Palliative Sedation via Intraosseous Vascular Access: A Safe and Feasible Way to Obtain a Vascular Access End of Life

Introduction

Intraosseous (IO) access is normally reserved for emergencies and critical care conditions when venous cannulation is not possible. There are very few references in the literature regarding other uses of an IO route, despite relatively few contraindications, and an IO access is rarely used outside the field of emergency medicine.^1–3

In specialized palliative care, many drugs are administrated subcutaneously but some medication and medical conditions need a reliable and stable venous access, which can be difficult to achieve in terminally ill patients. Propofol, used for sedation, has a short half-life and is, therefore, dependent on a stable intravenous infusion. ⁴

Palliative sedation with propofol is effective^4,5 but still not very often used in Swedish palliative care. However, we have used propofol safely both in our hospice and in different home care settings. The following case report describes how we performed a palliative sedation with propofol, in a home care setting, despite no venous access.

Case Description

A 56-year-old man, tetraplegic for many years due to progressive spinal muscular atrophy, was referred to our Advanced Home Health Care service due to intractable pain and immense suffering.

The patient was confined to his bed due to pain and repeated osteoporotic stress fractures. He was cared for sitting in a repositioning sling-seat connected to a caring bar and a ceiling lift that he could control through a mouth-operated joystick. He required nighttime pressure-controlled ventilation through a continuous positive airway pressure machine and was dependent on two personal caregivers nonstop.

His pain had deteriorated over the course of months and to a point where he could no longer lie down but he sat floating in the sling around the clock. He reported intolerable pain throughout the whole body but most specifically in the pelvis region and in his lower limbs. The slightest movement caused severe pain and the daily care was almost impossible to perform, a life situation that he described as torture. He had ceased to eat and drank very little. He had explicit requests for euthanasia, as soon as possible, and thus to be relieved of his intolerable suffering. The patient expressed a strong wish to stay and to be allowed to die in his home. We concluded that it was neither ethical nor possible, due to his medical condition, to refer him further but to care for him in his home.

At the time of referral, he was prescribed a daily dose of 50 mg methadone orally. Initially, we changed the route of administration from oral to subcutaneous (SC) and gave methadone 10 mg/mL (50 mg/day) and added midazolam 5 mg/mL (10 mg/day) for continuous infusion using a Niki^® T34 syringe driver (Caesarea Medical Electronics GmbH, Lichtenstein). In addition, the patient received a CADD-Legacy^® patient-controlled analgesic (PCA) pump (Smiths Medical International, Inc., MN) with hydromorphone10 mg/mL with a continuous SC infusion of 1 mg per hour, and with the freedom for his caregivers to give him a bolus dose of 4 mg (with a lockout interval of 15 minutes) whenever he needed.

He was in our care for 11 days, with multiple evaluations daily, with the attempt of optimizing his symptoms and well-being and at the same time being a psychological support to his caregivers. Despite increasing doses of methadone (70 mg/day) and midazolam (25 mg/day) through the syringe driver, and hydromorphone of 4 mg per hour with bolus doses of 6 mg through the PCA pump, he still described his situation as unbearable. We had repeated discussions with him and his caregivers about palliative sedation, as an ethically acceptable way to alleviate the burden of his intractable suffering and not as a means to hasten his death.

His condition continued to deteriorate in general, but above all with escalating levels of pain and anxiety. The situation got increasingly difficult to handle in the homecare setting. We concluded that his symptoms were refractory and that the only possible way to gain symptom control was to reduce his consciousness level through palliative sedation. Since the level of anxiety had been escalating, despite adequate doses of midazolam, we seriously started to consider whether it could be due to a paradoxical reaction to midazolam. We, therefore, decided to use propofol, instead of midazolam, as a sedative. However, the patient had no vascular access. Several attempts at peripheral intravenous catheter placement failed and a central venous catheter placement was not possible due to contractures and scoliosis. We, therefore, decided, although somewhat hesitantly, to use an IO cannulation to gain a stable vascular access. The patient immediately accepted this proposal.

An IO cannula (15 G 45 mm) was inserted, at the first attempt, into the patient's right humeral head^6,7 using an EZ-IO driver (Vidacare, San Antonio, TX). The placement of the IO cannulation was associated with some pain and the patient experienced a moderate level of local pain when the infusion of propofol was started. This local pain was effectively treated with bolus doses of SC hydromorphone. Propofol was administered through an Alaris^® GW CareFusion (CareFusion Corporation, Switzerland). The cannula worked smoothly during the four days it was used and did not need replacement.

Palliative sedation was initiated with a bolus dose of 20 mg propofol (10 mg/mL). When the first dose was given, the patient slept quietly for five minutes and reported a pleasant sleep on waking up. A repeat bolus dose of 20 mg was given and then a continuous infusion of 20 mg (2 mL) per hour was started. On this dose, the patient did not sleep continuously but was capable of some level of communication and of intake of oral fluids whenever requested. The continuous dose was subsequently increased, to maintain symptom relief, and titrated to 60 mg per hour at the end of the first 24 hours, to 90 mg on day 2, and to 110 mg per hour on the last day of his life.

We continued the SC infusion of methadone (70 mg/day) and gradually increased the hydromorphone dose, to meet and ensure an adequate pain relief, and during the last days of his life, the continuous infusion was 4 mg per hour and the bolus dose of 6 mg.

The patient died after 11 days of palliative care, of which the last 4 days were with palliative sedation with propofol using an IO cannula as a vascular access. No complications were noted from this route of administration.

Discussion

This case posed us some dilemmas.

The patient's immediate and explicit requests for euthanasia probably made us defer initiating palliative sedation. Extreme distress is a medical emergency and this patient's suffering was intolerable. However, given the profound interindividual variability of responsiveness to palliative interventions, an assessment whether an intervention is futile or appropriate is often very difficult to make. We initially focused on assessing any potentially overlooked and reversible causes to his distress and on finding an adequate treatment. Nonetheless, despite our efforts, the patient's distress escalated rather than subsided and only when we lacked other methods of palliation did we initiate palliative sedation.

Moreover, once we decided to start palliative sedation, we were challenged with the fact that the patient lacked a venous access. We were able to place the IO needle at the first attempt. We could then immediately start the palliative sedation with propofol and perform and monitor it safely without any complications. We were concerned whether the procedure to place an IO cannula would cause further acute or ongoing pain or even worse cause a fracture in his fragile bone. We were also worried that the needle would stop working during the continuous sedation or whether an extended sedation over several days would lead to a local infection. Osteoporosis is reported to be a semicontraindication for IO placement due to the obvious risk for iatrogenic fractures. Likewise, prolonged IO access use is linked to an increased risk for serious complication as osteomyelitis, cellulitis, and skin abscesses and should be restricted to a few hours until vascular access is achieved without exceeding 24 hours.^1–3 In our case, we had no choice but to continue with an IO infusion, even though it meant an increased risk of serious events, since we could not find any better treatment alternatives for his refractory distress. We would not hesitate to use an IO cannula, as a means of gain a vascular access, in a similar situation again. Hence, we would like to advocate the use of IO placement in the palliative care of terminally ill patients, when other possibilities have failed to obtain a venous access.

Conclusion

Palliative sedation using an IO cannulation is a safe and feasible way to obtain vascular access end of life.