Comparison of manual data coding errors in two hospitals.

The routine manual encoding of pathological data, using the SNOP and SNOMED systems at two London teaching hospitals, was reviewed. The error rates in the two departments were compared and the causes analysed. The relative merits of SNOP and SNOMED were considered. Methods to optimise the efficiency of manual encoding are suggested and the importance of accuracy in coding is emphasised.     

Détermination de la tortuosité pour la détection des rétinopathies vasculaires sur les images d''angiographie rétinienne

Les rétinopathies vasculaires peuvent être accompagnées d'une modification de la forme des vaisseaux. Les patients atteints courent le risque de perdre la vue. Jusqu'ici, les méthodes utilisées par les ophthalmologistes pour analyser ces modifications sont qualitatives. Nous proposons une méthode nouvelle d'analyse de la tortuosité des vaisseaux de la rétine en vue d'élaborer par la suite un outil d'aide au diagnostic des rétinopathies vasculaires. La première étape de la méthode est une représentation symbolique de la structure vasculaire de la rétine. La deuxième étape appelée reconstitution 2D est la classification des entités vasculaires en gros vaisseaux (artères ou veines) et en petits vaisseaux (artérioles ou veinules) en utilisant la moyenne du diamèetre et du niveau de gris des segments de vaisseau comme paramètres discriminants. La dernière étape consiste à calculer l'excentricité des vaisseaux pour décrire leur tortuosité. Pour s'assurer de la capacité de l'excentricitéà décrire les petites modifications de forme, un second paramètre a été défini à titre comparatif. Les résultats obtenus permettent de quantifier les modifications de forme des vaisseaux.     

Activated ras oncogenes in human thyroid cancers

Human thyroid epithelial (follicular) cells give rise to two malignant tumors--"follicular" carcinomas, which metastasize almost exclusively via the bloodstream, and "papillary" carcinomas, which metastasize predominantly via lymphatics (Williams, E. D. In: W. Duncan (ed.), Recent Results in Cancer Research: Thyroid Cancer, pp. 47-55. Berlin: Springer-Verlag, 1980). We have investigated whether this contrast in biological behavior might be associated with different patterns of oncogene activation. DNA transfection analysis of five follicular and ten papillary cancers indeed showed a statistically significant difference in the pattern of genes responsible, activated ras oncogenes being found in 80% of follicular tumors but only 20% of papillary tumors. In addition, in follicular cancers we have found activation of all three ras oncogenes (H-ras, K-ras, and N-ras), the first time that this has been demonstrated in a primary human tumor type (as opposed to cell lines). We suggest therefore that ras activation may be an important determinant of metastatic capability in these epithelial cancers.     

Serum concentrations of oestradiol-17beta, progesterone, relaxin and chorionic gonadotrophin during blastocyst implantation in natural pregnancy cycle and in embryo transfer cycle in the rhesus monkey

The present study was undertaken to assess the temporal association between the profiles of serum concentrations of oestradiol-17beta, progesterone, chorionic gonadotrophin (CG) and relaxin in pregnancies established naturally, and after embryo transfer, as well as in failed pregnancies in rhesus monkeys. In naturally mated cycles (group 1) a conception rate of 75% was obtained. In group 1, the mean day of CG detection in serum was 11.5 +/- 1.9 day post-ovulation, and for relaxin, 9.0 +/- 2.5 day post-ovulation. In group 2, embryo transfer to synchronous, non-mated surrogate recipients was performed; seven embryo transfer cycles yielded three pregnancies which were allowed to continue to term and normal infants were delivered. In embryo transfer cycles the mean day of CG detection was 14.8 +/- 1.8 day post- ovulation, and for relaxin, 11.4 +/- 2.6 day post-ovulation. A delay of about 3 days was observed in the appearance in circulation of CG (P < 0.05) and also of relaxin (P < 0.05) between natural mated and embryo transfer conception cycles. Significant differences (P < 0.05 for progesterone and P < 0.03 for oestradiol) were obtained for the areas under the curves for progesterone and oestradiol between days 12 and 16 in conception cycles compared with failed pregnancies. These data provide the first observation of the normal hormonal signals associated with maternal recognition of transferred embryos during the peri- implantation period, and suggest that the use of such an experimental primate embryo transfer model may help to elucidate components of maternal and embryonic signal-response mechanisms during embryo implantation.