Abstract
We report on the case of an elderly patient in a palliative care unit who fell and sustained a hip fracture. Hip fractures are potentially terminal events for elderly patients with other life-limiting conditions. The case highlights the need for more evidence to determine the best approach to care for hip fractures in patients who are in the final weeks or months of life.
Keywords
An 84-year-old man with a history of glioblastoma multiforme was admitted to a Toronto, Ontario hospital palliative care inpatient unit (PCU) for end-of-life care. There was no further active management possible for his disease, and his life expectancy was estimated to be 1 to 3 months based on his disease burden and Palliative Performance Scale (50%). 1 His past medical history consisted of prostatic hypertrophy and dementia. His goals of care on admission to the unit were to receive symptomatic care for his disease and to remain as independent as possible. A family member was acting as his substitute decision-maker for complex decisions as he had cognitive impairment that was mild initially but progressed during his stay. The expectation of the patient and his family was that he would die in the PCU.
The patient was identified as being at high risk for falls early in his admission. He was ambulatory in his room and the hallway but observed to be unsteady. He was seen by the physiotherapist who recommended the use of a walker, which he refused to use. He had discussions with several care team members about strategies to reduce his risk of falls. He was encouraged to ask for help when ambulating but refused to do this. He had bed alarms in place so that staff could come to his assistance when he exited the bed. He had a comprehensive discussion about risk of falls with the attending physician on the second week of his admission. He indicated that he would rather die of a hip fracture than be forced to use a walker. At the time of that discussion, the physician felt he was competent to understand the implications of his decision. The physician discussed the fall risk with his family and all were in agreement with a plan to try to mitigate the risk of falls (through use of the bed alarm, physiotherapy, and having the walker available), but allowing the patient to walk as per his wishes for reasons of quality of life.
The patient had an unwitnessed fall from his own height post admission while attempting to stand at the bedside. An X-ray showed a right nondisplaced intertrochanteric hip fracture.
The team and family considered both operative and conservative management of the patient’s hip fracture. At the time of his fall, the patient’s revised prefracture life expectancy was thought to be 1 to 2 months. This was due to increasing difficulties with cognition, mobility, and agitation and an overall decline in function. At the time of his fall, he was incompetent to make both simple and complex decisions. After the fall, his Palliative Performance Scale was 30%, and he became increasingly delirious and was unable to meaningfully interact with visitors and staff. With or without surgery, the PCU physician estimated the postfracture life expectancy to be 2 to 4 weeks. This was because of the added disease burden of hip fracture, the possible complications of surgery, the patient’s diminished ability to participate in rehabilitation, and the obligate period of immobility (even if the surgery was corrective). 2 His family was concerned about operative management due to the pain and discomfort of an acute care visit and the fact that surgery (and the following recovery period) would be sources of pain and anxiety in of themselves. This concern was balanced by a desire to have his pain controlled at the end of life. The family agreed to a transfer to acute care for orthopedic consultation as there were no acute care services available at the patient’s home hospital.
Three orthopedic surgeons in 3 separate teaching hospitals were consulted by phone (for administrative and resource availability reasons). All were in agreement that the sole indication for surgery would be pain relief. This was in keeping with the goals discussed with the PCU team and the family. It was felt that the likelihood of the patient becoming independently ambulatory given his disease burden, rehabilitation potential, and prognosis was remote. One surgeon indicated that he would not recommend surgery at all, the second surgeon saw surgery as the only viable option, and the third thought surgery should be considered only if the team thought that the patient’s life expectancy exceeded 1 month.
The patient was sent to the emergency department for an in-person consultation with the on-call orthopedic surgeon. He was sent back to the PCU by ambulance having been deemed not appropriate for surgery by the orthopedics service. The reasons cited were his comorbidities and poor prognosis. The recommendation was for symptomatic nonsurgical management.
Conservative treatment in the PCU included the use of rapidly titrated doses of hydromorphone and midazolam to control pain and anxiety. The patient was opiate naive and had limited benzodiazepine use previous to the fracture. One week postfracture, he was receiving 14 mg of hydromorphone and 20 mg of midazolam by subcutaneous infuser every 24 hours. He was maintained on those doses until his death. While at rest, the patient was noted to be calm and comfortable. There was no facial grimacing or calling out. He did experience significant pain with care, and this presented the greatest challenge to the team. Solutions were found by administering 100 µg of sublingual fentanyl prior to care and dexamethasone subcutaneously for adjuvant pain control. On the recommendation of the physiotherapist, his right lower limb was packed with sandbags to minimize movement. Nursing practice was adjusted to reduce his pain with movement. Procedures and care events were limited and grouped. Adapted 2 person transfers were used when movement was required. Patient reassurance and distraction during care were also noted to be helpful. With these adjustments, it was possible to deliver the necessary care with minimal patient discomfort. The patient died in the palliative care unit on the 11th day postfracture.
Discussion
In most patients with hip fractures, operative management within the first 48 hours is the standard of care. Surgery is generally preferred over conservative management due to superior outcomes in mobility, length of hospital stay, risk of Venous Thromboembolism, difficulty with nursing care, and pain control. 3 This approach may not be optimal for many patients in the palliative care setting.
Evidence is sparse for outcomes after hip fracture in the palliative population and severely lacking for patients in the final months or weeks of life. 4 One small study showed no improved mortality for hospice patients with a 6-month life expectancy with surgical versus conservative treatment. 5 A larger study showed a modest improvement in survival but this was in a population with life expectancy of 1 year. 4 For many patients in the palliative setting, life expectancies are limited to weeks or a few months. Surgery to improve survival might be considered moot when prolongation of life ceases to be a goal.
The time required for recovery from surgery is also at issue in patients with a limited prognosis pre fracture. Total recovery times for the most common surgical repairs can be on the order of months. 6 Optimal recovery may also involve rehabilitation programs that many palliative care patients could not participate in given poor functional status and comorbid conditions. When considering the most debilitated patients, surgery may not lead to any functional rehabilitation and thus may not be a reasonable goal.
Improvement in pain control is another possible goal for operative management in patients with limited life expectancies. Again, the evidence for degree of pain control in surgical versus conservative treatment in the subgroup of very debilitated patients is lacking. 5 The pain from the fracture itself must be weighed against possible increased pain in the postoperative period and the ability of pain to be managed without operative intervention. This was the greatest challenge for the clinical team in this case as multiple interventions (and staff resources) were needed to decrease pain with movement.
The ability of the patient’s care team to manage symptom burden must also be considered. In the present case, our patient was managed conservatively in a palliative care unit with access to physiotherapy services, specialized equipment, and around-the-clock nursing care. Like many palliative care units, our team regularly provides care for patients with pain from bone metastases and pathological fractures. These skills are transferable to patients with traumatic fracture. Patients receiving care from specialized teams similar to ours have shown improved pain outcomes when compared to those on standard acute care wards. 7
Patients with hip fracture will have altered prognosis from their prefracture state. 7 Disease burden from the fracture decreases the already poor prognosis of this severely ill patient population, and an adjusted prognosis must be considered when defining new goals of care.
Many patients at the end of life are cared for in settings without easy access to orthopedics consultation (home, hospice, nursing home). This was true in our setting. In such cases, possible morbidity associated with transfer must also be considered. For a severely ill patient at the end of life, a transfer of facilities for consultation or X-ray can cause significant discomfort. The time spent in investigation and evaluation may take up a significant percentage of the patient’s remaining life. The need for a transfer, may, in some cases, tip the balance in favor of conservative management.
In our case, initial indecision about the best treatment plan resulted in an ultimately unnecessary transfer and emergency stay for a patient in the last days of his life. In the end, the patient was successfully managed conservatively in the PCU and would have perhaps been better served by simply remaining there and forgoing the in-person orthopedics consultation. Both clinical teams (palliative care and orthopedics) and the patient’s family were in agreement that pain control was the goal. The only matter of debate was whether or not pain control would have been superior to surgery. In this case, it is difficult to know the answer. Given the patient’s ultimately short survival and the success of his conservative symptom management, in hindsight, nonsurgical management did meet the goals of care.
Conclusion
The literature is lacking in clear direction on operative versus conservative treatment for hip fracture in the last weeks or months of life. Our experience demonstrates that expert opinion can also be varied.
Decisions on how to treat arising acute conditions at the end of life need to take into account the patient’s goals, abilities, and prognosis. Since these are highly variable in a palliative population, a “one size fits all” guideline is difficult to formulate. Cases such as this one illustrate the need to evaluate the whole needs of the patient on a case-by-case basis.
This case illustrates that pain control can be adequately delivered without surgery and conservative management should be considered as a viable treatment option for dying patients.
Those who are most suited to conservative management are likely to be: in the last days to weeks of life, in a setting where expert symptom management can be delivered, have pain control (and not mortality or mobility) as a goal, those who would incur increased morbidity from surgery and/or surgical evaluation (for reasons of disease burden, functional status, or geography).
It is prudent to consider that the management of hip fracture in this subset of patients remains controversial and decisions are likely to be made on an individualized basis in conversation with patients, families, and multiple clinical services.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
