Marga Hoogendoorn

43 Cares and different patient characteristics. Although all studies used the standardized NAS-scoring system, these differences still make reliable comparisons between the studies more difficult. Finally we evaluated the ability of the workload scoring systems to quantify the nursing capacity in daily practice, mostly translated into a Nurse:Patient-ratio. The Nurse:Patient- ratio is important because this is the translation of ‘abstract’ points into nursing capacity in daily practice, and can so be helpful for Intensive Care management. It enables to plan the needed numbers of nurses per shift and it enables a nurse to know at the beginning of his/her shift how many patients are under his/her responsibility. The Nurse:Patient- ratio was validated for only half of the systems (56%), of which only three systems used objective time-measurements (11%). Given the fact that the ultimate aim of a scoring system should be supporting the planning of nursing capacity, it is disappointing that the accuracy of translating the scores of a system into needed nursing time was only assessed to a limited extent and even then often in an inadequate way. Comparison with time- measurements is only described for the TISS, the PINI and the NWL Patient Category Scoring System. In all three studies the categorization to a Nurse:Patient-ratio led to an overestimation of needed time 34, 42, 46 . Comparison among different scoring systems also gave disappointing results with large differences in the reported Nurse:Patient-ratio, with examples of doubling the needed nursing staff 69 . In a number of articles, the calculated Nurse:Patient-ratio from a scoring system was compared to the actual available or planned number of nurses. However, a description of how the actual or planned number of nurses was determined was lacking. This information is crucial to interpret the results of the comparisons made. Without a validation by time-measurements it is impossible to assess the accuracy of both the actual planned staff as well as the planned staff according to a scoring system. Implications for research Regarding the validation of the systems, the low number of systems that were validated with the gold standard, i.e. time-measurements, is striking. The implications of the absence of the gold standard becomes clear when interpreting the results of the second- best method for validation; comparing two different systems often show large variation 22, 33, 48, 57, 58 . In these cases it is hard to tell which scoring system agrees with reality, due to the absence of time-measurements. Studies in which the systems-based nursing capacity was compared with the actual nursing staff show the same weakness. A higher indication for needed nursing staff by a system compared to the actual present staff would suggest that the workload of the nurses is too high and should be lowered. However, without information on the accuracy of both the actual planned and system-based calculated