|Year : 2008 | Volume
| Issue : 4 | Page : 419
Comparison of Invasive and Noninvasive Mechanical Ventilation for Patients with COPD:Randomised Prospective Study
Ivo Matic1, Visnja Majeric-Kogler2, Katarina Sakic-Zdravcevic2, Matija Jurjevic1, Ivan Mirkovic1, Zlatko Hrgovic3
1 Department of Anaesthesiology and Intensive Care, General Hospital "Dr. Josip Bencevic", Slavonski Brod,
2 Clinic of Anaesthesiology and Intensive Care, KBC Rebro, Zagreb, Croatia,
3 Clinical Hospital University Medical Schol of Osijek "J. J. Strossmayer"
|Date of Acceptance||29-Apr-2008|
|Date of Web Publication||19-Mar-2010|
Kaiserstraße 15, 60311 Frankfurt
Source of Support: None, Conflict of Interest: None
Acute respiratory failure due to chronic obstructive pulmonary disease presents an increasing problem for both health and economics in the modern world. The goal of this study was to compare invasive and noninvasive mechanical ventilation for patients with COPD. A prospective, randomized trial was performed in a multidisciplinary intensive care unit. Of 614 patients requiring mechanical ventilation (MV) longer than 24h, after excluding those who didn't meet the inclusion criteria, 72 patients with COPD remained the research sample. The MV procedure was performed using standard methods, applying two MV methods: invasive MV and noninvasive MV. Patients were randomized into two groups for MV application using closed, non transparent envelopes. Comparison was made based on patient characteristics, objective parameters 1h, 4h, 24h, and 48h after admission and finally treatment outcome. In patients with COPD NIMV had statistically better outcome compared to IMV with MV duration NIMV:IMV 102:187h, p <0.001, time spent in ICU 127:233h, p <0.001. Need for intubation/reintubation 16 (42.1%):34 (100%)/4 (11.8%), p <0.001, hospital pneumonia 2 (5.3%):18 (52.9%), p =0.001. Applying strict application protocols, and based on comparison of objective parameters of pulmonary mechanics, biochemistry and finally treatment outcome, high advantage of NIMV method was confirmed.
Keywords: COPD, Respiration, Mechanical ventilation, Invasive, Noninvasive
|How to cite this article:|
Matic I, Majeric-Kogler V, Sakic-Zdravcevic K, Jurjevic M, Mirkovic I, Hrgovic Z. Comparison of Invasive and Noninvasive Mechanical Ventilation for Patients with COPD:Randomised Prospective Study. Indian J Anaesth 2008;52:419
|How to cite this URL:|
Matic I, Majeric-Kogler V, Sakic-Zdravcevic K, Jurjevic M, Mirkovic I, Hrgovic Z. Comparison of Invasive and Noninvasive Mechanical Ventilation for Patients with COPD:Randomised Prospective Study. Indian J Anaesth [serial online] 2008 [cited 2019 Jun 26];52:419. Available from: http://www.ijaweb.org/text.asp?2008/52/4/419/60656
| Introduction|| |
Acute respiratory failure (ARF) due to chronic obstructive pulmonary disease (COPD) presents an increasing problem for both health and economics in the modern world  . COPD is defined as a disease state characterized by airflow limitation that isn't fully reversible. The airflow limitation is usually both progressiveand associated with an abnormal inflammatory response of the lungs to noxious particles or gases  . It encompasses chronic bronchitis and emphysema.
Therapeutic management of COPD includes drugs (antibiotics for those with purulent sputum, systemic corticosteroids, ß 2 agonists), oxygen, physiotherapy and mechanical ventilation  . Mechanical ventilation is often applied in the late stages of COPD or in patients with rapid clinical deterioration  . Applying the standard invasive mechanical ventilation (IMV) means confronting the patient with all the side effects and complications following endotracheal intubation  . The most important complication of this form of MV is high ratio of hospital pneumonia  . Other complications include: damage to the trachea caused by endotracheal tube producing ulceration, oedema and haemorrhage that can lead to tracheal stenosis  . Noninvasive mechanical ventilation (NIMV) presents an alternative to conventional IMV through an endotracheal tube, both in early stage of ARF as well as in patients with severe disease , . It includes similar techniques for alveolar ventilation improvement to those of IMV, but without endotracheal intubation  . Avoiding endotracheal intubation complications, improving patient comfort and maintaining airway defence mechanisms, speech and swallowing are the most important NIMV advantages , . There are also potential complications of this MV method eg damage to the face and nose skin, gastric distension with aspiration risk, sleeping disorders and conjuctivitis  . Conditions that limit NIMV application are: coma, unstable respiratory centre, swallowing disorders, mental immaturity, face deformations, shock and cardiorespiratory arrest.
There is still no unified approach towards specific MV method indications , . Kramer et al  in their research in 1995 report that NIMV is a method of choice for patients with COPD because of fast blood gas exchange improvement and the posibility to avoid endotracheal intubation (success ratio was 74%). At the same time, Brochard et al  report that NIMV is a bad choice for patients with COPD because only 29% of these patients are suitable for NIMV and therefore can NIMV only be researched as an alternative procedure to IMV. Squadrone et al  in their study from 2004. present similar results. They had high failure ratio with NIMV for patients with COPD (40 out of 64 patients required endotracheal intubation).
Guerin et al  in their study in 1998 report that 39% of patients with COPD that had NIMV applied eventually had to be endotracheally intubated.
Mountain et al  in their research in 1998. recommend endotracheal intubation for severe cases of asthma because of respiratory muscle fatigue or incidence of life threathening complications.
Jones et al  in their research in 1995 report that NIMV is suitable as first line intervention in carefully chosen patients with COPD facing ARF. In their research, they report that COPD patients with pneumonia and congestive heart failure are not suitable for this MV method.
Ambrosino et al  in their research in 1995 report that patients with COPD who had high Acute Physiology and Chronic Health Evaluation (APACHE II) score  , also had high NIMV failure rate. This was accompanied with all negative NIMV application complications.
Opposite to that research, Fernandez et al  in their research in 1993 report that APACHE II score has no influence on prediction of NIMV success in patients with COPD.
Through careful examination of published studies, we have established that influence and importance of objective pulmonary mechanics parameters: tidal volume (Vt), respiratory rate (f), f/Vt ratio, maximal inspiratory pressure (Pimax), plato pressure (Pplato), static pulmonary compliance (Cst) and airway resistance (R), on the choice of MV method and treatment outcome has not yet been researched. Also, influence of blood gas exchange parameters: partial pressure of arterial oxygen and carbon dioxide (PaO 2 and PaCO 2 ), negative logarithm of H+ concentration (pH), arterial oxygen tension/inspiratory oxygen fraction (PaO 2 /FiO 2 ) ratio, bicarbonate concentration and blood oxygen saturation (SatO 2 ) on the choice of MV method and treatment outcome has not been confirmed.
Clinical research point to many scientific dillemmas and contradictions on NIMV indications. Advantage of either MV method has not yet been established by neither clinical nor laboratory parameters.
The goal of our study was to determine the relationship and influence of objective parameters of pulmonary mechanics and biochemistry (Vt, f, f/Vt ratio, Pimax, Pplato, static pulmonary compliance, airway resistance, PaO 2 , PaCO 2 , pH, PaO 2 /FiO 2 ratio, bicarbonate concentration and SatO 2 ) on the choice of MV method and treatment outcome in patients with COPD. Furthermore, based on specific MV method treatment outcome (MV duration, time spent in ICU, need for intubation/reintubation, need for tracheostomy, incidence of life threatening complications and hospital pneumonia)we wanted to determine the superiority of a specific MV method (NIMV or IMV) in COPD patients.
Patients. A prospective randomized study was conducted at the medical-surgical intensive care unit (ICU) at "Dr. Josip Bencevic" General Hospital in Slavonski Brod, Croatia, between January 2004 and February 2006. The study sample consisted of 614 eligible adult patients requiring MV for more than 24 hours. After exclusion of patients who did not meet the strict defined criteria and/or did not have COPD, 72 patients remained and they present the research sample. These patients were than randomized for either NIMV or IMV using closed-non transparent envelopes choosen by a third person, not included in the research. Exclusion criteria were: MV duration shorter than 24h, use of MV on admission to the ICU, patients scheduled for organ donation, patients who died during the research because of another associated disorder, patients in coma and those who had cardio-respiratory arrest (because these conditions present contraindications for NIMV application and including them in the research would influence patient randomization), patients with unstable respiratory centre and patients in shock.
On admission, the following patient data were collected: sex, age, comorbidities (hypertension, diabetes mellitus, and congestive heart disease) and previous use of MV.
Severity of illness was assessed using Acute Physiology and Chronic Health Evaluation (APACHE II) score both on admission and during MV procedure. This scoring system includes: Glasgow Coma Score (GCS), acute physiologic patient's state, age and presence of chronic illness.
Objective patient data was measured and recorded on admission to the ICU, 1h, 4h, 24h and 48h. After that, every 24h of ICU stay, and if necessary more often than that. The following objective data was measured and recorded during MV procedure: f, Vt, SatO 2 , pH, PaO 2 , PaCO 2 and bicarbonate blood level, PaO 2 /FiO 2 ratio, f/Vt ratio, Pimax, Pplato, R, Cst.
Methods. The following objective parameters were used as indications for the use of one of the MV methods: worsening of clinical status, spontaneous respiratory frequency > 25/min, spontaneous tidal volume <5ml/kg, PaO 2 <60mmHg, PaCO 2 >45mmHg, pH <7.30, SatO 2 <88%, PaO 2 /FiO 2 <200, f/Vt >100, restless patient. If a patient had met at least 6 of these criteria, he was then randomized into either IMV or NIMV group.
During the research, the following data was also collected for each MV method: total MV duration, length of ICU stay, success of MV, need for intubation or reintubation, life threatening complications, need for tracheostomy, incidence of hospital pneumonia and ICU mortality.
During ICU stay, patients received all necessary treatment required by their condition. All laboratory and clinical parameters were evaluated and corrected if necessary. Enteral nutrition was preferred to parenteral whenever possible, applied through nasogastric tube or perorally  .
IMV protocol: Patient with respiratory insufficiency, who had IMV applied, was orotracheally intubated. During IMV patients received the lowest respiratory support level that secured SatO 2 >90% with FiO 2 <0.6, satisfying CO 2 elimination (PaCO 2 <45mmHg) and stabile hemodynamic patient condition. The weaning process was conducted using pressure support ventilation (PSV)  . Pressure support level was 2-4cmH 2 O depending on the patient's clinical status and values of the measured parameters of pulmonary mechanics, biochemistry and circulation. Patients were extubated after fulfilling general extubation criteria, where 5cmH 2 O of pressure support was enough for spontaneous patient breathing. After the conclusion of the weaning process, continuous patient monitoring was performed. Side effects and complications as well as need for further MV were observed and evaluated. Patients were continuously monitored until they were either dismissed from ICU, or death occurred.
NIMV protocol: The patient's head-pad was raised to 45°, necessary equipment was prepared next to the patients head and the procedure that follows and was explained. An appropriate face mask was chosen and connected to the respirator. Starting respirator parameters were set to: continuous positive air pressure (CPAP) to 0cmH 2 O, PSV 10cmH 2 O and FiO 2 was adjusted to reach SatO 2 >90%. The nose was protected using strapping to prevent skin damage. The patient was reassured and the mask was gently held to the patient's face, simultaneously trying to harmonize the patient's respiration with the respirator. The mask was then secured to the face with plaster strips. The respirator was then set to: CPAP 3-5cmH 2 O and PSV 10-25cmH 2 O in order to reach tidal volume >5ml/kg and respiratory frequency <25/min. After that, the alarms on the respirator and respiratory support level were set. Clear communication with the patient was ensured so the patients could signal in case of complications occurrence. Due to small number of patients, only face masks were used so the study results wouldn't be influenced
According to clinical status and objective parameters, respiratory support level was reduced until MV could be discontinued. Endotracheal intubation was performed in case of respiratory arrest, loss of consciousness, severe psychomotor agitation that requires sedation, hemodynamic instability, failure to reach SatO 2 >90% with FiO 2 =0.6 and PaCO 2 >60mmHg.
In all patients included in the research, MV was administered by use of Evita Drager dura 2 respirators (Drager, Lubeck, Germany), with software option for NIMV and Puritan Bennet 7200 respirators (Puritan Bennet, Carlsbad, CA, USA). Nasal and face masks were applied for NIMV (Respironics Inc, Herrsching, Germany). Parameters of pulmonary mechanics were directly measured on the respirator. For patients with spontaneous breathing, tidal volume and respiratory frequency were measured with a spirometer (Ohmeda Biox, Louisville, CO, USA), a maximal inspiratory pressure with a manometer (Ohmeda Biox). Pulmonary biochemistry parameters were measured using blood gas analysis on Ciba Corning (Ciba Corning, Halsted, England). Cardio-respiratory functions were continuously monitored using Datex monitors (Datex Ohmeda, Helsinki, Finland), and ventilation and oxygenation using Datex Engstrom AS3 and CS3 Compact monitor (Datex Ohmeda, Helsinki, Finland).
Statistical analysis. Qualitative and numerical data were analyzed with descriptive statistic parameters: median, minimum value, maximum value, IQ range. Frequency tables were used to present qualitative data. Contingency tables with chi-square test were employed for comparison of two independent for qualitative variables. In case of small sample size, Fisher's exact test was used. The Mann-Whitney test was employed for the comparison of two independent groups on numerical data. Normality of distribution was tested by the Kolmogorow-Smirnov test. P <0.01 was considered statistically significant.
The study was carried out in line with ethical principles and was approved by the Hospital Ethics Committee.
| Results|| |
Over 26 months a total of 1311 patients were treated, 614 of them requiring MV. After randomisation and exclusion of patients who didn't have COPD and/ or didn't meet the inclusion criteria, 72 COPD patients remained and they represent the research sample. After randomisation, 38 patients had NIMV and 34 had IMV applied. In the NIMV group, the procedure was successful for 22 (57.9%) COPD patients. In the IMV group, the procedure was successful for 16 (47.1%) patients.
[Table 1] shows the demographical data for COPD patients with NIMV and IMV.
[Table 2] shows comparison of NIMV and IMV based on parameters of pulmonary mechanics. Shown parameters are: tidal volume, respiratory frequency, f/ Vt ratio, Pmax, Pplato, R and Cstat through 1h, 4h, 24h and 48h.
[Table 3] shows comparison of NIMV and IMV based on pulmonary biochemistry parameters ( pH, PaO 2 , PaCO 2 , PaO 2 /FiO 2 , bicarbonates and SatO 2 through 1h, 4h, 24h and 48h).
[Table 4] shows comparison of NIMV and IMV for patients with COPD based on treatment outcome.
| Discussion|| |
The results of this research have shown that both NIMV and IMV are suitable methods for securing respiratory support in ARF. During the research, 614 patients with ARF were treated in a multidisciplinary ICU. All patients with MV shorter than 24h, patients admitted to the ICU already mechanically ventilated or those transferred to some other hospital for treatment continuation were all excluded from the research. Furthermore, patients scheduled for organ donation, patients in coma or with cardiorespiratory arrest were also excluded. These conditions present absolute contraindications for NIMV application and including them in the research would fail research randomization. Patients who died during the research of some other illness, patients with unstable respiratory centre and patients in shock were also excluded from the research.
NIMV has proven superior in terms of MV duration which of course influences the total ICU length of stay and the need for tracheostomy in facilitating the weaning process. Furthermore, hospital pneumonia was recorded in only 2 patietns in NIMV group compared to 12 in the IMV group. Even so, no statistical difference in mortality rates was recorded between groups. This is possibly due to relatively small sample size but also because most severe patients have similar mortality rates, although some studies have shown a reduction in mortality rates in favor of NIMV
Similar results was presented Keenan et al  in their study in 2003. They confirmed the superiority of NIMV as a method that, compared to IMV in patients with acute exacerbation of COPD, reduces the need for endotracheal intubation to 28%, lowers mortality rate from 15% to 10% and reduces MV duration from 6.83 to 4.57 days.
In contrast, Squdrone et al  in their research from 2004., had a high failure ratio with NIMV applied in acute exacerbations of COPD. They had to intubate 40 of 64 patients. MV duration, mortality rate and treatment duration were similar in both groups. When NIMV was successful, complication incidence, mortality rate were lower and treatment duration was shorter, but the NIMV procedure overall had high failure ratio, and those patients, who had failed NIMV had worse endoutcome results than patients who were treated with IMV from the beginning.
Brochard et al  in a multicentric study from 1995, conducted on 85 COPD patients with ARF (pH 7.27±0.1, PaCO 2 70±12mmHg, reported rapid improvement in PaO 2 , and slower correction of PaCO 2 . Patients randomized for NIMV had significantly lower intubation rates. They report less complications (14%:45%, p <0.001) and reduced mortality with NIMV (9%:29%, p=0.02), as well as shorter hospital treatment duration (23±17:35±33 days, p =0.02). Even so, they conclude that NIMV is not a good choice for patients with COPD because only 29% of patients in their study were suitable for successful NIMV. They recommend that NIMV can only be considered as an alternative procedure to IMV.
Meduri et al  in their study in 1996 had 91 patients with COPD, 51 in acute exacerbation, 27 with pneumonia and 13 with congestive heart failure. 18% of patients with acute exacerbation had to be intubated (78% of those because of the inability to improve arterial blood gas pressures). 41% of patients with pneumonia had to be intubated (45% due to inadequate gas exchange), and 46% of patients with congestive heart disease had to be intubated (67% due to inadequate gas exchange).
Robino et al  in their study in 2003 researched the success of NIMV in other than patients with decompensated chronic obstructive and restrictive pulmonary diseases. The research was conducted on 64 patients with COPD and 20 with chronic restrictive pulmonary disease. In the group of patients with chronic restrictive pulmonary disease, the number of patients with successful NIMV was smaller compared to COPD group, and need for endotracheal intubation was higher (67:35%, p =0.01). After 12h of NIMV application, similar results were observed for respiratory frequency, minute volume and arterial blood gases. pH and PaCO 2 improvement after 12h was better in COPD patients and they report this as a predictor of success, but that isn't the case with restrictive pulmonary disease patients. Success with NIMV in ARF is therefore less in chronic restrictive pulmonary diseases.
| Conclusion|| |
In this study with well defined ventilation protocols and comparison based on the objective parameters of pulmonary mechanics, biochemistry and treatment outcomeNIMV was superior compared to IMV.
| References|| |
|1.||American Thoracic Society. Standard for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995;152:77-121. |
|2.||Pauwels R A, Buist A S, Calverley P. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global Initiative of Chronic Obstructive Lung Disease (GOLD) Workshop Summary. Am J Respir Crit Care Med 2001;163:1256-1276. |
|3.||Barnes PJ, New therapies for chronic obstructive pulmonary disease. Thorax 1998;53:137-147. |
|4.||Kondili E, Alexopoulou C, Prinianakis G, Xirouchaki N, Georgopoulos D. Pattern of lung emptying and expiratory resistance in mechanically ventilated patients with chronic obstructive pulmonary disease. Intensive Care Med 2004;30:1311-1318. [PUBMED] [FULLTEXT] |
|5.||Nevins ML, Epstein SK. Predictors of outcome for patients with COPD requiring invasive mechanical ventilation. Chest 2001;119:1840-1849. [PUBMED] [FULLTEXT] |
|6.||Luyt CE, Chastre J, Fagon JY. Value of the clinical pulmonary infection score for the indentification and management of ventilator-associated pneumonia. Intensive Care Med 2004;30:844-852. [PUBMED] [FULLTEXT] |
|7.||Conti G, Antonelli N, Navalesi P. Noninvasive versus conventional mechanical ventilation in patients with chronic obstructive pulmonary disease after failure of medical treatment in the ward: a randomized trial. Intensive Care Med 2002;28:1701-1707. |
|8.||Elliot MW. Non-invasive ventilation in acute exacerbation of chronic obstructive pulmonary disease: a new gold standard? Intensive Care Med 2002;28:1691-1694. |
|9.||Brochard L. Non-invasive ventilation for acute exacerbation of COPD: a new standard of care. Thorax 2000;55:817-818. [PUBMED] [FULLTEXT] |
|10.||Lightowler JV, Wedzicha JA, Elliott MW, Ram FSF. Noninvasive positive pressure ventilation to treat respiratory failure resulting from exacerbations of chronic obstructive pulmonary disease: Cochrane systematic review and meta-analysis. BMJ 2003;326:185-190. |
|11.||Gorini M, Ginanni R, Villella G, Tozzi D, Augustynen A, Corrado A. Noninvasive negative and positive pressure ventilation in the treatment of acute on chronic respiratory failure. Intensive Care Med 2004;30:875-881. [PUBMED] [FULLTEXT] |
|12.||Brochard L. Mechanical ventilation: invasive versus noninvasive. Eur Respir J 2003;22:Suppl.47,31-37. |
|13.||Girault C, Briel A, Hellot MF, Tamion F, Woinet D, Lesoy S, et all. Noninvasive mechanical ventilation in clinical practice: A 2-year experience in a medical intensive care unit. Crit Care Med 2003;31:552-559. |
|14.||Consenssus conference. Clinical Indications for noninvasive positive pressure ventilation in chronic respiratory failure due to restrictive lung disease, copd, and nocturnal hypoventilation. -A Consesus Conference Report. Chest 1999;116:521-534. |
|15.||Liesching T, Kwok H, Hill N. Acute applications of noninvasive positive pressure ventilation. Chest 2003;124:699-713. |
|16.||Kramer N, Meyer TJ, Meharg J. Randomised, prospective trial of noninvasive positive pressure ventilation in acute respiratory failure. Am J Respir Crit Care Med 1995;151:1799-1806. |
|17.||Brochard L, Mancebo J, Wysocki M, Lofaso F, Conti G, Rauss A, et all. Noninvasive ventilation for acute exacerbation of chronic obstructive pulmonary disease. N Engl J Med 1995;333:817-822. [PUBMED] [FULLTEXT] |
|18.||Squadrone E, Frigerio P, Fogliati C, Gregoreti C, Conti G, Antonelli M, et al. Noninvasive vs invasive ventilation in COPD patients with severe acute respiratory failure deemed to require ventilatory assistance. Intensive Care Med 2004;30:1303-1310. |
|19.||Guerin C, Girard R, Chemorin C, De Varax R, Fournier G. Facial mask noninvasive mechanical ventilation reduces the incidence of nosocomial pneumonia. Intensive Care Med 1998;24:1024-1032. |
|20.||Mountain RD, Sahn SA. Clinical features and outcome in patients with acute asthma presenting with hypercapnia. Am Rev Respir Dis 1998;138:535-539. |
|21.||Jones DJ, Paul EA, Jones PW. Nasal pressure support ventilation plus oxygen therapy alone in hypercapnic COPD. Am J Respir Crit Care Med 1995;152:538-544. |
|22.||Ambrosino N, Foglio K, Rubini F. Noninvasive mechanical ventilation in acute respiratory failure due to chronic obstructive pulmonary disease: Correlates for success. Thorax 1995;50:755-757. |
|23.||Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818-829. [PUBMED] |
|24.||Fernardes R, Blanch LI, Valles J. Pressure support ventilation via mask in acute respiratory failure in hypercapnic COPD patients. Intensive Care Med 1993;19:456-461. |
|25.||Dhalival R, Jurewitsch B, Harrietha D, Heyland DK. Combination enteral and parenteral nutrition in critically ill patients: harmful or beneficial? A systematic review of the evidence. Intensive Care Med;2004:30:1666-1671. |
|26.||Matic I, Majeric-Kogler V. Comparison of pressure support and t-tube weaning from mechanical ventilation: Randomized prospective study. Croatian Medical Journal 2004;45:162-166. |
|27.||Keenan SP, Sinuff T, Cook DJ, Hill NS. Which patients with acute exacerbation of chronic obstructive pulmonary disease benefit from noninvasive positive pressure ventilation? A systematic review of the literature. Annals of Internal Medicine 2003;138:861-70. [PUBMED] [FULLTEXT] |
|28.||Meduri GU, Turner RE, Abou-Shala N. Noninvasive positive pressure ventilation via face mask: First-line intervention in patients with acute hypercapnic respiratory failure. Chest 1996;109:179-193. |
|29.||Robino C, Faisy C, Diehl JL, Rezgui N, Labrousse J, Guerot E. Effectiveness of non-invasive positive pressure ventilation differs between decompensated chronic restrictive and obstructive pulmonary disease patients. Intensive Care Med 2003;29:603-610. [PUBMED] [FULLTEXT] |
[Table 1], [Table 2], [Table 3], [Table 4]