Impact of Different Sampling Strategies on Score Results of the Nine Equivalents of Nursing Manpower Use Score (NEMS)

Abstract

Prospective observational study to assess the impact of two different sampling strategies on the score results of the NEMS, used widely to estimate the amount of nursing workload in an ICU. NEMS scores of all patients admitted to the surgical ICU over a one-year period were automatically calculated twice a day with a patient data management system for each patient day on ICU using two different sampling strategies (NEMS(individual): 24-hour intervals starting from the time of admission; NEMS(8 a.m.): 24-hour intervals starting at 8 a.m.). NEMS(individual) and NEMS(8 a.m.) were collected on 3236 patient days; 687 patients were involved. Significantly lower scores were found for the NEMS(8 a.m.) (25.0 +/- 8.7) compared to the NEMS(individual) (26.1 +/- 8.9, p < 0.01); the interclass correlation coefficient (ICC) was good but not excellent: 0.78. The inter-rater correlation between the two NEMS scores was high or very high (kappa = 0.6-1.0) for six out of nine variables of the NEMS. Different sampling strategies produce different score values, especially due to the end of stay. This has to be taken into account when using the NEMS in quality assurance projects and multi-center studies.