Handgrip Force and Maximum Inspiratory and Expiratory Pressures in Critically Ill Patients With a Tracheostomy

This was a retrospective observational study conducted from May 2017 through January 2018 with a sample of 20 ICU patients. Data were collected from the patients’ medical records.

The study included patients more than 18 years of age who were hospitalized in the ICU for more than 11 days (ICU-AW occurs after 10 days of hospitalization^23,24 ). Patients included were hemodynamically stable and were able to perform commands and breathing exercises (Glasgow Coma Scale score of 15). Patients included in the study also had a good level of communication, a level of consciousness sufficient to execute orders, and intact limbs. Patients were excluded if they had any neurological syndrome (central or peripheral), craniocerebral injury, or condition that could affect peripheral muscle strength or were taking any medication that could affect perception or muscle tone.

The MIP and the MEP are valid indicators of the condition of the respiratory muscles. The most commonly used and reliable tool for recording MIP and MEP is the membrane manometer.²⁵  We assessed HGF with a handheld dynamometer (Baseline hydraulic dynamometer, Medline Industries).²⁶ 

We collected the following data from patients’ medical records: MIP, MEP, HGF, age, weight, height, Glasgow Coma Scale score,²⁷  creatinine level, receipt of inotropic medications, PaO₂, platelet count, bilirubin level, oxygen supply (fraction of inspired oxygen), and day of and reason for hospitalization in the ICU. The Sequential Organ Failure Assessment (SOFA) was used to determine the severity of each patient’s condition.

Descriptive statistics are presented as means with SDs, medians with interquartile ranges, or percentages as appropriate. The Shapiro-Wilk test was used to determine whether the variables of interest followed a normal distribution. Correlations were examined with the Pearson correlation coefficient (r). A correlation coefficient of less than 0.20 indicates a very weak correlation; of 0.21 to 0.40, a weak correlation; of 0.41 to 0.60, a moderate correlation; of 0.61 to 0.80, a strong correlation; and of 0.81 to 1, a very strong correlation.^28,29  We used linear regression analysis to predict HGF from MIP and MEP. The regression analysis was performed in hierarchical order to control for the confounding influence of sex on HGF. We constructed 3 models. Model 1 used both MIP and MEP to predict HGF, model 2 used only MEP to predict HGF, and model 3 used only MIP to predict HGF. Sex was included in the first block of predictors for each of the 3 models. We used χ² tests of independence to compare the frequency of strong MEP with strong and weak HGF. The significance level was set at a P of .05. We used SPSS version 20.0 (IBM) for all data analyses.

Participants’ demographics and descriptive data regarding day of and primary reason for ICU hospitalization are presented in Table 1. Age, sex, body mass index (calculated as weight in kilograms divided by height in meters squared), height, and weight were not significantly correlated with MIP or MEP (P > .05). Handgrip force was not significantly correlated with SOFA score (r = 0.04, P > .05), ICU day (r = −0.28, P > .05), or body mass index (r = −0.37, P > .05). Maximum inspiratory pressure was not significantly correlated with SOFA score (r = 0.01, P > .05), ICU day (r = 0.27, P > .05), or body mass index (r = 0.03, P > .05). Maximum expiratory pressure was not significantly correlated with SOFA score (r = −0.16, P > .05) or ICU day (r = −0.32, P > .05) but was significantly correlated with body mass index (r = −0.53, P = .02). In all cases, a strong HGF (>10 kg) corresponded to a strong MEP (>37 cm H₂O) (χ²₁ = 8.1, P = .008).

Significant positive correlations emerged between HGF and MIP (r = 0.45, P = .04), between HGF and MEP (r = 0.78, P = .001), and between MIP and MEP (r = 0.62, P = .004) (see Figure).

The constructed model 1 (R = 0.80, R² = 0.63, adjusted R² = 0.57), model 2 (R = 0.80, R² = 0.63, adjusted R² = 0.59), and model 3 (R = 0.47, R² = 0.28, adjusted R² = 0.13) significantly fit to the data overall (P < .05). In model 1, MIP was not found to be a significant predictor of HGF (P = .97), potentially because of its high correlation with MEP (r = 0.59, P = .006). The models are presented in Table 2.

The main finding of this study is that both MIP and MEP were correlated with HGF. The correlation of MIP with HGF was moderate, and the correlation of MEP with HGF was strong.

Tracheostomy is commonly performed in critically ill patients, especially those in need of prolonged mechanical ventilation because of acute respiratory failure and airway issues.³⁰  For this reason, the clinical condition of patients with tracheostomies may be more serious than that of other ICU patients. However, in studies by Lai et al³¹  and Bragança et al,³²  the MIP, MEP, and HGF of patients with tracheostomies did not significantly differ from the MIP, MEP, and HGF of other patients in the ICU.

Measurement of HGF may provide a simple and accurate alternative to the Medical Research Council score for the diagnosis of ICU-AW.³²  In some studies, the mean (SD) expected HGF, depending on sex and age, was 33.1 (11.1) kg, whereas in our study the mean (SD) HGF was 11.4 (7.4) kg (P < .001).^33,34 

Men are considered to have ICU-AW when the HGF is less than 11 kg; the corresponding value for women is less than 7 kg.³⁵  According to this classification, 12 patients in our study had ICU-AW. The others had moderate weakness well below (<50% of) the reference value; for healthy people between 75 and 85 years old, normal HGF is 33.4 kg for men and 16.6 kg for women.^33,34 

In addition to HGF and the Medical Research Council sum score, MIP can be used to diagnose ICU-AW. A MIP threshold of 36 cm H₂O can indicate ICU-AW (sensitivity, 88%; specificity, 76%).^11,36  This fact suggests that ICU-AW involves all skeletal muscles.^11,37 

The strong correlation between HGF and MEP (which corresponds to the maximum strength of the abdominal and internal intercostal muscles, which participate in coughing) suggests that an individual who is able to grip the hand tightly can also cough. Thus a simple command to perform a strong handshake could serve as an indirect cough evaluation. Cough strength is a predictor of extubation outcome, morbidity, and mortality in patients in the ICU.³⁸  Additionally, forced vital capacity (a cough strength indicator) measured before extubation closely correlates with forced vital capacity after extubation and may serve as an objective predictor of postextubation respiratory failure.³⁹ 

The moderate correlation between MIP and MEP may explain the interaction between these 2 indicators in patients with ICU-AW. Our result does not differ greatly from the result of a study by Park et al⁴⁰  in which MIP and MEP were significantly correlated in healthy patients (r = 0.597, P < .001).

Maximum inspiratory pressure, which is highly dependent on diaphragm strength, had a significant but weak correlation with the maximum force of peripheral skeletal muscles (ie, HGF) in critically ill patients in our study. This finding contrasts with the strong correlation that has been previously observed in healthy participants.⁴¹  The reason for this discrepancy is probably that the diaphragm does not follow the same degenerative course as other skeletal muscles in patients with ICU myoneuropathy, possibly because in ICU patients the diaphragm is exercised during breathing but the peripheral muscles remain less active.

Our regression analysis provides additional information about the associations between HGF, MIP, and MEP when the confounding influence of sex is controlled for. Handgrip force was a significant predictor of both MIP and MEP. However, when MIP and MEP were combined in a single model, the predicting ability of MIP was lost, possibly because both MIP and MEP contribute substantially to the prediction of HGF. Maximum inspiratory pressure can predict HGF, but our results suggest that MEP is preferred for predicting HGF and accounts for much of the observed variance.

Our study has 2 clinical implications. In our study, HGF was correlated more strongly with MEP than with MIP in ICU patients (in the healthy population, MIP correlates strongly with HGF). This result may indicate that in patients with ICU-AW, diaphragm degeneration is delayed as compared with other peripheral muscles. In addition, a strong handhold (handgrip) could indicate a strong expiratory muscle system (MEP), so patients with a strong handhold may have a stronger cough.

Because our study included more men than women, our findings are more applicable to men than to women. However, men generally outnumber women in the ICU, and most ICU patients with tracheostomies are also men.⁴²  Therefore, although our sample consisted mainly of men, it was representative of the ICU population.⁴³  We adapted our analysis to control for the confounding influence of sex because of the physical difference in grip strength between sexes, and our hierarchical regression model included sex in the first block of predictors.

Selecting patients with established ICU-AW (diagnosed either by HGF or by Medical Research Council sum score) would result in greater homogeneity in the study sample. We determined stages of ICU muscle weakness by using HGF and not the Medical Research Council sum score because HGF data were collected on the day MIP and MEP were measured. Measurement of Medical Research Council sum score requires only 1 assessor, whereas measurements of HGF are more reproducible among different examiners.

Cough strength is a predictor of extubation outcome, morbidity, and mortality, especially in patients in the ICU. Handgrip force and the strength of expiratory muscles involved in coughing are strongly correlated and most likely follow a parallel course in ICU-AW. A strong handhold could be an indication of a strong cough. Further studies with larger sample sizes, as well as empirical classification of HGF, are required to better characterize associations between HGF and patient outcomes.