14
Surgical Outcome Prognosis for the Patients with Degenerative Lumbar Spine Disease Y.I. Zhuravlev, G.I. Nazarenko, A.M. Cherkashov, V.V. Ryazanov, A.G. Nazarenko

Surgical Outcome Prognosis for the Patients with Degenerative Lumbar Spine Disease Y.I. Zhuravlev, G.I. Nazarenko, A.M. Cherkashov, V.V. Ryazanov, A.G

Embed Size (px)

Citation preview

Surgical Outcome Prognosis for the Patients with Degenerative Lumbar Spine Disease

Y.I. Zhuravlev, G.I. Nazarenko, A.M. Cherkashov, V.V. Ryazanov, A.G. Nazarenko

«Knowledge to foresee in order to be able»

To develop algorithms for outcome prognosis for the patients with degenerative lumbar spine disease

Auguste Comte

Objective:

Groups of Patients

1-stgroup

Radiofrequency destructionof facet nerves (RFG)

185 patients,

average age 54.7 years

2-ndgroup

Percutaneous laser discectomy of lumbar discs with clinically significant protrusion detected (LD)

3-rdgroup

Microdiscectomy at one or more levels with intervertebral disc extrusion without stabilization of vertebral segments (MLD)

4-thgroup

Discectomy followedby stabilization, (Fusion)

39 patients,

average age 44.6 years

182 patients,

average age 44.3 years

105 patients,

average age 47 years

511 patients with degenerative lumbar spine diseases were operated on from 1997 through 2010

4-thgroup

Groups of Patients

2-ndgroup

1-stgroup

3-rdgroup

Surgical OutcomeSurgical outcome was assessed according to the modified criteria of

Kawabata et al. (1973)

Marked improvementand no disabilities

(minor sensory deficitand paresis grade 4/5

with improvementat least 1 grade)

I good (class 1)

Some residual symptoms (minor complaints)

and abnormal objective findings (minor sensory

deficit, mild atrophy, and paresis grade 3/5 or 4/5 with

improvementon at least one grade)

II fair(class 2)

Poor, no improvement (major complaints, marked deficits

and atrophy)or deterioration

III poor(class 3)

II poor(class 2)

I good

(class 1)

Association of Classes

I good

(class 1)

II fair(class 2)

III poor(class 3)

N List of symptomsPoints

0 1 2

1 Low back pain none sometimes always

2 Cross Lasegue symptom negative 45-600 < 450

3 Low extremities muscles paresis none grade 3-4 grade 0-2

4 Knee-jerk reaction normal assymetric absent on both sides

5 Achilles-jerk reaction normal assymetric absent on both sides

7 X-ray features of degenerative disease no 1 segment 2 and > segments

8 Degeneration on MRI and CT no 1 segment 2 and > segments

9 Disc herniation no protrusion extrusion/sequestration

Measures for the patient evaluation

Measures for the Patient Evaluation, 28 Symptoms

I good

(class 1)

II poor(class 2)

Data was analyzed with original “RECOGNITION” software, allowing the application of the patients’ identification algorithm to the possible outcomes

Recognition Software

DATABASERecognition

Software

For prognosis building following methodshave been applied:

Prognosis MethodsR

ecog

nit

ion

Sof

twar

e

Test Recognition Algorithm (TRA)

Decision Trees (DT)

Logical Patterns (LP)

Outcome prognosis was based on collective decision of these methods («voting»)

Voting PrincipleR

ecog

nit

ion

Sof

twar

e

Test Recognition Algorithm (TRA)

Decision Trees (DT)

Logical Patterns (LP)

II poor(class 2)

I good (class 1)

II poor(class 2)

II poor(class 2)

Prognostic Accuracy (rate of correct answers at cross-validation)

GroupI

For radiofrequency destruction of facet nerves 89.7 – 100%

GroupII

GroupIII

GroupIV

For laser discectomy 89.7 – 100%

For microdiscectomy 93.8 – 99.2%

For spinal stabilization 87.9 – 96.9%

Registry Architecture

Conclusion

• Algorithms obtained can be used for the outcome prognosis of the arbitrary new patients.

• The prognostic system allows to select an optimal operation type for the patients with degenerative lumbar spine disease

• Prognosis module is a part of «Russian Spine Registry»

• Prognosis module of «Russian Spine Registry» is available in English language

• www.spineregistry.ru