Research strategy and selection of studies

The research strategy obeyed the key concepts defined by the following PICO model(18):

The research was carried out, on January 20, 2022, in the databases PubMed (MEDLINE), Cumulative Index of Nursing and Allied Health (CINAHL), Cochrane Central Register of Controlled Trials (CENTRAL), Physiotherapy Evidence Database (PEDro), Clinical Trials.gov, WHO International Clinical Trials Registry Platform and EU Clinical Trials Register.

The terms “early ambulation”, “walking” or “mobility” were searched, combined with “intensive care units”, “critical care” or “critical illness”, according to the strategy presented in table 1.

In order to gather as much information as possible, this review, in addition to including randomized studies, also included quasi-experimental, case-control and cohort studies (including quality improvement programs with before and after design) published in English, Spanish and Portuguese. An analysis period of 15 years prior to the research was defined, including studies published between 2004 and 2019, in order to include articles that may have been the basis of current recommendations for early mobilization.

Clinical cases, systematic reviews, review articles and any studies that analyzed the effect of mobilization programs in conjunction with other interventions (e.g. ABCDE bundle) were excluded. After eliminating duplicate articles through the Mendeley program, two independent reviewers used the Rayyan application to analyze the title and abstract of the articles. The same reviewers subsequently analyzed the full text of articles that passed the first review phase to decide whether they met the defined criteria. The data relating to the studies to be included in the review were extracted into an Excel that also served as the basis for the risk of bias analysis of each one of them. Disagreements were resolved by discussion between reviewers.

Characteristics of the studies

Of the 18 studies included in the review (20-37) (Table 1), 3 are randomized controlled trials (RCTs) and 15 are observational studies. Many of these observational studies are related to quality improvement programs and use a design that assesses patients before and after implementation of the early mobilization program. It should be noted that 15 of the 18 studies included in the review result from data collected in the United States of America (USA) and that only one includes data collected in Europe.

Quality and risk of bias

The evaluation of the methodological quality and the risk of bias of the studies included in the review was independently performed by two researchers, using the criteria of the Newcastle-Ottawa Scale (NOS - Newcastle-Ottawa Scale)³⁸⁾ for cohort studies (Table 3) and Risk of Bias 2 (RoB 2) ³⁹ for randomized trials (Figure 2). Disagreement cases were resolved by discussion between the two evaluators.

Figure 2 Evaluation of Risk of Bias 2 (RoB 2) 

Table 4 Comparison of the results of early progressive mobilization programs 

Mortality

Of the studies included in this review, 12 analyzed the topic mortality. Of these, five ones refer to ICU mortality^22-24,28,37, six to in-hospital mortality^{20,27,31,33,35,36} and one presents 30-day mortality results²⁹.

Analyzing Table 4, it can be seen that mortality in the ICU varied between 0 and 17.2%, with no significant differences being verified between the group of patients undergoing the mobilization program and the control group. Still, of the five studies that analyzed mortality in the ICU, four reported slightly higher mortality rates in the mobilization group.

With regard to hospital mortality, it varied between 0 and 25%. Only in an analysis of the results of a quality improvement project through the implementation of a progressive mobilization protocol ³¹, a significant reduction in hospital mortality was identified in patients submitted to the early mobilization program (11 vs 24%, p < 0 ,01). In three studies^21,27,35 hospital mortality was higher in the mobilization group and in another 3 it was higher in the control group^31,33,36.

With regard to 30-day mortality²⁹, the mobilization group was 4.5% lower than the control group, although this difference cannot be considered significant (p = 0.12).

In none of the three RCTs under analysis ^28,35,36 there are significant differences in mortality between groups.

Ventilation time

There are 12 articles that compare mean ventilation times between patients in an early mobilization program and usual treatment (Table 4). The mean ventilation times reported ranged between 3³¹ and 30.9³⁴ days, with no significant increase in ventilation time associated with the development of early progressive mobilization programs being verified in any of the studies analyzed. On the other hand, in 10 of the analyzed studies, there was a reduction in ventilation time in the early progressive mobilization group, in relation to the control group, with statistically significant difference in four of them^30,31,34,36.

The two randomized studies that analyzed this outcome^28,36 both show a reduction in ventilation time in the early mobilization group, but only in one³⁶ was this difference statistically significant.

ICU length of stay

The ICU length of stay presented by the different studies is quite heterogeneous, varying between 3.97²⁴ and 24.9³⁴ days.

In 15^{20,22,23,26-37} of the 18 studies analyzed (Table 4) there was a reduction in the length of stay in the group of patients included in the mobilization program, in relation to the control group. Of these, in seven studies this difference is statistically significant^29-31,34-37. On the other hand, in three studies^21,24,25) the group of patients included in the mobilization program had a longer hospital stay than the control group, although only in one²⁵ were these differences statistically significant.

In all three RCTs under analysis^28,35,36 there was a reduction in the length of stay in the early mobilization group, compared to the control group, with this difference being statistically significant in one of the studies³⁵.

Readmission in ICU

The definition of readmission varies in the different studies analyzed, one of them refers to hospital readmission of critically ill patients²⁰, while the other two refer to readmission in the ICU within 30 days after discharge from that unit^26,27. Despite this difference in concepts, in the three studies analyzed there is a reduction in the number of readmissions in the early mobilization group, with significant differences reported by two of them^20,27.

Length of hospital stay

The average length of hospital stay was quite heterogeneous in the different studies under analysis, varying between 6.5²⁶ and 46.5 days³².

When the differences between groups were observed, it was also verified that the results are also heterogeneous. On the one hand, nine studies describe a reduction in the length of hospital stay in the early mobilization group^{22,23,26,28-31,33,35} and of these, four show significant differences in relation to the control group ^23,29,33,35. On the other hand, six studies^{21,24,25,27,36} report an increase in the length of hospital stay in the early mobilization group, although only one study²⁵ reports statistically significant difference.

Focusing the analysis on randomized studies^28,35,36, the evidence remains inconsistent. Only one of the studies³⁵ reports a significant reduction in hospitalization time in patients mobilized early.

ICU high functionality

Functionality upon discharge from the ICU, in addition to not being a frequently analyzed outcome, was measured in different ways in the various investigations that included it.

When evaluating the functionality through the Barthel index²⁷, significantly better results were verified in the group of patients mobilized early (85 vs 63 points, with p<0.001). It was also found that more patients in the early mobilization group improved their sitting balance, from a first assessment to discharge (75% vs 36.7%, with p=0.008)³².

In the study that evaluated functionality through the Katz index³⁷, there was also better functionality at discharge in the early mobilization group, however the differences were not statistically significant.

Regarding the only RCT that evaluated this outcome³⁵, it was found that significantly more patients in the intervention group achieved the ability to walk, which corresponds to a score on the SOMS scale (Surgical intensive care unit Optimal Mobilization Score) equal to or greater than a 4 (52% vs 25%, with p<0.001) and that, on average, the levels of functionality at discharge, assessed by the mmFIM scale (mini-modified functional independence measure score) were significantly higher in patients included in the mobilization group early, a difference also observed by the authors at the time of hospital discharge.

In common, all studies identify functional improvements in the early mobilization group, in relation to the control.

Muscle strength at ICU discharge

Only one of the studies included in this review³⁷ evaluated muscle strength at ICU discharge. For this purpose, the MRC Index was used from the bilateral evaluation of four muscle groups, in which strength was scored from 0 to 5, with a final score variable between 0 and 40. The average score did not show significant differences between the groups. of early mobilization (22.4 points) and the control (26 points), with p=0.643.

Discharge to home

Of the 11 studies included in the analysis, only one³⁷ analyzed the destination for discharge from the ICU and not the destination for hospital discharge like the other investigations. Comparing the intervention groups with the respective controls, we found that in most studies the number of patients discharged home is greater in the early mobilization group, with two studies^29,35 reporting significant differences.

Incidence of ICU-acquired infections

The articles included in this review, all of them using non-randomized groups, analyze possible effects of early mobilization on the incidence of the following infections: pneumonia^22,23,33, ventilator-associated pneumonia^29,37, urinary catheter-associated infection ²⁷, bloodstream infection²⁹ and sepsis²³. In all areas analyzed, the early mobilization group had a lower rate of infection compared to the control group, with significant differences in the incidence of pneumonia²³, urinary catheter-associated infection²⁷ and bloodstream infections²⁹.

Incidence of pressure ulcers

The incidence of pressure ulcers (PU) was an outcome evaluated in seven of the investigations included in this review, all of them non-randomized. Of these, four studies^20,27,29,37 report a reduction in the incidence of PU in the early mobilization group, and two of them^20,29 identify significant differences in relation to the control group. In addition, an article²⁷ mentions a significant reduction in “quality outcomes” (which include: falls, ventilator-associated events, PU and urinary catheter-associated infections).

Otherwise, three studies^23,25,26 refer to an increase in the incidence of PU in the group of patients mobilized early, and one of them²⁵ identifies a significant increase in relation to patients in the control group.

Incidence of thromboembolic phenomena

Of the studies included in this review, six analyzed the incidence of thromboembolic phenomena, all of them non-randomized. Some studies analyzed thromboembolic events as a whole^22,37, others analyzed the incidence of deep venous thrombosis and pulmonary embolism separately^26,33, and other studies analyzed only the incidence of deep venous thrombosis^23,29. From the analysis of the results, three studies that report a non-significant increase in thromboembolic events^29,33,37 and three that report a reduction^22,23,26. In two cases^22,23 there were statistically significant differences, both favoring the early mobilization group.

Study limitations

This study had several limitations. First, the low proportion of randomized studies included in the analysis compared to observational studies. A significant number of historically controlled studies were included, and the results could be due to factors other than just the mobilization program. We also found discrepancies in the level of care provided to patients in the different control groups, which may give greater or lesser emphasis to the benefits of the program under analysis.

Contributions to the area

When it is compared with other previous reviews, this research, in addition to presenting an update of knowledge, is innovative for it concretely analyzes the effect of early progressive mobilization programs. Previous reviews refer to early mobilization more broadly, including different forms of early mobilization or rehabilitation.

As it’s evident in this article, early progressive mobilization programs are safe for patients, with no negative outcome resulting from their application.

The benefits found reinforce the importance of investing in this type of program and permanently change the paradigm of mobilization vs. immobilization of critically ill patients, where the rehabilitation nurse plays an important role. It is advisable to associate them with a wider range of interventions, such as the well-known ABCDEF bundle.