Evaluation criteria for the assessment of occupational diseases of the lumbar spine - how reliable are they? -

Low back pain is one of the most common health problems worldwide [1, 2], but leads to a particularly significant healthcare burden in industrial countries [2, 3]. In Germany, 26% of all individuals participating in the national health insurance system require medical treatment for back pain at least once a year [4]. In addition to anatomical abnormalities and psychological disorders, occupational activity has been recognized as a major risk factor for the development of chronic low back pain [5‐8]. Overall, approximately 20% of all occupational activities appear to be associated with pathological spinal load, which could lead to early degenerative lumbar disc disease followed by chronic pain [9]. In Germany, the degree of load-related degeneration of the lumbar spine is relevant in the determination of worker compensation. Therefore, in 1993, the Federation of Commercial Occupational Insurance Associations (HVBG)—subsequently renamed as the German Association of Occupational Accident Insurance Funds (DGUV)—defined occupational disease no. 2108 that describes the development of deep lumbar back pain as a result of degeneration of the intervertebral discs due to non-physiological occupational load [10]. However, there are very few objective and reliable parameters for assessing the severity of occupational lumbar spinal diseases.

To overcome this problem, in 2000, 2 projects were initiated to improve our scientific knowledge regarding the evaluation of occupational spinal diseases. First, the German Spine Study evaluated the epidemiological dose-effect-relationship between occupational working exposure and the development of degenerative spinal diseases related to the intervertebral disc [11]. Based on these data, the “Mainz-Dortmund-Dose-Model” was developed as a uniform algorithm to correlate work load and degenerative changes of the lower back [12‐14]; this model is still currently used. Second, an interdisciplinary working group developed consensus recommendations regarding the following topics [10]: clinical und morphological parameters that correlate with load-related symptoms; causal factors related to the development of occupational spinal diseases; criteria for evaluating whether a patient needs to discontinue the stressful occupational activity; and criteria for assessing a reduction in earning capacity The recommendations of this working group were published in 2005, and remain in use at present [10]. The researchers defined 5 radiologic parameters for the assessment of pathologies of the intervertebral disc, evaluated using plain anteroposterior (AP) and lateral lumbar spinal radiographs, including intervertebral osteochondrosis, vertebral osteosclerosis, anterolateral and posterior spondylosis, and spondyloarthritis.

In Germany, orthopedic and trauma surgeons regularly examine patients with work-related pain and offer expert opinions to the DGUV on the severity of work-related degenerative spinal diseases. Thus, the spinal radiographs of thousands of patients are evaluated annually for the presence of intervertebral osteochondrosis, vertebral osteosclerosis, anterolateral and posterior spondylosis, and spondyloarthritis, to assess for the presence of occupation-related pathologies of the lumbar intervertebral discs. Nevertheless, despite their frequent use, the reliability of these parameters has not yet been evaluated. In the present study, we aimed to evaluate the inter- and intra-observer reliability of the radiological parameters, including intervertebral osteochondrosis, vertebral osteosclerosis, anterolateral and posterior spondylosis, and spondyloarthritis, using plain AP and lateral standing radiographs of the lumbar spine.

In the present study, the evaluation of intervertebral osteochondrosis by measuring the intervertebral disc spaces of the lumbar spine showed good inter-observer reliability and excellent intra-observer reliability. The classification of vertebral osteosclerosis showed moderate kappa values for both inter- and intra-observer reliability. These data suggest that the 2 parameters, including intervertebral osteochondrosis and vertebral osteosclerosis have sufficient reliability to permit decision making regarding the severity of degenerative occupational diseases of the lumbar spine. In contrast, the classification of lumbar antero-lateral and posterior spondylosis and the evaluation of the spondyloarthritis stage showed poor inter- and intra-observer reliability, which makes these methods unsuitable for daily clinical use.

Based on the DGUV recommendations, intervertebral osteochondrosis should be used as the main parameter for assessing degenerative occupational spinal diseases. Vertebral osteosclerosis, antero-lateral spondylosis, posterior spondylosis, and spondyloarthritis are considered as secondary co-factors that may develop as a result of decreased intervertebral disc spaces, but are not always present. These secondary co-factors should only be used for decision-making in individual cases where the measurement of the intervertebral disc space alone does not lead to a conclusion [10]. Based on our findings, we believe that vertebral osteosclerosis is the only reliable secondary co-factor in addition to intervertebral chondrosis for the assessment of the severity of spinal degeneration.

However, in clinical practice, the measurement and interpretation of intervertebral disc spaces using plain lateral radiographs has certain limitations. First, the radiological projection of vertebral bodies in plain radiographs significantly depends on the position of the central x-ray beam in relation to the spinal segment [19, 20]. Anderson et al. have shown that a lateral tilt of > 10° and an axial rotation of the spine of > 20° can result in significant inter-observer variation when measuring intervertebral disc spaces [19]. Therefore, in patients with lumbar scoliosis, a three-dimensional imaging technique may offer more reliable results relative to plain radiographs. To overcome this problem in the present study, we excluded patients with Cobb angles > 10°.

Similarly, in patients with a history of vertebral fractures, Scheuermann’s disease, or spondylolisthesis, it may be difficult to accurately identify the endplates. Some authors have proposed that in healthy people, the central disc spaces may slightly increase with increasing age, as a result of degenerative microfractures of the endplates that lead to a deformation of the vertebral body [15, 21]. This effect may result in the presence of constant disc spaces over time, even though the patient may have progressive spinal degeneration. Furthermore, there is significant variation between the disc spaces of healthy individuals, which makes it difficult to define reference values for the interpretation of reduced disc spaces in patients with spinal pathologies [15]. Finally, studies have found that intervertebral disc spaces vary significantly during the day depending on the activity pattern of individuals. The delivery of load to the spine during daily activities results in a reduction of water in the nucleus, which reduces the intervertebral space; this effect is reversible during the night [22]. Thus, the measurement of intervertebral chondrosis may produce reduced disc spaces when assessed using radiographs later in the evening, as compared to that on radiographs of the same patient obtained earlier in the morning.

To overcome these problems, Bolm-Audorff et al. developed an algorithm to assess the relative disc space [10, 16]. Therefore, the absolute intervertebral space of a lumbar segments is measured on the lateral x-ray. The values are multiplied with a correction factor to equalize the magnification effects due to radiological projection [15].

The segment with the highest disc space is used as the standard reference. The height of the remaining 4 segments is then calculated as a percentage of the segment with the highest disc space.

In the presented evaluation, inter- and intra-observer reliability of the spaces between the fourth and fifth vertebra were greater than those of other segments. These findings may be consistent with that noted in previous studies, and can be explained by the greater orthogonal geometrical exposure of this segment in relation to the x-ray beam as compared to the others segments [20, 23]. In the present study, the inter- and intra-observer reliability of the intervertebral disc space measurement via plain radiography showed lower ICC values, as compared to those obtained via MRI and ultrasound [24, 25]. Nevertheless, the reliability of the presented method was sufficient for clinical decision making.

However, besides reliability a diagnostic method also requires adequate validity to allow for use in clinical practice. Several trials have shown, that degenerative changes of the lumbar spine do not necessarily correlate with clinical symptoms [26, 27]. Even advanced imaging techniques fail to clearly identify the source of symptoms in patients with low back pain [28]. In addition, recent studies have shown that brain functional changes such as abnormal prefrontal cortex connectivity may also have significant impact on the development and maintenance of chronic back pain [29]. Therefore, future administration of workers compensation for low back pain may require a more comprehensive assessment of co-factors affecting this pathology.

The measurement of intervertebral osteochondrosis and classification of vertebral osteosclerosis showed an adequate inter- and intra-observer reliability when evaluated using plain radiography. However, the classification of antero-lateral and posterior spondylosis and the spondyloarthritis stage showed insufficient reliability for clinical use. Therefore, these 3 parameters cannot be recommended for the assessment of occupational diseases of the lumbar spine. Thus, we believe that the consensus recommendations of DGUV for the evaluation of occupational diseases of the lumbar spine may require additional revision.