Classification of substandard factors in perinatal care: development and multidisciplinary inter-rater agreement of the Groningen-system

Mariet Th. van Diem, Albertus Timmer, Sanne J. Gordijn, Klasien A. Bergman, Fleurisca J. Korteweg, Joke Ravise, Ellen Vreugdenhil, Jan Jaap H.M. Erwich

Perinatal audit is an established method for improving the quality of perinatal care. In audit meetings substandard factors (SSF) are identified in cases of perinatal mortality and morbidity. To our knowledge there is no classification system specifically designed for the classification of substandard factors. Such a classification may help to standardise allocation of substandard factors to categories. This will help to prioritise, guide and implement actions in quality improvement programs. A classification system of 284 substandard factors (SSF) identified in perinatal audit meetings between 2007 and 2011 was drawn up using the WHO Conceptual Framework for the International Classification for Patient Safety as a starting point. Discussions were held on inter-rater disagreements, inclusion of items, format and organisation and definitions of the main- and subcategories. A guideline was developed. An independent multidisciplinary group tested the classification. Independent of inter-rater agreement the allocations to categories were counted. For the counts in the subcategories one and two, we used the allocations in the main category as reference. The chance corrected agreement between classifiers was tested with Cohen's kappa statistic. The classification consists of 9 main categories with one or two subcategories. The main categories are (1) Equipment and Materials, (2) Medication, (3) Additional tests/ investigations, (4) Transportation , (5) Documentation, (6) Communication, (7) Medical practice, (8) Other and (9) non classifiable. Of 3663 allocations by 13 classifiers 1452 SSF's were allocated (40 %) to 'medical practice' and 1247 (34 %) to 'documentation'. 118 (3 %) times SSF's were not classifiable, mainly due to unclear phrasing of the SSF. The chance corrected agreement of 284 substandard factors in the main category was 0.68 (95 % CI 0.66-0.70) and 0.57 (95 % CI 0.54-0.59) for the CDG and the IGD respectively. Classifying substandard factors has given insight into problem area's in perinatal care and can give direction to medical, political and financial quality improvement measures. The Groningen-system has well defined categories and subcategories and the guidelines and examples are clear. The multidisciplinary inter-rater agreement is moderate to good. Improvement of the phrasing of the substandard factors is expected to improve inter-rater agreement.