Oxygen Therapy in COPD

Abstract

Long-term oxygen therapy (LTOT) is a mainstay treatment for patients with severe resting hypoxemia secondary to chronic respiratory conditions including COPD. The evidence for LTOT is based on two trials that are now several decades old but have been insufficiently revisited. Therefore, many questions remain about precisely which patients experience the most benefit from LTOT, as well as how to define that benefit. Most studies have examined LTOT’s effect on longevity rather than its impact on quality of life. In addition, many challenges exist in training both clinicians and patients on best practices for LTOT and associated equipment. Reimbursement policies have reduced the kinds of equipment available to the LTOT patient community, presenting additional challenges. This paper will review the current evidence for LTOT in COPD, the challenges involved with providing optimal therapy, and potential avenues of modernizing this essential intervention

Introduction

Oxygen is fundamental to life. Long-term oxygen therapy (LTOT) is nearly as fundamental to the treatment of advanced COPD. In total, over 1.5 million people in the United States have been prescribed LTOT for a chronic respiratory condition, with the majority of cases related to COPD.¹ This represents somewhere between 5–10% of the diagnosed COPD population. Despite the relatively small proportion of oxygen therapy users, LTOT is a significant driver of health care costs. Medicare spends well in excess of $2 billion per year on LTOT equipment, accounting for approximately one-quarter of the program’s spending on durable medical equipment (DME) as a whole.²

Despite the fundamental and impactful nature of this therapy, there are significant knowledge gaps regarding its application. The fundamental evidence supporting the concept of LTOT is now over 40 years old yet still underpins our therapeutic guidelines and research. As noted in the last LTOT comprehensive review by Branson, more recent studies have been complicated by difficulties in recruitment and potentially the shift from objective outcome measures to patient-reported ones.³ The role of LTOT outside of severe chronic hypoxemia continues to elude definitive understanding. Education on the proper use and assessment of LTOT is severely lacking for clinicians and patients alike. Technological improvement has been largely stagnant for decades, save some bursts of innovation as we fight to emerge from the COVID-19 pandemic.

It is against this backdrop that this paper will examine the current state of LTOT, including clinical trials including both severe and moderate hypoxemia, the physiologic and psychologic impact of LTOT on users, administrative and prescriptive concerns, and the use of oxygen in special circumstances (eg, during COPD exacerbations). In addition, opportunities for future research to better elucidate the most appropriate use of LTOT will be discussed.

LTOT in COPD With Severe Resting Hypoxemia

The use of LTOT in people with COPD living with resting hypoxemia (defined as P_aO2 ≤ 55 mm Hg while taken at rest on room air or P_aO2 ≤ 59 mm Hg with evidence of heart failure, pulmonary hypertension, or erythrocythemia) is the most well-defined and evidence-based application of oxygen therapy. The Nocturnal Oxygen Treatment Trial (NOTT)⁴ and Medical Research Council (MRC)⁵ Trial together form the evidentiary foundation for LTOT in this population, as well as the basis for oxygen therapy qualification for most insurance providers in the United States (including Medicare).

NOTT was the first of the trials and at 203 subjects was the larger as well. This multi-center trial based in the United States recruited subjects with a spirometrically confirmed diagnosis of COPD and known to have resting hypoxemia as previously defined.⁴ In order to ensure these subjects were hypoxic in their stable state, study inclusion criteria also required 2 oxygenation evaluations at least one week apart during an observation period lasting 3 weeks, during which they also had to be free of exacerbations. Subjects also had to be oxygen therapy naïve at inclusion. Once included, subjects were randomized to receive either continuous oxygen therapy or 12 h of oxygen therapy at night. The subjects were then followed for one year, with mortality as the primary end point. Additional outcome measurements, including pulmonary function data and quality-of-life evaluations, were also obtained. Results demonstrated a clear mortality benefit for continuous oxygen therapy (11.9% compared with 20.6% mortality in the nocturnal oxygen–only group). This benefit was also seen in the 37 continuous therapy subjects and 29 nocturnal therapy subjects who survived to 24 months (resulting in a mortality rate of 22.4% and 40.8%, respectively). Continuous therapy subjects were also found to have improvements in hematocrit measurements and pulmonary vascular resistance, but no other outcomes were found to be statistically significant.

While clearly a milestone trial, there are several important factors to consider while contemplating the role of NOTT in the present day. The study was plagued by recruitment issues, including potential subjects improving during the evaluation period to the point where they were no longer hypoxemic enough to enroll.⁶ This had a detrimental impact on enrollment rate, an issue that continues to plague LTOT-related clinical trials to this day. In fact, the study did not meet initial recruitment goals, which the study group also admitted were couched in assumption and supposition regarding the number of subjects needed to appropriately power it. In addition, per the study design, each subject was seen in their domicile by a nurse practitioner weekly and seen in a follow-up clinic monthly for the first 6 months of the trial. After that period, they received a domiciliary visit monthly, with clinic visits occurring every 3 months at minimum. The group was clear to point out that some improvements may have been attributable not to LTOT but the above-average number of touchpoints each subject received by health care providers. This model, obviously, does not persist in today’s health care landscape, which may have an impact on current outcomes.

Often described as a companion study, the MRC trial in fact had several important differences from NOTT and more closely resembled the nocturnal therapy arm of NOTT rather than the study as a whole. Eighty-seven subjects in the United Kingdom were enrolled and randomized to receive either no supplemental oxygen or oxygen at 2 L/min via nasal cannula for at least 15 h per day, then followed for a total of 5 y, with the primary study end point again being mortality. All subjects had chronic air flow obstruction as determined by spirometry, and all were concurrently diagnosed with congestive heart failure. A significant number of subjects continued to smoke during the trial, whereas smoking status was not included in the NOTT data. Despite the differences in design, the MRC trial results still complement those from NOTT. The oxygen therapy group mortality rate was 45%, whereas mortality in the control group was 67%. This was held to be a significant benefit, especially considering the overall greater disease severity in the MRC cohort. The 2 studies combined suggest that the dose of oxygen (at least the duration of exposure) plays a significant role in clinical benefit.

Additional research into the best application of LTOT continued throughout the 1980s. Weitzenblum and colleagues⁷ found that application of continuous oxygen for at least 15 h per day appeared to halt the progression of pulmonary hypertension in severely hypoxemic people with COPD. A series of right heart catheterizations was performed on a cohort of 16 subjects, beginning approximately 4 y (mean 47 months ± 28 months) before the initiation of LTOT. The catheterization was repeated immediately before initiation and once more approximately 2 and a half y (mean 31 months ± 19 months) of therapy. Twelve of the 16 subjects saw improvements in mean pulmonary arterial pressure, although it did not generally return to baseline. There were no detected changes in oxygenation status, cardiac output, or pulmonary capillary wedge pressure, meaning the stabilization of pulmonary hypertension was likely attributable to reduced vascular resistance. A subsequent study with a larger cohort (95 subjects) over a longer time (6 y) revealed similar findings.⁸ LTOT for at least 14 h per day slightly reduced mean pulmonary arterial pressure over the course of the first 2 y, then stabilized it over the course of the next 4. However, this study was again plagued by many subjects lost to follow-up, largely due to high mortality associated with chronic hypoxemia. Only 39 subjects were recatheterized at the 2-y mark, 31 at 4 y, and 12 at end of the full 6-y study. The group found that despite these hemodynamic improvements, other COPD-related parameters (such as degree of air flow obstruction) continued to progress.

Taken together, these studies form the basis for clinical practice guidelines (and often payer reimbursement policies) around the world. However, that basis is now over 40 years old. Continuous LTOT for people with severe hypoxemia clearly has a longevity benefit; indeed, it is one of the precious few COPD therapies we have that can make that claim. However, it remains an open question as to whether concurrent advances in other COPD therapies (including pulmonary rehabilitation, noninvasive ventilation, and pharmacologic interventions) over the past 4 decades may enhance this survival benefit or if certain subtypes of COPD may respond better to LTOT than others. That may be a difficult signal to detect given the relatively blunt instrument of today’s guidelines. Unfortunately, there appears to be little appetite for this type of research at present.

LTOT in COPD With Moderate Resting Hypoxemia

Given the clear benefit of LTOT for severe hypoxemia as defined by the NOTT and MRC studies, there has been great interest over the years to determine whether the therapy could also be as effective for less severe impairment. Unlike in the setting of severe hypoxemia, this question has been studied numerous times over the past few decades. However, the answer has almost invariably been, “no.”

Two of the earliest studies of this population were published in the late 1990s. Górecka and associates⁹ evaluated subjects with P_aO2 levels between 56–65 mm Hg who were prescribed continuous LTOT, similar to previous research. The group found no difference in survival between the control and experimental groups. Notably, adherence to therapy was somewhat problematic, with the mean duration of oxygen therapy < 14 h per day (compared to the prescribed 17 h or more daily). The group could detect no differences in survival between those who adhered to therapy for more than 15 h per day (consistent with previous studies) and those who did not adhere, although sample sizes were again quite small (68 subjects in the experimental group). Chaoaut et al¹⁰ took a slightly different approach, intentionally using only nocturnal oxygen and also examining pulmonary arterial pressure as an additional outcome. However, the conclusions were essentially the same, with no significant improvement to survival, time to continuous LTOT prescription, or hemodynamic indicators. The team did explain their mortality findings due to the ever-present low sample size (76 at randomization, with only 46 undergoing hemodynamic follow-up at the end of the 2-y project). That uncertainty, coupled with the relatively favorable risk-benefit profile of LTOT, prompted many clinicians to prescribe LTOT outside of practice guidelines in an attempt to provide subjective dyspnea relief. Indeed, one study found that provision of palliative supplemental oxygen provided some perceived benefit in dyspnea level, albeit no different than that provided by medical air at the same flow.¹¹

The possibility that LTOT might yet provide some benefit in certain people led to the Long-Term Oxygen Treatment Trial or LOTT. LOTT was envisioned as the largest study yet done to evaluate the use of oxygen therapy in general. Initially, the study was designed to solely examine the survival benefit in people with moderate hypoxemia at rest as measured by pulse oximetry (specifically, S_pO2 between 89–93%). However, after the first 7 months of enrollment, only 34 subjects had been randomized. At that point, the study was deemed not feasible as implemented and redesigned to allow subjects who experienced moderate exercise-associated desaturations (defined as S_pO2 < 90% for at least 10 s but remaining > 80% for at least 5 min during a 6-min walk test [6MWT]). In addition, time to first hospitalization was combined with death to create a composite end point. Those subjects randomized to the treatment group with moderate resting hypoxemia were prescribed continuous supplemental oxygen, whereas those with exercise-induced hypoxemia were prescribed oxygen during activity and sleep.

After the redesign, a total of 738 subjects were enrolled in the study. Notably, this was over twice the number of subjects as the NOTT and MRC trials combined. These subjects were followed for up to 6 y, with oxygen flow prescription evaluated annually. However, no significant difference was noted between the control group and either intervention group in mortality or hospitalization. Secondary outcomes, including quality-of-life measures, mental health status, and activity capacity (measured by the 6MWT), were also not significantly different between the groups. The investigators attempted to find any benefit whatsoever in various subgroups but were unable to find any significant difference in any outcome, even given that overall adherence in the intervention group was consistent with rates consistent with positive outcomes in previous studies like NOTT.

The study is not without its detractors, perhaps unsurprising given the need to redesign it after initiation. Other notable concerns reported by the investigators themselves included selection bias against people who were already using oxygen therapy and perceived benefit (therefore chose not to enroll) and non-standardized oxygen equipment across the study sites. In addition, concern has been raised about the lack of blinding to low-saturation alarms on oximeters (which may have caused test subjects to, for example, adjust their walking pace during the 6MWT).¹² However, a subsequent meta-analysis that included 6 studies published between 1992–2020 (including NOTT, MRC, and LOTT itself) appears to confirm the lack of benefit at least in terms of 3-y mortality rate.¹³

LTOT in COPD for Exertional and/or Nocturnal Hypoxemia

As with moderate hypoxemia, the role of supplemental oxygen for intermittent desaturations is controversial. LOTT (and the subsequent meta-analysis by Lacasse et al) demonstrated that this therapy is not generally indicated. In addition, the National Emphysema Treatment Trial, another large-scale study including LTOT, saw no significant difference in survival in subjects with exercise-associated desaturation (defined as any desaturation S_pO2 < 90% by pulse oximetry, a looser definition than used in LOTT). Several smaller studies (as well as a Cochrane meta-analysis) have suggested that supplemental oxygen may improve exercise capacity specifically, which may theoretically lead to improvements over time in quality-of-life measures, if not longevity.^14-16 However, a more recent study casts some doubt on whether it is oxygen or merely gas flow that provides the benefit. Alison and associates¹⁷ studied 111 subjects with S_pO2 ≤ 89% during a 6MWT by placing them either on oxygen or medical air at 5 L/min via nasal cannula. The investigators found that after an 8-week course of increasing exercise intensity both study groups had improvements in exercise capacity as measured by endurance shuttle walk test, with no significant difference between the groups. Six months after the end of the exercise program (and the flow supplementation) both groups still demonstrated improvement in exercise capacity, although this had dropped below the level of significance. Both groups also showed improvements in quality of life as measured by the Chronic Respiratory Disease Questionnaire. Notably, the intervention was blinded to the exercise trainers, the investigators, and the participants. In addition, the equipment was externally identical, removing virtually all potential for bias.

Similarly, the role of LTOT in people with COPD who desaturate during sleep remains elusive. Most of the research that has been conducted thus far has been predicated on the notion that treating hypoxemia is desirable and individuals who expereince nocturnal desaturation may evade conventional detection. Thus, it has focused on estimating prevalence and designing potential screening programs.^18-20 However, as with exercise-induced hypoxemia, this assumption comes under suspicion the more it is researched. The International Nocturnal Oxygen (INOX) Trial attempted to determine the actual benefit of intervention through a randomized double-blind trial.²¹ In this study, subjects were identified through use of 2 home oximetry studies. Those who desaturated to < 90% for at least 30% of the recording time were randomized to receive either supplemental oxygen via concentrator or room air via a sham device. All devices were externally identical; the sham concentrators simply had their sieve beds bypassed by the manufacturer. The treatment group had their oxygen flow titrated to maintain a saturation of ≥ 90%, whereas the placebo group underwent a sham titration to maintain blinding. Despite the quality design of this study, the recruitment and retention issues that frequently plague LTOT research forced early termination, as only 243 of the needed 600 subjects were recruited after 4 y.²¹ The INOX group concluded that nocturnal oxygen had no impact on either survival or advancement to continuous supplemental oxygen. Due to the recruitment issues, the authors did offer the caveat that the study was severely underpowered. In addition, this study excluded subjects who experienced exercise desaturation.

LTOT outside of severe resting hypoxemia remains frequently prescribed, despite the lack of evidence to support it either continuously for moderate resting desaturation or intermittently for specific causes. Some studies suggest that as many as 40% of subjects prescribed oxygen technically do not qualify.²² The reasons for this are varied but are generally similar to the aforementioned view that considering the low amount of risk involved in oxygen therapy, it is worth a shot to determine benefit. In addition, many patients do perceive some benefit, even in physiological markers that oxygen generally does not impact (such as breathlessness).²³ In some cases, there may actually be objective benefit in particular patients, but those signals become lost in the noise of the overall study cohorts. Fortunately, there appears to be renewed interest in determining the best way to individualize oxygen therapy for maximum benefit. The LOTT trial publication was accompanied by an editorial by Swedish pulmonologist Magnus Ekström that suggested a pragmatic approach, with a blinded exercise test with oxygen or medical air to determine efficacy in an individual with exercise-induced desaturation, an approach also favored by Hatipoğlu and Stoller.^24,25 A systematic review to more carefully look at the role of oxygen therapy to prevent exercise-induced desaturation has been proposed by Kawachi and associates.²⁶ A similar review looking more broadly at survival in people with COPD who are prescribed LTOT outside the usual practice guidelines and recommendations is being undertaken by Alexandre et al.²⁷ It should also be noted that many exercise-related studies have relied upon the 6MWT as the preferred tool for inducing desaturation, despite the test not being designed for this purpose. A recent study by Gupta, Ruppel, and Espiritu highlighted this potential shortcoming and suggested that diffusion capacity of the lung for carbon monoxide may be a better predictor of such desaturation. Lacasse et al also suggest that pulse oximetry may not be a sufficient tool to capture all hypoxemic subjects. Their cross-sectional study correlating S_pO2 with arterial blood gas measurements found 40% of these subjects could have been denied LTOT based on current practice guidelines using pulse oximetry despite being severely hypoxemic based on arterial oxygen saturation (S_aO2). Given the growing body of evidence suggesting pulse oximetry may not be accurate in various non-white skin tones (thereby producing additional falsely negative results),^28-31 hypoxemia assessment protocols may need to be completely rethought.

These studies should allow clinicians a better understanding of when oxygen therapy is actually potentially helpful rather than simply not harmful.

Oxygen as Rescue Therapy

Many patients are first introduced to LTOT as part of a COPD exacerbation, and short-term oxygen therapy is an important, well-established component of exacerbation treatment plans, especially for exacerbations requiring hospitalization.³² Supplemental oxygen is also frequently provided in cases of non-pulmonary events potentially causing respiratory compromise, such as angina or cerebrovascular accident. However, regardless of the process in play, appropriate dosing of oxygen has been frequently debated since its introduction.

Over 60 years ago, one of the most well-known papers in respiratory care history was published by Campbell³³ describing a potential relationship between hyperoxia and hypercapnic respiratory failure in subjects with COPD receiving supplemental oxygen. This work was based on some earlier preliminary studies with very small cohorts in which certain subjects appeared to suffer reduced respiratory effort when administered supplemental oxygen at relatively high levels. This phenomenon became known as the hypoxic drive theory and despite limited confirmatory research became the standard of care in COPD for the next several decades. Whereas there is some evidence that excessive oxygen administration does have an impact on blood gas chemistry, it is important to remember that these alterations are due to other processes including ventilation/perfusion mismatching and the Haldane effect, which is a rightward shift of the CO₂ dissociation curve secondary to deoxygenated hemoglobin’s higher affinity for CO₂ molecules.³⁴ In addition, clinicians must remember that not all people with COPD retain carbon dioxide and certainly not all to the same degree. Clinicians should not be afraid to administer oxygen at whatever dose is necessary to relieve hypoxemia, particularly in emergency situations. In fact, Sandau and associates³⁵ recently found that any hypoxemic event (S_pO2 < 88%) within the first 24 h of admission was associated with significantly poorer outcomes over the course of 14 d, whereas hyperoxic events (S_pO2 > 92%) were not, even at 30 d. In addition, patients with concurrent acute issues, such as myocardial infarction brought on by an exacerbation, may benefit from higher doses of oxygen due to those complicating factors.³⁶ It is important to recognize that the role of supplemental oxygen for the treatment of cardiac events in the setting of normoxia is increasingly controversial.³⁷

Other studies have demonstrated different results, however. Echevarria et al demonstrated that subjects with S_pO2 > 92% on admission had higher in-patient mortality even in the setting of normocapnia. This risk appeared to be dose dependent as well, with the group of subjects with S_pO2 > 97% having a similar mortality risk than those with S_pO2 < 88% upon arrival to the hospital. These findings are consistent with other previous work from New Zealand, which found hyperoxemia (defined in this case as S_pO2 > 96%) significantly increased the risk of hypercapnic respiratory failure, noninvasive ventilation, and/or mechanical ventilation compared to normoxia and even hypoxia.³⁸ Austin et al³⁹ had similar findings in a randomized controlled study that compared supplemental oxygen titrated to an S_pO2 between 88–92% and oxygen administered by non–rebreather mask in subjects with a history of COPD transported to hospitals by paramedic services.

Some thought has been given to the use of heated high-flow nasal cannula (HFNC) given the ability to precisely set F_IO2 as well as the flow enhancements that seem to be of at least subjective benefit to many with COPD. Small studies have shown to improve both oxygenation status and hypercarbia, likely to the enhanced washout of physiological dead space and some mild end-expiratory positive pressure.⁴⁰ To be clear, these studies have been very heterogeneous to date and continue to have small sample sizes. A multi-center pilot study by Plotnikow and associates⁴¹ determined that the application of HFNC to people with COPD upon admission to an ICU was successful in reducing breathing frequency by 27%. However, no outcomes data were collected, and the investigators noted an 18% treatment failure rate.

Whereas it is somewhat unclear why some studies demonstrate harm by hyperoxia and others do not, it seems prudent to err on the side of safety and establish appropriate therapeutic recommendations for the use of short-term oxygen therapy for exacerbations of COPD. Such a pragmatic approach can also mitigate resource issues such as the occasional oxygen supply shortages seen internationally during the COVID-19 pandemic by preventing the unnecessary waste of oxygen. Several organizations call for maintaining saturation levels between 88–92%, the generally agreed-upon range between hypoxia and hyperoxia in these studies, including the Global Initiative for Chronic Obstructive Lung Disease³² and the British Thoracic Society.⁴² The American Thoracic Society(ATS)/European Respiratory Society clinical practice guideline on COPD exacerbations recognizes the role of oxygen therapy in these cases but are silent on titration range.⁴³ This seems to be an area for quality improvement in order to standardize care and maximize patient safety. However, many clinicians may desire to see more definitive data elucidating the populations greatest at risk (or a better explanation of outlier studies) prior to changing their established practices. This is likely also true in other situations featuring advanced dyspnea (such as end of life), where evidence supporting the use of oxygen in non-hypoxic states remains elusive.

Ambulatory Oxygen Equipment

Understanding when to apply LTOT is only part of the equation; clinicians must also understand how to apply the therapy. There is a wide range of equipment available for use outside the hospital, each of which has a variety of pros and cons (Table 1). Hardavella et al⁴⁴ published a concise summary of the general features of each type of equipment in 2019, much of which remains useful today. However, it is important to review overall issues and considerations for recommending and/or prescribing specific equipment.

Table 1.

Oxygen Delivery System Comparison

Oxygen Therapy Prescription and Education

In the United States, the Centers for Medicare and Medicaid Services (CMS) is generally considered the main driver of oxygen access policy, as most patients prescribed LTOT are on Medicare. The National Coverage Determination document provides guidance to clinicians for coverage and prescribing (Table 2).⁴⁵ The document was recently updated in an effort to reduce administrative overhead by eliminating the need for certificates of medical necessity (CMNs). CMNs were additional forms that contain information such as diagnosis, prognosis, and duration of need and were previously required for complex DME including oxygen devices. However, much of the information contained in a CMN was also included either in the oxygen prescription itself or the patient’s medical records. Thus, CMS opted to discontinue this requirement in favor of chart audits.⁴⁶ Whereas many clinicians celebrated the elimination of redundancy, others (as well as a variety of professional and patient advocacy organizations, including the American Association for Respiratory Care and the COPD Foundation) have raised concerns about the reliability and consistency of the chart audit process given the wide variety of electronic medical records and other documentation systems.⁴⁷

Table 2.

Medicare Requirements for Ambulatory Oxygen Therapy Qualification

Because paperwork problems can cause delayed access to needed equipment, it is, therefore, essential the prescriber thoroughly document the patient’s needs in their chart notes and complete the oxygen prescription accurately. In order for oxygen equipment to be covered, the prescriber must evaluate the patient via a face-to-face (or telehealth) encounter within 30 days of the written order. The order must also contain a description of the item being ordered (in this case, the specific type of oxygen delivery system needed), the prescribed flow, the relevant diagnosis code, the frequency and duration of use (eg, 15 h a day or continuously; “as needed” or “prn” are not acceptable), and the estimated duration of need (eg, 3 months or lifetime) (Table 3). If any of these items are missing, delivery may be delayed or the claim eventually rejected.

Table 3.

Medicare Requirements for Oxygen Written Order

It should be noted that whereas the oxygen prescription covers many aspects of equipment and overall therapy application, it lacks coverage for education services. In addition, selection of specific devices is generally left to the DME provider, which means the clinician may not be able to assess how well the equipment actually meets the needs of the patient during their activities of daily living. Unlike other disease states (such as diabetes⁴⁸ or advanced renal disease⁴⁹), Medicare does not provide any reimbursement for disease management education sessions, nor does it pay for any type of respiratory therapist service. This places the burden of teaching safe and effective use of oxygen equipment on the DME provider. Prior to 2011, these services were often provided by a staff respiratory therapist at the expense of the DME company. However, in 2011, CMS implemented the Competitive Bidding Program (CBP) in an effort to reduce overall program costs. The CBP essentially pits DME companies against each other to see who could provide equipment at the lowest possible cost.⁵⁰ By the time the program was fully implemented in 2016, reimbursement rates for DME were approximately halved on average.⁵¹ This had a massive impact on DME suppliers around the country, inducing them to reduce staff (including patient educators and other respiratory care practitioners). Many patients report no longer receiving appropriate training on their equipment. A survey from the ATS Nursing Assembly Oxygen Working Group found that 35% of respondents felt somewhat or wholly unprepared to safely or effectively operate their new equipment.⁵² Most respondents reported receiving what training they had from delivery personnel, with only 8% stating a clinician provided education and 10% reporting getting no training at all.

Competitive bidding has had significant impacts on other aspects of DME supply as well. Along with educators, delivery staff and infrastructure have historically been major cost centers for DME suppliers. This is particularly true with modalities that are labor intensive to deliver (such as compressed gas tanks) or require special vehicles or driver’s license endorsements (such as liquid oxygen [LOX]). These costs vary widely in the United States based on region but have increased substantially across the country during the COVID-19 pandemic.⁵³ The implementation of certain technologies, such as stationary concentrators that can also refill compressed gas cylinders (known as transfill systems), has been used to cover service gaps; but as we will discuss in the next section, other modalities have become increasingly difficult to access.

Oxygen Delivery Equipment

Liquid oxygen.

LOX has been one of the major casualties of competitive bidding. These systems provide significant advantages over other portable oxygen systems (as well as some over stationary ones). LOX is stored at much higher pressures than gaseous oxygen, allowing far greater quantities to be transported easily. One liter of LOX is approximately the equivalent of 860 L of oxygen gas. That capacity allows for either very high flows (in excess of the 2–3 L/min portable concentrators are capable of) or extended time between refills, allowing for greater mobility.⁵⁴ Mobility is of great concern to many on LTOT, with the ATS Oxygen Working Group survey also finding that almost 80% of respondents felt their portable gaseous system or portable concentrator limited their activity outside their home. Unfortunately, LOX systems can be expensive to service and maintain, owing to the extremely low temperatures and high pressures needed to contain it safely. That extra cost has led many DME suppliers to drop LOX supply altogether in the face of competitive bidding. A recent review of Medicare data found that between 2013–2019 the number of beneficiaries using LOX systems decreased 89%, with the equipment no longer available in large sections of the United States.⁵⁵

Compressed gas.

Compressed oxygen cylinders were one of the first oxygen sources and remain a common method to provide portable and backup oxygen supplies. They require no electricity to operate, making them a good choice for emergencies, particularly in areas prone to disruptions to the power grid. Unfortunately, cylinders in sizes practical for travel outside the home also tend to be rather heavy, a major concern in the demographics commonly associated with COPD. Al-Mutairi et al⁵⁶ surveyed 311 people with COPD on LTOT and found portable oxygen cylinders were a major source of concern, as many respondents were concerned the cumbersome nature of the tanks would cause them to become home bound. In addition, as a cost-saving measure, many DME suppliers have limited the number of cylinder deliveries available to patients, further limiting the time they can spend away from their stationary system.

Stationary oxygen concentrators.

Oxygen concentrators are currently the most common device for LTOT in the home. These devices use a molecular sieve to filter out nitrogen molecules from ambient air, then deliver the purified oxygen to the patient. They are able to provide continuous flow at up to 10 L/min, and the gas they deliver is generally between 90–96% pure. Whereas there have been some advances in materials science that have reduced the weight and noise of these units, they still often weigh as many as 45 pounds. Most have caster wheels to facilitate movement from room to room, but they are otherwise not generally considered portable. Positioning of the concentrator is, therefore, critical, as they tend to generate heat and noise levels that may be disruptive to sleep. However, they can use up to 100 feet of oxygen extension tubing to ameliorate this issue.⁵⁷

Portable oxygen concentrators.

The underlying technology of oxygen concentrators has not changed significantly since their invention in the late 1970s, but advances in electronics and other sciences have allowed the miniaturization of key components to the point where some degree of portability is possible. These devices also rely upon a molecular sieve to filter oxygen, but their portable nature limits the size of the sieve bed. Thus, their maximum continuous flows are limited to 2–3 L/min, with the higher end only achievable by wheeled, pull-along devices. To overcome this limitation, as well as improve battery life, portable concentrators often deliver a brief bolus of oxygen near the beginning of inhalation, then pause to allow additional concentration. This is known as pulse-dose delivery. (Notably, some gaseous cylinder regulators are also capable of this to extend the life of the cylinder’s contents.)

There is a great deal of confusion among both clinicians and patients regarding the use of pulse delivery. During the development of portable concentrators, gaseous regulators were used as the predicate device to facilitate the Food and Drug Administration review process. Therefore, settings for various bolus sizes were assigned numerical values under the assumption they would roughly correlate with continuous liter flow settings. However, the actual amount of oxygen inhaled during each breath depends on a variety of factors, including breathing frequency and tidal volume.⁵⁸ In addition, bolus size can vary between manufacturers, meaning a setting of “2” on one device will likely not carry over exactly should the patient change equipment.⁵⁹

Adding to potential complications is the notion that due to the intermittent nature of the oxygen flow patients may need to adjust their breathing pattern to adapt to the device. These devices are also not able to compensate sufficiently for changes in minute ventilation or oxygen demand, such as during activity. Using nasal airway simulators, Chen et al⁶⁰ found that at breathing frequencies and tidal volumes consistent with exercise (800 mL per breath at 22 breaths/min) volume averaged F_IO2 for pulse dose devices could be up to one-third less than continuous flow. Importantly, this study also suggested that the devices tested also did not consistently provide sufficient oxygen during sleep. This team also found that continuous flow oxygen routinely delivered a greater absolute volume of oxygen per breath compared with pulse dose.⁶¹ Clearly, more differentiation between continuous flow settings and the inconsistent, dimensionless settings on pulse dose devices is needed.

There are additional policy issues surrounding the use of portable concentrators. CMS policies differentiate between stationary and portable devices but do not specify what type of portable device should be provided. In the current environment, most DME suppliers provide a stationary concentrator and an allotment of cylinders. Those people that prefer portable concentrators are, therefore, forced to either give up their continuous flow device and become reliant upon a single device that may not meet all their needs or purchase their portable concentrator out of pocket. These devices represent a significant expense, generally $2,000–$4,000, a steep price to pay for enhanced mobility. These costs have led to many LTOT users to seek lower-cost solutions from online retailers. Over the past 2 years, devices have begun to appear on popular online marketplaces advertised as portable oxygen generators and concentrators despite not having gone through any regulatory evaluation. However, they are substantially less expensive than true concentrators, which makes them quite appealing to many. Casaburi and associates⁶² studied oxygen delivery characteristics for 3 such devices using a metabolic simulator and discovered some models put out oxygen at subclinical purity. In addition, the team found no customer support after the sale for these devices in case of malfunction or other issues. Both of these problems present significant safety issues to the LTOT community.

The Road Ahead

Oxygen therapy is a key component of COPD management, particularly for those individuals with severe resting hypoxemia. However, significant barriers exist for patients attempting to access the most appropriate devices for their lifestyle. In response to this, an interprofessional coalition has assembled to push for significant reforms to CMS oxygen policy. Known as the Four Pillars of Oxygen Reform, these policy principles would ensure that future users of LTOT are able to use their equipment safely and effectively, enabling them to maximize their quality of life. First, supplemental oxygen therapy must be patient centric. This includes concepts like using the phrase “ambulatory oxygen” in policy documents rather than “home oxygen,” to reflect the modern reality that people on oxygen are not simply sitting in their domicile awaiting their demise, as well as a patients’ bill of rights. This will ensure care is focused on the needs of the user rather than the equipment. It may reflect the notion that oxygen devices are not simply equipment like a wheelchair or even a nebulizer but more akin to a prosthetic device that needs to be fitted to the individual and adjusted to their support their lifestyle.

The second pillar is restoring access to LOX and ensuring that it remains available when it is medically necessary. As discussed, LOX is a critical modality for many in terms of quality of life; and in other disease states, there are simply no other portable options. Third, the time has long since come for respiratory therapy services to be reimbursed. It is unfair to prescribers to expect them or their staffs to be as intimately familiar with oxygen therapy and devices as a respiratory therapist, and it is unfair to DME suppliers to force them to provide an unfunded yet essential service. Access to respiratory therapists can ensure a high level of patient confidence and self-efficacy,⁶³ and a recent simulation study suggested long-term cost savings with regular visits by a respiratory specialist.⁶⁴ Finally, the organizations call for standardized national documentation requirements that utilize a template rather than medical records or redundant paperwork. This final pillar is intended to make reimbursement more predictable and reduce fraud while keeping prescriber workloads reasonable.

In addition to policy improvements, educational efforts must continue to improve prescribing patterns. Without clear evidence, LTOT for situations like moderate hypoxemia and intermittent desaturation states appears to be overprescribed. This places a burden not only on the health care system but on the individuals receiving this unnecessary care with potentially cumbersome equipment. Certain prescribing practices, like hyperoxygenation in the perihospital setting, may even increase the risk of adverse outcomes. It would likely also be helpful to improve public health education surrounding oxygen therapy to potentially reduce stigma and improve adherence, and self-management education should absolutely be encouraged (and reimbursed by payers).

Finally, long-held assumptions like heavy reliance upon pulse oximetry for diagnosis and assessment must be questioned. Given the increased attention being paid to the limitations of this technology, clinicians may wish to reconsider its application for initiation of LTOT, as well as assessing its efficacy. This paradigm shift may also have implications for research, and future investigators may well find better ways to determine who benefits from LTOT and who does not. That may finally answer some of the concerns that have remained unaddressed despite being raised by the last two Respiratory Care Journal conferences focusing on oxygen therapy.

Discussion

MacIntyre: You touched on it briefly, but I think it’s one of the focuses of this meeting, and I’d like to have a few more thoughts on it – and that is who should be on supplemental O₂ and how do we find these people? My sense is that when people do pulse oximetry in the clinic it’s usually when they are sitting in a chair as part of taking their blood pressure and pulse and whatnot. And I think with COPD and other diseases, exercise hypoxemia is woefully underdiagnosed. I’ll just share our experience at Duke. We wrote an article¹ a couple of years ago in Chest about a standardized procedure to assess exercise and O₂ needs. What was striking to me was that of 222 sequential patients we studied, 101 of the 165 subjects (61%) on no baseline O₂ had desaturation during 6 min of walking. I’m struck by that. You have a wonderful presentation here on access, but I’m concerned that we may be underdiagnosing what I call episodic hypoxemia. I’m talking not only about exercise but sleep as well. And I’m concerned that we’re not paying enough attention to that as perhaps we should. It’s only going to compound the problems you brought up

Mike Hess: Without question we are underdiagnosing it. That’s part of the issue with a standardized assessment and prescription form. And speaking of terminology, we often conflate it. We call it the 6MWT, and it’s very strictly defined as a measure of distance, and you can desaturate and rest and all that sort of thing. But in order to make somebody desaturate, it’s not just about walking them around. I have patients in the clinic, and I could walk them around the clinic floor all afternoon and their saturation won’t do a thing, but take them up those 5 stairs to the registration desk they’ll pass out. So that’s the rub, how do we do that? Do we do it with a remote patient monitor? Can we use a Fitbit-type device like we might do with a Holter monitor? Can we send them home and tell them to go about their normal daily activities and we’re going to record for 48 h and then come back to the clinic or mail this back, and we’ll look at it and see you had this level of exertion here; what were you doing? And that way we can actually more finely tune our titration, and we can capture more of those people who do need that exercise or exertion where we might not be able to see it in the clinic environment. That is absolutely true.

Haynes: One of the tools you can use to find patients who are hypoxemic during exercise is D_LCO. There are data suggesting that D_LCO < 50% predicts exercise desaturation, although this seems to be more powerful in pulmonary fibrosis but also in emphysema. We had a protocol in my previous lab that any patient with a D_LCO < 50% predicted automatically did an informal walking desaturation test and, if they could do it, climbed a flight of stairs. Using this protocol, we captured a lot of people who had a similar D_LCO for years and were never walked, so their exercise desaturation had not been discovered. The other thing is that a lot of people view a 6MWT and an O₂ titration test as the same thing, and they’re clearly not. When you do a 6MWT, as you mentioned, you’re looking at how many meters the patient can walk, so the protocol is to push them, and as I showed in my presentation, you can achieve V̇_O2max during the 6MWT.² An O₂ titration test should be at the patient’s usual activity level. We always make that clear to our patients, and often after you’ve completed a 6MWT, you can ask the patient, “would you ever walk this hard or would you be symptom limited?” And they often say, “I would never push myself that far.” So you run the risk of finding hypoxemia at a level of exercise that the patient would never do on their own. The other point is that usually O₂ titration studies are done on continuous flow O₂, and as you mentioned, then they’re put on that flow with a pulse dose; and then when you walk these people, you may not hear the O₂ pulse clicking on because the patient has reverted to mouth breathing as their ventilation increases.

Mike Hess: Those are great points. If you have the capacity, the patient is willing to do a D_LCO test, that’s a great way to find those patients. Unfortunately, when we get out into rural areas or other clinics that don’t have access to a pulmonologist or a PFT lab, then we lose that opportunity for screening. The rub with O₂ saturation testing is yes it’s true that they might not walk that fast, but as they go about their daily life they’re probably engaging those other muscle groups and probably using the same level. I don’t have the data to support this, but as they’re doing laundry, carrying groceries, as they’re going about their actual activities of daily living (ADL) that are not represented in the clinic, they’re probably having a similar O₂ demand.

MacIntyre: Just to clarify, the 6MWT was developed because it was reflective of typical ADL. It’s not a 12-min walk, it’s a 6-min walk, and you’re allowed to walk at your own pace. I’m not as nihilistic. I think the 6MWT is a reasonable reflection of ADLs, but I do agree that the O₂ titration may require a longer walk time. In our protocol, you have to walk at least 6 min and be stable for at least 3 min on the final dose. One other comment on this. Jerry and I and others were part of the Long-Term Oxygen Treatment Trial (LOTT), and while these episodic hypoxemic subjects had better saturations when they used O₂, the long-term outcome (ie, mortality) was not reflected. But I think that’s reflecting ultimate outcomes. I think you will definitely get functional improvement in exercise tolerance if you exercise without desaturating.

Mike Hess: Right. It may not be adding years to your life, but what about adding life to your years? Is that of some benefit? I would argue yes, but others may say no.

*Dean Hess: Should every patient with COPD on home O₂ have a pulse oximeter and be taught how to titrate their O₂ according to their saturation?

Mike Hess: They should have a good pulse oximeter, but yes.

Orr: I think the problem there is a good one is around $150, but you can get one from CVS for $20 that at least on the surface appears to do the same thing. Insurance almost never covers an oximeter; we have patients on home NIV, and they won’t pay for a pulse oximeter. Is there any move to try and change that and try to provide some coverage for an FDA-cleared pulse oximeter?

MacIntyre: Before you answer that, I’d like you to finish Dean’s question. Should patients be monitoring themselves at home and perhaps adjusting O₂ according to their saturation?

Mike Hess: Yes, absolutely. They should have a good pulse oximeter, and they should be measuring; again, this is an issue with some of our ordering protocols. Historically we have a liter flow with activity, at rest, and during sleep.

Orr: I thought it was rhetorical because it’s like putting someone on antihypertensives and then never taking their blood pressure again.

*Dean Hess: I think the issue is the quality of the pulse oximeter that is used, but even the best quality pulse oximeter has some error associated with it.

Mike Hess: Yes, and you have to know how to interpret it too. You have to be stable for a minute; it’s not an instantaneous reading. It comes back to access and education again. You’ve been in the trials and read the data, and you know O₂ works in the right circumstances, but how do you get to those circumstances? But yes, everybody should do that. I have a good friend and O₂ user whose favorite phrase is, “titrate and migrate” because with activity he adjusts up and then he adjusts down.

*Dean Hess: All that said, I don’t know if that has been studied and been shown that there are better outcomes associated with that.

Carlin: That’s the problem. When you talk about desaturation with exercise, without a proven morbidity or more importantly a mortality benefit, who’s going to pay for this? No one. Unfortunately, there aren’t really any studies out there. It certainly does make clinical sense to “titrate to saturate.”

*Dean Hess: It seems right, but I don’t know that we have the data to support it.

Carlin: I don’t know that we’ll ever have the data. This conversation is interesting because I remember being at a meeting over 10 years ago in Atlanta where this same subject was a topic of interest, and here we are over a decade later still discussing it.

Mike Hess: I’ll come back to that; let me finish Jeremy’s point. It’s on the list, but again, everything is prioritized. Right now the O₂ task force, such as it is, is focused on access to liquid O₂ and the external stuff that CMS isn’t touching. But we’re hopeful to get there because, yes, everyone should have access to a nice, good quality, FDA-cleared pulse oximeter, and they should be taught how to use it. But we’re not there yet.

Orr: One follow-up on that. I think there are a lot more of these drugstore pulse oximeters out there now in the post-COVID era. I’m a little worried about what that might do with false reassurance and people titrating or acting upon information that can be fairly questionable.

Mike Hess: You and me both. To follow up with Brian, it’s interesting you should say that because when I was first looking at doing this talk I pulled up O₂ therapy from the last Journal conference, and I was going to do “everything that’s old is new again.” We haven’t really made any progress, so maybe I’ll talk about things we can make some progress on. But again, going back to the picture of the 1970s-era O₂ concentrator, it’s the road to nowhere. We’ve gone nowhere with this.

Criner: The lack of innovation over 40 years, is it solely related to lack of reimbursement? Medicare spends over $4 billion a year on O₂ therapy; is it the wrong players in industry or in your view why is there such stagnation in innovation and what could be done to prompt the right players?

Mike Hess: I think it’s multifactorial, and I do think a big part of it is reimbursement. In fact, I’ve been told by some people in industry that there’s no money in it so why would we do it? There’s no reimbursement, and it’s not getting better, so it’s not cost effective. I think another big component of it is that COPD users are the largest population. They don’t necessarily have the highest O₂ demand, but they’re the largest user population, and there is traditional stigma with COPD therapy that they brought it on themselves and they should have known better, etc, and so there has been this sort of de-prioritization of COPD in particular. But I would argue that—and this is my personal opinion—for some reason as a health care system and as a society we don’t take lung conditions very seriously. We take diabetes seriously, heart problems seriously, neurological issues seriously, but lungs? Eh, you’ll be fine, you just can’t run. It’s viewed as not a very big deal, and that is something where we really need a groundswell change to affect all the factors here, not only in COPD but other lung conditions as well. So the two big things are, yes, reimbursement, but we also need in my view more of a societal shift where we place that emphasis on lung conditions because that will drive demand and that will drive innovation.

Garvey: You did a great job with your talk. Don’t you find it alarming that the FDA has never been held accountable for approving portable O₂ concentrators that don’t use a standardized method for O₂ flow? I’ve had pulmonologists scream at me when I tell them that the dial doesn’t mean L/min. I’ve looked at 10 different manuals for the most popular portable concentrators, and 6 of them say nothing about O₂ flow. All the FDA cared about was electrical safety. How is it possible that they could do that, and the patients are out there thinking this device is the answer to controlling their oxygen levels, yet many max out at less than 1 L/min?

Mike Hess: I’m still learning about the ins and outs of device approval. I will admit that while O₂ has been a focus of mine since I’ve been at the Foundation, prior to that I was an inhaler technique guy. It’s been a learning experience, but it’s my understanding that the device approval process is a lot more lenient than say pharma approval or implanted devices and things like that. So yeah, it’s problematic that they haven’t been held accountable; and yes, it’s problematic that the industry hasn’t settled on a standard and there are probably technical limitations and so forth, but they haven’t landed on a standard bolus. It’s unfortunate that they’ve chosen to use numbers rather than letters, shapes, or colors. I don’t know that it’s ever been explicitly advertised to end users as a direct analogue to continuous liters of flow, but it’s an obvious assumption. The two questions I get asked the most are, “Why am I short of breath when my pulse oximeter is high?” and “Why isn’t 2 on my portable the same as 2 on my stationary?”

Carlin: Can you comment on the recent revision of the Certificate of Medical Necessity (CMN)?

Mike Hess: Yes, there was a recent revision to the reimbursement standard for Medicare that added cluster headaches as an approved indication for O₂ therapy, and it removed the requirement for CMN. It was a little bit controversial because as of now we’re relying on chart audits to confirm the need for O₂. So there’s the thought that they could really game that system to reduce the amount of O₂ claims even further, even though there are some other standardized O₂ orders out there they have not been approved for use. So it’s something that we pushed back on; ACCP, ATS, we had a whole group letter that said, “please, we know the CMN is terrible but we need something to replace it and not just rely on chart audits.” That’s where CMS opted to go.

*Dean Hess: To follow up on Chris’ comment, I think we have work to do with the durable medical equipment (DME) companies as well. A physician orders a patient to go home on 2 L/min of O₂; the DME company gets the prescription; they call the patient’s primary care physician and ask if they can substitute for a device on a number 2. They may not even talk to the physician. They talk to a nurse or a medical assistant in the office; one device set on 2 is substituted for another device set on 2, and all of a sudden, the patient is at home on suboptimal therapy. I would suggest many times it’s not intentional; it’s just a lack of understanding.

Mike Hess: That’s where these hot spots come into play here with documenting and ordering, the transition, education, the introduction to the equipment, and things like that. Again, this is when you’re diagnosed in the hospital, but it’s very similar when you’re diagnosed in the ambulatory care setting. And you have all these breakdowns. Remember earlier one of the biggest physician-reported barriers is unable to coordinate with DME. That’s exactly right, nobody knows this stuff. And we can go round and round about the why; do we need respiratory therapists (RTs) in primary care; do we need better education; do we need x, y, or z; and those do exist, but there is room for improvement.

*Dean Hess: For we RTs who work in the hospital, this is an opportunity. We can work with our physicians who send patients home on O₂, and we, the hospital-based RT, can test the patient in the hospital on the device and make sure that the patient goes home with the best device suited to the patient.

Mike Hess: In the best-case scenario, absolutely.

#Goodfellow: Thank you for the presentation. I want to ask a question related to clinical practice guidelines (CPGs). In your work with the COPD Foundation, where are some of the venues that you see clinically that RTs (and Dean just mentioned one) could be involved with getting the right device and the right flow? And in your opinion do you think a CPG might be helpful if it’s evidence-based?

Mike Hess: It can be a little bit tough because, as we’ve discussed, we sometimes don’t have the evidence to base anything on, so it goes into that realm of consensus statement or expert opinion. I do think we could use some revisions. I know we have an O₂ CPG for both acute care and for ambulatory care, although I think it’s home and alternative settings or something like that. It’s getting a little long in the tooth, so we might want to look at that again. Personally I’d love to see, and I don’t know how we’d formulate this into a guideline, but more things along the lines of care coordination because I think that’s what it really comes down to. You may not necessarily have a floor RT doing that training, but maybe you have an RT educator. At one of our hospitals in Kalamazoo, we have a pulmonary care coordinator, an RT who does the work of a standard nurse care coordinator, care transitions, that sort of thing. That’s their job to work with DME, to work with primary care offices, to work with the patient, to work with the hospitalist, to work with all of these groups and pull everything together. They are the pit crew chief that we saw, but not every place has that or knows the framework for how to put that together. So, and again I don’t know if it’s necessarily a guideline; but if we had a framework or model for that, I think that would go a long way not only for O₂ but for NIV and getting people to pulmonary rehab. With closing a lot of these gaps, we have that crew chief. And on the ambulatory side of things, I could speak all day about tobacco treatment and disease management. I really hope we get that disease management credential from NBRC because that’s what our patients really need.

†Branson: So Mike, being the person who wrote the last Journal conference paper³ on home O₂ therapy for COPD, you’re correct in that nothing much has changed. I will say that serendipitously yesterday the FDA put out a statement warning about pulse oximeter use in the home and warning patients on a patients page, and I can share the link with everybody, mentioning some of these issues (https://www.fda.gov/medical-devices/safety-communications/pulse-oximeter-accuracy-and-limitations-fda-safety-communication?utm_medium=email&utm_source=govdelivery. Accessed June 21, 2023). I like the idea of monitoring patients’ S_pO2 at home, but I’m also struck by the recent study in the New England Journal of Medicine where patients with COVID-19 at home who monitored their O₂ saturation at home didn’t have any better outcome or any positive effects of the oximetry than those who just came in when their dyspnea got worse.⁴ So maybe it’s not helpful. On the other hand, and I don’t know who would pay for it, but closed-loop control where you wear an O₂ sensor that’s Bluetooth connected to your concentrator that increases or decreases the pulse volume or flow may be useful and could do so many other things like report the flow, tell the office that the O₂ requirement increased, and much more. There’s so much that could be done, but it’s all limited by reimbursement. And then the last thing, and I’m not trying to promote this at all, but the group from the hypoxia lab at UCSF have a page called Open Oximetry (https://openoximetry.org/oximeters/. Accessed June 21, 2023), and it’s open access, and they evaluate all these over-the-counter oximeters, and they’ll let you know the precision of the device for any one that’s been purchased. I could talk about this all day. Dean and I discussed an online video regarding home use of oximetry where a nurse comes on screen and she’s demonstrating how to use the oximeter. Unfortunately, she has these fake nails that were painted all these different colors, and we were watching it, and you know, the devil is in the details. I’m sorry, I’m not sure there’s a question there.

Mike Hess: I do have a response though, especially to the closed-loop control thing. Those devices are under development. The barriers are reimbursement; it’s going to be costly. The other barrier is going to be, and again this is what I’ve been told by one of the companies that’s developing it, FDA is really sensitive about anything where you remove the human from the control loop it has to be bullet proof. That’s one of the things that when Ruth Tal-Singer, the CEO of the COPD Foundation, told me, “I want you to lead this project” it’s the first thing I said we should look at because it should be an easy thing we could develop a smart concentrator. We have all these other smart things, refrigerators, microwaves; we have all these things; why don’t we have a smart concentrator where we could develop a better pulse oximeter that would fit a little more conveniently in a lifestyle and control that concentrator, and then we remove the need for settings and education about 2 not being 2 on the concentrator. But as you said, the devil is in the details, and the FDA are very picky about approving that kind of software.

*Dean Hess: I’ll just say we’re talking about this as though it would be better, but I think that is a hypothesis. And we’d need to study it and determine if it is indeed better.

Mike Hess: Virtually every solution proposed here is a hypothesis because again other than we have O₂, we know it works, and we can monitor it, we still have so many unanswered questions. We have some that are answered, but we have quality-of-life things, mobility, those sorts of things remain unanswered. And as Brian mentioned, they may never be answered because who’s going to pay for the study?

Carlin: Certainly pulmonary rehabilitation programs, even given the limited number available, are ideal places to continue the provision of this education for all involved. Several years ago there was the ideal of providing oxygen clinics (similar to anticoagulation clinics), but this never materialized. I think all of this can be very effectively provided within a pulmonary rehabilitation program.

Mike Hess: Another possible solution would be some of the pharmacy chains are starting to move into more advanced things; there’s one chain that does CPAP clinics now. Those are right in the community; the people often know the pharmacists or whoever is in charge. We could provide education through those venues; we have a lot of, for lack of a better term, community health professionals who are underutilized. One program in North Carolina has been sending out specially trained paramedics to be an RT’s eyes and ears in the field. The RT care coordinator there calls them the “care-a-medics,” and they can do pulse oximetry and video chat to do that kind of evaluation. It truly takes a village to care for an O₂ patient, a COPD patient, however you want to look at it. But we have to start utilizing our resources better.

†Branson: What do you think about direct-to-consumer selling of concentrators?

Mike Hess: I think it is unfortunate.

†Branson: I agree. I believe that the disappearance of liquid oxygen is primarily due to cost reasons, but another reason was because it was overprescribed. If I’m the patient, I want liquid; even if I’m on 1 L/min, I want liquid because it extends my time away from home to be involved in activities that make life worthwhile. I think that bringing back liquid would take specific requirements; especially with patients with interstitial pulmonary fibrosis who need 4–5 L/min, liquid is really their only answer to any kind of quality of life.

Mike Hess: That’s where the focus is. I have my own quibbles about what we should prioritize because I think even if you bring liquid back, the infrastructure will take a long time; there are a lot of non-policy barriers to overcome. But that is where the focus is right now for things like interstitial pulmonary fibrosis and the high flow user. I think the cutoff they’re looking at right now is 4 L or more.

†Branson: I’m involved in another group looking at home O₂ and have spoken to some of the bigger liquid O₂ providers who still provide it in hospitals, and they say we could get it back up and running in 3 months. It would be no problem as far as technology. Distribution would be a different story. My feeling is that all concentrators should have the volume rather than the numerical setting of a dimensionless number. Does anybody else think that would be helpful?

Carlin: I agree. It’s so confusing right now, and we’ve not done a great job of educating the care providers about how these delivery systems work and note that one system is truly different from another.

†Branson: I’ve worked in an academic center for 43 years, so I am careful to speak about business. But it almost seems like there’s a business opportunity. You could create RT O₂ and have a van and drive around to primary physician’s offices and help them fill out the paperwork for home O₂, help them educate the patient, aerosol therapy, all the other home self-management techniques that can improve quality of life. I think the number one thing we have to do is remind everyone of something we already know, which is that O₂ is a drug.