Geographic Variation in Long-term
Oxygen Therapy in Denmark*
Factors Related to Adherence to Guidelines for
Long-term Oxygen Therapy
Thomas J. Ringbaek, MD; Peter Lange, MD; and Kaj Viskum, MD
Study objectives: To evaluate regional differences in adherence to guidelines for long-term
oxygen therapy (LTOT) in Denmark and to determine factors related to compliance with these
guidelines.
Design: Cross-sectional study and analysis of a nationwide database (Danish Oxygen Register).
Setting: Denmark.
Patients: In November 1994, 1,354 COPD patients were receiving LTOT in Denmark.
Measurements and results: Among 16 counties, the prevalence of LTOT for COPD varied from 14
to 53 per 100,000. The prevalence was highest in counties where general practitioners (GPs) were
prescribing LTOT. Adherence to national guidelines for LTOT was found in 34.4% of the patients
for the whole of Denmark and varied regionally from 14 to 63%. Mean compliance with
guidelines was 5.3 (range, 2.9 to 9.1) times as likely if the oxygen was prescribed by a pulmonary
department compared to LTOT initiated by a GP.
Conclusions: Marked geographic variations in compliance with LTOT guidelines are present even
in a small country as Denmark. In general, the adherence to the guidelines is poor, especially
when non-chest physicians prescribe LTOT. We therefore recommend that local and national
thoracic societies together with health organizations responsible for treatment should play a
more forceful role in implementing the guidelines. This could be done by enhanced educational
efforts, by monitoring of adherence, or even by centralizing the prescription right to departments
with pulmonary physicians.
(CHEST 2001; 119:1711–1716)
Key words: COPD; home care; guidelines; monitoring; oxygen therapy; respiratory failure
Abbreviations: GP ⫽ general practitioner; LTOT ⫽ long-term oxygen therapy
uidelines for long-term oxygen therapy (LTOT)
G are
Most
based on three controlled studies.
1–3
guidelines focus on three key issues: COPD patients
eligible for LTOT should have hypoxemia in a stable
condition, they should be nonsmokers, and they should
use oxygen for at least 15 h/d (use of oxygen 24 h/d is
the optimal goal).4 The best overall information on
adherence to the guidelines comes from Sweden and
France.5,6 In addition, there are regional surveys from
Poland, United Kingdom, Spain, Italy, and the Netherlands.7–16 In most studies,5–15 only 40 to 70% of the
patients used oxygen ⱖ 15 h/d. Some studies8 –10 have
*From the Department of Respiratory Medicine (Drs. Ringbaek
and Lange), University Hospital of Copenhagen, Hvidovre Hospital; and Department of Respiratory Medicine (Dr. Viskum),
University Hospital of Copenhagen, Gentofte Hospital.
Supported by the Danish Lung Association.
Manuscript received February 1, 2000; revision accepted January
23, 2001.
Correspondence to: Thomas J. Ringbaek, MD, Moseskraeten 17,
DK-3140 Aalsgaarde, Denmark; e-mail: [email protected]
found that more patients were not fulfilling the recommendations for LTOT if their treatments were initiated
by a general practitioner (GP), compared with a pulmonary physician. Yet, Waterhouse et al11 found that
patients with treatment initiated by GPs were prescribed oxygen the same number of hours per day as
patients with treatment initiated by hospital doctors.
A regional difference in the use of oxygen has been
shown in two studies.5,17 However, to our knowledge
regional differences in adherence to guidelines for
LTOT have not previously been studied. The aims of
the present study were to investigate regional differences and to evaluate factors related to compliance
with criteria for LTOT.
Materials and Methods
Collection of Information
The Danish Oxygen Register was established in November
1994. The patients were identified through the oxygen suppliers.
CHEST / 119 / 6 / JUNE, 2001
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21963/ on 06/18/2017
1711
Except for one supplier, all oxygen suppliers provided information on patients receiving LTOT, their prescriptions, and their
oxygen systems. One supplier, who covered 2% of the Danish
population, provided only information on the number of patients
(which covered some 4% of all patients receiving LTOT). From
the patients’ hospital files or GP files, we achieved information on
cause of hypoxemia, medical treatment, smoking habits, and
arterial blood gas tensions. The highest Pao2 values (on room air)
between May 1993 and November 1994 were registered.
On November 1, 1994, a total of 2,190 patients received home
oxygen therapy. Of these patients, 1,835 adult patients had
hypoxemia due to cardiopulmonary disease, including 1,354
patients (73.8%) who had hypoxemia due to COPD (Table 1).
The files of 16 patients with COPD could not be found.
In November 1994, the patients received a questionnaire by
mail. The questionnaire requested information on daily use of
oxygen (hours per day) according to the patient and was answered
by 72% of the COPD patients (responders). The most pronounced differences between responders and nonresponders
were the 3-month mortality rate (6.5% vs 24.4%) and prevalence
of current smokers (17.2% vs 32.8%; Table 2). In the present
study, only one variable (oxygen used according to the patient)
stemmed from the questionnaire, while all other information
originated from the hospital files or GP files. The regional ethical
committees and the Data Inspection Board have approved the
study.
Table 2—Differences Between COPD Patients Who
Answered the Questionnaire (Responders) and
Nonresponders*
p
Responders Nonresponders Value
Variables
Total patients, No. (%)
Age mean, yr
Gender, % female
Median duration of LTOT, mo
Prescribed at least 15 h/d, %
Flow mean, L/min
Concentrator, %
Mobile system, %
Current smoking, %
Pao2 on room air ⱕ 55 mm Hg, %
Pao2 on room air ⬎ 55 mm Hg, %
3-mo mortality, %
973 (72)
69.2
58.1
12.7
49.8
1.3
57.2
31.5
17.2
50.6
19.5
6.5
381 (28)
70.8
54.3
10.5
47.8
1.2
51.4
19.7
32.8¶
43.2
27.5
24.4
†
‡
†
‡
‡
‡
§
§
†
㛳
§
*Data are presented as No. unless otherwise indicated.
†p value ⬍ 0.05.
‡p value ⬎ 0.05.
§p value ⬍ 0.001.
㛳p value ⬍ 0.01.
¶Data from the hospital or GP file.
Guidelines and Organization of LTOT in Denmark
Statistics
The recommended guidelines for prescribing LTOT, issued by
the Danish Society of Respiratory Medicine, are presented in
Table 3.18 Denmark consists of 16 counties with populations
ranging from 45,000 to 614,000. Each county has its own
organization of LTOT. In eight counties, the GPs took part in
prescribing home oxygen. However, as GPs in Denmark are not
able to test blood gases, they either must rely on previous tests at
the hospital or must assume that the patient had hypoxemia when
he or she had severe symptoms and perhaps poor lung function.
Liquid oxygen became available in July 1995. In general, the
suppliers were able to choose between a concentrator and gas
cylinders. In cases of large consumption of oxygen, it was more
favorable for the suppliers to deliver a concentrator. All costs of
LTOT, including consumption of electricity by oxygen concentrators, are covered by the local hospital. This is also the case if a
GP prescribes LTOT.
Analyses were performed using software (Statistical Package
for the Social Sciences, version 8.0; SPSS; Chicago, IL). The ␹2,
two-sample t tests, and Mann-Whitney U tests were used as
appropriate to compare differences between groups. A multiple
logistic regression analysis was employed with “adherence to
guidelines” as the dependent variable, and entering gender and
the type of doctor (GP, internist, chest physician at nonpulmonary department and at pulmonary department) as covariables. A
two-sided p value of ⬍ 0.05 was considered significant.
Table 1—Selection of Patients to the Danish Oxygen
Register Receiving LTOT on November 1, 1994*
Variables
All patients receiving LTOT
Unknown due to incomplete information
from one supplier
Cluster headache
Age ⬍ 18 yr
Diagnosis not available
Adults with cardiopulmonary disease
COPD, No. (%)
Lung cancer, No. (%)
Cardiac disease, No. (%)
Other diseases†, No. (%)
Data
2,190
95
231
23
6
1,835
1,354 (73.8)
166 (9.0)
92 (5.0)
223 (12.2)
*Data are presented as No. unless otherwise indicated.
†Lung fibrosis, pneumoconiosis, pulmonary sarcoidosis, neuromuscular diseases, kyphoscoliosis, cystic fibrosis, pulmonary embolism,
adiposity, and sequelae from pulmonary tuberculosis.
Results
The prevalence of LTOT varied from 14 to 53 per
100,000 in the 16 counties. Characteristics of the
patients are shown in Table 4. Chest specialists at
pulmonary departments initiated treatment in 439
patients (32.5%), chest specialists working at general
medical departments initiated treatment in 285 patients (21.1%), internists initiated treatment in 270
patients (20.0%), and GPs initiated treatment in the
remaining 358 patients (26.5%). Most patients with
Table 3—Danish Guidelines for LTOT
Pao2 on room air ⱕ 55 mm Hg in clinically stable and optimally
treated condition
A marked increase in Pao2 after oxygen supply (aiming at 60 to 75
mm Hg) without a substantial increase in Paco2
No smoking
Patient motivated for and able to manage the therapy
Oxygen administered and used for at least 15 h/d
Control every 3 to 6 mo
1712
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21963/ on 06/18/2017
Clinical Investigations
Table 4 —Characteristics of Patients With COPD Receiving LTOT in Denmark and in Counties With Lowest and
Highest Value (Geographic Range), and Differences Between Treatment Modality and Patient Characteristics in
Patients With Treatments Prescribed by GPs and Hospital Doctors*
Characteristics
No. per 100,000
Age, yr (n ⫽ 1,354)
Female patients, %
(n ⫽ 1,354)
LTOT prescribed by GPs, %
(n ⫽ 1,352)
Duration of LTOT, mo
(n ⫽ 1,353)
Flow of oxygen, L/min
(n ⫽ 1,279)
Patients prescribed oxygen
(n ⫽ 1,354), %
ⱖ 15 h/d
⬍ 15 h/d
As needed
No information
Used oxygen according to
the patient (n ⫽ 937), %
As needed
⬍ 15 h/d
15–24 h/d
Mobile oxygen, %
(n ⫽ 1,347)
Concentrator, % (n ⫽ 1,347)
Stationary cylinder, %
(n ⫽ 1,347)
Current smokers, %
(n ⫽ 1,290)
All COPD Patients
in Denmark
(n ⫽ 1,354)
Geographic
Range
Prescribed by GPs
(n ⫽ 358†)
Prescribed by HDs
(n ⫽ 994)
p Value
14–53
65–72
48–69
—
71.4
47.2
—
69.0
60.6
⬍ 0.001
⬍ 0.001
26.5
0–60
—
—
12.1 (0.0–146.1)
6–31
16.0
11.5
⬍ 0.001
0.5–1.5
1.5
1.2
⬍ 0.001
49.3
11.7
24.3
14.7
5–90
0–26
3–68
0–79
31.3
20.4
31.6
16.8
55.8
8.7
21.7
13.8
⬍ 0.001
⬍ 0.001
⬍ 0.001
⬎ 0.05
5.9
34.9
59.2
28.1
0–21
13–56
35–86
4–55
10.8
47.8
41.4
20.2
4.1
30.2
65.7
31.0
⬍ 0.001
⬍ 0.001
⬍ 0.001
⬍ 0.001
55.6
47.3
25–93
6–84
39.2
61.5
⬍ 0.001
21.1
8–41
23.8
20.2
⬎ 0.05
26.8
69.6 (8.7)
57
1.3 (0.7)
*Data are presented as mean (SD) or median (range) unless otherwise indicated. HD ⫽ hospital doctor.
†The specialty of the prescribing doctor was not known in two cases.
COPD were female (57%). Based on existing hospital files, the Danish Oxygen Register recorded blood
gas measures after the first of May 1993. Oxygen
therapy was initiated before that date in 536 patients
(39.6%). Only half of the patients were prescribed
oxygen at least 15 h/d, and a corresponding number
of patients reported use of oxygen at least 15 h/d.
Among the responders, we observed a positive significant association between correct prescription of
LTOT (15 to 24 h/d) and daily oxygen usage according to the patient (p ⬍ 0.001). Nearly one of four
patients (24.3%) was prescribed oxygen as needed,
while only 5.9% of the patients stated that they used
oxygen as needed. Fortunately, despite incorrect
prescription (⬍ 15 h/d), 111 patients (34.9%) were
receiving oxygen at least 15 h/d. Although the guidelines recommend that LTOT should be administered
only to nonsmokers, 21.1% of the patients admitted
that they were still smoking (Table 4).
Altogether, of 535 patients with sufficient information, only 34.4% fulfilled the following criteria:
objective verified hypoxemia, blood gases measured
with oxygen supply, nonsmoking, and use of oxygen
at least 15 h/d (compliance with guidelines).
Geographic Variation Among Patients With COPD
Receiving LTOT
Even among patients with the same diagnosis,
there was a considerable difference among the counties (Table 4). In the counties with the largest cities,
more than half of the patients were prescribed
oxygen at least 15 h/d; in accordance with this,
patients stated that they used oxygen at least 15 h/d.
In the rural counties, less than half of the patients
were prescribed ⱖ 15 h/d, and patients stated that
they used oxygen fewer hours. Compliance with
guidelines varied from 14 to 63% in the 16 counties.
The prevalence of LTOT was significantly higher
in those eight counties where GPs took part in LTOT
prescribing, compared with the rest of the counties:
34.4 per 100,000 vs 19.2 per 100,000 (p ⬍ 0.001).
Generally, if GPs had initiated LTOT, the patients
were older, more often male, more often prescribed
CHEST / 119 / 6 / JUNE, 2001
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21963/ on 06/18/2017
1713
oxygen as needed or ⬍ 15 h/d, more often using
oxygen ⬍ 15 h/d, more seldom delivered a concentrator and a mobile system, had more seldom detected hypoxemia, and were prescribed a higher
oxygen flow (Tables 4, 5). Incorrect prescriptions of
LTOT by GPs were, however, not able to explain all
the geographic differences. For instance, in one
county, where only hospital doctors were prescribing
LTOT, 41% of the patients were current smokers
and hypoxemia was only documented in 34% of the
patients who started after May 1993. In a univariate
analysis, female patients were more often compliant
with guidelines than male patients: 39% vs 28%
(p ⫽ 0.015). However, when a multiple regression
analysis was employed to determine whether gender
and the type of doctor (chest specialist, internist, or
GP) had an impact on compliance with guidelines,
the influence of gender was no longer significant.
When GPs and male gender were set as reference,
the odd ratios (95% confidence intervals) for good
compliance were as follows: female gender, 1.4 (1.0
to 2.1); internist, 1.6 (0.9 to 2.9); chest specialist
at nonpulmonary department, 2.1 (1.1 to 3.9);
and chest specialist at pulmonary department, 5.3
(2.9 to 9.1).
Discussion
The prevalence of LTOT and the treatment modality varied markedly among the different counties
of Denmark. A great deal of this variation was due to
prescription by the GPs in some counties. The
prevalence of LTOT in patients with COPD varied
from 14 to 53 per 100,000. The hospital admission
rates for patients with COPD in the different counties vary only with a factor of 1.66, and are not
correlated to the prevalence of LTOT. Thus, the
variation in the prevalence of LTOT documented in
this study seemed not to reflect the prevalence of
severe COPD, as judged by hospital admissions. GPs
were frequently prescribing oxygen fewer hours than
recommended and started LTOT when the disease
was less advanced (Pao2 on room air ⬎ 55 mm Hg)
in comparison with chest physicians. This is in
accordance with previous findings.8 –10,12
The positive correlation between correct prescription (15 to 24 h/d) and daily oxygen usage suggests a
better compliance with treatment in those patients
whose treatment was initiated by a doctor who
adhered to guidelines. Fortunately, in our study,
35% of the patients with inadequately prescribed
number of hours with oxygen reported an adequate
usage. A similar finding has also been reported in
two small studies from the United Kingdom.12,19 The
explanation for this may be that some patients have
read or have been told by others to use oxygen more
hours per day, or that patients prescribed oxygen
⬍ 15 h/d or as needed may actually have needed
oxygen ⬎ 15 h/d.
An important limitation of our study is that compliance with usage of oxygen was based on selfreported data. We have previously investigated the
reliability of self-reported compliance in a single
county.20 Of 125 patients, the reported and the
actual number of spent hours with oxygen were, on
average, 17.7 h/d and 17.4 h/d, respectively. However, of patients reporting compliance with usage of
oxygen (15 to 24 h/d), the actual usage based on the
oxygen consumption was ⬍ 15 h/d in 29%. Contrary,
of patients reporting noncompliance, the calculated
daily number of hours with oxygen exceeded 15 h in
25%. The study was not able to answer whether this
Table 5—Characteristics of COPD Patients Who Started LTOT After May 1993, Presented for the Whole Country
and in Counties With Lowest and Highest Value (Geographic Range); Secondly, Differences Between Treatment
Modality and Patient Characteristics in Patients Prescribed by GPs and Hospital Doctors*
Characteristics
Whole
Denmark
(n ⫽ 817†)
Geographic
Range
Prescribed
by GPs
(n ⫽ 186)
Prescribed
by HDs
(n ⫽ 631)
p Value
10.9
12.4
69.1
10.9
0–35
0–33
37–85
0–26
20.1
8.9
45.8
25.1
10.7
7.2
75.8
6.8
⬍ 0.001
⬎ 0.05
⬍ 0.001
⬍ 0.001
57.5
25.4
17.1
34.4
40–81
0–45
5–30
14–63
33.5
33.0
33.5
17.6
64.3
23.2
12.4
39.0
⬍ 0.001
⬍ 0.01
⬍ 0.001
⬍ 0.001
Measured blood gases (n ⫽ 807), %
Pao2 (on room air) measured
Pao2 with O2 measured
Pao2 with O2 and on room air
Not measured
Pao2 on room air (n ⫽ 807), %
ⱕ 55 mm Hg detected
⬎ 55 mm Hg detected
Not measured
Compliant with guidelines (n ⫽ 547), %
*See Table 4 for expansion of abbreviation.
†The date of initiation of LTOT was not known in one case.
1714
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21963/ on 06/18/2017
Clinical Investigations
underestimation by the patients was real or was
reflecting a wasted consumption of oxygen. Two
other studies7,12 have also found overestimation of
compliance based on self-reported data. In the British study,12 the mean difference between reported
and actual number of hours per day was 1.5. In the
other study,7 16 of 54 patients had misreported
compliance. Assuming a similar degree of patient
overestimation of compliance in our population
makes the performance of all physicians with regard
to achieving proper compliance with LTOT even
worse. Another limitation of our study is that data
were obtained 6 to 7 years ago. Although there have
been no changes in guidelines for prescribing LTOT
since this time, it is possible that there could have
been changes in the way the guidelines were interpreted and/or implemented.
Despite the fact that all national guidelines exclude current smokers as candidates for LTOT, this
criterion is obviously enforced differently. It has
been reported that 8 to 20% of the patients were still
smoking. Our finding of approximately 21% places
Denmark in the top.5,6,8,12–15 As patients were not
checked for tobacco smoking routinely, these figures
may even be underestimated. Neither the British,
the American, nor our guidelines on LTOT have
stated how smoking cessation should be ensured.4,18,21 We suggest that testing for carboxyhemoglobin in either the expired air or the blood
should be recommended in future guidelines on
LTOT.
Gender Differences in LTOT
More female than male patients were receiving
home oxygen. Although a lot of Danish women have
been heavy smokers, the prevalence of COPD is still
higher in men. In Sweden, more female patients
were receiving home oxygen as well.5 An explanation
could be that female patients were seeking doctors
more frequently, and that they were more willing to
accept the inconveniences of LTOT (smoking cessation and use of oxygen for at least 15 h/d). Supporting this hypothesis, we found that when the data
were unadjusted for type of prescribing doctor, a
significantly higher percentage of male patients than
female patients fulfill the criteria of home oxygen.
Although only hospital doctors were prescribing
LTOT in Sweden, The Swedish Oxygen Register has
also shown that the prevalence of LTOT varied
regionally from 3.6 to 13.3 per 100,000, and even
more at county level.5 They concluded that a substantial part of the variation had to be explained by
differences in knowledge and habits of physicians
treating patients with respiratory failure. A markedly
uneven consumption of oxygen has been found in a
region of England.17 Neither the Swedish nor the
English study has focused on regional differences in
treatment modality and adherence to guidelines.
According to the guidelines, it is unacceptable that
GPs initiate LTOT without measuring the blood
gases. Besides, in a decentralized organization of
LTOT, most GPs will never get sufficient experience
with LTOT because they will treat very few patients.
More patients adhered to guidelines when chest
specialists prescribed LTOT— especially when the
doctor worked at a pulmonary department. A similar
difference in adherence to guidelines between chest
physicians and internists was found in a minor
district of United Kingdom.12 In Denmark, even in
rural areas, patients are living in the vicinity of a
chest physician. It is therefore feasible to recommend that only chest physicians should be able to
prescribe LTOT. In the areas of the university
hospitals, only doctors at the pulmonary departments
should be allowed to initiate this expensive therapy.
We acknowledge that recommendation on centralized organization is difficult to implement in countries with long distances between chest physicians. In
this situation either internists or GPs capable of
measuring blood gases could care for patients receiving LTOT. Even with a centralized organization of
LTOT, several studies, including ours, found considerable room for improvement. Therefore, in addition, the local thoracic societies should play a more
forceful role in implementing the guidelines through
education of all LTOT prescribers, frequent monitoring of adherence, and by the possibility of refused
reimbursement if the guidelines are not followed.
Conclusion
This study of almost all patients treated with
LTOT in Denmark showed substantial geographic
differences among the different counties, which in
part was explained by prescription of LTOT by GPs
in some but not in other counties. Adherence to the
guidelines was poor, especially when GPs initiated
LTOT. We therefore recommend that local and
national thoracic societies together with health organizations responsible for the treatment should play a
more forceful role in implementing the LTOT
guidelines. This could be done by enhanced educational efforts, monitoring of adherence, or even by
centralizing the prescription right to departments
with pulmonary physicians.
References
1 Nocturnal Oxygen Therapy Trial Group. Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive lung
CHEST / 119 / 6 / JUNE, 2001
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21963/ on 06/18/2017
1715
disease. Ann Intern Med 1980; 93:391–398
2 Medical Research Council Working Party. Report of longterm domiciliary oxygen therapy in chronic hypoxic cor
pulmonale complicating chronic bronchitis and emphysema.
Lancet 1981; 1:681– 685
3 Gorecka D, Gorzelak K, Sliwinski P, et al. Effect of long term
oxygen therapy on survival in patients with chronic obstructive pulmonary disease with moderate hypoxemia. Thorax
1997; 52:674 – 679
4 ATS statement: standards for the diagnosis and care of
patients with chronic obstructive pulmonary disease. Am J
Respir Crit Care Med 1995; 152(suppl 5):S88 –S93
5 Ström K, Boe J. A national register for long-term oxygen
therapy in chronic hypoxia: preliminary results. Eur Respir J
1988; 1:952–958
6 Pépin J-L, Barjhoux CE, Deschaux C, et al. Long-term
oxygen therapy at home. Chest 1996; 109:1144 –1150
7 Granados A, Escarrabill J, Borràs JM, et al. The importance of
process variables analysis in the assessment of long-term
oxygen therapy by concentrator. Respir Med 1997; 91:89 –93
8 Kampelmacher MJ, Kersteren RG, Alsbach GPJ, et al. Prescription and usage of long-term oxygen therapy in patients
with chronic obstructive pulmonary disease in the Netherlands. Respir Med 1999; 93:46 –51
9 Dilworth JP, Higgs CMB, Jones PA, et al. Prescription of
oxygen concentrators: adherence to published guidelines.
Thorax 1989; 44:576 –578
10 Bellone A, Venanzi D, De Angelis G, et al. Who should
prescribe long-term oxygen in patients affected by chronic
arterial hypoxemia? Monaldi Arch Chest Dis 1994; 49:396 –
398
11 Waterhouse JC, Nichol J, Howard P. Survey on domiciliary
12
13
14
15
16
17
18
19
20
21
oxygen by concentrator in England and Wales. Eur Respir J
1994; 7:2021–2025
Walshaw MJ, Lim R, Evans CC, et al. Prescription of oxygen
concentrators for long term oxygen treatment: reassessment
in one district. BMJ 1988; 297:1030 –1032
Restrick LJ, Paul EA, Braid GM, et al. Assessment and
follow-up of patients prescribed long term oxygen treatment.
Thorax 1993; 48:708 –713
Morrison D, Skwarski K, MacNee W. Review of the prescription of domiciliary long term oxygen therapy in Scotland.
Thorax 1995; 50:1103–1105
Zielinski J, Sliwinski P, Tobiasz M, et al. Long-term oxygen
therapy in Poland. Monaldi Arch Chest Dis 1993; 48:479 –
480
Okubadejo AA, Paul EA, Wedzicha JA. Domiciliary oxygen
cylinders: indications, prescription and usage. Respir Med
1994; 88:777–785
Williams BT. Geographic variation in the supply of domiciliary oxygen. BMJ 1981; 282:1941–1943
Pallisgaard G, Petersen JT, Viskum K. Dansk Lungemedicinsk Selskabs rekommandation: iltbehandling i hjemmet ved
kronisk respirationsinsufficiens [letter]. Ugeskr Læger 1988;
150:3206
Howard P, Waterhouse JC, Billings CG. Compliance with
long-term oxygen therapy by concentrator. Eur Respir J 1992;
5:128 –129
Ringbaek T, Lange P, Viskum K. Compliance with LTOT and
consumption of mobile oxygen. Respir Med 1999; 93:333–337
BTS guidelines for the management of chronic obstructive
pulmonary disease: the COPD Guidelines Group of the
Standards of Care Committee of the BTS. Thorax 1997;
52(suppl 5):S1–28
1716
Downloaded From: http://publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21963/ on 06/18/2017
Clinical Investigations