Behaviour management in dementia

Case description: part 1

Mr J.C. is an 84-year-old widower who has been a nursing home resident for the past 2 years. He was admitted with mild dementia, which has progressively worsened. His nurse calls because he has been becoming more agitated and is physically aggressive at times. His sleep has deteriorated; occasionally he disturbs other residents at night with shouting and calling out. He is not oriented to time or place, and rarely remembers visits from family or friends. Mr J.C.’s mobility has declined; he spends most waking hours sitting. He has been moved within the nursing home several times owing to complaints from other residents and families.

Mr J.C.’s history includes hypertension, type 2 diabetes mellitus, and osteoarthritis (OA). He has known benign prostatic hyperplasia, with urinary urgency and dribbling at times. He was briefly treated with anticholinergic medication for urgency, but his behaviour worsened. He has a remote history of heavy alcohol use and a 40-pack-year smoking history, although he has not indulged either habit for the past 3 years. His medications include 10 mg of ramipril daily, 500 mg of metformin twice daily, and intermittent ibuprofen or acetaminophen for arthritis. He no longer complains that his joints are painful. He occasionally becomes constipated and requires laxatives.

On examination, Mr J.C. seems a bit agitated and resistant. He does not appear to recognize his regular family doctor. His pulse is 82 beats/min and regular. Blood pressure is 142/78 mm Hg. Temperature is 37.2°C. Physical findings include generalized joint changes typical of OA, early contractures of his knees and hips, mild lower abdominal tenderness, and constipated stool on rectal examination. Complete blood count and electrolyte levels are essentially normal. Results of a spot blood sugar measurement reveal a level of 7.9 mmol/L, and a urine dipstick reveals moderate leukocytes.

The nurse on duty is asking for something to help sedate Mr J.C. so that his behaviour is less disruptive.

Bringing evidence to practice

• Look for triggering factors when evaluating new behavioural issues—for example, infection, medications, constipation, or depression^1,2 (Box 1^2,3).

• Consider whether pain could be causing or exacerbating the behaviour, and, if so, treat the cause (eg, urinary retention, constipation) or manage the pain with analgesics as appropriate (eg, a trial of regular acetaminophen for osteoarthritis).^4,5

• Use nondrug treatments first where applicable^6,7 (Box 2³).

• Pharmacotherapy can be added when a treatable cause cannot be identified and nonpharmacologic measures have been tried and are inadequate owing to risk of harm to self or others.⁶ The clinical value of drugs in behaviour management is limited owing to a modest benefit that is offset by serious adverse events.

Case description: part 2

Mr J.C. is found to have a urinary tract infection, which is successfully treated. He is also placed on a regular bowel care regimen, and acetaminophen is given on a schedule (650 mg, 4 times daily) for his OA. His behaviour, although not completely settled, improves. A few weeks later, however, his agitation worsens again; this is treated with intermittent doses of lorazepam. Over time, nursing staff become more concerned about lewd remarks and some groping of staff. After 6 months of deteriorating behaviour, 0.5 mg of risperidone twice daily is added to his medical therapy and the lorazepam is discontinued.

Bringing the evidence to practice

• Consider whether the symptoms need pharmacologic treatment. Some symptoms might be tolerated if they are not harmful to the patient or to others.

• There is some evidence that a stepwise protocol of analgesics considerably improves overall neuropsychiatric symptoms and pain in nursing home residents with moderate to severe dementia and agitation.⁴

• Benzodiazepines can cause confusion, disinhibition, and a worsening of symptoms.⁸ They might also increase the risk of falls and fractures.^9–11 They should generally be reserved for short-term treatment of acute or procedural anxiety.

• There is some evidence that antipsychotic medications might be of benefit in controlling certain behavioural symptoms, specifically severe agitation, aggression, and psychosis. The effect of antipsychotic medication is modest and the clinical value is limited owing to serious adverse events, including a possible increase in mortality^12–14 (Table 1^8,15–17). Beneficial effects might be greater for those with more severe behavioural issues. Reassess beneficial and adverse effects of antipsychotic medications within 3 to 7 days of starting them. Patients with Lewy body disease (eg, visual hallucinations) typically have increased sensitivity and intolerance to antipsychotic drugs.

• Attempts at withdrawing antipsychotic drugs should be made regularly (eg, taper and stop if possible after 3 months).⁸ Follow-up to the DART-AD (Dementia Antipsychotic Withdrawal) trial found that mortality risk was reduced from 71% to 46% over 2 years in patients whose antipsychotic medications were stopped (number need to treat was 4).¹⁷

• Other drugs, such as anticonvulsants, are sometimes tried, although data for efficacy are lacking and adverse events are common.¹⁸ More information is available from CFPlus.*

Case description: part 3

Mr J.C. does not tolerate risperidone owing to excessive drowsiness, which substantially interferes with his ability to eat. A short trial of quetiapine (25 to 50 mg at bedtime) is better tolerated but largely ineffective for behaviour modification. Thus, a decision is made to discontinue the antipsychotic and focus on better pain management with nondrug interventions in addition to his regular acetaminophen. Additional pain relief options, such as a topical nonsteroidal anti-inflammatory drug or a low-dose regular opioid, could be added if necessary.¹⁹ Efforts to control the pain help to reduce Mr J.C.’s agitation. A meeting is held with the family to discuss the natural progression of dementia and the benefits and risks of therapeutic options. The family and nursing staff agree that tolerating behaviour that is not dangerous to the patient, other residents, or staff is preferable to causing adverse events related to medications.