Withdrawing medication

Goals of care

Margaret is married and lives with her husband of 43 years; they were unable to have children. She is a retired office receptionist, and her husband has retired from the military. They live in their own home. They have a few close friends and there are no family members living in their area.

Margaret and her husband return to discuss the diagnosis and prognosis. It is a very difficult conversation and they are both overwhelmed by the news. You refer her to an oncologist to further discuss the diagnosis, prognosis, and options for treatment. She is informed of the potential side effects and substantial morbidity that might result with treatment. After much thought and angst, given the poor prognosis with or without treatment, she decides to forgo any further life-prolonging treatment.

As physicians we need to be aware of patients’ spiritual, religious, and cultural beliefs, as these factors might affect their decision making when it comes to withdrawing or withholding medication or other treatments at the end of life. When initiating these discussions, the decisions to decline treatment or withdraw medication might occur all at once as a single and complete change in direction, or they might take place over time, with specific treatments, medications, and devices (eg, implantable cardioverter defibrillator, continuous positive airway pressure) being gradually withdrawn.2,3

Little guidance exists to help physicians consistently manage chronic comorbidities during this period of change. Frequent review and amendments to goals of care are necessary when managing medical comorbidities. The need for certain medications, as well as their route and time of administration, should be evaluated regularly. We must always be looking at what is best for the patient.4

One month later, you see Margaret on a house visit. It is too difficult for her to visit your office now. It has been more than 2 months since her diagnosis, and she spends most of her days in bed. She has a poor appetite, with occasional nausea but no vomiting. Her husband comments on the number of pills she has to take and how it has become a chore. She denies any pain and complains of shortness of breath with slight exertion but not at rest. On physical assessment, she has lost 25 lb (she now weighs 105 lb) and her mucous membranes are dry. Air entry is decreased in her right lung. Respiration is at a rate of 26 breaths/min and slightly laboured while talking. She needs assistance to get up out of bed and with her activities of daily living. There has been a substantial functional decline in Margaret in a very short time, with her palliative performance scale version 2 score at 40%.5 New medications are initiated to help manage her symptoms of dyspnea and nausea. You recognize that the time has come to initiate discussion about the need for a number of her medications at this stage of her disease.

Period of change

A patient facing a palliative diagnosis and prognosis, and coming to grips with a limited life expectancy of weeks to a few months, deserves to have discussions on goals of care introduced. It is best to have these discussions with patients and their families. As illness progresses, a catabolic state develops as fat, carbohydrate, and protein metabolism is altered. This catabolic state coupled with the release of various cytokines adds to the triad of anorexia, weight loss, and fatigue. Consequently weight loss, possible cachexia, and dehydration arise. During end-of-life care, it is common for new medications to be added for symptoms such as nausea, dyspnea, and pain, to name a few. There is a great potential for polypharmacy and its consequent adverse effects if these medications are added to existing medications for comorbidities.6

You and a home care nurse visit Margaret and her husband to talk about medication withdrawal. (Table 16,7 summarizes factors to consider when withdrawing medications.) During this period of change, communication, support, and reassurance are essential; it is important to discuss and redefine the goals of care and treatments. Based on the evidence, Margaret and her husband can make informed decisions. Given the evidence (number need to treat and time until benefit) and her considerable physical and functional decline (palliative performance scale score at 40%), it seems reasonable to discontinue atorvastatin and alendronate.8,9 After discussing evidence on primary and secondary prevention with antihypertensives, she agrees to stop using ramipril and hydrochlorothiazide. She continues to take metoprolol as long as she is able, as it might prevent morbidity (angina and tachycardia) given her history of coronary artery disease and atrial fibrillation.10

After reviewing her anticoagulation therapy, you decide it is important to consider some form of anticoagulation with her high risk of venous thromboembolism and resulting morbidity. Margaret is informed of the increased risk of adverse consequences of warfarin such as bleeding complications, especially in her state of nutritional compromise and with the interference of competing medications. She and her husband are frustrated by the frequency of phlebotomy, as her international normalized ratio is often out of therapeutic range. She is informed of the risks of stopping her warfarin and given the option of no further anticoagulation versus injections of low-molecular-weight heparin, which do not require monitoring. The evidence is weak but does support low-molecular-weight heparin in palliative care even in advanced disease; however, it is clearly a decision based on benefit versus burden.11–14 Margaret decides to forgo any further treatment with anticoagulants, and you support her in this decision.

The UKPDS15 (UK Prospective Diabetes Study) has formed the cornerstone of current guidelines in the management of type II diabetes. Extrapolating these guidelines and applying them to a population with a very limited life expectancy might not be practical. Looking at number needed to treat and time to effect with conventional diabetic treatments suggests we do not need aggressive treatment or tight glucose control. A glucose level of about 15 mmol/L or an undesirable symptom resulting from hyperglycemia might be cause to consider initiating oral hypoglycemics. Intensive treatments with insulin increase the risk of hypoglycemia, not to mention the discomfort and inconvenience of insulin injections.16

Discussions are initiated regarding Margaret’s diabetic medication. She is anorexic and has had no hypoglycemic episodes, but there is concern about the risk if she continues taking her diabetic medication. After discussing the options, she agrees with stopping her metformin. She agrees to occasional glucose testing at her and her husband’s discretion based on symptoms.

Two weeks later the palliative home support nurse receives a call from Margaret’s husband. Margaret’s condition has deteriorated substantially. She is unable to take any oral medications and only has sips of liquid. From this point all of her oral medications are stopped. The focus is on her palliative symptoms and great attention is given to good pain and symptom management, as well as emotional support for both her and her husband. She dies peacefully 4 days later.17