A case of topical opioid-induced delirium mistaken as behavioural and psychological symptoms of dementia in demented state


Correspondence: Dr Go Ito MD, Department of Psychiatry, Aichi Medical University, 1-1 Yazako Karimata, Nagakute 480-1195, Japan. Email: go-ito@umin.net


In Japan, indications for opioid analgesics, once exclusively used as pain killers for patients suffering from malignant cancer, have been expanded for a wide range of pain. Herein we report a patient with opioid-induced delirium associated with the administration of buprenorphine patches that was well below the indicated therapeutic range limit. An 82-year-old woman was referred to us from an orthopaedic practitioner for uncontrollable behavioural problems apparently caused by the beginning of dementia; the patient had gradually developed disorientation, visual hallucinations, and delusions. Laboratory and imaging findings excluded common causes of delirium including Alzheimer's disease and diffuse Lewy body disease. Detailed questioning revealed that the patient's confused state appeared following a buprenorphine patch dose increase and subsequently disappeared after administration was stopped. Delirium has not been reported as a side-effect in clinical trials of buprenorphine patches. However, our findings in this case show that even topical opioids can precipitate the development of a delirious state in elderly patients.


Opioids have been used for thousands of years for palliative care.[1] The Sumerians in Mesopotamia were among the first to start cultivating opium poppy flowers around 3400 BCE. After this time, they spread to every major civilization, including those in Europe and Asia, and have been used to treat pain around the world. Since the extraction of morphine from opium by the German chemist Friedrich Serturner in 1803, opioid drugs rapidly prevailed in medical practice as potent painkillers.[2] However, problems with indiscriminate use of opioids soon became apparent because of such life-threatening side-effects as respiratory depression, as well as serious risks of recreational abuse and addiction. As a result, opioids were legally banned except for cases of cancerous pain and as adjunctive agents for surgical intervention.[3] At the advent of the 21st century, regulations on opioid use and prescription began to be loosened after the development of several new types of medications that are safer than conventional preparations, though that assertion remains controversial.[4]

Herein we report a unique case of an elderly Japanese woman with opioid-induced delirium, which was initially confused with other common behavioural problems in elderly individuals. Our findings alert to a possible drastic increase of similar cases as a result of loosened regulations for opioid use in the near future.

Case Report

The patient was 82-year-old woman who lived independently. She was referred to our department because of the recent development of severe behavioural problems. She was admitted to our hospital 2 days after abnormal behaviour was pointed out (day X). Previously, the patient had suffered from chronic back pain and was treated by an orthopaedic practitioner. About 6 weeks prior to admission (day X – 43), 37.5 mg tramadol and 325 mg acetaminophen were initiated to relieve pain. On day X – 24, tramadol was suddenly stopped and switched to buprenorphine patches (5 mg), possibly as a result of the strong nausea caused by tramadol. On day X – 12, the buprenorphine patch dose was increased to 10 mg. Although the backache was ameliorated with the increased dose of buprenorphine, the patient became increasingly inactive and lethargic. On day X – 5, the patch was stopped due to a local skin eruption, and the original mixture of tramadol and acetaminophen was restarted (Fig. 1).

Figure 1.

Opioid use before admission. X, day of admission.

On day X – 4, the patient began to assert that insects were wriggling on the ceiling, unknown children came to her bedside, and a couple of birds flew into her bedroom through the window. On day X – 3, she was brought to the emergency room of our hospital (Aichi Medical University Hospital, Nagakute, Japan) because of these delusional complaints. Computed tomography scanning revealed no particular findings except for slightly enlarged lateral ventricles on both sides (Fig. 2). Tramadol and acetaminophen were stopped the same day and she was sent home. On day X – 2, the patient misrecognized a car incidentally parked in front of her house as a bathroom and attempted to break into it while carrying toilet paper. On day X – 1, she walked away from her house in the early morning and was found several blocks away in her pyjamas. On day X, her family brought her to us to seek protective custody as well as possible medical help.

Figure 2.

No obvious haemorrhage or mass was observed on head computed tomography.

Findings upon admission included body temperature of 36.6°C, pulse at 80/min, and blood pressure of 120/70 mmHg. Laboratory tests were non-contributory, except for slightly elevated creatine kinase, possibly from the long-distance walking that morning. To our surprise, the patient could tell us where she was, what day it was, and who brought her to the hospital; she also had a score of 23/30 on the Mini-Mental State Examination. Furthermore, no neurological signs were found. However, at the second visit on the same day, the patient failed to completely answer the same questions. While 123I-metaiodobenzylguanidine scintigraphy showed no sign of decreased accumulation (Fig. 3), electroencephalogram findings revealed generalized slowing (Fig. 4). No particular pathological signs suggestive of dementia with Lewy bodies or Alzheimer's disease were noted in magnetic resonance imaging or single-photon emission computed tomography with technetium-99m-ethyl cysteinate diethylester examinations (Fig. 5, 6).

Figure 3.

123I-metaiodobenzylguanidine scintigraphy showed 2.929 of early H/M ratio and 3.227 of delay H/M ratio, indicating no decreased accumulation. H/M, heart-to-mediastinum.

Figure 4.

On day X + 3, electroencephalogram showed a generalized diffuse slow wave, 8∼9 Hz, with no epileptiform discharges. X, day of admission.

Figure 5.

Low intensity by T1 and high intensity by T2 (left) and FLAIR (right) at both sides of the putamen and around the lateral ventricle, which might be the result of ageing. High intensity by T2 and low intensity by T1 and FLAIR at both sides of the nucleus basalis that indicate enlargement of vessels (middle). FLAIR, fluid-attenuated inversion recovery.

Figure 6.

Brain single-photon emission computed tomography with technetium-99m-ethyl cysteinate diethylester was performed. It revealed a mottled pattern of hypoperfusion at both frontal lobes and slight hypoperfusion at left parietal lobes, which is thought to be the result of ageing. No obvious hypoperfusion at the occipital lobes or hippocampus was observed.

Until day X + 6, the patient's cognitive ability fluctuated greatly. For example, the patient mistook the bed of another patient as her own and slipped into it. She wandered around and muttered to herself in the middle of the night. Her sudden, unprovoked or provoked explosions of anger often startled the staff and required temporary physical restraint on several occasions. Loitering, muttering, and emotionally lability as well as unpredicted behaviours continued. Gradually, the behavioural aberrations became inconspicuous, though on day X + 8 she still had delusory ideas, such as meeting school friends and family members who lived far from the hospital. Although the patient behaved in a more controlled manner as before the introduction of opioids, she could not recall any events until day X + 15. Electroencephalogram findings on day X + 22 revealed a restoration of normal background rhythm of 10–11-Hz waves (Fig. 7). On day X + 31, the patient was discharged after a successful tentative sleepover at her son's house.

Figure 7.

On day X + 22, electroencephalogram was performed with findings of 10–11-Hz waves. X, day of admission.


In the present case, imaging studies such as magnetic resonance imaging and single-photon emission computed tomography failed to show any signs suggestive of central nervous system diseases, except for non-specific frontal lobe hypoperfusion, which might explain the acute deterioration of behaviour. Laboratory tests did not reveal any common causes of a delirious state in elderly individuals, such as infection, electrolyte imbalance, or other metabolic disorders. No alcohol abuse was reported and no drugs that might precipitate delirium such as benzodiazepines, anticholinergic agents, or antidepressant agents had been prescribed. While all other plausible causes possibly leading to delirium were excluded, the timing of the opioid administrations corresponded well with the appearance and disappearance of a delirious state, as well as with the slowing of the electroencephalogram. Our findings indicated the opioids as the most likely culprit.

Our patient was given two types of opioids, buprenorphine and tramadol. A buprenorphine patch is designed to provide continuous relief from moderate to severe pain for up to 7 days and has a half-life of 15 ± 6 h for the 5-mg dose and 30 ± 21 h for 10 mg. The patches became available in some countries in July 2003 and began to be distributed in Japan in August 2011.[5] Tramadol, with a half-life of 5.1 ± 0.8 h (37.5 mg), became available in some countries in December 2008 as a tablet mixed with acetaminophen and began to be distributed in Japan in July 2011.[6]

No behavioural problems known as opioid-related side-effects were recognized when tramadol was prescribed, thus it is likely that tramadol was not the cause of the psychotic behaviour. Furthermore, the short half-life of tramadol (5 h) supports its exclusion as the responsible agent, while inactivity and lethargy in the patient were only recognized after increasing the dose of the buprenorphine patch from 5 to 10 mg. Taken together, administration of the buprenorphine patch seems to be responsible for the delirious state.

It should also be noted that a literature search failed to reveal any case of delirium caused by buprenorphine patch use. However, the long recovery time (approximately 4 weeks), which far exceeded the expected time for drug elimination, in this case merits special attention. This extended period may be explained by a slow removal via lowered metabolic function as a result of lipophilic drug in association with advanced age.[7] Alternatively, rapid switching from tramadol to buprenorphine, as in the present case, may also have contributed to a delay in recovery because of the buprenorphine's high affinity to the endorphin receptor. This high affinity can cause buprenorphine to displace other opioids from receptors, thus triggering withdrawal in patients who are physically dependent on opioids.[8]

A buprenorphine patch is considered to be an excellent choice for patients with osteoarthrosis and other types of lumbago when pain continues despite adequate administrations of non-opioid analgesics. Delirium has not been reported in clinical trials of buprenorphine patch thus far. The present case shows that even topical opioids can precipitate seriously delirious states in elderly patients. Furthermore, when used as an analgesic, it is important to taper down gradually when neurological or other side-effects are observed. It should be also noted that, relative to other analgesics, greater attention is needed when patients switch opioids. Finally, opioids should be prescribed only with paramount caution, especially in elderly patients.