Rapid Shallow Breathing Index and Ultrasonographic Diaphragmatic Parameters as Predictors of Weaning Outcome in Critically Ill Patients on Mechanical Ventilation

Abstract

Background:

Successful weaning is a crucial element in care toward critically ill patients on mechanical ventilation. An attempt was made to propose and assess a reliable predictor of weaning outcome.

Materials and Methods:

A prospective observational study was conducted on 76 patients on mechanical ventilation, assessed by Acute Physiology and Chronic Health Evaluation II (APACHE II) score. For all these patients we calculate Rapid shallow breathing index (RSBI), Ultrasonographic diaphragmatic parameters namely diaphragmatic excursion (DEx), diaphragmatic thickening fraction (DTF) and diaphragmatic contraction velocity (DCV). Values were compared among patients with two groups of successful and failed weaning outcomes, respectively, and statistically analyzed.

Results:

Of 76 patients included in the study, with ultrasonographic diaphragmatic parameters being measured 30 min into SBT, 71 patients tolerated spontaneous breathing test (SBT) for 2 h and were extubated. Of these, 61 patients did not require reintubation or any form of ventilatory support within 48 h after extubation. There was a statistically significant difference in APACHE II scores, duration of ventilation, oxygen saturation levels, RSBI, DEx, DTF, and DCV between groups of patients who showed successful and failed weaning from mechanical ventilation. There were a significant positive correlation between the duration of ventilation and the RSBI and a significant negative correlation between DEx, DCV, DTF, and duration of ventilation. As predictors of weaning outcome, RSBI showed the best validity, followed by DCV, DTF, and DEx.

Conclusion:

RSBI can be reliably used as a predictor of weaning outcome in critically ill patients on mechanical ventilation.

INTRODUCTION

Mechanical ventilation is a lifesaving procedure in most of the critically ill patients admitted to intensive care units (ICUs). The procedure has gotten even greater attention and application during the recent COVID-19 pandemic, and rates of mechanical ventilatory support being received by the ICU patients have been reported to range from 29.1% to 89.9% in various epidemiological studies.[1,2] However, a patient put on a mechanical ventilatory support needs to be liberated from the support and intubation and is an essential and unavoidable element of the care toward critically ill patients. Withdrawing the mechanical ventilatory support, gradually or abruptly, as demanded by the clinical condition or stability of the patient, is known as weaning. However, it might be associated with severe complications or costs, often proportional to the duration of support, followed by extension of ventilatory support and ICU stays of the patients, and even mortality.[3]

An appropriately chosen time for liberation from mechanical ventilation reduces the risks of both prolonged ventilatory support and premature weaning. An unnecessary prolonging of the mechanical ventilation may lead to diaphragmatic atrophy, venous thromboembolism, pneumonia, or delirium, whereas premature weaning may lead to an added respiratory or cardiovascular stress, contributing toward higher morbidity and mortality.[4,5] Thus, assessment of weaning preparedness and prediction of the clinical outcomes of weaning are important clinical considerations. Various weaning predictors have been introduced so far, few of them being the ultrasonographic diaphragmatic parameters and the rapid shallow breathing index (RSBI).[4,6]

One of the main reasons behind an unsuccessful weaning is the diaphragmatic dysfunction, due to polyneuropathy and myopathy, and is found in 33%–95% of the patients weaned from mechanical ventilation. Even short durations of ventilatory support may reduce the force needed for diaphragmatic capacity, which may worsen with disuse atrophy of the myofibers of the diaphragm after the administration of a neuromuscular blocking agent.[5,6,7] Assessing the diaphragmatic function may thus help the intensivists predict the time and outcome of weaning. Recently ultrasonography becoming an invaluable tool for intensivists. Diaphragmatic functions such as diaphragmatic excursion (DEx), diaphragmatic thickening fraction (DTF), and diaphragmatic contraction velocity (DCV) have been assessed by ultrasonography to predict the weaning outcomes.[6,8] Similarly, to aid this prediction, many weaning indices have been introduced. RSBI, described in 1991, is the ratio of respiratory rate (RR) to tidal volume (TV). While it can be calculated manually, the RSBI obtained by direct ventilometry is displayed in ventilator algorithms as well, easing the efforts and vision of the intensivists.[4,9]

Several studies have used ultrasonographic diaphragmatic parameters and RSBI to predict the weaning outcomes. However, there have been varied and conflicting propositions regarding their implications in the predictability of the weaning outcomes.[10,11,12,13] Noting the similarities and differences among the previous studies, this study was planned to evaluate the utility of the ultrasonographic diaphragmatic parameters, namely DEx, DTF, and DCV versus RSBI in predicting the weaning outcomes. We determine the best one among them as a predictor of weaning and their cutoff values for a successful weaning. If possible, correlation between the number of days the patients on mechanical ventilation with these weaning parameters are done.

MATERIALS AND METHODS

This prospective observational study was conducted on mechanically ventilated adult patients (>18 years of age) admitted to the ICU, of a tertiary care center, from November 2019 to July 2021, using a universal sampling method. The study was also registered with the Clinical Trials Registry of India with CTRI registration number: CTRI/2020/12/030097. Patients ventilated with endotracheal tube for ≥24 h who met the criteria for Spontaneous breathing test (SBT) were included in the study. Patients with the clinical conditions such as pneumothorax, ascites, unilateral or bilateral diaphragmatic palsy, pregnancy, history of neuromuscular disease, or tracheostomy were excluded from the study. Considering the sensitivity of DEx in predicting extubation failure from data of a previous study[11] as 95%, with 5% precision and 5% alpha error, the minimum sample size was calculated to be 73 patients. The study was approved by the institutional ethical committee, and written informed consent was obtained from the patients or their relatives, depending on the ability of the patients to give the consent.

All patients were assessed clinically by Acute Physiology and Chronic Health Evaluation II (APACHE II) score at the time of admission in the ICU. The decision to wean the patients from mechanical ventilation was made by the critical care physicians and the primary ICU team. Patients were subjected to spontaneous breathing trial (SBT) under low-level pressure support (pressure support ≤8 and positive end-expiratory pressure [PEEP] ≤5 on pressure support ventilation mode) for 2 h, when the following criteria were met:

1. Resolution of the underlying pathology that led to intubation and ventilatory support

2. The patient was calm, cooperative, and arousable

3. PaO₂ >60 mmHg and PaCO₂ <50 mmHg as assessed by arterial blood gas analysis

4. FiO₂ ≤0.4

5. PEEP ≤5

6. PaO₂/FiO₂ >300

7. RR <35/min

8. Hemodynamically stable without the need for vasopressor support.

Right hemi-diaphragmatic parameters, namely DEx, time of peak inspiration (CT), diaphragmatic thickness at end inspiration (DTI), and Diaphragmatic thickness at end expiration (DTE), were measured 30 min into the 2-h SBT using GE LOGIQ P6 PRO ultrasound system. The methodology followed to measure these parameters was referred from the study conducted by Matamis et al.[14] Three measurements each were taken for all the parameters during tidal breathing, and measurements during deeper or smaller breaths were excluded. The critical care physician and the primary ICU team were blinded to the measurements. DTF and DCV were calculated using the formulae suggested by Matamis et al.[14] and Palkar et al.[10]

DTF = (DTI − DTE) ×100

DCV = Diaphragm excursion (mm)/time to peak inspiration (s).

Vitals of patients (heart rate, systolic, and diastolic blood pressure), TV, RR, and minute ventilation (MV) were recorded, and RSBI (RR/TV) was calculated.

The criteria for SBT failure were defined as:

1. Alteration of mental status defined as Glasgow Coma Scale score <10

2. RR >35/min

3. Systolic blood pressure >180 mmHg or <90 mmHg

4. Heart rate >120 bpm or increase in 20% from the baseline value.

If a patient failed SBT within 30 min, diaphragmatic parameters were immediately measured. Patients who passed the SBT without any clinical deterioration were extubated under the supervision of the critical care physician and the primary ICU team. All patients who successfully passed SBT were extubated and followed up for 48 h to assess any weaning failure and a need for reintubation. Weaning failure was defined as the requirement of reintubation or noninvasive ventilatory (NIV) support within 48 h of extubation. The decision for reintubation or to initiate NIV support was made by the critical care physician and primary ICU team based on their clinical and parametrical judgment.

Based on these observations, the patients were divided into two groups: the success group, with patients who were successfully weaned of the ventilatory support, passed the SBT, and extubated and did not require any NIV support or reintubation within 48 h after extubation, and the failure group, with patients who did not pass the SBT or passed the SBT but required NIV support or reintubation within 48 h after extubation.

Data were entered into a Microsoft Excel data sheet and were analyzed using SPSS 22 version software, (IBM, Washington DC, USA). Categorical data were represented in the form of frequencies and proportions. The Chi-square test was used as test of significance for qualitative data. Continuous data were represented as mean and standard deviation. The Independent t-test or Mann–Whitney U-test was used as test of significance to identify the mean difference between two quantitative variables. The Pearson correlation was done to find the correlation between two quantitative variables.

RESULTS

A total of 97 patients on mechanical ventilation were assessed for eligibility to be included in the study, with 76 of them meeting the inclusion criteria and consenting to participate in the study. Of the 76 patients included in the study, 71 patients tolerated the SBT for 2 h and were extubated, with ultrasonographic diaphragmatic parameters being measured 30 min into SBT. Of these, 61 patients did not require reintubation or any form of ventilatory support within 48 h after extubation, two patients needed reintubation within 48 h after extubation, and eight patients required NIV support within 48 h after extubation [Figure 1 and Table 1]. No statistically significant difference was observed between the patients with successful and failed weaning [Table 2]. However, there was a statistically significant difference in the APACHE II scores of both the groups of patients [Table 3]. Similarly, there was a statistically significant difference in the number of ventilator days among both the groups of patients, with patients exhibiting successful weaning having less ventilator days as against the ones with failed weaning [Table 4]. However, there were no significant differences in the vital parameters, namely heart rate and systolic and diastolic blood pressure among both the groups, except for a significant difference in the oxygen saturation levels [Table 5]. A significant difference was observed in the respiratory parameters, namely the TV and RR among both the groups of patients [Table 6]. The values of MV, RSBI, DEx, CT, DCV, DTI, DTE, and DTF were also statistically significantly different among the patients who exhibited a successful and failed weaning [Table 7].

Figure 1.

Consort flowchart. MV = Minute ventilation, SBT = Spontaneous breathing test, NIV = Noninvasive ventilatory

Table 1.

Weaning outcome among the study patients

Group	Count	Percentage
Success	61	80.3
Failure
Failed SBT	5	19.7
Reintubation within 48 h	2
NIV within 48 h	8
Total (n)	76	100.0

SBT = Spontaneous breathing test, NIV=Noninvasive ventilatory

Table 2.

Demographic comparison among the patients with successful and failed weaning

	Group						P
	Success			Failure
	Mean	SD	Median	Mean	SD	Median
Age (years)	48.33	17.76	50.00	57.27	17.67	55.00	0.085
Height (m)	1.61	0.07	1.64	1.57	0.07	1.55	0.089
Weight (kg)	65.75	8.43	68.00	61.33	5.21	62.00	0.057
BMI (kg/m2)	25.35	2.32	25.71	24.82	2.02	25.08	0.414

BMI=Body mass index, SD=Standard deviation

Table 3.

Comparison of Acute Physiology and Chronic Health Evaluation II scores among patients with successful and failed weaning

	Group						P
	Success			Failure
	Mean	SD	Median	Mean	SD	Median
APACHE II	20.75	3.32	21.00	25.53	2.77	25.00	<0.001*

*P value <0.001 is statistically significant. APACHE II=Acute Physiology and Chronic Health Evaluation II, SD=Standard deviation

Table 4.

Comparison of ventilator days among patients with successful and failed weaning

	Group						P
	Success			Failure
	Mean	SD	Median	Mean	SD	Median
VD (days)	2.74	1.38	2.00	5.27	1.16	5.00	<0.001*

*P value <0.001 is statistically significant. VD=Ventilator days, SD=Standard deviation

Table 5.

Comparison of vital parameters during Spontaneous breathing test among the patients with successful and failed weaning

	Group						P
	Success			Failure
	Mean	SD	Median	Mean	SD	Median
Heart rate (bpm)	84.15	10.30	82.00	86.80	8.96	85.00	0.363
SBP (mmHg)	126.75	11.19	128.00	129.20	12.55	134.00	0.461
DBP (mmHg)	82.82	9.90	82.00	83.60	10.23	90.00	0.787
Oxygen saturation (%)	97.95	1.91	98.00	96.53	1.77	96.00	0.011*

*P value <0.001 is statistically significant. SBP=Systolic blood pressure, DBP=Diastolic blood pressure, SD=Standard deviation

Table 6.

Comparison of respiratory parameters among the patients with successful and failed weaning

	Group						P
	Success			Failure
	Mean	SD	Median	Mean	SD	Median
TV (mL)	377.59	52.68	371.00	309.93	20.47	310.00	<0.001*
RR (/s)	19.90	4.02	18.00	31.27	2.12	32.00	<0.001*

*P value <0.001 is statistically significant. SD=Standard deviation, TV=Tidal volume, RR=Respiratory rate

Table 7.

Comparison of minute ventilation, rapid shallow breathing index, diaphragmatic excursion, CT, diaphragmatic contraction velocity, diaphragmatic thickness at end inspiration, diaphragmatic end-expiration, and diaphragmatic thickening fraction between the patients with successful and failed weaning

	Group						P
	Success			Failure
	Mean	SD	Median	Mean	SD	Median
MV (L/min)	7.45	1.51	7.24	9.69	0.91	9.92	<0.001*
RSBI (breaths/min/L)	53.66	15.41	48.00	100.80	8.99	101.00	<0.001*
DE×1 (mm)	24.10	6.03	25.90	11.00	2.90	10.20	<0.001*
DE×2 (mm)	24.34	5.86	25.70	10.93	2.93	10.40	<0.001*
DE×3 (mm)	24.18	6.07	26.10	10.96	2.91	10.40	<0.001*
DEx mean (mm)	24.25	5.97	25.90	10.97	2.90	10.50	<0.001*
CT1 (s)	1.39	0.29	1.40	1.05	0.19	1.00	<0.001*
CT2 (s)	1.44	0.33	1.40	1.07	0.21	1.00	<0.001*
CT3 (s)	1.47	0.35	1.50	1.07	0.19	1.00	<0.001*
CT mean (s)	1.44	0.31	1.40	1.07	0.17	1.00	<0.001*
DCV (mm/s)	16.49	3.74	17.00	9.93	3.10	9.00	<0.001*
DTI1 (mm)	3.23	0.33	3.20	2.80	0.19	2.80	<0.001*
DTI2 (mm)	3.19	0.37	3.10	2.78	0.25	2.80	<0.001*
DTI3 (mm)	3.20	0.37	3.10	2.78	0.27	2.80	<0.001*
DTI mean (mm)	3.21	0.35	3.10	2.79	0.23	2.80	<0.001*
DTE1 (mm)	2.40	0.27	2.40	2.19	0.16	2.20	<0.001*
DTE2 (mm)	2.48	0.24	2.40	2.26	0.17	2.20	0.002*
DTE3 (mm)	2.49	0.24	2.40	2.24	0.18	2.20	<0.001*
DTE mean (mm)	2.46	0.23	2.40	2.23	0.17	2.20	<0.001*
DTF (%)	30.27	5.17	30.70	24.75	5.28	23.80	<0.001*

*P value <0.001 is statistically significant. DTF=Diaphragmatic thickening fraction, DTE=Diaphragmatic end expiration, MV=Minute ventilation, RSBI=Rapid shallow breathing index, DEx=Diaphragmatic excursion, DCV=Diaphragmatic contraction velocity, DTI=Diaphragmatic thickness at end inspiration, SD=Standard deviation, CT=Comparison of time to peak inspiration

There was a significant positive correlation between the duration of ventilation and the RSBI, with an increase in RSBI with the increasing duration of ventilation. Similarly, there was a significant negative correlation between the DEx, DCV, and DTF and the duration of ventilation, with a decrease in these parameters with increasing duration of ventilation [Table 8]. As predictors of weaning outcome, RSBI showed the best validity, followed by DCV, DTF, and DEx [Table 9].

Table 8.

Correlation between the duration of ventilation and the rapid shallow breathing index and ultrasonographic diaphragmatic parameters

	VD (days)
RSBI (breaths/min/L)
Pearson correlation	0.601**
P	<0.001*
n	76
DEx (mm)
Pearson correlation	−0.579**
P	<0.001*
n	76
DCV (mm/s)
Pearson correlation	−0.433**
P	<0.001*
n	76
DTF (%)
Pearson correlation	−0.399**
P	<0.001*
n	76

*In the present study there was a significant negative correlation between duration of ventilation and DEx, DCV and DTF i.e., with increase in duration of ventilation there was decrease in DEx, DCV and DTF. **In the present study there was a significant positive correlation between duration of ventilation and RSBI. i.e., with increase in duration of Ventilation there was an increase in RSBI. RSBI=Rapid shallow breathing index, DEx=Diaphragmatic excursion, DCV=Diaphragmatic contraction velocity, DTF=Diaphragmatic thickening fraction, VD=Ventilator days

Table 9.

Comparison of rapid shallow breathing index and ultrasonographic diaphragmatic parameters in predicting weaning outcome

Order of validity	Parameters	AUC	SE	95% CI
1	RSBI	0.988	0.0125	0.931–1.000
2	DCV	0.855	0.0785	0.741–0.933
3	DTF	0.737	0.106	0.608–0.843
4	DEx	0.514	0.127	0.381–0.645

RSBI=Rapid shallow breathing index, DEx=Diaphragmatic excursion, DCV=Diaphragmatic contraction velocity, DTF=Diaphragmatic thickening fraction, SE=Standard error, CI=Confidence interval, AUC=Area under the curve

DISCUSSION

RSBI is the simplest and most widely adopted index for predicting weaning outcomes and reflects the respiratory muscle endurance. However, RSBI cannot be depended on alone due to the possibility of fatigue of the nondiaphragmatic muscles in their attempt to compensate for a failing diaphragm. Thus, it has always been better to also monitor the diaphragmatic function in the era of excellent ultrasonographic assistance.

In the present study, 80.61% of the patients were successfully weaned of mechanical ventilation, with the remaining either failing the SBT, or requiring a reintubation or NIV support within 48 h after the weaning. The evaluated APACHE II scores between both the groups of patients were significantly different, and a similar conclusion was drawn in the study conducted by Farghaly and Hasan[15] Similarly, a significant difference observed in the duration of ventilatory support required in the patients undergoing successful and failed weaning (2.74 ± 1.38 days and 5.27 ± 1.16 days, respectively) was similar to that observed by Banerjee and Mehrotra (4.21 ± 1.15 days and 6.62 ± 1.12 days, respectively),[11] Spadaro et al. (57 h and 169 h, respectively),[16] and Ferrari et al. (26 h and 37 h, respectively).[17]

A significant difference was found in the RSBI among the patients who underwent successful and failed weaning, with a higher RSBI in the latter group of patients. We found RSBI to be an excellent predictor of weaning failure with a sensitivity of 93.33%, specificity of 100%, positive predictive value (PPV) of 10%, negative predictive value (NPV) of 98.4%, and area under receiver operating curve (AUROC) of 0.988 for a cutoff of 90 breaths/min/L. This was comparable to Bahgat et al.[18] who reported that RSBI was an excellent predictor of weaning outcomes with a sensitivity of 97%, specificity of 64%, PPV of 78%, NPV of 95%, and an AUROC of 0.89 for a cutoff value of 105 breaths/min/L. Similarly, Banerjee and Mehrotra[11] reported a statistically significant difference in mean RSBI between the successful and failed weaning groups (28.43 ± 4.33 breaths/min/L vs. 8.46 ± 4.58 breaths/min/L, respectively) and concluded RSBI to be an excellent marker of extubation outcome with a high sensitivity (100%) and specificity (100%) and AUROC of 0.996 for a cutoff of 104 breaths/min/L. Khan et al.[12] reporting a statistically significant difference in mean RSBI between the successful and failed weaning groups (54.05 ± 7.00 breaths/min/L vs. 63.14 ± 7.64 breaths/min/L, respectively) proposed a 79% sensitivity, 64% specificity, and an AUROC of 0.815 for a cutoff of 59 breaths/min/l. The lower cutoff value obtained in their study can be attempted to be justified by the fact that they considered RR ≥30 as a criterion to discontinue SBT, as against RR ≥35/min in our study.

However, our results contradicted the propositions of Huang et al.[13] who reported RSBI to be a comparatively poor predictor of weaning success, with a sensitivity of 58.33%, specificity of 85.71%, positive likelihood ratios of 4.08, negative likelihood ratios of 0.49, and AUROC of 0.695 for a cutoff value >68 breaths/min/L. Furthermore, they observed no significant variation in mean RSBI between patients with and without diaphragmatic dysfunction (61.1 ± 10.39 breaths/min/L vs. 61.89 ± 19.1 breaths/min/L, P = 0.634). This contradiction can be due to only geriatric patients aged >80 years being included in their study, as against the mean age of patients in our study being 50.09 ± 17.98 years. Moreover, they calculated RSBI before starting SBT, and an RSBI >105 during SBT was used as a criterion to discontinue SBT. However, in our study, rather than RSBI, RR ≥35 was the criterion to discontinue SBT. Similarly, our results were in contradiction to those of Farghaly and Hasan[15] who reported no statistically significant difference between the median RSBI of patients who were successfully extubated and those who failed extubation. However, this difference could be due to the reason that the failure group consisted of patients who passed as well as failed SBT, as against the present study where only SBT-passing patients were included.

In our study, although we found DEx values to be significantly higher in patients with successful weaning as compared to failed weaning, it was a poor predictor of weaning outcome, and this was comparable to the results by Palkar et al.,[10] Khan et al.,[12] and Farghaly and Hasan.[15] However, Huang et al.[13] and Bahgat et al.[18] got DEx to be a good predictor of weaning outcome in their studies on geriatric patients and premature infants respectively.

We found DTF and DCV to be a good predictor of weaning outcome similar to that proposed by Banerjee and Mehrotra,[11] Bahgat et al.,[18] Farghaly and Hasan,[15] and Ferrari et al.,[17] although the cutoff value for DTF in the last study in the series was higher compared to our study, which could be due to the fact that measurements in their study were done during deep inspiration to total lung capacity and expiration to residual volume and all measurements in our study were done during normal quiet breathing. However, Palkar et al.[10] suggested DCV to be a poor indicator of extubation failure. The authors studied only the change in DCV between A/C mode of ventilation and SBT, whereas we measured the DCV during SBT. A/C mode of ventilation is associated with a lesser mechanical load on the diaphragm and a lesser RR when compared to SBT.

Our study demonstrated RSBI to be the best predictor of weaning outcome when compared with the ultrasonographic diaphragmatic parameters for weaning outcome and a positive correlation of the RSBI and a negative correlation of the diaphragmatic parameters with the duration of ventilation.

CONCLUSION

Weaning indices such as RSBI and assessment of diaphragmatic functions are critical in aiding the intensivists in deciding the appropriate time of weaning the critically ill patients from the ventilator and predicting the weaning outcomes. We concluded RSBI to be the foremost indicator of weaning outcomes, with a cutoff value of 90 breaths/min/L. Among the ultrasonographic diaphragmatic parameters, DCV was found to be the best predictor, followed by DTF, and may serve as valuable adjuncts to RSBI in predicting weaning outcomes.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.