The impact of mobilization protocols on the length of postoperative hospitalization among cardiac surgery patients

Materials and methods. Participants – adult patients of 2018–2019 with less than 24-hour artificial lung ventilation (ALV). The first group were treated according to the early mobilization protocol (EM, patients of 2018); the second group were treated according to the EEM protocol (patients of 2019). Design: a retrospective analysis. Settings: cardiosurgical unit. Interventions: the major difference is that the resources of patient mobilization team have expanded since 2019, namely it included a physical therapist, which made it possible to modify the EM protocol (standing on the 2 postoperative day (POD), activation with the help of medical staff, respiratory exercise) to the EEM protocol (standing on the 1 POD following consultation with an anesthesiologist, exercises with a physical therapist, respiratory exercise). The main outcomes: LICU, LPOU and total postoperative hospitalization (LTPO) (number of nights).

One of the major factors of iatrogenic harm in patients is hospital-induced immobility, the so-called "pajama paralysis" [1]. Specifics and benefits of mobilizing critically ill patients of intensive care unit (ICU) and surgical patients are important physical therapy issues. A great deal of research focuses on the study of an early mobilization impact on ICU patients, including critically ill patients being the most studied population in this regard [2][3][4]. The barriers to mobilizing critically ill patients include cultural competence in ICU staff and a lack of resources [5].
However, a specialized mobilization team headed by a physical therapist (PT) may be a good option to take the hurdle [5,6]. Safety and effectiveness of ICU early mobilization have been proved and strategies to eliminate barriers have been studied in detail [7,8]. At the same time, the efficacy and feasibility of early mobilization have been criticized [9].
Historically, cardiac surgery (CS) patients were on bed rest for several postoperative days (POD) in the ICU to prevent complications and promote recovery. Over time, the benefits of early mobilization were acknowledged [10]. CS patients have been found to have a high risk for immobility-related complications [11][12][13]. CS patient mobilizing practice includes exercises to increase a range of motion, dangling at the bedside and transition to a chair on the operation day, if a patient is able. The first POD entails transferring from the bed to the chair 2-3 times and walking in a room, if possible [10].
Early mobilization is a key part of the post-operative physical therapy performed in the ICU for CS patients [14][15][16], although the issue of its desirable amount has not yet been resolved.
The obstacles to its implementation require increase in time spent with a patient and a decrease in patient/physical therapist ratio.

Aim
To determine the impact of implementing the extra early mobilization protocol (EEM) on the length of ICU (LICU) stay and postoperative unit (LPOU) stay and to assess the role of age, heart contractility, functional class and surgical outcomes.

Materials and methods
Design: retrospective analysis of electronic medical records.
Setting: cardiosurgical unit. Participants: adult patients who underwent CS in the Department of Congenital and Acquired Heart Defects (Department for Adults) during the period from 2018-2019. The analysis of 784 patients included their age, sex, body weight and length, body surface area (BSA), Euroscore II, cardiac ultrasound findings (ejection fraction, pulmonary hypertension), angina pectoris (functional class, unstable or painless type), comorbidities (arterial hypertension, diabetes mellitus, chronic obstructive pulmonary disease, chronic rheumatic heart disease), heart failure stage, NYHA functional class, cardiac surgery volume and duration, anesthesia, artificial lung ventilation (ALV), cardiopulmonary bypass, aortic compression duration and postoperative arrhythmia. The data of patients with ALV duration of less than 24 hours were analyzed.
The following indicators were studied: the length of ICU stay (LICU, number of nights), postoperative unit stay (LPOU, number of nights) and total postoperative hospital stay (LTPO, number of nights).
Organization of the study. The patients were divided into two groups: patients of the first group were treated according to the early mobilization protocol (EM, patients of 2018), whereas patients of the second group were treated according to the EEM protocol (patients of 2019). The major difference between the groups was the patient mobilization team extension in 2019 including PT, who enabled implementing the EEM protocol.
Interventions. Nurses and cardiologists put the EM protocol into practice through their assistance in mobilization and stimulation of patients' motor activity. The EM patient practice included: 1 -sitting on the bed during ICU stay (1 POD); 2 -upright standing, walking on the spot, walking across a ward in the postoperative unit (usually on the 2 POD if no restrictions existed opposing this); 3 -walking in the hospital corridor (usually on the 3 POD). If a patient was not transferred to the ward in the postoperative unit on the 2 POD, sitting time on the bed was increased, whereas upright standing was not performed in the ICU. Тhe EEM patient practice included: 1 -sitting at the side of bed (legs dangling), upright standing on the 1 POD (getting up with assistance and under the PT control, using a rolling walker; as agreed by an anaesthesiologist) and walking on the spot, if feasible; 2 -walking on the spot, walking across a ward in the postoperative unit (usually on the 2 POD unless contraindicated); 3 -walking in the hospital corridor on the 2-3 PODs, doing therapeutic exercises with the PT. If the patient was not transferred to the ward in the postoperative unit on the 2 POD, the length of sitting on the bed, standing and walking on the spot was increased, whereas w walking in the hospital corridor was not performed at all. Sessions (about 20 minutes each) with the PT were conducted 2 times a day on the 1-3 PODs, 1 time on the 4-6 PODs. When found appropriate (patient's condition, the need to motivate), the PT could increase the number and length of the sessions. PTs implemented the EEM protocol during the working week, and nurses performed patient mobilization according to the EM protocol on weekends. Before the surgery, the patients were briefly consulted by the PT on the aims and content of the рhysical therapy and activation algorithm after the surgery.
Physical therapy of both groups included breathing exercises, tapping massage techniques to stimulate mucous clearance and promote motor activity.
Statistical analysis. The materials of the study were processed in a program of statistical analysis IBM SPSS 21. Mathematical processing of the numerical data was fulfilled with the help of variation statistics. The data were assessed for normal distribution using the Shapiro-Wilk (W). But the data mostly did not fit normal distribution. Тherefore, nonparametric statistics was used. Wherever the quantitative indicators were not normal, we presented median (Me) and upper/lower quartiles (25 %; 75 %). Mann-Whitney U tests were used to assess significance of differences. The following criteria were also used: Phi and Cramer's (V), Spearman's rank correlation coefficients.

Results
The vast majority of indicators were initially tested and found to be non-normally distributed; only BSA indicator was normally distributed in both groups. Body weight and length indicators had the normal distribution only in one of the groups -in the EM (n = 385) and the EEM (n = 399), respectively. Age and anthropometric indicators did not differ between groups ( Table 1). Male-to-female ratio was 63.2 %/36.8 % in the EM group and 71.4 %/28.6 % in the EEM group, which influenced the statistical difference between the samples. The groups also were statistically different in the Euroscore II indicators. However, the differences revealed could not be considered critical taking into account Me (25 %; 75 %) indicator of the Euroscore II and the absence of Me indicator crossing at a 5 point level in both groups ( Table 1).
Other specificities of a past medical history and comorbidity were the same for both groups ( Table 2).
Besides, a correlation analysis revealed absence, low and very low relations between LICU, LPOU, LTPO indicators and criteria that distinguished between the groups of 2019 and 2018. Given that the correlations were similar in the EEM and EM groups, Table 3 demonstrates relations in the total patient sample (n = 784). Patient sex had no correlations with LICU, LPOU, or LTPO. Thus, the revealed differences between the EEM and EM groups were not meaningful when comparing LICU, LPOU, and LTPO values in the samples. It should be noted that LICU and LPOU correlation was very weak and inverse (which was seemingly surprising excepting the factors requiring hospital but not ICU stay). LICU did not prolong LPOU in the total sample of CS patients, provided that ALV duration was less than 24 hours.
Taking into account a 5-day working week of PTs, in order to obtain more accurate results, the analysis excluded 2019 patients operated on Friday (or on the day before public holidays) because the PTs could not perform their standing upright on the 1 POD. However, this fact did not result in differences between the EEM (n = 329) and EM   At the same time, the analysis revealed no differences between the EEM and EM groups in LICU, LPOU, and LTPO when age limits were set for the analysis (over 50, 60, 65, 70 years). This showed that the EEM and EM protocols had equal influence on these indicators regardless of the age. This corresponded to very weak correlations between the age and LICU, LPOU, LTPO ( Table 3).
Following the previous exclusion criterion for 2019 patients and considering that the vast majority of patients stayed in the ICU for two nights (the EEM 57.1 % with n = 329; the EM 52.5 % with n = 385), we assessed the impact of the EEM applying on LPOU and LTPO among patients who spent only two nights in the ICU. Such an exclusion revealed significant differences between the EEM (n = 188) and EM (n = 202) samples: LPOU (7 (6; 10) vs. 8 (7; 10); P = 0.031), LTPO (9 (8; 12) vs. 10 (9; 12); P = 0.031). Therefore, we concluded that the EEM protocol had reduced LPOU and, accordingly, LTPO for the patients with ALV of less than 24 hours and a two-night ICU stay after the surgery.
The analysis excluded the EEM patients who were operated on Friday (or on the day before public holidays) in the same way as the analysis algorithm for the general groups. However, this did not find differences between the EEM (n = 101) and EM (n = 55) patients. The next step (including patients with 2 night-LICU) did not result in the EEM (n = 66) superiority over the EM (n = 81) patients: LPOU (7 (6; 9) vs. 7.0 (6.5; 10.0); P = 0.102), LTPO (9 (8; 11) vs. 9.0 (8.5; 12.0); P = 0.102). Considering the last results obtained and the previous the EEM superiority (Friday excluding and 2 night-LICU filters -n = 188) over the EEM (LICU = 2 filter -n = 202), we came to the conclusion that there was a group or groups of patients (with 2 night-LICU filter), in whom the EEM protocol reduced LPOU (and LTPO as a result), but these groups did not include patients with CABG alone.
This warrants a more detailed identification of patient groups where the EEM protocol reduced LPOU and indirectly LTPO. Accordingly, the analysis included other patients (the EEM group with Friday excluding and 2 night-LICU filters; the EEM group with 2 night-LICU filter) grouped by the type of CS interventions (distinct or combined) to identify the EEM patients benefits. However, the analysis did not identify such groups of patients.
Thus, we can summarize that the comparison between the EEM and EM patients using exclusion factors (Friday excluding and 2 night-LICU filters for the EEM patients (n = 188); 2 night-LICU filter (n = 202) for the EM patients) revealed the statistically significant positive impact of the EEM protocol primarily on LPOU and indirectly on LTPO. However, this difference lost its significance when these groups were divided by the type of CS interventions due to the decrease in the number of patients in the groups and the increase in the difference threshold, which was necessary to detect the statistical difference.

Discussion
The PT's role is evidently important in improving the postoperative management of recovery processes in patients and their faster mobilization based on the EEM protocol.   However, the effectiveness of the EEM protocol seems doubtful in terms of reducing LICU, LPOU, and LTPO as compared to the standardized the EM protocol.
The results obtained actualize the issue of rational using PT time, since the assistance of a patient in standing upright requires a considerable amount of time in addition to overcoming and taking into account many barriers -wires, catheters, drainages.
The analysis confirmed the EEM protocol security. It did not reveal any critical cases leading to falls and traumas or necessity of additional surgery, which proved consistent cooperation of PTs and anesthesiologists in addressing the possibility of patient standing upright in the ICU. This confirmed safety and feasibility of early mobilization of critical and elderly patients in the ICU [7,[17][18][19].
At the same time, results revealed very weak correlations between age and LICU, LPOU and LTPO indicators ( Table 3) and no differences between the EEM and EM groups in LICU, LPOU, and LTPO indicators when setting the age limits for the analysis (over 50, 60, 65, 70 years), which proved the EEM protocol sufficiency regardless of patient's age.
Similar to this study, the work of S. Floyd et al. [11] aimed to determine the impact of progressive mobility protocol in cardiovascular ICU, particularly on LICU, LTPO. The researchers did not mention the difference, but for some reason emphasized clinical significance and cost savings of the implemented progressive mobility. At the same time, the analysis of this work showed large LICU and LTPO ranges with small samples, which suggests incorrect criteria of patient selection.
Interestingly, Dubb et al. [7] share our opinion that additional specialists including one PT were involved to implement the EM protocol and overcome barriers. This confirms the relevance of PTs involvement in the ICU. At the same time, PTs mobilized critically ill patients to higher levels compared with nurses [3].
Systematic review of P. M. R. Santos et al. [20] provided evidence of early mobilization impact after CS interventions to prevent postoperative complications and reduce LTPO. The authors [20] identified a variety of methods used for mobilization, as well as the periods considered as early. At the same time, P. M. R. Santos noted that mobilization groups had better results than non-intervention groups and, as a rule, these advantages did not differ among intervention groups (no superiority of the EEM group over the EM group was revealed in this study as well). The researchers failed to perform meta-analysis due to variability of interventions suggested as early mobilization.
In this respect, the conclusion of T. Castelino et al. [21] should be also mentioned: "Few comparative studies have evaluated the impact of early mobilization protocols on outcomes after abdominal and thoracic surgery. The quality of these studies was poor and the results were conflicting. Although bed rest is harmful, there is little evidence available to guide clinicians in effective early mobilization protocols that increase mobilization and improve outcomes".
A recent study has revealed that patients in the early mobilization group had shorter duration of hospitalization after CS [22].
Early mobilization of CS patients is obviously very important in general, but in some studies, this mobilization was even earlier than it was in the EEM group analyzed in the current study. For instance, J. K. Brown et al. [23] state: "It is important to emphasize to patients the importance of an early mobilization plan with daily goals for time out of bed and distance started as soon as the day of surgery". It seems interesting that the importance of being out of bed and distance walked begins as early as the day of the surgery. However, the following factors should be considered as well: surgery start time (8:00 AM, 9:00 AM, or 2:00 PM), duration of surgery, anesthesia and ALV, the duration of PT working day, the patient status.
Brown J. K. [23], on the other hand, showed the absence of Enhanced Recovery After Surgery (ERAS) impact on the length of hospital stay with reference to I. O. Fleming et al. [24]. A significant number of the physical therapy protocols [25][26][27] reported mobilization from the bed on the 1 POD, walking within the ward on the 2 POD, walking in the corridor on the 3 POD, but there are studies indicating that standing upright was normally performed on the 3 POD [28,29]. PT survey [30] also confirmed that there were slight variations in mobilization supervised by PTs following CS.
It has long been noted: "Lastly, it is important to realize that early extubation does not necessarily mean early intensive care unit or hospital discharge" [31]. It should be realized that physical therapy and standing upright on the 1 POD do not guarantee reduction in LICU, LPOU and LTPO. There are many factors that make these terms prolonged, namely the results of blood tests, rhythm disturbances, pericarditis, dysfunction of kidney and other internal organs, the need to determine the optimal dose of pharmacological drugs.

Conclusions
1. The EEM protocol implementation did not reduce the length of stay in the intensive care unit and postoperative unit, as well as the total length of hospital stay as compared to the standardized EM protocol.
2. The studied approaches to early mobilization were equally effective, despite the fact that the EEM protocol was implemented by physical therapists, and the EM protocolby nurses and cardiologists. Correlation analysis revealed absence, weak and very weak relations between the LICU, LPOU, LTPO indicators and age, ejection fraction, ALV.
Study limitations. The study failed to conduct a fullscale examination of the postoperative condition severity in patients. Taking into account a quite large number of patients in the groups, exclusion criteria (ALV duration of less than 24 hours) and the fact that the examined patients were enrolled from two consecutive years, the possibility of the groups to differ in postoperative severity indicators was minimized.
Prospects for future research. Taking into account that mobilization and walking are performed on the day of surgery in some studies, these factors raise the following questions. If surgery starts in the morning, should PT ambulate patients, operated on the same day, every day after 3:00 PM? If surgery starts in the afternoon, when should PT perform mobilization? Will transition to the chair, sitting and walking be maximally aidless on the day of the surgery? Will PT perform it instead of the patient? Will patient and PT be willing for such mobilization on the day of surgery? Will the patient remember any physical therapy goals if he/she has just recovered from anesthesia and just wants to sleep or drink? What percentage of anaesthesiologists, cardiologists, and other staff members will consider such activation rational? Is it possibly better to wait a day or even two? How much time does this mobilization require on the day of surgery? Is it possible to use this time more rationally?
There is the need to study the EEM protocol impact on the dynamic of strength, stamina, endurance, external respiration function, and satisfaction with physical therapy in postoperative patients.