乌鸦传媒

Older patients prescribed an atypical antipsychotic for hospitalization-related delirium more often stayed on it past the recommended duration than if prescribed a typical antipsychotic, researchers reported.

Among nearly 6,000 patients 65 and older without prior psychiatric disorders, 13.5% were newly prescribed haloperidol and 86.5% initiated an atypical antipsychotic to treat delirium around an infection-related hospitalization, according to Kueiyu Joshua Lin, MD, of Brigham and Women's Hospital in Boston, and colleagues.

But treatment stopped within 30 days of initiation for more than half of haloperidol users compared with only about one in 10 of those taking an atypical antipsychotic (52.1% vs 11.4%, P<0.001).

By 180 days later, 93.7% of haloperidol users and 76.3% of atypical antipsychotic users discontinued the new meds.

"This is contrary to clinical recommendations to discontinue both types of antipsychotic medications [APMs] as soon as the delirious state or acute behavioral disturbance has resolved," wrote the researchers in .

"Clinicians may be reluctant to actively discontinue the ongoing treatment after the patient's condition is stabilized, which may explain the low rates of APM discontinuation after delirium onset," they added.

Lin's group also took note of some long-term trends over the study period from 2004 to 2022. During this timeframe, discontinuation rates for haloperidol users increased by 5% (95% CI 3%-7%) each year. However, discontinuation rates for atypical antipsychotics held steady from 2004 through 2022.

Rates of medication discontinuation were similar across all the atypical antipsychotics prescribed in the cohort: aripiprazole (Abilify), risperidone (Risperdal), quetiapine (Seroquel), and olanzapine (Zyprexa).

Older patients who were hospitalized longer than 30 days were significantly less likely to discontinue use compared with those who had an inpatient stay under 7 days long for either type of antipsychotic (adjusted HR 0.61 for haloperidol, 95% CI 0.45-0.84, and 0.86 for atypical antipsychotics, 95% CI 0.77-0.97).

Patients with dementia were also less likely to discontinue haloperidol or atypical antipsychotics than those without it (aHR 0.71, 95% CI 0.58-0.87, and aHR 0.80, 95% CI 0.74-0.86, respectively). Patients with diabetes were also less likely to discontinue haloperidol, while those with a history of stroke or ischemic heart disease were less likely to stop taking atypical antipsychotics.

On the other hand, patients with pneumonia, heart failure, or gastrointestinal bleeding were more likely to quit haloperidol, while there was a higher rate of atypical antipsychotic discontinuation among those with pneumonia, cancer, or osteomyelitis/septic arthritis.

The researchers said these findings underscored the need for more "proactive interventions to facilitate discontinuation of these potentially inappropriate medications in older adults."

Listing off a few suggestions, Lin's group highlighted the benefits of both patient and clinician education, use of tapering or deprescribing plans guided by healthcare professionals, and utilizing monitoring protocols.

"Specific strategies to support behavioral change include adding visual cues to the environment (e.g., deprescribing algorithms, medication checklists, etc) and building prompts into routine workflow (e.g., electronic health record alerts, reminder letters or messages, etc)," they recommended.

It's important to discontinue antipsychotics if no longer medically necessary, Lin's group explained, as both older and second-generation classes carry specific risks. They explained that although some studies have indicated a higher risk of death with typical antipsychotics, other data have suggested antipsychotic medications (APMs) carry just as high risks for serious adverse events like death and cardiac arrhythmias.

All antipsychotics currently carry a boxed warning on their labels about the risk of mortality in elderly patients treated for dementia-related psychosis.

"Given multiple serious adverse reactions associated with APM use, our findings call for effective interventions to proactively discontinue APMs when they are no longer indicated," urged the researchers.

The average age in the retrospective cohort study was 81.5 among the 790 new haloperidol users and 79.8 among the 5,045 new atypical antipsychotic users. Data came from Optum's U.S.-based, deidentified Clinformatics Data Mart database of large commercial and Medicare Advantage health plans. All patients were 65 and older without a prior psychiatric disorder and dispensed a prescription for a new oral antipsychotic within 30 days after discharge from an infection-related hospitalization.

The database didn't have data on inpatient prescribing or on whether the post-discharge use was on an as-needed basis, which were limitations of the study.

"A prior study showed that new initiation of APMs in a hospital setting is a good proxy for the presence of delirium," the researchers noted. "Although delirium is undercoded in administrative databases, it constitutes a majority of indications for APM initiation in the hospital."

Some 30% of patients initiating an APM during hospitalization are discharged on one, and the discharging clinician typically prescribes it to ensure that patients have an adequate medication supply until the first postdischarge follow-up appointment, the researchers noted.

The study excluded patients with prior use of or chronic indications for APMs, including schizophrenia and other psychotic disorders, bipolar disorder, and depression at any time before cohort entry to reduce confounding.

The most common infection during hospitalization was urinary tract, followed by pneumonia, soft tissue, bacteremia, and COVID-19. More than half had dementia at baseline.

Haloperidol users tended to be older and were more likely to have bacteremia, cancer, heart failure, gastrointestinal bleeding, anemia, liver disease, or end-stage kidney disease. They were also less likely to be of Black or other race or ethnicity, be mildly frail, or have dementia or COVID-19.

Comment