Limited evidence on the long-term prescribing of opioids in elders

Référence

O'Brien MDC, Wand APF. A systematic review of the evidence for the efficacy of opioids for chronic non-cancer pain in community-dwelling older adults. Age Ageing 2020;49(2):175-183.

Sommaire

These authors searched several databases for studies published since 2003 that evaluated opioid use in adults older than 65 years who were being treated for chronic (at least 3 months) noncancer pain. They supplemented their search by hand-checking the reference lists of included studies, but didn't look at clinical trials databases or other sources of unpublished studies. This is the first paper in my recollection for which the authors found more studies in the reference lists than in their structured search. Maybe they should have engaged the help of a librarian. The authors used standardized tools, specific to various research designs, to assess the quality of the included studies. Ultimately they found only 7 studies with between 10 and 10,372 patients. Two of the studies took place in care facilities and 5 in community settings. Four of the studies were cross sectional, 2 were prospective and 1 was a chart review. Please note—the authors found no randomized trials. Given this kind of variability of study designs, the authors wisely chose to not pool the results. The quality of the studies ranged widely; overall, they were of modest quality. For pain relief, the study findings ran the gamut from high rates of pain relief but with significant residual pain to chronic undertreatment of pain, especially among those with significant cognitive impairment. One study found that opioids improved the sleep quality of patients with chronic osteoarthritis pain but did not improve physical or social function. However, an additional study found that opioid use in nursing home residents was associated with significant improvement in activities of daily living and social engagement. Overall, the studies have significant confounding issues that limit our ability to draw many conclusions. The authors also didn't report on the harms of opioid use in this potentially vulnerable population.