Intraexaminer and interexaminer reliability of mastoid fossa readings using a temporal artery thermometer

Abstract

Objective

The temperature of the mastoid fossa region has been measured by chiropractors since the 1950s using various instruments. The reliability of the procedure is largely unknown for the various instruments used for this purpose. This study assessed the reliability of a thermal instrument designed to measure forehead temperature while having an alternate landmark as the mastoid fossa.

Methods:

Thirty students were recruited to participate. Three blinded examiners twice scanned each student with the temporal artery thermometer. Intraexaminer and interexaminer reliability was assessed using the intraclass correlation coefficient. Differential margins of error were also calculated.

Results

Intraexaminer reliability was acceptable in 6 of 6 assessments, whereas interexaminer reliability was acceptable in 3 of 4 assessments. The maximum mean differential margin of error was 0.54 for intraexaminer reliability and 0.46 for interexaminer reliability.

Discussion

Combining the mean maximum differential margin of error for intraexaminer (0.54°) and interexaminer (0.46°) and then averaging these 2 ([0.54 + 0.46] / 2) results in a maximum mean differential margin of error of 0.50°.

Conclusion

In this study, intraexaminer reliability ranged from fair to strong. For these examiners, fossa differentials of 0.5 can be considered within the margin of error.

Introduction

Obtaining heat readings near the atlas transverse process (mastoid fossa) has been a procedure used by chiropractors at least since the 1950s.¹ The first instrument used for this purpose seems to be the single-probe chirometer—a skin contact instrument. Initially designed as a companion instrument for the dual-probe neurocalograph,² which is a dual-probe skin contact instrument that detects heat by conduction, the chirometer can provide a spot-temperature reading of any particular area, but is typically used in the mastoid fossa. The heat readings are recorded on all patient visits to see if a persistent pattern is present. If a pattern is present, this finding is incorporated with other findings to determine the presence or absence of the neurological component of a vertebral subluxation. This study proposes to determine the reliability of a handheld temporal artery thermometer in reading mastoid fossa heat, a component of pattern analysis. The temporal artery thermometer, if shown to be reliable and valid, could be used as a relatively inexpensive and portable device for patterning the mastoid fossa readings.

Methods

The study was approved by the institutional review board at Sherman College. Three examiners, selected via convenience sample, each scanned 30 student participants, also recruited via the convenience sample method. Each of the 3 examiners twice performed a fossa reading, first on the right fossa then on the left. The examiners blinded themselves by holding the instrument so that the reading was not visible, and a fourth person recorded the readings. Each participant was seated in a regular plastic chair commonly used in classrooms. The first scan was done immediately after the participant was seated. The instrument used in this study was the Temporal Artery Thermometer, TAT 2000C (Exergen Corp, Watertown, MA). The instrument has a 0.1° resolution (centigrade or Fahrenheit), operating range of 60.0° to 107.6°F, and a 0.04-second response time.³

Statistical analysis was performed for agreement between examiners using the intraclass correlation coefficient (ICC), 2-way mixed, single measures, in SPSS version 14.0 (SPSS Inc, Chicago, IL). The following levels were established: 0.90 to 0.99 was considered as strong reliability, 0.80 to 0.89 as good, 0.70 to 0.79 as fair, and less than 0.69 as poor. ⁴ An ICC was considered as acceptable if it has at least a fair rating (0.70-0.79). The Kolmogorov-Smirnov and Shapiro-Wilk statistics revealed that the data were not normally distributed. Therefore, the nonparametric Wilcoxon test was used to assess the data for significant differences (P < .05). For intraexaminer and interexaminer differences, the first scan (round 1) was compared with the second scan (round 2) on the same side. The readings were recorded in Fahrenheit and entered into a spreadsheet and then pasted into SPSS. A mean margin of error was calculated for intraexaminer differences by subtracting round 1 temperature readings from those of round 2 for each fossa and then averaging the differences.

A mean margin of error was also calculated for interexaminer differences by (a) averaging the combined temperature readings for all 3 examiners for each round and for each fossa, (b) subtracting each examiner's mean temperature for that round and for that fossa, and (c) averaging the mean differential from each round for each fossa. For example, for round 1 right fossa, if the mean temperature was 97.0 and examiner A's mean temperature finding was 96.8, a 0.2 difference was noted. If examiner A's mean temperature finding for round 2 right fossa was 96.4, then a difference of 0.6 was noted. Consequently, in this example, a mean interexaminer difference of 0.4 ([0.2 + 0.6] / 2) for examiner A for the round 1 right fossa would be noted. If the mean interexaminer difference for round 2 right fossa was 0.2, for example, then a subfinal mean interexaminer difference of 0.3 would be noted for examiner A for the right fossa. Hypothetically, if examiners B and C had subfinal mean interexaminer differences of 0.1 and 0.2, respectively, also for the right channel, then a final interexaminer difference of 0.2 ([0.3 + 0.1 + 0.2] / 3) would be reported for mean interexaminer margin of error for the right fossa. The same method would be used for the left fossa. In practice, only one decimal place is observed with this instrument.

Results

For intraexaminer reliability, there was one ICC score that was strong, 4 that were good, and one that was fair (ICC range, 0.760-0.937). For interexaminer reliability, there were 3 ICC scores that were considered fair (ICC range, 0.668-0.744) and one that was considered to be poor (Table 1). There were no significant differences according to the Wilcoxon test for intraexaminer and interexaminer comparisons (all analyses had a P value > .05) (Table 2).

Table 1.

Intraexaminer and interexaminer agreement using ICC for 3 examiners and 30 participants

Examiner	ICC	P
Intraexaminer ICC
Right Fossa
A1-A2	0.932	.000
B1-B2	0.869	.000
C1-C2	0.743	.000
Left Fossa
A1-A2	0.898	.000
B1-B2	0.877	.000
C1-C2	0.851	.000
Interexaminer ICC
Right Fossa
A1, B1, C1	0.748	.000
A2, B2, C2	0.671	.000
Left Fossa
A1, B1, C1	0.708	.000
A2, B2, C2	0.727	.000

Table 2.

Intraexaminer and interexaminer differences using ICC for 3 examiners and 30 participants

Examiner	P
Intraexaminer Differences
Right Fossa
A1-A2	.1
B1-B2	.9
C1-C2	.6
Left Fossa
A1-A2	.3
B1-B2	.5
C1-C2	.8
Interexaminer Differences
Right Fossa
A1-B1	.1
A2-B2	.8
A1-C1	.6
A2-C2	.7
B1-C1	.4
B2-C2	.7
Left Fossa
A1-B1	.2
A2-B2	.1
A1-C1	.6
A2-C2	.5
B1-C1	.1
B2-C2	.5

For actual temperature readings, the mean intraexaminer difference between rounds 1 and 2 was 0.29° for the right fossa and 0.25° for the left fossa (total maximum mean differential margin of error was 0.54°F; Table 3). The mean interexaminer difference was 0.23° each for the left and right fossa (total maximum mean differential margin of error of 0.46°F; Table 3). The combining of the intraexaminer and interexaminer maximum mean differential margins of error results in a final maximum mean differential margin of error of 0.50 ([0.54 + 0.46] / 2).

Table 3.

Mean differentials

Intraexaminer Differentialsa
Examiner	Fahrenheit
Right Fossa
A1-A2	0.25
B1-B2	0.25
C1-C2	0.38
Left Fossa
A1-A2	0.27
B1-B2	0.21
C1-C2	0.28
Intraexaminer Mean Differential for Right Fossa: 0.29
Intraexaminer Mean Differential for Left Fossa: 0.25
Intraexaminer Maximum Mean Differential Margin of Error: 0.54
Interexaminer Differencesb
Examiners	Mean for All 3 Examiners	Mean for All 3 Examiners Minus Individual Examiner Mean
Right Fossa, First Round
A, B, and C	97.86
A		0.24
B		0.19
C		0.21
Mean Differential		0.21
Right Fossa, Second Round
A, B, and C	97.74
A		0.25
B		0.24
C		0.26
Mean Differential		0.25
Left Fossa, Round 1
A, B, and C	98.00
A		0.24
B		0.21
C		0.29
Mean Differential		0.25
Left Fossa, Round 2
A, B, and C	97.87
A		0.24
B		0.21
C		0.19
Mean Differential		0.21
Interexaminer Mean Differential for Right Fossa: 0.23
Interexaminer Mean Differential for Left Fossa: 0.23
Interexaminer Maximum Mean Differential Margin of Error: 0.46

^aRound 1 minus round 2.

^bMean of all 3 examiners. Each examiner's mean was then subtracted from the mean of all 3 examiners.

Discussion

Although reliability may be considered as acceptable in this study, the margin of error should be an important consideration. Thus, when a patient displays a differential at the margin of error and below (≤0.50), the differential may actually be nonexistent or balanced. Consequently, differentials ≥0.51 should be considered as significant. In pattern analysis, further significance would exist only if the warm side persisted (≥0.51°) from visit to visit. Further studies are indicated with other examiners to see if the margin of error may be less than the 0.50 found in the present study.

Conclusion

For these 3 examiners, a fossa differential of 0.5 is considered within the mean margin of error. Further research is needed to confirm this finding.