Sunteți pe pagina 1din 5

Acta Neurochir (Wien) (1996) 138:4044

Acta Neurochirurgica 9 Springer-Verlag1996 Printed in Austria

A Critical Assessment of Clinical Diagnosis of Disc Herniation in Patients with Monoradicular Sciatica
M. J. Albeck
University Clinic of Neurosurgery, Rigshospitalet, Copenhagen

Summary
The diagnostic power or clinical parameters in the diagnosis of lumbar disc herniation in patients with monoradicular pain was evaluated in a prospective study with a 100% verification of the diagnosis. Eighty patients with monoradicular pain corresponding to the fifth lumbar or the first sacral nerve root were included. Pre-operatively a number of clinical parameters were recorded and compared to the intra-operative finding of a disc herniation. The parameters were analysed by receiver operating characteristic (ROC) curves. Results from the available literature were analysed by ROC curves for comparison. In 76% of the cases a disc herniation was discovered. The level of the disc herniation was correctly predicted in 93% of these cases by the location of the pain alone or supplemented by neurological signs. Apart from radicularly distributed pain, all parameters in the present study and in the literature had no or low diagnostic accuracy. Thus, in patients with monoradicular sciatica further clinical parameters do not add to the diagnosis of lumbar disc herniation.

recent years, this is n o t the case for clinical e x a m i n a tions. Most studies on clinical tests were m a d e m a n y years ago, with different scientific traditions from those of today. To establish the value of a diagnostic test the popu l a t i o n m u s t be well-defined. W h e n a p p l y i n g n o s o logical probabilities, the sensitivity and specificity m u s t be c o n s i d e r e d j o i n t l y [1, 4]. A final diagnosis is often o n l y o b t a i n e d in patients with positive tests, w h i c h i n t r o d u c e d v e r i f i c a t i o n bias, a flaw in m a n y studies. In the present study all the patients had m o n o radicular sciatica, and were all operated on i n d e p e n dent of any i m a g i n g e x a m i n a t i o n . The results were assessed by receiver operating characteristic (ROC) curves to express the true positive fraction (TPF) equal to sensitivity as a f u n c t i o n of the false positive fraction (FPF) equal to 1-specificity.

Keywords: Intervertebral disc displacement; neurological examination; quality of test; receiver operating characteristics curve.

Material and Methods


The prospective material comprises 80 patients (32 females and 48 males) with monoradicular pain from 5th lumbar or 1st sacral root. Conservative treatment should have failed. Patients with previous low back surgery or age below 18 years and above 60 years were excluded, as were patients with the need for acute surgery. Pain should be present also without provocation from straight leg raising. Pain was accepted as evidence of L5 root compression when distributed to the antero-lateral aspect of the calf and to the dorsum of foot, and as evidence of S 1 root compression when referred to the posterior portion of the calf extending to the heel and the lateral aspect of the foot. If pain did not extend below the ankle, at least one additional neurological sign was demanded for the patient to be included. These signs were for the L5 root: a) hypaesthesia in the dorsum of foot, b) weakness of dorsiflexion of foot or first toe, c) impaired medial hamstring reflex. For the S 1 root the signs were: a) hypaesthesia at the lateral aspect of the foot, b) weakness of the plantarflexion of the foot or first toe, c) impaired Achilles tendon

Introduction
T h e m e d i c a l history and signs are i m p o r t a n t in the d i a g n o s t i c clarification of patients with low back p a i n and sciatica and in the d e c i s i o n on i m a g i n g e x a m i n a tions and specialist referral [5]. E v i d e n c e of l u m b a r disc h e r n i a t i o n is d i s c o v e r e d in 2 0 - 3 0 % of a s y m p t o m a t i c p e r s o n s on m y e l o g r a p h y , computed tomography and magnetic resonance imaging. This indicates that the therapeutic d e c i s i o n in patients with low b a c k trouble and sciatica can n o t be b a s e d o n r a d i o l o g i c a l e x a m i n a t i o n alone. W h i l e m a n y studies have b e e n p u b l i s h e d about the d i a g n o s t i c potentials of r a d i o l o g i c a l m e t h o d s in the

M. J. Albeck: A Critical Assessment of Clinical Diagnosis of Disc Herniation reflex. 57 patients were included on the basis of monoradicular pain referred to the foot alone and 23 on the basis of radicular pains in combination with other signs. The median age was 40 (21-59). The medical history and clinical findings were carefully recorded prior to surgery. Sciatica or simultaneous onset of low back pain and sciatica contrary to low back pain preceding sciatica were considered indicative of disc herniation. Claims for workers compensation was regarded as speaking against a disc herniation provided except pain was positive if leg pain was aggravated by coughing, laughing or defaecation. Segmental spasm was the visual interpretation of impaired movement between the lower lumbar segments. Trunk list was examined with the patient in the standing position. The test was considered positive if there was a list of the trunk to either side during flexion of the back. The finger-floor distance was considered suggestive of disc herniation if it was only to the knee level or above. The straight leg raising test was described as positive only if radicular pain was elicited. Hypaesthesia should have a dermatomal distribution to be considered. Impaired reflexes were compared with the nonsymptomatic side. All the patients were investigated using myelography, CT- and MR-scanning. These examinations were available at surgery, but were not included in the decision to include the patient in the study. The level according to the root syndrome was always explored. In 9 cases the neighbouring disc level was explored as well, either because of a negative disc exploration at the clinical level or because paraclinical examination rendered suspicion of herniation at this level. The surgical findings were carefully recorded and considered the definitive diagnosis. A disc herniation was defined as an extruded nucleus pulposus tissue through a defect in the annulus fibrosus. A bulging disc alone was not considered as a positive finding. Statistics: The true positive fraction, TPF (sensitivity) and the false positive fraction, FPF (1-specificity) was calculated as p {T + I D +} and p {T + I D-}, i.e., the probability of a positive test given presence and absence of a disc herniation, respectively. The ROC diagram was plotted as TPF as a function of FPF for varying decision thresholds. For dichotomous variables only a single point was plotted. Points close to the 45 ~ line from (0, 0) to (1, 1) are of no diagnostic value. An optimal test will be close to the upper left corner (0, 1) [20, 24]. To get the ROC for a combination of all the parameters were added. All positive parameters were weighted equally and assigned the value one. The resulting ROC curve for varying decision thresholds, i.e., number of positive parameters, was drawn. The significance level was based on the Mann-Whitney statistic [14]. As significance level X2 = 3.84, df = 1 corresponding to a significance level of 0.05 was used [13]. Available values of sensitivity and specificity from the literature [11, 12, 18, 19, 22] were plotted in a ROC diagram. These figures include sciatica, scoliosis, hypaesthesia, motor weakness, and straight leg raising test. The study design included informed consent and the patients accepted to be operated upon on the basis of clinical judgement alone, even if paraclinica! investigations showed no herniation. The study was approved by the Scientific-Ethical Committee of Copenhagen.

41

predicted in 57 (93%) of these 61 cases. The four cases in whom the diagnosis with regard to level was incorrect were two at the 4th lumbar disc and two at the 5th lumbar disc. In Fig. 1 the values are presented in the ROC-diagram. All values are below or close to the X2 = 3.84 curve which implies that all the tests have no or very low diagnostic accuracy. The ROC curve after adding the number of positive results for all the parameters in Fig. 1 is displayed in Fig. 2. This curve is close to the 45 ~ line. If the analysis was restricted to the 57 patients who were included on the findings of sciatica only, this would not change the results.
True positive fraction
1 S~
0.8

Onset i Provided pain x Workers c o m p e n s a t i o n


Scoliosis

9 Segmental s p a s m Z T r u n k list 9 Finger-floor d i s t a n c e 0.4 / 0.2 /// 0 v Straight l e g r a i s i n g


Paresis

N Muscle wasting
9 I, i m p a i r e d reflex

-- H y p a e s t h e s i a o V" , 0
, , , , ! , ,

0,2

0.4

0.6

. . . . 0.8

False positive fraction

Fig. 1. Receiver operating characteristic (ROC) diagram of twelve clinical parameters in patients operated on for suspected lumbar disc herniation. Points above the dashed line are significantly diagnostic (p < 0.05)

True positive fraction


1

0.8

0.6

//

. ..-

0.4

///

."

0r2

,'

0.2

0.4

0.6

0.8

False positive fraction

Results
At operation a disc herniation was exposed in 61 cases. The level of the disc herniation was correctly

Fig. 2. Receiver operating characteristic (ROC) curve (dotted line) of the combination of twelve clinical parameters in patients with suspected lumbar disc herniation. The dashed line represent the p = 0.05 limit

42
T r u e positive fraction 1.0 -L

M.J. Albeck: A CriticalAssessmentof Clinical Diagnosisof Disc Herniation ence of sciatica has a high predictive value of a disc herniation. But it is not possible to say anything about the negative predictive value of sciatica. To see if the conclusion is valid for other studies, corresponding values of sensitivity and specificity from the available literature were plotted in a ROC diagram. It was only possible to find few valid figures, as either the figures for patients with negative test were not provided or the figures for the test were not provided for patients with negative disc exploration. Values for straight leg raising sign [12, 16], sensory disturbances [16, 19], paresis [16, 19], scoliosis [19], and sciatica [5] were obtained. As the size of the population differs between the various studies it is not possible to plot a single significance curve, but it is apparent that all tests, except for sciatica, are close to the 45 ~ line, which means that they are diagnostically uninformative. The only parameter placed in the upper left corner is sciatica [6, 22]. A strong correlation between radicular pain distribution an a myelogram indicative of disc herniation has been demonstrated by pain drawing [25]. When the clinician evaluates a patient he does not rely on a single parameter alone, but will summarize elements from the medical history and physical examination into an imaginary likelihood that the patient has a disc herniation and will benefit from surgery. This has resulted in construction of several rating scales with a more or less arbitrary weighting of the parameters [8, 15, 23]. The relative weight of each parameter may vary according to different clinicians, but assuming the weight of one for every parameter is probably close to common clinical practice. When the parameters are simply added it is obvious from the ROC diagram that this does not increase the diagnostic ability as the resulting ROC plot is close to the 45 ~ line. As all the parameters have a low diagnostic accuracy, it is unlikely that a more elaborated rating scale will be better. Values of sensitivity and specificity, including ROC curves, must always be considered together with the frequency of disc herniation amongst new cases in the population in question. If the sensitivity of sciatica is 0.95 and the specificity is 0.88 [8], and the relative frequency of disc herniation is 75%, which is likely in a neurosurgical department, then the corresponding positive predictive value will be 0.96 and the negative predictive value 0.85. On the other hand, if the patient is seen by the general practitioner the relative frequency of disc herniation can be assumed

0.8

Straight leg raising


0.6 + +

+ Hypaesthesia :~ Paresis 9 Scoliosis

0.4

x Sciatica

0.2

0.0 0.0

. . .0.2 . . .

014 '

'

' 016 '

'

018 '

'

' 1.0

False positive fraction

Fig. 3. Receiver operating characteristic (ROC) diagram of five clinicalparameters obtainedfrom the literaturein patients with suspected lumbar disc herniation. The ROC diagram of values of sensitivity and specificity from the literature is shown in Fig. 3. The values for hypaesthesia, motor weakness, straight leg raising and scoliosis are all close to the 45 ~ line. The only parameter which is situated in the upper left corner is sciatica.

Discussion The clinical history and physical examination play a major role in the decision to perform surgery in patients with radiologically confirmed disc herniation. Yet, the scientific foundation of most parameters is deficient. Common flaws in the current literature are: 1) the results of the test are only given for patients where a disc herniation is discovered [7, 17, 18, 22, 26]; 2) only operative findings for patients with a positive test are given [12]; 3) no final diagnosis is obtained [2]; 4) only patients with a positive test are included [21]; 5) exclusively accuracy or sensitivity are presented. When presenting the diagnostic value of a parameter the sensitivity and specificity must be considered together. In the present study the conclusion for most of the parameters would be, that these parameters are suitable to make the diagnosis of a disc herniation (high sensitivity) and poor for excluding the diagnosis (low specificity). When looked at in the ROC-diagram, it is apparent that the tests have no or a low diagnostic power. The patients are selected on the basis of monoradicular sciatica and it is consequently not possible to calculate the specificity of this symptom, but the high incidence of disc herniation suggests that the pres-

M. J. Albeck: A Critical Assessment of Clinical Diagnosis of Disc Herniation to be 1% and consequently the positive predictive value will be 0.07 and the negative predictive value 0.99. In spite of these limitations, R O C curves are valuable for testing the relative a c c u r a c y o f different tests. It m a y be questioned if a different result w o u l d have been f o u n d in a population without obvious monoradicular pain. The figures f r o m the literature are compiled f r o m various populations. It has been demonstrated by Hakelius [12] that there is no difference in the incidence o f a disc herniation in patients with m o n o - and polyradicular syndromes. It is thus likely that the result is true for m o s t patients with sciatica. The situation m a y be different for a population of patients without leg pain, but as virtually all patients w h o are candidates for surgery have pain in the legs, this is o f small relevance. Twenty-three of the patients in the present study were included on the basis o f leg pain together with at least one neurological sign, as their pain did not extend below the ankle. This introduces some selection bias. However, excluding these patients f r o m the analysis did not change the results. Kortelainen [18] was able to localize the level o f disc herniation in 93% by pain projection and found that pain referred to the 5th lumbar distribution was more reliable for "level" diagnosis than pain referred to the first sacral distribution. In the present study it was possible f r o m the pain distribution to predict the correct level in 93% of patients with a disc herniation, but there was no difference in the reliability in level prediction between the L5 root and the S 1 root. Other authors suggest that it is not possible to distinguish between an affection of the 5th lumbar and 1 st sacral root f r o m the distribution o f pain alone [3, 10]. The results o f the tests in this and most other studies reflect the examiners' subjective interpretation. A more objective examination by instruments m a y improve the diagnostic accuracy of sensory [26] and m o t o r c h a n g e [9], but are tess suitable for routine examination.

43

against it. E v e n though it does not add to the diagnosis o f a disc herniation, the medical history should be obtained and a precise clinical examination carried out to u n c o v e r other disorders like n e o p l a s m and poly n e u r o p a t h y and to expose non-organic problems.

Acknowledgement
Supported by Kathrine & Vigo Skovgaards Foundation and The Danish Hospital Foundation for Medical Research, Region of Copenhagen, The Faroe Islands and Greenland.

References
1. Beam CA, Sostman HD, Zheng J-Y (1991) Status of clinical MR evaluations 1985-1988: baseline and design for further assessments. Radiology 180:265-270 2. Blower PW (198I) Neurological patterns in unilateral sciatica. A prospective study of 100 new cases. Spine 6:175-179 3. Centor RM, Schwartz JS (1985) An evaluation of methods for estimating the area under the receiver operating characteristic (ROC) curve. Med Decis Making 5:149-156 4. Cooper LS, Chalmers TC, McCally M, Berrier I, Sacks HS (1988) The poor quality of early evaluations of magnetic resonance imaging. JAMA 22:3277-3280 5. Deyo RA, Rainville J, Kent DL (1992) What can the history and physical examination tell us about low back pain. JAMA 268:760-765 6. Deyo RA, Tsui-wu YJ (1987) Descriptive epidemiology of low back pain and its related medical care in the United States. Spine 12:264-268 7. Edgar MA, Park WM (1974) Induced pain patterns on passive straight leg raising in lower lumbar disc protrusion. J Bone Joint Surg 56B: 658-667 8. Finneson BE, Cooper VR (1979) A lumbar disc surgery predictive score card. A retrospective evaluation. Spine 4:141-144 9. Finsterbush A, Frankel U, Pharm B, Arnon R (1983) Quantitative power measurement of extensor hallucis longus. Spine 8: 206-210 10. Friis ML, Gulliksen GC, Rasmussen P, Husby I (1977) Pain and spinal root compression. Acta Neurochir (Wien) 39: 241-249 11. Hakelius A, Hindmarsh J (1972) The comparative reliability of preoperative diagnostic methods in lumbar disc surgery. Acta Orthop Scand 43:234-238 12. Hakelius A, Hindmarsh H (1972) The significance of neurological signs and myelographic findings in the diagnosis of lumbar root compression. Acta Orthop Scand 43:239-246 13. Hanley JA, McNeil BJ (1982) The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 143:29-36 14. Hanley JA, McNeil BJ (1983) A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148:839-843 15. Herron LR, Turner J (1985) Patient selection for lumbar laminectomy and discectomy with a revised objective rating system. CIin Orthop 199:145-152 16. Jensen OH (1987) The level-diagnosis of a lower lumbar disc herniation: the value of sensibility and motor testing. Clin Rheumatol 6(4): 564-569

Conclusion
The best clinical indicator of a disc herniation is sciatica. Further clinical examination does not add significantly to the diagnosis. All patients with sciatica, w h o do not respond to conservative treatment, should be referred for CT or MRI. Patients with m o n o radicular leg pain and a positive imaging examination can safely be referred for surgery even if there are no positive signs, unless, e.g. n o n - o r g a n i c signs speak

44

M. J. Albeck: A Critical Assessment of Clinical Diagnosis of Disc Herniation 23. Spengler DM, Freeman CW (1979) Patient selection for lumbar disctomy. Spine 4:129-134 24. Swets JA (1979) ROC analysis applied to the evaluation of medical imaging techniques. Invest Radiol 14:109-120 25. Ud'n A, Landin LA (1987) Pain drawing and myelography in sciatic pain. Clin Orthop 216:I24-130 26. Weise MD, Garfin SR, Gelberman RH, Katz MM, Thorne RP (1985) Lower-extremity sensibility testing in patients with herniated lumbar intervertebral discs. J Bone Joint Surg 67A: 1219-1224

17. Khuffash B, Porter RW (1989) Cross leg pain and trunk list. Spine 14:602-603 18. Kortelainen P, Puranen J, Koivisto E, L~hde S (1985) Symptoms and signs of sciatica and their relation to the localization of the lumbar disc herniation. Spine 10:88-92 19. Kosteljanetz M, Espersen JO, Halaburt H, Miletic T (1984) Predictive value of clinical and surgical findings in patients with lumbago-sciatica. A prospective study (Part I). Acta Neurochir (Wien) 73:67-76 20. Metz CE (1978) Basic principles of ROC analysis. Semin Nucl Med 8:283-298 21. Porter RW, Miller CG (1986) Back pain and trunk list. Spine 11: 596-600 22. Spangfort EV (1972) The lumbar disc herniation. Acta Orthop Scand [Suppl] 142:1-95

Correspondence: Michael J. Albeck, M.D., 431 Neurosurgical Department, Hvidovre University Hospital, Ketteg~rds All~ 30, DK-2650 Hvidovre, Denmark.

S-ar putea să vă placă și