Kopeček M., Bareš M., Brunovský M., Novák T., Stopková P., Šóš P., KoženýJ., Krajča V., Höschl C. EEG cordance as a predictor of response to antidepressivemedication - pooled analyses. XIV. World Congress of psychiatry, 20-25, September, 2008, Prague, Czech Republic. Česká a Slovenská Psychiatrie, 104; 2008, S2:1193.

25. 9. 2008


P-03-091
EEG CORDANCE AS A PREDICTOR OF RESPONSE TO
ANTIDEPRESSIVE MEDICATION - POOLED ANALYSES
INSTITUTIONS
1. 3rd Faculty Of Medicine, Charles University and Prague Psychiatric Centre, Department of Psychiatry and Medical Psychology, Prague, Czech
Republic
2. Bulovka Hospital, Department of Neurology, Prague, Czech Republic
3. University of North Carolina, Department of Psychiatry and CERT-BD, Chapel Hill, United States
AUTHORS
1. Miloslav Kopecek1,2,3, MD, PhD, kopecek@pcp.lf3.cuni.cz, 2. Martin Bares1, MD, bares@pcp.lf3.cuni.cz
3. Martin Brunovsky1,2, MD, PhD, brunovsky@pcp.lf3.cuni.cz, 4. Tomas Novak1, MD, novak@pcp.lf3.cuni.cz
5. Pavla Stopkova1, MD, PhD, stopkova@pcp.lf3.cuni.cz, 6. Petr Sos1, MD, sos@pcp.lf3.cuni.cz, 7. Jiri Kozeny1, Dr., CSc
8. Vladimir Krajca2, Ing., CSc, 9. Cyril Hoschl1, MD, DrSc., FRCPsyc
Objectives: Three small-scale studies have shown that a decrease of
theta prefrontal EEG cordance after 1 week on antidepressant medication
can predict clinical response to treatment. To obtain information
about basic predictive characteristics of prefrontal EEG cordance
decrease, we pooled data from published and unpublished studies.
Methods: We used a categorization of decrease to predict response
and non-decrease to predict non-response. There was the total of
107 patients. Thirty-seven outpatients were treated in the USA and
70 inpatients were treated in the Czech Republic. Eighty-two out of
the 107 patients did not respond to a minimal 1 antidepressant trial.
Eighty-seven patients were treated with antidepressants and 20 with
rTMS monotherapy.
Results: In 40 out of the 47responders, a decrease of prefrontal cordance
was detected. There was no decrease of prefrontal cordance in
43 patients out of 60 non-responders. Computed sensitivity of prefrontal
cordance decrease was 0.851, (95% confident interval [CI]
0.720-0.929), specificity was 0.716; (95% CI 0.591-0.815), positive
predictive value was 0.701; (95% CI 0.572-0.805) and negative predictive
value was 0.860; (95% CI 0.735-0.933). The Number Need to
Diagnosis (NND) was 1.76; (95% CI 1.45-2.52). Conclusion: Pooled
analysis results support preliminary findings showing that QEEG
cordance might be a useful test in early prediction of antidepressive
response. The EEG cordance test are at leat comparable to results of
exercise ECG for coronary ischaemia. The prospective interventional
study is needed to evaluate real cost and benefit in daily clinical
practice. Supported by the grant MZCR NR9330