To compile a database on the clinical features of the diseases presenting as jaundice in the various countries, using a common proforma with common terminology and definitions.
To provide quality control of the large and disparate data by circulating all cases to all centres, allowing them to monitor their performances and resolve anomalies as soon as they arise.
To use the Community scientific communications networks for concurrent exchange of data, comment and analysis.
To determine the accuracy of diagnosis using the clinical data from adequate numbers (>100 cases) of each disease, and thus establish the baselines for technological aid in these diseases.
To encourage concurrent analysis of the data in all centres by such techniques (statistical, pattern-recognition, expert systems) as are available.
To make the benefits of the technology itself available to each participating centre from the outset, and to such other centres as are willing to accept it.
A database has been compiled of the clinical features of diseases presenting as jaundice in over 128 centres in 26 European countries. The database covers adequately at least 45 disease conditions. The agreement among 4 observers examining a case according to the protocol was measured on 111 doctors in 8 countries. It was consistently 84-89% regardless of language. Norms for the 12 common blood tests used in diagnosing jaundice were based on the results of 75000 tests. Normal ranges were obtained on 220 controls (11 centres) and laboratory variation on 6 standard samples tested in 22 centres. A problem in standards was discovered. Audit feedback on their cases in comparison to the database as a whole was provided to the centres, listing missing data, mix of diseases, symptom frequencies and the diagnostic accuracy of the program on their cases (a pan European audit), and summaries of their case. The use of the protocol and of feedback to the observers improved the diagnostic quality of the data by 12%, the diagnostic accuracy of the doctors by up to 50% and the omission rate of the data 3-fold. Diagnostic programs were prepared by Bayesian, pattern recognition, likelihood ratio, neural net and knowledge based techniques. The crude Bayesian version based on quality case achieved 72% accuracy. Circulated for field test it attained 50% while a trial algorithm reached 75% (106 cases reported by 12 centres to data) Neural net and hybrid knowledge approaches outperformed Bayes. The threshold for assessment of the value added by diagnostic technology was reproducibly set by computer diagnosis. The added value of each blood test was demonstrated. The residual scope for new technologies was also established, but may be affected by refinement of computer diagnosis and development of more user friendly interface.
During the course of the project reciprocal participation has occurred with other European projects in related fields.
The project was in effect a clinical trial of clinical data from a clinical and bioengineering point of view, reflecting practice on the ground. It is not an epidemiological or statistical exercise.
The over-riding considerations were the European dimension, the uniformity of case gathering, adequate numbers of cases and an effective diagnostic method.
The participating hospitals were self-chosen, unfunded and dependent on local conditions. They included university, large municipal, district and small hospitals, and medical, surgical, hepatology, and mainly, gastro-enterology units. All instruments (data-set, disease-set, proforma, brochure, video; data entry, edit and diagnostic programs; feedback and database were constructed in wide consultation, field-tested in all centres, amended, retested in the field and adopted.