Key points are not available for this paper at this time.
We investigated the quality of diagnostic coding in relation to cerebrovascular disease (CVD) in a computerized clinical information system by comparing the codes on the computer printout, obtained from casemix database on 166 patients with a diagnosis of cerebrovascular disease, with the codes allocated by the investigators after examining the case notes. Overall, the diagnostic coding was incorrect in 44 (26%) cases. These inaccuracies included 9 (5%) patients who did not have CVD and 35 (21%) patients who were incorrectly coded as "acute but ill-defined stroke". The inadequacy of information in discharge letters was the main reason for such discrepancies (21 cases). There is a need for adequate education and training of doctors and coding clerks and for a constant dialogue between them if the deficiencies in coding are to be rectified and the value of hospital information system in the audit, costing and health care planning is to be improved.
Hasan et al. (Fri,) studied this question.