Nursing workload measurement for urban hospitals in China

Objectives. The objective of this study was to develop a nursing workload measurement framework, which was relevant in deciding nurse staffing in hospitals in China. Design. This exploratory study was conducted as three phases. Phase One of the study used time and motion data to examine nursing workload in the two hospitals. Phase Two of the study used ICD-9 data from the same two hospitals to analyze the patient severity levels. Based on the findings from the first two phases, Phase Three of the study proposed a severity adjusted nursing workload measurement framework for urban hospitals in China. Population, sample, settings, years . Phase One used secondary data from a previous project. The data was collected in two urban hospitals in China from 1995 to 1996. Phase two data was collected by the investigator from the same two hospitals' patient record departments in 1999. Variables. Variables included in the Phase One of the study are hospital characteristics, hospital departments, patient demographics, patient's level of care, selected nurse demographics, and GRASP activity time measures. Variables in the Phase Two of the study are patient length of stay and severity of illness. Methods. For Phase One, the first step was to identify the range, mean and standard deviation of the time it takes nurse to perform each nursing activity and for nursing activity groups; and to examine task group performed time variation. Next, the degree to which nursing activities and nursing activity groups varied based on nurses' ages, patients' gender, patients' ages, and levels of care was identified. For Phase Two, first the overall average LOS of the ICD-9 codes was examined for the two hospitals. Second, the difference between the average LOS for each primary ICD-9 at the two hospitals was tested. Finally, Alemi's LOS adjusted severity of illness formula was used to examine the differences between severities of patients' in the two hospitals. Findings. Phase One data analysis was stopped after descriptive analysis was run because problems with the quality of the data were identified. The results of Phase Two showed there was a statistically significant difference (p < .05) in LOS among 62 percent of the matched primary ICD-9 codes in the two hospitals. Hospital A had a shorter mean LOS than Hospital B, excepting two codes. Alemi's LOS adjusted severity of illness also indicated among the same ICD-9 codes, the patients were sicker in Hospital A than in Hospital B. Conclusions. According to study results, patients are sicker in hospital A. Patients with higher severity ratings required more nursing care. Nurses were found to spend more time performing the same task on sicker patients. Based on this conclusion, nursing workload measurement should consider patients' severity when staffing to meet patient required care. It is not appropriate to have a single national nurse staffing standard.