Midaxillary-to-midclavicular diaphragmatic mobility ratio predicts weaning success in invasive mechanical ventilation: an observational cohort study

Introduction

Invasive mechanical ventilation (IMV) is a core life-saving intervention in critical care, yet the rate of weaning failure remains as high as 20–35%, which is significantly associated with prolonged intensive care unit (ICU) stay, increased healthcare resource consumption, and elevated mortality (1). As the most important inspiratory muscle, the diaphragm contributes 60–79% of ventilatory function. However, factors such as IMV (especially controlled ventilation modes), analgesics and sedatives, and malnutrition can rapidly induce ventilator-associated diaphragmatic dysfunction (VIDD). In recent years, significant progress has been made in understanding the molecular mechanisms of VIDD. A review by Powers systematically showed that increased mitochondrial reactive oxygen species generation within diaphragmatic myofibers accelerates proteolysis and inhibits protein synthesis via redox signaling cascades, collectively promoting the development of diaphragmatic atrophy and contractile dysfunction (2).

Diaphragmatic ultrasound has become an important tool for assessing diaphragmatic function due to its non-invasive, bedside, and reproducible advantages. Current clinical studies have mainly focused on diaphragmatic mobility (DM) and diaphragmatic thickening fraction (DTF). A meta-analysis of 26 studies involving 1,204 patients showed that the pooled sensitivity and specificity of DM for predicting weaning success were 0.80 and 0.80, respectively, whereas those of DTF were 0.85 and 0.75, indicating moderate predictive value for both parameters (3). However, there is regional heterogeneity in anatomy and kinematics between the diaphragmatic dome and the posterior wall. Moreover, most existing studies have measured only a single anatomical plane, which makes it difficult to account for inter-individual baseline differences and fails to fully reflect the global functional status of the diaphragm (4).

To address this gap, the present study for the first time proposes the ratio of DM measured at the midaxillary line (MAL, representing dome mobility) to that measured at the midclavicular line (MCL, representing posterior wall mobility) as a novel individualized parameter for predicting weaning success. The physiological rationale for this ratio is grounded in respiratory mechanics: the diaphragm’s pressure-generating capacity depends on the integrated action of its constituent regions, which exhibit heterogeneous muscle fiber type composition—the domed region is predominantly type I (slow-twitch, fatigue-resistant) fibers, whereas the posterior wall (assessed at MCL) contains a higher proportion of type II (fast-twitch, more susceptible to disuse atrophy) fibers. Accordingly, isolated single-site measurements may be biased by inter-individual variability and regional susceptibility to VIDD, whereas the ratio between the two regions could serve as an internal control, reflecting relative functional reserve and potentially offering superior predictive value for weaning outcomes. We therefore hypothesize that the MAL/MCL ratio will outperform traditional single-site measurements (MAL, MCL) and the partial pressure of oxygen/fraction of inspired oxygen (PaO₂/FiO₂) ratio in predicting weaning success from IMV. To date, no established reference standard exists for this ratio; therefore, we compared its predictive performance against traditional single-site measurements (MAL, MCL) and PaO₂/FiO₂ as benchmarks. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0933/rc) (5,6).

Methods

Study design and participants

This observational cohort study was conducted from 18 September 2025 to 10 April 2026 in the ICU of Zhangzhou Affiliated Hospital of Fujian Medical University. The study strictly adhered to the requirements of the Declaration of Helsinki and its subsequent amendments. The study protocol was approved by the Ethics Committee of Zhangzhou Affiliated Hospital of Fujian Medical University (No. 2025LWB429) and registered on the ClinicalTrials.gov PRS website (No. NCT07235956). Since this study involved no commercial interests, all data analyses were anonymized, and no interventions were performed, the requirement for informed consent was waived by the Ethics Committee. The inclusion criteria were as follows: age ≥14 years, admission to the ICU requiring IMV due to medical necessity, and, to ensure completeness of data collection, a duration of ventilator-assisted breathing >72 hours. The exclusion criteria were as follows: (I) severe traumatic brain injury with absence of spontaneous consciousness (as spontaneous breathing trial (SBT) requires the patient to be conscious); (II) pre-existing severe diaphragmatic dysfunction; (III) patients with advanced malignancy; and (IV) unclear ultrasound imaging.

Sample size was determined using power analysis (α=0.05, β=0.2, power =80%). Based on a pilot study of 2011–2014 patients from our ICU (unpublished), we assumed a 30% difference in the primary outcome (MAL/MCL ratio at SBT) between the success and failure groups. With an estimated effect size of 0.8, the required sample was 98 patients. To account for potential dropouts, we enrolled 112 patients.

Data collection and variables

Demographic and clinical data were collected from electronic medical records, including age, sex, Acute Physiology and Chronic Health Evaluation version II (APACHE II) score, and length of ICU stay. The primary grouping variable was the success or failure of the first attempt at removing IMV. Weaning success was defined as the ability to maintain spontaneous breathing for at least 48 hours after extubation without requiring non-invasive ventilation or reintubation. Weaning failure was defined as the need for reintubation or resumption of IMV within 48 hours after extubation, or the inability to tolerate the SBT leading to termination of the weaning attempt. Ultrasound measurements were performed at three time points: the first 24-hour examination during hospitalization (24h), 72 hours after admission (72h), and the examination during the SBT. The following parameters were assessed: MAL, MCL, left ventricular ejection fraction (LVEF), prealbumin (Pa), PaO₂/FiO₂, and serum potassium (K⁺).

Ultrasound examination protocol

All ultrasound examinations were performed by two experienced sonographers (each with >5 years of critical care ultrasound experience) using a Sonosite EDGE device (FUJIFILM Sonosite, Bothell, WA, USA) with a 2–5 MHz convex array probe. The sonographers were blinded to the patients’ clinical data and outcomes.

Examinations were conducted in accordance with the Technical Specification for Clinical Application of Critical Ultrasound in China. “Intra-rater and inter-rater reliability were assessed in a random sample of 20 patients by having two independent sonographers repeat measurements on the same day and one sonographer repeat measurements 24 hours later. The intraclass correlation coefficients (ICCs) for MAL were 0.94 (intra-rater) and 0.90 (inter-rater); for MCL, 0.92 and 0.88, respectively. All these patients were treated according to the standard of care.

The patients were placed in a supine position with the head of the bed elevated 0–30°, both arms positioned alongside the body, and the abdomen relaxed. The ultrasound device was equipped with M-mode functionality. The right diaphragm (using the liver as an acoustic window) was preferred, and the left diaphragm (using the spleen as an acoustic window) was also examined. At least three consecutive respiratory cycles were measured, and the mean value was calculated (7). The following description takes the ultrasound measurement of the right diaphragm as an example:

• For MAL measurement, the transducer was placed at the right midaxillary line, between the 7^th and 9^th intercostal spaces, with the probe marker oriented cephalad. Scanning along the intercostal space, the hyperechoic bright line of the diaphragmatic dome was identified in B-mode, with the ultrasound beam directed almost perpendicularly to the dome. In M-mode, the spatial position of the dome was marked at end-expiration, and again at the same point at end-inspiration. The linear distance between these two points was defined as DM (8) (Figure 1).
• For MCL measurement, the transducer was placed at the right subcostal region along the midclavicular line, with the probe marker oriented cephalad and the angle between the probe and the abdominal wall plane >45°. The diaphragm was identified, and the transducer was adjusted so that the ultrasound beam was directed perpendicularly to the posterior third of the diaphragmatic dome, where the echo signal was strongest and motion was most clearly visualized (2). Switching to M-mode, the sampling line was aligned perpendicularly to the diaphragmatic echo line. At least three respiratory cycles were stably recorded, and the vertical distance between the lowest point of the curve at end-expiration and the highest point at end-inspiration was taken as DM (9) (Figure 2).
• Patients were considered ready for SBT when they met standard criteria: conscious state, hemodynamic stability, PaO₂/FiO₂ ≥150; minimal doses of vasopressors, analgesics, and sedatives were allowed. The SBT was performed using pressure support ventilation with a pressure support of 8–10 cmH₂O and positive end-expiratory pressure of 5 cmH₂O for 30 minutes. It is recognized that positive pressure ventilation may affect diaphragmatic excursion; therefore, all SBT ultrasound measurements were taken under identical ventilator settings to ensure consistency.

Figure 1 Measurement of MAL. For MAL measurement, the transducer was placed at the right MAL, between the 7^th and 9^th intercostal spaces, with the probe marker oriented cephalad. Scanning along the intercostal space, the hyperechoic bright line of the diaphragmatic dome was identified in B-mode, with the ultrasound beam directed almost perpendicularly to the dome. In M-mode, the spatial position of the dome was marked at end-expiration, and again at the same point at end-inspiration. The linear distance between these two points was defined as DM. DM, diaphragmatic mobility; FPS, frames per second; L, line; MAL, diaphragmatic mobility measured at the midaxillary line; RMAL, right MAL.

Figure 2 Measurement of MCL. For MCL measurement, the transducer was placed at the right subcostal region along the MCL, with the probe marker oriented cephalad and the angle between the probe and the abdominal wall plane >45°. The diaphragm was identified, and the transducer was adjusted so that the ultrasound beam was directed perpendicularly to the posterior third of the diaphragmatic dome, where the echo signal was strongest and motion was most clearly visualized. Switching to M-mode, the sampling line was aligned perpendicularly to the diaphragmatic echo line. At least three respiratory cycles were stably recorded, and the vertical distance between the lowest point of the curve at end-expiration and the highest point at end-inspiration was taken as DM. DM, diaphragmatic mobility; FPS, frames per second; L, line; MCL, diaphragmatic mobility measured at the midclavicular line; RMCL, right MCL.

Statistical analysis

Data were analyzed using the software SPSS 25.0 (IBM Corp., Armonk, NY, USA). The Shapiro-Wilk test was used to assess the normality of continuous variables. Normally distributed variables were presented as mean ± standard deviation and compared using Student’s t-test. Non-normally distributed variables were presented as median (interquartile range) and compared using the Mann-Whitney U test. Categorical variables were presented as frequencies (percentages) and compared using the Chi-squared test or Fisher’s exact test.

Repeated measures analysis of variance (ANOVA) with Bonferroni correction was used for pairwise comparisons of ultrasound parameters at different time points within each group. Receiver operating characteristic (ROC) curves were constructed to evaluate the predictive value of ultrasound parameters for 28-day mortality. The area under the curve (AUC) was calculated, and optimal cut-off values were determined using the Youden index.

Calibration of the logistic regression models was assessed to evaluate the agreement between predicted and observed probabilities of weaning failure. For each predictor, a univariable logistic regression model was constructed with weaning failure as the outcome. The outcome was coded as 1 for weaning failure and 0 for successful weaning. Predicted probabilities were obtained from the fitted logistic models.

Model calibration was evaluated using a locally estimated scatterplot smoothing (LOESS)-smoothed calibration curve, the Brier score, and the Hosmer-Lemeshow goodness-of-fit test. The calibration curve compared the predicted probability of weaning failure with the observed probability. The Brier score was used as a measure of overall prediction error, with lower values indicating better accuracy. The Hosmer-Lemeshow test was used to assess goodness of fit, with a significant result suggesting imperfect calibration.

For graphical presentation, the calibration curve of the SBT MAL/MCL ratio was displayed because this parameter showed the best discriminatory performance in ROC analysis.

A two-sided P value <0.05 was considered statistically significant. To address potential confounders, multivariate logistic regression analysis was performed, adjusting for age and APACHE II.

Results

Patient characteristics

A total of 112 patients were included in this study, comprising 79 males and 33 females. Based on the success or failure of the first attempt to remove IMV, patients were divided into two groups: a success group (n=76) and a failure group (n=36). Baseline characteristics, including age, APACHE II score, 24h-PaO₂/FiO₂, 24h-MAL, 24h-MCL, 24h-MAL/MCL, 24h-Pa, 24h-LVEF, and 24h-K⁺ were compared between the two groups. A statistically significant difference was observed in 24h-MAL/MCL between the success and failure groups (P<0.05) (Table 1).

In the examination conducted on the 72^nd hour after hospitalization, there was no significant difference (P>0.05) between the two groups in terms of 72h-PaO₂/FiO₂, 72h-MAL, 72h-MCL, 72h-MAL/MCL, 72h-Pa, 72h-LVEF, and 72h-K⁺. There were no statistically significant differences between the two groups in terms of Pa, LVEF, K⁺, or PaO₂/FiO₂, indicating that during standardized treatment, the common parameters affecting diaphragm function were managed uniformly. Furthermore, no significant differences were observed between the two groups in 72h-MAL, 72h-MCL, or the 72h-MAL/MCL ratio. This may be attributable to incomplete resolution of the underlying disease and the continuous use of standardized sedative and analgesic agents (fentanyl and midazolam), which are known to suppress diaphragmatic activity and thereby mask functional differences (Table 2).

There was no significant difference (P>0.05) between the two groups in terms of SBT-PaO₂/FiO₂, SBT-MAL, SBT-MCL, SBT-Pa, SBT-LVEF, and SBT-K⁺. A statistically significant difference was observed in SBT-MAL/MCL between the success and failure groups (P<0.05) (Table 3).

ROC curve analysis

In this study, weaning success was defined as the positive outcome. During the spontaneous breathing test (Figure 3), ROC curve analysis was performed to evaluate the predictive value of PaO₂/FiO₂, MAL, MCL, and MAL/MCL. Table 4 shows the diagnostic accuracy of each parameter for predicting successful extubation, including the AUC, optimal cut-off value, Youden index, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The results showed that the AUC of MCL was 0.74 [95% confidence interval (CI): 0.647–0.839], which was superior to that of MAL (0.636; 95% CI: 0.540–0.725) and PaO₂/FiO₂ (0.741; 95% CI: 0.650–0.819). However, all three parameters exhibited only moderate predictive performance. In contrast, the MAL/MCL ratio achieved an AUC of 0.945. At the optimal cut-off value of 1.11, the sensitivity was 82.89%, specificity 94.44%, Youden index 0.802, PPV 96.9%, and NPV 72.3%, indicating that this ratio has near-perfect discriminatory ability for distinguishing successful from failed weaning. Pairwise comparison of the ROC curves (Table 5) further revealed that the AUC of the MAL/MCL ratio was significantly higher than that of MAL (difference =0.309; P<0.001), MCL (difference =0.181; P<0.001), and PaO₂/FiO₂ (difference =0.204; P<0.001). The AUC of MCL was also significantly higher than that of MAL (difference =0.128; P<0.001). No significant differences were observed between MAL and PaO₂/FiO₂ (P=0.154) or between MCL and PaO₂/FiO₂ (P=0.751).

Figure 3 ROC curves of MAL, MCL, MAL/MCL, PaO₂/FiO₂ at the SBT. The results showed that the AUC of MCL was 0.74 (95% CI: 0.647–0.839), which was superior to that of MAL (0.636; 95% CI: 0.540–0.725), and PaO₂/FiO₂ (0.741; 95% CI: 0.650–0.819). However, all three parameters exhibited only moderate predictive performance. In contrast, the MAL/MCL ratio achieved an AUC of 0.945. At the optimal cut off value of 1.11, the sensitivity was 82.89%, specificity 94.44%, Youden index 0.802, PPV 96.9%, and NPV 72.3%, indicating that this ratio has near-perfect discriminatory ability for distinguishing successful from failed weaning. AUC, area under the curve; CI, confidence interval; MAL, diaphragmatic mobility measured at the midaxillary line; MCL, diaphragmatic mobility measured at the midclavicular line; NPV, negative predictive value; PaO₂/FiO₂, arterial partial pressure of oxygen/fraction of inspired oxygen; PPV, positive predictive value; ROC, receiver operating characteristic; SBT, spontaneous breathing trial.

The predicted probability of weaning failure was derived from a univariable logistic regression model using the SBT MAL/MCL ratio. The diagonal dashed line represents perfect calibration, and the solid curve represents the LOESS-smoothed observed probability. Although the SBT MAL/MCL ratio showed excellent discrimination, the calibration curve demonstrated deviation from the ideal line. The Brier score was 0.189, and the Hosmer-Lemeshow goodness-of-fit test suggested imperfect calibration of the univariable model (Figure 4).

Figure 4 Calibration curve of the SBT MAL/MCL-based logistic model for predicting weaning failure. The predicted probability of weaning failure was derived from a univariable logistic regression model using the SBT MAL/MCL ratio. The diagonal dashed line represents perfect calibration, and the solid curve represents the LOESS-smoothed observed probability. Although the SBT MAL/MCL ratio showed excellent discrimination, the calibration curve demonstrated deviation from the ideal line. The Brier score was 0.189, and the Hosmer-Lemeshow goodness-of-fit test suggested imperfect calibration of the univariable model. AUC, area under the curve; HL, Hosmer-Lemeshow; LOESS, locally estimated scatterplot smoothing; MAL, diaphragmatic mobility measured at the midaxillary line; MCL, diaphragmatic mobility measured at the midclavicular line; SBT, spontaneous breathing trial.

Discussion

The weaning failure rate remains persistently high in patients undergoing IMV, with approximately 30% of patients developing ventilator dependence, leading to prolonged ICU stay, increased medical costs, and elevated mortality. In patients with difficult weaning from mechanical ventilation, the assessment of diaphragmatic function has been highlighted as of paramount importance (10). As the primary inspiratory muscle, the functional status of the diaphragm directly affects weaning success. A previous magnetic resonance imaging (MRI) study (11) demonstrated that the mobility of the posterior region of the normal diaphragm is significantly greater than that of the domed top (4.5±0.2 vs. 2.6±0.1 cm), and the posterior mobility is approximately 1.73 times that of the domed region. Therefore, we hypothesize that under pathological conditions, measuring MAL and MCL by diaphragmatic ultrasound is superior to single-site mobility, a superiority that may arise from the following two mechanisms. First, the dome (assessed by MAL) and the posterior wall (assessed by MCL) have different fiber type compositions: the dome is richer in type I fibers adapted for sustained low-force contractions, whereas the posterior wall has a higher proportion of type II fibers that are more vulnerable to disuse atrophy. Second, the zone of apposition—the area where the diaphragm abuts the rib cage—affects regional mechanics; the posterior wall lies largely within this zone, making its excursion more dependent on abdominal pressure and muscle length. The ratio captures the balance between these two regions, which is lost when only absolute mobility is measured.

Previous studies on diaphragmatic ultrasound have mainly focused on DM and DTF, both of which are limited by large inter-individual variability in baseline values and inconsistent contractile amplitudes across different anatomical regions, leading to considerable fluctuations in predictive performance across studies. A systematic review and meta-analysis published in 2022, including 12 studies with 828 patients, showed that the pooled AUC of DM for predicting weaning success was 0.84 (95% CI: 0.80–0.87), and that of DTF was 0.85 (95% CI: 0.81–0.88) (12). Another meta-analysis published in 2024 further confirmed that the AUC ranged from 0.70 to 0.90 for DM and from 0.75 to 0.95 for DTF (13). These findings indicate that although the absolute mobility measured at a single site has some predictive value, it is far from reaching the ideal level.

This study, for the first time, proposes the ratio of DM measured at the MAL (representing dome mobility) to that measured at the MCL (representing posterior wall mobility) as a predictor of weaning success, effectively eliminating inter-individual differences in baseline diaphragmatic function. The results showed that the MAL/MCL ratio measured during SBT achieved an AUC of 0.945 for predicting weaning success, with a specificity of 94.44% at a cutoff value >1.11, which was significantly superior to traditional parameters such as MAL, MCL, and PaO₂/FiO₂. This result remains advantageous compared with the AUC of 0.986 reported by Kumbhar et al. in 2025 for pre-intubation DM in predicting complex weaning (14) and the AUC of 0.924 reported by Luo et al. for predicting weaning success (15). This excellent performance is attributed to the “self-calibration” effect of the ratio parameter on inter-individual variability: the ratio eliminates inter-individual differences because both MAL and MCL are influenced by common factors such as overall diaphragmatic strength, thoracic geometry, and ventilator settings. By dividing one by the other, these shared sources of variability cancel out, leaving a value that reflects the relative function of the dome vs. the posterior wall, thereby possessing greater pathophysiological specificity. This is analogous to using a ratio index in other physiological contexts [e.g., forced expiratory volume in 1 second/forced vital capacity (FEV₁/FVC)]. Therefore, a patient with a low baseline mobility in both regions can still have a normal ratio, and it is the deviation from the expected ratio that signals regional diaphragmatic dysfunction. Nevertheless, the excellent performance of the MAL/MCL ratio in this single-center study should be interpreted with caution. Our ICU follows a strict weaning protocol, and the patient population (excluding neurological and prolonged ventilation cases) may not represent the full heterogeneity of ICUs elsewhere. Therefore, the reported AUC of 0.945 is likely an overestimate of the true performance in unselected populations.

This study further revealed the dynamic trends of the MAL/MCL ratio between the weaning success and failure groups. This finding has important clinical implications: in the success group, the dome mobility increased faster relative to the posterior wall, suggesting that the dome muscle fibers are more sensitive to rehabilitative stimuli; in the failure group, the posterior wall mobility decreased more prominently relative to the dome, which may reflect a higher susceptibility of the posterior wall region to ventilator-induced diaphragmatic injury, consistent with previous findings (11). A prospective study by Ağaoğlu et al. confirmed that diaphragmatic thickness decreased significantly within 48 hours after the initiation of mechanical ventilation, and the magnitude of decrease was associated with ventilation duration and weaning failure (16). Subsequent studies further found that the decrease in diaphragmatic thickness is not uniform, with more pronounced changes in the posterior wall region (17). The present study validates this regional difference from the perspective of DM and, for the first time, quantifies it as a ratio parameter for weaning prediction.

Traditional weaning predictors such as the rapid shallow breathing index (RSBI) often have a low PPV, frequently leading to unnecessary reintubation or delayed weaning. Song et al. combined diaphragmatic ultrasound with RSBI and proposed the diaphragmatic-RSBI and DTF-RSBI, which predicted weaning failure with AUCs of 0.813 and 0.859, respectively; however, certain false-positive rates remained (18). The specificity of the MAL/MCL ratio at a cutoff value >1.11 during SBT reached 94.44% in this study, which allows clinicians to make weaning decisions with greater confidence, particularly in patient populations at high risk of weaning failure.

A prospective observational cohort study suggested that a combination of ultrasound parameters of the heart, lungs, and diaphragm could predict the outcome of critically ill patients when they were weaned off mechanical ventilation (19). We performed that the exploratory linear regression analyses showed weak positive correlations between K⁺ and DM (MAL, MCL) and between LVEF and PaO₂/FiO₂. These findings are consistent with the notion that maintaining K⁺ concentration, LVEF, and other factors are synergistic factors influencing weaning success. Previous studies have identified LVEF <40% as a risk factor for weaning failure (20); in the present study, LVEF in both groups was >50% (51.85±4.43 vs. 55.83±3.79), and therefore no statistically significant difference was observed. As stated in clinical guidelines, “Correction of severe electrolyte disturbances” has become a standard prerequisite in all weaning protocols (21); accordingly, K⁺ levels in the two groups of this trial were well maintained (3.99±0.33 vs. 4.09±0.27 mmol/L), and no statistically significant difference was found.

Notably, at the time of SBT, previous studies have shown that when MAL >1 cm, the weaning success rate can reach 70% (22). However, our findings showed that some patients with both MAL and MCL >2 cm still experienced weaning failure when their MAL/MCL ratio was below 1.11, whereas other patients with both MAL and MCL <2 cm achieved successful weaning when their MAL/MCL ratio exceeded 1.11. The possible explanation is that although both MAL and MCL mobility are increased, this may indicate unfavorable factors such as excessive inspiratory effort or recruitment of accessory respiratory muscles. In contrast, the MAL/MCL ratio better reflects the degree of coordinated contraction between different anatomical regions of the diaphragm. This finding may provide a useful predictor for patients with pre-existing diaphragmatic atrophy, such as those with chronic obstructive pulmonary disease (COPD). Therefore, this explains why the absolute values of MAL and MCL showed no statistically significant differences between the success and failure groups (P=0.686 and 0.205, respectively at SBT), whereas their ratio differed highly significantly (P=0.047). Furthermore, in the weaning success group, the MAL/MCL ratio measured at 24 hours after admission was also statistically different from that measured at SBT, suggesting that this ratio may serve as an early predictor of weaning success. This phenomenon further reinforces the advantage of the ratio parameter: when inter-individual baseline differences are large, single-site mobility is insufficient to distinguish weaning outcomes, whereas after eliminating individual differences by taking the ratio, the underlying pathophysiological differences become apparent. This finding also explains why previous studies on DM have shown considerable heterogeneity (23,24).

This study had several limitations. (I) As this was a single-center study, the near-perfect AUC (>0.94) may be explained by the modest sample size (n=112) and the small number of patients with underlying diaphragmatic dysfunction (e.g., acute exacerbation of COPD) enrolled, which could have introduced selection bias and limited the generalizability of the findings. Hence, the reported AUC and cutoff value need to be validated in larger cohorts. (II) DTF was not simultaneously assessed in this study, although DTF and DM provide complementary information (25); future studies could explore a combined model incorporating the MAL/MCL ratio and DTF. (III) Patients with neurological critical illness or prolonged mechanical ventilation (>14 days) were not included, limiting the generalizability of the findings. (IV) Although this study proposed that a cutoff value >1.11 for the MAL/MCL ratio is a reliable predictive parameter, whether an upper limit of the ratio exists requires further investigation. (V) The clinicians managing weaning were not blinded to the ultrasound results because the ultrasound measurements were performed as part of clinical care. This may have introduced bias, and we cannot exclude the possibility that knowledge of the ratio influenced the decision to extubate.

The present study found that the SBT MAL/MCL ratio had excellent discriminatory performance for predicting weaning failure. However, calibration analysis showed that the univariable logistic model based on SBT MAL/MCL was not optimally calibrated. Therefore, the clinical value of SBT MAL/MCL may lie mainly in identifying patients at high risk of weaning failure rather than in providing precise individualized risk probabilities. Future studies with larger sample sizes and external validation cohorts are needed to determine whether multivariable models incorporating SBT MAL/MCL and relevant clinical variables can achieve better calibration and more reliable individualized risk prediction.

Conclusions

MAL/MCL is a novel diaphragmatic ultrasound parameter, which by eliminating inter-individual differences in baseline diaphragmatic function, indicated excellent discriminatory ability in this single-center cohort. It is recommended that clinicians consider measuring MAL and MCL and calculating their ratio during weaning assessment, but calibration analysis revealed that the clinical value of the SBT MAL/MCL ratio lies primarily in identifying patients at high risk of weaning failure, rather than in providing precise individualized risk probabilities, so these findings require validation in larger, multicenter studies before routine clinical adoption.

Acknowledgments

We would like to thank all the patients and their families who participated in this study.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0933/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0933/dss

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0933/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study protocol was approved by the Ethics Committee of Zhangzhou Affiliated Hospital of Fujian Medical University (No. 2025LWB429), and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Khazaei O, Laffey CM, Sheerin R, McNicholas BA, Pham T, Heunks L, Bellani G, Brochard L, Simpkin AJ, Laffey JGWEAN SAFE Investigators. Impact of comorbidities on management and outcomes of patients weaning from invasive mechanical ventilation: insights from the WEAN SAFE study. Crit Care 2025;29:114. [Crossref] [PubMed]
2. Powers SK. Ventilator-induced diaphragm dysfunction: phenomenology and mechanism(s) of pathogenesis. J Physiol 2024;602:4729-52. [Crossref] [PubMed]
3. Parada-Gereda HM, Tibaduiza AL, Rico-Mendoza A, Molano-Franco D, Nieto VH, Arias-Ortiz WA, Perez-Terán P, Masclans JR. Effectiveness of diaphragmatic ultrasound as a predictor of successful weaning from mechanical ventilation: a systematic review and meta-analysis. Crit Care 2023;27:174. [Crossref] [PubMed]
4. Santana PV, Cardenas LZ, de Albuquerque ALP, de Carvalho CRR, Caruso P. Diaphragmatic ultrasound findings correlate with dyspnea, exercise tolerance, health-related quality of life and lung function in patients with fibrotic interstitial lung disease. BMC Pulm Med 2019;19:183. [Crossref] [PubMed]
5. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453-7. [Crossref] [PubMed]
6. Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, Timpson NJ, Higgins JPT, Dimou N, Langenberg C, Loder EW, Golub RM, Egger M, Davey Smith G, Richards JB. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ 2021;375: [Crossref] [PubMed]
7. Haaksma ME, Smit JM, Boussuges A, Demoule A, Dres M, Ferrari G, Formenti P, Goligher EC, Heunks L, Lim EHT, Mokkink LB, Soilemezi E, Shi Z, Umbrello M, Vetrugno L, Vivier E, Xu L, Zambon M, Tuinman PR. EXpert consensus On Diaphragm UltraSonography in the critically ill (EXODUS): a Delphi consensus statement on the measurement of diaphragm ultrasound-derived parameters in a critical care setting. Crit Care 2022;26:99. [Crossref] [PubMed]
8. Le Neindre A, Philippart F, Luperto M, Wormser J, Morel-Sapene J, Aho SL, Mongodi S, Mojoli F, Bouhemad B. Diagnostic accuracy of diaphragm ultrasound to predict weaning outcome: A systematic review and meta-analysis. Int J Nurs Stud 2021;117:103890. [Crossref] [PubMed]
9. Song F, Liu H, Ma H, Chen X, Wang S, Qin T, Liang H, Huang D. AI Model Based on Diaphragm Ultrasound to Improve the Predictive Performance of Invasive Mechanical Ventilation Weaning: Prospective Cohort Study. JMIR Form Res 2025;9:e72482. [Crossref] [PubMed]
10. Fritsch SJ, Siemer AG, Dreher M, Simon TP, Marx G, Bickenbach J. Diaphragm ultrasound in patients with prolonged weaning from mechanical ventilation. Quant Imaging Med Surg 2024;14:3248-63. [Crossref] [PubMed]
11. Dong Z, Zhao M, Sun X, Huang B. Beyond feasibility: standardization and reproducibility as the next frontier for respiratory muscle ultrasound. Crit Care 2025;29:314. [Crossref] [PubMed]
12. Mahmoodpoor A, Fouladi S, Ramouz A, Shadvar K, Ostadi Z, Soleimanpour H. Diaphragm ultrasound to predict weaning outcome: systematic review and meta-analysis. Anaesthesiol Intensive Ther 2022;54:164-74. [Crossref] [PubMed]
13. Marques MR, Pereira JM, Paiva JA, de Casasola-Sánchez GG, Tung-Chen Y. Ultrasonography to Access Diaphragm Dysfunction and Predict the Success of Mechanical Ventilation Weaning in Critical Care: A Narrative Review. J Ultrasound Med 2024;43:223-36. [Crossref] [PubMed]
14. Kumbhar PG, Tejpal M, Karna ST, Singh P, Saigal S, Waindeskar V. Sonographic diaphragmatic parameters as a predictor of weaning failure in critically ill patients in need of invasive mechanical ventilation: a prospective observational cohort study. Monaldi Arch Chest Dis 2025; Epub ahead of print. [Crossref]
15. Luo W, Zhang H, Chen Y, Luo W, Lin X. The 30-min diaphragm movement change rate for predicting weaning success in severe pneumonia patients requiring invasive ventilation. Front Med (Lausanne) 2025;12:1595814. [Crossref] [PubMed]
16. Ağaoğlu RT, Bayrak AÇ, Akbulut ÖV, Kindan A, Tüfekçioğlu E, Kaya M, Özgürlük İ, Vural Yilmaz Z, Yakut Yücel K. Diaphragmatic Function in Late-Onset FGR Insights From Ultrasound-Based Quantification: Insights from Ultrasound-Based Quantification. J Ultrasound Med 2025;44:1893-901. [Crossref] [PubMed]
17. Song J, Luo Q, Lai X, Hu W, Yu Y, Wang M, Yang K, Chen G, Chen W, Li Q, Hu C, Gong S. Combined cardiac, lung, and diaphragm ultrasound for predicting weaning failure during spontaneous breathing trial. Ann Intensive Care 2024;14:60. [Crossref] [PubMed]
18. Song J, Qian Z, Zhang H, Wang M, Yu Y, Ye C, Hu W, Gong S. Diaphragmatic ultrasonography-based rapid shallow breathing index for predicting weaning outcome during a pressure support ventilation spontaneous breathing trial. BMC Pulm Med 2022;22:337. [Crossref] [PubMed]
19. Zhou Q, Xu Y, Chen J, Gu Q, Kong W. Prediction of weaning outcomes from mechanical ventilation in critically ill patients based on the combination of ultrasound parameters of the heart, lung, and diaphragm: a prospective observational cohort study. Quant Imaging Med Surg 2025;15:7697-710. [Crossref] [PubMed]
20. Sanfilippo F, Di Falco D, Noto A, Santonocito C, Morelli A, Bignami E, Scolletta S, Vieillard-Baron A, Astuto M. Association of weaning failure from mechanical ventilation with transthoracic echocardiography parameters: a systematic review and meta-analysis. Br J Anaesth 2021;126:319-30. [Crossref] [PubMed]
21. Trudzinski FC, Neetz B, Bornitz F, Müller M, Weis A, Kronsteiner D, Herth FJF, Sturm N, Gassmann V, Frerk T, Neurohr C, Ghiani A, Joves B, Schneider A, Szecsenyi J, von Schumann S, Meis J. Risk Factors for Prolonged Mechanical Ventilation and Weaning Failure: A Systematic Review. Respiration 2022;101:959-69. [Crossref] [PubMed]
22. Sarkar K, Chaudhury M, Bahinipati P, Das S. Assessment of Diaphragmatic Dysfunction in Mechanically Ventilated Patients with Ultrasonography. Ann Afr Med 2025;24:22-7. [Crossref] [PubMed]
23. Karageorgos V, Proklou A, Vaporidi K. Lung and diaphragm protective ventilation: a synthesis of recent data. Expert Rev Respir Med 2022;16:375-90. [Crossref] [PubMed]
24. Li R, Zhou Y, Chen W, Lyu L, Qiu G, Pan C, Tang Y. Speckle tracking ultrasound as a new tool to predict the weaning outcome of mechanical ventilation patients: a prospective observational study. Front Med (Lausanne) 2024;11:1449938. [Crossref] [PubMed]
25. Rauseo M, Cotoia A, Padovano FP, Squiccimarro E, Paparella D, Vetuschi P, Da Lima S, Mollica G, Mirabella L, Paternoster G, Cinnella G. Weaning From Mechanical Ventilation in Cardiac Surgery Patients: Current Strategies, Monitoring Innovations, and Future Perspectives. J Cardiothorac Vasc Anesth 2026; Epub ahead of print. [Crossref]