Nursing allocation in isolation wards of COVID-19 designated hospitals: a nationwide study in China

In this study, 132 COVID-19 designated hospitals from 31 provincial administrative districts in China were selected using stratified sampling. The head nurses and nurses in the isolated wards of these hospitals were asked to complete a self-reported nursing human resource allocation questionnaire that includes three parts: (1) informed consent form; (2) general data of hospitals and participants; (3) evaluation of the status quo and its effect on the isolated nursing workforce allocation. The inclusion criteria for the participants were as follows: (1) being the head nurse or a nurse in the isolation ward of a COVID-19 designated hospital, and being engaged in caring for COVID-19 patients for more than 1 month; (2) having volunteered to participate in this study.

From April 11 to July 31, 2020, a retrospective cross-sectional online survey to explore frontline nursing staff allocation in the isolation wards of COVID-19 designated hospitals was conducted. First, participants provided informed consent on the survey platform before they could proceed to the survey items. Next, the questionnaire was sent to each participant via a mobile phone text message or Wechat, and the respondent could only complete the questionnaire once to avoid repeated submissions. One week after the questionnaire was sent, the researcher verified the number and quality of completed questionnaires. In case of any ambiguities, the researcher could contact the participants by telephone to verify the accuracy of the data.

In this study, a total of 229 head nurses and 1378 nurses from 117 COVID-19 designated hospitals were surveyed. Among head nurses, 223 (97.4%) were women with a median age of 39 years, 85.2% had a bachelor’s degree, 55.5% had a professional title of ‘nurse supervisor’, and the median length of employment was 20 years. Among frontline nurses, 1285 (93.3%) were women with an average age of 31.48 years, 76.0% had a bachelor’s degree, and 76.0% had a professional title of ‘nurse’. More details are presented in Table 1.

From Table 2, it can be seen that: (1) the average bed-to-nurse ratio of different types of isolation wards was different (Z = 36.742, P = 0.000). Multiple comparisons show that there was no statistically significant difference in bed-to-nurse ratio between the wards of suspected cases and confirmed cases (1:0.56 vs 1:0.45, P = 0.077). However, differences in other pairwise comparisons were statistically significant, among which the bed-to-nurse ratio in the ICU ward was the highest, with an average of 1:1.88; (2) work hours of nurses per shift in different types of isolation wards were different (Z = 8.468, P = 0.014). Multiple comparisons showed that there was no significant difference in the average work hours per shift between the suspected case wards and confirmed case wards (P = 0.433), but differences in other pairwise comparisons were all statistically significant. The average work hours per shift in the severe ICU ward was 5 h, while in the suspected case and confirmed case wards were both 6 h; (3) in terms of nursing staff structure, there was a significant difference in the ratio of nurses with junior college degrees in different wards (Z = 8.693, P = 0.013). Multiple comparisons showed that there was no significant difference in the proportion of nurses with junior college degrees between the suspected case and confirmed case wards (40% vs 38%, P = 0.256), but differences in other pairwise comparisons were all statistically significant. The ratio of nurses with junior college degrees in the severe ICU ward was 20%, and the proportion of nurses with bachelor’s degrees was as high as 73%.

We found that the overtime work of nurses was proportional to the average work time per shift (γ_s = 0.146, p = 0.028) and inversely proportional to the weekly rest time (γ_s = −0.195, p = 0.003). The utilization of nursing human resources was significantly positively correlated with the proportion of senior nurses (γ_s = 0.136, p = 0.039). Physical fatigue of nurses was positively correlated with the ratio of nurses with bachelor-level and above degrees (γ_s = 0.133, p = 0.044), but was significantly negatively correlated with the weekly rest time (γ_s = − 0.155, p = 0.019). The ratio of nurses with junior college degrees was significantly positively correlated with teamwork (γ_s = 0.142, p = 0.032), but significantly negatively correlated with the satisfaction of head nurses (γ_s = 0.137, p = 0.040). The proportion of nurses who have been working for more than 10 years were significantly negatively correlated with both teamwork and psychological problems (γ_s = − 0.135, p = 0.041; γ_s = − 0.203, p = 0.002). Additional details regarding this analysis are presented in Tables 3 and 4.

First, we performed a multiple stepwise regression analysis to identify potential predictors of first-line nurses’ satisfaction, yielding the following model equation: Satisfaction = 4.703 - (0.051 × nurse work hours per shift) (see Table 5 for details). This model demonstrated that only the nurse work hours per shift was predictive of nurses’ satisfaction, accounting for 5.2% of the variance in these values. Next, the average length of stay (LOS) of patients was considered as another evaluation index, and we found that the proportion of nurses and the work hours of nurses per shift were both independent influencing factors, and the regression equation was formulated as follows: LOS = 19.581 - (6.72 × nurse ratio) - (0.632 × nurse work hours per shift) (Table 6).

In this study, we found that the bed-to-nurse ratios of the suspected case and confirmed case wards were approximately 1:0.56 and 1:0.45, respectively, and that for the confirmed severe case ward (ICU) was approximately 1:1.88. These results were consistent with those of Cao et al. [11], i.e., the bed-to-nurse ratio of the suspected case and confirmed case wards were both 1: 0.45, while the bed-to-nurse ratio was lower than the reported ratio of 1: ≤ 2.5 in the ICU. Wang et al. [12] refined the allocation of nursing workforce per shift, pointing out that the nurse-to-patient ratio per shift was 1:5 for the suspected cases ward, 1: 3 for the confirmed cases ward, and 1:1 for the ICU. Besides, it was identified that the nurse-to-patient ratio should be 1:1 per shift for patients who have prone position ventilation, invasive mechanical ventilation, or hemodynamic instability. This is consistent worldwide [8]. Therefore, if we encounter a pandemic similar to COVID-19 again, it is suggested that designated hospitals refer to this practical experience for the preliminary allocation of the nursing workforce, and thereafter combine the characteristics of each hospital with the number of patients for a dynamic allocation of the nursing workforce. In summary, the allocation of nursing human resources is the basis for the prevention and control of public health emergencies such as the COVID-19 pandemic. In the future, real-time nurse-to-patient ratios of clinical decision-making systems can be used to appropriately allocate the workforce. Also, based on the preset nurse-to-patient ratio, combined with the number of real-time patients, the reasonable number of nurses and the number of surplus nurses in each ward can be calculated, which can help in the dynamic allocation of the nursing workforce [13].

This study found that the structure of nurses in different isolation wards was consistent in the distribution of professional titles and working years, i.e., the professional titles of nurses were dominated by the title of ‘senior nurse’ which accounted for an average of 43 to 50% of all nurse titles. The proportion of nurses working for 0–2 years was less than 10%, while the average distribution was in the medium or many years of service, with 3–5 years of service accounting for about 30%, and ≥ 10 years of service accounting for 5–9%, respectively. The differences were mainly in the educational background of nursing staff. The proportion of nurses with junior college degrees in different wards was different (Z = 8.693, P = 0.013). The proportion of nurses with junior college degrees in the suspected case and confirmed case wards were similar (40% vs 38%, P = 0.256), while in the ICU ward it was only 20%, and the proportion of nurses with bachelor’s degrees was as high as 73%. This indicates that nurses must have strong professional nursing and isolation protection skills, especially nurses in the ICU, since they are in close contact with patients and provide care for patients with serious conditions or high failure-to-rescue rates. Thus, nursing care during the COVID-19 pandemic is beyond the standard of regular nursing care and has higher requirements for professional literacy [14]. At present, there is a lack of research on nurse structural configurations in various forefronts of the COVID-19 pandemic. The results of this study can provide a reference for hospitals globally on the response to the COVID-19 pandemic and to optimize the mix of skills among nurses.

A good nurse rostering should not only ensure the quality of clinical nursing but also meet the resting needs of nurses, to ensure the safety and continuity of nursing care and maximize the efficiency of nursing workforce allocation and nursing service quality. The results of this study showed that nurses in the suspected case and confirmed case wards worked an average of 6 h per shift, which was higher than that of 109 COVID-19 first-line nurses who worked 5 h per shift, while their desired work hours per shift was 4 h [10]. The average duration of shift for nurses in the ICU ward of confirmed cases in this study was 5 h per shift, which was higher than the 4 h recommended by the National Health Commission of the People’s Republic of China [15]. These numbers illustrate that on the one hand, there is a lack of nurses, while on the other hand, it may be due to the shortage of protective equipment, such as masks and isolation gowns. Hospitals have to extend nurses’ work hours per shift to reduce the use of new protective equipment. However, excessively long work hours will not only lead to increased fatigue and psychological pressures for nurses but may also affect the quality of clinical nursing care. In addition, the results of regression analyses in this study showed that the work hours per shift was the main factor influencing nurses’ satisfaction and the average hospitalization days of patients. Thus, it can be seen that nurse rostering is one of the important indicators of scientific and reasonable allocation of the nursing workforce. Some studies recommend that direct nursing care should be 4 h per shift, while indirect nursing care could be 6 to 8 h per shift [12, 16]. Other hospitals have suggested that the general isolation ward could adopt an 8-h shift and a 4-h shift, inside and outside the isolation area, respectively. Each shift could have a resting time of 2 or 3 days per week. Besides, the ICU could adopt the “4 × 6-1” model of nurse rostering, i.e., each shift would be 4 h a day, for 6 days in a row, with 1 day off [11]. Therefore, nurse managers may attempt to change the general nursing work mode, according to the type and content of the assignment, adjust the work hours of each shift, and implement flexible duty shifts.

We found that the overall nursing workforce allocation in the isolation wards of COVID-19 designated hospitals in China is better than stipulated above. For example, the satisfaction of nurses is high, the utilization rate of nursing human resources is sufficient, and the team cooperation is good. However, the problem of long work hours for nurses does exist. In particular, 27.7% of hospitals had nurses working overtime, with nurses working more than 60% of their usual work hours. Correlation analysis identified that the overtime work of nurses was proportional to the average work time per shift (γ_s = 0.146) and inversely proportional to the weekly rest time (γ_s = − 0.195), which indicates that for longer work hours of nurses, the occurrence of overtime work is more serious. The reason could be heavy workloads or the lack of nursing staff. Therefore, to allocate the nursing workforce effectively, it is necessary to accurately consider the needs of patients, and the number of nurses, and their physical and mental needs, to ensure that nursing care can be conducted continuously and effectively.

We also found that the overall rate of nursing human resource utilization was proportional to the proportion of senior nurses in the ward (r_s = 0.136), implying that the greater the proportion of senior nurses in the ward, the higher the utilization rate of nursing human resources, which further illustrates that senior nurses play an important role during the COVID-19 pandemic. In addition, the ratio of nurses with junior college degrees was significantly positively correlated with teamwork (r_s = 0.142), which indicates that nurses with junior college degrees were weak in teamwork and communication skills. The ratio of nurses who have worked for more than 10 years was inversely proportional to teamwork and nurses’ psychological problems (r_s = 0.135, r_s = 0.203), which indicated that nurses with more work experience had stronger abilities for team cooperation, communication skills, and coping skills under pressure. But research conducted by Wang et al. [17] pointed out that the proportion of nurses with work experience of ≤5 years was a protective factor for preventing adverse events. Therefore, the proportion of nurses at all levels should be taken into account in the allocation of the nursing workforce in response to public health emergencies in the future.