Printer Friendly

A comparison of two nasal irrigation solutions with no treatment in chronic sinusitis sufferers.


We compared the efficacy of two nasal irrigation solutions with each other, using two 3-week, blind trials in 10 participants with chronic sinusitis. To control for regression toward the mean, we also included an untreated control group of 18 participants. All participants continued their customary over-the-counter or prescribed treatments. As measured by the Sino-Nasal Outcome Test (SNOT-20), no hypertonic, buffered saline reduced sinus symptoms by 20% relative to controls (p = 0.37, 10 participants). For a commercial, isotonic saline solution with natural antimicrobial agents, the reduction was a statistically significant 27% (p = 0.046, 10 participants); however, the difference between the two solutions was not statistically significant in this small study (p = 0.43). These treatment benefits are similar to those reported by others using long-term pharmacological and surgical treatments. Surprisingly, three weeks after participants stopped using sinus irrigation, there was no significant relapse of symptoms. Improvements in 18 untreated participants (regression to the mean) ranged from 3% to 16% by 4 measures; this finding shows the importance of including untreated participants in future studies, in lieu of a placebo group.


Chronic sinusitis is one of the most commonly reported reasons for patient visits to a family physician. During 2004, 14% of adults in the US (31 million) reported having sinusitis. It caused 73 million restricted-activity days per year during 1990 to 1992, and incurred $5.8 billion in direct medical costs in 1996. In a prospective study of adult patients, the most common severe symptoms were nasal obstruction and facial congestion, often accompanied by headache and fatigue. Average annual treatment involved 2.7 antibiotic courses and use of steroid nasal sprays and prescription antihistamines for 18 weeks and 16 weeks, respectively. Persisting symptoms lead to about 200,000 sinus surgeries per year, costing about $800 million. Sinusitis often occurs in the setting of chronic allergic or vasomotor rhinitis in which a viral cold will infect the upper respiratory tract and cause obstruction. Once obstruction has occurred, the stagnated environment favors the growth of bacteria and fungi. Repeated courses of antibiotics, while reducing the bacterial count, often do not change the obstructive symptoms. In fact, fungal overgrowth secondary to recurrent use of antibiotics can complicate the picture and prolong symptoms.

Self-irrigation of the nasal passageway with saline solution has been used for many years to treat sinusitis. ft may help by removing and thinning sinus mucus and by decreasing inflammation caused by antigens, bacteria, and fungi. Both isotonic and hypertonic saline have been studied as adjunctive therapy for acute and chronic or recurrent sinusitis. Authors of several studies have concluded that saline irrigation improves sinusitis symptoms and is safe and acceptable to users. However, uncertainties remain about the preferred salinity and pH of the solution, the best methods of irrigation, and which kinds of sinusitis are most responsive. Rabago et al. advocate irrigation with 2% (hypertonic) saline buffered with sodium bicarbonate, 150 mL administered daily through each nostril by gravity flow from a commercially available nasal cup. This method is based on an ancient cleansing ritual of yoga.

One commercial nasal irrigation solution attempts to improve on isotonic, buffered saline solution by adding natural antifungal and antibacterial ingredients. The product's supplier commissioned us to evaluate its efficacy compared with the well-studied, buffered 2% hypertonic saline. According to anecdotal reports received by the supplier, its patented product seems clearly superior to plain saline for many persons. It has been on the market for 4 years, has over 30,000 regular users, and is sold under the name Sinus Cleanser.

We report here the results of a small, blind comparison of Sinus Cleanser with plain, buffered 2% hypertonic saline solution, administered according to the manufacturer's instructions. We also compared both solutions with an untreated control group, a comparison that is rarely performed but which is necessary to control for regression toward the mean. Finally, we measured symptoms after stopping sinus irrigation, to gain new information about the rate of relapse of symptoms.


The study was approved by the institute's institutional review board. The study population consisted of participants who responded either to a leaflet at a local lecture on chronic sinusitis or to invitations from the lead investigator in his general medical office. About 100 interested persons were interviewed by telephone. Those who met the inclusion criteria (over age 18, in reasonably good health, sinusitis symptoms for at least 3 months, failure to receive benefit from at least one over-the-counter or prescription medication for sinusitis, and no smoking within the past 6 months) were invited to attend one of two scheduled meetings (10 days after the lecture). Forty persons attended. The study and informed consent form were explained in depth. All 40 attendees agreed to participate. They signed the consent form, completed an initial (baseline) 20-question Sino-Nasal Outcome Test (SNOT-20), and gave a blood sample for analysis of white blood cells. They were assigned in the order of their arrival at the meeting to a rotating sequence of 4 experimental groups (Group 1 = participants 1, 5, 9, etc.; Group 2 = participants 2, 6, 10, etc. This method assured nearly equal numbers of participants in each group, with effective randomization.

The study consisted of two phases, each 3 weeks long. All participants, including the control participants, were instructed to continue any other treatments that they were using for sinusitis symptoms. At the meeting, participants in Groups 1 and 2 received their initial supply of nasal irrigation solution and detailed written and pictorial instructions from the supplier for its use. They were asked to use the solution on themselves. Trained study coordinators evaluated each participant's technique. The importance of proper technique was strongly emphasized. Participants were given three bottles (4 fluid ounce, 118 mL) of solution supplied by the study sponsor, labeled Alpha or Beta, according to their group assignments. The identity of the solutions was not revealed to participants until after they completed phase 2 (and the third and final SNOT-20 questionnaire), and no evaluations of signs or symptoms were performed by staff who were aware of the identity of the solutions. Additional bottles were supplied to a few participants as needed.

The active, isotonic (Active) solution contained wild indigo (Baptisia tinctoria, a homeopathic agent), oregano, grapefruit seed extract, and colloidal silver in buffered, isotonic saline. The Saline solution was 2% hypertonic solution buffered with sodium bicarbonate, as used by Rabago.

At the end of phase 1 (3 weeks), all participants returned to a second meeting at which they repeated the SNOT-20 questionnaire and blood sample and received instructions for phase 2. Groups 3 and 4 received their initial supply of sinus irrigation solution and instructions for its use, as described for Groups 1 and 2. At the end of phase 2, all participants returned for a third and final SNOT-20 questionnaire and blood sample.

At the end of both phases, participants were asked to write on the back of the SNOT-20 form the number of bottles of solution that they used and any comments about their experience. Four months after the end of phase 2, participants were asked by mail or telephone to complete a questionnaire about their evaluation and continued use of sinus irrigation.

Except where noted, we report SNOT-20 scores as the sum of 20 symptom ratings, where each rating ranges from 0 to 5 (so the sum can range from 0 to 100). This method permits direct comparison with results reported by Rabago et al. Alternatively, one can calculate the average rating per symptom reported, with a possible range of 0 to 5, as suggested by the authors of the SNOT-20 questionnaire. Results of the two methods are equivalent, differing by a factor of exactly 20, except when participants do not rate all 20 symptoms (infrequently in our study). We also briefly mention what we call SNOT-5 scores, which are limited to the 5 symptoms designated at baseline as most important by each participant.

The written instructions given for the use of the nasal irrigation solutions were as follows:

1. Lie on the floor or a bed on your back.

2. Open your mouth and slowly breathe through your mouth.

3. Begin by instilling just a few drops of the cleansing solution into one of your nostrils.

4. It is OK if some of the solution goes down the back of your throat. Continue breathing and slowly instill the solution into both of your nostrils (you may alternate, back and forth between nostrils).

5. Try to sense closing the back of your throat (soft palate), as when you swim under water.

6. Squeeze the cleansing solution into your nostrils until they are almost full (starting to drip out).

7. Hold for 60 seconds while breathing through your mouth. (Not harmful if some is swallowed.)

8. With thumb and forefinger in each nostril, stand up and bend head over sink, holding the solution in this position for an additional 60 seconds.

7. Then remove fingers and expel all fluids into sink and finish by gently blowing your nose.

8. Take a shower or arrange to catch the residual nasal drainage for the next few minutes.

9. Repeat this procedure daily.

10. Wait at least two hours between this procedure and using any over-the-counter or prescription nasal sprays. (Use this treatment procedure at a time so as to not dilute or rinse away any medical treatment that you are concurrently receiving.)

Following the sinus cleansing procedure step by step--as directed in your consent form--is of paramount importance. The validity of this study depends on the participant's willingness to follow the instructions as directed.

If you are unable to follow the procedure as directed, withdrawal from the study is requested.

Statistical Analysis

To compare changes in SNOT-20 scores between treatment groups, we used Mann-Whitney tests. We chose these non parametric tests rather than the commonly used two-sample t-tests, because of evidence for significantly skewed distributions of symptom changes. Similarly, we used nonparametric sign tests to assess uncontrolled changes in SNOT-20 scores within groups (instead of one-sample t-tests). To calculate percentage changes of SNOT-20 scores compared with control groups, we reduced the baseline scores in treated groups by the decrease found in the control groups. All p-values are two-tailed, from NCSS 2004 statistical software (Kaysville, Utah).


The participants' mean age was 63 (range 41 to 85). Slightly over half were female (58%). They reported histories of seasonal allergies (80%), deviated septum (23%), nasal surgery (20%), asthma (17%), and known nasal polyps (9%).

Two of 40 entering participants did not complete phase 1 and are not included in the analyses; neither was receiving study treatment then (Groups 3 and 4). All other participants (95%) completed both phases. Follow-up questionnaires were obtained from 35 of the 38 participants (92%).

Randomization produced baseline group average SNOT-20 scores close to the overall mean of 36.4, with no significant differences between groups. As expected for regression toward the mean, SNOT-20 scores declined slightly (-1.2 points or -3.4%) in the 18 control participants in phase 1.

Relative to the decline of -1.2 in control participants, SNOT-20 scores declined in phase 1 by an average -10.5 (-27%) in 10 Active-solution participants (Group 1, p = 0.046) and by -7.9 (-20%) in 10 Saline-solution participants (Group 2, p = 0.37). Similarly, when the Group 3 and 4 participants received Active and Saline treatments in phase 2, larger average declines occurred in Active participants than in Saline participants. Combining these two, independent comparisons of the Active and Saline solutions (Table 2), the Active solution in 19 participants decreased SNOT-20 scores by an average of 3.8 points more than the Saline solution in 19 participants. However, this 3.8 difference has a relatively large standard error of 4.9 and is not statistically significant (p = 0.39 by Mann-Whitney test).

Although Saline treatment decreased SNOT-20 scores by an average of -7.0 in 19 participants (relative to the decline of -1.2 in 18 untreated controls), and although this same hypertonic saline treatment is proven effective in other studies, we did not have sufficient participants for statistical significance. However, we did reach statistically significance for the Active solution relative to control participants: -10.5 (-27%) in 10 participants in phase 1 (p = 0.046) and -10.8 (-28%) in the combined 19 participants (p = 0.023).

Nearly all previous studies of sinusitis treatments are based on comparing symptom ratings before and after treatment, without comparison to an untreated control group. If we analyze our data in this uncontrolled, "before vs. after" way, the effects of sinus irrigation appear to be slightly larger and considerably more statistically significant than we report. For example, the uncontrolled decrease in SNOT-20 scores in phase 1 was -11.7 for Active solution and -9.1 for Saline solution. The corresponding p-values decrease from 0.046 and 0.37 reported above, to 0.002 and 0.17 by sign test. According to commonly used paired t-tests, the p-values would be 0.011 and 0.12, but t-tests are questionable for our changes in SNOT-20 scores, which appear substantially skewed (skewness ~ -1.0).

When Group 1 and 2 participants stopped sinus irrigation in phase 2, we found that SNOT-20 scores changed very little on average during 3 weeks of follow-up. The 3-week changes of 1.5 and -0.3 are much smaller than the treatment benefits in phase 1 and are statistically indistinguishable from zero.

The available reports of bottles used during 3 weeks averaged 1.3 [+ or -] 0.5 SD in 14 Active participants (range 0.8 to 2.5), only about half of the average 2.5 [+ or -] 1.2 SD in 12 Saline participants (range 1 to 4.5; p = 0.003 by t-test). Five of 20 Active participants mentioned stinging, burning, or side effects compared with only one mention among 18 Saline participants (p = 0.18 by Fisher's exact test). Two Active participants reported diminished sense of taste and smell. These findings suggest that the Active solution was less comfortable and accepted than the Saline solution (though it might be more effective).

The ~4-month follow-up questionnaire was completed by 35 of 38 participants (92%). Most respondents reported continuing to use sinus irrigation (71 cY0), usually with the leftover Active solution that we provided. The continuing percentage was 61% among 18 Active participants and 76% among 17 Saline participants, but this difference was not statistically significant (p = 0.47 by Fisher's exact test). Four participants were using various other commercially available products. Frequency of use was "daily" (8 participants), "a few times a week" (6), "weekly" (6), and "less than weekly" (4). Those who quit did so nearly all at the end of the experiment, mostly citing lack of benefit. One mentioned time constraints and one cited side effects. While they were in the experiment, large percentages reported satisfaction with the solutions (71%), feeling better overall (78%), breathing easier overall (77%), experiencing less severe symptoms (70%), and they would recommend the treatment to others (94%). Of 28 participants using antibiotics or nasal sprays, 36% reported reducing their use (5 Active and 5 Saline).

The blood cell counts were performed on the ABBOTT Cell-DYN 1 700 and slides stained with Protocol Hema-Quik II Wright Giemmsa stain, Fisher Scientific Company. The number of eosinophils and basophils were noted per 100 white blood cells. cell counts at baseline (mean [+ or -] SD for 38 participants): white blood cells 6.6 [+ or -] 1.4 (1,000 per [micro]L), eosinophils 2.6 [+ or -] 2.2 (%), and basophils 0.9 [+ or -] 0.9 (%). There were no significant changes or consistent trends after sinus irrigation.

For purposes of evaluating regression toward the mean in 18 control participants (phase 1), we calculated its size for 4 variants of SNOT scores.


Three weeks of sinus irrigation reduced sinusitis symptoms on average by 20% to 27% relative to controls, and was well rated by most participants. Many participants reduced their use of other treatments. These results are similar to previous reports. In a prior, uncontrolled study of 2% hypertonic saline irrigation, average SNOT-20 scores declined in 12 months from 43.5 [+ or -] 5.7 (SE) to 28.4 [+ or -] 4.8, a 35% decline (14 participants, average age about 43). This decline compares to our uncontrolled decline in 3 weeks of 24% with the same irrigation solution, but in less-symptomatic and older participants. Other treatments used by a university otolaryngology clinic reduced uncontrolled SNOT-20 per-symptom scores by 32% after 6 months and 26% after 12 months. Thus, self-treatment with sinus irrigation may quickly reduce symptoms by average amounts similar to long-term medical treatment.

In our study, two independent comparisons of plain, 2% hypertonic Saline with the Active, isotonic saline showed consistent trends in favor of the Active solution containing added antimicrobial ingredients. However, the differences were not large enough to be statistically significant (p = 0.43 for 10 participants each in phase 1 and p = 0.39 for 19 participants in each group). Previous studies suggest that hypertonic saline may produce superior mucociliary clearance and symptom relief. Thus, it is possible that the lower tonicity of the Active solution reduced its effectiveness relative to our Saline solution. From previous studies of other additives it is not clear yet whether additives are beneficial.

With one exception known to us, no previous study of sinus irrigation (or any study using SNOT-20 data) included a control group. Uncontrolled studies can measure only changes in symptom scores before and after treatment, the method suggested, for example, by authors of the SNOT-20 questionnaire. Thus uncontraed studies include in "treatment effects" the possible contribution of placebo effects and regression toward the mean (RTM). RTM is caused by a form of selection bias. It is expected whenever study participants are selected, or self-selected, on the basis of study variables (sinus symptoms in this case). The more their symptoms may be at a temporary high, the more likely they are to be selected, or to volunteer, for a study of their symptoms. After entry, their symptoms are likely to regress somewhat, even without treatment.

It seems difficult to imagine a valid placebo control for sinus irrigation. However, an untreated control group controls for RTM and is quite feasible. It may serve most of the purpose of a placebo group, because RTM often may account for most or all of what are commonly called "placebo effects." (Only rare reports of "placebo effects" include both a placebo group and an untreated group that are needed to show a difference between them.)

Under our conditions, RTM during 3 weeks in 18 control participants was about -3% for SNOT-20 scores and about -6% for SNOT-20 per-symptom scores. Understandably, RTM was much greater, -15% and -16%, for each participant's 5 most important symptoms (SNOT-5 and SNOT-5 per-symptom scores). The 5 symptoms judged most important at baseline presumably played the greatest role in each participants' decision to join the study and were also the most likely to regress substantially without treatment. These substantial RTM effects strongly support the routine use of untreated control groups in studies of sinusitis treatments. Uncontrolled, "before vs. after" comparisons are expected to overestimate treatment effects by these amounts, and to greatly overestimate their statistical significance. Uncontrolled comparisons also overestimate treatment effects by the amount of any true placebo effects not included in RTM.

We know of no other reports of RTM for SNOT-20 scores. However, Rabago et al. reported data that allow us to calculate RTM for three other assessment measures in their study of sinus irrigation: Rhinosinusitis Disability Index (RSDI), Single-Item Symptom Severity Assessment (SIA), and Medical Outcomes Survey Short Form 12 (SF-12). After 1.5 months in 24 untreated control participants, RTM was -9% for RSDI, -0.5% for SIA, and -9% for SF-12. At 3 months and 6 months all values were also negative, cumulatively averaging -8% for RSDI, -6% for SF-12, and -3% for SIA. These values are similar to our values for SNOT-20, but smaller than we find for SNOT-5, as one might expect. However, such comparisons are only approximate, because RTM is expected to vary with each experiment, depending on many factors, including the nature of the assessment and the time interval between participant selection and baseline measurements. (Which is why results from a control group in one experiment are ill suited for use in another.)

After participants used sinus irrigation for 3 weeks and then stopped, there was surprisingly little increase in symptom scores during 3 weeks of follow-up. We know of only one comparable report. After a 4-month comparison of 3.5% and 0.9% saline irrigation solutions in 30 children with chronic sinusitis (aged 3 to 16 years), the authors write in their abstract, "Clinical observation 1 month after the end of the study showed no change compared with the end of the study in both groups." However, there is no related data or mention in the text. About 4 months after the end of our study, substantial numbers of participants reported continuing to use sinus irrigation at intervals of daily, a few times per week, weekly and less than weekly. Apparently there is some perceived benefit not reflected in SNOT-20 scores during 3 weeks of non-use, and wide individual variability in the irrigation frequency needed to achieve this benefit.

Similarly, we suggest that there may be no single tonicity or pH that is best for all. Though 2% to 5% hypertonic saline solutions may be superior to isotonic saline in objective ways, they may be more likely to cause burning or stinging, and thus be less used by sensitive participants. For those who may find 2% hypertonic solution uncomfortable, they might be advised to try dilution with an equal quantity of water, and then to experiment to find the best concentration for them individually (possibly more than 2% for some).


In this study of participants who continued their usual treatments for sinusitis, addition of nasal irrigation with the commercial solution gave statistically significant relief of symptoms compared with a control group who did not use sinus irrigation. Participants who added 2% hypertonic nasal irrigation solution benefited somewhat less (a difference not statistically significant in this small study, compared with either the control participants or the commercial solution). Nasal irrigation was well accepted by most participants, and 71% of them continued to use it 4 months after the end of the study. About one-third of participants in both groups were able to reduce their use of antibiotics and other treatments. Given the high incidence and morbidity of sinusitis, nasal irrigation seems underutilized by family physicians and otolaryngologists. Further studies are warranted to better determine the most effective irrigation solutions and the best means of using them.

Untreated control groups are needed to control for regression toward the mean (caused by participants who selectively enter studies near the peak of their symptoms and subsequently improve somewhat without treatment).


We thank Michele Barton, Christina Mackay, and Rachel Carney for carefully coordinating participant contacts and data collection. We thank the Bio-Center Laboratory personnel who worked overtime to perform blood cell counts at each of the participant meetings, including lab director James Jackson (slide preparation, and interpretation), technicians Sharon Neatherly and Karen Hart (phlebotomy, slide preparation and staining and CBC testing), and Megan Powers (phlebotomy and slide preparation).

Ronald E. Hunninghake, MD, and Donald R. Davis, PhD

Bio-Communications Research Institute

3100 N. Hillside Ave.

Wichita, Kansas 67219

by Ronald E. Hunninghake, MD, and Donald R. Davis, PhD

Dr. Hunninghake is the chief medical officer of the Riordan Clinic. He is a 1976 graduate of the University of Kansas School of Medicine. Dr. Hunninghake trained alongside Dr. Hugh D. Riordan for 17 years. He is current member of several boards of directors, including the International Schizophrenia Foundation, Japanese College of Intravenous Therapies, Pure North S'Energy Foundation, and the Riordan Clinic. Dr. Hunninghake is the 2011 recipient of the Orthomolecular Doctor of the Year award presented by the International Society for Orthomolecular Medicine, and lectures nationally and internationally on a range of topics, including IVC and cancer.

Dr. Davis is a nutritional biochemist and expert on food nutritional quality. He published his first paper on the measurement of food quality in 1976 with R.J. Williams in the American Journal of Clinical Nutrition. It is titled "Potentially Useful Criteria for Judging Nutritional Adequacy" (29:710-715).
COPYRIGHT 2012 The Townsend Letter Group
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Chronic Sinusitis
Author:Hunninghake, Ronald E.; Davis, Donald R.
Publication:Townsend Letter
Article Type:Report
Geographic Code:1USA
Date:Dec 1, 2012
Previous Article:Case report: COPD and anxiety resolved by integrative medicine.
Next Article:Treating influenza with vitamin C: results and mechanisms.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters