| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departments of Clinical Epidemiology and Biostatistics (Dr. Guyatt and Ms. Austin) and Medicine (Mr. Weaver), McMaster University, Hamilton, Ontario; the Department of Medicine (Dr. McKim and Ms. Norgren), University of Ottawa, Ottawa, Ontario; and the Department of Medicine (Dr. Goldstein and Mr. Bryan), University of Toronto, Toronto, Ontario, Canada.
Correspondence to: Roger S. Goldstein, MD, FCCP, Division of Respiratory Medicine, West Park Hospital, 82 Buttonwood Ave, Toronto, Ontario M6M 2J5, Canada;
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Design: Cross-sectional survey.
Participants: Two hundred thirty-seven patients receiving domiciliary oxygen in the Ontario Ministry of Health Home Oxygen Program (HOP).
Methods: A respiratory therapist visited the patients’ homes and administered questionnaires, obtained resting arterial blood gas measurements, and conducted a standardized home exercise test while monitoring oxygen saturation using an oximeter.
Measures of outcome: We evaluated the extent to which patients met HOP criteria that are based on the inclusion criteria of randomized trials showing the life-prolonging effects of domiciliary oxygen. We also assessed the extent to which the patients’ oxygen prescription was consistent with the results of rest and exercise testing.
Results:
Ninety-six of 237 participants (40.5%; 95% confidence interval, 34.3
to 46.8) did not meet criteria for home oxygen. Patients aged 
70
years were more likely to meet criteria (71 of 105 patients; 67.9%)
than those > 70 years old (70 of 132 patients; 53.0%). The
proportion of patients meeting criteria was similar whether the
referring physician was a specialist (71 of 112 patients; 62.5%) or a
primary-care physician (69 of 123 patients; 56.1%). A very important
health benefit from oxygen was identified among 82% of those who met
criteria and 88% of those who did not. Patients received higher flow
rates than our criteria suggested were appropriate. Agreement between
the independently assessed oxygen prescription at rest and the
patients’ report of oxygen use was extremely poor (chance-corrected
agreement [
], 0.17), as was agreement concerning optimal exercise
flow rates (, 0.26).
Conclusions: Current procedures for administration and reimbursement of home oxygen result in a large proportion of recipients not meeting criteria, as well as the prescription of excessive oxygen flow rates. These results are likely to apply to many jurisdictions and suggest a large potential for more efficient resource allocation.
Key Words: audit • home assessment • long-term oxygen
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Different jurisdictions show a varying willingness to generalize beyond COPD patients with rest hypoxemia. Whatever their eligibility criteria, third-party payers rely on physicians to make appropriate referrals for domiciliary oxygen. Appropriate referral requires an understanding of the criteria for LTOT, an awareness of the necessity for optimal treatment, and a period of stability before referral.
In our jurisdiction (Ontario, Canada; population, 11 million), physicians referring patients for government-reimbursed LTOT must provide evidence that patients meet criteria, including providing the results of a resting arterial blood gas measurement. Documentation of the persisting necessity for home oxygen relies on the documentation of oxygen saturation. This assessment is not standardized either in terms of the exercise employed or the target level of saturation.
In Ontario, > 15,000 patients receive LTOT under the Home Oxygen Program (HOP) of the Ministry of Health. Costs for this program rose from $18 million (Canadian) in 1985 to $55 million in 1995. We undertook this study to establish the proportion of patients currently receiving oxygen in whom its administration is warranted, and to evaluate the appropriateness of the oxygen dose they receive. This study is of importance to any health-delivery system that includes the provision of domiciliary oxygen.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patient Assessment
A respiratory therapist (RT) visited consenting patients in
their homes, measured simple spirometry, and obtained information
regarding medical history, use of oxygen, degree of activity, and
extent to which the patients experienced dyspnea during daily living.
Our criteria for adequate management of COPD patients included their
receiving an inhaled ß-agonist, an inhaled anticholinergic agent,
and, if they experienced important nocturnal dyspnea, a trial of an
oral theophylline. We stipulated that patients with heart failure
should be receiving an angiotensin-converting enzyme inhibitor.
We considered patients to be unstable if they had been discharged from an acute-care hospital, or had had an exacerbation (defined as an increase in their breathing medication or the administration of antibiotics due to their breathing being worse) within 2 months, or if they said they were feeling much worse than usual. If patients were unstable, we delayed our assessment until they were stable.
Resting oximetry was measured after the patient had been seated for 20 min while breathing room air. If an Allen’s test demonstrated adequate collateral flow and if the patient was not receiving an anticoagulant, the RT obtained an arterial blood gas sample. For patients whose PaO2 was 56 to 59 mm Hg and who did not have ankle edema, the RT obtained from the physician’s office the hematocrit prior to the administration of home oxygen.
If the saturation by pulse oximeter
(SpO2) was 90% on room air,
patients received supplemental oxygen in increments of 1 L/min at
20-min intervals until the lowest oxygen flow rate to keep the
SpO2 > 90% was established.
The patients then underwent a standardized exertional oximetry test in
which they walked on the spot for 15 s alternating with 15 s
of walking at their usual pace for a total of 5 min during which
SpO2 was monitored. The 5-min
exercise period was repeated while patients received increments of 1
L/min in 20-min intervals until the lowest oxygen flow rate such that
the SpO2 
90% for > 80% of the
duration of the test was established. During each repetition of the
test, we measured the mean SpO2
during each of the 5 min of the test, and we also documented whether
criteria (saturation 88% for 2 continuous min) for exercise
hypoxemia were met.
Patient Categorization
With respect to the indications for receiving oxygen, we
classified patients in the following categories:
1. PaO2 55 mm Hg breathing room
air at rest.
2. PaO2 of 56 to 59 mm Hg, with
either edema or increased hematocrit ( 55).
3. Positive result on Allen’s test (thus not possible to do arterial
blood gas measurements) and a resting arterial saturation of oxygen of
88%.
4. PaO2 of 56 to 60 mm Hg and cor pulmonale or pulmonary hypertension as specified by physician.
5. PaO2 of 56 to 60 mm Hg and
saturation 88% for 2 min on exercise, correctable with oxygen.
6. Oxygen prescribed for sleep hypoxemia.
7. No indication for oxygen.
The first three categories represent the strongest indication for domiciliary oxygen: the Nocturnal Oxygen Therapy Trial and Medical Research Council criteria for life-prolonging oxygen or the equivalent in patients for whom we could not obtain arterial blood gas measurements. In addition to patients meeting these strict trial criteria, the HOP accepts patients in categories 4, 5, and 6 for home oxygen.
We also made note of individuals who, although they did not meet criteria, were funded by the HOP on "compassionate grounds." The "compassionate use" criterion is a response to specific physician appeals for exceptions to the set standards. A consultant reviews these appeals and decides whether they should be accepted.
Two raters, the RT who had collected audit data and the research associate coordinating the study, classified patients on the basis of the record of data collected during the assessment. For patients for whom the raters disagreed, one of the investigators (GHG) made the final decision.
The same raters also judged the appropriateness of oxygen prescriptions at rest and during exercise. Prescriptions could range from 0 to 5 L/min. For patients in whom saturation fell to < 85% for > 1 min of the test period despite receiving 4 L/min of oxygen, the raters were instructed to recommend further assessment. Again, a third reviewer (GHG) adjudicated any disagreements.
Statistical Methods
For each variable, we used standard methods to calculate
proportions, means, and SDs. We calculated chance-corrected agreement
()4
between our two raters regarding whether there was
any indication for oxygen (categories 1 to 6 vs category 7). We also
calculated between the raters concerning rest and exercise
prescription using a weighted with quadratic weights.5
We also used weighted with quadratic weights to assess agreement
between the patient’s report of rest and exercise oxygen prescription
and the record of rest and exercise prescription in the files of the
HOP, as well as between both of these reports and the independently
measured rest and exercise oxygen prescription. To explore a number of
possible predictors of the appropriateness of oxygen use, we used
logistic regression for independent variables (patient age, gender,
city, urban or suburban, diagnosis, and whether the prescribing
physician was a specialist or primary-care physician) and dependent
variable (whether the prescription of oxygen was appropriate
[categories 1 to 6]).
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Table 1 presents a comparison of the 237 patients who participated and the 181 eligible patients who did not participate for variables for which we have data from both groups. In both populations, most patients had COPD. Of the variables in which we had data for both groups, none showed a statistically significant difference between participants and nonparticipants.
|
We did not find any patients who were receiving suboptimal medications. In addition to the eight patients who remained unstable throughout the study, two patients who were initially unstable were reassessed after they had become stable. Of the 237 patients, 69 patients (29.1%) had been receiving home oxygen for < 18 months, 89 patients (37.6%) for 19 to 24 months, and 79 patients (33.3%) for > 2 years.
Appropriateness of Oxygen Administration
Table 2
presents our final assessment of the appropriateness of oxygen. The
for agreement on whether patients met any of the criteria for home
oxygen was 0.79 (95% confidence interval [CI], 0.71 to 0.87). Those
who clearly met criteria for life-prolonging oxygen constituted 50.2%
of the patients. Weaker criteria for home oxygen applied to 9.3% of
the population. Of the 237 patients, 96 patients (40.5%; 95% CI, 34.3
to 46.8) met no criteria for home oxygen. Of these, 20 patients were
receiving oxygen on compassionate grounds. Thus, 76 of 237 patients
(32.1%; 95% CI, 26.1 to 38.0) did not meet criteria and had not been
funded on compassionate grounds.
|
70 years old (67.9%) and in 70 of 132 patients > 70
years old (53.0%; odds ratio, 1.85; 95% CI, 1.1 to 3.2; p = 0.02).
Specialist physicians, most of whom were chest physicians, showed a
similar rate of appropriate prescribing (71 of 112 patients; 62.5%) to
primary-care physicians (69 of 123 patients; 56.1%; odds ratio, 1.35;
95% CI, 0.80 to 2.29; p = 0.26). Among the 141 patients (59.5%) who met at least one criterion for appropriateness of home oxygen, 140 responded to a question concerning the degree of benefit. The distribution of responses for these patients was as follows: very beneficial, 82.1%; moderately beneficial, 15.7%; slightly beneficial, 2.1%; no benefit, 0%; and made condition worse, 0%. The distribution of responses to the same question among the 96 patients (40.5%) who did not meet any of the criteria for appropriateness for home oxygen was as follows: very beneficial, 87.5%; moderately beneficial, 8.3%; slightly beneficial, 3.1%; no benefit, 1.0%; and made condition worse, 0%.
Appropriateness of Oxygen Prescription
The weighted for our agreement on oxygen prescription at rest
was 0.90 (95% CI, 0.84 to 0.95) and during exercise was 0.94 (95% CI,
0.91 to 0.97). Table 3
presents the proportion of patients receiving each oxygen prescription
as defined by the three criteria: HOP documentation, patient report,
and independent audit assessment. These would include those with no
criteria for home oxygen, indications of oxygen only on exercise, and
those with oxygen recommended for nocturnal hypoxemia. In general, our
criteria would result in patients receiving less oxygen than they were
using. We would recommend 2 L/min of oxygen at rest for only 57
patients (24.1%) of the population. Results were very similar
for exercise oxygen prescription: our recommended prescription
was less than the patient was receiving, and less than that documented
in the HOP application.
|
, 0.52; 95% CI, 0.37 to 0.68) between
the participant’s report of how much oxygen they were receiving at
rest and the HOP documentation of what they had been prescribed. The
agreement between our recommended oxygen prescription at rest and
either patients’ report of oxygen use (, 0.17; 95% CI, 0.10 to
0.25) and the documentation provided to the HOP (, 0.08; 95% CI,
0.01 to 0.15) was extremely poor. Patient reports of oxygen
prescription for exercise showed moderate agreement (, 0.63; 95%
CI, 0.13 to 0.76) with the HOP; but again, agreement between our
recommended oxygen prescription for exercise and either patients’
report (, 0.26; 95% CI, 0.16 to 0.37) or the documentation provided
to the HOP (, 0.15; 95% CI, 0.05 to 0.2.4) was very poor (Table 4
).
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our major finding is that 40.5% (95% CI, 34.3 to 46.8) of patients receiving home oxygen met neither the criteria of the third-party funder for home oxygen nor any criteria that strongly mandate domiciliary oxygen use. Subtracting those funded by the third party on compassionate grounds, there remained 32.1% (95% CI, 26.1 to 38.0) who did not meet known criteria. Despite this, each patient had been accepted by the HOP based on information at the time of their application. What are the possible reasons for this discrepancy?
One of the most plausible explanations for initial eligibility of those who proved to be ineligible during the audit is that patients were tested while still experiencing the adverse effects of an exacerbation. When the exacerbation was fully resolved, such patients may no longer have required oxygen. For example, Levi-Valensi and colleagues6 found that 30% of patients meeting criteria for domiciliary oxygen after 1 month of apparent stability no longer met the same criteria after an additional 3 months of observation. Another possibility is that prolonged oxygen therapy may result in cardiopulmonary vascular changes that improve ventilation/perfusion matching and result in normoxia at rest.7 Evidence for this phenomenon is weak, but we cannot exclude it as a possibility. Other possibilities include conscious or unconscious bias during testing for eligibility, given that testing protocols are not standardized. For instance, patients tested with overly vigorous exercise or without an adequate rest period prior to obtaining measurements may yield inappropriately reduced oxygen levels. Equipment calibration and reproducibility might be another source of error.
Almost all patients who did not meet criteria experienced at least moderate subjective benefit from home oxygen. Their experience represents either a placebo effect or an oxygen-related decrease in exertional dyspnea in daily activities.8 Patients with COPD with a resting PaO2 > 60 mm Hg and exercise hypoxemia show inconsistent improvement in laboratory exercise capacity with supplemental oxygen.9 The only randomized trial addressing the effect of home oxygen on symptoms and health-related quality of life in patients with exercise hypoxemia who did not meet randomized trial inclusion criteria found no benefit.10 These results suggest that placebo effects may play an important role in the subjective benefit of home oxygen.
We found not only a high rate of inappropriate oxygen use, but an unstandardized approach to determining oxygen flow rates. Flow rates tended to be higher than our criteria would dictate, and bore a very weak relationship with our suggested flow rates. Thus, it is likely that many patients were receiving greater flow rates than required, and some lower flow rates than would be optimal.
Smaller studies in other jurisdictions have also shown a high incidence of PaO2 > 60 mm Hg in patients receiving domiciliary oxygen11 and a tendency for patients to use more oxygen than necessary.12 Ours is the first study to systematically examine the frequency of inappropriate oxygen use in a jurisdiction that had already implemented criteria based on the results of the British Medical Research Council trial and the Nocturnal Oxygen Therapy Trial.
Our results suggest that the HOP program could realize substantial savings (a 30% budget reduction would represent $16.5 million), by denying oxygen to those who are not eligible. A more appropriate choice of oxygen flow rates might achieve further cost reductions. On the one hand, we could view the reduction of "inappropriate oxygen" with satisfaction, noting not only cost savings but also an improved quality of life from the discontinuation of an unnecessary piece of equipment. On the other hand, we do not know whether denying patients home oxygen would result in a higher utilization of physician visits, emergency services, or hospital admissions. This important issue requires a trial of alternative organizational systems of oxygen delivery that would include health-services utilization as an outcome measure.
The Ontario criteria, and methods of establishing the need for oxygen therapy, resemble those of many other jurisdictions worldwide. The need for clinical stability is stated in almost all program criteria. Stipulation of an initial blood gas measurement is common, as are additional criteria to accommodate exercise and sleep hypoxemia. Rigorous follow-up as represented by this study is very uncommon. It is therefore likely that our findings are applicable to domiciliary oxygen programs in other countries. Our results highlight the considerable costs associated with LTOT, and the desirability of standardized methods for establishing oxygen prescriptions, particularly exercise prescriptions. Our findings suggest that objective evaluations of the necessity of oxygen administration some months after patients begin therapy are likely to identify a large proportion who do not meet current criteria. The cost of such additional visits must be included in the evaluation of this approach. We would estimate these costs to be below $75.00 per visit (United States) and therefore as being trivial relative to the potential benefits of appropriate selection. Education of referring physicians and patients as to the often-transient need for oxygen at home, particularly at a time of expedient discharge from hospital, would also be helpful. Specific criteria and standardized approaches to assess oxygen during exercise will also clarify the benefits of LTOT for individuals with respiratory conditions. However, changes in the organizational structure of assessment and delivery may well be necessary to achieve optimal health resource allocation in home oxygen programs throughout the world.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
= chance-corrected agreement;
LTOT = long-term oxygen therapy; RT = respiratory therapist;
SpO2 = saturation by pulse oximeter Received for publication December 1, 1999. Accepted for publication June 20, 2000.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y Lacasse, J LaForge, and F Maltais Got a match? Home oxygen therapy in current smokers Thorax, May 1, 2006; 61(5): 374 - 375. [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Lacasse and F. Maltais From the authors Eur. Respir. J., October 1, 2005; 26(4): 747 - 747. [Full Text] [PDF]
|
|
|
|
 |
|
 D Gorecka Liquid oxygen, is it the gold standard? Chronic Respiratory Disease, October 1, 2005; 2(4): 181 - 182. [PDF]
|
|
|
|
 |
|
 G. H. Guyatt, M. Nonoyama, C. Lacchetti, R. Goeree, D. McKim, D. Heels-Ansdell, and R. Goldstein A Randomized Trial of Strategies for Assessing Eligibility for Long-Term Domiciliary Oxygen Therapy Am. J. Respir. Crit. Care Med., September 1, 2005; 172(5): 573 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Chaney, K. Jones, K. Grathwohl, and K. N. Olivier Implementation of an Oxygen Therapy Clinic to Manage Users of Long-term Oxygen Therapy Chest, November 1, 2002; 122(5): 1661 - 1667. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gershon, C. K. N. Chan, and G. Guyatt Home Oxygen Requalification Chest, July 1, 2001; 120(1): 318 - 319. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
