Correlation of the Hand Grip Force With Maximum Inspiratory and Maximum Expiratory Pressure, in Critical Ill Patients

Design of the study This is a random retrospective observational study concerned the period from 5/2017 to 1/2018. The study was conducted for patients of the 2nd General University Intensive Care Unit, of Attikon University General Hospital. The data studied as a sample are derived from file records of adult male and female ICU patients.

The inclusion - exclusion criteria

• Adult ICU patients
• Stable haemodynamic
• Who were able to perform commands from breathing exercises in their hospitals in the ICU (Glasgow Comma Scale 15) good communication and level of consciousness in order to be able to execute orders
• Have integers limbs
• Were tracheostomy (measurements in this case result from a special adapter in the tracheostomy that ensures minimization of losses)
• Their stay in the ICU should have been more than 11 days (the incidence of ICU muscle weakness occurs after 7 days of hospitalization19).
• In order to ensure the randomness of the sample, records of patients who were only hospitalized on an even bed number.
• All requirements for patient recruitment were solely for the day of the study
• They should not suffer from any neurological syndrome (central or peripheral type) other than the ICU acquired weakness, or even craniocerebral injury that could have an effect on the peripheral muscle strength.
• Without drugs that could affect perception or muscle tone Measurements Τhe blindness of the study was ensured because both evaluators and patients don't know were participating in a study. The data from records on patient files related to the following: PEmax, PImax, HG, Age, Weight, Height, points in Glasgow Coma Scale, creatinine, use of inotrope drugs, partial pressure of oxygen (PaO2), platelets, bilirubin, oxygen supply (FiO2), cause of admission and day of hospitalization in the ICU. The last eight values are taken to calculate the SOFA to determine the severity of the condition of these patients. Finally, due to the use of historical data from the patient file, it was not considered necessary to obtain a paper consensus.

Statistical analysis Descriptive statistics are presented as means ± SD, medians and interquartile ranges, or percentages when appropriate. The Shapiro -Wilk test was used to determine whether the variables of interest followed a normal distribution. For the calculation of the correlation between parametric variables Pearson's correlation test was used, and for the calculation of the correlation between non parametric variables Spearman's rho was used. Rho coefficient correlation of \<0.20 is considered to indicate very weak correlation, 0.21-0.40 indicates weak correlation, 0.41- 0.60 indicates moderate correlation and 0.61-0.80 indicates strong correlation. Finally a linear regression analysis was run to predict HGF from PEmax.