ORIGINAL ARTICLE Postoperative cognitive dysfunction: computerized and conventional tests showed only moderate inter-rater reliability Finn M. Radtke • Martin Franck • Norbert Papkalla • Tim S. Herbig • Edith Weiss-Gerlach • Robin Kleinwaechter • Klaus D. Wernecke • Claudia D. Spies Received: 9 December 2009 / Accepted: 7 April 2010 / Published online: 28 May 2010 Ó Japanese Society of Anesthesiologists 2010 Abstract Purpose The incidence of postoperative cognitive dys- function (POCD) depends on the test battery and calcula- tion method used. The measurements may be performed with a paper and pencil test battery or with a computerized test battery. The objective of this study was to measure the incidence and congruence of POCD by comparing a computerized test battery with a paper and pencil test battery in the same patient population. Methods In total, 67 patients were included: 30 consec- utive in-patients undergoing surgery under general anes- thesia and 37 non-surgical out-patients as control. Patients were tested with a paper and pencil test battery and a computerized test battery on inclusion in the study and 7 days later. Both test batteries covered the cognitive domains: visual attention, visual learning, memory, and speed of processing. Results The computerized test battery classified 10.0% (95% CI 3.5–25.6%) of the patients as suffering from POCD whereas the paper and pencil test battery classified 30.0% (95% CI 16.7–47.9%) as suffering (95% CI for difference 3.9–36.5%, P = 0.03). The inter-rater reliability between both test batteries showed moderate agreement (Cohen’s kappa of 0.41). All patients identified by use of the computerized test battery were also identified with the paper and pencil test battery. The paper and pencil test battery identified 6 additional cases. Conclusion In our study we demonstrated that the inci- dence of POCD measured with computerized test battery and paper and pencil test battery showed moderate inter- rater reliability. Use of neuropsychological test batteries theoretically covering the same cognitive domains does not automatically lead to the same classification of POCD. Keywords Cognitive deficit  Complication  Recovery  Neurological  Postoperative Introduction Postoperative cognitive dysfunction (POCD) is associated with higher mortality [1]. Symptoms may range from a mild cognitive deterioration that might not even be rec- ognized by the patients themselves [2] to longer-lasting more apparent dysfunction of memory, information pro- cessing, and perception. Subjective self-reported cognitive symptoms do not substitute for objective cognitive testing, because they correlate poorly [3]. Large studies have reported an incidence of approxi- mately 25% 1 week after surgery and approximately 10% after 3 months [2, 3]. It may occur in all age groups but elderly patients seem to be at increased risk [2, 4]. For measurement of POCD different aspects of cogni- tion (advertence, concentration, verbal abilities, learning, and memory etc.) need to be assessed. Specific require- ments for test selection have been recommended [5] in order to introduce uniformity into the assessment and definition of POCD. F. M. Radtke and M. Franck contributed equally. F. M. Radtke  M. Franck  N. Papkalla  T. S. Herbig  E. Weiss-Gerlach  R. Kleinwaechter  C. D. Spies (&) Department of Anesthesiology and Surgical Critical Care Medicine, Campus Charite ´ Mitte and Campus Virchow-Klinikum, Charite ´, Universita ¨tsmedizin, Charite ´platz 1, 10117 Berlin, Germany e-mail: claudia.spies@charite.de K. D. Wernecke SOSTANA, Wildensteiner Straße 27, 10318 Berlin, Germany 123 J Anesth (2010) 24:518–525 DOI 10.1007/s00540-010-0952-z