Methods for measuring the volume and weight of saliva are divided into those that measure secretions from specific glands and those that measure whole or pooled saliva. Such methods can be performed under conditions of unstimulated (resting) or stimulated flow. Methods of collecting saliva from the major glands include the use of a collecting cup such as the Carlson-Crittenden cup or Lashley cup^5,6 at the orifices of the major excretory ducts, the use of custom-made collection devices or "segregators" that fit in the floor of the mouth for collection of sub-mandibular and sublingual gland secretions,⁷ and the use of micropipetting techniques.⁸

These tests are used most often in salivary research and provide accurate volumetric measurements. Furthermore, the collected sample can be tested for buffering capacity and salivary constituents. They are, however, time-consuming techniques that require special equipment and trained personnel, and they are unwieldy as screening procedures for xerostomia and hyposalivation in either a medical or dental office.

Filter paper strips of a predetermined size can be used to blot secretions from minor glands over a fixed time, with moisture levels being measured with a calibrated Periotron (Harco Electronics, Winnipeg, Manitoba, Canada), which is a device that measures small volumes of fluids.⁹ Others have used filter paper strips attached to custom molds.¹⁰

Whole or pooled saliva consists of secretions that generally collect in the floor of mouth and is composed of saliva from major and minor glands, gingival crevicular fluid, desquamated epithelial cells, bacteria, food debris and leukocytes. Assessment of whole or pooled saliva may be a more accurate indicator of overall salivary dysfunction.¹¹

Making whole or pooled saliva measurements consists of collecting saliva that passively drains from a patient’s mouth, is spat out, is suctioned out or is absorbed onto preweighed swabs over a fixed period.^12–14 These methods have been shown to produce comparable results, with draining and spitting methods being the simplest and most reproducible.¹⁵ Whole saliva measurements are easier to perform, requiring only simple collection and weighing devices, and are less time-consuming than taking individual gland measurements if compositional data are not desired.

Filter paper strips of a predetermined size can be used to blot secretions from minor glands over a fixed time.

Another study uses a semiquantitative test called the "wafer test" to screen patients who may have hyposalivation. This test measures the time it takes to dissolve a standardized 37-millimeter diameter wheat flour wafer placed on the dorsum of the tongue.¹⁶

The Schirmer test is used routinely by ophthalmologists to measure the tear film wetness.¹⁷ This test usually is done in both eyes without anesthetic and consists of placing the rounded wick end of a strip of filter paper at the junction of the middle and temporal one-third of the lower eyelid margin. The degree of wetness is read from the calibrated strip after five minutes. A reading of less than 10 mm indicates dry eyes.

Two studies have measured saliva secretions using paper strips in a manner similar to the Schirmer test. One used the Schirmer test strip placed between two tongue depressors, with a protruding 3-mm end placed against the patient’s parotid papilla for five minutes.¹⁸ The other used a paper strip measuring 1 x 17 centimeters encased in a sterilized polyethylene bag.¹⁹ One end was extracted from the bag and placed in the patient’s mouth for five minutes.

The objectives of our study were to determine whether the Schirmer test could be modified for use as a rapid, objective screening test to measure resting whole saliva wetting and whether this modified Schirmer test (MST) could distinguish between healthy adults who do not have xerostomia or hyposalivation and those who have xerostomia and hyposalivation resulting from receiving head and neck radiation, oral chronic graft-versus-host disease (CGVHD) or both.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
We studied two groups of subjects: a control group of subjects without xerostomia and a test group of subjects with severe xerostomia.

The control group. This group consisted of 41 healthy adult volunteers who did not have xerostomia and who were not taking any medications. However, we included post-menopausal women who were or were not receiving hormone replacement therapy. One examiner (Y.W.) performed all of the testing. She performed the MSTs between 8 a.m. and noon and evaluated each control subject twice, at least one week apart and no more than one month apart. In addition, she asked 14 of these subjects to spit saliva into a container over a five-minute interval, and then she weighed the container.

The test group. The examiner evaluated once in the manner described previously the 10 subjects with xerostomia who had received 50 to 70 grays of radiation for head and neck cancer and the 10 subjects with xerostomia who had oral and systemic CGVHD (four-16 months postallogenic bone marrow transplantation).

We performed the MST with the ColorBar Schirmer Tear Test (Eagle Vision, Memphis, Tenn.). The test strip is a 4-cm strip of filter paper (calibrated in 1-mm intervals from 5-35 mm along its length) that has been impregnated with a blue dye at the 0-mm wick end that has a rounded, notched tab. When the wick end contacts moisture, the blue dye travels up the strip, and its length can be read at designated intervals.

In our test, subjects sat upright in a dental chair. We asked each subject to swallow once to clear secretions in the mouth. We held the strip between cotton pliers and touched the wick end to the floor of the subject’s mouth either to the right or the left of the lingual frenum, with the subject’s tongue slightly raised and gently retracted so as not to inadvertently wet the strip (Figure 1). We read the length of wetting on the strip, as indicated by the level of blue dye on the strip, in millimeters at one-, two- and three-minute intervals (Figure 2). We removed the strip for two to three seconds to take the reading. We recorded the readings immediately, since the dye continues to travel for a few millimeters after the strip is removed. Readings ranged from 5 to 35 mm. When the reading was less than 5 mm, we recorded the measurement as 5 mm, and when the reading was greater than 35 mm, we recorded the measurement as 35 mm. This truncation of the readings at 5 mm and 35 mm may have reduced the significance of the test results but it did not invalidate the statistical inference.

View larger version (100K): [in this window] [in a new window]   Figure 1. Schirmer test strip in position.

View larger version (163K): [in this window] [in a new window]   Figure 2. Unused Schirmer test strip (left) compared with strip with dye at 25 millimeters (right).

We made comparisons between the control group and the two groups of subjects with xero-stomia and between men and women in the control group with the Wilcoxon rank sum test. All three tests are nonparametric in the sense that no distributional assumptions regarding the data need to be made. Owing to the small sample sizes, we computed exact P values for all tests using StatXact 4 for Windows (Cytel Software Corp. Cambridge, Mass.). We applied a Bonferroni correction to all test results because of the multiple tests performed.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
The nonxerostomic control group included 41 healthy adults (aged 22-55 years; male:female ratio [M:F] = 22:19). The test group included 10 subjects who had undergone radiotherapy (aged 48–74 years; M:F = 4:6) and 10 subjects who had oral CGVHD (aged 25–46 years; M:F = 8:2), all of whom had xerostomia.

The difference between the MST measurements for the first visit of the control subjects and the MST measurements for the subjects with xerostomia is striking (Figure 3 and Table 1). The mean MST reading at three minutes for the control group was 29.5 mm (standard deviation [SD] = ± 4.3 mm), whereas the mean MST reading at three minutes for the test subjects was 6.9 mm (SD = ± 2.6 mm). Table 2 includes the exact P values from the Wilcoxon rank sum test. As indicated in Table 2 and Figure 3, the differences between the control subjects and the test subjects are highly significant even after a Bonferroni correction (P < .00005 for all tests). We found no significant differences in MST values between the subjects who received radiation for head and neck cancer and the subjects who had oral CGVHD (Table 2).

View larger version (71K): [in this window] [in a new window]   Figure 3. Plot of modified Schirmer test measurements at one minute, two minutes and three minutes for the control subjects (first visit) and subjects with xerostomia. mm: Millimeters.

View this table: [in this window] [in a new window]   TABLE 2 TEST RESULTS FOR MST* COMPARISONS BETWEEN CONTROL AND TEST SUBJECTS, USING WILCOXON RANK SUM TEST.

View larger version (36K): [in this window] [in a new window]   Figure 4. Plots of paired observations of control subjects for readings at one minute. A 45-degree line has been added to indicate the extent of concordance between the paired observations. mm: Millimeters.

View larger version (33K): [in this window] [in a new window]   Figure 5. Plots of paired observations of control subjects for readings at two minutes. A 45-degree line has been added to indicate the extent of concordance between the paired observations. mm: Millimeters.

View larger version (33K): [in this window] [in a new window]   Figure 6. Plots of paired observations of control subjects for readings at three minutes. A 45-degree line has been added to indicate the extent of concordance between the paired observations. mm: Millimeters.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
These results show that the MST readily distinguished the resting whole saliva flow between healthy asymptomatic subjects and subjects who experienced profound xerostomia and hyposalivation. Nonxerostomic healthy subjects had a mean reading of approximately 30 mm at three minutes, while subjects with xerostomia had a mean reading of 7 mm or less. López-Jornet and colleagues¹⁹ used a paper strip and had a mean reading of 43.3 mm at five minutes; assuming a steady rate of wetting, this would interpolate to approximately 26.0 mm at three minutes. The wetting rate obtained using paper strips is close to the results in our study. However, both López-Jornet and colleagues’¹⁹ study and our study had faster wetting rates compared with the wetting rate from Davis and Marks’¹⁸ study; they used the Schirmer test strip for five minutes instead of three minutes as we did and found mean readings of between 18 and 26 mm depending on the age and sex of the subject. These mean readings had a mean wetting distance of between 10.6 and 15.6 mm when interpolated to a three-minute time point, which was lower than the readings obtained in this study. This slower wetting rate can be attributed to the fact that Davis and Marks¹⁸ took saliva measurements from the flow of a single parotid duct and not the whole saliva as was done in our study and López-Jornet and colleagues’¹⁹ study. The mean weight of saliva we collected from the 14 control subjects was 0.55 g/minute; this is similar to the results obtained in a study (0.47 g/minute) using the same spitting technique.¹⁵

The difference between the two visits of the control subjects was not significant, but there still was some intersubject variation (Figures 4–6). A possible explanation for this variation was the lack of standardization of subjects before we administered the tests. This was done to simulate normal clinical encounters in the dental office. First, we did not place any restrictions on food or fluid intake before the test. Second, even though we asked subjects to swallow, there is, nevertheless, variability in the amount of residual saliva left in the mouth.

Collection of saliva from individual glands is important, especially for measuring changes in pH and flow. For a quick screening evaluation of xerostomia and hyposalivation, however, traditional techniques are cumbersome and time-consuming, and they often require equipment or devices that normally would not be available to dental or medical practitioners in general practice. According to the results of our study, the MST has the following advantages. First, the entire procedure takes less than five minutes to perform. Second, the strip is inexpensive, readily available in sterilized packs and requires no other equipment, though, occasionally, cotton pliers may be used to hold the strip. Third, it has wide patient acceptability. The use of catheters and suction devices in some sialometric tests may cause local irritation of tissues that already are inflamed as a result of hyposalivation. Fourth, it is a simple test that potentially can be performed routinely in any office setting.

Further studies using larger cohorts to compare the MST with standard sialometric tests are warranted. It is unclear how sensitive the MST would be in patients who do not experience profound hyposalivation; future studies may be able to provide range values for such patients.

The MST may have application in routine screening of patients who complain of xerostomia such as in burning mouth syndrome and suspected Sjögren’s syndrome. It also may be useful for monitoring the effect of medications on mouth dryness, the response of patients with hyposalivation to sialogogues and the evolution of chronic diseases that primarily affect saliva production and flow.

	CONCLUSIONS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
In this preliminary study, we used the MST to evaluate 41 healthy volunteer control subjects, 10 subjects with severe xerostomia and hyposalivation after receiving head and neck radiation therapy and 10 subjects with severe xerostomia and hyposalivation from oral CGVHD. The healthy volunteers had significantly higher mean readings at all three time points (one, two and three minutes) using the MST than did the patients who had received radiation and the patients with oral CGVHD (all P < .00005).

The MST we used in this study is inexpensive, easy to perform and well-tolerated by patients. It readily distinguishes between patients who are healthy and asymptomatic and those who experience profound xerostomia and hyposalivation.