Prolonged Invasive Ventilation Following Acute Ventilatory Failure in COPD* Weaning Results, Survival, and the Role of Noninvasive Ventilation

CHEST Prolonged Invasive Ventilation Following Acute Ventilatory Failure in COPD* Weaning Results, Survival, and the Role of Noninvasive Ventilation Timothy G. Quinnell, MRCP; Samantha Pilsworth, BSc; John M. Shneerson, MA, DM; and Ian E. Smith, MA, MD Original Research ACUTE VENTILATORY FAILURE Background: Invasive ventilation for COPD has significant mortality, and weaning can be difficult. At Papworth Hospital, we provide a specialist weaning service using noninvasive ventilation (NIV) for patients requiring prolonged invasive ventilation after recovery from acute illness. We analyzed our results for patients with COPD to identify factors associated with weaning outcome and survival. Methods: A retrospective analysis was conducted of COPD patients admitted for weaning from invasive ventilation, from 1992 to 2003. Weaning success and survival were assessed. Associations were sought between these outcomes and age, sex, spirometry, arterial blood gas levels, APACHE (acute physiology and chronic health evaluation) II score, length of stay (LOS), and the use of NIV and long-term oxygen therapy. Results: Sixty-seven patients were identified, all of whom were receiving tracheostomy ventilation on transfer to the Respiratory Support and Sleep Centre (RSSC). Sixty-four patients (95.5%) were weaned, and 62 patients survived to hospital discharge. NIV was used in weaning 40 patients and in the long term in 25 patients. Median survival was 2.5 years (interquartile range, 0.7 to 4.6 years). One-year, 2-year, and 5-year survival rates were 68%, 54%, and 25%, respectively. Long-term survival was inversely associated with age and LOS in the ICU and the RSSC. The provision of maintenance NIV after weaning was associated with better long-term survival, independent of age and LOS (hazard rate, 0.48; p 0.03). Conclusions: These results demonstrate that a specialist multidisciplinary approach, including the use of NIV, can be successful in weaning most COPD patients from prolonged invasive ventilation. The data also suggest that long-term NIV may improve survival in selected patients. (CHEST 2006; 129:133 139) Key words: COPD; invasive mechanical ventilation; noninvasive ventilation; tracheostomy; weaning Abbreviations: ABG arterial blood gas; APACHE acute physiology and chronic health evaluation; CI confidence interval; IMV invasive mechanical ventilation; IQR interquartile range; LOS length of stay; LTOT long-term oxygen therapy; NIV noninvasive ventilation; PCP Progressive Care Programme; PSV pressure support ventilation; Ptco 2 transcutaneous Pco 2 ; RSSC Respiratory Support and Sleep Centre; USB unsupported spontaneous breathing Acute ventilatory failure is common in exacerbations of COPD that require hospital admission. Despite the advent of noninvasive ventilation (NIV), *From Papworth Hospital, Papworth Everard, Cambridge, UK. Manuscript received February 2, 2005; revision accepted June 22, 2005. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). invasive mechanical ventilation (IMV) is still required by 15 to 26% of patients. 1,2 The prognosis for COPD patients requiring IMV for ventilatory failure is better than for those intubated for other reasons, 3 5 but hospital survival ranges from 76% to as low as 31%. 3,4,6 9 The pro- Correspondence to: Tim Quinnell, MRCP, Respiratory Support and Sleep Centre, Papworth Hospital, Papworth Everard, Cambridge, UK CB3 8RE; e-mail: Tim.Quinnell@papworth.nhs.uk www.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 133

portion of ventilation time spent weaning is greater for COPD than for many other conditions. 10 In series 6,11 describing patients requiring prolonged IMV, 46 to 59% have had COPD. One-year survival for patients requiring IMV ranges from 57 to only 15%. 3,4,6 9 The Respiratory Support and Sleep Centre (RSSC) at Papworth Hospital, Cambridge offers a specialist weaning service for patients requiring prolonged IMV at other hospitals. This Progressive Care Programme (PCP) has been reported previously 12 and is applied by a multidisciplinary team with the goal of achieving independence from IMV. In this article, we present outcomes for COPD patients entered into the PCP from 1992 to 2003, we identify factors associated with weaning outcome and survival, and we discuss those aspects of our weaning strategy that may have contributed to the results achieved. Materials and Methods Patients are referred to the PCP when weaning has failed at the referring ICU. Those accepted for transfer to the RSSC have recovered from the acute illness that precipitated their intubation. Most receive ventilation via tracheostomy and are hemodynamically stable without need of inotropic support. General Care Following transfer to our unit, sedating drugs are withdrawn at the first opportunity, and a normal diurnal routine is reinforced. Respiratory function is optimized with bronchodilators and corticosteroids as indicated. Respiratory tract secretions are actively managed with physiotherapy, tracheal suctioning, and the appropriate use of antibiotics. The management of coexisting medical problems is reviewed, and metabolic abnormalities are corrected. Tube enteral feeding is used to optimize nutritional status, and a normal diet and supplements are introduced when swallowing is safe. Weaning from IMV On arrival, IMV is continued, using pressure support ventilation (PSV) with positive end-expiratory pressure aiming to normalize arterial blood gas (ABG) levels. Ventilatory weaning involves the introduction of unsupported spontaneous breathing (USB). Initial brief episodes (up to 30 min) are lengthened according to individual patient progress. This is guided by close clinical observation for respiratory distress, and the use of pulse oximetry, transcutaneous Pco 2 (Ptco 2 ) monitoring, and ABG levels. If, despite optimizing other medical therapies, the patient fails to increase the duration of USB without respiratory distress or hypercapnia (Paco 2 or Ptco 2 7.5 kpa), then NIV is introduced (NIPPY 2 ventilator; B&D Electromedical; Stratford-on- Avon, UK). This is applied via a nasal or full-face mask with the tracheostomy tube capped off and its cuff deflated. NIV is guided by continuous oximetry and Ptco 2, and daytime ABG levels, aiming for normalization of Paco 2. For COPD patients, inspiratory pressure is titrated to a maximum of 36 cm H 2 O and expiratory pressure up to 6 cm H 2 O. Only when USB is effective for most of the day is IMV withdrawn at night. The provision of nocturnal NIV is determined by the same criteria guiding its daytime introduction (ie, respiratory distress and/or hypercapnia 7.5 kpa during USB). Oxygen is only prescribed for nocturnal use if the mean sleeping oxygen saturation drops below 80% and NIV (if indicated) is already optimized. Long-term oxygen therapy (LTOT) is added if the daytime Pao 2 is 7.3 kpa or 8 kpa with cor pulmonale or polycythemia. If hypercapnia ( 7.5kPa) develops when the patient is receiving LTOT alone, maintenance nocturnal NIV is started. Laryngeal Function and Tracheostomy Management Effective laryngeal function is a priority. Speaking is encouraged as soon as self-ventilation via a one-way tracheostomy speaking valve is possible without risk of aspiration. Swallowing safety is first assessed with the tracheostomy tube cuff inflated, using oropharyngeal administration of blue food dye followed by tracheal suctioning. If there is no aspiration, dyed fluids and food are tried. Once successful, these steps are repeated with the cuff deflated. Mild aspiration usually improves with swallowing and speaking practice. Regular suctioning can be required, and anticholinergic agents are used if oropharyngeal secretions are profuse. The cuff remains inflated until there is little or no aspiration. The tracheostomy tube is only removed when it is no longer needed for airway protection, secretions access, or ventilation. Patient Selection Patients with a discharge diagnosis of COPD were sought from a database of all patients entered into the PCP. The diagnosis was reviewed on the basis of patient history, referring hospital notes, clinical assessment, spirometry, and radiography. Coexisting pulmonary, neuromuscular, or chest wall disorders, or obstructive sleep apnea were exclusion criteria. Data Collection A retrospective case note review was performed, and the following data were extracted: age at PCP admission; sex; length of stay (LOS) at referring ICU; LOS in the RSSC; FEV 1 ; FVC; ABG levels at intubation, at transfer to, and discharge from the PCP; APACHE (acute physiology and chronic health evaluation) II score at ICU admission; ventilation mode and inspired oxygen concentration at RSSC discharge; and discharge destination. The most recent ventilator settings were sought for patients maintained on NIV after weaning. The Intensive Care National Audit and Research Centre provided APACHE II scores kept on their database when unavailable from our records. Waiver-of-confidentiality agreements were made with the respective referring units prior to data release. Outcomes Weaning success was defined as complete withdrawal of IMV, with or without continuing requirement for nocturnal NIV. Hospital and long-term survival were also measured. Patients with a tracheostomy retained solely for suction were regarded as weaned. Failure to wean was defined as death during the weaning process, or discharge from the RSSC with a continuing requirement for IMV. Long-term survival was assessed using case notes, the hospital computerized patient administration system, and the National Health Service strategic tracing service, which provides information on a patient s vital status accurate to within 2 weeks. 13 134 Original Research

Associations were sought between each of the three outcomes and the following variables: age, sex, FEV 1, FVC; APACHE II score at ICU admission; ABG levels at intubation and RSSC arrival; and LOS in the ICU. Associations were also sought for long-term survival with RSSC LOS, ABG levels at RSSC discharge, and the provision of maintenance NIV or LTOT after discharge. Statistical Analysis Data were analyzed using statistical software (SPSS version 12.0 for Windows; SPSS; Chicago, IL). Long-term survival was assessed with Kaplan-Meier charts and life tables. Associations with survival were sought using Cox regression, with grouped data used when statistically significant but nonlinear relationships were found for continuous variables. Normally distributed data were compared using independent t tests, and nonparametric data were compared with the Mann-Whitney U test. Results We accepted 436 patients into the PCP between June 1992 and December 2003, 97 of whom had a discharge diagnosis of COPD. Thirty patients were excluded from the analysis due to coexisting pulmonary, chest-wall or neuromuscular disorders, or obstructive sleep apnea. Demographics, ICU, and weaning data of the remaining 67 patients (36 male) are presented in Table 1. All 67 patients were receiving tracheostomy ventilation on arrival to the RSSC. Three patients died in the RSSC while receiving IMV, leaving 64 patients (95.5%) who were weaned successfully, 2 of whom retained a minitracheostomy for secretion management. All those who were weaned successfully also survived to discharge from the RSSC. Fifty-two Table 1 Patient Demographic, ICU, and Weaning Data Variables Age at PCP admission, yr FEV 1,L FVC, L FEV 1 /FVC, % ph at ICU admission APACHE II score* Intubation Pao 2, kpa Paco 2, kpa Transfer to PCP Pao 2, kpa Paco 2, kpa Discharge from PCP Pao 2, kpa Paco 2, kpa LOS in ICU, d LOS in RSSC, d Median (IQR Range) 66 (55 73) n 67 0.60 (0.45 0.9) n 48 1.93 (1.5 2.63) n 48 34 (24 50) n 48 7.21 (7.14 7.28) n 30 25.5 (7.9) n 42 7.5 (5.9 10.8) n 41 10.3 (9 12.7) n 44 8.6 (7.2 10.5) n 67 7.5 (6.6 9.1) n 67 8.8 (8.1 9.4) n 64 6.1 (5.5 6.7) n 64 24 (14 38) n 66 27 (16 45) n 67 *Mean (SD). Spirometry data taken from up to 1 year after weaning. patients went directly home, and one patient went to a residential home. Eleven patients were transferred back to their referring hospital, 2 of whom died before discharge. Overall survival to discharge from hospital was 62 of 67 patients (92.5%). APACHE II scores were available for 42 patients (63%); the mean score was 25.5 (SD, 7.9), giving a predicted survival to hospital discharge of 45% compared to an actual survival of 93% for this subgroup. NIV was used in weaning 40 of 67 patients, including 2 of the 3 patients who failed to wean. Twenty-five patients continued with long-term NIV (Table 2). The mean inspiratory pressure (n 21) was 28 cm H 2 O (SD, 4.7 cm H 2 O). Five long-term NIV patients were also discharged with LTOT, and 11 patients were discharged on LTOT alone. Ninety-day survival for the whole group from the date of admission to the weaning program was 88.1%, and median survival was 2.5 years (interquartile range [IQR], 0.7 to 4.6 years). Annual survival rates at 1 to 5 years were 68%, 54%, 40%, 31%, and 25%, respectively (Fig 1). For those weaned from IMV, 1- to 5-year survival rates from PCP admission were 71%, 56%, 41%, 33%, and 26%, respectively; median survival was 2.6 years (IQR, 0.9 to 5.4 years). Patient Characteristics and Correlation With Longterm Survival Neither weaning success nor hospital survival were associated with age; sex; FEV 1 ; FVC; APACHE II score on ICU admission; ABG levels at ICU or PCP admission; LOS in the ICU; or use of NIV during weaning. Long-term survival was inversely associated on univariate analysis with age and LOS in the ICU and the RSSC. Patients aged 70 years on entry to the PCP had a 2.7-fold hazard of dying subsequently, compared to those 70 years old (95% confidence interval [CI], 1.5 to 4.9; p 0.001). Patients who had spent 28 days on ICU were more likely to die in the follow-up period than those in the ICU for 14 days (hazard rate, 2.4; 95% CI, 1.1 to 5.1; p 0.02). Likewise, staying on the RSSC for 28 days was Table 2 Maintenance NIV Patients Compared to Non- NIV Users* Variables NIV No NIV p Value Age at PCP admission, yr 64.5 (53 69) 67.3 (56.9 73.9) 0.16 LOS in ICU, d 31 (12.5 39) 23 (14 37) 0.85 LOS in RSSC, d 27 (17 46.5) 21 (16 40) 0.61 Discharge Paco 2, kpa 6.6 (0.7) 5.8 (0.9) 0.001 Discharge Pao 2, kpa 8.3 (1.0) 9.0 (1.1) 0.006 FEV 1,L 0.5 (0.4 0.8) 0.7 (0.5 0.9) 0.08 FVC, L 1.8 (1.3 2.6) 2.2 (1.6 2.8) 0.13 *Data are presented as median (IQR) or mean (SD). www.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 135

term survival was not associated with sex; FEV 1 ; FVC; APACHE II score at ICU admission; ABG levels at ICU or PCP admission, or RSSC discharge; use of NIV during weaning; or provision of LTOT. Discussion Figure 1. Kaplan-Meier plot of cumulative survival for all patients (n 67) from admission to the PCP. associated with poorer long-term survival than a stay of 14 days (hazard rate, 4.0; 95% CI, 1.5 to 10.7; p 0.005). Long-term NIV was associated with a survival advantage on univariate analysis (hazard rate, 0.50; 95% CI, 0.26 to 0.96; p 0.04) and after correcting for age and LOS (hazard rate, 0.48; 95% CI, 0.24 to 0.95; p 0.03). Median survival for patients receiving maintenance NIV was 3.3 years (IQR, 1.3 to 7.9 years) and 1.6 years (IQR, 0.48 to 4.2 years) for those discharged without it (Fig 2). Long- Figure 2. Kaplan-Meier plot of cumulative survival according to discharge NIV status (n 64). Exacerbations of COPD requiring hospitalization are associated with substantial hospital and longerterm mortality. 3,4,6 9,14 Those cases complicated by acute ventilatory failure frequently require IMV, and weaning can be difficult. 1,2,10,15 We report a 95.5% success rate for weaning COPD patients from IMV and 92.5% survival to hospital discharge. These subjects had recovered from the acute illness that precipitated their intubation, were hemodynamically stable, and received ventilation via tracheostomy. The median survival for those patients weaned from IMV is 2.6 years. Survival rates at 1, 2, and 5 years are 71%, 56%, and 26%, respectively. These outcomes represent the application of a process of care, including the use of NIV, by a specialist multidisciplinary team. The interpretation of our data is limited by their retrospective and uncontrolled nature. This is also the case for most other published reports of COPD patients receiving invasive ventilation (Table 3). Our results appear to compare favorably with these, which may be explained by the patients in our series having less severe disease. However, the data for lung function, ph at acute presentation, and Paco 2 at various stages of admission (Table 1) confirm that our patients had significant airflow obstruction with severe ventilatory failure. Only Menzies et al 7 provide equivalent data to allow comparison and, on the basis of FEV 1 and ABG levels at acute presentation, our patients had a similar degree of airflow obstruction and respiratory failure. Our results do not only compare well with data describing COPD patients needing IMV. In the 2003 National COPD Audit (First Report of the 2003 National COPD Audit [mail-merge document sent out by the Royal College of Physicians to UK hospitals, 2004]) the median Paco 2 on admission was 5.9 kpa (IQR, 5.0 to 7.3 kpa). In 1,815 episodes (23%) arterial ph was 7.35 at some point, but ventilatory support (IMV or NIV) was instituted in just 39% of these cases. There may be several explanations for this, but it demonstrates that in the United Kingdom the threshold for intubating patients with COPD is high. By contrast, all the patients in our series were intubated for ventilatory failure. When considered alongside the APACHE II score, lung function, and ABG data at intubation (Table 1), it suggests our series has a more severely 136 Original Research

Table 3 Published Survival Rates at 1 Year for COPD Patients Who Required IMV Source Year Patients Receiving IMV, No. % Weaned % Survival to Discharge 1-yr Survival, % RSSC 2004 67 95.5 92.5 69 Schönhofer et al 6 * 2002 239 76 24 Nava et al 9 * 1994 42 55 71 57 Stauffer et al 3 1993 67 81 67 44 Menzies et al 7 1989 95 76 34 Spicher and White 4 1987 13 54 31 15 Witek et al 8 1985 11 82 64 36 *Weaning center. One-year survival from start of weaning. One-year survival from time of discharge. One-year survival from intubation. One-year survival from successful weaning. ill group of patients than that described in the National COPD Audit. Nonetheless, the hospital and 90-day mortality rates (7.5% and 11.9%, respectively) are similar to those from the audit (7.5% and 15.5%). The results for our series may be due to selection bias. Those patients who were referred for weaning had survived an acute severe illness and prolonged IMV at their referring ICU. We could therefore be describing a population of survivors, although our data suggest that a greater LOS in the ICU was associated with a worse longer-term outcome. Further exploration of this would require mortality data from the referring ICUs for all patients with COPD. Comparing each patient transferred to the RSSC with a COPD patient identified from their referring ICU, matching for factors including age, APACHE II score, and ICU LOS, could clarify whether admission to the PCP confers a weaning and survival advantage. Without this information, selection bias cannot be excluded, but the condition of these patients was poor enough to prevent them responding to standard management including conventional critical care unit weaning techniques. It is also possible that the weaning results and survival data represent the positive impact of our process of care. IMV has a number of complications, 16 particularly when prolonged, with an increasing risk of ventilator-associated pneumonia and consequent mortality. 17,18 Both Nava et al 9 and Schönhofer et al 6 reported better survival for those patients successfully weaned from IMV. Another series 19 describing COPD patients receiving prolonged IMV reported a median survival of only 5 months. It seems likely that weaning the majority of our patients had a positive impact on both hospital and long-term survival. There are few studies to guide the choice of ventilation mode during weaning. Esteban et al 20 demonstrated faster weaning using spontaneous breathing than for either PSV or intermittent mandatory ventilation. However, criteria for determining intolerance of weaning differed between the spontaneous breathing and PSV groups, which may have selectively delayed weaning in the latter. Brochard et al 21 found an advantage for PSV over both synchronized intermittent mandatory ventilation and T- piece trial in terms of weaning failures. These studies employed stepwise reduction of pressure support. Neither study examined the combination of an optimal level of PSV with USB. Optimal PSV supports the spontaneous inspiratory effort, allowing diaphragmatic activity and possibly training, while preventing its fatigue. 22,23 Combining this with USB may have contributed to the high rate of weaning success in our series. NIV is a relatively new weaning technique. Of the series in Table 3, only Schönhofer et al 6 and ourselves used NIV during weaning. We introduce NIV when patients fail to progress with self-ventilation due to the development of respiratory distress or hypercapnia. Similarly, Schönhofer et al 6 administered NIV to 55 COPD patients who acquired hypercapnia when off IMV. There was no report of NIV influence on weaning outcome, but the 39 patients who tolerated it continued with long-term NIV. There are other data supporting the use of NIV in weaning. In a trial 24 of 50 COPD patients still requiring IMV after 48 h, 25 patients were randomized to extubation and NIV and 25 patients to weaning without NIV. In the NIV study arm, there were reductions in duration of IMV, ICU LOS, and incidence of ventilator-associated pneumonia; and increases in 60-day survival, probability of survival during weaning, and weaning success. Hilbert et al 25 compared 30 COPD patients receiving NIV for ventilatory failure after extubation with 30 historical control subjects. The NIV group had lower reintubation rates, duration of ventilatory support, and ICU LOS. A larger randomized trial 26 compared www.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 137

NIV to standard medical therapy for respiratory failure after extubation. The trial was stopped early due to a higher rate of death in the NIV study arm, apparently due to mortality differences in those who were reintubated. Reintubation rates were similar, but time to reintubation was longer in those receiving NIV (p 0.02), suggesting that it led to delay in reinstitution of IMV. Not all patients had hypercapnia, and only 23 of the 221 patients were intubated for COPD. Of those, the rate of reintubation was lower in the NIV study arm (not significant), but the sample was too small to allow further analysis. The applicability of these results to COPD patients intubated for ventilatory failure is therefore limited. Although our retrospective series is uncontrolled and comprises patients who had a more protracted ICU admission, the results suggest a role for the selective use of NIV when weaning COPD patients from IMV. Maintenance NIV after discharge was associated with better long-term survival in our series. Schönhofer et al 6 found that survival duration was associated with the use of NIV (p 0.0001), but there was no subgroup analysis for COPD. Case series have reported relatively good survival for COPD patients treated with NIV for chronic ventilatory failure. 27,28 Randomized controlled trials 29,30 have not shown a survival benefit, but it is not clear they achieved effective nocturnal ventilation. The mean inspiratory pressures were set at only 12 to 14 cm H 2 O, and the efficacy of nocturnal ventilation was assessed either clinically 29 or with oximetry. 30 Nocturnal carbon dioxide tensions were not assessed, and low oxygen saturations were corrected by increasing inspired oxygen rather than the level of ventilatory support. Higher inspiratory pressures are required to normalize Paco 2. 31 We used nocturnal oximetry and Ptco 2 plus daytime ABG levels to guide pressure titration, aiming to normalize Paco 2. The mean inspiratory pressure required was 28 cm H 2 O. The allocation of long-term NIV in our group was based on clinical criteria and not randomized. The non-niv group, although similar to those receiving NIV in terms of age, LOS, and lung function, had statistically significantly better ABG levels at discharge (Table 2). The survival advantage, despite worse ABG levels, for those receiving maintenance NIV suggests that even if it does not fully correct ventilatory failure, NIV may still improve survival. The study by Chu et al 14 examined outcomes of COPD patients who received hospital NIV for acute ventilatory failure. They found that 63% had a further episode of acute hypercapnia requiring assisted ventilation (NIV or IMV) within 1 year of discharge, and 49% had died. All the patients in our series had a history of acute severe ventilatory failure. It may be that domiciliary NIV protects these patients during subsequent exacerbations of COPD, reducing the degree of acute ventilatory deterioration and its associated mortality. We report a 95.5% weaning rate and 92.5% survival to discharge for COPD patients dependent on IMV who were admitted to our center. These outcomes are better than predicted from the APACHE II scores or in previously reported series. Bias in the referral of patients to our service cannot be excluded, but all patients had severe COPD and had failed to progress with weaning in their local ICU. Our results were achieved through the application of a process of care by a multidisciplinary team, which includes the combination of PSV with unsupported spontaneous breathing and the use of NIV. The survival advantage demonstrated for patients discharged on maintenance NIV suggests it has a role in the management of a subgroup of patients with severe COPD and a history of difficulty in weaning from IMV. ACKNOWLEDGMENT: The authors thank the Intensive Care National Audit and Research Centre for their help in providing APACHE II data, and Fay Cafferty for statistical input. References 1 Brochard L, Mancebo J, Wysocki M, et al. Noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease. N Engl J Med 1995; 333:817 822 2 Plant PK, Owen JL, Elliott MW. Early use of non-invasive ventilation for acute exacerbations of chronic obstructive pulmonary disease on general respiratory wards: a multicentre randomised controlled trial. Lancet 2000; 355:1931 1935 3 Stauffer JL, Fayter NA, Graves B, et al. Survival following mechanical ventilation for acute respiratory failure in adult men. Chest 1993; 104:1222 1229 4 Spicher JE, White DP. Outcome and function following prolonged mechanical ventilation. Arch Intern Med 1987; 147:421 425 5 Swinburne AJ, Fedullo AJ, Bixby K, et al. Respiratory failure in the elderly: analysis of outcome after treatment with mechanical ventilation. Arch Intern Med 1993; 153:1657 1662 6 Schönhofer B, Euteneuer S, Nava S, et al. Survival of mechanically ventilated patients admitted to a specialised weaning centre. Intensive Care Med 2002; 28:908 916 7 Menzies R, Gibbons W, Goldberg P. Determinants of weaning and survival among patients with COPD who require mechanical ventilation for acute respiratory failure. Chest 1989; 95:398 405 8 Witek TJ, Schachter EN, Dean NL, et al. Mechanically assisted ventilation in a community hospital. Arch Intern Med 1985; 145:235 239 9 Nava S, Rubini F, Zanotti E, et al. Survival and prediction of successful ventilator weaning in COPD patients requiring mechanical ventilation for more than 21 days. Eur Respir J 1994; 7:1645 1652 10 Esteban A, Alia I, Ibanez J, et al. Modes of mechanical ventilation and weaning: a national survey of Spanish hospitals. Chest 1994; 106:1188 1193 138 Original Research

11 Gracey DR, Viggiano RW, Naessens JM, et al. Outcomes of patients admitted to a chronic ventilator-dependent unit in an acute-care hospital. Mayo Clin Proc 1992; 67:131 136 12 Smith IE, Shneerson JM. A progressive care programme for prolonged ventilatory failure: analysis of outcome. Br J Anaesth 1995; 75:399 404 13 NHS Strategic Tracing Service. Available at: www.nsts. nhsia.nhs.ukhttp://www.nsts.nhsia.nhs.uk. Accessed December 5, 2005 14 Chu CM, Chan VL, Lin AW, et al. Readmission rates and life-threatening events in COPD survivors treated with noninvasive ventilation for acute hypercapnic respiratory failure. Thorax 2004; 59:1020 1025 15 Connors AF, Dawson NV, Thomas C, et al. Outcomes following acute exacerbation of severe chronic obstructive lung disease. Am J Respir Crit Care Med 1996; 154:959 967 16 Tobin MJ. Mechanical ventilation. N Engl J Med 1994; 330:1056 1061 17 Fagon JY, Chastre J, Domart Y, et al. Nosocomial pneumonia in patients receiving continuous mechanical ventilation: prospective analysis of 52 episodes with use of a protected specimen brush and quantitative culture techniques. Am Rev Respir Dis 1989; 139:877 884 18 Kollef MH, Silver P, Murphy DM, et al. The effect of late-onset ventilator-associated pneumonia in determining patient mortality. Chest 1995; 108:1655 1662 19 Votto J, Brancifort JM, Scalise PJ, et al. COPD and other diseases in chronically ventilated patients in a prolonged respiratory care unit: a retrospective 20-year survival study. Chest 1998; 113:86 90 20 Esteban A, Frutos F, Tobin M, et al. A comparison of four methods of weaning patients from mechanical ventilation. N Engl J Med 1995; 332:345 350 21 Brochard L, Rauss A, Benito S, et al. Comparison of three methods of gradual withdrawal from ventilatory support during weaning from mechanical ventilation. Am J Respir Crit Care Med 1994; 150:896 903 22 Brochard L, Harf A, Lorino H, et al. Inspiratory pressure support prevents diaphragmatic fatigue during weaning from mechanical ventilation. Am Rev Respir Dis 1989; 139:513 521 23 Shneerson J. Muscle dysfunction and non-invasive ventilation. Monaldi Arch Chest Dis 2003; 59:234 237 24 Nava S, Ambrosino N, Clini E, et al. Noninvasive mechanical ventilation in the weaning of patients with respiratory failure due to chronic obstructive pulmonary disease: a randomized, controlled trial. Ann Intern Med 1998; 128:721 728 25 Hilbert G, Gruson D, Portel L, et al. Noninvasive pressure support ventilation in COPD patients with postextubation hypercapnic respiratory insufficiency. Eur Respir J 1998; 11:1349 1353 26 Esteban A, Frutos-Vivar F, Ferguson ND, et al. Noninvasive positive-pressure ventilation for respiratory failure after extubation. N Engl J Med 2004; 350:2452 2460 27 Jones SE, Packham S, Hebden M, et al. Domiciliary nocturnal intermittent positive pressure ventilation in patients with respiratory failure due to severe COPD: long term follow up and effect on survival. Thorax 1998; 53:495 498 28 Sivasothy P, Smith IE, Shneerson J. Mask intermittent positive pressure ventilation in chronic hypercapnic respiratory failure due to chronic obstructive pulmonary disease. Eur Respir J 1998; 11:34 40 29 Casanova C, Celli BR, Tost L, et al. Long-term controlled trial of nocturnal nasal positive pressure ventilation in patients with severe COPD. Chest 2000; 118:1582 1590 30 Clini E, Sturani C, Rossi A, et al. The Italian multicentre study on noninvasive ventilation in chronic obstructive pulmonary disease patients. Eur Respir J 2002; 20:529 538 31 Windisch W, Vogel M, Sorichter S, et al. Normocapnia during nippv in chronic hypercapnic COPD reduces subsequent spontaneous Paco 2. Respir Med 2002; 96:572 579. www.chestjournal.org CHEST / 129 / 1/ JANUARY, 2006 139