four weeks; and c) fluticasone propionate 200 µg twice daily for eight weeks. Malm 1976 used 200 µg, 400 µg and 800 µg daily doses of beclomethasone dipropionate in a cross‐over study design. Malm 1981 , in comparison to the previous study, used budesonide at daily doses of 50 µg, 200 µg or 800 µg, also in a cross‐over study design. Webb 2002 (parallel‐group study) used fluticasone propionate respectively at 200 µg and 400 µg daily dosage. Finally, Scadding 1995 used both different doses of fluticasone and another intranasal corticosteroid ‐ beclomethasone. Specifically, they used fluticasone propionate 200 µg once daily, 200 µg twice daily and beclomethasone dipropionate 200 µg twice daily for 12 weeks.
Rescue medication
Some studies allowed for rescue medications to be used concurrently in all study groups ( Day 1990 ; Havas 2002 ; Lundblad 2001 ; Malm 1981 ; Spector 1980 ).
Outcomes
Primary outcomes
Disease severity as measured by patient‐reported symptom score
Thirty‐four studies reported a patient‐reported disease severity score ranging from one symptom to a total nasal symptom score or an overall disease severity score. These scores differed greatly in the method of reporting, ranging from a mean of symptoms to individual scales for up to 10 symptoms. The summary scores were also all constructed differently. A summary of the scales is shown in Table 1 .
|
Study ID |
Symptoms measured |
Score for each symptom |
Summation (total
|
Notes |
|
Nasal obstruction |
Measured on a VAS: 0 to 10, 0 is better |
Completed prior to trial and at 1, 2 and 3 months (range 0 to 10); as only one symptom no summation needed |
No summary data reported to allow inclusion in the meta‐analysis for this outcome |
|
|
1. Obstruction 2. Rhinorrhoea 3. Sneezing |
Scale unclear, low indicates fewer symptoms |
Total scores (for the last 7 days of a 2‐week treatment period) represented the means of scores for the 3 symptoms (range unclear) |
Unclear scale for individual symptoms |
|
|
A. Total nasal score (sum score of blockage, sneezing and rhinorrhoea) 1. Blockage 2. Sneezing 3. Rhinorrhoea B. Overall intensity of total nasal symptoms Does not consist of individual symptoms |
A. Measured on a scale of 0 to 3 (3 means worse) B. Measured on a VAS: 0 to 10 (0 is better) |
A. Presented as mean sum score of 3 symptoms for 1 week (range 0 to 3) B. Measured at 2 weeks pre‐treatment, 4 weeks after first batch of treatment, 8 weeks after treatment (range 0 to 10); overall intensity, therefore no summation of symptoms |
We used B as a total nasal symptom score because a VAS is a more established measurement |
|
|
Combined nasal symptom score (after emailing author):
1. Nasal blockage
|
Measured on a VAS scale of 0 to 10 (10 is worse) |
Measured pre‐intervention and at 2 weeks |
— |
|
|
Total nasal symptom score, unclear definition |
Unclear |
Unclear |
— |
|
|
1. Blocked nose 2. Itchy nose 3. Runny nose 4. Sneezing |
Measured on a scale of 0 to 3 (0 is better) |
Mean change in total combined symptom score (range 0 to 3) from end of treatment to baseline (week 4 versus week 0) |
— |
|
|
Congestion |
Measured on a scale of 0 to 4 (0 is better) |
Reported after 8 weeks of treatment and 16 weeks of post‐treatment follow‐up (range 0 to 4); as only one symptom no summation needed |
We evaluated the data at the end of 8 weeks of treatment as this was the most common method of measurement among other studies |
|
|
Total nasal symptom score; no breakdown in individual symptoms |
Scale unclear |
Measured pre‐intervention and at 2 weeks; range and summation unclear |
— |
|
|
Nasal congestion |
Measured on a VAS of 0 to 100 (0 is better) |
Reported for all days, 0 to 14 days, in the morning and in the evening (range 0 to 100); as only one symptom no summation needed |
No differences between morning and evening data. Reported morning data at day 13. |
|
|
1. Rhinorrhoea 2. Nasal blockage 3. Sneezing 4. Headache 5. Post‐nasal drip 6. Sore throat |
Measured on a VAS of 0 to 5 (0 = no symptoms), separately for each side |
Sum of mean values of rhinorrhoea, nasal blockage and sneezing (range 0 to 30); measurement at end of treatment |
We did not use the reported aggregate relief score, but instead calculated a total nasal symptom score out of rhinorrhoea, nasal blockage and sneezing (mean values and SDs per symptom were provided) |
|
|
Total nasal symptom score 1. Sneezing 2. Rhinorrhoea 3. Nasal pruritis 4. Blocked nose 5. Itchy eyes 6. Watery eyes 7. Red eyes Individual symptom scores |
Both measured on a scale of 0 to 3 (0 = nil, 1 = mild, 2 = moderate and 3 = severe) |
Total nasal symptom score: reported as a mean of 4‐week treatment period |
No separate data reported for non‐allergic rhinitis, only responder/non‐responder data |
|
|
Overall severity of rhinitis |
Measured daily on a scale of 1 to 4 (0 is better) |
Reported mean at end of week 2, 4, 6 and 8 (range 0 to 4) |
We did not use the reported individual rhinitis symptoms to calculate a total nasal symptom score, as only P values were reported |
|
|
1. Congestion 2. Rhinorrhoea 3. Post‐nasal drip |
Measured twice daily on diary cards using a
|
TNSS is sum of the 3 symptom scores (range 0 to 9). Change in TNSS (range 0 to 9) from end of treatment (week 4) to baseline. |
We used daily reflective TNSS and not morning instantaneous TNSS |
|
|
1. Nasal secretion 2. Sneezing 3. Nasal blockage |
All 3 symptoms were measured daily on a scale of 0 to 3/4 (0 is better) |
Reported as sum score of 2‐week treatment period (range 0 to 126/168) |
Unclear whether maximum scale was 3 or 4 |
|
|
1. Rhinorrhoea 2. Congestion 3. Itching 4. Sneezing 5. Anosmia 6. Conjunctivitis |
Scale of 0 to 4 (0 is better) |
Most likely ‐ although not explicitly cited ‐ sum of all individual symptoms. Reported for 2 weeks after treatment. Reported as mean change from baseline. |
— |
|
|
1. Nasal obstruction 2. Nasal itch 3. Rhinorrhoea 4. Sneezing |
Unclear scale |
Unclear total range of symptoms |
— |
|
|
1. Rhinorrhoea 2. Nasal stuffiness/congestion 3. Nasal itching 4. Sneezing |
Measured on a scale of 0 to 3 (0 is better) |
Range 0 to 12; reported most likely as sum at the end of treatment |
Converted into dichotomous outcome: improved versus unimproved. Improvement defined as a reduction of at least 1 point in the overall symptom score. No numerical data on original TNSS, so not included in meta‐analysis for this outcome |
|
|
1. Nasal catarrh 2. Blockage 3. Nasal itching 4. Sneezing |
Measured on a scale of 0 to 3 (0 is better) |
Range 0 to 12 |
No numerical data on original TNSS, therefore not included in meta‐analysis. There are data on 'non‐responder/improvement/responder' |
|
|
1. Nasal obstruction 2. Rhinorrhoea 3. Sneezing 4. Eye irritation |
Measured on a scale of 0 to 3 (0 is better) |
TNSS not reported, but calculated for meta‐analysis: sum of mean values of nasal obstruction, rhinorrhoea and sneezing (range 0 to 9); measurement at end of treatment) |
We calculated a total nasal symptom score out of rhinorrhoea, nasal blockage and sneezing (mean values and SDs per symptom were provided) for the dosages 200 µg, 400 µg and 800 µg |
|
|
1. Nasal obstruction 2. Nasal secretion 3. Sneezing |
Measured on a scale of 0 to 3 (0 is better) Reported as mean ± SEM of at last 3 days of the patients symptom score in each treatment period for nasal obstruction and secretion. For sneezing: scale of 0 to 3 (0 is good: no sneezing = 0; 1 to 5 sneezes = 1 point; 6 to 15 = 2 points; more than 15 sneezes = 3 points) |
TNSS not reported, but calculated for meta‐analysis: sum of mean values of nasal obstruction, rhinorrhoea and sneezing (range 0 to 9) |
We calculated a total nasal symptom score out of rhinorrhoea, nasal blockage and sneezing (mean values and SDs per symptom were provided) for the dosages 50 µg, 200 µg and 800 µg |
|
|
Nasal symptom score, unclear definition |
Unclear |
Unclear |
Unclear |
|
|
1. Nasal obstruction 2. Discharge 3. Post‐nasal drip 4. Sneezing |
Measured by patient at 2 and 4 weeks after treatment on a scale of 0 to 3 (0 is better) |
Not reported |
We were unable to calculate a SD for rhinorrhoea (secretion) and post‐nasal drip as the P values for these symptoms were not reported, therefore not included in meta‐analysis for this outcome |
|
|
1. Nasal obstruction 2. Anterior rhinorrhoea 3. Posterior rhinorrhoea 4. Sneezing 5. Facial pain |
Measured on a scale of 0 to 5 (0 is better) |
Composite score for all 5 symptoms, range 0 to 25 |
Only P values reported, therefore not included in meta‐analysis for this outcome |
|
|
A. 1. Nasal blockage on waking 2. Nasal blockage during the rest of the day 3. Sneezing 4. Rhinorrhoea B. 1. Overall assessment of symptoms by patient 2. Overall assessment of symptoms at clinic visit |
A. Measured by patient on a scale of 0 to 3 on a daily card (0 is better) B. 1 . Measured by patient on a scale of 0 to 3 on a daily card (0 is better) 2. VAS, 0 to 10 cm (10 = worst symptoms) |
A. Range 0 to 12 B.1. Range 0 to 3 B.2. Range 0 to 10 |
No data reported for non‐allergic rhinitis subgroup separately, therefore not included in meta‐analysis. The study states that there were no differences between allergic and non‐allergic rhinitis patients |
|
|
1. Sneezing 2. Stuffy nose 3. Runny nose 4. Nose blowing 5. Post‐nasal drip |
Duration in hours per symptom measured |
All 5 symptoms combined to determine overall duration of patients' symptoms. Reported as percentage of days during which all 5 of the symptoms lasted 1 hour or less and the percentage of days during which at least one of the 5 symptoms lasted 4 hours or more |
No data reported for non‐allergic rhinitis subgroup separately for this outcome, therefore not included in meta‐analysis. There are data on 'responder/non‐responder'. |
|
|
— |
— |
— |
Not included in meta‐analysis as TNSS data are reported in relationship to cold dry air provocation |
|
|
A. Overall VAS B. Individual symptom VAS: 1. Congestion 2. Sneezing 3. Itching 4. Rhinorrhoea |
Measured on a VAS: 0 to 10, 0 is better |
Range (0 to 10), measurement at week 8 |
— |
|
|
1. Sneezing 2. Stuffiness 3. Runny nose 4. Nose blowing 5. Post‐nasal drip |
Patient evaluation, sum of 5 symptoms numerically assessed as absent (1), mild (2), moderate (3), or severe (4); range 1 to 4 |
Range (5 to 20), measurement at week 4 |
— |
|
|
1. Rhinorrhoea 2. Nasal congestion 3. Post‐nasal drip 4. Eye itching/burning 5. Eye tearing/watering 6. Eye redness |
4‐point categorical scale of 0 to 3 (none, mild, moderate, severe), measured by patient on paper diary card, in AM and PM, as instantaneous (i) and over previous 12 hours (reflective). Instantaneous score measured in AM, and reflective in both AM and PM ‐ measured during screening and treatment periods |
Combined 3 reflective individual nasal symptom scores (range 0 to 9) Measured in AM and PM, which were averaged to arrive at the final daily value (daily rTNSS) |
Compared data from week 4 to baseline to arrive at change from baseline TNSS. We included the reflective TNSS (rTNSS) in the meta‐analysis, not the instantaneous TNSS (iTNSS). |
|
|
Individual rhinitis symptoms: 1. Daily and nightly sneezing 2. Nasal congestion 3. Rhinorrhoea Total nasal symptoms |
0 if absent, 1 if they lasted less than 30 minutes, 2 if between 30 minutes and 2 hours, and 3 if longer than 2 hours |
Total nasal symptoms: the sum of the nasal symptom scores (sneezing, congestion, and rhinorrhoea, as well as the total) |
No separate data for non‐allergic rhinitis subgroups |
|
|
1. Sneezing 2. Runny nose 3. Stuffy nose 4. Eye itch 5. Throat itch |
Measured on 0 to 6 scale (0 is better) |
Sum score of symptoms scores for sneezing, runny nose, stuffy nose, eye itching and throat itching measured post‐treatment, each measured on 0 to 6 scale (range 0 to 30). To this was added the number of tablets and nasal sprays required to control nasal symptoms for the preceding 12‐hour period; possible scores range from 0 to 40 |
Unusual method of measurement of TNSS |
|
|
1. Nasal itching 2. Sneezing 3. Rhinorrhoea 4. Nasal obstruction |
Measured on 0 to 3 scale (0 is better) |
Range 0 to 12; assessed on day 1 and at 8 weeks |
Reported TNSS as continuous data and change from baseline. We have included the first approach |
|
|
Overall severity of symptoms Individual rhinitis symptoms: 1. Sneezing 2. Runny nose 3. Nose blowing 4. Post‐nasal drip |
Overall severity of symptoms: scale from 0 to 3 Individual nasal symptoms: scale from 0 to 4 |
— |
— |
|
|
1. Nasal obstruction
2. Post‐nasal drip
|
Measured on a VAS, 0 to 100 (0 is better) |
Range 0 to 300, measured at both 2 weeks and 4 weeks |
TNSS reported as change from baseline. Combination of 3 studies |
TNSS: total nasal symptom score
iTNSS: instantaneous total nasal symptom score
rTNSS: reflective total nasal symptom score
SD: standard deviation
The individual symptom scores varied and included nasal obstruction, nasal congestion, rhinorrhoea, post‐nasal drip, sneezing, itchy nose, facial pain, anosmia, itchy eyes, watery or red eyes, headache, cough, mucus production and sore or itchy throat. These were most commonly measured on a scale ranging from 0 to 3 to 0 to 6, or a visual analogue scale (VAS) ranging from 0 to 5 to 0 to 100.
We used the symptoms rhinorrhoea (secretion), congestion (obstruction) and sneezing to calculate a total nasal symptom score in cases where only individual symptom scores were reported.
The majority of studies reported an overall symptom score ( Balle 1982 ; Blom 1997 ; Boechat 2019 ; Day 1990 ; Guo 2015 ; Havas 2002 ; Incaudo 1980 ; Jacobs 2009 ; Kalpaklioglu 2010 ; Löfkvist 1976 ; O'Reilly 1991 ; Scadding 1995 ; Schulz 1978 ; Song 2018 ; Tantilipikorn 2010 ; Turkeltaub 1982 ; Varricchio 2011 ; Webb 2002 ). Most studies combined individual symptom scores into a sum score of total nasal symptom score ( Blom 1997 ; Day 1990 ; Havas 2002 ; Jacobs 2009 ; Löfkvist 1976 ; O'Reilly 1991 ; Schulz 1978 ; Tantilipikorn 2010 ; Turkeltaub 1982 ; Varricchio 2011 ; Webb 2002 ). Balle 1982 used a mean of individual symptom scores. Blom 1997 measured intensity of nasal symptoms on a VAS from 0 to 10. Boechat 2019 and Song 2018 measured a combined nasal symptom score on a VAS from 0 to 10. Incaudo 1980 assessed overall severity of rhinitis on a scale of 1 to 4. Kalpaklioglu 2010 evaluated a total nasal symptom score on a scale of 0 to 4. Finally, Scadding 1995 reported overall assessment of symptoms by patients on a scale of 0 to 3, and at clinic visits on a VAS of 0 to 10.
Significant adverse events: epistaxis
Eight studies reported on the significant adverse event 'epistaxis' ( Arikan 2006 ; Incaudo 1980 ; Jacobs 2009 ; Lin 2017 ; Lundblad 2001 ; Malm 1981 ; Scadding 1995 ; Tantilipikorn 2010 ). In almost all studies this adverse event was reported as the number of cases of epistaxis at the end of follow‐up, either actively asked for by the investigator and/or spontaneously reported by the patient or recorded daily in a diary or on a question form ( Jacobs 2009 ). Scadding 1995 only mentioned "generally minor adverse events" and reported no explicit numbers.
The risk of epistaxis was reported in five studies included in the meta‐analysis ( Incaudo 1980 ; Jacobs 2009 ; Lundblad 2001 ; Malm 1981 ; Tantilipikorn 2010 ).
Secondary outcomes
Disease‐specific health‐related quality of life
Six studies measured quality of life ( Behncke 2006 ; Boechat 2019 ; Kalpaklioglu 2010 ; Lin 2017 ; Lundblad 2001 ; Song 2018 ). Behncke 2006 used the Rhinitis Quality of Life Questionnaire (RQLQ) (but reported no numerical data on the non‐allergic rhinitis group separately). Boechat 2019 used the SNOT‐22. Kalpaklioglu 2010 used the mini‐Rhinitis Quality of Life Questionnaire (mini‐RQLQ) (but reported no numerical data on the non‐allergic rhinitis group separately). Lin 2017 used the SF‐12v2 and Lundblad 2001 did not report on how quality of life was measured. Song 2018 used the SF12‐v2 to measure quality of life. Boechat 2019 and Song 2018 were included in our analyses.
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
Ten studies objectively measured nasal airflow ( Boechat 2019 ; Ellegård 2001 ; Hallén 1997 ; Jessen 1990 ; Kalpaklioglu 2010 ; Malm 1981 ; O'Reilly 1991 ; Singh 2017 ; Spector 1980 ; Tarlo 1977 ). Boechat 2019 measured peak nasal inspiratory flow (PNIF) (L/min). Ellegård 2001 measured a blockage index ((PEF‐nPEF)/PEF) and acoustic rhinometry. Hallén 1997 measured rhinostereometry, acoustic rhinometry (MCA2 area) and PNIF (L/min). Jessen 1990 measured rhinomanometry during inclusion of patients but it was not used to objectively measure airflow after treatment. Kalpaklioglu 2010 measured nPIFR (nasal peak inspiratory flow rate). Malm 1981 measured rhinomanometry (in degrees). O'Reilly 1991 measured rhinomanometry using the Brom's method. Singh 2017 used the minimal cross‐sectional area (MCA) before and after cold dry air (CDA) provocation. Spector 1980 used the nasal peak expiratory flow rate (PEFRn), the mouth peak expiratory flow rate (PEFRm) and the blockage index. Tarlo 1977 measured nasal airway resistance used the method of Taylor and Shivalkar.
Only three studies provided numerical data for objective airway measurements for non‐allergic rhinitis patients that we could use in our analysis ( Boechat 2019 ; Malm 1981 ; Spector 1980 ). The other studies assessed another comparison than intranasal corticosteroids versus placebo or reported no numerical data for the non‐allergic rhinitis subgroup.
Other adverse events: local irritation, discomfort
Nineteen studies included 'adverse events' (besides epistaxis) as an outcome ( Arikan 2006 ; Day 1990 ; Incaudo 1980 ; Jacobs 2009 ; Jessen 1990 ; Kalpaklioglu 2010 ; Lin 2017 ; Lundblad 2001 ; Malm 1976 ; Malm 1981 ; Miller 1969 ; O'Reilly 1991 ; Scadding 1995 ; Song 2018 ; Spector 1980 ; Tantilipikorn 2010 ; Tarlo 1977 ; Turkeltaub 1982 ; Varricchio 2011 ). In almost all studies these adverse events were reported as the number of cases of adverse events at the end of follow‐up, either actively asked for by the investigator and/or spontaneously or recorded daily by the patient in a diary or on a question form ( Day 1990 ; Jacobs 2009 ). Of these studies four were included in the meta‐analysis ( Jacobs 2009 ; Lundblad 2001 ; Song 2018 ; Tantilipikorn 2010 ). The studies that were not included in the meta‐analysis either did not report data on non‐allergic rhinitis patients separately, did not report numerical data or were excluded from the meta‐analysis for other reasons (for example, too low or too high an intranasal corticosteroid dosage or unclear dosage subgroup).
Excluded studies
In total we excluded 43 studies (see Characteristics of excluded studies ).
Astafieva 2012 compared two types of intranasal corticosteroids, brand versus generic, which was not a comparison included in our protocol and therefore we excluded this study.
Celiker 2011 compared intranasal corticosteroids with radiofrequency ablation of the inferior turbinate for nasal obstruction. It was excluded because the comparison intranasal corticosteroids versus radiofrequency ablation was not defined in our protocol.
We excluded studies with high risks of bias such as Synnerstad 1996 . Besides obvious high risks of bias for allocation concealment, blinding of participants and personnel, and blinding of outcome assessors, the study had minor issues with incomplete outcome data, and some with selective outcome reporting (individual nasal symptoms measured but not thoroughly reported, total nasal symptoms reported but not included in methods). This study was supported by a grant from Astra Draco AB, Lund, Sweden, and the second author worked for the company. The company provided budesonide (Rhinocort). The study suggested that budesonide was better than beclomethasone. There are significant grounds to suspect high risk of bias. Based on these observations, we decided to exclude this study.
The Small 1982 study (comparing beclomethasone with placebo in non‐allergic rhinitis patients) did not provide the results for the placebo group and was therefore excluded.
We also excluded 38 studies that were performed in patients with perennial rhinitis and did not present results for the non‐allergic rhinitis subgroup separately ( Adamopoulos 1995 ; Arbesman 1983 ; Balle 1982b ; Basran 1995 ; Berger 2012 ; Bernstein 1997 ; Blair 1977 ; Bunnag 1992 ; Chatterjee 1974 ; Dieges 1978 ; Dockhorn 1999 ; Gibson 1974 ; Hansen 1974 ; Harding 1976 ; Hartley 1985 ; Haye 1993 ; Jones 1979 ; Joubert 1983 ; Juniper 1993 ; Kakumanu 2003 ; Kivisaari 1998 ; Kohan 1989 ; Lahdensuo 1977 ; Lau 1990 ; Lebowitz 1993 ; Malmberg 1988 ; McAllen 1969 ; McAllen 1980 ; Negreiros 1975 ; Price 2013 ; Rusnak 1981 ; Scadding 1991 ; Shaw 1979 ; Svendsen 1989 ; Sy 1979 ; Turner Warwick 1980 ; Webb 1977 ; Weckx 2001 ; Wight 1992 ). We contacted the authors of the studies in an attempt to obtain these results, without success.
One excluded study was a meta‐analysis with the only relevant study already included in our review ( Zucker 2019 ).
Besides the 43 excluded studies, two other studies did not present results for the non‐allergic rhinitis subgroup separately but also did not have any authors listed. These studies were considered 'discarded'.
Risk of bias in included studies
We included 34 studies in this review. Our judgements about risk of bias are presented as a 'Risk of bias' graph in percentage form for all included studies combined ( Figure 2 ). The risk of bias in individual studies in shown in a 'Risk of bias' summary ( Figure 3 ).
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Most studies described a random component in the sequence generation process but with no more information, so we judged them to have an unclear risk of bias. The exceptions are Havas 2002 and Miller 1969 , which had a high risk of bias due to pseudo‐randomisation and quasi‐randomisation. Incaudo 1980 , Löfkvist 1976 , Malm 1976 and O'Reilly 1991 also have a high risk of bias because they did not describe randomisation at all although the study type is very suggestive of a randomised trial. Boechat 2019 (randomisation by a computer‐generated code), Day 1990 (balanced and stratified randomisation), Lin 2017 (computer software) and Song 2018 (number table method) have a low risk of selection bias.
Allocation concealment was unclear in most studies, with the exception of Havas 2002 , which had a high risk of bias (pseudo‐randomisation), Lin 2017 (had a non‐random component: day/order of admission) and Varricchio 2011 (allocation was not concealed, single‐blinded study). In addition, Incaudo 1980 , Löfkvist 1976 , Malm 1976 and O'Reilly 1991 also had a high risk of bias as they did not describe randomisation at all although the study type is very suggestive of a randomised trial. Miller 1969 had a low risk of bias for allocation concealment as the authors described allocation concealment in detail (i.e. "over‐printed on a tear‐off portion of the label which was attached to the case report form").
Blinding
Most studies reported blinding of patients and physicians but did not give more information on the blinding process so had an unclear risk of bias. Arikan 2006 had a low risk of bias as one of the main outcomes (CT scoring) was at low risk because of blinding of the radiologist. Boechat 2019 , Havas 2002 , Lin 2017 , Singh 2017 , Song 2018 and Varricchio 2011 had high risk of bias for blinding either because of no reporting of blinding and different treatment strategies per group making blinding complicated, pseudo‐randomisation, no randomisation or single‐blinding of the study.
Incomplete outcome data
Nineteen studies had a low risk of attrition bias because data for all included participants were reported. In 10 studies, the risk of attrition bias was high due to incomplete outcome data reporting or violation of the intention‐to‐treat protocol ( Behncke 2006 ; Jessen 1990 ; Malm 1981 ; Meltzer 1994 ; O'Reilly 1991 ; Scadding 1995 ; Spector 1980 ; Tarlo 1977 ; Turkeltaub 1982 ; Warland 1982 ). In eight studies, the risk of attrition bias was unclear because only a very small amount of data was not reported and this had an unclear (and most likely low) effect on clinical outcome ( Ellegård 2001 ; Hallén 1997 ; Hillas 1980 ; Jacobs 2009 ; Löfkvist 1976 ; Schulz 1978 ; Singh 2017 ; Webb 2002 ).
Selective reporting
Fifteen studies had a low risk of selective reporting bias because all of the outcomes described in the methods section could be found in the results. We were not able to find a study protocol for any of the included studies.
In eight studies the risk of reporting bias was unclear due to incomplete presentation of all outcomes ( Balle 1982 ; Hillas 1980 ; Jessen 1990 ; Lin 2017 ; Löfkvist 1976 ; Scadding 1995 ; Schulz 1978 ; Webb 2002 ). In the remaining 12 studies the risk of selective reporting bias was high due to major lack of reporting of significant outcomes, which could influence the conclusions ( Behncke 2006 ; Blom 1997 ; Ellegård 2001 ; Guo 2015 ; Lundblad 2001 ; Malm 1976 ; Meltzer 1994 ; O'Reilly 1991 ; Singh 2017 ; Tarlo 1977 ; Warland 1982 ).
Other potential sources of bias
The risk of other bias was high in four studies ( Incaudo 1980 ; Jessen 1990 ; Malm 1976 ; Spector 1980 ). Incaudo 1980 included only male patients. In Jessen 1990 , it was unclear if blinding was compromised for patients to report medication safety. In addition, the scale up to "3 or 4 for severe symptoms" is vague. Finally, it was not clear which groups the patients (5 of 24) co‐treated with xylometazoline belonged to. In Malm 1976 , the cross‐over study design had no washout period, leaving it possible for there to be a carry‐over effect. In Spector 1980 , women of childbearing potential were excluded, making the study biased.
Several studies received funding from a pharmaceutical company without clarifying their role. Another extra bias in some studies resulted from limited ways of reporting data, for example without mean and standard deviation. Only 15 studies had a low risk of other potential sources of bias ( Boechat 2019 ; Blom 1997 ; Ellegård 2001 ; Guo 2015 ; Havas 2002 ; Hillas 1980 ; Jacobs 2009 ; Kalpaklioglu 2010 ; Lin 2017 ; O'Reilly 1991 ; Schulz 1978 ; Song 2018 ; Tantilipikorn 2010 ; Varricchio 2011 ).
Effects of interventions
See summary of findings Table for the main comparison for the main comparison: 'Intranasal corticosteroids versus placebo'.
Intranasal corticosteroids versus placebo
Thirteen studies (2045 participants) comparing intranasal corticosteroid treatment with placebo provided data that could be used in our analyses ( Arikan 2006 ; Balle 1982 ; Blom 1997 ; Day 1990 ; Incaudo 1980 ; Jacobs 2009 ; Lundblad 2001 ; Malm 1976 ; Malm 1981 ; Spector 1980 ; Tantilipikorn 2010 ; Turkeltaub 1982 ; Webb 2002 ). Twelve included studies could not be used in the analyses ( Ellegård 2001 ; Hallén 1997 ; Lin 2017 ; Löfkvist 1976 ; Meltzer 1994 ; Miller 1969 ; O'Reilly 1991 ; Scadding 1995 ; Schulz 1978 ; Tarlo 1977 ; Varricchio 2011 ; Warland 1982 ).
Different types of intranasal corticosteroids were used (budesonide, beclomethasone, flunisolide, fluticasone propionate, fluticasone furoate, dexamethasone, mometasone furoate).
Among the studies treatment dosage varied from 50 µg to 2000 µg daily. Most of the studies that compared different dosages of intranasal corticosteroids used a cross‐over study design, with the exception of Blom 1997 and Webb 2002 , which used a parallel‐group study design. In the cross‐over studies the same patients were treated with different dosages of intranasal corticosteroids, with a short (one‐week) or no washout, complicating a clear comparison between these dosage subgroups ( Balle 1982 ; Malm 1976 ; Malm 1981 ). Only Balle 1982 showed a dosage effect for two nasal symptom score outcomes. Malm 1976 and Malm 1981 showed no significant difference between the dosage subgroups. The two parallel‐group studies both concluded that there were no statistically significant differences among the different intranasal corticosteroid dosage subgroups ( Blom 1997 ; Webb 2002 ). In the parallel‐group studies different dosage subgroups contained different patients but were compared with the same control group. To prevent counting the same patients or controls more than once, we decided to include one intranasal corticosteroids dosage in the meta‐analysis. The most common intranasal corticosteroid dosage was 200 µg. A test for subgroup differences showed no significant difference ('no dosage effect') between 200 µg and 400 µg. We therefore included studies in the meta‐analysis with an intranasal corticosteroid dosage range of 200 µg to 400 µg.
Treatment vehicles varied and included spray, aerosol, nebuliser, pressured canister and atomised bottle. Frequency of usage varied from once daily to four times daily.
Disease severity, as measured by patient‐reported total nasal symptom score
Eleven studies presented data for disease severity using a number of different scales that could be used in meta‐analysis ( Balle 1982 ; Blom 1997 ; Day 1990 ; Incaudo 1980 ; Jacobs 2009 ; Malm 1976 ; Malm 1981 ; Spector 1980 ; Tantilipikorn 2010 ; Turkeltaub 1982 ; Webb 2002 ). Table 1 shows the different scales used. Some studies provided us with a total nasal symptom score (TNSS). In studies that did not provide a total nasal symptom score, we calculated this score based on individual rhinitis symptom scores, i.e. rhinorrhoea (secretion), congestion (obstruction) and sneezing. Due to the differences in the scales used, we used a standardised mean difference (SMD) in the analysis.
Outcomes were measured at up to four weeks follow‐up in four studies and at more than four weeks (six weeks to three months) follow‐up in three studies. Outcomes were also measured as change from baseline in another four studies.
Up to four weeks follow‐up
We were able to pool data from four studies that reported a patient‐reported total nasal symptom score (or individual scores that could be calculated into a total nasal symptom score) with a follow‐up of up to four weeks ( Balle 1982 ; Malm 1976 ; Malm 1981 ; Spector 1980 ). These studies showed that patients treated with intranasal corticosteroids had lower total nasal symptom scores compared to placebo (SMD ‐0.74, 95% confidence interval (CI) ‐1.15 to ‐0.33; 4 studies; 131 participants; I 2 = 22%) ( Analysis 1.1 ) (low‐certainty evidence). This represents a medium effect size ( Cohen 1988 ). Spector 1980 was the only study that did not report an improvement of total nasal symptom score with intranasal corticosteroids.
The heterogeneity in this analysis is mainly the result of Spector 1980 . Removing this study reduces the heterogeneity to 0%.
There were not enough data to carry out our planned subgroup analyses to assess the differences between different dosages (see above), types, vehicles or frequencies of intranasal corticosteroid treatment.
More than four weeks follow‐up (six weeks to three months)
Three studies reported a patient‐reported total nasal symptom score with a follow‐up of more than four weeks ( Blom 1997 ; Incaudo 1980 ; Turkeltaub 1982 ). The follow‐up period varied between six weeks and three months.
These studies showed that patients treated with intranasal corticosteroids had no difference in nasal symptom scores compared to placebo but the evidence is very uncertain (SMD ‐0.24, 95% CI ‐0.67 to 0.20; 3 studies; 85 participants; I 2 = 0%) ( Analysis 1.2 ) (very low‐certainty evidence).
Blom 1997 studied four different treatment regimens with different intranasal corticosteroid dosages. The authors concluded that there were no statistically significant differences among the four treatment regimens in the investigators' assessments of symptoms and rhinoscopy at clinic visits.
There were not enough data to carry out our planned subgroup analyses to assess the differences between different dosages (see above), types, vehicles or frequencies of intranasal corticosteroid treatment.
Change from baseline, up to four weeks follow‐up
Four studies reported on the change from baseline of a patient‐reported total nasal symptom score, with a follow‐up of up to four weeks ( Day 1990 ; Jacobs 2009 ; Tantilipikorn 2010 ; Webb 2002 ).
These studies showed that patients treated with intranasal corticosteroids had a significant difference in total nasal symptom scores compared to placebo (SMD ‐0.15 95% CI ‐0.25 to ‐0.05; P = 0.004; 4 studies; 1465 participants; I 2 = 35%) ( Analysis 1.3 ) (low‐certainty evidence). This represents a small effect size.
These results use adjusted data from Jacobs 2009 that are different from the original results presented and published by the study authors. The Jacobs 2009 study reports a very unlikely standard deviation (SD) value that does not match with the presented means, n and P values. Using the data presented in Jacobs 2009 resulted in very high heterogeneity (I 2 = 96%). The data would make more sense if the presented SD values were actually standard error of the mean (SEM), which we confirmed by a re‐analysis. We were not able to obtain a response from the study authors to our questions about the data. Jacobs 2009 is one of the larger studies and is also one of the most well‐known and frequently cited. We therefore decided to retain the study in the meta‐analysis. Using the formula SD = SEM * SQRT (n) we calculated the adjusted SD values assuming the SD values presented in the study were actually SEM. We used these adjusted SD values in Analysis 1.3 .
Webb 2002 studied two daily dosages (200 µg and 400 µg) in a parallel‐group study, with nearly the same effect on total nasal symptom score change from baseline. Only the highest dosage (400 µg) from Webb 2002 was included in the meta‐analysis.
Webb 2002 also reports an improvement in favour of intranasal corticosteroids, however this is less certain than in Jacobs 2009 . The amount of improvement is more clinically relevant than in Jacobs 2009 , i.e. around a 10% improvement in total nasal symptom score. In general, the data from Webb 2002 seem to be far more reliable than the data from Jacobs 2009 .
Webb 2002 studied different daily dosages (200 µg and 400 µg) and concluded that there were no statistically significant differences.
There were not enough data to carry out our planned subgroup analyses to assess the differences between different dosages (see above), types, vehicles or frequencies of intranasal corticosteroids treatment.
For the outcome total nasal symptom score change from baseline there were no studies reporting a follow‐up of more than four weeks.
Twelve studies that did report nasal symptom score(s) could not be included in the meta‐analysis ( Arikan 2006 ; Meltzer 1994 ; Miller 1969 ; Lin 2017 ; Lundblad 2001 ; Löfkvist 1976 ; O'Reilly 1991 ; Scadding 1995 ; Schulz 1978 ; Tarlo 1977 ; Varricchio 2011 ; Warland 1982 ). Arikan 2006 and Lundblad 2001 are, however, included in the meta‐analysis for other outcomes (adverse events). Miller 1969 and Varricchio 2011 are not included in the total nasal symptom score(s) meta‐analysis because they used an intranasal corticosteroid dosage higher than 200 µg to 400 µg daily. See Table 2 for a summary of the results from these 12 studies for nasal symptom score(s).
|
Study |
Findings |
|
Concluded that treatment with fluticasone propionate provided significantly greater relief from the symptom of nasal obstruction compared with placebo over the entire 3‐month treatment period. Patients' subjective assessments of nasal obstruction after medical treatment correlated with the results of objective testing. |
|
|
This study was not included in the meta‐analysis because of lack of quality of the study data. Firstly, the study presents unexpected data, with disappearance of the benefit of intranasal corticosteroids with longer follow‐up. Secondly, including the study in the meta‐analysis resulted in a high level of heterogeneity. The SD values that are presented in the study do not match with the presented means, n and P values. The data make more sense if the as‐presented SD values should actually be standard error of the mean (SEM), which was confirmed by a re‐analysis. As the authors did not reply to our question regarding the above, we decided to not include this study in the meta‐analysis. The study did show a beneficial effect of intranasal corticosteroids over placebo, however this effect disappeared with longer follow‐up. |
|
|
This study was not included in the meta‐analysis for this outcome as it did not report numerical data for the non‐allergic rhinitis subgroup. They did conclude that fluticasone propionate reduces total symptoms, improves individual symptoms (mainly obstruction) and achieves a significant overall improvement in non‐allergic rhinitis compared to placebo. |
|
|
Reported a statistically significant difference in symptoms in favour of intranasal corticosteroids (it did not report P values for rhinorrhoea and post‐nasal drip so we were not able to calculate a SD). |
|
|
Reported no numerical data on original TNSS, so it could not be included in the meta‐analysis for this outcome. It did report data that could be translated into proportions of 'responders/non‐responders'. The study did not find significant differences between intranasal corticosteroids and placebo. The study converted TNSS into a dichotomous outcome: improved versus unimproved. Improvement was defined as a reduction of at least 1 point in the overall symptom score. No numerical data on original TNSS was provided, therefore this study was not included in the meta‐analysis for this outcome. |
|
|
This study was not included in the meta‐analysis as no data on TNSS were reported. The study did report data on 'responders/non‐responders' with 29/39 responders in the intranasal corticosteroids group and 12/39 responders in the placebo group, favouring intranasal corticosteroids with an odds ratio (OR) of 0.44 (95% confidence interval (CI) 0.24 to 0.64). |
|
|
This study was not included in the meta‐analysis as only P values were reported. Patients reported subjective symptom scores on a scale of 0 to 5 for nasal obstruction, anterior rhinorrhoea, posterior rhinorrhoea, sneezing and facial pain. When the composite scores for all 5 symptoms were compared, there was a significant difference between beclomethasone dipropionate and baseline (P = 0.01) and beclomethasone dipropionate and placebo (P = 0.02) in favour of beclomethasone dipropionate. |
|
|
This study on 2 types of intranasal corticosteroids versus placebo in perennial rhinitis patients (allergic and non‐allergic rhinitis) reported a number of individual rhinitis symptoms and an overall assessment of symptoms, but no separate data on non‐allergic rhinitis patients were presented. However, the study does state that there were no differences between allergic and non‐allergic rhinitis. This study reported a significant improvement with intranasal corticosteroids versus placebo in perennial allergic rhinitis, with fluticasone propionate aqueous nasal spray (200 µg) as effective as beclomethasone dipropionate µg twice daily. |
|
|
This study was not included in the meta‐analysis as no data on TNSS were reported. The study did report data on responders/non‐responders with 6 of 14 responders in the intranasal corticosteroids group and 8 of 18 responders in the placebo group with an OR of 0.02 (95% CI 0.33 to 0.36), therefore not a significant difference. |
|
|
This study was not included in the meta‐analysis for this outcome as it did not report enough numerical data for the non‐allergic rhinitis subgroup. They concluded that after 6 months 6 of 9 non‐allergic rhinitis patients were successfully treated with intranasal corticosteroids and 3 of 9 non‐allergic rhinitis patients were unsuccessfully treated with intranasal corticosteroids. They concluded that their results (in favour of intranasal corticosteroids over placebo) in those in whom a possible allergic component could be identified were not different from those of the whole group. |
|
|
This study was not included in the meta‐analysis because we decided to only include studies with an intranasal corticosteroid dosage of 200 µg to 400 µg. This study uses an intranasal corticosteroid dosage of 2000 µg. The study did report a significant improvement in nasal symptoms in non‐allergic rhinitis after an 8‐week treatment period with intranasal flunisolide. |
|
|
This study was not included in the meta‐analysis for this outcome because it did not report numerical data for the non‐allergic rhinitis subgroup. They concluded that flunisolide nasal solution seems to be effective in both allergic rhinitis and vasomotor rhinitis patients, although it seems to be more effective in an allergic state. |
SD: standard deviation
TNSS: total nasal symptom score
Significant adverse events: epistaxis
The risk of epistaxis was reported in four studies included in the meta‐analysis, two with a follow‐up of up to four weeks ( Malm 1981 ; Tantilipikorn 2010 ) and two with a follow‐up of more than four weeks ( Jacobs 2009 ; Lundblad 2001 ). Malm 1981 studied different dosages of intranasal corticosteroids in a cross‐over study design. The daily dosage of 200 µg was included in the meta‐analysis (see reasons above).
We decided to combine the four studies and not to separate them into up to four weeks and more than four weeks follow‐up. All studies showed a significantly higher risk of epistaxis in the intranasal corticosteroids group compared to placebo (risk ratio (RR) 2.10, 95% CI 1.24 to 3.57; 1174 participants; 4 studies; I 2 = 0%) (moderate‐certainty evidence). The absolute risk difference for epistaxis was 0.04 ( Analysis 1.4 ), with a number needed to treat for an additional harmful outcome (NNTH) of 25 (95% CI 16.7 to 100).
Three of the studies included in the meta‐analysis that reported on the risk of epistaxis showed no significant difference between intranasal corticosteroids and placebo ( Jacobs 2009 ; Malm 1981 ; Tantilipikorn 2010 ). For these studies the NNT, NNTB (number needed to treat for an additional beneficial outcome) and NNTH are as follows: Tantilipikorn 2010 had a NNT of 25, with NNTB 10 and NNTH 50. Jacobs 2009 had a NNT of 50, with NNTB 20 and NNTH 100. Malm 1981 had a NNT of 10, with NNTB 6.25 and NNTH 14.29. Finally, Lundblad 2001 did show a significant difference with a NNT of 14.29 (95% CI 7.69 to 100).
Three studies reported on epistaxis but were not included in the meta‐analysis because the study did not report numerical data for the non‐allergic rhinitis subgroup ( Scadding 1995 ), due to lack of quality of the study ( Lin 2017 ), or because no events were observed in either group ( Arikan 2006 ). Scadding 1995 reported "generally minor" adverse events in the intranasal corticosteroids group and Lin 2017 reported two cases of epistaxis in a total group of 22 patients treated with budesonide versus no cases of epistaxis in the placebo group. Arikan 2006 reported no epistaxis in either the intervention group or the control group.
Disease‐specific health‐related quality of life
No studies except Lin 2017 reported numerical data on (health‐related) quality of life for non‐allergic rhinitis patients. Lin 2017 used the Short Form 12 (SF‐12v2) questionnaire to measure quality of life (scale range 0 to 800). This study was not included in the analysis because of lack of quality of the study data (see reasons above). Lin 2017 reported a higher quality of life in the intranasal corticosteroids group versus the placebo group after one month (152.3 versus 145.6); however, while this difference was clear at one‐month follow‐up it was barely noticeable at three months follow‐up (148.4 versus 145.6) (low‐certainty evidence).
Lundblad 2001 reported no numerical data on quality of life but did report narratively that there was no significant difference in quality of life between the intranasal corticosteroids group and the placebo group.
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
Only two studies provided data for objective airway measurements that we could use in our analyses ( Malm 1981 ; Spector 1980 ), one using peak flow expiratory rate ( Spector 1980 ) and one using rhinomanometry ( Malm 1981 ). The study using peak flow expiratory rate did not find a significant difference for flunisolide over placebo (SMD 0.78, 95% CI ‐0.47 to 2.03; 11 participants) ( Analysis 1.5 ). For rhinomanometry there was also no significant difference (SMD ‐0.46, 95% CI ‐1.06 to 0.14; 44 participants) ( Analysis 1.6 ) ( Malm 1981 ). This evidence is of very low certainty.
Ellegård 2001 was not included in the meta‐analysis as it compared intranasal corticosteroids versus placebo in a single separate subgroup of non‐allergic rhinitis patients, i.e. participants with pregnancy rhinitis. Ellegård 2001 reported a blockage index ((PEF‐nPEF)/PEF) to objectify airflow after treatment. The mean blockage index after eight weeks of treatment in the fluticasone group was 0.39 (SD 0.16) and the mean blockage index in the placebo group was 0.41 (SD 0.15), therefore there was no significant difference between the intranasal corticosteroids and placebo groups.
Hallén 1997 was not included in the meta‐analysis as it compared intranasal corticosteroids versus placebo in a single separate subgroup of non‐allergic rhinitis patients, i.e. participants with rhinitis medicamentosa. Hallén 1997 reported both acoustic rhinometry and PNIF after 13 days of treatment. The mean acoustic rhinometry outcome in the intranasal corticosteroids group was 0.28 cm 2 (SD 0.19) and the mean acoustic rhinometry outcome in the placebo group was 0.03 cm 2 (SD 0.17) (SMD ‐1.34, 95% CI ‐0.35 to ‐2.33). The mean PNIF outcome in the intranasal corticosteroids group was 121.2 L/min (SD 69.0) and the mean PNIF outcome in the placebo group was 128.7 L/min (SD 40.4) (SMD ‐0.13, 95% CI ‐0.75 to 1.00), i.e. there was no significant difference.
O'Reilly 1991 was not included in the meta‐analysis because it only reported P values and there was too wide a variation between baseline and placebo values.
Tarlo 1977 was not included in the meta‐analysis because it reported no numerical data for the non‐allergic rhinitis subgroup.
Jessen 1990 used rhinomanometry during the inclusion of patients but it was not used to objectively measure airflow after treatment and could therefore not be used in the meta‐analysis.
Kalpaklioglu 2010 was included in the meta‐analysis for the comparison of intranasal corticosteroids versus ipratropium bromide.
The objective airflow measurements of Singh 2017 could not be included as they were related to cold dry air exposure.
Other adverse events
The outcome 'other adverse events' was defined as adverse events other than epistaxis, for example pharyngitis, nasal dryness/crusting and headache.
Three studies included in the meta‐analysis reported on 'other adverse events' besides epistaxis ( Jacobs 2009 ; Lundblad 2001 ; Tantilipikorn 2010 ). We decided to combine the three studies and not to make a separation into up to four weeks and more than four weeks follow‐up. Intranasal corticosteroids probably result in little or no difference in the risk of other adverse events compared to placebo (RR 0.99, 95% CI 0.87 to 1.12; 3 studies; 1130 participants; I 2 = 0%) ( Analysis 1.7 ) (moderate‐certainty evidence).
Lin 2017 was not included in the meta‐analysis due to lack of quality of the study data (see above). Miller 1969 was not included in the meta‐analysis as it used a dosage of dexamethasone of 672 µg to 1008 µg per day and only studies with an intranasal corticosteroid dosage of 200 µg to 400 µg were included in the meta‐analysis (see above). Malm 1981 was not included in the meta‐analysis as it was unclear in which intranasal corticosteroid dosage subgroup the other adverse events occurred. Other studies describing 'other adverse events' as an outcome in their 'Materials and methods' sections did not report actual numbers/data in the 'Results' section or did not report for the non‐allergic rhinitis group separately and were therefore not included in the review.
Lin 2017 was not included in the meta‐analysis but reported seven cases of other adverse events in a total group of 22 patients treated with budesonide versus no other adverse events in the placebo group.
Miller 1969 described two (non‐epistaxis) adverse events in the intranasal corticosteroids group and two adverse events in the placebo group.
Malm 1981 reported four 'other adverse events' in the intranasal corticosteroids group versus none in the placebo group.
Varricchio 2011 reported no clinically relevant adverse events in either the treatment or control group.
Subgroups/phenotypes
Within the comparison of intranasal corticosteroids versus placebo there were not enough data to perform a subgroup analysis for different subgroups/phenotypes of non‐allergic rhinitis. Ellegård 2001 evaluated intranasal corticosteroids in pregnancy rhinitis patients (see also under 'Objective measurement of airflow'). Overall, the study did not find a beneficial effect of intranasal corticosteroids over placebo. Hallén 1997 evaluated intranasal corticosteroids in rhinitis medicamentosa patients (see also 'Objective measurement of airflow'). They concluded that the symptom scores for nasal stuffiness showed a marked reduction during the treatment period in both groups, but there was a faster onset of symptom reduction after treatment with fluticasone propionate.
Intranasal corticosteroids versus saline
One four‐armed study compared intranasal budesonide nasal spray 256 µg once daily and nasal saline irrigation 100 mL of 3% saline per nostril combined with intranasal budesonide nasal spray 256 µg once daily to nasal saline alone and no treatment ( Lin 2017 ). This study was not included in the meta‐analysis due to lack of quality of the study data (see above).
Disease severity, as measured by patient‐reported total nasal symptom score
This study reported a total nasal symptom score using a visual analogue scale (VAS) (unclear scale per individual symptom and unclear total range of symptoms).
There was a significant difference between budesonide (from VAS 5.91 to VAS 5.68 after three months) and saline (from VAS 5.96 to VAS 4.80 after three months) in favour of saline (t‐test, P < 0.05).
Significant adverse events: epistaxis
The risk of epistaxis was higher in the intranasal corticosteroids group (two participants with epistaxis) compared to the saline group (no participants with epistaxis).
Disease‐specific health‐related quality of life
The quality of life measurement (SF‐12v2: range 0 to 800) also showed a significant effect (t‐test, P < 0.05) in favour of saline (SF‐12v2 increase from 146 at baseline to 151.30 after three months) compared to budesonide (SF‐12v2 increase from 146 to 148.40).
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
The study did not report objective measurements of nasal airflow.
Other adverse events
There were seven other adverse events in the budesonide treatment group (pharyngitis and nasal dryness/crusting) and no other adverse events in the saline group.
Intranasal corticosteroids versus intranasal antihistamine
Three studies reported on intranasal corticosteroids versus an intranasal antihistamine ( Behncke 2006 ; Kalpaklioglu 2010 ; Song 2018 ). Kalpaklioglu 2010 and Song 2018 were included in the meta‐analysis. Kalpaklioglu 2010 compared triamcinolone acetonide nasal spray 220 µg once daily to azelastine hydrochloride and Song 2018 compared budesonide 200 µg two times per day to azelastine 200 µg two times per day.
Disease severity, as measured by patient‐reported total nasal symptom score
In Song 2018 , there was a non‐significant difference in combined nasal symptom VAS score in favour of budesonide nasal spray (mean difference (MD) ‐0.25, 95% CI ‐0.69 to 0.19; 80 participants) ( Analysis 2.1 ).
In Kalpaklioglu 2010 , there was a non‐significant difference in total nasal symptom score mean change from baseline in favour of triamcinolone acetonide nasal spray (MD ‐0.50, 95% CI ‐1.92 to 0.92; 63 participants) ( Analysis 2.2 ). The study reports a significant improvement in sneezing with triamcinolone in patients with non‐allergic rhinitis (P < 0.01), as well as conjunctivitis.
Significant adverse events: epistaxis
Epistaxis was not evaluated.
Disease‐specific health‐related quality of life
Kalpaklioglu 2010 assessed quality of life with the mini‐Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ) but these results were not reported for non‐allergic rhinitis participants separately.
Song 2018 assessed quality of life with the SF12‐v2 questionnaire (a higher score indicates better quality of life). There was a non‐significant difference in favour of azelastine in quality of life (MD ‐1.30, 95% CI ‐3.60 to 1.00; 80 participants) ( Analysis 2.3 ).
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
Kalpaklioglu 2010 also reported on the inspiratory peak flow rate (change from baseline) and showed a small, non‐significant difference in favour of triamcinolone acetonide (MD ‐6.17, 95% CI ‐15.25 to 2.91; 63 participants) ( Analysis 2.4 ).
Other adverse events
Song 2018 reported a higher risk of 'other adverse events' (such as dryness of the nasal mucosa) in the budesonide group compared to azelastine (RR 2.00, 95% CI 0.19 to 21.18; 80 participants) ( Analysis 2.5 ).
Behncke 2006 was not included in the meta‐analysis because it reported no numerical data for non‐allergic rhinitis participants separately. This study does conclude that there is no difference in effectiveness between intranasal corticosteroids and intranasal antihistamines. The authors conclude that azelastine nasal spray and fluticasone nasal spray improve RQLQ scores and rhinitis symptom scores in geriatric patients with either allergic or non‐allergic rhinitis.
Intranasal corticosteroids versus capsaicin
One study provided data for this comparison. Havas 2002 compared budesonide nasal spray applied twice daily (256 µg daily dosage) to capsaicin 2.616 µg once weekly.
Disease severity, as measured by patient‐reported total nasal symptom score
There was a large significant difference in mean total nasal score in favour of capsaicin (MD 1.60, 95% CI 0.03 to 3.16; 40 participants) ( Analysis 3.1 ). A total nasal symptom score was calculated as the mean sum of rhinorrhoea, congestion and sneezing (VAS 0 to 5 for each symptom, per side; range 0 to 30).
Significant adverse events: epistaxis
Epistaxis was not evaluated.
Disease‐specific health‐related quality of life
Quality of life was not evaluated.
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
Objective measurements of airflow were not evaluated.
Other adverse events
Other adverse events were not evaluated.
Intranasal corticosteroids versus sodium cromoglycate
One study provided data for this comparison ( Hillas 1980 ). This study compared sodium cromoglycate 2% six times daily to beclomethasone dipropionate 400 µg daily. It reported no numerical data for the non‐allergic rhinitis group separately and was therefore not included in the meta‐analysis.
Disease severity, as measured by patient‐reported total nasal symptom score
Intranasal corticosteroids relieved symptoms in 76.9% of patients versus 50% of patients treated with sodium cromoglycate, a significant difference. The total symptom score (mean of a total of seven symptoms that were scored on a range from 0 to 3) at the end of treatment was 4.12 in participants treated with sodium cromoglycate and 2.37 in participants treated with intranasal corticosteroids, a significant difference.
Significant adverse events: epistaxis
The study reported that occasionally (no numerical data) patients using intranasal corticosteroids had blood spotting while blowing their noses. This was not reported in the group treated with sodium cromoglycate.
Disease‐specific health‐related quality of life
Quality of life was not evaluated.
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
Objective measurements of airflow were not evaluated.
Other adverse events
No numerical data on other adverse events were reported, however some patients experienced sneezing after using intranasal corticosteroids. No significant adverse events were reported for cromoglycate sodium.
Intranasal corticosteroids versus ipratropium bromide
One study provided data for this comparison ( Jessen 1990 ). This cross‐over study compared beclomethasone aerosol, twice daily (total daily dose 400 µg) with ipratropium bromide 160 µg.
Disease severity, as measured by patient‐reported total nasal symptom score
There was no significant difference between the treatments in total nasal symptom score (MD ‐1.50, 95% CI ‐12.24 to 9.24; 48 participants) ( Analysis 4.1 ).
Significant adverse events: epistaxis
Epistaxis was not evaluated.
Disease‐specific health‐related quality of life
Quality of life was not evaluated.
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
The study used rhinomanometry during the inclusion of patients but it was not used to objectively measure airflow after treatment.
Other adverse events
No other adverse events were evaluated.
Intranasal corticosteroids versus intranasal corticosteroids combined with intranasal antihistamine
Three studies provided data for this comparison ( Guo 2015 ; Singh 2017 ; Song 2018 ).
Guo 2015 compared fluticasone dipropionate nasal spray in an unknown dosage of two sprays in each nostril once daily with fluticasone dipropionate nasal spray 100 µg combined with azelastine in an unknown dosage in each nostril twice daily.
Singh 2017 reported no numerical data and was therefore not included in the meta‐analysis.
Song 2018 compared budesonide nasal spray 200 µg two times per day with budesonide nasal spray 200 µg two times per day combined with azelastine nasal spray 200 µg two times per day.
Disease severity, as measured by patient‐reported total nasal symptom score
There was a significant difference between intranasal corticosteroids alone and intranasal corticosteroids combined with intranasal antihistamine for nasal symptom score (SMD 0.75, 95% CI 0.48 to 1.02; 2 studies; 242 participants) ( Analysis 5.1 ).
Guo 2015 reported a small but significant difference in total nasal symptom score (unclear scale and range) in favour of fluticasone dipropionate nasal spray combined with azelastine after six weeks of treatment (SMD 0.37, 95% CI 0.06 to 0.68; 162 participants).
Song 2018 also reported a significant difference in total symptom VAS score (range 0 to 10) in favour of budesonide nasal spray combined with azelastine after eight weeks of treatment (SMD 0.75, 95% CI 0.48 to 1.02; 80 participants).
Significant adverse events: epistaxis
Guo 2015 and Song 2018 did not report any cases of epistaxis in either treatment group.
Disease‐specific health‐related quality of life
In Guo 2015 , quality of life was not evaluated.
Song 2018 did evaluate quality of life by means of the SF12‐v2 questionnaire (a higher score indicating a better quality of life). It showed a significantly higher quality of life in the group treated with budesonide combined with azelastine nasal spray compared to budesonide nasal spray alone (MD ‐7.20, 95% CI ‐9.77 to ‐4.63; 80 participants) ( Analysis 5.2 ).
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
In Guo 2015 and Song 2018 objective measurements of airflow were not evaluated.
Other adverse events
Two studies included in the meta‐analysis reported on 'other adverse events' ( Guo 2015 ; Song 2018 ). Both studies showed a nearly significant higher rate of other adverse events in the combined intranasal corticosteroid and intranasal antihistamine group (RR 0.26, 95% CI 0.07 to 1.01; 2 studies; 242 participants; I 2 = 15%) ( Analysis 5.3 ).
Guo 2015 reported more adverse events (five reporting fatigue and bitter taste) in the fluticasone dipropionate with azelastine group than in the fluticasone dipropionate alone group (no adverse events) (RR 0.09, 95% CI 0.00 to 1.54; 162 participants) ( Analysis 5.3 ).
Song 2018 also reported more adverse events (dryness of nasal mucosa, dry throat discomfort, bitter taste, slight erosion of nasal mucosa) in the budesonide with azelastine group than in the budesonide alone group (RR 0.50, 95% CI 0.10 to 2.58; 80 participants ( Analysis 5.3 ).
Singh 2017 compared intranasal corticosteroids combined with azelastine to placebo instead of intranasal corticosteroids. It could not be included in the meta‐analysis because the total nasal symptom score was not reported numerically. Total nasal symptom score and objective measurement of airflow were both related to cold dry air provocation. The study did report that there were no statistically significant differences between the two treatments.
Intranasal corticosteroids versus intranasal corticosteroids combined with saline irrigation
One study provided data for this comparison ( Lin 2017 ). This study was not included in the meta‐analysis due to lack of quality of the study data (see above). It compared nasal saline irrigation (100 mL of 3% saline per nostril twice a day) with intranasal budesonide nasal spray 256 µg once daily (two sprays per nostril per day, 64 µg per spray) to intranasal budesonide alone.
Disease severity, as measured by patient‐reported total nasal symptom score
There was a significant difference between combination therapy (from VAS 6.18 to VAS 4.48 after three months) and budesonide (from VAS 5.91 to VAS 5.68 after three months) in favour of combination therapy (t‐test, P < 0.05).
Significant adverse events: epistaxis
The combination therapy group had one patient with epistaxis, while the budesonide group had two patients with epistaxis.
Disease‐specific health‐related quality of life
The quality of life measurement (SF‐12v2) also showed a significant effect in favour of combination therapy (increase from 146 at baseline to 152.9 after three months) over budesonide (increase from 146 to 148.40) (t‐test, P < 0.05).
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
The study did not report on objective measurements of nasal airflow.
Other adverse events
The combined therapy group had eight participants with other adverse events (pharyngitis, nasal dryness/crusting), while the budesonide group had seven participants with other adverse events.
Intranasal corticosteroids with saline spray versus saline spray alone
One study provided data for this comparison ( Boechat 2019 ). It compared mometasone furoate 200 µg daily with isotonic saline spray to isotonic saline spray alone.
Disease severity, as measured by patient‐reported total nasal symptom score
There was a non‐significant difference between intranasal corticosteroid spray combined with isotonic saline spray (VAS score 4.1 (SD 2.4)) and isotonic saline spray alone (VAS score 5.4 (SD 2.1)) after two weeks (MD ‐1.30, 95% CI‐2.97 to 0.37; 28 participants) ( Analysis 6.1 ).
The pre‐treatment VAS score for intranasal corticosteroid spray combined with isotonic saline spray was 5.2 (SD 2.0) and for nasal spray alone was 5.3 (SD 2.5). Although the combined treatment showed better symptom improvement compared to saline alone the reduction was non‐significant (P = 0.056).
Significant adverse events: epistaxis
The study reported no adverse events.
Disease‐specific health‐related quality of life
The quality of life measurement (SNOT‐22 questionnaire with a lower score indicating a better quality of life) showed no significant difference between intranasal corticosteroid spray combined with isotonic saline spray (24.3 (SD 16.5)) and isotonic nasal spray alone (32.3 (SD 15.2)) after two weeks (MD ‐8.0, 95% CI ‐19.75 to 3.75; 28 participants) ( Analysis 6.2 ). Pre‐treatment quality of life (SNOT‐22) for intranasal corticosteroid spray combined with isotonic saline spray was 30.0 (SD 15.2) and for nasal spray alone was 38.1 (SD 19.9). The reduction was non‐significant (P = 0.095).
Inspiratory peak flow levels, rhinomanometry or other objective measurements of airflow
The peak nasal inspiratory flow (PNIF) measurements showed no significant difference between intranasal corticosteroid spray combined with isotonic saline spray (72.9 L/min (SD 25.5)) and isotonic nasal spray alone (82.1 L/min (SD 39.8)) after two weeks (MD ‐9.20, 95% CI ‐33.96 to 15.56; 28 participants) ( Analysis 6.3 ). Pre‐treatment PNIF for intranasal corticosteroid spray combined with isotonic saline spray was 77.1 (SD 25.8) and for nasal spray alone was 90.7 (SD 38.5). The reduction was non‐significant (P = 0.688).
Other adverse events
The study reported no adverse events.
Discussion
Summary of main results
See summary of findings Table for the main comparison .
We included 34 studies with a total of 4452 participants in this review, reporting on our main comparison (intranasal corticosteroids versus placebo) and eight further comparisons: intranasal corticosteroids versus saline, versus intranasal antihistamine, versus capsaicin, versus cromoglycate sodium, versus ipratropium bromide, versus intranasal corticosteroids and intranasal antihistamine, versus intranasal corticosteroids with saline and intranasal corticosteroids with saline versus saline alone. We were able to analyse data from 18 studies for the eight different comparisons.
Intranasal corticosteroids versus placebo
We were only able to identify a significant number of studies (25) for the main comparison, intranasal corticosteroids versus placebo; 13 of these studies could be included in the meta‐analysis. However, the evidence was limited by the fact that most studies had only small numbers of patients and there was a high degree of variance in their results. The two largest studies did show a significant improvement in symptom scores ( Jacobs 2009 ; Webb 2002 ). However, the study data in Jacobs 2009 are unlikely to be credible. The study presented very unlikely standard deviation (SD) values, which did not match the presented mean, n and P values , resulting in very high heterogeneity of the data for the outcome 'Total nasal symptom score, change from baseline'. Most likely the standard deviation (SD) values should have been standard error of the mean (SEM) values, as we confirmed by re‐analysis. When we used the adjusted SD values (assuming the SD values presented were actually SEM), the improvement in symptom score in Jacobs 2009 was no longer significant.
There may be an improvement in patient‐reported disease severity as measured by a total nasal symptom score (TNSS) with intranasal steroids but we are uncertain because we assessed the certainty of the evidence as low to very low due to high imprecision and risk of publication bias due to small patient numbers. There were too few data to draw conclusions on any differences according to type of intranasal corticosteroids, dosage, vehicle used, frequency of usage or duration of treatment.
There is probably a higher risk of epistaxis with intranasal corticosteroids compared to placebo (moderate‐certainty evidence).
One study assessed quality of life ( Lin 2017 ). This study showed that quality of life was better in the intranasal corticosteroids group compared to the placebo group. However, while this difference was significant at one‐month follow‐up it was barely noticeable at three‐month follow‐up. Lin 2017 was not included in the meta‐analysis because of lack of quality of the study data. Firstly, the study presents unexpected data with disappearance of benefit of intranasal corticosteroids with longer follow‐up. Secondly, including the study in the meta‐analysis resulted in a high level of heterogeneity. Finally, the SD values that are presented in the study do not match with presented means, n and P values. The data make more sense if the SD values presented are actually standard error of the mean (SEM), which we again confirmed by a re‐analysis.
As only two studies evaluated objective measurements of airflow and the data could not be pooled due to the different methods used, we cannot draw conclusions on this outcome. Neither study found a difference between intranasal corticosteroids and placebo.
Intranasal corticosteroids probably result in little or no difference in the risk of other adverse events compared to placebo (moderate‐certainty evidence).
Other comparisons
For the following comparisons it is uncertain whether there are differences between intranasal corticosteroids and the comparator group for any of the outcomes because only one study assessed each comparison and in each case the certainty of the evidence was very low: intranasal corticosteroids versus saline irrigation; intranasal corticosteroids versus intranasal antihistamine; intranasal corticosteroids versus capsaicin; intranasal corticosteroids versus cromoglycate sodium; intranasal corticosteroids versus ipratropium bromide; intranasal corticosteroids versus intranasal corticosteroids combined with intranasal antihistamines; intranasal corticosteroids versus intranasal corticosteroids combined with saline irrigation; and intranasal corticosteroids with intranasal isotonic nasal spray versus isotonic nasal spray alone.
Three studies compared an intranasal corticosteroid with an intranasal corticosteroid combined with an intranasal antihistamine. Two studies reported a significant difference in favour of intranasal corticosteroids combined with an intranasal antihistamine versus intranasal corticosteroids alone ( Guo 2015 ; Song 2018 ). The difference in favour of the combined treatment strategy in these two studies was significant. The third study reported no statistically significant differences between the two treatments ( Singh 2017 ).
Overall completeness and applicability of evidence
The types and dosages of intranasal corticosteroids used in the studies were in keeping with manufacturers' recommendations and are applicable to the population being studied. The phenotype/endotype population of patients with non‐allergic rhinitis studied most likely varied among studies. As discussed in the Background , one would expect the inflammatory non‐allergic rhinitis endotypes (local allergic rhinitis/non‐allergic rhinitis with eosinophilia syndrome (NARES)) to benefit more from intranasal corticosteroids than the neurogenic or idiopathic endotypes.
Quality of life, which is one of the most important outcomes for patients, was only included in three studies as an outcome measure ( Boechat 2019 ; Lin 2017 ; Song 2018 ). There is too little information, therefore, to establish whether intranasal steroids have an impact on patients' quality of life.
Quality of the evidence
The certainty of the evidence (GRADE assessment) for our primary outcome, disease severity as measured by patient‐reported symptom score (total nasal symptom score) was in general low because most studies had small participant numbers resulting in high imprecision and high risk of publication bias. One of the only two studies with a large number of participants was Jacobs 2009 . Unfortunately this study presented unlikely standard deviation values that were most likely to be standard error of the mean values.
It is likely that the variety of different intranasal corticosteroid treatment strategies (type of intranasal corticosteroids, dosage, method of delivery), the differences in included non‐allergic rhinitis pheno‐ and endotypes and the differences in the ways of measuring disease severity scores that were used contributed to the heterogeneity in the study results. There was great variety in the methods used to measure symptom severity scores and many scales were not validated. Not all studies defined non‐allergic rhinitis endotypes, for example presence or absence of inflammatory cells such as eosinophils (NARES), which complicated subgroup analyses based on pheno‐ and endotypes. A higher proportion of inflammatory cells (eosinophils) might improve the chances of treatment effectiveness.
This low certainty of evidence is in contrast to the epistaxis adverse event outcome, where we can be more certain that there is probably an increased risk in the intranasal corticosteroids group compared to placebo (moderate‐certainty evidence).
For quality of life and objective measurements of airflow there was not enough information to draw conclusions (low‐certainty evidence).
There was moderate‐certainty evidence (large number of patients, low heterogeneity) that intranasal corticosteroids probably result in little or no difference in the risk of other adverse events compared to placebo.
Potential biases in the review process
The primary outcome total nasal symptom score consisted of different nasal symptoms in different studies, measured on different measurement scales. When only individual nasal symptom scores were reported but no total nasal symptom score, we calculated a total nasal symptom scores for rhinorrhoea (secretion), nasal obstruction (blockage) and sneezing: the most common symptoms used to calculate a total nasal symptom score in the other included studies ( Table 1 ). We decided not to include itching, as a previous study by our research group has shown that ocular itch plays a less dominant role in non‐allergic rhinitis compared to allergic rhinitis ( Segboer 2018 ). However, given that itching was included in the total nasal symptom score of a few other studies, this may have resulted in a potential bias.
In some cases, the studies did not report enough information for us to analyse the results further. Therefore for some studies we manually measured pixels from graphs to calculate mean values and imputed standard deviations based on the P values reported.
Some studies did include both allergic and non‐allergic rhinitis participants but did not provide (enough) separate data for non‐allergic rhinitis participants to calculate a mean and standard deviation (SD).
In the meta‐analysis we included only studies with an intranasal corticosteroid dosage range of 200 µg to 400 µg. The reason for this was to prevent double counting of the same patients or controls (see Differences between protocol and review ). This leads to a potential bias in the review as the meta‐analysis is limited to certain dosages. However, only one study was excluded from the meta‐analysis because of this dosage limitation ( Varricchio 2011 ).
Among studies the daily dosage of intranasal corticosteroids varied from 50 µg to 2000 µg. Most of the studies that compared different dosages of intranasal corticosteroids used a cross‐over study design with the exception of Blom 1997 and Webb 2002 , which used a parallel‐group study design. In the cross‐over studies the same participants were treated with different dosages of intranasal corticosteroids, with a short (one‐week) or no washout, complicating a clear comparison between these dosage subgroups ( Balle 1982 ; Malm 1976 ; Malm 1981 ). Only Balle 1982 showed a dosage effect for two nasal symptom score outcomes. Malm 1976 and Malm 1981 showed no significant difference between the dosage subgroups. The two parallel‐group studies both concluded that there were no statistically significant differences between the different intranasal corticosteroid dosage subgroups ( Blom 1997 ; Webb 2002 ). In the parallel‐group studies the different dosage subgroups contained different participants but were compared with the same control group. To prevent counting the same patients or controls more than once, we decided to include one intranasal corticosteroid dosage in the meta‐analysis. The most common intranasal corticosteroids dosage was 200 µg. A test for subgroup differences showed no significant difference (no 'dosage effect') between 200 µg and 400 µg of intranasal corticosteroids. We therefore included studies in the meta‐analysis with an intranasal corticosteroid dosage range of 200 µg to 400 µg.
Agreements and disagreements with other studies or reviews
There are no previous published Cochrane Reviews on intranasal corticosteroids in non‐allergic rhinitis. There are, however, some position papers on non‐allergic rhinitis, such as Hellings 2017 . This paper states that the inflammatory group of non‐allergic rhinitis patients (occupational and drug‐induced rhinitis) may benefit from anti‐inflammatory treatment such as nasal/oral corticosteroids. However, they conclude that most randomised controlled trials evaluating local corticosteroids in non‐allergic rhinitis patients have shown a lack of efficacy. The PRACTALL report suggests that intranasal corticosteroids could be effective in two phenotypes of non‐allergic rhinitis, i.e. NARES and possibly rhinitis medicamentosa, but it does not mention effectiveness for other pheno‐ or endotypes of non‐allergic rhinitis ( Papadopoulos 2015 ).
Process for sifting search results and selecting studies for inclusion.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 1 Total nasal symptom score, follow‐up ≤ 4 weeks.
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 2 Total nasal symptom score, follow‐up > 4 weeks.
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 3 Total nasal symptom score (change from baseline), follow‐up ≤ 4 weeks.
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 4 Significant adverse event: epistaxis.
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 5 Objective measurement of airflow: peak flow rate (expiratory).
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 6 Objective measurement of airflow: rhinomanometry.
Comparison 1 Intranasal corticosteroids versus placebo, Outcome 7 Other adverse events.
Comparison 2 Intranasal corticosteroids versus intranasal antihistamine, Outcome 1 Total nasal symptom score, follow‐up > 4 weeks.
Comparison 2 Intranasal corticosteroids versus intranasal antihistamine, Outcome 2 Total nasal symptom score (change from baseline), follow‐up ≤ 4 weeks.
Comparison 2 Intranasal corticosteroids versus intranasal antihistamine, Outcome 3 Quality of life (SF‐12v2).
Comparison 2 Intranasal corticosteroids versus intranasal antihistamine, Outcome 4 Objective measurement of airflow: inspiratory peak flow rate (change from baseline).
Comparison 2 Intranasal corticosteroids versus intranasal antihistamine, Outcome 5 Other adverse events.
Comparison 3 Intranasal corticosteroids versus capsaicin, Outcome 1 Total nasal symptom score, follow‐up ≤ 4 weeks.
Comparison 4 Intranasal corticosteroids versus ipratropium bromide, Outcome 1 Total nasal symptom score, follow‐up ≤ 4 weeks.
Comparison 5 Intranasal corticosteroids (INCS) versus INCS + intranasal antihistamine, Outcome 1 Total nasal symptom score, follow‐up > 4 weeks.
Comparison 5 Intranasal corticosteroids (INCS) versus INCS + intranasal antihistamine, Outcome 2 Quality of life (SF12‐v2).
Comparison 5 Intranasal corticosteroids (INCS) versus INCS + intranasal antihistamine, Outcome 3 Other adverse events.
Comparison 6 Intranasal corticosteroids + saline versus saline, Outcome 1 Total nasal symptom score, follow‐up ≤ 4 weeks.
Comparison 6 Intranasal corticosteroids + saline versus saline, Outcome 2 Quality of life (SNOT‐22).
Comparison 6 Intranasal corticosteroids + saline versus saline, Outcome 3 Objective measurement of airflow: peak nasal inspiratory flow.
|
Intranasal corticosteroids compared to placebo for non‐allergic rhinitis |
|||||||
|
Patient or population:
adults and children > 12 with non‐allergic rhinitis
|
|||||||
|
Outcomes |
Anticipated absolute effects * (95% CI) |
Relative effect
|
№ of participants
|
Certainty of the evidence
|
Comments |
||
|
Risk with placebo |
Risk with intranasal corticosteroids |
||||||
|
Disease severity as measured by patient‐reported symptom score (total nasal symptom score) |
Follow‐up ≤ 4 weeks |
— |
SMD 0.74 lower
|
— |
131
|
⊕⊕⊝⊝
|
Intranasal corticosteroids may improve patient‐reported disease severity at a follow‐up of up to 4 weeks compared to placebo. The mean difference in disease severity score was 0.74 standard deviations lower (1.15 lower to 0.33 lower) with intranasal corticosteroids compared to placebo. This represents a medium effect size ( Cohen 1988 ). |
|
Follow‐up > 4 weeks |
— |
SMD 0.24 lower
|
— |
85
|
⊕⊝⊝⊝
|
It is uncertain whether intranasal corticosteroids improve patient‐reported disease severity with a follow‐up of more than 4 weeks compared to placebo, because the certainty of the evidence is very low. |
|
|
Change from baseline Follow‐up ≤ 4 weeks |
— |
SMD 0.15 lower (0.25 lower to 0.05 lower) |
— |
1465
|
⊕⊕⊝⊝
|
Intranasal corticosteroids may slightly improve patient‐reported disease severity change from baseline with a follow‐up of up to 4 weeks compared to placebo. The SMD of 0.15 represents a small effect size. There are two large studies ( Jacobs 2009 ; Webb 2002 ). Jacobs 2009 reports with a high degree of certainty a small improvement in favour of intranasal corticosteroids. Webb 2002 reports a less certain clinically relevant improvement in favour of intranasal corticosteroids. Jacobs 2009 has an adjusted SD value (presented SD are most likely SEM). |
|
|
Significant adverse event: epistaxis Follow‐up: 2 weeks to 33 days |
Study population |
RR 2.10 (1.24 to 3.57) |
1174
|
⊕⊕⊕⊝
|
There is probably a higher risk of epistaxis with intranasal steroids compared to placebo. |
||
|
(31 per 1000 |
65 per 1000
|
||||||
|
Disease‐specific health‐related quality of life Short Form 12 (SF‐12v2) (range 0 to 800) Follow‐up: 1 month to 3 months |
Just one study reported quality of life ( Lin 2017 ). Quality of life was better in the intranasal corticosteroids group versus the placebo group, however while this difference was significant at a follow‐up of 1 month (152.3 versus 145.6) it was barely noticeable at a follow‐up of 3 months (148.4 versus 145.6). |
49 (1 RCT) |
⊕⊕⊝⊝ low 5 |
There is not enough information (1 study) to conclude whether there is a difference. |
|||
|
Objective measurement of airflow: peak flow rate (expiratory) Follow‐up: 2 weeks to 4 weeks |
Just two studies reported objective measurement of airflow ( Malm 1981 ; Spector 1980 ). However, they used different outcome measurements to measure outflow: rhinomanometry and expiratory peak flow rate (PEFR). Neither found a significant difference between groups: rhinomanometry ( Malm 1981 ) SMD ‐0.46, 95% CI ‐1.06 to 0.14; 44 participants; PEFR ( Spector 1980 ) SMD 0.78, 95% CI ‐0.47 to 2.03; 11 participants. |
55
|
⊕⊝⊝⊝ very low 6 |
There is not enough information (2 studies with different methods of measurement) to conclude whether there is a difference. |
|||
|
Other adverse events Follow‐up: 1 month to 6 weeks |
Study population |
RR 0.99
|
1130
|
⊕⊕⊕⊝
|
Intranasal corticosteroids probably result in little or no difference in the risk of other adverse events compared to placebo. |
||
|
454 per 1000 |
450 per 1000
|
||||||
|
*
The risk in the intervention group
(and its 95% confidence interval) is based on the assumed risk in the comparison group and the
relative effect
of the intervention (and its 95% CI).
|
|||||||
|
GRADE Working Group grades of evidence
|
|||||||
|
1 Due to the small sample size we downgraded once for imprecision and once due to the risk of publication bias. The I 2 value in this pooled analysis was 22% so there was no reason to downgrade for heterogeneity. 2 We downgraded twice for serious imprecision due to the small sample size and because the confidence interval includes both meaningful benefit and harm. We downgraded once for risk of publication bias due to the small sample size. 3 We downgraded twice for serious inconsistency because the I 2 value was 96% when the original SD values presented in the Jacobs 2009 study were used and in that case the confidence interval for Jacobs 2009 does not overlap with the other studies. Jacobs 2009 has an unlikely SD value, which does not match the mean, n and P values. It is likely that the originally SD value presented should actually be a standard error of the mean (SEM), therefore we re‐calculated the SD values. 4 We downgraded once due to study limitations (risk of bias) because there were unclear blinding domains, which could have influenced the significant adverse events (epistaxis) outcome. The I 2 value in this pooled analysis is 0% so there is no reason to downgrade for heterogeneity. We judged that there were no other reasons to downgrade. 5 Due to the small sample size we downgraded once for imprecision and once due to the risk of publication bias. 6 We downgraded once for inconsistency (when the two studies are combined the I 2 value is 67%). The two studies used different methods for measuring objective airflow, which contributed to the heterogeneity. Due to the small sample size we downgraded once for imprecision and once due to the risk of publication bias. 7 We downgraded once due to study limitations (risk of bias) as there were unclear blinding domains, which could have influenced the adverse events outcome. The I 2 value in the pooled analysis is 0% so there is no reason to downgrade for heterogeneity. We judged that there were no other reasons to downgrade. |
|||||||
|
Study ID |
Symptoms measured |
Score for each symptom |
Summation (total
|
Notes |
|
Nasal obstruction |
Measured on a VAS: 0 to 10, 0 is better |
Completed prior to trial and at 1, 2 and 3 months (range 0 to 10); as only one symptom no summation needed |
No summary data reported to allow inclusion in the meta‐analysis for this outcome |
|
|
1. Obstruction 2. Rhinorrhoea 3. Sneezing |
Scale unclear, low indicates fewer symptoms |
Total scores (for the last 7 days of a 2‐week treatment period) represented the means of scores for the 3 symptoms (range unclear) |
Unclear scale for individual symptoms |
|
|
A. Total nasal score (sum score of blockage, sneezing and rhinorrhoea) 1. Blockage 2. Sneezing 3. Rhinorrhoea B. Overall intensity of total nasal symptoms Does not consist of individual symptoms |
A. Measured on a scale of 0 to 3 (3 means worse) B. Measured on a VAS: 0 to 10 (0 is better) |
A. Presented as mean sum score of 3 symptoms for 1 week (range 0 to 3) B. Measured at 2 weeks pre‐treatment, 4 weeks after first batch of treatment, 8 weeks after treatment (range 0 to 10); overall intensity, therefore no summation of symptoms |
We used B as a total nasal symptom score because a VAS is a more established measurement |
|
|
Combined nasal symptom score (after emailing author):
1. Nasal blockage
|
Measured on a VAS scale of 0 to 10 (10 is worse) |
Measured pre‐intervention and at 2 weeks |
— |
|
|
Total nasal symptom score, unclear definition |
Unclear |
Unclear |
— |
|
|
1. Blocked nose 2. Itchy nose 3. Runny nose 4. Sneezing |
Measured on a scale of 0 to 3 (0 is better) |
Mean change in total combined symptom score (range 0 to 3) from end of treatment to baseline (week 4 versus week 0) |
— |
|
|
Congestion |
Measured on a scale of 0 to 4 (0 is better) |
Reported after 8 weeks of treatment and 16 weeks of post‐treatment follow‐up (range 0 to 4); as only one symptom no summation needed |
We evaluated the data at the end of 8 weeks of treatment as this was the most common method of measurement among other studies |
|
|
Total nasal symptom score; no breakdown in individual symptoms |
Scale unclear |
Measured pre‐intervention and at 2 weeks; range and summation unclear |
— |
|
|
Nasal congestion |
Measured on a VAS of 0 to 100 (0 is better) |
Reported for all days, 0 to 14 days, in the morning and in the evening (range 0 to 100); as only one symptom no summation needed |
No differences between morning and evening data. Reported morning data at day 13. |
|
|
1. Rhinorrhoea 2. Nasal blockage 3. Sneezing 4. Headache 5. Post‐nasal drip 6. Sore throat |
Measured on a VAS of 0 to 5 (0 = no symptoms), separately for each side |
Sum of mean values of rhinorrhoea, nasal blockage and sneezing (range 0 to 30); measurement at end of treatment |
We did not use the reported aggregate relief score, but instead calculated a total nasal symptom score out of rhinorrhoea, nasal blockage and sneezing (mean values and SDs per symptom were provided) |
|
|
Total nasal symptom score 1. Sneezing 2. Rhinorrhoea 3. Nasal pruritis 4. Blocked nose 5. Itchy eyes 6. Watery eyes 7. Red eyes Individual symptom scores |
Both measured on a scale of 0 to 3 (0 = nil, 1 = mild, 2 = moderate and 3 = severe) |
Total nasal symptom score: reported as a mean of 4‐week treatment period |
No separate data reported for non‐allergic rhinitis, only responder/non‐responder data |
|
|
Overall severity of rhinitis |
Measured daily on a scale of 1 to 4 (0 is better) |
Reported mean at end of week 2, 4, 6 and 8 (range 0 to 4) |
We did not use the reported individual rhinitis symptoms to calculate a total nasal symptom score, as only P values were reported |
|
|
1. Congestion 2. Rhinorrhoea 3. Post‐nasal drip |
Measured twice daily on diary cards using a
|
TNSS is sum of the 3 symptom scores (range 0 to 9). Change in TNSS (range 0 to 9) from end of treatment (week 4) to baseline. |
We used daily reflective TNSS and not morning instantaneous TNSS |
|
|
1. Nasal secretion 2. Sneezing 3. Nasal blockage |
All 3 symptoms were measured daily on a scale of 0 to 3/4 (0 is better) |
Reported as sum score of 2‐week treatment period (range 0 to 126/168) |
Unclear whether maximum scale was 3 or 4 |
|
|
1. Rhinorrhoea 2. Congestion 3. Itching 4. Sneezing 5. Anosmia 6. Conjunctivitis |
Scale of 0 to 4 (0 is better) |
Most likely ‐ although not explicitly cited ‐ sum of all individual symptoms. Reported for 2 weeks after treatment. Reported as mean change from baseline. |
— |
|
|
1. Nasal obstruction 2. Nasal itch 3. Rhinorrhoea 4. Sneezing |
Unclear scale |
Unclear total range of symptoms |
— |
|
|
1. Rhinorrhoea 2. Nasal stuffiness/congestion 3. Nasal itching 4. Sneezing |
Measured on a scale of 0 to 3 (0 is better) |
Range 0 to 12; reported most likely as sum at the end of treatment |
Converted into dichotomous outcome: improved versus unimproved. Improvement defined as a reduction of at least 1 point in the overall symptom score. No numerical data on original TNSS, so not included in meta‐analysis for this outcome |
|
|
1. Nasal catarrh 2. Blockage 3. Nasal itching 4. Sneezing |
Measured on a scale of 0 to 3 (0 is better) |
Range 0 to 12 |
No numerical data on original TNSS, therefore not included in meta‐analysis. There are data on 'non‐responder/improvement/responder' |
|
|
1. Nasal obstruction 2. Rhinorrhoea 3. Sneezing 4. Eye irritation |
Measured on a scale of 0 to 3 (0 is better) |
TNSS not reported, but calculated for meta‐analysis: sum of mean values of nasal obstruction, rhinorrhoea and sneezing (range 0 to 9); measurement at end of treatment) |
We calculated a total nasal symptom score out of rhinorrhoea, nasal blockage and sneezing (mean values and SDs per symptom were provided) for the dosages 200 µg, 400 µg and 800 µg |
|
|
1. Nasal obstruction 2. Nasal secretion 3. Sneezing |
Measured on a scale of 0 to 3 (0 is better) Reported as mean ± SEM of at last 3 days of the patients symptom score in each treatment period for nasal obstruction and secretion. For sneezing: scale of 0 to 3 (0 is good: no sneezing = 0; 1 to 5 sneezes = 1 point; 6 to 15 = 2 points; more than 15 sneezes = 3 points) |
TNSS not reported, but calculated for meta‐analysis: sum of mean values of nasal obstruction, rhinorrhoea and sneezing (range 0 to 9) |
We calculated a total nasal symptom score out of rhinorrhoea, nasal blockage and sneezing (mean values and SDs per symptom were provided) for the dosages 50 µg, 200 µg and 800 µg |
|
|
Nasal symptom score, unclear definition |
Unclear |
Unclear |
Unclear |
|
|
1. Nasal obstruction 2. Discharge 3. Post‐nasal drip 4. Sneezing |
Measured by patient at 2 and 4 weeks after treatment on a scale of 0 to 3 (0 is better) |
Not reported |
We were unable to calculate a SD for rhinorrhoea (secretion) and post‐nasal drip as the P values for these symptoms were not reported, therefore not included in meta‐analysis for this outcome |
|
|
1. Nasal obstruction 2. Anterior rhinorrhoea 3. Posterior rhinorrhoea 4. Sneezing 5. Facial pain |
Measured on a scale of 0 to 5 (0 is better) |
Composite score for all 5 symptoms, range 0 to 25 |
Only P values reported, therefore not included in meta‐analysis for this outcome |
|
|
A. 1. Nasal blockage on waking 2. Nasal blockage during the rest of the day 3. Sneezing 4. Rhinorrhoea B. 1. Overall assessment of symptoms by patient 2. Overall assessment of symptoms at clinic visit |
A. Measured by patient on a scale of 0 to 3 on a daily card (0 is better) B. 1 . Measured by patient on a scale of 0 to 3 on a daily card (0 is better) 2. VAS, 0 to 10 cm (10 = worst symptoms) |
A. Range 0 to 12 B.1. Range 0 to 3 B.2. Range 0 to 10 |
No data reported for non‐allergic rhinitis subgroup separately, therefore not included in meta‐analysis. The study states that there were no differences between allergic and non‐allergic rhinitis patients |
|
|
1. Sneezing 2. Stuffy nose 3. Runny nose 4. Nose blowing 5. Post‐nasal drip |
Duration in hours per symptom measured |
All 5 symptoms combined to determine overall duration of patients' symptoms. Reported as percentage of days during which all 5 of the symptoms lasted 1 hour or less and the percentage of days during which at least one of the 5 symptoms lasted 4 hours or more |
No data reported for non‐allergic rhinitis subgroup separately for this outcome, therefore not included in meta‐analysis. There are data on 'responder/non‐responder'. |
|
|
— |
— |
— |
Not included in meta‐analysis as TNSS data are reported in relationship to cold dry air provocation |
|
|
A. Overall VAS B. Individual symptom VAS: 1. Congestion 2. Sneezing 3. Itching 4. Rhinorrhoea |
Measured on a VAS: 0 to 10, 0 is better |
Range (0 to 10), measurement at week 8 |
— |
|
|
1. Sneezing 2. Stuffiness 3. Runny nose 4. Nose blowing 5. Post‐nasal drip |
Patient evaluation, sum of 5 symptoms numerically assessed as absent (1), mild (2), moderate (3), or severe (4); range 1 to 4 |
Range (5 to 20), measurement at week 4 |
— |
|
|
1. Rhinorrhoea 2. Nasal congestion 3. Post‐nasal drip 4. Eye itching/burning 5. Eye tearing/watering 6. Eye redness |
4‐point categorical scale of 0 to 3 (none, mild, moderate, severe), measured by patient on paper diary card, in AM and PM, as instantaneous (i) and over previous 12 hours (reflective). Instantaneous score measured in AM, and reflective in both AM and PM ‐ measured during screening and treatment periods |
Combined 3 reflective individual nasal symptom scores (range 0 to 9) Measured in AM and PM, which were averaged to arrive at the final daily value (daily rTNSS) |
Compared data from week 4 to baseline to arrive at change from baseline TNSS. We included the reflective TNSS (rTNSS) in the meta‐analysis, not the instantaneous TNSS (iTNSS). |
|
|
Individual rhinitis symptoms: 1. Daily and nightly sneezing 2. Nasal congestion 3. Rhinorrhoea Total nasal symptoms |
0 if absent, 1 if they lasted less than 30 minutes, 2 if between 30 minutes and 2 hours, and 3 if longer than 2 hours |
Total nasal symptoms: the sum of the nasal symptom scores (sneezing, congestion, and rhinorrhoea, as well as the total) |
No separate data for non‐allergic rhinitis subgroups |
|
|
1. Sneezing 2. Runny nose 3. Stuffy nose 4. Eye itch 5. Throat itch |
Measured on 0 to 6 scale (0 is better) |
Sum score of symptoms scores for sneezing, runny nose, stuffy nose, eye itching and throat itching measured post‐treatment, each measured on 0 to 6 scale (range 0 to 30). To this was added the number of tablets and nasal sprays required to control nasal symptoms for the preceding 12‐hour period; possible scores range from 0 to 40 |
Unusual method of measurement of TNSS |
|
|
1. Nasal itching 2. Sneezing 3. Rhinorrhoea 4. Nasal obstruction |
Measured on 0 to 3 scale (0 is better) |
Range 0 to 12; assessed on day 1 and at 8 weeks |
Reported TNSS as continuous data and change from baseline. We have included the first approach |
|
|
Overall severity of symptoms Individual rhinitis symptoms: 1. Sneezing 2. Runny nose 3. Nose blowing 4. Post‐nasal drip |
Overall severity of symptoms: scale from 0 to 3 Individual nasal symptoms: scale from 0 to 4 |
— |
— |
|
|
1. Nasal obstruction
2. Post‐nasal drip
|
Measured on a VAS, 0 to 100 (0 is better) |
Range 0 to 300, measured at both 2 weeks and 4 weeks |
TNSS reported as change from baseline. Combination of 3 studies |
|
|
TNSS: total nasal symptom score
|
||||
|
Study |
Findings |
|
Concluded that treatment with fluticasone propionate provided significantly greater relief from the symptom of nasal obstruction compared with placebo over the entire 3‐month treatment period. Patients' subjective assessments of nasal obstruction after medical treatment correlated with the results of objective testing. |
|
|
This study was not included in the meta‐analysis because of lack of quality of the study data. Firstly, the study presents unexpected data, with disappearance of the benefit of intranasal corticosteroids with longer follow‐up. Secondly, including the study in the meta‐analysis resulted in a high level of heterogeneity. The SD values that are presented in the study do not match with the presented means, n and P values. The data make more sense if the as‐presented SD values should actually be standard error of the mean (SEM), which was confirmed by a re‐analysis. As the authors did not reply to our question regarding the above, we decided to not include this study in the meta‐analysis. The study did show a beneficial effect of intranasal corticosteroids over placebo, however this effect disappeared with longer follow‐up. |
|
|
This study was not included in the meta‐analysis for this outcome as it did not report numerical data for the non‐allergic rhinitis subgroup. They did conclude that fluticasone propionate reduces total symptoms, improves individual symptoms (mainly obstruction) and achieves a significant overall improvement in non‐allergic rhinitis compared to placebo. |
|
|
Reported a statistically significant difference in symptoms in favour of intranasal corticosteroids (it did not report P values for rhinorrhoea and post‐nasal drip so we were not able to calculate a SD). |
|
|
Reported no numerical data on original TNSS, so it could not be included in the meta‐analysis for this outcome. It did report data that could be translated into proportions of 'responders/non‐responders'. The study did not find significant differences between intranasal corticosteroids and placebo. The study converted TNSS into a dichotomous outcome: improved versus unimproved. Improvement was defined as a reduction of at least 1 point in the overall symptom score. No numerical data on original TNSS was provided, therefore this study was not included in the meta‐analysis for this outcome. |
|
|
This study was not included in the meta‐analysis as no data on TNSS were reported. The study did report data on 'responders/non‐responders' with 29/39 responders in the intranasal corticosteroids group and 12/39 responders in the placebo group, favouring intranasal corticosteroids with an odds ratio (OR) of 0.44 (95% confidence interval (CI) 0.24 to 0.64). |
|
|
This study was not included in the meta‐analysis as only P values were reported. Patients reported subjective symptom scores on a scale of 0 to 5 for nasal obstruction, anterior rhinorrhoea, posterior rhinorrhoea, sneezing and facial pain. When the composite scores for all 5 symptoms were compared, there was a significant difference between beclomethasone dipropionate and baseline (P = 0.01) and beclomethasone dipropionate and placebo (P = 0.02) in favour of beclomethasone dipropionate. |
|
|
This study on 2 types of intranasal corticosteroids versus placebo in perennial rhinitis patients (allergic and non‐allergic rhinitis) reported a number of individual rhinitis symptoms and an overall assessment of symptoms, but no separate data on non‐allergic rhinitis patients were presented. However, the study does state that there were no differences between allergic and non‐allergic rhinitis. This study reported a significant improvement with intranasal corticosteroids versus placebo in perennial allergic rhinitis, with fluticasone propionate aqueous nasal spray (200 µg) as effective as beclomethasone dipropionate µg twice daily. |
|
|
This study was not included in the meta‐analysis as no data on TNSS were reported. The study did report data on responders/non‐responders with 6 of 14 responders in the intranasal corticosteroids group and 8 of 18 responders in the placebo group with an OR of 0.02 (95% CI 0.33 to 0.36), therefore not a significant difference. |
|
|
This study was not included in the meta‐analysis for this outcome as it did not report enough numerical data for the non‐allergic rhinitis subgroup. They concluded that after 6 months 6 of 9 non‐allergic rhinitis patients were successfully treated with intranasal corticosteroids and 3 of 9 non‐allergic rhinitis patients were unsuccessfully treated with intranasal corticosteroids. They concluded that their results (in favour of intranasal corticosteroids over placebo) in those in whom a possible allergic component could be identified were not different from those of the whole group. |
|
|
This study was not included in the meta‐analysis because we decided to only include studies with an intranasal corticosteroid dosage of 200 µg to 400 µg. This study uses an intranasal corticosteroid dosage of 2000 µg. The study did report a significant improvement in nasal symptoms in non‐allergic rhinitis after an 8‐week treatment period with intranasal flunisolide. |
|
|
This study was not included in the meta‐analysis for this outcome because it did not report numerical data for the non‐allergic rhinitis subgroup. They concluded that flunisolide nasal solution seems to be effective in both allergic rhinitis and vasomotor rhinitis patients, although it seems to be more effective in an allergic state. |
|
|
SD: standard deviation
|
|
|
Outcome or subgroup title |
No. of studies |
No. of participants |
Statistical method |
Effect size |
|
1 Total nasal symptom score, follow‐up ≤ 4 weeks Show forest plot |
4 |
131 |
Std. Mean Difference (IV, Random, 95% CI) |
‐0.74 [‐1.15, ‐0.33] |
|
1.1 Budesonide |
2 |
74 |
Std. Mean Difference (IV, Random, 95% CI) |
‐0.83 [‐1.30, ‐0.35] |
|
1.2 Flunisolide |
1 |
15 |
Std. Mean Difference (IV, Random, 95% CI) |
0.11 [‐0.90, 1.13] |
|
1.3 Beclomethasone dipropionate |
1 |
42 |
Std. Mean Difference (IV, Random, 95% CI) |
‐0.96 [‐1.60, ‐0.32] |
|
2 Total nasal symptom score, follow‐up > 4 weeks Show forest plot |
3 |
85 |
Std. Mean Difference (IV, Random, 95% CI) |
‐0.24 [‐0.67, 0.20] |
|
2.1 Fluticasone propionate |
1 |
31 |
Std. Mean Difference (IV, Random, 95% CI) |
‐0.17 [‐0.87, 0.54] |
|
2.2 Flunisolide |
2 |
54 |
Std. Mean Difference (IV, Random, 95% CI) |
‐0.28 [‐0.84, 0.27] |
|
3 Total nasal symptom score (change from baseline), follow‐up ≤ 4 weeks Show forest plot |
4 |
1465 |
Std. Mean Difference (IV, Fixed, 95% CI) |
‐0.15 [‐0.25, ‐0.05] |
|
3.1 Fluticasone furoate (110 µg fluticasone furoate equals around 200 µg FP, BUD or BDP) |
2 |
794 |
Std. Mean Difference (IV, Fixed, 95% CI) |
‐0.06 [‐0.20, 0.08] |
|
3.2 Budesonide |
1 |
20 |
Std. Mean Difference (IV, Fixed, 95% CI) |
‐0.74 [‐1.65, 0.17] |
|
3.3 Fluticasone propionate |
1 |
651 |
Std. Mean Difference (IV, Fixed, 95% CI) |
‐0.24 [‐0.40, ‐0.09] |
|
4 Significant adverse event: epistaxis Show forest plot |
4 |
1174 |
Risk Difference (M‐H, Fixed, 95% CI) |
0.04 [0.01, 0.06] |
|
4.1 Fluticasone furoate (110 µg fluticasone furoate equals around 200 µg FP, BUD or BDP) |
2 |
801 |
Risk Difference (M‐H, Fixed, 95% CI) |
0.02 [‐0.00, 0.05] |
|
4.2 Budesonide |
1 |
44 |
Risk Difference (M‐H, Fixed, 95% CI) |
0.05 [‐0.07, 0.16] |
|
4.3 Mometasone furoate |
1 |
329 |
Risk Difference (M‐H, Fixed, 95% CI) |
0.07 [0.01, 0.13] |
|
5 Objective measurement of airflow: peak flow rate (expiratory) Show forest plot |
1 |
11 |
Std. Mean Difference (IV, Random, 95% CI) |
0.78 [‐0.47, 2.03] |
|
6 Objective measurement of airflow: rhinomanometry Show forest plot |
1 |
44 |
Std. Mean Difference (IV, Fixed, 95% CI) |
‐0.46 [‐1.06, 0.14] |
|
7 Other adverse events Show forest plot |
3 |
1130 |
Risk Ratio (M‐H, Fixed, 95% CI) |
0.99 [0.87, 1.12] |
|
7.1 Fluticasone furoate (110 µg fluticasone furoate equals around 200 µg FP, BUD or BDP) |
2 |
801 |
Risk Ratio (M‐H, Fixed, 95% CI) |
0.97 [0.81, 1.16] |
|
7.2 Mometasone furoate |
1 |
329 |
Risk Ratio (M‐H, Fixed, 95% CI) |
1.02 [0.87, 1.19] |
|
Outcome or subgroup title |
No. of studies |
No. of participants |
Statistical method |
Effect size |
|
1 Total nasal symptom score, follow‐up > 4 weeks Show forest plot |
1 |
80 |
Mean Difference (IV, Fixed, 95% CI) |
‐0.25 [‐0.69, 0.19] |
|
1.1 Budesonide |
1 |
80 |
Mean Difference (IV, Fixed, 95% CI) |
‐0.25 [‐0.69, 0.19] |
|
2 Total nasal symptom score (change from baseline), follow‐up ≤ 4 weeks Show forest plot |
1 |
63 |
Mean Difference (IV, Random, 95% CI) |
‐0.5 [‐1.92, 0.92] |
|
2.1 Triamcinolone acetonide |
1 |
63 |
Mean Difference (IV, Random, 95% CI) |
‐0.5 [‐1.92, 0.92] |
|
3 Quality of life (SF‐12v2) Show forest plot |
1 |
80 |
Mean Difference (IV, Fixed, 95% CI) |
‐1.30 [‐3.60, 1.00] |
|
3.1 Budesonide |
1 |
80 |
Mean Difference (IV, Fixed, 95% CI) |
‐1.30 [‐3.60, 1.00] |
|
4 Objective measurement of airflow: inspiratory peak flow rate (change from baseline) Show forest plot |
1 |
63 |
Mean Difference (IV, Random, 95% CI) |
‐6.17 [‐15.25, 2.91] |
|
4.1 Triamcinolone acetonide |
1 |
63 |
Mean Difference (IV, Random, 95% CI) |
‐6.17 [‐15.25, 2.91] |
|
5 Other adverse events Show forest plot |
1 |
80 |
Risk Ratio (M‐H, Fixed, 95% CI) |
2.0 [0.19, 21.18] |
|
5.1 Budesonide |
1 |
80 |
Risk Ratio (M‐H, Fixed, 95% CI) |
2.0 [0.19, 21.18] |
|
Outcome or subgroup title |
No. of studies |
No. of participants |
Statistical method |
Effect size |
|
1 Total nasal symptom score, follow‐up ≤ 4 weeks Show forest plot |
1 |
40 |
Mean Difference (IV, Fixed, 95% CI) |
1.60 [0.03, 3.16] |
|
1.1 Budesonide |
1 |
40 |
Mean Difference (IV, Fixed, 95% CI) |
1.60 [0.03, 3.16] |
|
Outcome or subgroup title |
No. of studies |
No. of participants |
Statistical method |
Effect size |
|
1 Total nasal symptom score, follow‐up ≤ 4 weeks Show forest plot |
1 |
48 |
Mean Difference (IV, Fixed, 95% CI) |
‐1.5 [‐12.24, 9.24] |
|
1.1 Beclomethasone |
1 |
48 |
Mean Difference (IV, Fixed, 95% CI) |
‐1.5 [‐12.24, 9.24] |
|
Outcome or subgroup title |
No. of studies |
No. of participants |
Statistical method |
Effect size |
|
1 Total nasal symptom score, follow‐up > 4 weeks Show forest plot |
2 |
242 |
Std. Mean Difference (IV, Fixed, 95% CI) |
0.75 [0.48, 1.02] |
|
1.1 Fluticasone propionate (dosage unclear) |
1 |
162 |
Std. Mean Difference (IV, Fixed, 95% CI) |
0.37 [0.06, 0.68] |
|
1.2 Budesonide |
1 |
80 |
Std. Mean Difference (IV, Fixed, 95% CI) |
1.85 [1.32, 2.38] |
|
2 Quality of life (SF12‐v2) Show forest plot |
1 |
80 |
Mean Difference (IV, Fixed, 95% CI) |
‐7.20 [‐9.77, ‐4.63] |
|
2.1 Budesonide |
1 |
80 |
Mean Difference (IV, Fixed, 95% CI) |
‐7.20 [‐9.77, ‐4.63] |
|
3 Other adverse events Show forest plot |
2 |
242 |
Risk Ratio (M‐H, Fixed, 95% CI) |
0.26 [0.07, 1.01] |
|
3.1 Fluticasone propionate (dosage unclear) |
1 |
162 |
Risk Ratio (M‐H, Fixed, 95% CI) |
0.09 [0.00, 1.54] |
|
3.2 Budesonide |
1 |
80 |
Risk Ratio (M‐H, Fixed, 95% CI) |
0.5 [0.10, 2.58] |
|
Outcome or subgroup title |
No. of studies |
No. of participants |
Statistical method |
Effect size |
|
1 Total nasal symptom score, follow‐up ≤ 4 weeks Show forest plot |
1 |
28 |
Mean Difference (IV, Fixed, 95% CI) |
‐1.30 [‐2.97, 0.37] |
|
1.1 Mometasone furoate |
1 |
28 |
Mean Difference (IV, Fixed, 95% CI) |
‐1.30 [‐2.97, 0.37] |
|
2 Quality of life (SNOT‐22) Show forest plot |
1 |
28 |
Mean Difference (IV, Fixed, 95% CI) |
‐6.00 [‐19.75, 3.75] |
|
2.1 Mometasone furoate |
1 |
28 |
Mean Difference (IV, Fixed, 95% CI) |
‐6.00 [‐19.75, 3.75] |
|
3 Objective measurement of airflow: peak nasal inspiratory flow Show forest plot |
1 |
28 |
Mean Difference (IV, Fixed, 95% CI) |
‐9.20 [‐33.96, 15.56] |
|
3.1 Mometasone furoate |
1 |
28 |
Mean Difference (IV, Fixed, 95% CI) |
‐9.20 [‐33.96, 15.56] |