| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Dept of Respiratory Medicine, Hvidovre Hospital, University Hospital of Copenhagen, Hvidovre and 2 Dept of Respiratory Medicine, Gentofte University, University Hospital of Copenhagen, Gentofte, Denmark
CORRESPONDENCE: T.J. Ringbaek, Krogebakke 2 B, DK-3140, Aalsgaarde, Denmark. Fax: 45 36323716. E-mail: ringbaek@dadlnet.dk
Keywords: home care, hospitalisation, obstructive lung diseases, oxygen inhalation therapy
Received: October 2, 2001
Accepted February 6, 2002
This study was supported by The Danish Lung Association.
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
15 h daily versus noncontinuous oxygen therapy (NCOT) <15 h daily). A total 246 COPD patients were studied, with each patient acting as their own control. Patients were divided into four groups: 125 patients on COT who started LTOT in conjunction with hospitalisation, 37 patients on COT who started LTOT as outpatients, 58 patients on NCOT who started LTOT in conjunction with hospitalisation and 26 patients on NCOT who started LTOT as outpatients. Admission rates, days spent in hospital and number of patients with at least one hospitalisation (ever hospitalised) were compared in two periods of 10 months before and after initiation of LTOT.
Overall during the LTOT period, in comparison with the preoxygen period, the admission rates, hospital days and "ever hospitalised" were reduced by 23.8%, 43.5% and 31.2%, respectively. Among patients who started LTOT as outpatients, a tendency towards a higher effect in the compliant group was observed.
This study shows that in hypoxaemic chronic obstructive pulmonary disease patients, long-term oxygen therapy is associated with a reduction in hospitalisation.
Long-term oxygen therapy (LTOT) is a widely used treatment that improves survival in chronic obstructive lung disease (COPD) with severe hypoxaemia 1–3. Since the cost of home oxygen therapy is high, demonstration of any economical advantage from a reduction in hospitalisation related to use of LTOT is important 4. An early study from the 1970s and two recent studies, all including a small number of patients acting as their own control, have indicated that LTOT decreases hospitalisations 5–7. However, a Medical Research Council (MRC) study, which had a randomised-control group, failed to confirm this advantage of LTOT 1. Today, it is considered unethical to undertake placebo-controlled studies in COPD patients with chronic hypoxaemia. Conversely, studies with patients acting as their own control may be biased by the fact that frequent hospitalisations and the decision to prescribe LTOT are interrelated; physicians may be more likely to initiate LTOT in patients with frequent hospitalisations rather than in patients with a stable condition. A reduction in hospitalisations after initiation of LTOT could therefore simply reflect a "regression to the mean phenomenon", a bias that has not been focused on in previous studies 5–7. Taking this into account, the effect of LTOT on hospitalisations, in a larger study with patients acting as their own control, was investigated. In addition, given that not all patients use oxygen for the recommended number of hours (at least 15 daily), the authors investigated whether compliance with hours spent on oxygen had an impact on hospitalisation.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The study comprised 246 COPD hypoxaemic patients who were divided into four groups depending on their use of oxygen (continuous oxygen therapy (COT) (patients using oxygen 15 h daily) versus noncontinuous oxygen therapy (NCOT) (patients using <15 h daily)) and the circumstances of initiation of oxygen therapy (start during hospitalisation versus start as an outpatient) (fig. 1
). Subclassification according to circumstances of initiation of oxygen therapy was owing to a hypothesis that an effect of oxygen therapy in patients who started home oxygen therapy as outpatients was less likely to be derived from a "regression to the mean phenomenon". Patients with NCOT were either prescribed oxygen <15 h daily or were prescribed COT but used it insufficiently (poor compliance). The two groups of patients with NCOT did not differ from each other regarding days spent in hospital and number of admissions before and after the start of home oxygen therapy. A previous Danish study has shown a high correlation between the actual use of oxygen and the reported use of oxygen 8. Except for usage of oxygen and prescribed mobile oxygen, small differences were found between the COT and NCOT patients (table 1). All patients had a clinical diagnosis of COPD, which in most cases was confirmed by a lung function test. Only the best spirometry value obtained during the study period was included in this study. Hospital admission rates and numbers of days in hospital were compared over two 10-month periods, before (preoxygen period) and after (oxygen period) initiation of LTOT therapy. If oxygen therapy was stopped within 10 months (n=49 patients), the preoxygen period was reduced correspondingly, omitting days from the beginning of the period. The average observation time before and after LTOT was 283 days, median time 304 days (range 46–304 days). Although reassessment is recommended 1–3 months after the start of treatment, only 50 of 147 (34%) patients who started LTOT immediately after a hospitalisation and received oxygen for at least 3 months had hypoxaemic status certified 1–3 months after start of treatment.
|
|
The Danish entry criteria for LTOT are in accordance with international guidelines, and re-assessment is recommended four times annually. The regional ethical committees and the Data Inspection Board approved the study.
Statistics
The Wilcoxon test was used to compare admission rates and days in hospital in pre- and postoxygen periods. The Sign test was used to compare "ever hospitalised". The Mann-Whitney U-test was used to compare the effect of home oxygen therapy on spent days in hospital. Two-sided tests were employed. A p<0.05 was considered significant.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
).
|
30–40% during the oxygen period (table 2
). When the effect of COT on days spent in the hospital was investigated in patients who started oxygen therapy as an outpatient, a similar reduction was found in days spent in hospital and number of patients with a hospitalisation, however the admission rates were not reduced (table 3).
|
). Patients who started COT as outpatients seemed to benefit more from oxygen therapy compared to those with NCOT, but this difference did not reach statistical significance (table 3). |
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Home oxygen therapy had a similar effect on patients in both the COT and NCOT groups, this could reflect either a bias from a connection between frequent hospitalisation and initiation of LTOT or a true effect of oxygen therapy. The majority of the COT patients in this study (77.2%) started oxygen immediately after a hospitalisation. Similarly, in a large study of all new home oxygen users in the USA in 1992, 68% of the patients had been hospitalised up to 90 days prior to oxygen therapy 12. Thus, potentially, the observed decrease in hospital days following initiation of COT may reflect a "regression to the mean phenomenon" rather than an effect of oxygen per se. However, when selecting patients who started COT in the outpatient clinic, and are therefore considered more clinically stable, the magnitude of beneficial effect on bed days was still present, although it did not reach statistical significance because of the small numbers. The reduction in admission rates was only present among patients who started COT after hospitalisation. This indicates that the decrease in hospital days, but not the decrease in number of hospitalisations, was related to COT and not to a bias from a connection between frequent hospitalisation and initiation of oxygen therapy. Correspondingly, in a small study of 10 patients on LTOT, a reduction in hospital days was observed, which probably indicated a real effect of LTOT on hospitalisations since the patients did not have increased numbers of hospital days prior to initiation of treatment 6. These 10 patients, however, may not be representative of the majority of LTOT patients, and a bias from a connection between frequent hospitalisation and initiation of oxygen therapy (even as outpatients) is impossible to role out definitively using a cohort study.
The impact of number of hours on oxygen is less clear. Previously, only the Nocturnal Oxygen Therapy Trial (NOTT) study has focused on hospitalisations where oxygen was used <15 h daily 2. As in the NOTT study, when patients who started home oxygen therapy during a clinically stable condition were investigated, it was found that using oxygen for 18 h daily was slightly better than 8–12 h in terms of reducing hospitalisation in hypoxaemic COPD patients.
According to the study by Silverman et al. 12 the most common causes of hospitalisation in patients on LTOT are heart failure, COPD exacerbations and lung infections. In hypoxaemic patients, any effect from oxygen therapy on hospitalisation could be derived from alleviating attacks of severe desaturations and from stabilising pulmonary haemodynamics 2, 13, 14. It is reasonable to expect that both NCOT and COT are able to reduce dyspnoea by reducing the effort of breathing and preventing the development of fear and anxiety, thus preventing hospitalisation of some COPD patients in a stable condition even without permanent hypoxaemia 15. Perhaps COT and NCOT combined with home care may result in an even more marked reduction of hospitalisation. In support of this hypothesis, introduction of a home-care programme with frequent visits to patients on LTOT has been shown to reduce hospitalisation rates by one-half when compared with both a historical control period and a randomised control group 10, 11. In addition, home management has been proven to reduce hospitalisation in COPD patients with exacerbations of COPD 16. This kind of home management of patients on NCOT or LTOT was not in practise in Denmark at the time of this study.
In the MRC study, the authors speculated that the failure of LTOT to reduce the hospital stay might reflect the close clinical supervision of the patients in both the oxygen treatment and the placebo group 1. Similarly, one reason for the observed reduction of hospitalisation in the studies using historical comparisons could be intensified visits to the outpatient clinic with adjunctive measures and treatment. However, in the current study only 57 (35.2%) of the COT patients were seen in the outpatient clinic within the first 6 months on oxygen therapy. Another explanation for the failure of LTOT to reduce hospitalisation when compared with placebo, could be a substantial effect of placebo. Several studies have found that placebo (sham oxygen) improves dyspnoea and exercise tolerance 17, 18.
Most of the patients in this study started LTOT immediately after hospitalisation, during a clinically unstable condition. Only 50 (34%) of the 147 patients who had LTOT for >3 months and were started on LTOT in connection with a hospitalisation had hypoxaemic status certified 1–3 months after initiation of home oxygen therapy. Previous studies have shown that 40% of these clinically unstable patients would not meet the criteria for Pa,O2 if retested 3 months later 2, 19. Therefore, it is likely that not all of the patients classified as hypoxaemic had continuous hypoxaemia. However, regarding the effect of oxygen therapy on hospitalisation, unstable patients with certified hypoxaemic status did not differ from those unstable patients without a follow-up.
To conclude, the current study shows that in hypoxaemic chronic obstructive pulmonary disease patients, long-term oxygen therapy is associated with a reduction in days spent in hospital. Furthermore, the beneficial effect of long-term oxygen therapy on hospitalisation seems to reflect an effect of therapy per se and not a "regression to the mean phenomenon".
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. S. Torok, J. D. Leuppi, F. Baty, M. Tamm, and P. N. Chhajed Combined Oximetry-Cutaneous Capnography in Patients Assessed for Long-term Oxygen Therapy Chest, June 1, 2008; 133(6): 1421 - 1425. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. D. Ramsey and F. D. R. Hobbs Chronic Obstructive Pulmonary Disease, Risk Factors, and Outcome Trials: Comparisons with Cardiovascular Disease Proceedings of the ATS, September 1, 2006; 3(7): 635 - 640. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. L. Croxton, W. C. Bailey, and for the NHLBI Working Group on Long-term Oxygen Tr Long-term Oxygen Treatment in Chronic Obstructive Pulmonary Disease: Recommendations for Future Research: An NHLBI Workshop Report Am. J. Respir. Crit. Care Med., August 15, 2006; 174(4): 373 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Cerasoli Jr Developing the Ideal Inhaled Corticosteroid Chest, July 1, 2006; 130(1_suppl): 54S - 64S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Scott, P Walker, and P M A Calverley COPD exacerbations {middle dot} 4: Prevention Thorax, May 1, 2006; 61(5): 440 - 447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T J Ringbaek, P Fabricius, and P Lange The effect of home oxygen therapy on hospitalization in moderate hypoxaemic COPD Chronic Respiratory Disease, April 1, 2005; 2(2): 107 - 108. [Abstract] [PDF]
|
|
|
|
|
|
P. M. A. Calverley Reducing the Frequency and Severity of Exacerbations of Chronic Obstructive Pulmonary Disease Proceedings of the ATS, April 1, 2004; 1(2): 121 - 124. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
