Early rehabilitation relieves diaphragm dysfunction induced by prolonged mechanical ventilation: a randomised control study

This prospective randomised controlled trial was conducted in accordance with CONSORT (Additional file 1) in a 16-bed general ICU of the Affiliated Hospital of Qingdao University from June 2019 to March 2020. The trial was approved by the Ethics Committee of the Affiliated Hospital of Qingdao University (No. QYFY WZLL 25526) and is registered in the Chinese Clinical Trial Registry (No. ChiCTR1900024046). Written informed consent was obtained from the patients or their relatives.

Patients who were eligible for inclusion were required to meet the following criteria: (1) Prolonged MV (> 72 h); (2) stable oxygen saturation, fraction of inspired oxygen ≤ 55%, and positive end expiratory pressure (PEEP) ≤ 8 cmH₂O; (3) dose of dopamine < 10 μg/kg/min and dose of epinephrine < 0.4 μg/kg/min; (4) mean arterial pressure > 75 mmHg and urine output > 1 mL/kg/h; (5) good healing of the incision after surgery; (6) normal cognitive function; and (7) no history of chronic mental illness or chronic obstructive pulmonary disease.

The exclusion criteria were as follows: (1) Inability to perform physical activities; (2) long-term MV prior to admission; (3) neurological comorbidities involving muscles; (4) irreversible disorders with a 6-month mortality rate of > 50% according to Acute Physiology and Chronic Health Evaluation II (APACHEII); (5) unsound limbs; (6) administration of glucocorticoids (prednisone or other corticosteroid dose equivalents > 20 mg/day) for at least 20 days prior to admission; (7) cardiopulmonary resuscitation before admission to the ICU; (8) radiotherapy or chemotherapy within the previous 6 months; (9) presence of comorbidities, including acute myocarditis, deep venous thrombosis/embolism, and cerebrovascular accident; and (10) unstable fractures.

Eighty patients who met the inclusion criteria were enrolled in this study. A computer-generated allocation order with a block size of 4 was used with patients allocated at a 1:1 ratio to an early rehabilitation group or a control group. Finally, 39 patients were assigned to the early rehabilitation group (treated with early rehabilitation therapy in the ICU) and 41 patients were assigned to the control group (treated with standard care).

The early rehabilitation program was designed in accordance with the physical and psychological conditions of the patients. Rehabilitation therapy consisted of six levels: level 0, turning over once every 2 h for unconscious patients with unstable vital signs; level 1–2, in addition to turning over, maintaining joint range of motion to prevent muscle atrophy, and placing normal limb position for conscious patients who could sit up for at least 20 min, 3 times a day; level 3, similar to level 2, but sitting on the edge of the bed for patients who could perform upper-limb anti-gravity training; level 4, similar to level 3, but standing up or sitting in a chair for at least 20 min a day for patients who could perform lower-limb anti-gravity training; and level 5, patients actively moved from the bed and walked bedside. Level 1–5 exercises were suitable for conscious patients. Generally, level 1–3 exercises were selected for patients with tracheal intubation, while level 3–5 exercises were selected for patients with tracheotomy. For level 0 exercises, the patients were placed in a supine or orthopnoea position. The connecting wires of the electrocardiogram (ECG) monitor were separated, the drainage tubes were clamped, and all pipes were placed on the right side of the bed. Enteral nutrition or constraints were temporarily stopped. For patients using a ventilator, the condensate water in the ventilator pipe was dumped to prevent airway infection because of the backflow of water during turning over. The connection between the monitoring lead wire and the ventilator pipe was loosened appropriately to prevent pulling the patient. Two nurses stood on the same side of the bed, one holding the patient's shoulders and waist, and the other holding the patient's buttocks and popliteal fossa. The nurses simultaneously lifted the patient to the proximal side, before gently turning the patient to the opposite side. The patients were given percussion on the back, except for those who were contraindicated for percussion. A soft pillow was placed on the back, chest, and between the knees according to the requirements of the lateral position. Enteral nutrition was continued, and the patient’s vital signs were monitored to comprehensively evaluate the effect of turning over.

Rehabilitation therapy was terminated when the following criteria were met: (1) mean arterial pressure < 65 mmHg or > 120 mmHg; systolic or diastolic blood pressure of patients with pre-existing renal diseases decreased by more than 10 mmHg after treatment; (2) heart rate < 50 beats/min or > 140 beats/min; (3) occurrence of new arrhythmia or taking > 1 μg/kg/min of norepinephrine to maintain blood pressure; (4) inhaled oxygen concentration of 60.0% accompanied by PaO₂ < 70 mmHg; (5) PEEP > 8 cmH₂O; (6) pulse blood oxygen saturation (SpO₂) decreased by 10% or < 85%; (7) respiration rate > 35 breaths/min; (8) body temperature > 38 °C; and (9) after rehabilitation exercise, the patient’s condition deteriorated or the patient became unconscious as a result of issues such as new sepsis, gastrointestinal bleeding, and chest pain. Reassessment of the above conditions was necessary on the second day of rehabilitation. After 3 days of rehabilitation training, patients in the rehabilitation group continued to undergo rehabilitation exercises until successful weaning from MV.

US measurement of the diaphragm was performed for all patients as previously described [17, 18]. One well-trained investigator (Ying L) performed the US measurements after 24 h of MV (at 1-day after MV) and 3 days of rehabilitation training (at 4-day after MV) using continuous positive airway pressure (CPAP) mode. The measurement was stopped immediately if the patients displayed any sign of discomfort or distress.

Diaphragm function and activity were evaluated using two diaphragmatic parameters, DE and DTF. DE reflects the amplitude of the movement of the diaphragm during the respiratory cycle. The broadband linear array probe was placed at the junction of the midclavicular line or the anterior axillary line and the lower edge of the costal arch. The liver or spleen was set as the diaphragm acoustic window. The probe was pointed to the head and back, and DE was displayed under M-mode (time-motion mode). Three subsequent measurements were averaged to taken as the final value. DTF shows varied thickness of the diaphragm at end-expiration and end-inspiration. The maximum and minimum values of each breathing cycle were taken as the end-inspiratory diaphragm thickness (DTei) and the end-expiratory diaphragm thickness (DTee), respectively. DTF was calculated by DTF = (DTei − DTee)/DTee × 100%. The values for 3 consecutive respiratory cycles were recorded and the average value was taken as the final value. All patients underwent diaphragm US in a semi-recumbent position (bed slope of 45°); a right anterior subcostal view was preferred.

The sample size was calculated based on the following formula, with DE as the primary outcome:where σ = 0.4; Z_α/2 = 1.96 when α = 0.05; Z_β = 0.84 when β = 0.2. μ₁ in the rehabilitation group was estimated to be 1.5, and μ₂ in the control group was estimated to be 1.25. Therefore, 40 participants per group were needed. Considering a dropout rate of 10%, the minimal sample size of each group was calculated to be 44. The sample size was minimized to 80 after an interim analysis.

All data were analysed using the SPSS ver. 20.0 MacIntosh package. Normality distribution of all data was checked by Shapiro—Wilk normality test. For the data conform to the normality distribution, parametric test was used to calculate the difference; otherwise, nonparametric test was utilized. Parametric and nonparametric data of descriptive statistics are expressed as mean ± standard deviation and median (interquartile range), respectively. DE and DTF at different times in the same group were compared using the paired t-test. Differences between rehabilitation group and control group were compared by independent samples t test. Nonparametric tests (Wilcoxon rank-sum test) were employed to compare continuous data between the groups. Categorical variables were compared using Fisher’s exact test. The level of statistical significance was set at p < 0.05.

Eighty patients who were admitted to the ICU during the study period and met the inclusion criteria were enrolled in the study. The patients were randomly dichotomized into a rehabilitation intervention group (n = 39) and a control group (n = 41). The baseline characteristics of the two groups are shown in Table 1. There were no statistically significant differences in age, BMI, APACHEII score, and sex between the patients in the rehabilitation and control groups (p < 0.05). Males accounted for more than half of these patients (rehabilitation intervention group, 64.10%; control group, 56.10%). There was no statistically significant difference in the mean BMI between the two groups (rehabilitation group, 23.18 ± 3.32 kg/m²; control group, 23.22 ± 3.67 kg/m²).

After 1 day of MV, rehabilitation therapy was administered to the rehabilitation intervention group for 3 days, whereas only standard care was given to the control group. All 39 patients in the rehabilitation intervention group completed the rehabilitation exercises. One patient experienced SpO₂ < 90% and was dropped out from the study. The SpO₂ was improved to the normal range, and the vital signs were stable following bed rest. No accidental extubation or other adverse events were reported during the study period.

The two groups were assessed for DE and DTF using US at 1-day and 4-day MV. DE and DTF were significantly decreased in all patients at 4-day MV compared to those at 1-day MV (1.42 ± 0.53 vs. 1.30 ± 0.44, p < 0.001; 0.16 ± 0.06 vs. 0.13 ± 0.06, p < 0.001) (Fig. 1; Table 2). This suggests that prolonged MV impairs diaphragm function.

Before rehabilitation therapy, there were no significant differences in DE and DTF between both groups at 1-day MV (p-value > 0.05) (Table 3). At 4-day MV, the rehabilitation group had significantly decreased DTF compared to the control group (0.15 ± 0.06 g vs. 0.12 ± 0.05 g, p-value = 0.008) (Table 4). Changes of DE and DTF in rehabilitation group and control group were displayed in Fig. 2. Although a decline in DE was observed in the rehabilitation group relative to that in the control group, the decline was marginally significant (p = 0.541) (Table 4). We further found that the magnitude of the decrease in DTF before and after 3 days of rehabilitation therapy was obviously smaller in the rehabilitation group than in the control group (0.017 ± 0.030 vs. 0.034 ± 0.036, p = 0.026) (Table 4). These results reveal that early rehabilitation training can ameliorate diaphragm dysfunction caused by prolonged MV.

The clinical outcomes of the rehabilitation intervention and control groups were compared (Table 5). The rehabilitation group had a significantly shorter duration of ventilator use (7.49 ± 2.59 days vs. 9.41 ± 5.32 days, p-value = 0.045) and a significantly shorter duration of intubation (8.31 ± 2.80 days vs. 10.37 ± 5.32 days, p = 0.037) than the control group (Table 5). The two groups were comparable in terms of duration of ICU stay, proportion of patients undergoing tracheotomy, and proportion of recovered patients (p-value > 0.05) (Table 5).