Marga Hoogendoorn

41 D I SCUSS I ON This review shows that over the years classification of nursing workload on an Intensive Care has been a topic of continuous attention. Our aimwas to identify the existing scoring systems in literature. A high number of scoring systems has been developed and used for planning of care since the first system was published in 1974. In total we included 27 different systems for measuring nursing workload on an Intensive Care in this review. Remarkable is the continuous use of this first developed system, the TISS. Although developed in 1974, the TISS is still used in daily Intensive Care practice as well as for development and validation of other scoring systems. Table 1 shows an increasing number of new systems between 1980 and 2000. The continuous use of those systems since their development shows that quantification of nursing care is still actual and considered important. The next important part of this reviewwas the evaluation of the validity and the reliability of the scoring systems. Although we found many articles about validation and reliability of the different systems, none of the finally included 27 systems that claimed to quantify needed nursing staff satisfied all our pre-set validity and reliability criteria. Only a few satisfied a majority of our pre-set criteria. The content validity of almost all 27 included systems was good; most systems were developed by nurses or a multidisciplinary team of nurses and physicians. Only the items of the TISS concerned mainly medical interventions exclusively selected by physicians, which can be explained by the fact that the original aim of this system was to classify severity of illness and not nursing workload 15 . It is therefore remarkable that the TISS has become one of the most commonly used scoring systems to measure nursing workload. Moreover the TISS itself or items of the TISS have also been used in the development of six other systems for measuring nursing workload, namely the NISS, TISS-28, CritScore, NEMS, NAS, and the NWL Patient Category Scoring System. The inter- and/or intra-rater reliability was tested in less than half of the systems (44%). If described, the results were generally moderate to good. Only the results on the inter-rater reliability of the NAS evaluated in several different studies showed a large variability with weak to good results 45 . In particular the inter-rater reliability of nursing activities which included an estimate of the duration of that activity, such as monitoring and titration, hygiene procedures, support and care of patient and relatives, administrative tasks and mobilization, showed a large variability. For example, the inter-rater reliability of the item “Mobilization and positioning” resulted in an agreement of 49% (Kappa 0.16) if rated by a nurse and a manager. If rated by a nurse and a physician, the agreement was 39,6% (Kappa 0.020) 45 . On the other hand, a medical intervention like oxygen showed a 100%