GB2453022A - Lumbar flexion measurement device - Google Patents

Abstract

A device for measuring lumbar flexion comprises a support body 4, a pair of spaced apart first locating means 5, a flexible measuring means 2 mounted on the support body, and a second locating means 1 attached to the measuring means 2. The measuring means 2 is extendible to accommodate movement of the second locating means 1 relative to the first locating means 5. The locating means 1 and 5 may define cavities to arrange over features on a patient's back and may be cupped. The measuring device 2 may comprise a tape, be retractably mounted on the support body 4, mounted on a reel or be slidable within the support body 4. The first locating means 5 may be symmetrical about the measuring means 2 with the distance from the measuring means 2 adjustable. The measuring means 2 may also have a sensing means and a display.

Description

LUMBAR FLEXION MEASUREMENT

The present invention relates to measurement of lumbar flexion.

Evaluating impairment of the lumbar spine according to the range of movement (RUM) is well accepted in the clinical practice setting and is often subsequently used to select appropriate treatment techniques and monitor patient's progress. Trunk flexion is usually recorded from subjective observation (Maitland 1986), but objective techniques do exist. One of the more commonly used measures is a skin distraction technique, originally described by Schbber (1937; Cited in Williams et al 1993) for the assessment of lumbar flexibility. This technique seems to be an ideal measurement tool because it is convenient, easy to use and not costly (Chok 2001).

Williams et al (1993) cite the original Schober technique to have involved using a tape measure held directly over the spine, between the lumbosacral junction and a point 10cm above, with the patient in the neutral standing position. When the patient moves forward into fall lumbar flexion, the lordosis is reduced and the increase in distance between the marks is suggested to give an estimate of spinal flexion ROM. It is important to note that a variety of modifications to the Schober's test exist, some of which are referred to as modified Schober's test (MST) or modified modified Schober's test (MMST), but it is not always clear which method is used since these terms are often miss-used. Table 1 summarises some of the modifications of Schober's test described in the literature, but the list is not exhaustive.

Author Inferior mark Superior mark Schober(1937) Lumbosacral junction 10cm above Macrae and Wright (1969) 5 cm below lumbosacral 10 cm above lumbosacral junction Junction Magee(1992) Midway between PSISs 10cm above midway between PSISs Merrittet a! (1986)* 5cm below midway 10cm above midway between PSISs between PSISs Williams etal (1993) Midway between PSISs 15cm above midway between PSISs Macrae and Wright (1969) suggested during Schober's test both skin marks moved upwards relative to the spinous processes on anterior flexion because the skin was poorly tethered to deep structures over the lumbar spine and sacrum. They noted that the skin was relatively well tethered and indeed appeared to stretch from the coccygeal area. As a result of their observations, a point lower down on the sacrum was thought to offer an improvement and measurements were taken from a skin mark 10cm above to another mark 5cm below the lumbosacral junction -a distance of 15cm in the erect position. To support their modification, Macrae and Wright (1969) investigated the relationship between clinical and x-ray identification of the lumbosacral junction by placement of skin markers over the estimated junction of the lumbar and sacral spines. The identification of the lumbosacral junction in this study was found to be subject to an error of approximately 2cm when checked radiographically.

Further subjects (of unknown number) were studied to investigate how much error was introduced if marks were placed 2cm above or below the correct location for both Schober's test and the modification. With Schober's method, placing the markers 2cm too high resulted in an underestimate of up to 15° in spinal flexion measurement, and placing them 2cm too low caused an overestimate of up to 14°. With their suggested MST i.e the lower mark 5cm below lumbosacral junction, the errors were very much smaller, 5° and 3° respectively.

Macrae and Wright (1969) then attempted to demonstrate whether both the original Schobers test and their modification did what they said they did, i.e. give a true indication of the amount of spinal mobility. They used radiography as the gold standard' since they could accurately measure the angle of lumbar flexion in the same sagittal plane of Schobers test and its modification. Having already identified that faulty placement of skin marks impaired the accuracy of Schobers test more than the modified method, the resulting data on the tests validity reflected this. Although both methods were found to demonstrate a linear relationship between skin distraction marks and true forward flexion of the lumbar spine, the modified method appeared to improve the accuracy. The original Schober method demonstrated close association with a correlation coefficient of 0.9 and a standard error of 6.2° whilst the modified Schobers test displayed a similar correlation coefficient of 0.97 but the standard error was smaller at 3.250. These calculations highlight the importance of reporting error since this tells us more about the validity of a test in real terms, especially when the correlation coefficients of two sets of data are very similar. Although radiography seems to be accepted as an appropriate way to validate such measurements it is a two-dimensional procedure that does not take into account shifts of the spine laterally which a tape measure applied directly to the spine may accommodate. This friay be of more significance when validating the Schober test measurements of spines affected by pathology and pain when lateral or rotational movements may not be identified on a lateral radiograph of forward flexion.

Gracovetsky (1990) examined to what extent it is possible to obtain information on the spine movement from the kinematics of markers placed on the skin during Schobers test. The basic protocol used for this investigation consisted of taping 3mm steel balls over the spinous processes and the iliac crest of volunteers (n-25) to be x-rayed. To control for image magnification of the x-ray machine, a one-inch diameter steel ring is taped above multifidus as close to L3 as possible. By taking several AP and lateral radiographs of the subjects, measurements were apparently taken in a variety of postures but Gracovetsky (1990) only discusses the measure of lordosis. Calculations from the data obtained from the radiographs demonstrated that the angles of lordosis, as measured from the skin markers and relative position of vertebrae T12/L1 to L5/S1, were fairly linear indicating that the motion of external markers placed on the skin does contain information relevant to the estimation of the true lumbosacral angle. Descriptions of skin marker placement upon the spinous processes were not provided and since the exact location in relation to the spinous process should be consistent, observer error in placement may seriously effect the outcome validity. How the reliability of carrying out a procedure effects its validity has been demonstrated by Macrae and Wright (1969) but Gracovetsky (1990) does not reflect upon this possible source of error. A summary of the experimental results presents the average error between true and effective intersegmental mobility measurements, but the overall error in the lumbar spine angle and its estimate with skin markers is not presented. Thus the absence of satisfactory measures of accuracy between the true and measured values limits interpretation of the validity of the research. Gracovetsky (1990) concluded that it is not unreasonable to use skin markers to approximate the motion of the spine, but suggests that caution must be taken in obese people, presumably because of difficulties in locating bony landmarks. The population the author studied is not described in terms of height or weight restrictions such that one might question the arriving at expressing caution for a particular subgroup. To be more explicit, further validity studies on marker placement on larger subjects should be performed to demonstrate error calculations.

Most recently, Mahadevi and Lai (1998) attempted to re-validate Macrae and Wright's (1969) MST by determining the correlation between MST measurements for lumbar flexion and corresponding radiographic measurements, for six subjects. The raw data is presented in the article for both methods of measurements of lumbar flexion and the authors used a Pearson's correlation coefficient to determine the validity. A correlation coefficient of 0.98 was calculated between the angle and MST for flexion, indicting excellent correlation, but there was no reference to the measurement error. The article is clear that for this part of the research, one physiotherapist conducted the MST measurements on the subjects, but it is not stated who measures the angle on the radiograph, and more importantly, whether the same person takes each of the radiographic measurements. The consistent measurement of the angle is of importance in such a study aimed at validating a procedure, since if different observers are involved another variable is added.

To obtain information about the validity of a tool, the variables of the experiment must all be fixed to reduce the introduction of error i.e. use the same subject, the same tool with a standardised protocol, and the same observer to make the measurements. Placement of the tape measure to record lumbar flexibility is in itself subject to unreliable application, difficulties possibly arising in locating anatomical landmarks and reapplying the tape measure with accuracy before and after movement. Therefore the present invention aims to provide a tool used to measure lumbar flexion which not only precise, but also can be applied reliably. The present invention therefore provides a device for measuring lumbar flexion with a modification of Schober's test that relies upon using bony prominences to locate the placement of the instrument.

The concept of reliability encompasses repeatability and precision of measures and in measuring human movement, there is a requirement for measurement tools and the operators of measurement tools to perform reliably under a number of different conditions, otherwise the data collected are meaningless (Durward et a! 1999). In order to judge whether the mechanical testing procedure is reliable, the equivalence and stability should be analysed in terms of agreement rather than association (Sim and Wright 2000). In order to investigate the inter-and intra-rater reliability it stands to reason that the subject being measured is consistent i.e. one subject who has a fixed amount of lumbar flexibility which may be controlled in the testing procedure by having the subject flex forward to reach a designated point. Having established that the tool and its protocol are precise and reproducible, the observer's measurement can truly be observed and the introduction of bias and inaccuracies to the resulting data will have been limited to those introduced by the observer. However, much of the research investigating the reliability of Schober's test has clearly not standardised each subjects ROM for each measurement.

Merritt et al (1986) investigated, among others, the reproducibility of Macrae and Wright's (1969) MST with 25 healthy, pain and pathology free subjects.

They measured the intra-rater consistency by having each of three investigators examine a subgroup of approximately 8 subjects on 3 different days. In a further 25 subjects inter-rater consistency was measured by having a different observer perform each of three examinations. Calculations of the coefficients of variance (CV) were performed for the three examinations of each subject and then the reproducibility over all subjects was summarised by calculating the mean and median CV over each individual CV. Care should be taken when interpreting the mean intra-rater CV in this study since the value is across three investigators for approximately 8 subjects, so the subject being tested is not fixed and neither are the examiners. Although the CV is an appropriate measure of dispersion of values relative to the mean in a distribution (Sim and Wright 2000), by amalgamating data and only presenting the mean CV it is impossible for the reader to judge the meaningfulness of the results. Rankin and Stokes (1998) consider using the CV to calculate reliability to be inappropriate since taking the mean score of all subjects has the potential to provide misleading estimates. Therefore, the good reproducibility' conclusions drawn from Merrittt et at (1986) are inappropriate when based upon the mean intra-and inter-rater CV values of 6.6% and 6.3% respectively. For the same reasons, the results from Gill et al (1988) (who determined the intra-exarniner repeatability of MST to be consistently good with a CV of 0.9%) must also be discarded. With no additional supportive data the reader has no knowledge of the boundaries and spread of the measurements and thus the usefulness of the piece of research must be questioned.

The MMST was concluded by Williams et a! (1993) to be a reliable method for measuring lumbar flexion in patients with low back pain. In this study, three therapists took lumbar flexion measurements on two occasions, 2 days apart, for a total of 15 subjects with low back pain, with or without leg pain.

The sample group inclusion and exclusion criteria were presented along with the sample group statistics. Emphasis was made on standardising the protocol and ensuring the accuracy of the tape measure for the research which similar studies negated to do. However, measuring reliability of any mechanical testing procedure on subjects with back pain introduces much room for variability since pain is infrequently constant and likely to vary from day-to-day. Low back pain severity may be reflected in the amount of flexibility a subject may, or may not have in the lumbar spine as well as being affected by additional extraneous factors including environment, psychological and social.

So if the subjects could have a large degree of variation in the amount of lumbar flexibility from day to day, the reliability of the tester would be difficult to assess. The authors make no comment upon this important area but focus rather on the Pearson Correlation Coefficients of 0.89, 0.78 and 0.83 for intra-rater reliability. Although the authors conclude their calculations to suggest moderate' reliability values, the Pearson's correlation coefficient is inappropriate because it measures the strength of linear association, not agreement i.e. the good correlation value does not guarantee that the results of the two measurements are almost the same. Indeed, in one set of measurements, the observer may have given more encouragement to the subject to flex further resulting in consistently larger measurements -but this will not affect the statistical calculations of correlation since the values may remain associated even though they do not agree. The inter-rater reliability of the therapists measurements were calculated more appropriately from analysis of variance (ANOVA)-derived intraclass correlation coefficients (ICCs) (3,1) to be 0.72. From this figure, the authors suggest that their results may be generalised to other raters of the MMST within the same setting and if the raters are trained to the same standard in the use of the same protocol.

However, Rankin and Stokes (1998) suggest that interpretation of the ICC in isolation is not that simple since the coefficient is just one point estimate of reliability and without the magnitude of confidence interval being calculated a true picture of reliability is not obtained.

Hyatianen et al (1991) used three different observers with appropriate background knowledge and skills to measure spinal mobility using nine different tests, including the Macrae and Wright (1969) modified Schober's test. Thirty male subjects (unknown low back pain/pathology status) were examined three times by three experienced physiotherapists during one day to obtain inter-rater reliability information. A further examination by one of the physiotherapists was carried out one week later to identify intra-rater reliability.

It is not known what the subjects did between testing procedures and the length of time between measurements on the same day is not clarified. Means of the measurements taken by each of the physiotherapists were calculated for the thirty subjects and correlation coefficients of the tests were calculated to be intra-rater r=0.88 and inter-rater r=0.87. However, this study has failed to account for controlling the variables within the subject(s) being tested i.e. each subject maybe able to flex their lumbar spine varying amounts according to external environment, temperature, time of day, pain etc. as well as according to pathological status and age etc. Thus the correlation calculations presented do not reflect this possible variability between measurements and cannot be used as a measure of reliability.

In much of the aforementioned studies, the repetition of whole sets of tests may have caused a systematic improvement in the results of a measure probably because tissues stretch more when warmed, or the subject learns to the test better during repeated performances. Equally, variation in the starting position of a test and variation in how each observer interprets and instructs subjects on how to perform the procedure protocol may contribute to errors. Hyatianen et al (1991) recognises that individual properties of the subjects, such as motivation, fatigue and pain experience affect the test results even though participants had participated voluntarily in the study, but had additional statistical tests been performed, they may have taken these factors into account. The available evidence regarding the reliability of the Schober' s test demonstrates the importance of supplying raw data, applying the correct statistical tests and also in providing an indication of the standard error of measurement/limits of agreement. In addition, standardising the protocol for the experimental procedure and exploring ways of employing a random sample of raters from a larger population to reduce the potential for bias in rater's ability would improve the generalizability of any further research.

Ultimately, the presence of consistent patterns of correlation between skin motion and vertebral motion demonstrate that skin motion cannot be random and suggests this motion to contain information characterising both the spine and its surrounding soft tissues (Gracovetsky et a! 1995). Since both may be affected by pathology, it must follow that skin motion is also affected by the presence of pathology but no research to date has confirmed this. Schober's test and its modifications should be validated for each subgroup of the population as determined by specific pathology and body type (i.e. obese subjects) in order to assess the differing effects upon the ratio between skin motion and true spinal mobility. If Schober's test is being used clinically as an objective marker of lumbar flexibility then what it actually measures may be of less importance than if the test were employed to evaluate the effect of pain and/or pathology upon function in order to formulate a dependable diagnosis.

The present invention provides a device for measuring lumbar flexion comprising a support body, a pair of spaced apart first locating means each arranged to locate on corresponding spaced apart features on a patient's back, a flexible measuring means mounted on the support body, and a second locating means attached to the measuring means and arranged to locate on a further feature of a patient's back, wherein the measuring means is extendible to accommodate movement of the second locating means relative to the first locating means.

The spaced apart features may be bone features, for example the PSIS (posterior superior iliac spine). The other feature may be a bone feature, for example a spinous process, such as the Li spinous process.

The first locating means may define respective cavities arranged to locate over said features. For example they may be cupped.

The second locating means may define an opening arranged to locate over said feature. For example it may comprise a ring or a cup or hook.

The measuring means may comprise a tape, and may have markings on it to allow extension or retraction of the measuring means to be determined.

The measuring means may be retractably mounted on the support body, for example on a reel.

The first mounting means may be located symmetrically about a mounting of the measuring means. The distance between the first locating means may be adjustable. The adjustment mechanism for the first locating means may be arranged to maintain their symmetry about the mounting for the measuring means during adjustment. For example, the adjustment mechanism may be arranged to control movement of the first locating means so that they move in opposite directions relative to the mounting body during adjustment. Alternatively the first locating means may be elongated in the horizontal direction so as to accommodate differences in distance between the features on which they locate.

The present invention further provides a method of measuring lumbar flexion of a patient the method comprising providing a device according to any foregoing claim, locating the first locating means on features of the patient's back, locating the second locating means on a feature of the patient's back, flexing the patient's back, and measuring extension of the measuring means.

Preferred embodiments of the present invention will now be described by way of example only with reference to the accompanying drawings in which: Figure 1 is a front view of a measuring device according to a first embodiment of the invention; Figure 2 is perspective view of a measuring device according to a second embodiment of the invention; Figure 3 is a front view of a measuring device according toa third embodiment of the invention; Figure 4 is a side view of the device of Figure 3 Figure 5 is a transverse section through the device of Figure 3; and Figure 6 is an enlargement of part of the section of Figure 5.

Referring to Figure 1, a measuring device according to a first embodiment of the invention comprises a support housing 4 which houses a reel 3 on which a tape measure 2 is wound. The reel 3 is provide with a return spring so that the tape measure can be pulled from the support housing 4, and when released will be retracted into the support housing 4.

A ring 1 is attached to the free end of the tape measure 2 and is marked with a scale 2a indicating the distance along the tape measure, from a point in the centre of the ring 1. The ring is arranged to be placed over, and hence located on, the spinous process Li in a patient's spine.

A pair of soft rubber cups 5 are each supported on the end of a respective support bar 5a, the support bars extending out of opposite sides of the support housing 4 perpendicular to the tape measure 2. Each of the support bars 5a has a ratchet 6 formed on it, and the two ratchets engage on opposite sides of a cog 7 so that the support bars 5a can move symmetrically in the direction transverse to the tape measure 2 so as to adjust the distance between the cups 5, whilst keeping them equidistant from the tape measure 2. The cups 5 are arranged to be placed under the PSISs of the patient.

The method of use of the device of Figure 1 includes the following steps: 1. Using the tool of Figure 1, measure the distance from LI to midpoint of PSISs with the subject in an upright position.

2. Instruct patient to flex fully forward, replace measuring tool and re-measure.

3. The difference between the two measurements may provide an indication of the lumbarsacral spine movement since it fixes to anatomical bony landmarks, not skin distraction.

A further embodiment is disclosed in Figure 2. This operates generally in the same way as the first embodiment. However the location cups 25 are not adjustable, but rather elongated in the horizontal direction so as to accommodate differences in spacing between the locating bone features on the patient's back. Also the reel on which the tape 22 is wound within the support body 24 is connected to a marked wheel 28 which rotates as the tape 22 is extended or retracted and is marked so that the amount of rotation of the wheel 28 is displayed, giving an indication of the amount by which the tape 22 has been extended or retracted. Furthermore the marked wheel 28 can be set to a zero reading with the tape 22 at any point. This means that the upper and lower locators, i.e. the ring 2land cups 25, can be located on the relevant features with the patient's back straight, and the reading then zeroed. The patient then bends their back to extend the tape 22, and the wheel 28 will indicate the increase in extension of the tape 22 resulting from the bending of the patient's back.

It will be appreciated that the display can be controlled electronically, and the measuring tape can be of any flexible material with some form of detectable marking on it, such as optical, magnetic, or shaped, that can be detected by an associated sensor to measure the degree of extension of the flexible material. This enables the display to be re-set, for example using a re-set input button, when the patient is in the upright position, so that the increase in length, or extension, of the flexible measure, can be measured by the sensor and displayed on the display.

Referring to Figures 3 to 6, in a third embodiment a measuring device a support body 34 is formed of moulded plastics material and is hollow and rectangular in cross section as shown in Figures 5 and 6. A flexible measuring strip 32, enclosed within a flexible sleeve 32a, extends through the support body 34 so that the sleeve can slide in the support body and the measuring strip can slide within the sleeve. The measuring strip 32 has a circular aperture 31 through one end arranged to form the upper locator.

The measuring strip 32 has a first scale marked on it, which is visible through an aperture 36 in the sleeve 32a, which forms a reference marker on the sleeve enabling the position of the measuring strip relative to the sleeve 32a to be indicated. A second scale is marked on the top surface of the sleeve 32a, and is visible through a hole 37 in the top of the support body 34, which forms a reference marker on the support body enabling the position of the sleeve relative to the support body to be indicated. A support bar 35a is mounted on the back of the support body 34, extending perpendicular to the measuring strip 32, and has a pair of cups 35 slidably mounted on it, so that their separation can be adjusted.

In use this device is similar to the previous embodiments. The lower supports 35 are located on the PSISs, the measuring strip 32 is slid into the sleeve 32a to zero the first measuring scale, and the sleeve 32a is then adjusted in the support body 34 until the aperture 31 is located on the Li spinous process, with the patient upright. The patient then bends over, while the sleeve 32a is held stationary in the support body 34 and the measuring strip 32 slides upwards in the sleeve 32a. When the patient is fully flexed forwards, the first scale indicates the degree of extension.

It will be appreciated that the provision of two sliding members, in this case the sleeve 32a and strip 32, allows one of them to be moved to adjust the device to the patient and the other to be moved to measure extension during flexing, and the exact way in which this is achieved could vary in other embodiments.

References Chok B 2001 A comparison of the flexion range of motion between persons with low back pain and pain free individuals using the modified Schober's test.

Physiotherapy Singapore 4(1): 10-12 Durwood BR, Baer GD, RowePJ 1999 Functional human movement: Measurement and analysis. Butterworth Heinemann, Oxford Gill K, Krag MH, Johnson GB, Haugh LD, Pope MH 1988 Repeatability of four clinical methods for assesment of lumbar spineal motion. Spine 13(1) :50-53 Gracovetsky S 1990 Musculoskeletal function of the spine. In: Winters JM, Woo SLY 1990 Eds Multiple muscle systems:Biomechanics and movement organization, Springer-Verlag, New York Gracovetsky S. Newman N, Pawlowsky M, Lanzo V. Davey B, Robinson L 1995 A database for estimating normal spinal motion derived from noninvasive measurements. Spine 20(9): 1036-1046 Hyytiainen K, Salmi nen JJ, Suvitie T, Wickstrom G, Pentti J 1991 Reproducibtlity of nine tests to measure spinal mobility and trunk muscle strength. Scand J Rehab Med 23:3-10 Macrae TF, Wright V 1969 Measurement of back movement. Annuls of RJheumatological Diseases 28:284-289 Magee D 1992 Orthopaedic physical assessment 2nd edn. WB Saunders Company, Philadelphia Mahadevi, Lai A 1998 Validity and reliability of the modified Schober's test in measuring lumbar flexion and extension in normal subjects. Physiotherapy Singapore 1(1):13-17 Maifland GD 1986 Vertebral manipulation 5th edn. Butterworth Heinemann, Oxford Memt JL, McLean TJ, Erikson RP, Offord KP 1986 Measurement of trunk flexibility in normal subjects: Reproducibility of three clinical methods. Mayo Clinic Proceedings 61:192-197 Moll JMH, Wright V 1971 Normal range of spinal mobility: an objective clinical study. Ann Rheum Dis 30:381-386. Cited in: Merrit JL, McLean TJ, Erikson RP, Offord KP 1986 Measurement of trunk flexibility in normal subjects: Reproducibility of three clinical methods. Mayo Clinic Proceedings 61:192-197 Rankin G, Stokes M 1998 Reliability of assessment tools in rehabilitation: an illustration of appropriate statistical analyses. Clinical Rehabilitation 12:187-199 Schober P 1937 The lumbar vertebral column in backache. Munchener Medizinisch Wochenschrift 84:336-338. Cited in.Williams R, Binkley, Bloch R, Goldsmith CH, Munik T 1993 Reliability of the modified-modified schdber and double inclinometer methods for measuring lumbar flexion and extension.

Physical Therapy 73(1) :26-37 Sim J, Wright C 2000 Research in healthcare: Concepts, designs and methods.

Stanley Thomes Ltd, Gloscestershire Williams R, Binkley, Bloch R, Goldsmith CH, Munik T 1993 Reliability of the modified-modified schober and double inclinometer methods for measuring lumbar flexion and extension. Physical Therapy 73(l):26-37 Winters JM, Woo SLY 1990 Eds Multiple muscle systems:Biomechanics and movement organization. Springer-Verlag, New York

Claims (17)

1. A device for measuring lumbar flexion comprising a support body, a pair of spaced apart first locating means each arranged to locate on corresponding spaced apart features on a patient's back, a flexible measuring means mounted on the support body, and a second locating means attached to the measuring means and arranged to locate on a further feature of a patient's back, wherein the measuring means is extendible to accommodate movement of the second locating means relative to the first locating means.

2. A device according to claim 1 wherein the first locating means define respective cavities arranged to locate over said features.

3. A device according to claim 2 wherein the first locating means are cupped.

4. A device according to any foregoing claim wherein the second locating means defines an opening arranged to locate over said feature.

5. A device according to any foregoing claim wherein the measuring means comprises a tape.

6. A device according to any foregoing claim wherein the measuring means is retractably mounted on the support body.

7. A device according to claim 6 wherein the measuring means is mounted on a reel.

8. A device according to any of claim 1 to 4 wherein the measuring means is slidable within the support body.

9. A device according to claim 8 wherein the measuring means comprises two members which are movable relative to each other wherein one of the members includes the second locating means.

10. A device according to claim 9 wherein the two members comprise a sleeve and a measuring strip slidable within the sleeve.

11. A device according to any foregoing claim wherein the first mounting means are located symmetrically about the measuring means.

12. A device according to any foregoing claim wherein the distance between the first locating means is adjustable.

13. A device according to claim 12 when dependent on claim 11 wherein the adjustment mechanism for the first locating means is arranged to maintain their symmetry about the mounting for the measuring means during adjustment.

14. A device according to claim 13 wherein the adjustment mechanism is arranged to control movement of the first locating means so that they move in opposite directions relative to the mounting body during adjustment.

15. A device according to claim 1 including sensing means for sensing extension of the measuring means.

16. A device according to claim 15 further comprising a display arranged to indicate the degree of extension of the measuring means, wherein the display can be re-set when the measuring means is at an initial extension.

17. A method of measuring lumbar flexion of a patient the method comprising providing a device according to any foregoing claim, locating the first locating means on features of the patient's back, locating the second locating means on a feature of the patient's back, flexing the patient's back, measuring extension of the measuring means.