Reliability of the modified-modified Schober and double inclinometer methods for measuring lumbar flexion and extension.

Low back pain (LBP) is one of the most common medical conditions in the Western world, afflicting approximately 85% of all persons at some time in their lives.[1,2] As LBP has been shown to alter lumbar range of motion ROM),[3-5] there is a need to identify reliable methods for assessing lumbar ROM in patients. Lumbar ROM is often used to evaluate spinal function, select appropriate treatments, and monitor the patient's progress.[6-8] it is difficult to accurately measure spinal movement because bony landmarks are often difficult to palpate due to excess soft tissue, normal curves of the spine vary from individual to individual, and the presence of hip motion may confound the spinal movement measurement.[9]

One method of measuring spinal ROM is with the use of a tape measure.[10] This technique, known as the skin distraction method, was originally described by Schober.[11] This technique involves using a tape measure held directly over the spine between points 10 cm above the lumbosacral junction with the patient in the neutral standing position. When the patient moves into full lumbar flexion, the increase in distance between the marks gives an estimate of spinal flexion ROM.

Macrae and Wright[12] modified the original Schober method by marking a point 5 cm below and 10 cm superior to the lumbosacral junction. The rationale for this modification was an observation that on forward flexion, both the lumbosacral and 10-cm superior skin marks tended to move superiorly relative to the spinous processes and the skin was more firmly tethered at the point 5 cm lower on the sacrum. In addition, they suggested that the effect of inaccurate identification of the lumbosacral junction was minimized by adding the landmark 5 cm below the lumbosacral function. These investigators attempted to validate this method by comparing the lumbar flexion measurements of the Schober and modified Schober techniques with radiographic lumbar flexion measurements obtained from 11 subjects with and without spinal disease. One radiograph was taken while the subject was in the standing position, and the second radiograph was taken with the subject's lumbar spine flexed. Pearson Product-Moment Correlation Coefficients of .97 and .90 were found for the modified Schober and the original Schober techniques, respectively. Reliability was not addressed. Moll and Wright[10] suggested that the modified Schober method might also be useful for assessing lumbar extension by measuring the attraction of the skin marks as they approach each other during backward bending.

Several authors[13-15] have reported the reliability of lumbar measurements obtained by the modified Schober method. Reynolds[13] examined the merits of measuring back movements using the spondylometer, the goniometer, and the skin distraction methods on 30 subjects (10 subjects with arthritic disorders and 20 volunteers who were either impatients from a general medical ward or medical students). Interobserver error was calculated by comparing the results of two observers on 10 subjects. The skin distraction results (Pearson correlation coefficients) were .59 for lumbar flexion and .75 for extension. Intraobserver error of the skin distraction method was determined by calculating coefficients of variation (CVs) for each measurement performed on the same subject on 10 separate occasions. Values of 11.65% and 21.57% were obtained for lumbar flexion and extension, respectively. The CV is a measure that is often used to describe the amount of variation in a sample.16 It reflects all variation, regardless of the source, in a series of measurements, not just measurement error. Reliability is the consistency or repeatability of measurements, that is, the degree to which measurements are error-free and the degree to which repeated measurements will agree.17 Therefore, reliability is not characterized by the CV.

When measuring the lumbar ROM of 17 physical therapy students with the modified Schober method, Fitzgerald et al[14] reported an interobserver reliability (Pearson correlation coefficients) of 1.00 for lumbar flexion and .88 for lumbar extension. As these subjects were young and healthy, generalizations of these findings to older subjects with lumbar dysfunction may be difficult. Beattie et al[15] reported high reliability (intraclass correlation coefficient [ICC]=.95) on 200 subjects using the modified Schober method of measuring lumbar extension. Intrarater reliability was assessed on two subgroups. One group consisted of 100 subjects with significant LBP, and the other group consisted of 100 subjects with no significant LBP. The ICCs for the subjects with significant LBP and for the subjects without significant LBP were .93 and .90, respectively. Interrater reliability (ICC=.94) was reported for 11 subjects without significant LBP. As the interrater reliability assessment did not include subjects with significant LBP, these results may not be generalizable to other patient groups.

Van Adrichem and van der Korst[18] evaluated lumbar flexion measurements using a tape measure and the landmarks of the posterior superior iliac spines (PSISs) and marks 5, 10, 15, and 20 cm above the PSISs in five men, 20 to 25 years of age, without LBP. With the subject standing, the position of the spinous process of the fifth lumbar vertebra was assessed by the intersection of the line connecting the PSISs with the midline of the back. The point of intersection was marked on the skin. Four additional marks were placed on the dorsal midline of the back 5, 10, 15, and 20 cm above the first mark. With a steel tape measure pressed against the skin, the subjects bent forward as far as possible and the distances between the first (lowest) mark and each of the four superior marks were measured. These measurements were taken seven times at 1-week intervals. Further investigations on 248 healthy subjects, 6 to 18 years of age, compared the increase in distance during lumbar flexion at the 5-cm intervals superior to the intersection of the line connecting the PSISs with the midline of the back. The results indicated that the more superior the landmarks, the smaller the increase in length during lumbar flexion. In the lowest 5-cm interval, the mean increase in length was less than 3 cm, whereas the increase in the upper segment (between 15 and 20 cm superior to the intersection of the line connecting the PSISs with the midline of the back) was about 1 cm. The investigators concluded that including the segment between 15 and 20 cm superior to the midline of the PSISs would contribute little to an overall measure of lumbar flexion. They suggested that because the length of the lumbar spine was approximately 15 cm, a 15-cm mark superior to the midline of the PSISs was an appropriate landmark for measuring lumbar flexion ROM. Thus, the PSISs, as opposed to the lumbosacral junction as advocated by Macrae and Wright,[12] were justified as being viable, identifiable landmarks for the starting point in measuring lumbar ROM. Contrary to van Adrichem and van der Korst's[18] contention, however, the PSISs are landmarks for the spinous process of the second sacral vertebra rather than the lumbosacral junction.[19,20]

An inclinometer, another instrument that is used to measure spinal motion, is a hand-held, circular, fluid-filled disk with a weighted gravity pendulum indicator that remains oriented in the vertical direction.[10,21,22] Two inclinometers are used to measure lumbar ROM. With the patient standing, one inclinometer is placed on the sacrum to evaluate hip motion while the other is placed on the first lumbar vertebra to measure hip and lumbar ROM. The patient is asked to bend forward maximally while the readings of the two inclinometers are recorded. Lumbar ROM can then be estimated as the difference in the two measurements.

Several authors[23-26] have used the double inclinometer (DI) technique for measuring lumbar ROM. in a sample of 51 subjects with and without chronic spinal dysfunction, Mayer et al23 evaluated the DI method for measuring lumbar ROM. Flexion and extension radiographs were taken on a subsample of 12 subjects with chronic LBP in the standing position and were compared with inclinometer measurements. With each subject in forward flexion, inclinometer measurements were determined and, with the subject holding the position, the flexion radiograph was taken. Similarly, the subject extended fully and, while holding the position from which the inclinometer measurements were taken, an extension radiograph was taken. Subtraction of inclinometer measurements derived from the full flexion radiograph and from the radiograph taken at the neutral position resulted in a radiologic estimate of lumbar motion. The investigators concluded that there was no difference between the two measurement techniques and that the DI technique was useful for assessing the spinal ROM of patients with LBP. Unfortunately, these investigators used means and standard deviations, and not correlations, to compare the two techniques. Reliability was not addressed.

Using the DI technique, Keeley et al[24] reported interrater reliability values (Pearson correlation coefficients) of .92 for 9 subjects with chronic LBP and .90 for 11 subjects without LBP. Merritt et al[25] studied the lumbar flexibility of 25 healthy subjects by examining lumbar flexion measurements using the fingertip-to-floor and the modified Schober methods and lumbar flexion and extension measurements using the single inclinometer (placed 15 cm above the sacrum) technique. These authors concluded that the inclinometer extension test showed poor reproducibility (CV=65.4% for interexaminer reproducibility and 50.7% for intraexaminer reproducibility) and that the modified Schober demonstrated good reproducibility (CV=6.3% for interexaminer reproducibility and 6.6% for intraexaminer reproducibility). As noted previously, the CV reflects all sources of variation, not just measurement error. In addition, these results may not reflect reproducibility, because the measurements were taken three times within 1 week and the assessors may have been influenced by memories of their previous measurements.

Gill et al[26] compared the repeatability of fingertip-to-floor, modified Schober, DI, and photometric techniques in 10 healthy subjects. They concluded that the modified Schober method was most repeatable (CV=0.9% for flexion and 2.8% for extension), followed by the fingertip-to-floor method (CV=14.1%) and the DI method (CV=33.9% and 3.6% for upper inclinometer flexion and extension, respectively, and CV=9.3% and 4.7% for lower inclinometer flexion and extension, respectively).

Reports on the reliability of lumbar flexion and extension ROM measurements obtained with the tape measure and the inclinometer techniques are variable, Thus, we decided to conduct a study to determine the reliability of measurements obtained with the most frequently used methods of clinical measurement. We chose to study the modified-modified Schober (MMS) technique, which was described by van Adrichem and van der Korst,[18] rather than the modified Schober method. The MMS method uses two landmarks: a line intersecting the line connecting the PSISs with the midline of the back and a mark drawn 15 cm superiorly.

The MMS technique was selected for several reasons. First, the landmarks in the modified Schober technique[12] require the identification of the lumbosacral junction and a mark 5 cm below this landmark to include lumbosacral motion and to minimize skin movement. Based on our clinical experience, however, a distance 5 cm below the lumbosacral junction is difficult to accurately identify as it lies in the upper part of the natal cleft on most individuals. The use of the PSISs as the inferior landmark in the MMS technique has several advantages. A mark placed midway between the PSISs is at the second sacral level.[19,20] This mark is on the sacrum, which is an inflexible bone, and no motion would be expected to be gained by using an additional mark below this level. This landmark is also easily identifiable and eliminated the need for an additional landmark 5 cm below this landmark. Second, the superior landmark for the MMS technique was identified as a measured distance (15 cm) above the inferior landmark of the line intersecting the line connecting the PSISs, rather than the first lumbar vertebra. This method was proposed to minimize error in identifying the first lumbar vertebra. Van Adrichem and van der Korst[18] suggested that a 15-cm distance superior to the line intersecting the line connecting the PSISs was an accurate representation of the actual length of the lumbar spine. The final reason for selection of the MMS technique was to provide common landmarks for the MMS and DI methods. The MMS technique would be more comparable to the DI method because of the placement of the upper inclinometer on the first lumbar vertebra, which should correlate more closely with the mark 15 cm above the lumbosacral junction.

The main purpose of this study was to calculate the test-retest and interrater reliability of lumbar flexion and extension measurements using the MMS and DI methods on patients with LBP. A secondary purpose was to determine the time taken to perform these measurements.

Method

Subjects

Sample size calculations indicated that with intrarater and interrater ICCs that exceeded 0.7, an alpha level of 5%, a power of 80%, and an effect size of 0.5, 15 subjects were required.27 The subjects in this study were volunteers who were patients in the Physical Therapy Department, Hamilton Civic Hospitals, Henderson General Division, Hamilton, Ontario, Canada. Inclusion criteria for the study required that the subjects (1) be diagnosed as having LBP, with or without leg pain, (low back pain was defined as pain on the posterior aspect of the trunk below the basilar costal margins or above the greater trochanter[15]); (2) be 18 years or older; (3) be able to stand for at least 15 minutes; and (4) be able to tolerate repeated forward and backward bending. Patients were excluded from the study if they had known nonmechanical back pain (eg, spinal neoplasm, infection, Paget's disease, inflammatory back pain), neurological impairment (reflex changes or lower-extremity muscle weakness) attributable to lumbar pathology, pain at less than 30 degrees during straight leg raising, or pain of visceral origin or were pregnant. Informed consent was obtained from all patients before admission to the study. The patients (8 women, 7 men) ranged in age from 25 to 53 years ([bar] X=35.7, SD=9.9) and were described as having chronic LBP, with a mean duration of back pain of 17.6 months (SD=18.0) (Tab. 1).

Table 1. Characteristics of the Sample(a)
Variable                   [bar] X    SD     Range
Age (y)                      35.7    9.9    25.0-53.0
Height (cm)                 171.3    9.8   155.0-188.0
Weight (kg)                  72.5   16.0    55.3-106.0
Duration of back pain (mo)   17.6   18.0     2.0-48.0
(a) Sample consisted of 15 patients (7 men, 8 women) with chronic low back pain.

Three physical therapists on the staff of the Physical Therapy Department, Hamilton Civic Hospitals, Henderson General Division, participated in the study. The therapists' clinical area of expertise was orthopedics, and their clinical experience varied from 3 to 12 years (X=8.3, SD=4.7). Prior to the start of the study, the therapists were given a written description of the MMS and DI techniques. Standardization of the protocols for both methods was established with the investigators and the therapists in a group session. The results obtained in this study, therefore, will only be generalizable to those settings in which all examiners undergo similar training in the use of a standardized protocol. Two practice sessions using the standardized protocols on peers and patients were also included. The therapists were encouraged to use the techniques as much as possible between the practice sessions.

Instrumentation

Three tape measures (Fig. 1) marked in 1-mm increments were used to obtain measurements with the MMS method. Six MIE Medical Research Ltd inclinometers(*) (Fig. 2) marked in 0.1-degree increments were used to obtain measurements with the DI technique.

To ensure that the length of the tape measures did not change with repeated use, the accuracy of the tape measures was checked by comparing the centimeter scale on the tape with a wooden meter stick. The inclinometers were also checked by a therapist who was not involved in the study by using them on patients with LBP to ensure that they were in working order. The accuracy of the instruments was checked 1/2 hour prior to the start of the study.

Procedure

Lumbar measurements were taken by each of the three physical therapists using both the MMS and the DI methods on two occasions, 2 days apart. Instructions to the therapists are outlined in Appendixes 1 through 4. The measurement sequence, that is, the order of patients, movements (flexion and extension), and techniques (MMS and DI), was randomly determined using a table of random numbers. Random allocation for each session was determined a priori. Two days later, the same therapists measured the same patients using the same procedures (with the sequence changed randomly).

After the subjects completed the consent form, they entered an examination room, removed their shoes, and disrobed, exposing their backs from the gluteal fold to the midthoracic spine. All subjects were instructed to stand erect with their eyes directed horizontally, their arms at their sides, and their feet placed on a set of paper footprints that were secured to the floor (the heels of the footprints were about 15 cm apart).

The Modified-modified Schober Flexion Technique

The examiner knelt behind the standing subject and identified the PSISs by marking the inferior margins of the subject's PSISs with his or her thumbs. An ink mark was drawn along the midline of the lumbar spines horizontal to the PSISs. Another ink mark was made 15 cm above the original mark (Fig. 3). The tape measure was then lined up between the skin markings. With the tape measure pressed firmly against the subject's skin and while holding the tape measure with his or her fingertips, the therapist instructed the subject to bend forward. When the subject bent forward into full lumbar flexion (Fig. 4), the new distance between the superior and inferior skin markings was measured. The distance between these marks after trunk flexion was measured, and the change in the difference between the marks was used to indicate the amount of lumbar flexion. Each therapist recorded measurements to the nearest 1 mm, the difference in the initial length between skin markings (15 cm) and the length measured in forward lumbar flexion. After each measurement, all skin marks were removed with rubbing alcohol.

The Modified-Modified Schober Extension Technique

The same landmarks and procedures described for the MMS flexion technique were used for measuring lumbar extension. With the subject in the erect standing position, the therapist lined up the tape measure between the markings, While holding the tape measure, placed firmly against the subject's skin, the therapist instructed the subject to place the palms of the hands on the buttocks and to bend backward. When the subject bent backward into full lumbar extension (Fig. 5), the new distance between the superior and inferior skin markings was measured. The distance between these marks after trunk extension was measured using the tape measure, and the change in the difference between the marks was used to indicate the amount of lumbar extension. Each therapist recorded measurements to the nearest 1 mm, the difference in the initial length between skin markings (15 cm) and the length measured in forward lumbar extension. All skin marks were removed with rubbing alcohol.

The Double Inclinometer Flexion Technique

Two inclinometers were used to measure lumbar flexion and extension. The examiner knelt behind the standing subject. The skin was marked along the midline of the spine horizontal to the PSIS, and one mark was made on the spinous processes, 15 cm superior to the PSIS line. Both inclinometers were placed over the skin marks. The inclinometers were set as close to 0 degrees as possible. While holding the inclinometers, the therapist instructed the subject to bend forward (Fig. 6). When the subject bent forward into full lumbar flexion, the angles on each inclinometer were recorded to the nearest degree. The angle measured on the upper inclinometer indicated gross flexion motion of the lumbar spine and hips. The angle measured on the lower inclinometer indicated hip flexion alone. Lumbar flexion ROM was determined by the subtraction of the lower inclinometer reading from the upper inclinometer reading. The skin marks were removed with rubbing alcohol.

The Double Inclinometer Extension Technique

The same landmarks and procedures that were described for the DI flexion technique were used for measuring lumbar extension. Both inclinometers were placed over the skin marks. While holding the inclinometers, the therapist instructed the patient to bend backward (Fig. 7) and recorded the angles on the inclinometers. Skin marks were removed with rubbing alcohol.

Time

The time to perform each technique was calculated by taking a random sample of the time in seconds for 10 different measurements (lumbar flexion and extension) performed by each therapist for each technique (MMS and DI). A randomization scheme using a table of random numbers was determined a priori.

Data Analysis

Pearson Product-Moment Correlation Coefficients and analysis of variance (ANOVA)-delived ICCS (3,1)[28] were used to calculate the test-retest and interrater reliability, respectively, of the therapists' measurements. To check the assumption of equality of variances, Hardey's test[29] was used. This test is useful for detecting whenever one variance is smaller or larger than the others.

Richman et al[30] have suggested the following reliability coefficient values: .80 to 1.00 is very reliable, .60 to .79 is moderately reliable, and .59 and below is of questionable reliability. These values were used to characterize the quality of the reliability coefficient results of the study. A two-way repeated-measures ANOVA was used to compare the time that it took for the therapists to perform each measurement technique.

Results

Lumbar flexion and extension measurements obtained by the therapists are shown in Table 2. A table of the percentage of total variance explained by various factors based on variance components[6] is presented in Table 3. Table 4 shows the results of the Pearson correlation coefficients for the test-retest reliability of the therapists' measurements. For therapist 1, using the MMS technique, Pearson correlation coefficients were .89 for flexion and .79 for extension. Using the DI technique, therapist 1's Pearson correlation coefficients were .87 for flexion and .28 for extension. The Pearson correlation coefficients for therapist 2 for the MMS technique were .78 for flexion and .91 for extension. For the DI technique, therapist 2's Pearson correlation coefficients were .76 for flexion and .66 for extension. For therapist 3, the Pearson correlation coefficients for the MMS technique were .83 for flexion and .69 for extension. Therapist 3's Pearson correlation coefficients for the DI method were .13 for flexion and .55 for extension. [TABULAR DATA 2 OMITTED]

Table 3. Percentage of Total Variance Explained by Various Factors Based on
Variance Components[16]
              Percentage of Variance Explained
Method(a)     Subjects   Therapists              Residual
MMS
  Flexion       62.0         0.9                    37.2
  Extension     77.9         1.3                    20.2
DI
  Flexion       61.4         0.0                    38.6
  Extension     51.1         3.0                    45.1
(a) MMS=modified-modified Schober method, DI=double inclinometer method.
Table 4. Pearson Product-Moment Correlation Coefficient Estimates of the
Test-Retest Reliability of the Therapists' Measurements
            Therapist
Method(a)      1     2     3
MMS
  Flexion     .89   .78   .83
  Extension   .79   .91   .69
DI
  Flexion     .87   .76   .13
  Extension   .28   .66   .55
(a)MMS=modified-modified Schober method, DI=double inclinometer method.

The ANOVA ICCs for the interrater reliability of the therapists' measurements are presented in Table 5. in all cases, the values for the reliability of the measurements were statistically significant (P<.001) and the subjects accounted for most of the variation observed in the data. The ICCs for the therapists using the MMS technique were .72 (flexion) and .76 (extension); with the DI method, the ICCs were .60 (flexion) and.48 (extension).

Table 5. Analysis of
Variance-Derived Intraclass Correlation
Coefficients[28] (ICCs) for Interrater
Reliability of the Therapists' Measurements
Method(a)                                  ICC
MMS
  Flexion                                  .72
  Extension                                .76
DI
  Flexion                                  .60
  Extension                                .48
(a) MMS=modified-modified Schober method,
DI=double inclinometer method.

Using Hartley's test,[29] the values for the MMS lumbar flexion measurements showed that the variances for the subjects were not significantly different from each other; however, the DI lumbar flexion variances were significantly different (Tab. 6). The values for the lumbar extension measurements with the MMS and the DI methods demonstrated that the variances for the subjects were not significantly different from each other.

Table 6. Hartley's Test for Homogeneity of Variances Among 15 Subjects29 (Six
Scores Contributed to the Mean of Each Subject)
                      SD
Method(a)     Maximum     Minimum     Fm.(14,5)       P
MMS
Flexion        1.00        0.34           8.75       >.05
Extension      0.74        0.22          11.58       >.05
DI
Flexion       13.34        1.76          57.42       <.05
Extension      7.74        1.94          15.89       >.05
(a) MMS=modified-modified Schober method, DI=double inclinometer method.

The time taken to perform each of the measurement methods is shown in Table 7. These results showed that the mean time to obtain lumbar flexion and extension measurements for the MMS and the DI methods were 10.2 seconds and 23.1 seconds, respectively. The ANOVA revealed that the time taken to obtain the measurements with the MMS technique was significantly less (F = 140.22; df = 1, 100; P<.001) than for the DI method. There was no difference between the time that it took to measure lumbar flexion versus the time that it took to measure lumbar extension (F=1.76; df= 1, 100; P =. 187). Thus, the MMS procedure, as described in this study, is a faster method of measuring lumbar flexion and extension ROM compared with the DI method.

Table 7. Time (in Seconds) Taken to Perform the Measurements
Method(a)       n      [bar]X      SD       Range
MMS
  Flexion       30        9.3      3.0      4-9.3
  Extension     30       11.1      3.9      5-11.1
DI
  Flexion       30       25.0     10.5      10-25.0
  Extension     30       21.1      6.0      11-21.1
(a) MMS=modified-modified Schober method, DI=double inclinometer method.

Discussion

Reliable methods of measuring lumbar ROM are important in assessing spinal function in patients with LBP.[7,10,13,23,31,32] According to the reliability coefficient values discussed by Richman et al,30 the results of this study show that the MMS method of measuring lumbar flexion and extension ROM has moderate reliability (.72 and .76, respectively) and that the DI technique has questionable reliability (.60 and .48, respectively). Although the time that was calculated to perform both techniques was less than 30 seconds, this study demonstrates that significantly less time was required to obtain lumbar flexion and extension measurements with the MMS technique. Thus, the MMS technique appears to be a reliable and time-efficient method of measuring lumbar flexion and extension in comparison with the DI technique.

It may be difficult to compare our results for the MMS technique with those of van Adrichem and van der Korst,[18] who reported using the MMS method on healthy male subjects aged 20 to 25 years. Our study consisted of male and female subjects with LBP, 25 to 53 years of age. For the DI technique, our results are dissimilar to those of Keeley et al,[24] who reported high reliability values on subjects with and without LBP. This discrepancy is perhaps due to the fact that in the Keeley et al study, there may have been an order effect or a sequencing effect,[33] as all patients were measured by the therapists in the same order. Order effects have implications for measurement results, as the ROM measurements obtained may be the result of the previous movement interacting with the movement being measured.[33] In our study, the subjects, techniques, and movements were randomly allocated to the therapists. Another difference may be that in the Keeley et al study, the inclinometers were placed on the first lumbar spinous process and the sacrum. It is difficult to palpate the first lumbar vertebra when the patient is in the neutral standing position.

A potential drawback of using the measured distance of 15 cm, as described for the MMS technique, is that height and body dimension variations may affect the placement of a measured mark relative to the location of the first lumbar vertebrae. This is also applicable for the measured distance of 10 cm with the modified Schober technique. Van Adrichem and van der Korst[18] showed that the upper lumbar levels contributed least to the overall lumbar flexion ROM measurement and that the segment from 15 to 20 cm superior to the PSISs contributed minimally. We believe, therefore, that height variations, which could change the actual upper lumbar segments included in the 15-cm distance, would have a minimal effect on the overall measurement. The advantages of using a measured distance for the superior landmark are minimization of error in identification of the level of the first lumbar vertebrae and speed. We believe these advantages outweigh the potential disadvantages. According to Schober's original article,[11] his method was a rough estimate based on the increasing distance between the examiner's thumb and index finger, placed on two random points of the patient's spine, during lumbar flexion. Macrae and Wright[12] developed a standardized version of this procedure. The only difference is that Macrae and Wright measured the distance from 5 cm below the line connecting the PSISs to 10 cm above it, whereas in our study, we proposed to measure the distance between the midpoint of the line connecting the PSISs and a point 15 cm above it. The MMS method does not include an immobile segment of 5 cm, as in the modified Schober technique.

In our study, the fact that the variances for the DI flexion measurements were detectably different makes the usefulness of the DI technique questionable. The therapists in our study stated that it was difficult to secure the inclinometers on the subjects as the subjects bent forward and backward. One therapist complained that her knees were sore because she had to bear weight on them to establish the landmarks on the subjects. One suggestion would be to have the therapist seated on a movable stool. This position would allow the therapist to be closer to the patient's lumbar spine and would enable the therapist to secure the instruments to the patient while the patient moves through the ROMs. The therapists seemed to prefer using the tape measures. They stated that the tape measures were less cumbersome to use, easier to secure to the subject, and easier to read compared with the inclinometers. Tape measures are also less expensive and more readily available. Merritt and associates[25] concluded that the inclinometer method shows promise but requires more time, training, and calculations to master the technique. An advantage of the DI technique is that its results are reported in degrees of flexion and extension. This may make assessing ROM function easier because this assessment is based on degrees of movement. The values obtained with both techniques may be used to measure progress.

The concepts of reliability and validity are inextricably related. If we do not have construct validity, it is irrelevant whether our instruments are reliable.[34] Even if our constructs are valid, we must use measures that yield reliable measurements if we wish to draw any conclusions about the effects of our treatments.[34]

This study was the first step in determining the usefulness of the MMS and DI methods. The validity of the MMS and the DI methods has not been studied. The usefulness of the MMS and the DI measures, however, depends on their construct validity and, ultimately, the criterion-related validity of measurements obtained with these methods. The construct validity or theoretical basis for the use of the MMS method seems reasonable because the distance measured is a measurement of flexion and extension of the lumbar spine. The criterion-related validity of MMS could be assessed by comparing lumbar ROM measurements with spinal ROM measurements obtained radiographically. The MMS method, as described in this study, seems to have moderate reliability[30] and therefore may be appropriate for any future study addressing the criterion-related validity of the MMS measurements. The results obtained in our study may be generalized to those settings in which the examiners undergo similar training in the use of a standardized protocol.

Conclusions

The MMS method, as described in this study, appears to produce reliable measurements of spinal flexion and extension in patients with LBP, whereas the DI technique needs improvement. The therapists in this study seemed to prefer the MMS method over the DI technique, as they stated that the MMS measurements are easier and quicker to obtain. Our results also indicate that less time is required to obtain lumbar flexion and extension measurements with the MMS method. Although the DI technique shows promise, it may require more time and training to master.

The results of this study, however, do not provide evidence as to whether the MMS is valid in assessing a patient's ability to flex or extend the lumbar spine. Further study is needed to determine the criterion-related validity of the MMS measurements and to predict the change in the amount of lumbar flexion and extension. Appendix 1. Modified-Modified Schober Technique for Measuring Lumbar Flexion 1. A set of paper footprints is secured to the floor,

with the heels of the footprints about 15 cm apart. 2. Patient position: The patient stands erect, eyes focusing

horizontally, arms at side, with feet placed on the paper footprints. 3. Skin markings: The therapist stands behind the patient and identifies

the posterior superior iliac spines (PSISs) by marking the PSISS with

his or her thumbs. Make an ink mark on the midline of the lumbar

spines horizontal to the PSIS. Make another mark on the spinous

processes 15 cm superior to the PSIS line (to the nearest millimeter).

The distance between the superior and inferior skin marks on the

spinous processes is 15 cm. 4. Align the tape measure between the two skin marks, with zero at the

inferior skin mark and 15 cm at the superior skin mark. Keep the tape

measure firmly against the patient's skin while the patient bends forward. 5. Instructions to the patient: "Bend forward as far as you can while keeping your knees straight." 6. Measurement procedure: When the patient has bent forward, the

new distance between the superior and inferior skin markings is

measured (to the nearest millimeter) with the patient positioned in full lumbar flexion. 7. Instructions to the patient: "You can come back to a comfortable standing position." 8. On the sheet provided, record the measurement. Flexion range of motion

is the difference between the initial length between skin markings

(15 cm) and the length measured in full forward flexion. 9. Remove all skin marks with rubbing alcohol. What to do if the patient cannot attain or maintain position: Allow 1 minute of rest in the patient's most comfortable position. Try again. If that fails, abort the test. Record "N/A" and indicate the reason for failure. Indications to discontinue test: Failure on second attempt. Position in which to leave the patient: In his or her most comfortable position. Appendix 2. Modified-Modified Schober Technique for Measuring Lumbar Extension 1. A set of paper footprints is secured to the floor,

with the heels of the footprints about 15 cm apart. 2. Patient position: The patient stands erect, eyes focusing horizontally,

arms at side, with feet placed on the paper footprints. 3. Skin markings: The therapist stands behind the patient and identifies

the posterior superior iliac spines PSISS) by marking the PSISS with

his or her thumbs. Make an ink mark on the midline of the lumbar

spines horizontal to the PSIS. Make another mark on the spinous

processes 15 cm superior to the PSIS line (to the nearest millimeter).

The distance between the superior and inferior skin marks on the

spinous processes is 15 cm. 4. Align the tape measure between the two skin marks, with zero at the

inferior skin mark and 15 cm at the superior skin mark. Keep the tape

measure firmly against the patient's skin while the patient bends backward. 5. Instructions to the patient: "Place the palms of your hands

on your buttocks and bend backward as far as you can." 6. Measurement procedure: When the patient has bent backward, the

new distance between the superior and inferior skin markings is

measured (to the nearest millimeter) with the patient positioned in full lumbar extension. 7. Instructions to the patient: "You can come back to a comfortable standing position." 8. On the sheet provided, record the measurement. Extension range of

motion is the difference between the initial length between skin

markings (15 cm) and the length measured in full lumbar extension. 9. Remove all skin marks with rubbing alcohol. What to do if the patient cannot attain or maintain position: Allow 1 minute of rest in the patient's most comfortable position. Try again. If that fails, abort the test. Record "N/A" and indicate the reason for failure. Indications to discontinue test: Failure on second attempt. Position in which to leave the patient: In his or her most comfortable position. Appendix 3. Double Inclinometer Technique for Measuring Lumbar Flexion 1. A set of paper footprints is secured to the floor,

with the heels of the footprints about 15 cm apart. 2. Patient position: The patient stands erect, eyes focusing horizontally,

arms at side, with feet placed on the paper footprints. 3. Skin markings: The therapist stands behind the patient and identifies

the posterior superior iliac spines (PSISs) by marking the PSISs with

his or her thumbs. Make an ink mark on the midline of the lumbar

spines horizontal to the PSIS. Make another mark on the spinous

processes 15 cm superior to the PSIS line (to the nearest millimeter).

The distance between the superior and inferior skin marks on the

spinous processes is 15 cm. 4. Placement of inclinometers: With one inclinometer in each hand, the

therapist places the inclinometers on the skin marks, ensuring that the

indicators on the inclinometers are set at zero. Hold the

inclinometers in place. 5. Instructions to the patient: "Bend forward as far as you can." 6. Measurement procedure: When the patient has bent forward, the

angles on each inclinometer are read "to the nearest degree). 7. Instructions to the patient: "You can come back to a comfortable

standing position." 8. On the sheet provided, record the measurement. The angle measured on

the superiorly placed inclinometer indicates gross flexion motion

of the lumbar spine and hips. The angle measured on the inferiorly placed

inclinometer indicates hip flexion alone. Lumbar flexion motion is

determined by the subtraction of the hip flexion angle from the total flexion angle. 9. Remove all skin marks with rubbing alcohol. What to do if the patient cannot attain or maintain position: Allow 1 minute of rest in the patient's most comfortable position. Try again. If that fails, abort the test. Record "N/A" and indicate the reason for failure. Indications to discontinue test: Failure on second attempt. Position in which to leave the patient: In his or her most comfortable position. Appendix 4. Double Inclinometer Technique for Measuring Lumbar Extension 1. A set of paper footprints is secured to the floor,

with the heels of the footprints about 15 cm apart. 2. Patient position: The patient stands erect, eyes focusing horizontally,

arms at side, with feet placed on the paper footprints. 3. Skin markings: The therapist stands behind the patient and identifies

the posterior superior iliac spines (PSISs) by marking the PSISs with

his or her thumbs. Make an ink mark on the midline of the lumbar

spines horizontal to the PSIS. Make another mark on the spinous

processes 15 cm superior to the PSIS line (to the nearest millimeter).

The distance between the superior and inferior skin marks on the

spinous processes is 15 cm. 4. Placement of inclinometers: With one inclinometer in each hand, the

therapist places the inclinometers on the skin marks, ensuring that the

indicators on the inclinometers are set at zero. Hold the inclinometers in place. 5. Instructions to the patient: "Place the palms of your hands

on your buttocks and bend backward as far as you can." 6. Measurement procedure: When the patient has bent backward, the

angles on each inclinometer are read (to the nearest degree). 7. Instructions to the patient: "You can come back to a comfortable

standing position." 8. On the sheet provided, record the measurement. The angle measured

on the superiorly placed inclinometer indicates gross extension of

the lumbar spine and hips. The angle measured on the inferiorly placed

inclinometer indicates hip extension alone. Lumbar extension

motion is determined by the subtraction of the hip extension angle

from the total extension angle. 9. Remove all skin marks with rubbing alcohol. What to do if the patient cannot attain or maintain position: Allow 1 minute of rest in the patient's most comfortable position. Try again. If that fails, abort the test. Record "N/A" and indicate the reason for failure. Indications to discontinue test: Failure on second attempt. Position in which to leave the patient: In his or her most comfortable position.

References

[1] Nachemson AL. The lumbar spine: an orthopaedic challenge. Spine. 1976;1:59-71. [2] Beals RK, Hickman NW. Industrial injuries of the back and extremities, comprehensive evaluation--an aid in prognosis and management: a study of one hundred and eighty patients. J Bone Joint Surg [Am]. 1972;54:1593-1611. [3] McKenzie RA. Manual correction of sciatica scoliosis. NZ Med J. 1972;76:194-199. [4] Lankhorst GJ, van der Stadt RJ, van der Korst JK. The natural history of idiopathic low back pain. Scand J Rehabil Med. 1985; 17:1-14. [5] Kapp JR, Alexander AH, Turocy RH, et al. The use of lumbar extension in the education and treatment of patients with an acute herniated nucleus pulposus. Clin Orthop. 1986;202: 211-218. [6] McKenzie RA. The Lumbar Spine: Mechanical Diagnosis and Therapy. Waikanae, New Zealand: Spinal Publications Ltd; 1981. [7] Maitland GD. Vertebral Manipulation. 5th ed. London, England: Butterworth & Co (Publishers) Ltd; 1986. [8] Mayer TG, Gatchel RJ. Functional Restoration for Spinal Disorders: The Sports Medicine Approach. Philadelphia, Pa: Lea & Febiger; 1988. [9] Joint Motion: Method of Measuring and Recording. Chicago, Ill: American Academy of Orthopaedic Surgeons; 1965:46. [10] Moll J, Wright V. Measurement of spinal movement. In: Jayson M, ed. The Lumbar Spine and Back Pain. New York, NY: Grune & Stratton Inc; 1976:96. [11] Schober P. The lumbar vertebral column in backache. Munchener Medizinisch Wochenschrift. 1937;84:336-338. [12] Macrae IF, Wright V. Measurement of back movement. Ann Rheum Dis. 1969;28:584-589. [13] Reynolds PMG. Measurement of spinal mobility: a comparison of three methods. Rheumatol Rehabil, 1975;14:180-185. [14] Fitzgerald GK, Wynveen KJ, Rheault W, Rothchild B. Objective assessment with establishment of normal values for lumbar spinal range of motion. Phys Ther. 1983;63:1776-1781. [15] Beattie P, Rothstein JM, Lamb RL. Reliability of the attraction method for measuring lumbar spine backward bending. Phys Ther. 1987;67: 364-369. [16] Snedecor GW, Cochran WG. Statistical Methods. Ames, Iowa: Iowa State University; 1989:37-39, 237-253. [17] Task Force on Standards for Measurement in Physical Therapy. Standards for tests and measurements in physical therapy practice. Phys Ther. 1991;71:589-622. [18] Van Adrichem JAM, van der Korst JK. Assessment of the flexibility of the lumbar spine: a pilot study in children and adolescents. Scand J Rheumatol. 1973;2:87-91. [19] Hoppenfeld S. Physical Examination of the Spine and Extremities. East Norwalk, Conn: Appleton-Century-Crofts; 1976:241. [20] Maggee D. Orthopaedic Physical Assessment. Philadelphia, Pa: WB Saunders Co; 1984: 199. [21] Loebl W. Measurement of spinal posture and range of spinal movement. Ann Phys Med. 1967;9:103-110. [22] Mayer TG. Using physical measurements to assess low back pain. J Musculoskel Med. 1985; 2:44-59, [23] Mayer TG, Tencer AF, Kristoferson S. Use of noninvasive technique for quantification of spinal range-of-motion in normal subjects and chronic low back dysfunction patients. Spine. 1984;9:588-595. [24] Keeley J, Mayer TG, Cox R, et al. Quantification of lumbar function, part 5: reliability of range of motion measures in the sagittal plane and in vivo torso rotation measurement technique. Spine. 1986;11:31-35. [25] Merritt JL, McLean TJ, Erickson RP. Measurement of trunk flexibility in normal subjects: reproducibility of three clinical methods. Mayo Clin Proc. 1986;61:192-197. [26] Gill K, Krag MH, Johnson GB, et al. Repeat-ability of four clinical methods for assessment of lumbar spinal motion. Spine. 1988;13:50-53. [27] Cohen J. Statistical Power for the Behavioral Sciences. Rev ed. New York, NY: Academic Press Inc; 1977:384. [28] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86:420-428. [29] Pearson ES, Hartley HO. Biometrika Tables for Statisticians: Volume 1. New York, NY: Cambridge University Press; 1966:202. [30] Richman J, Mackrides L, Prince B. Research methodology and applied statistics, part 3: measurement procedures in research. Physiotherapy Canada. 1980;32:253-257. [31] Burdett RG, Brown KE, Fall MP. Reliability and validity of four instruments for measuring lumbar spine and pelvic positions. Phys Ther. 1986;66:667-684, [32] Marras WS, Wongsam PE. Flexibility and velocity of the normal and impaired lumbar spine. Arch Phys Med Rehabil. 1986;67:213-217. [33] Ottenbacher KJ. Evaluating Clinical Change: Strategies for Occupational and Physical Therapists. Baltimore, Md: Williams & Wilkins; 1986:129-130. [34] Oyster CK, Hanten WP, Llorens LA. Introduction to Research: A Guide to the Health Science Professional. Philadelphia, Pa: JB Lippincott Co; 1987:47.