The analysis incorporated 27 cases documented across 24 articles. Among these cases, the mean age at the onset of the hallucinations was 58.3 years (range 21–88). Females accounted for the majority, with nineteen patients (67.9%).

Supplementary Table S1 outlines the demographic and clinical characteristics of the cohort under study. Table 1 shows background diseases and conditions. Approximately one-third of the patients (33.3%) reported a history of reduced hearing or hearing loss. Psychiatric disorders were prevalent among the cohort, with nine individuals (33.3%) presenting conditions such as bipolar disorder, anxiety, depression, panic disorder, and dissociative disorder. Cancers or leukemia were reported in seven patients (25.9%), while neurodegenerative diseases, specifically Parkinson’s and Lewy Body Dementia, were observed in three elderly individuals (11.1%). Notably three otherwise healthy individuals (11.1%) experienced musical hallucinations following ketamine intravenous infusion in a placebo-controlled experiment (Powers et al., 2015).

Some patients had multiple background diseases together, which is why the total percentage is >100% in Table 1. These findings provide a comprehensive overview of the diverse clinical backgrounds associated with drug-induced musical hallucinations in the examined cohort.

Table 2 delineates the identified trigger drugs in 27 cases suspected of inducing musical hallucinations. Among these cases, 18.5% were associated with antidepressants, while 11.1% were linked to each of the categories: anti-Parkinson drugs, opioids, ketamine, and voriconazole. Benzodiazepines, ceftazidime, and vasodilators accounted for a total of 22.2% (each 7.4%). One instance (3.7%) was attributed to alcohol withdrawal. The remaining 11.1% involved singular triggers, namely, amiodarone, corticosteroids, and aspirin.

Nineteen patients provided details regarding the onset time of musical hallucinations following the administration of trigger drugs, ranging from less than 75 min to 240 days. The mean initiation time was 21.4 days, with a median of 2 days. A considerable proportion of patients, sixteen (59.3%), noted the onset occurring in less than 10 days after the administration of the trigger drug.

Beyond musical hallucinations, six patients simultaneously encountered visual hallucinations, and one patient reported disssociation. These additional manifestations further contribute to the diverse perceptual experiences associated with possible drug-induced side effects in the examined cohort.

As illustrated in Table 3, the characteristics of the music or songs experienced during musical hallucinations exhibited a diverse range. Patients showed familiarity with the majority of the songs from their youth, although some songs were novel to them. The repertoire included various genres, including holiday songs, church hymns, and patriotic melodies. This heterogeneity underscores the broad spectrum of musical content encountered in the experiences of musical hallucinations within the studied cohort.

Results from electroencephalography (EEG) were available for 9 (33.3%) patients, all of which were within the normal range (Fisman, 1991; Vallada and Gentil, 1991; Gilbert, 1993; Curtin and Redmund, 2002; Kumagai et al., 2003; Agrawal and Sherman, 2004; Allen, 2008; Gondim Fde et al., 2010; Powers et al., 2015). EEG findings excluded active seizures in those nine patients.

Computed tomography (CT) scans were performed on 6 (22.2%) patients (Fisman, 1991; Vallada and Gentil, 1991; Tomar and Cheung, 2007; Gondim Fde et al., 2010; Powers et al., 2015; Song and Jung, 2019). Five patients yielded either normal or slight atrophy consistent with normal aging (Vallada and Gentil, 1991; Tomar and Cheung, 2007; Gondim Fde et al., 2010; Powers et al., 2015; Song and Jung, 2019). However, one patient exhibited an abnormal CT scan, revealing “slight prominence of the left temporal horn, and the right lateral ventricle was minimally larger than the left” (Fisman, 1991).

Magnetic resonance imaging (MRI) was conducted on 3 (11.1%) patients (Gilbert, 1993; Kobayashi et al., 2004; Kataoka and Ueno, 2014), with only one patient displaying abnormal results (Kobayashi et al., 2004). The abnormal MRI findings included “generalized cortical atrophy with several small lesions in the deep white matter of the cerebral hemispheres” (Kobayashi et al., 2004).

These diagnostic images excluded brain tumors, and acute ischemic or hemorrhagic strokes.

As shown in Table 4, in response to the occurrence of musical hallucinations, various interventions were implemented: 1) The trigger drugs were terminated; 2) The dosage of trigger drugs was reduced; 3) The administration route was altered from intravenous (IV) to oral; 4) Extended-release pramipexole was switched back to immediate-release pramipexole in one case; 5) The trigger drugs were changed to another corticosteroid, sedative, or anti-Parkinson drug; 6) Besides termination of trigger drugs, sedatives, tricyclic antidepressants, or atypical antipsychotics were added; 7) Benzodiazepines were used to treat alcohol withdrawal in one case.

Twenty-four (88.9%) patients reported complete resolution of hallucination after the initiation of the above treatments. If the alcohol withdrawal case is not counted, then 23/27 (85.2%) fully recovered after modulating trigger drugs with or without additional pharmacotherapy. Among these 24 patients, only 21 patients specified when the musical hallucination stopped. The therapy for musical hallucination was effective within an average of 5.6 days (range 75 min–21 days). The total duration of musical hallucination in those 21 patients extended from 75 min to 23 days, with a mean duration of 8.7 days. Three patients whose musical hallucinations disappeared did not define the time.

As shown in Table 4, more than half of musical hallucinations completely disappeared by simple termination of trigger drugs (40.7%) and dose reduction (14.8%). Changing administration routes or formulas or switching to similar drugs resulted in further musical hallucination resolution (14.8%). If musical hallucinations persisted after stopping the trigger drugs, additional medications, such as risperidone, clonazepam, sultopride, and amitriptyline, led to symptom resolution (11.1%). A benzodiazepine (clonazepam) successfully controlled the musical hallucinations in a patient suffering from alcohol withdrawal (3.4%).

In the only case of aspirin-induced musical hallucination, the patient was an elderly female with hearing loss, wore bilateral hearing aids, and took a high dose (12 tablets of 300 mg) of aspirin daily for arthritis with a blood salicylate level higher than the therapeutic range (Allen, 2008). When the aspirin dose was reduced by half, her musical hallucination stopped a few days later (Allen, 2008). This supports that the overdose of aspirin caused musical hallucinations in this case.

Unfortunately, musical hallucinations were reduced but continued in 3 patients (11.1%) for as long as 1 year. Those three patients all had background diseases: depression + anxiety (1 patient) (Curtin and Redmund, 2002); depression + anxiety + panic attack + reduced hearing (1 patient) (Vallada and Gentil, 1991); bilateral hearing loss + chronic otitis media (1 patient) (Song and Jung, 2019).

These findings provide insights into the varied temporal aspects and outcomes associated with drug-induced musical hallucinations within the analyzed cases.

Among our case series focusing on drug-induced musical hallucinations, 33.3% of patients exhibited hearing loss or impairment, 33.3% had mental disorders, 25.9% presented with cancers, 11.1% with neurodegenerative disorders, and 7.4% with chronic otitis. Some patients had multiple medical conditions concurrently. Many drugs were suspected as trigger drugs, such as antidepressants, anti-Parkinson’s, opioids, etc. The exact pathophysiology remains elusive, however, we propose a multifactorial model with an imbalance of neurotransmitters/receptor activities contributing to the development of drug-induced musical hallucinations, as shown in Figure 2.

The diagnosis relies on the exclusion of alternative causes. Given that structural changes in the brain, such as tumors, ischemic strokes, and hemorrhages, account for 9% of patients with musical hallucinations (Golden and Josephs, 2015), brain imaging, including CT or MRI scans, is recommended. It is essential to rule out brain infections as well. Once structural changes or infections are excluded, suspicion of drug-induced musical hallucination arises when there is a temporal correlation between the onset of musical hallucination and alterations in the treatment regimen for the underlying disease.

The cessation or modification of trigger drugs, through dose reduction, switching to different routes or formulas, or switching to similar drugs, typically results in the resolution of musical hallucinations in most patients in less than 1 month. If the symptoms persist, atypical antipsychotics, tricyclic antidepressants, and benzodiazepines have proven effective. The rapid resolution of musical hallucinations after those treatments further supports the diagnosis of drug-induced musical hallucinations.

In our case series, three patients’ musical hallucinations continued for 6 months or longer than 1 year after the above-mentioned therapy. Those three patients had hearing loss, ottis, or psychiatric disorders. Because of the ineffectiveness of modulating the suspected trigger drugs, the diagnosis of drug-induced musical hallucination is in doubt. In these three patients, the underlying diseases may contribute dominantly to the musical hallucination occurrence and persistence.

To the best of our knowledge, the current study represents the initial comprehensive review of drug-induced musical hallucinations. However, there are inherent limitations to be acknowledged. As highlighted by Coebergh et al. (2015), none of the suspected trigger drugs have been investigated with placebo-controlled trials for musical halucination, raising concerns about the reliability of predicting the causative effects of trigger drugs. The absence of placebo-controlled information across all articles limits the robustness of the findings in our case series. Additionally, a low incidence rate of drug-induced musical hallucinations poses challenges in calculating prevalence, while the observational nature of the studies inhibits the generation of correlations. The inclusion criterion of articles with detailed patient information may introduce a potential bias, excluding other trigger drugs that were not published or lacked detailed information. Furthermore, the absence of experimental evidence supporting our hypothesis on the mechanism of drug-induced musical hallucination underscores the need for further research in this area.

In summary, the primary suspected trigger drugs for inducing musical hallucinations encompass antidepressants, antiparkinson drugs, opioids, ketamine, voriconazole, ceftazidime, and benzodiazepines. The diagnostic approach depends on exclusion. Patients with underlying conditions such as hearing impairment, psychiatric disorders, neurodegenerative diseases, or cancers exhibit a heightened susceptibility to drug-induced musical hallucinations. Imbalances in neurotransmitters or their receptor activities induced by trigger drugs are posited as contributing factors. Treatment strategies involve discontinuing or modifying trigger drugs, with or without additional pharmacotherapy. There is a pressing need for further research to identify possible trigger drugs and to delve into the underlying mechanism of drug-induced musical hallucinations.