Correlation of nasal symptoms with objective findings and surgical outcome measurement

Thesis submitted for the degree of Master of Surgery, University of London, 1993.
Published (excluding Chapter 9) 1996.
Recompiled HTML format June 2007
© 1993 – 2014 JW Fairley

Mr James W Fairley BSc MBBS FRCS MS
Consultant ENT Surgeon

Chapter 7
The relationship between symptom scores on a specially designed questionnaire and corresponding objective measurements: Postnasal drip, rhinorrhoea, nasal obstruction, cough and mucociliary clearance time

Thesis Contents

Summary

This study focuses on the difficult symptom of postnasal drip, to see if it has any organic basis.

There is evidence that postnasal drip is associated with chronic chest disease (Irwin et al, 1984).

Although the belief still persists that a post nasal drip actually causes cough and pulmonary infection, by infected mucus dripping down onto the larynx, a more likely explanation is that both are related to impaired mucociliary clearance.

If so, postnasal drip and cough should be more common in patients with abnormal mucociliary clearance. It might also be expected that rhinorrhoea and postnasal drip should occur together, especially in the presence of a mucociliary clearance abnormality.

This hypothesis was tested using the saccharin method on a group of 137 tertiary referrals to a nasal research clinic.

There was no relationship between the subjective symptom of postnasal drip and the objective measurement of mucociliary function.

Objectively impaired mucociliary function was associated with cough (p = .0005) and nasal obstruction (p = .0045).

There was a statistically significant relationship between the subjective symptom of postnasal drip and the symptom of feeling generally unwell (p = .0295).

It is possible that this may reflect a general complaining or “nasal neurotic” tendency of patients with postnasal drip, but this requires further study using psychometric techniques.

Introduction

This study focuses on the difficult symptom of postnasal drip, to see if it has any organic basis. There is evidence that postnasal drip is associated with chronic chest disease (Irwin et al, 1984). Although the belief still persists that a post nasal drip actually causes cough and pulmonary infection, by infected mucus dripping down onto the larynx, a more likely explanation is that both are related to impaired mucociliary clearance. If so, postnasal drip and cough should be more common in patients with abnormal mucociliary clearance. It might also be expected that rhinorrhoea and postnasal drip should occur together, especially in the presence of a mucociliary clearance abnormality. This hypothesis was tested on patients attending the Nasal Research Clinic with a variety of complaints.

Methods

The first 167 patients attending the Nasal Research Clinic were analyzed (See Data Appendices 1 and 6/7 for details). Mucociliary clearance time (MCCT) was measured on the first visit to the clinic using the saccharin test (Rutland and Cole, 1981; Stanley et al, 1984). The nature and purpose of the test were explained to the patient. Patients were instructed to sit quietly with the head flexed forward 10 degrees, to avoid sniffing, eating or drinking, and to avoid coughing and sneezing if possible. If a sneeze was inevitable, they were instructed not to cover the mouth with the hand and to leave the mouth wide open. A 0.5mm particle of saccharin was placed on the medial surface of the inferior turbinate, 1cm behind its anterior border, and the time taken to perceive a sweet taste was recorded in minutes. If no taste was perceived after 60 – 90 minutes, a saccharin particle was placed on the back of the tongue to check that the patient could taste sweetness. Symptom scores for postnasal drip, rhinorrhoea and cough were taken from the nasal symptom questionnaire, completed on the same occasion. For analysis, mucociliary clearance time was classified as follows:

Normal Up to 20 minutes
Prolonged 21 to 30 minutes
Severely prolonged 31 to 60 minutes
Grossly prolonged Over 60 minutes

Statistical Methods

MCCT and the subjective nasal symptom scores were examined using the SPSS-PC computer program, version 3.1. Each variable was looked at separately to determine which correlation statistics were appropriate. For contingency table analysis, Kendall’s Tau C, a statistic similar to the correlation coefficient but designed for use on ordinal measures such as the nasal symptom scores, was calculated with significance levels (Norusis, 1988a). Goodman and Kruskal’s lambda*, a measure of how well the knowledge of the independent variable predicts the value of the dependent variable, based on proportional reduction in error, was calculated with subjective nasal symptom scores as the dependent variable. A further test was done between the symptom score for feeling generally unwell, as a possible marker for neurotic traits or tendency to complain of exaggerated symptoms, and postnasal drip.

*Lambda is designed for ordinal variables. It ranges from 0 to 1. Zero means that predicting the dependent variable from knowledge of the independent variable is no better than guessing, i.e. the error is not reduced at all, while 1 means that the error is reduced by 100% i.e. prediction becomes perfect. Lambda of 0.5 means that error is reduced by 50%. Lambda allows for the fact that the dependent variable may already be easy to predict by guessing, for example if nearly all cases fall into one category.

Results

Full data was available on 137 cases, 30 had not had mucociliary clearance time measured. In four patients the saccharin test was not done because they had been referred with specific problems; two neurosurgical cases with tumours for endoscopic biopsy, one AIDS case for endoscopic examination to look for a Kaposi Sarcoma, one acute rhinitis of pregnancy. Three patients were seen before the measurement of MCCT became routine, and the remaining 23 because on some patients arriving toward the end of busier clinics this time-consuming measurement was omitted.

Mean MCCT was 28 minutes (Table 7.1) but this figure is not useful because patients who failed to taste saccharin were arbitrarily coded as 99 minutes. The stem & leaf plot (Table 7.2) shows that MCCT followed a bimodal distribution, with a large peak centred around 15 minutes, a long positive tail and a smaller peak at the positive extreme, around 60 minutes. The normal probability plot (Figure 7.1) was non-linear, showing systematic deviation from a normal distribution, and the Kolmogorov-Smirnov goodness-of-fit test confirmed this (K-S Z = 0.2048, 137 df, p = <.001). Therefore normal distribution based statistics were not applicable on the raw data, and all further statistics are based on the ordinal categories.

Table 7.1 Summary statistics on distribution of mucociliary clearance time (n = 137)

Mean     28.10  Std Err      1.99  Min     4   Skewness  1.93
Median   20.00  Variance   543.92  Max    99** S E Skew   .21
                Std Dev     23.32  Range  95   Kurtosis  3.27
                                   IQR    19   S E Kurt   .41

Table 7.2 Stem & leaf plot of
mucociliary clearance time (n = 137)

   Frequency    Stem &  Leaf

      14          0  .  &
      52          1  .  3567&&
      27          2  .  03&
      17          3  .  0&
      11          4  .  &
       3          5  .  &
       3    Extremes    (62), (67)
      10    Extremes    (68), (99*)

     Stem width:   10
     Each leaf:    11 cases
                    & denotes fractional leaves.

** 9 Patients coded as 99 minutes all failed to taste saccharin after 90 minutes. In each case, it was confirmed that sweetness could be detected by placing saccharin directly on the tongue.

Table 7.3  Crosstabulation of Mucociliary Clearance Time for 

Saccharin by Subjective PostNasal Drip (n=137)

+—————————————————————+

¦                  ¦   Subjective PostNasal Drip       ¦        ¦

¦ Mucociliary      +———————————–¦        ¦

¦ Clearance        ¦None /  ¦Mild    ¦Moderate¦Severe  ¦  Row   ¦

¦ Time      (mins) ¦Normal  ¦        ¦        ¦        ¦ Total  ¦

+——————+——–+——–+——–+——–+——–¦

¦ Normal     <=20  ¦    21  ¦    14  ¦    24  ¦    14  ¦    73  ¦

+——————+——–+——–+——–+——–+——–¦

¦ Abnormal  21-30  ¦    10  ¦     4  ¦     8  ¦     6  ¦    28  ¦

+——————+——–+——–+——–+——–+——–¦

¦ Severe    31-60  ¦     7  ¦     2  ¦     8  ¦     6  ¦    23  ¦

+——————+——–+——–+——–+——–+——–¦

¦ Gross      >60   ¦     5  ¦     3  ¦     2  ¦     3  ¦    13  ¦

+——————+——–+——–+——–+——–+——–¦

¦ Column Total     ¦    43  ¦    23  ¦    42  ¦    29  ¦   137  ¦

¦        Percentage¦  31.4  ¦  16.8  ¦  30.7  ¦  21.2  ¦ 100.0  ¦

+—————————————————————+

Lambda with PND dependent  0.0425

Kendall’s Tau C            -.00511   p = .4696

Discussion

These results show that there is no relationship between impaired mucociliary clearance, as measured by the saccharin time, and the symptoms of postnasal drip and rhinorrhoea. Even with gross mucociliary impairment, there is no trend toward increasing symptom score (Tables 7.3, 7.4). Impaired mucociliary clearance time was positively associated with cough and nasal obstruction (Tables 7.5, 7.6). Subjective postnasal drip was associated with feeling generally unwell (Table 7.7) but not with the objective measurement. It is striking that nearly all patients with a gross mucociliary defect (12 out of 13) scored either moderate or severe on nasal obstruction (Table 7.6) while there was no apparent relationship at all for postnasal drip and rhinorrhoea. Where statistically significant associations were demonstrated, as with other studies reported in this thesis (Chapters 5 & 6), the objective measurement accounts for only a small percentage of the variation, and is a poor predictor of subjective symptom severity.

There are several possible explanations for the failure to demonstrate any association between subjective postnasal drip, rhinorrhoea and the objective measurement of mucociliary function:

  1. The “objective” measurement may be a poor test of mucociliary function.
  2. The subjective symptom may be related to another pathological condition which is not associated with impaired mucociliary function.
  3. The subjective symptom may be entirely unrelated to any pathological condition, but reflect some neurotic trait. (This explanation is favoured by many practising ENT surgeons).
  4. The symptom questionnaire is invalid for postnasal drip and rhinorrhoea.
  5. There is a general objection (see introduction, Chapter 1) to the external validity of any negative crosstabulation result arising from this study on a selected sample of tertiary referrals.

To answer the first point, the saccharin test has been used extensively to investigate mucociliary clearance disorders, and is very sensitive in this respect (Stanley et al, 1984; Canciani et al, 1988; Armengot et al, 1990; Shone et al, 1990; Fisher et al, 1992). Like similar methods using radionuclide tracers (Simon, Drettner and Jung, 1977; Englender et al, 1990; Behrbohm and Sydow, 1991; Durak et al, 1991; Armengot and Basterra, 1991) it is a test of the whole mucosal surface transport system. An abnormal result may be due to problems with the mucus viscoelasticity (e.g. Young’s syndrome, cystic fibrosis) or the cilia. Cilia can be absent, present but immotile, have reduced motility, or exhibit uncoordinated motility. All of these patterns may coexist, and may be congenital or acquired. Further investigations are required to differentiate between them, including measurement of ciliary beat frequency and ultrastructural examination (Rutland and Cole, 1981; Greenstone et al, 1983). A slow saccharin test does not necessarily mean that the patient has a permanently damaged mucociliary transport mechanism, since infection itself can cause temporary disruption of mucociliary function (Wilson and Cole, 1988). There is a fairly high incidence of abnormal MCCT results in this study population (47%), so the technique clearly is working. I have also shown in this group of patients that impaired mucociliary clearance is positively associated with cough, an expected association, and with subjective nasal obstruction. The idea that the objective test is inadequate can therefore be discounted.

The second possible explanation, that postnasal drip is associated with some other pathological condition, is difficult to refute, except by asking “what condition?” The symptom of postnasal drip must be composed of at least two components, firstly the presence of mucus in the nasopharynx and secondly awareness of it. The mucus could be abnormal both in composition and amount. An abnormal sensation could be mediated by inflammatory mediators acting on mucosal receptors, or it could be due to an abnormality of the receptors themselves. I have not found any definite association between postnasal drip and any specific pathological state. It is also well known that postnasal drip is often refractory to otherwise successful surgical intervention in rhinosinusitis (Lund, 1988). Overall, the idea that we are “missing something” is possible, but seems unlikely.

The third explanation, that postnasal drip is unrelated to any pathology but due to some kind of general complaining or neurotic trait, was tested for by analyzing the relationship between the scores for postnasal drip and another subjective symptom on the questionnaire, “generally unwell”. The results show a weak but statistically significant relationship between them. The statistical significance (p = .0295) is due to the large number studied, 137 patients. The lambda figure of .08511 means that, if the patient’s score for feeling generally unwell is known, the likely score for postnasal drip can be predicted with an improvement in accuracy of 8.5% compared with a best guess, based on the knowledge of the proportions in each category of postnasal drip. This lends some support to the view that patients who complain of postnasal drip are in some way general complainers or “nasal neurotics”. It would however be preferable to investigate this further by comparing the results on postnasal drip with a formal psychological questionnaire to assess neurosis, complaining and hypochondriacal traits. It is equally possible that patients with postnasal drip really are in some way more generally unwell; there is no means of confirming or refuting this from my data.

The fourth possibility, that the symptom questionnaire is invalid for postnasal drip, must be addressed. The formal validation studies on the questionnaire (Chapter 4) focused on the internal consistency of the scale and its ability to discriminate nasal from non-nasal cases, and symptomatic from successfully treated rhinosinusitis patients. The validation did not involve detailed cross-checking of the score on every item with a structured interview by a clinician. However, I know from dealing personally with the nasal research clinic patients that those scoring highly on postnasal drip were complaining of that symptom, since I used the symptom score sheets as the basis for discussion in every case. What cannot be excluded is that there may be some patients who are troubled by post-nasal sticky catarrh, who do not score highly on the question as it is put, “Do you feel dripping at the back of the nose?”. Overall it is very unlikely that this has significantly affected the results, but if this study were to be repeated, it would be sensible to carry out some more specific validation work on the wording of this question. It may be that more than one symptom exists, one of “postnasal drip” and another of “sticky postnasal catarrh”; this could be tested for quite easily by questionnaire validation techniques (Powell, 1989).

Conclusions

In a group of 137 tertiary referrals to a nasal research clinic, there was no relationship between the subjective symptom of postnasal drip and the objective measurement of mucociliary function. Objectively impaired mucociliary function was associated with cough and nasal obstruction. There was a statistically significant relationship between the subjective symptom of postnasal drip and the symptom of feeling generally unwell. It is possible that this may reflect a general complaining or “nasal neurotic” tendency of patients with postnasal drip, but this requires further study using psychometric techniques.