Management of early mobilization in intensive care units: a multicenter cross-sectional study

The Management of Early Mobilization in Chinese Intensive Care Units study, a large multicenter research of Chinese ICUs, represents a meaningful step toward collaborative Chinese intensive care research on EM. An annual committee of nurses and nursing managers from various Chinese ICUs, formed in May 2017, invited nursing managers representing ICUs in their hospitals to participate through telephone contact, emails, and express delivery. The participants were encouraged to include as many as possible to minimize selection bias. Participation in the research was voluntary and unfunded.

Participating ICUs had at least six beds in each, they could be medical, surgical, or mixed, except for pediatric ICUs, according to “Hospital grade management standards defined by National Health and Family Planning Commission” in Jiangsu, China. We enrolled all nurses of recruited ICUs in their centers to participate through representing nursing managers. There was no sample size calculation in this survey.

In this research, EM was defined as any activity beyond range of motion performed by a care provider (nursing or physical therapist) for ICU patients, including active and passive, in-bed and out-bed exercises (positioning, bed mobility, sitting on side of bed, transfer to chair, walking, etc.).^{9, 15, 16} Our survey was developed through an iterative process involving the framework in the study by Cabana et al.,²¹ literature review, and critical care nurses.^{24, 25} A preliminary survey was carried out in several ICU nurses with critical care experience in a university hospital to access for feasibility. Modifications based on the nurses’ suggestions resulted in the final version of the survey. The survey was divided into two parts.

Part 1 addressed organizational characteristics, including type of hospital (government university and non-university), specialty of ICUs (medical, surgical, or mixed), number of beds, bed-to-nurse ratio in ICU, number of intensive nurses, nurse age, sex, education, nursing, and critical care experience.

Part 2 of the survey included process parameters as follows: EM implementation, especially for patients on mechanical ventilation (MV) or continuous renal replacement therapy (CRRT) or extracorporeal membrane oxygenation (ECMO). The part also recorded perceived barriers and knowledge regarding current practice in the ICUs.

This survey was conducted between July 1 to July 14, 2017. Data were recorded by one researcher and then independently checked by two other researchers. We checked the recorded data for statistical outliers and contacted the representing nursing managers for interpretation.

Categorical variables are given as number (percentage for implementation of EM), normally distributed numerical variables as mean (standard deviation), and other numerical variables as median (interquartile range). Categorical variables were analyzed with Chi-square test or Fisher’s exact test.

In assessing implementation of EM, nurses who practiced all or one of the EM protocols were deemed implementation.^{9, 15, 16} To identify the independent factors associated with implementation, we compared binary variable logistic regression analysis using models that included the implementation of EM protocols and all the variables collected for structure and process characteristics. Model fit was assessed with the Hosmer–Lemeshow goodness of fit test. SPSS 18.0 (SPSS, Inc, Chicago, IL, USA) was used for statistical analysis. A P-value of <0.05 is considered to be significant.

Most of the participating hospitals were university hospitals (85%), which had 11–20 ICU beds (46.2%) and bed-to-nurse ratio in ICU between 1 and 2 (55.4%). Overall, 74.2% (1,004/1,353) of patients reported EM practice, with 74.6% (894/1,199) of the university hospitals and 71.4% (110/154) of the non-university hospitals, respectively. The difference was not significant, although university hospitals were associated with a higher implementation of EM (each with a P value of ˃0.05). The five critical attending nursing factors that were related to the implementation of EM are as follows: nurses’ age less than 25 years, male gender, technical secondary education, 5–10 years of nursing experience, and critical attending experience less than 5 years, which reported higher EM rates in ICUs (each with a P value of ˂0.05). We identified six patient-specific treatments as follows: non-invasive mechanical ventilation (NIPPV), intubation, tracheotomy, CRRT-jugular vein catheter, CRRT-femoral vein catheter, and ECMO, which were associated with EM implementation in ICU. EM was less likely to be practiced in ECMO patients (P=0.000) (Table 1).

For the EM level, 98.1% (1,327) reported use of in-bed exercise, 5.7% (77) sitting on side of bed, 21.7% (294) transfer to chair, and 2.4% (33) walking. In university hospitals, 99.4% reported bed mobility training, 6.0% sitting on side of bed, 23.3% transfer to chair, 2.7% walking, whereas 94.1%, 3.3%, 11.2%, 1.3%, respectively, in non-university hospitals (Figure 1). For patients on MV or CRRT or ECMO, 72.5% of the nurses reported use of bed mobility training, 8.9% sitting on side of bed, 13.2% no experience, whereas transfer to chair and walking only 3.4% and 2.0%, respectively (Figure 2).

On logistic regression analysis, with models that included all variables in Table 1 and Figure 3, factors independently associated with EM implementation included nurses’ educational backgrounds (OR, 1.39; 95% CI, 1.06–1.82; P˂0.05), nursing experience (OR, 0.91; 95% CI, 0.85–0.99; P˂0.05), the lack of nursing resources (OR, 2.32; 95% CI, 1.20–4.49; P˂0.05), absence of physician (OR, 5.76; 95% CI, 1.11–29.85; P˂0.05), and the weakness of patients (OR, 0.06; 95% CI, 0.02–0.27; P˂0.01) (Table 2). There was no interaction between these targets, and the logistic regression models fit well.

The most frequently reported barriers to EM were unplanned extubation and lack of nursing resources (Figure 3). Most nurses indicated knowledge regarding EM; however, the agreement to transfer to chair and walking was only 69.6% (940) and 63.9% (863), respectively.

Figure 1

Figure 2

Figure 3

In the Chinese centers studied, the implementation rate for EM was 74.2%. Implementation rate was considerably higher, but still incomplete, for which the most reported reasons were bed mobility training and positioning. Both nurses and nursing managers identified unplanned extubation, lack of nursing resources, and absence of physical therapist as most important barriers to EM. Nursing knowledge regarding EM in ICUs still lacks.

Overall implementation of EM (74.2%) was higher in this study than in other multicenter research studies.²⁶ Recent implementation rates for EM were 45% in USA,²⁷ 71% in Canada,²⁸ and 54% in Australia.¹⁴ In this study, 74.6% of the university hospitals and 71.4% of the non-university hospitals reported EM practice (P >0.05).

As for the EM level, 98.1% reported use of bed mobility training, 5.7% sitting on side of bed, 21.7% transfer to chair, and 2.4% walking. We further analyzed organizational factors and six patient-specific treatments with the EM level. For patients on MV or CRRT or ECMO, transferring to chair and walking were only 3.4% and 2.0%, respectively. Similarly, a study of 116 German ICUs identified that only 6% of the patients with endotracheal tube were sitting and only 0.2% had stood out of bed or walked.¹⁰ The study of 38 ICUs in Australia and New Zealand had no patients who stood, sat out of bed, or started walking during MV.²⁹ Dean and Ross³⁰ found that even high-intensity exercises done in bed do not counteract the adverse effects of bed rest, recommending that upright positioning should be included in exercise plan of care. Taken together, these data showed that under the circumstances, the mainly achieved level of EM was in-bed exercise and more should be done to improve implementation of EM in ICUs.

In this study, it was found that hospital rank, ICU beds, and bed-to-nurse ratio in ICU are not associated with implementation rates for EM (each with a P value of >0.05). First, this could be because the low level of EM practice, especially for patients on MV or CRRT or ECMO, within 72.5% reported use of in-bed exercise. Second, in this study, the average bed-to-nurse ratio in ICUs is close to 1:2, which is in accordance with the status of nursing resource distribution in Chinese ICUs, approaching to the guidelines (1:2.5–3) for the evaluation of hospital management in the National Health and Family Planning Commission 2008.³¹ In our logistic regression analysis, the independent predictors of EM implementation were nurses’ educational backgrounds, years of nursing experience, the lack of nursing resources, absence of physician, and the weakness of patients (Table 2). Similarly, a recent systematic review revealed that factors associated with mobility practice included restrictive parameters, personnel resources, attitude, and perceptions.³²

EM for critically ill patients is becoming an established, evidence-based practice in the ICU.^{8, 9, 10, 11, 12, 13, 33, 34, 35, 36, 37} Feasibility, safety, and efficiency are confirmed.³⁷ However, implementing guideline into practice remains challenging. What about the whole standard and progressing EM? And how to practice?

Jason et al³⁸ suggested that barriers to adherence to the clinical practice guideline could be categorized into three areas as follows: knowledge, attitude, and behavior. We found that nurses aged less than 25 years, male, with technical secondary education, 5–10 years of nursing experience, and less than 5 years of nursing and critical care experience were more likely to practice EM (each with a P value of <0.05). Most nurses indicated knowledge regarding EM; however, only 63.9% reported agreement to walking and 69.6% transferring to chair. For patient on MV or CRRT or ECMO, the out-of-bed exercise (walking and transferring to chair) accounted for 5.4%. Similar research studies show that nursing knowledge and skill in achieving EM are essential.⁹ Implementation of a program^{39, 40, 41, 42, 43, 44, 45, 46, 47} and education of staff^{39, 46} are important factors in EM for critically ill patients in ICU. Doherty et al⁴⁸ and Garzo et al⁴⁹ suggested that better definitions of care roles may allow for more targeted surveys of attitudes and behaviors that may improve EM practice.

Safety and quality of care is an eternal topic. Most nurses reported their knowledge regarding safety concern. Our nursing respondents indicated unplanned extubation as the first barrier to EM and lack of nursing resources as the second barrier, whereas nursing managers reported lack of nursing resources as the primary barrier and unplanned extubation as the secondary barrier. This could be because of leadership’s systematic view of management. Nevertheless, both nurses and nursing managers identified unplanned extubation and lack of nursing resources as primary barriers to EM. To our relief, safety is confirmed.^{16, 37, 45, 50, 51, 52} On the basis of security awareness, EM will be a better practice. In addition, EM practice is affected by the availability of resources, such as nursing staffing resources, inadequate teamwork,^{53, 54, 55} and protocols.^{56, 57} Similar research studies found that the time and personnel necessary to mobility could be a deterrent to EM.^{1, 28, 30, 58} Some research studies have attempted to overcome this difficulty by prioritizing daily care routines to EM, simplified guidelines,^{58, 59, 60} and creating mobility teams.¹⁴ Bakhru et al²⁷ separated barriers into institutional barriers, patient barriers, and health-care providers barriers, suggesting a better understanding of potential benefits of EM and concerns about patients’ need of education and guidance on patient mobility techniques.

The IOEMCICUS study, a large multicenter study of Chinese ICUs, represents a meaningful step toward collaborative Chinese intensive care research on EM. To make the study more representative, we included a large number of participating centers, though because of a lack of funding this resulted in a need for compromise. This study is subject to selection bias and might not fully reflect intensive care through China. ICUs with less than six beds were excluded, many of the Jiangsu, Chinese Least Developed Cities did not participate, and university hospitals accounted for 85% of the centers. The latter two points stand out when one considers the fact that all participating units could perform positioning, bed mobility training, sitting on side of bed, transfer to chair, and walking, part of which are relative luxury in many resource poor areas in Jiangsu, China. Importantly, the number of nurses in the study sample was not proportionate to the number of nurses of China taking part in the study. About 89% of the nurses were enrolled in 56 university hospitals and only 11% in nine non-university hospitals. While some of these university hospitals have good resources, others (such as secondary hospitals) have relatively poor conditions, and results from the hospitals would have a disproportionate effect on overall implementation rates. While section bias is likely, we note that the difference in EM implementation rates was not significant between university and non-university hospitals in this study.

A current definition of EM refers to the application of physical activity within the first 2–5 days of critical illness or injury;²⁶ however, in our research, EM may be beyond the “window”. The net effect of “the window” bias could be an overestimate of EM implementation. Other limitations include potential interactions that may exist between survey questions that may introduce response bias. Finally, although we did not systematically check data accuracy, we did design the data entry forms to disallow missing fields and contact ICU representatives for clarification of any outlying data.

In summary, taking into account selection bias including the cross-sectional design, the under-representation of the Chinese poor-equipment hospitals, and the over-representation of university hospitals, the net effect of the “window” bias, true implementation rates for EM, is likely to be lower than our data. We suggest that given the conditions in China, the most appropriate strategy might be to focus on organizational structure (mobility team), process structure (protocols, risk management, education, checking feedback, and improvement), and outcome structure (the efficiency of intervention program, functional outcome of patients, and the costs and benefits).