Marga Hoogendoorn

84 The collection of data in the different hospitals was carried out over a period of 13 months. Because of this we think our results are not affected by seasonal influences and therefore representative for nursing workload all over the year. We included data from different hospitals with a digital nursing capacity-module and data from hospitals with manual registration of the Nursing Activities Score. No differences between the distribution of the Nursing Activities Score results of both types of hospital were found. There are also some limitations of this study. We included data from only eight Intensive Cares out of the 84 hospitals in the Netherlands. However, the included Intensive Cares were diverse in size (7 beds – 33 beds) and representative to Dutch Intensive Cares regarding hospital type (teaching, non-teaching) and geographical locations. We used data of 228 nurses in 226 different shifts, any future work might look at a (smaller) number of nurses over several shifts or comparing a number of nurses caring for the same patient over time for a clearer comparison. Another limitation is that registration burden of filling in Nursing Activities Score and NASA-Task Load Index might have influenced perceived workload. We measured the time for filling in the Nursing Activities Score for different nurses. The time for registration of Nursing Activities Score for a complex patient never reached more than 2 minutes. Nurses in hospitals using a digital registry needed about 1 minute per Nursing Activities Score registration. We do not expect a significant effect of this time on the perceived nursing workload. The time for recording the NASA-Task Load Index did not influence the Nursing Activities Score, because the timing of the NASA- Task Load Index was at the end of the shift after registration of Nursing Activities Score. Although our study is one of the larger studies in adult Intensive Cares comparing the NASA-Task Load Index with Nursing Activities Score, the number of observations is relatively low and this might cause lack of power to prove an association between the Nursing Activities Score and NASA-Task Load Index. In the study of Hoonakker et al. (2015) they measured the NASA-Task Load Index in 700 nurses in 7 hospitals in 17 Intensive Cares 10 . However, they analyzed only the association between NASA-Task Load Index and factors as kind of shift and nurse-patient ratio but not with patient factors. Therefore, further research with a larger study population is needed to confirm the generalizability of the results of our study. In our study we used the 6-scale NASA-Task Load Index. Recent research of Tubbs et al. showed that four of the six items (mental demand, physical demand, temporal demand, and effort) are strong and significant indicators for the overall nursing workload in Intensive Care nurses 16 . The study of Tubbs et al. was published after the data collection in our study, but we suggest that in a next study also the 4-scale version of the NASA-Task Load Index can be used for measuring nursing workload.