Pearls in Hyperbaric Oxygen Therapy: A Challenging Case Series

Most of the 15 indications for the use of HBOT can be lifesaving and limb sparing.¹ Providers who assist in treating patients who would benefit from HBOT need to be knowledgeable about the absolute and relative contraindications to its use. Untreated pneumothorax is clearly an absolute contraindication to HBOT. Relative contraindications include administration of doxorubicin during HBOT, use of bleomycin within the previous 6 months, use of disulfiram, the presence of implanted devices, the presence of epidural pain pumps, pregnancy, high fever or epilepsy, inability of an individual to clear their ears, history of thoracic surgery, eustachian tube dysfunction, claustrophobia, a history of eye surgery, and upper respiratory or sinus infection.¹ Urgent and emergent circumstances in which HBOT is administered include arterial or venous gas embolus, decompression illness, acute arterial insufficiency (including central retinal arterial occlusion), gas gangrene, certain necrotizing soft tissue infections, crush injury and compartment syndrome, intracranial abscess, compromised tissue flaps and skin grafts, severe anemia, thermal burns, idiopathic sudden sensorineural hearing loss, and carbon monoxide poisoning.²

The 2 primary mechanisms of action of HBOT are pressure and hyperoxygenation.³ The presence of a hyperbaric ambient pressure is in itself therapeutic in some instances, such as, gas embolus and decompression illness. In both cases, gas bubbles in blood and soft tissues contribute to the condition and the additional ambient pressure reduces the size of the bubbles, which ameliorates the condition.

Hyperoxygenation of tissues is the other primary mechanism of action in HBOT and is in itself therapeutic in several of the indications for HBOT. In severe anemia and carbon monoxide poisoning, for example, blood plasma is supersaturated with oxygen, which allows adequate oxygenation of tissues throughout the body even though the usual hemoglobin oxygen transportation system is unavailable.

However, HBOT also has several secondary effects owing to synergy between the high pressure and high oxygen content. It has an antimicrobial effect, which explains the use of HBOT in refractory osteomyelitis and necrotizing soft tissue infections. Additionally, HBOT blunts ischemia reperfusion injury, which is helpful in treating decompression illness, crush injuries, and carbon monoxide poisoning; it blunts vasoconstriction, which is directly related to edema control in several indications; and it improves wound healing via the aforementioned mechanisms as well as independent mechanisms that improve both the growth of new blood vessels from local endothelial cells, angiogenesis, and the recruitment of stem cells to form new blood vessels (ie, vasculogenesis).

Patients may have relative contraindications to HBOT, and it is the responsibility of the health care team to know what additional information is required to determine whether HBOT can be safely administered. Information on relative contraindications may be sparse, however. The cases presented herein include patients with relative contraindications who would otherwise benefit from HBOT.

Several scenarios are presented in this case series, including the use of ketamine for sedation in a patient undergoing HBOT. Other scenarios include the use of HBOT in the setting of cerebral vasoconstrictive syndrome, in a patient with an antimicrobial tracheal-esophageal device who had recently undergone thoracic surgery, in a pregnant woman with carbon monoxide poisoning, and in patients with various types of implantable devices. The purpose of this article is to educate providers on uncommon or unique situations in which HBOT may be indicated and to note the additional steps taken to manage each patient safely and successfully.

Case 1

The diagnosis and treatment of radiation-induced brain injury can be challenging. A 70-year-old female experienced neurological symptoms following noninvasive stereotactic radiosurgery and demonstrated improvement after undergoing HBOT.

The patient was diagnosed with a tentorial notch meningioma that caused noncommunicating hydrocephalus with poor balance, memory difficulty, and urinary incontinence. She underwent a ventriculostomy that resulted in temporary improvement of her symptoms, and 1 year later she underwent noninvasive stereotactic radiosurgery for definitive treatment. Her symptoms continued to improve, but then the patient began to deteriorate quickly. Her balance, memory, and concentration were particularly poor. Subsequent imaging showed decreasing ventricular size and stable meningioma.

The patient was diagnosed with soft tissue radionecrosis of the brain and referred to the hyperbaric medicine department. Her symptoms progressively worsened up until the start of HBOT. It was not possible to quantify the patient’s symptoms. The progression reported herein is based on serial examinations, medical records, the patient’s recollection, and the recollections of her family members. The decision was made to treat the patient at 2.0 ATA rather than the 2.5 ATA normally used at the treating institution in cases of soft tissue radionecrosis as a compromise between the need for HBOT and the patient’s increased risk of seizure owing to brain edema. Seizure is a rare adverse effect of HBOT.

Soon after HBOT was begun, the progression of symptoms halted, as determined based on the patient’s description of her symptoms during serial office visits. The notes acquired by her neurologist at a follow-up visit also indicated symptomatic improvement. Lingering symptoms included occasional problems related to balance and memory.

This patient's symptoms, which progressively worsened after the development of soft tissue radionecrosis, were halted and improved with HBOT. The patient was lost to follow-up from the hyperbaric medicine department. Her last known follow-up was with specialists in neurology, neurosurgery, and geriatrics for long-term care.

Case 2

Patients undergoing HBOT who require sedation often receive some combination of fentanyl, propofol, and benzodiazepines either in bolus form or continuously. None of these sedation agents is attractive for use in patients who present with hypotension, however. While ketamine has often been used in animal HBOT studies and in procedural sedation, particularly in children, no studies report the routine use of ketamine in humans undergoing HBOT.^4-6 In case 2, the tendency of ketamine to increase blood pressure was beneficial in sedating an intubated, hypotensive patient to undergo HBOT.

A 55-year-old intubated male with carbon monoxide poisoning was flown to a hyperbaric chamber facility; initially, he received continuous fentanyl and dexmedetomidine. En route to the facility, the patient’s mean arterial pressure dropped from the 70s mm Hg to the 50s mm Hg, and sedation was switched to ketamine with a resultant improvement in blood pressure. At the hyperbaric facility, blood pressure was adequate until sedation was switched to fentanyl and midazolam. Hypotension ensued, and IV fluids were given. Ketamine was administered again for sedation and HBOT commenced.

The patient completed 1 HBOT treatment on ketamine sedation with no further cardiovascular complications. Ketamine was administered 1 mg/kg IV followed by 0.5 mg/kg IV every 30 minutes as needed. Prior to HBOT and ketamine administration, systolic pressure was 82 mm Hg and diastolic pressure was 49 mm Hg. Systolic pressure was 148 mm Hg and diastolic pressure was 93 mm Hg at the end of treatment with IV fluid resuscitation alone and no vasopressors.

In patients who require HBOT and in whom hypotension is a concern, ketamine may be a valuable sedation agent. Intravenous ketamine onset occurs within seconds, with a half-life of approximately 45 minutes. Emergence phenomena may occur, including confusion, hallucinations, and delirium. Ketamine can increase intracranial pressure and blood pressure; thus, an alternative sedative should be used in cases in which ketamine is likely to pose a risk to the patient.

Case 3

The relationship between oxygen exposure and cerebral blood flow has been extensively studied.^7,8 The inverse relationship between them is dose- and time-dependent. There also exists a direct relationship between oxygen exposure and cerebral oxygen content that supersedes the potential loss of oxygenation due to vasoconstriction, such that HBOT has been suggested as a treatment for cerebral ischemic events.

A 58-year-old female presented for HBOT to manage mandibular ORN with cutaneous fistulization. Shortly after the initial consult and prior to HBOT, the patient presented to the emergency department with severe headaches of several days’ duration. Diffuse multifocal narrowing was detected at cerebral arteriography, and CVS was diagnosed. The patient underwent 40 HBOT treatments of 90 minutes each at 2.0 ATA over an 8-week period, and did not experience recurrence of CVS during the treatment period or during a 6-week follow-up period.

Cerebral vasoconstrictive syndrome is rare. To the authors’ knowledge, there are no previous reports in the literature of HBOT administration after this diagnosis. In this case, HBOT was used to treat the ORN, not the CVS. However, it was important to maintain awareness of the CVS because this comorbidity had the potential to complicate treatment by precipitating either another episode like the one that had brought the patient to the emergency department or, worse, an acute ischemic stroke. Based on the available published evidence, which appeared to show neither benefit nor harm from the use of HBOT in the setting of acute ischemic stroke,^9-11 the authors of this case report felt that it could be extrapolated that HBOT was not likely to precipitate it either. Additionally, because of the patient’s advanced ORN, it was felt that even if there was a small risk, that risk was outweighed by the benefit of HBOT for the treatment of ORN. Nonetheless, the patient was treated at 2.0 ATA rather than 2.5 ATA (which would otherwise have been used in the management of ORN), because vasoconstriction due to oxygen exposure is dose-dependent.

The patient experienced no further CVS episodes during the course of treatment or throughout 6-week follow-up with the hyperbaric medicine department. She was lost to follow-up thereafter.

Case 4

Barotrauma is a common adverse effect of HBOT and can occur where a closed, compressible compartment exists. Thoracic surgery opens such compartments, as do novel tracheal-esophageal devices. Patients often present with some combination of these potential contraindications to HBOT, and safety literature is lacking. This case illustrates treatment of a patient with a history of recent thoracic surgery as well as with an antimicrobial stoma button and an implanted voice prosthesis valve.

A 53-year-old female with a history of squamous cell carcinoma of the neck after chemoradiation and total laryngectomy presented for HBOT for the management of mandibular ORN. On presentation, she had a removable antimicrobial stoma button in situ and a permanent voice prosthesis valve. In addition, 1 month prior to her initial consult she underwent upper lobe division segmentectomy to manage a suspicious mass.

The patient was due for a new stoma button, and the authors of this case series asked that an antimicrobial button not be placed while the patient was undergoing HBOT. Initiation of HBOT was deliberately delayed until 12 weeks after the date of the patient’s thoracic surgery. With these and the authors’ standard precautions in place, the patient underwent 30 HBOT treatments of 90 minutes each at 2.5 ATA, with no adverse effects.

There are no guidelines for HBOT after thoracic surgery. Guidelines for commercial flying after thoracic surgery are consensus based, and they advocate waiting until 2 weeks to 6 weeks after surgery to fly.¹² The authors of this case series chose to recommend a more conservative 12-week waiting period after surgery before initiating HBOT because of the elective nature of HBOT and the lack of data to guide that decision. There are no guidelines regarding topical antimicrobial administration or voice prosthesis valves and HBOT, either. Given the choice, the authors opted for a non-antimicrobial stoma button due to concern for increased drug release with an antimicrobial button. The valve was presumed to be safe because of internalization and inability to create a closed compartment.

Case 5

Because carbon monoxide has a stronger affinity for hemoglobin than oxygen does, carbon monoxide binds easily to the hemoglobin molecule, blocks oxygen from binding, and prevents delivery of oxygen to the tissues.¹³ Severity of symptoms is dependent on the duration of the carbon monoxide exposure. Fetuses; infants; children; older persons; patients with cardiovascular disease, anemia, or pulmonary disease; and pregnant women are at increased risk of experiencing carbon monoxide poisoning.¹³ Symptoms of carbon monoxide poisoning include headache, weakness, dizziness, nausea or vomiting, shortness of breath, confusion, blurred vision, and loss of consciousness.³⁰ It is necessary to assess for clinical symptoms of carbon monoxide poisoning and to measure laboratory values to determine the percentage of carboxyhemoglobin in the blood.

A 29-year-old female who was 13 weeks’ gestation reported symptoms of intermittent headaches and dizziness of 2 days’ duration. She stated that she was having problems with her furnace and that the repair person who checked it informed her that it was leaking carbon monoxide. The patient presented to the emergency department for further evaluation. Her initial carboxyhemoglobin level was 8.5%.

Partial placental detachment was seen on ultrasonography. The patient was evaluated by a specialist in maternal fetal medicine, and HBOT was ordered. She underwent 2 HBOT treatments of 90 minutes’ duration each at 2.5 ATA per the protocols of the Undersea and Hyperbaric Medical Society.¹⁴ After treatment, the repeat carboxyhemoglobin level was 2%. Oxygen was administered overnight via a non-rebreather mask, and the repeat carboxyhemoglobin level was within normal limits. The patient was discharged that day with plans to stay with family until the furnace was fixed.

Even if maternal carboxyhemoglobin levels are not toxic, it is important to consider fetal involvement, because fetal carboxyhemoglobin levels can be higher than those of the mother and clearance of the carbon monoxide from the fetus is slower.¹³ The purpose of using HBOT aggressively in the pregnant mother with carbon monoxide poisoning is to protect the fetus.

Case 6

Polycythemia vera is a chronic myeloproliferative disorder that can cause thrombosis resulting from the disease process, which involves blood hyperviscosity.¹⁵ Management includes phlebotomy to decrease the risk of thrombohemorrhagic complications.

A 53-year-old female with polycythemia vera status post exsanguination with a hemoglobin level of 19 mg/dL developed altered mental status, which prompted a CT examination of the head. An arterial air embolus was found. The patient was treated with 3 HBOT treatments of 90 minutes’ duration at 2.8 ATA twice a day based on protocols of the Undersea and Hyperbaric Medical Society.¹⁴ She was monitored overnight and did well.

The occurrence of an intravascular arterial air embolus can result in neurologic complications, including seizures, loss of consciousness, altered mental status, and hemiplegia.¹⁶ Treatment includes use of HBOT to decrease the volume of the air bubble, remove nitrogen, and improve oxygenation.^17,18

Case 7

A 39-year-old male presented with necrotizing fasciitis. He underwent surgical debridement and was admitted to the intensive care unit. Treatment included the use of antimicrobial agents, and HBOT was requested. The patient had a history of hydrocephalus with placement of a ventriculopleural shunt. A chest CT examination was performed to rule out air bubbles in the shunt. The manufacturer was contacted to determine whether the shunt was soft or hard to avoid acute hydrocephalus during treatment with soft tubing. Ninety-minute HBOT treatments at 2.5 ATA based on protocols from the Undersea and Hyperbaric Medical Society¹⁴ were uneventfully completed 3 times daily for 3 days while the patient was in the intensive care unit.

There is scant published literature on the use of HBOT for patients with shunts, but the case report by Guha et al¹⁹ demonstrated uneventful outcomes for a patient undergoing HBOT with a ventriculopleural shunt. Recommendations include contacting the shunt manufacturer and requesting specific information and recommendations regarding the use of HBOT and the particular device.

Case 8

Pain management with an intrathecal pain pump is used for chronic pain therapy when nonsurgical management and interventions have been unsuccessful and surgical intervention has been ruled out.²⁰

A 43-year-old female with back pain after implantation of an intrathecal hydromorphone pump was evaluated for refractory osteomyelitis in the setting of diabetes. Hyperbaric oxygen therapy at 2.5 ATA was recommended. Prior to HBOT administration the device manufacturer was telephoned for recommendations. The level of fluid of the hydromorphone pump was checked daily before treatment to avoid interruption of pump supply during HBOT from fluid compression, and the patient underwent a course of thirty 90-minute HBOT treatments uneventfully.

Case 9

A 34-year-old male with a history of type 1 diabetes presented with a Wagner grade 3 diabetic foot ulcer that was unresponsive to conventional treatment. Adjunctive HBOT was deemed necessary. A continuous glucose monitor and insulin pump were in use to aid in the management of diabetes. The insulin pump was removed prior to HBOT, intermittent injectable insulin was started, and thirty 90-minute treatments of HBOT proceeded uneventfully.

The use of continuous blood glucose monitoring is helpful in the management of diabetes because of the ability to determine trends of glucose fluctuations, which helps predict and detect hypoglycemia and hyperglycemia, and aids in managing the treatment process to provide increased time in euglycemia.²¹ Insulin pumps are being used with increasing frequency, especially for persons with type 1 diabetes, and they can provide improved glycemic control and help reduce hypoglycemic events.²² While these tools aid in improved management of glucose control in the patient with diabetes, their potential interaction with HBOT must be considered. In the hyperbaric environment, the care team must be vigilant to prevent fire. For fire to occur, 3 conditions must be met: adequate fuel, sufficient oxygen, and an ignition source.²³ Electronic devices such as the insulin pump and continuous glucose monitor have the potential to create a spark; thus, it is recommended that these devices be removed prior to the HBOT treatment session.

Hyperbaric oxygen therapy is well tolerated by most patients. Careful patient selection and individualized treatment are paramount to risk mitigation; both tend to be straightforward in patients who are relatively healthy or whose characteristics match those of patients in large published studies. As this case series shows, however, many patients who might benefit from HBOT have unique presentations that require hyperbaric physicians to extrapolate from their knowledge, experience, and any existing supporting evidence to determine a reasonable treatment protocol.

In such cases, the mechanisms of action of HBOT must be considered in relation to the unique risk factor in question. Three discrete questions may be helpful in decision-making. Will HBOT cause the patient harm by diminishing the effect or enhancing the risk presented by a particular medication, device, or condition? Will that medication, device, or condition diminish the effect or enhance any of the risks inherent in HBOT? Is there an unacceptable opportunity cost to delivering HBOT in a particular situation?

Case 2 in this series primarily involved consideration of the first question, specifically, whether the administration of a sedative required to administer HBOT to a hypotensive, intubated patient in cardiogenic shock secondary to carbon monoxide poisoning places that patient at further risk of complete cardiovascular collapse and what can be done to mitigate that risk. Because this patient was intubated and could not be relied on to not cause himself harm in a monoplace hyperbaric chamber, some degree of sedation was required during treatment. Most sedatives commonly used in HBOT cause hypotension, which is not normally a problem because HBOT itself generally causes some degree of hypertension.¹³ The degree and timing of that side effect of HBOT is not predictable enough that it could have been relied on in this case to protect the patient, however, so it was necessary to use an alternative sedating agent. Ketamine is sedating and does not cause hypotension. It is often used safely worldwide in procedural sedation in humans.^4-6 Based on this related evidence, use of ketamine as a sedative was deemed appropriate in this case.

In case 3, it was the potential synergy between hyperbaric oxygen and CVS, both of which may cause vasoconstriction,^7,8,24 that posed a potential threat to the patient’s safety. It was hypothesized that, in the worst case, an ischemic cerebrovascular event might occur. Clinically, HBOT has not been shown to be detrimental in the setting of acute ischemic stroke^7,9,10,25; thus, it seems logical to extrapolate that it would not precipitate an acute ischemic stroke. To further mitigate the patient’s risk of stroke in this case, a lower oxygen pressure of 2.0 ATA was used rather than the standard 2.5 ATA normally used in the management of ORN.

Several of the cases in this series involved implants or other foreign devices. In each instance, it was necessary to determine whether or not HBOT could be detrimental to the device or vice versa. Helpful questions for determining this include, Will the device break? If it fails, is there a risk that it will fail unsafely? For example, external insulin pumps have been noted to deliver inappropriate doses of medication in the hyperbaric setting.²⁶ In case 9 in this series, the insulin pump was removed, which allowed safe delivery of HBOT without altering the treatment protocol. This is often the safest course of action when an external device can be removed. In case 4, although the antimicrobial stoma button could not be removed, replacing it with a nonmedicated version mitigated the risk of unsafe drug dosing.

Internal devices cannot be easily removed. Many medical devices used today are tested at pressure by manufacturers and are rated to withstand several atmospheres. The intrathecal drug pump used in case 8 and the ventriculopleural shunt in case 7 are examples of such devices. If technical manuals are not readily available online, manufacturers can be contacted, and they should have the information available. Alternately, they may have a physiologist on staff who is familiar with the device and who can at least offer an opinion on the safety of the device under hyperbaric conditions. Foreign objects that are completely internal, that do not occlude gas flow, and that are noncompressible are generally safe in hyperbaric conditions. This is because internal foreign objects are unlikely to interact with the hyperbaric oxygen in a way that would increase the risk of fire, nonocclusive objects are unlikely to cause barotrauma, and noncompressible objects are unlikely to cause the patient harm either directly or in a secondary fashion should they not work as expected. This is the case for orthopedic implants, stents, and the voice prosthesis valve in case 4.

Often it is difficult to draw a clear distinction between situations in which HBOT administration may be harmful because of a preexisting condition and situations in which the condition may enhance one of the risks inherent in HBOT administration. The distinction is not usually an important one to make. However, asking both questions can help ensure that a potentially dangerous situation is not missed. For example, in case 4 there is a distinct condition—the recent thoracic surgery—that is not jeopardized by the administration of HBOT but does create a potentially elevated risk of barotrauma, which is always a baseline risk of HBOT administration.²⁷ In that case, the solution was to use air travel guidelines¹² as a basis from which to advise a reasonable waiting period between surgery and HBOT administration. Although it is not known whether these guidelines exist to mitigate the risk of barotrauma (which is admittedly small in a pressurized commercial airplane cabin) or to mitigate the risk of tissue damage caused by postoperative physiologic increase in oxygen demand in a setting in which the partial pressure of oxygen is lower than at sea level (in which case administration of HBOT shortly after thoracic surgery should not be problematic), these were the only guidelines the authors of this case series could use to help form a basis for the treatment decision. Because the guideline was only remotely related to the situation to which it was applied, however, the authors opted for a more conservative waiting period and doubled the maximum waiting period from 6 weeks to 12 weeks after thoracic surgery before starting HBOT.

Limited data exist concerning the effect of most medical interventions on a pregnant patient and fetus, and hyperbaric oxygen therapy is no exception. In treating case 5, the study authors had few data to draw on in considering the potential danger of HBOT administration to the pregnancy and considering whether pregnancy would increase the risk of HBOT administration. To the authors’ knowledge, no prospective studies of HBOT administration in pregnant individuals have been published; however, retrospective reports agree that there appears to be no additional risk of HBOT administration due to pregnancy.^28,29 On the contrary, as previously noted, there is evidence that carbon monoxide binds preferentially to fetal hemoglobin and causes harm to the fetus even when maternal carboxyhemoglobin levels are low.^30,31 In case 5, there was a clear benefit to the patient and her fetus to the administration of HBOTwith an unknown, likely very low, associated risk. Thus, for the pregnant patient it is recommended to proceed with HBOT in emergency situations and defer treatment in elective situations.

Opportunity cost must be considered as well. In nearly every situation in which HBOT is considered for an emergent condition, the treating physician must ask whether this procedure will interfere with a more immediately advantageous procedure. Preparation for HBOT is time-consuming and the treatment itself is often time-consuming as well, particularly those protocols used in emergent situations. When time is short, providers must choose whether to perform HBOT or another procedure; if HBOT is chosen, then sometimes the opportunity to perform that other procedure is lost.

Sometimes even a chronic condition can precipitate the issue of opportunity cost, as in case 1. Cerebral radionecrosis is difficult to diagnose. Specifically, it is difficult to decide noninvasively whether a lesion assessed only via imaging represents recurrence of a tumor or edema and necrosis secondary to radiation. Misdiagnosis could mean the loss of many weeks during which treatment for the malignancy could be attempted. Failure to achieve the desired result via HBOT is not helpful in making a diagnosis after the fact, either, because this can represent either a failure of HBOT to treat the radionecrosis or a true misdiagnosis; it is difficult to make this distinction without performing a biopsy.

A limitation of the current report is that it is a review of a series of cases and not a controlled clinical trial. The case study was written by the authors with assistance from other clinical trials and studies mentioned in the references.

To treat any but the simplest cases in which HBOT is indicated, the hyperbaricist requires knowledge of the mechanisms of action of HBOT as well as sufficient experience to understand the practical and technical issues that commonly arise in its administration. It is also necessary to have a comparable degree of knowledge regarding conditions and devices present in patients who require care. A broad medical background is important in this task, but sometimes it is necessary to consult colleagues in other specialties. Sometimes, as is the case with implanted medical devices, the knowledge sought is outside the field of medicine entirely. Fortunately, device manufacturers usually test their products under a variety of conditions, including hyperbaric conditions, and the information is readily available from the manufacturers and vendors. Manufacturers and vendors also often employ physiologists who can help the provider understand the possible interactions of the particular device with an HBOT environment so that by pooling knowledge and experience, the provider and representative can extrapolate an individual patient’s risk. In this way, it is possible to reasonably advise patients and safely administer HBOT, even to those with complex presentations and novel implanted devices.