The appropriate indications for HBO2T are controversial and evolving. Practitioners in this area are in an unusual position. Unlike most branches of medicine, hyperbaric physicians do not deal with a range of disorders within a defined organ system (e.g., cardiology), nor are they masters of a therapy specifically designed for a single category of disorders. Inevitably, the encroachment of hyperbaric physicians into other medical fields generates suspicion from specialist practitioners in those fields. At the same time this relatively benign therapy, the prescription and delivery of which requires no medical license in most jurisdictions (including the United States), attracts both charlatans and well-motivated proselytizers who tout the benefits of oxygen for a plethora of chronic incurable diseases. This battle on two fronts has meant that mainstream hyperbaric physicians have been particularly careful to claim effectiveness only for those conditions where there is a reasonable body of supporting evidence.
In 1977, the UHMS systematically examined claims for the use of HBO2T in more than 100 disorders and found sufficient evidence to support routine use in only 12. The Hyperbaric Oxygen Therapy Committee of that organization has continued to update this list periodically with an increasingly formalized system of appraisal for new indications and emerging evidence (Table S11-1). Around the world, other relevant medical organizations have generally taken a similar approach, although indications vary considerably—particularly those recommended by hyperbaric medical societies in Russia and China where HBO2T has gained much wider support than in the United States, Europe, and Australasia. Nevertheless, there are now 29 Cochrane reviews summarizing the randomized trial evidence for 25 putative indications, including attempts to examine the cost-effectiveness of HBO2T. Table S11-2 presents a synthesis of these two approaches and lists the estimated cost of attaining health outcomes with the use of HBO2T. Any savings associated with alternative treatment strategies avoided as a result of HBO2T are not accounted for in these estimates (e.g., the avoidance of lower leg amputation in diabetic foot ulcers). Following are short reviews of three important indications currently accepted by the UHMS.
TABLE S11-1Current List of Indications for Hyperbaric Oxygen Therapy |Favorite Table|Download (.pdf) TABLE S11-1 Current List of Indications for Hyperbaric Oxygen Therapy
Air or gas embolism (includes diving-related, iatrogenic, and accidental causes)
Carbon monoxide poisoning (including poisoning complicated by cyanide poisoning)
Clostridial myositis and myonecrosis (gas gangrene)
Crush injury, compartment syndrome, and acute traumatic ischemias
Central retinal artery occlusion
Enhancement of healing in selected problem wounds
Exceptional blood loss (where transfusion is refused or impossible)
Necrotizing soft tissue infections (e.g., Fournier’s gangrene)
Osteomyelitis (refractory to other therapy)
Delayed radiation injury (soft-tissue injury and bony necrosis)
Skin grafts and flaps (compromised)
Sudden sensorineural hearing loss
TABLE S11-2Selected Indications for Which There Is Promising Efficacy for the Application of Hyperbaric Oxygen Therapy |Favorite Table|Download (.pdf) TABLE S11-2 Selected Indications for Which There Is Promising Efficacy for the Application of Hyperbaric Oxygen Therapy
|Diagnosis ||Outcome (Number of Sessions) ||NNT 95% CI ||Estimated Cost to Produce One Extra Favorable Outcome 95% CI (USD) ||Comments and Recommendations |
|Radiation tissue injury ||More information is required on the subset of disease severity, the affected tissue type that is most likely to benefit, and the time over which benefit may persist. |
| ||Resolved proctitis (30) ||3 ||22,392 ||Large ongoing multicenter trial |
| || ||2–11 ||14,928–82,104 || |
| ||Healed mandible (30) ||4 ||29,184 ||Based on one poorly reported study |
| || ||2–8 ||14,592–58,368 || |
| ||Mucosal cover in ORN (30) ||3 ||29,888 ||Based on one poorly reported study |
| || ||2–4 ||14,592–29,184 || |
| ||Bony continuity in ORN (30) ||4 ||29,184 ||Based on one poorly reported study |
| || ||2–8 ||14,592–58,368 || |
| ||Prevention of ORN after dental extraction (30) ||4 ||29,184 ||Based on a single study |
| || ||2–13 ||14,592–94,848 || |
| ||Prevention of dehiscence (30) ||5 ||36,480 ||Based on one poorly reported study |
| || ||3–8 ||21,888–58,368 || |
|Chronic wounds ||More information is required on the subset of disease severity or classification most likely to benefit, the time over which benefit may persist, and the most appropriate oxygen dose. Economic analysis is required. |
| ||Diabetic ulcer healed at 1 year (30) ||2 ||14,928 ||Based on one small study, more research required |
| || ||1–5 ||7464–37,320 || |
| ||Diabetic ulcer, major amputation avoided (30) ||4 ||29,856 ||Three small studies; outcome over a longer time period required |
| || ||3–11 ||22,392–82,104 || |
|ISSNHL ||No evidence of benefit >2 weeks after onset. More research is required to define the role (if any) of HBO2T in routine therapy. |
| ||Improvement of 25% in hearing loss within 2 weeks of onset (15) ||5 ||18,240 ||Some improvement in hearing, but functional significance unknown |
| || ||3–20 ||10,944–72,960 || |
|Acute coronary syndrome ||More information is required on the subset of disease severity and timing of therapy most likely to result in benefit. Given the potential of HBO2T in modifying ischemia-reperfusion injury, attention should be given to the combination of HBO2T and thrombolysis in early management and in the prevention of restenosis after stent placement. |
| ||Episode of MACE (5) ||4 ||4864 ||Based on a single small study; more research required |
| || ||3–10 ||3648–12,160 || |
| || ||6 ||7296 || |
| ||Incidence of significant dysrhythmia (5) ||3–24 ||3648–29,184 ||Based on a single moderately powered study in the 1970s |
|Traumatic brain injury ||Limited evidence that for acute injury HBO2T reduces mortality but not functional morbidity. Routine use not yet justified. |
|Mortality (15) ||7 ||34,104 ||Based on four heterogeneous studies |
| || ||4–22 ||19,488–58,464 || |
|Enhancement of radiotherapy ||There is some evidence that HBO2T improves local tumor control, reduces mortality for cancers of the head and neck, and reduces the chance of local tumor recurrence in cancers of the head, neck, and uterine cervix. |
| ||Head and neck cancer: ||5 ||14,592 ||Based on trials performed in the 1970s and 1980s. There may be some confounding by radiation fractionation schedule. |
| ||5-year mortality (12) ||3–14 ||8755–40,858 || |
| ||Local recurrence 1 year (12) ||5 ||14,592 ||May no longer be relevant to therapy |
| || ||4–8 ||11,674–23,347 || |
| ||Cancer of uterine cervix: ||5 ||24,320 ||As above |
| ||Local recurrence at 2 years (20) ||4–8 ||19,456–38,912 || |
|Decompression illnessa ||Reasonable evidence for reduced number of HBO2T sessions but similar outcomes when NSAID added. |
|Reduction of HBO2T treatment requirement by 1 || |
|N/R ||Single appropriately powered randomized trial |
LATE RADIATION TISSUE INJURY
Radiotherapy is a well-established treatment for suitable malignancies. In the United States alone, ~300,000 individuals annually will become long-term survivors of cancer treated by irradiation. Serious radiation-related complications developing months or years after treatment (late radiation tissue injury [LRTI]) will significantly affect between 5 and 15% of those long-term survivors, although incidence varies widely with dose, age, and site. LRTI is most common in the head and neck, chest wall, breast, and pelvis.
Pathology and Clinical Course
With time, tissues undergo a progressive deterioration characterized by a reduction in the density of small blood vessels (reduced vascularity) and the replacement of normal tissue with dense fibrous tissue (fibrosis). An alternative model of pathogenesis suggests that rather than a primary hypoxia, the principle trigger is an overexpression of inflammatory cytokines that promote fibrosis, probably through oxidative stress and mitochondrial dysfunction, and a secondary tissue hypoxia. Ultimately, and often triggered by a further physical insult such as surgery or infection, there may be insufficient oxygen to sustain normal function, and the tissue becomes necrotic (radiation necrosis). LRTI may be life-threatening and significantly reduce quality of life. Historically, the management of these injuries has been unsatisfactory. Conservative treatment is usually restricted to symptom management, whereas definitive treatment traditionally entails surgery to remove the affected part and extensive repair. Surgical intervention in an irradiated field is often disfiguring and associated with an increased incidence of delayed healing, breakdown of a surgical wound, or infection. HBO2T may act by several mechanisms to improve this situation, including edema reduction, vasculogenesis, and enhancement of macrophage activity (Fig. S11-3). The intermittent application of HBO2 is the only intervention shown to increase the microvascular density in irradiated tissue.
The typical course of HBO2T consists of 30 once-daily compressions to 202.6–243.1 kPa (2–2.4 ATA) for 1.5–2 h each session, often bracketed around surgical intervention if required. Although HBO2T has been used for LRTI since at least 1975, most clinical studies have been limited to small case series or individual case reports. In a review, Feldmeier and Hampson located 71 such reports involving a total of 1193 patients across eight different tissues. There were clinically significant improvements in the majority of patients, and only 7 of 71 reports indicated a generally poor response to HBO2T. A Cochrane systematic review with meta-analysis included six randomized trials published since 1985 and drew the following conclusions (see Table S11-2 for numbers needed to treat): HBO2T improves healing in radiation proctitis (relative risk [RR] of healing with HBO2T 2.7; 95% confidence interval [CI] 1.2–6) and after hemimandibulectomy and reconstruction of the mandible (RR 1.4; 95% CI 1.1–1.8); HBO2T improves the probability of achieving mucosal coverage (RR 1.4; 95% CI 1.2–1.6) and the restoration of bony continuity with osteoradionecrosis (ORN) (RR 1.4; 95% CI 1.1–1.8); HBO2T prevents the development of ORN following tooth extraction from a radiation field (RR 1.4; 95% CI 1.08–1.7) and reduces the risk of wound dehiscence following grafts and flaps in the head and neck (RR 4.2; 95% CI 1.1–16.8). Conversely, there was no evidence of benefit in established radiation brachial plexus lesions or brain injury.
A problem wound is any cutaneous ulceration that requires a prolonged time to heal, does not heal, or recurs. In general, wounds referred to hyperbaric facilities are those where sustained attempts to heal by other means have failed. Problem wounds are common and constitute a significant health problem. It has been estimated that 1% of the population of industrialized countries will experience a leg ulcer at some time. The global cost of chronic wound care may be as high as U.S. $25 billion per year.
Pathology and Clinical Course
By definition, chronic wounds are indolent or progressive and resistant to the wide array of treatments applied. Although there are many contributing factors, most commonly these wounds arise in association with one or more comorbidities such as diabetes, peripheral venous or arterial disease, or prolonged pressure (decubitus ulcers). First-line treatments are aimed at correction of the underlying pathology (e.g., vascular reconstruction, compression bandaging, or normalization of blood glucose level), and HBO2T is an adjunctive therapy to good general wound care practice to maximize the chance of healing.
For most indolent wounds, hypoxia is a major contributor to failure to heal. Many guidelines to patient selection for HBO2T include the interpretation of transcutaneous oxygen tensions around the wound while breathing air and oxygen at pressure (Fig. S11-4). Wound healing is a complex and incompletely understood process. While it appears that in acute wounds healing is stimulated by the initial hypoxia, low pH, and high lactate concentrations found in freshly injured tissue, some elements of tissue repair are extremely oxygen dependent, for example, collagen elaboration and deposition by fibroblasts, and bacterial killing by macrophages. In this complicated interaction between wound hypoxia and peri-wound oxygenation, successful healing relies on adequate tissue oxygenation in the area surrounding the fresh wound. Certainly, wounds that lie in hypoxic tissue beds are those that most often display poor or absent healing. Some causes of tissue hypoxia will be reversible with HBO2T, whereas some will not (e.g., in the presence of severe large vessel disease). When tissue hypoxia can be overcome by a high driving pressure of oxygen in the arterial blood, this can be demonstrated by measuring the tissue partial pressure of oxygen using an implantable oxygen electrode or, more commonly, a modified transcutaneous Clarke electrode.
Determining suitability for hyperbaric oxygen therapy (HBO2T) guided by transcutaneous oximetry around the wound bed. *In diabetic patients, <50 mmHg may be more appropriate. PtcO2, transcutaneous oxygen pressure.
The intermittent presentation of oxygen to those hypoxic tissues facilitates a resumption of healing. These short exposures to high oxygen tensions have long-lasting effects (at least 24 h) on a wide range of healing processes (Fig. S11-3). The result is a gradual improvement in oxygen tension around the wound that reaches a plateau in experimental studies at about 20 treatments over 4 weeks. Improvements in oxygenation are associated with an eight- to ninefold increase in vascular density over both normobaric oxygen and air-breathing controls.
The typical course of HBO2T consists of 20–30 once-daily compressions to 2–2.4 ATA for 1.5–2 h each session, but is highly dependent on the clinical response. There are many case series in the literature supporting the use of HBO2T for a wide range of problem wounds. Both retrospective and prospective cohort studies suggest that 6 months after a course of therapy, about 70% of indolent ulcers will be substantially improved or healed. Often these ulcers have been present for many months or years, suggesting that the application of HBO2T has a profound effect, either primarily or as a facilitator of other strategies. A recent Cochrane review included nine randomized controlled trials (RCTs) and concluded that the chance of ulcer healing improved about fivefold with HBO2T (RR 5.20; 95% CI 1.25–21.66; p = .02). Although there was a trend to benefit with HBO2T, there was no statistically significant difference in the rate of major amputations (RR 0.36; 95% CI 0.11–1.18).
CARBON MONOXIDE POISONING
Carbon monoxide (CO) is a colorless, odorless gas formed during incomplete hydrocarbon combustion. Although CO is an essential endogenous neurotransmitter linked to NO metabolism and activity, it is also a leading cause of poisoning death, and in the United States alone results in >50,000 emergency department visits per year and about 2000 deaths. Although there are large variations from country to country, about half of nonlethal exposures are due to self-harm. Accidental poisoning is commonly associated with defective or improperly installed heaters, house fires, and industrial exposures. The motor vehicle is by far the most common source of intentional poisoning.
Pathology and Clinical Course
The pathophysiology of CO exposure is incompletely understood. CO binds to hemoglobin with an affinity more than 200 times that of oxygen, directly reducing the oxygen-carrying capacity of blood, and further promoting tissue hypoxia by shifting the oxyhemoglobin dissociation curve to the left. CO is also an anesthetic agent that inhibits evoked responses and narcotizes experimental animals in a dose-dependent manner. The associated loss of airway patency together with reduced oxygen carriage in blood may cause death from acute arterial hypoxia in severe poisoning. CO may also cause harm by other mechanisms including direct disruption of cellular oxidative processes, binding to myoglobin and hepatic cytochromes, and peroxidation of brain lipids.
The brain and heart are the most sensitive target organs due to their high blood flow, poor tolerance of hypoxia, and high oxygen requirements. Minor exposure may be asymptomatic or present with vague constitutional symptoms such as headache, lethargy, and nausea, whereas higher doses may present with poor concentration and cognition, short-term memory loss, confusion, seizures, and loss of consciousness. While carboxyhemoglobin (COHb) levels on admission do not necessarily reflect the severity or the prognosis of CO poisoning, cardiorespiratory arrest carries a very poor prognosis. Over the longer term, surviving patients commonly have neuropsychological sequelae. Motor disturbances, peripheral neuropathy, hearing loss, vestibular abnormalities, dementia, and psychosis have all been reported. Risk factors for poor outcome are age >35 years, exposure for >24 h, acidosis, and loss of consciousness.
The typical course of HBO2T consists of two to three compressions to 2–2.8 ATA for 1.5–2 h each session. It is common for the first two compressions to be delivered within 24 h of the exposure. CO poisoning is one of the longest-standing indications for HBO2T—based largely on the obvious connection between exposure, tissue hypoxia, and the ability of HBO2T rapidly to overcome this hypoxia. CO is eliminated rapidly via the lungs on application of HBO2T, with a half-life of about 21 min at 2.0 ATA versus 5.5 h breathing air and 71 min breathing oxygen at sea level. In practice, however, it seems unlikely that HBO2T can be delivered in time to prevent either acute hypoxic death or irreversible global cerebral hypoxic injury. If HBO2T is beneficial in CO poisoning, it must reduce the likelihood of persisting and/or delayed neurocognitive deficit through a mechanism other than the simple reversal of arterial hypoxia due to high levels of COHb. The difficulty in accurately assessing neurocognitive deficit has been one of the primary sources of controversy surrounding the clinical evidence in this area. To date there have been six RCTs of HBO2T for CO poisoning, although only four have been reported in full. While a Cochrane review suggested there is insufficient evidence to confirm a beneficial effect of HBO2T on the chance of persisting neurocognitive deficit following poisoning (34% of patients treated with oxygen at 1 atmosphere vs 29% of those treated with HBO2T; odds ratio [OR] 0.78; 95% CI 0.54–1.1), this may have more to do with poor reporting and inadequate follow-up than with evidence that HBO2T is not effective. The interpretation of the literature has much to do with how one defines neurocognitive deficit. In the most methodologically rigorous of these studies (Weaver et al.), a professionally administered battery of validated neuropsychological tests and a definition based on the deviation of individual subtest scores from the age-adjusted normal values was used; if the patient complained of memory, attention, or concentration difficulties, the required decrement was decreased. Using this approach, 6 weeks after poisoning, 46% of patients treated with normobaric oxygen alone had cognitive sequelae compared to 25% of those who received HBO2T (p = .007; number needed to treat [NNT] = 5; 95% CI 3–16). At 12 months, the difference remained significant (32% vs 18%; p = .04; NNT = 7; 95% CI 4–124) despite considerable loss to follow-up.
On this basis, HBO2T remains widely advocated for the routine treatment of patients with moderate to severe poisoning—in particular in those aged >35 years, presenting with a metabolic acidosis on arterial blood-gas analysis, exposed for lengthy periods, or with a history of unconsciousness. Conversely, many toxicologists remain unconvinced about the place of HBO2T in this situation and call for further well-designed studies.
CURRENT CONTROVERSIES IN HYPERBARIC MEDICINE
The use of hyperbaric oxygen has been associated with controversy since it was first instituted in the 1950s. A vigorous debate has developed around the concept of performing sham controlled RCTs, particularly when studying possible indications where a placebo effect could significantly influence outcomes. The most popular method employed to achieve blinding of both staff and patients is the exposure of patients in the control arm to a modest pressure while breathing air in the chamber (between 1.1 and 1.3 ATA). While this strategy is effective in blinding the exposure, critics claim this exposure to air at pressure (equivalent to breathing around 27% oxygen at 1.0 ATA) is therapeutic in a way yet to be identified. These critics use this putative therapeutic effect to explain the modest measured benefits in patients with a range of chronic neurological conditions including cerebral palsy, autism spectrum disorders, and mild traumatic brain injury when exposed to either air at 1.1 to 1.3 ATA or 100% oxygen at 2.0 to 2.4 ATA (HBO2T) in a number of trials. These benefits have traditionally been interpreted as the result of a participation or placebo effect, with the various authors concluding there was no evidence of a specific effect for HBO2T in any of these conditions. The search continues for a convincing sham exposure that is universally regarded as inactive. Some workers claim this is not possible and that patient-blinded trials are therefore similarly unachievable. This impasse needs resolution.