antipsychotics in nursing homes

this is the clearest case i know of a warning that did not change practice and a measurement that did. it is also a case of what happens next when a number becomes a target. every figure here comes from the FDA, the HHS inspector general, CMS or the department of justice.

if someone you love is on one of these drugs, this page is not a reason to stop it. sometimes an antipsychotic is the least bad option for severe distress or danger, and stopping suddenly carries its own risks. it is a reason to ask what it is for, what was tried first, and when it will be reviewed.

the timeline

1. April 2005

   the FDA warns that antipsychotics raise the risk of death in elderly people with dementia about 1.6 to 1.7 times, from 17 trials in 5,106 patients. a boxed warning follows.^[1]

2. 2002 to 2005

   a drugmaker, the government later alleges, builds a sales force that calls “almost exclusively on nursing homes” to promote its antipsychotic for dementia-related psychosis. settled in 2007, $515 million across several allegations.^[3]

3. first half of 2007

   two years after the warning: 14 percent of elderly nursing home residents have Medicare antipsychotic claims. 88 percent of those claims are for residents with dementia, the condition the box warns about. half the claims are erroneous, $116 million.^[2]

4. 2009 and 2013

   two more drugmakers plead guilty to misdemeanors over promoting their antipsychotics, among other things, for elderly dementia patients: $1.415 billion and $2.2 billion.^[4]

5. 2012

   responding to the inspector general’s report, CMS launches a national partnership in early 2012 and, from july, publishes each home’s antipsychotic rate. the baseline, late 2011: 23.9 percent of long-stay residents.^[6]

6. 2015 and 2020

   the measure works: 17.4 percent in 2015, 14.1 percent by mid-2020, a 41 percent drop.^[7]

7. 2018

   the loophole: residents with schizophrenia are excluded from the measure. of 98,227 older residents reported with schizophrenia, 30 percent have no record of that diagnosis anywhere in their medicare claims.^[7]

8. January 2023

   CMS starts auditing schizophrenia coding and cutting star ratings for homes that miscode.^[8]

the warning that did not move the number

in april 2005 the FDA pooled 17 placebo-controlled trials of four antipsychotics in 5,106 elderly patients with dementia. fifteen showed more deaths on the drug, and several analyses put the increase at about 1.6 to 1.7 times.^[1] two years later, in the first half of 2007, the inspector general found that 304,983 elderly nursing home residents, 14 percent, had medicare claims for these drugs, costing $309 million in six months. 88 percent of the claims were for residents with dementia, the condition in the boxed warning, and 51 percent were erroneous under medicare’s own rules, $116 million.^[2]

the other direction, in the inspector general’s own words: “physicians are not prohibited from prescribing drugs for off-label conditions or in the presence of the condition(s) specified in the FDA boxed warning.” erroneous means the use was not supported by the reference books or not documented; it is not a finding that any resident was harmed.^[2]

the settlements

three drugmakers paid over how these drugs reached people with dementia. bristol-myers squibb settled civil allegations in 2007 that it promoted abilify for dementia-related psychosis and built a long-term-care sales force that “called almost exclusively on nursing homes”; its $515 million also covered unrelated pricing allegations.^[3] eli lilly pleaded guilty in 2009 and paid $1.415 billion over zyprexa promotion that included elderly dementia patients.^[4] janssen, part of johnson & johnson, pleaded guilty in 2013 as part of a $2.2 billion resolution over risperdal promotion to the same group.^[5] the full records, both directions, are on the aripiprazole, olanzapine and risperidone pages, and every federal resolution is on what the fraud settlements add up to.

the measure that did

in early 2012, in response to the inspector general’s report, CMS launched a national partnership to cut antipsychotic use in dementia care, and from july 2012 it published each home’s rate. at the end of 2011, 23.9 percent of long-stay residents were on an antipsychotic. by the third quarter of 2015 it was 17.4 percent.^[6] by mid-2020, 14.1 percent, a 41 percent drop, with every state lower.^[7] a warning printed on a label had not done that. a public number with a facility’s name on it did.

and then the loophole

the measure excludes residents with schizophrenia, huntington’s disease or tourette syndrome, because for them an antipsychotic is standard treatment. in 2018 nursing homes reported 98,227 residents aged 65 and older as having schizophrenia. 29,617 of them, 30 percent, had no record of that diagnosis in any medicare claim across two years, and 71 percent of those had an antipsychotic prescription. for-profit homes, 69 percent of all homes, accounted for 78 percent of the inconsistent entries.^[7]

the inspector general is careful, and so are we: some of those residents may truly have schizophrenia and simply not be getting care for it, which is its own problem.^[7] CMS was less careful. in 2023 it wrote that some homes “have erroneously coded residents as having schizophrenia, which can mask the facilities’ true rate of antipsychotic medication use,” leaving residents “subject to poor care and unnecessary antipsychotic medications,” and began auditing the coding and cutting star ratings.^[8] ownership by private equity, for what it is worth, showed no effect on antipsychotic use in the one study that measured it.^[9] the full ownership picture is on private equity in mental health care.

what it adds up to

a warning on a label changed little. a public, named measurement changed a lot. and once the number mattered, the exemption became the place to hide. that is not a nursing-home story. it is what happens to any number that gets attached to money or reputation, and it is why a measure is only as honest as the audit behind its exceptions.

questions people ask

sources

1. US Food and Drug Administration. Public Health Advisory: Deaths with Antipsychotics in Elderly Patients with Behavioral Disturbances, 11 April 2005 (archived copy). “Of a total of seventeen placebo controlled trials performed with olanzapine (Zyprexa), aripiprazole (Abilify), risperidone (Risperdal), or quetiapine (Seroquel) in elderly demented patients with behavioral disorders, fifteen showed numerical increases in mortality … These studies enrolled a total of 5106 patients, and several analyses have demonstrated an approximately 1.6-1.7 fold increase in mortality”; most deaths “due to heart related events … or infections (mostly pneumonia).” The boxed warning followed. https://web.archive.org/web/2016/http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm053171.htm
2. HHS Office of Inspector General. Medicare Atypical Antipsychotic Drug Claims for Elderly Nursing Home Residents, OEI-07-08-00150, May 2011. January to June 2007: “Of the 2.1 million elderly nursing home residents, 304,983 had at least 1 Medicare claim for an atypical antipsychotic drug”; “Claims for these residents amounted to $309 million”; 83 percent of claims were for off-label conditions and “Eighty-eight percent (1,263,641) of the drug claims were for residents diagnosed with dementia (the condition specified in the FDA boxed warning)”; “Fifty-one percent … were erroneous, amounting to $116 million.” Also: “Physicians are not prohibited from prescribing drugs for off-label conditions or in the presence of the condition(s) specified in the FDA boxed warning.” https://oig.hhs.gov/oei/reports/oei-07-08-00150.pdf
3. US Department of Justice. Bristol-Myers Squibb to Pay More Than $515 Million to Resolve Allegations of Illegal Drug Marketing and Pricing, #07-782, 28 September 2007. The government alleged that from 2002 through 2005 the company promoted Abilify “for pediatric use and to treat dementia-related psychosis, both off-label uses,” and “created a specialized long term care sales force that called almost exclusively on nursing homes, where dementia-related psychosis is far more prevalent than schizophrenia or bipolar disorder.” The $515 million also resolved unrelated pricing and kickback allegations; “over $25 million constitutes disgorgement of profits … resulting from BMSs illegal promotion of Abilify.” Civil allegations. https://www.justice.gov/archive/opa/pr/2007/September/07_civ_782.html
4. US Department of Justice. Eli Lilly and Company Agrees to Pay $1.415 Billion to Resolve Allegations of Off-label Promotion of Zyprexa, 15 January 2009. Guilty plea to a misdemeanor for promoting Zyprexa for uses including dementia in elderly patients. Full record on our olanzapine page. https://www.justice.gov/archives/opa/pr/eli-lilly-and-company-agrees-pay-1415-billion-resolve-allegations-label-promotion-zyprexa
5. US Department of Justice. Johnson & Johnson to Pay More Than $2.2 Billion to Resolve Criminal and Civil Investigations, 4 November 2013. Janssen pleaded guilty to a misdemeanor for promoting Risperdal for elderly dementia patients. Full record on our risperidone page. https://www.justice.gov/archives/opa/pr/johnson-johnson-pay-more-22-billion-resolve-criminal-and-civil-investigations
6. Centers for Medicare & Medicaid Services. S&C 16-28-NH, National Partnership to Improve Dementia Care in Nursing Homes: update, 3 June 2016. The partnership was established “in early 2012, in response to a 2011 Office of the Inspector General (OIG) report”; public reporting of each home’s antipsychotic measures began in July 2012. “In 2011 Quarter 4, 23.9 percent of long-stay nursing home residents were receiving an antipsychotic medication; since then there has been a decrease of 27 percent to a national prevalence of 17.4 percent in 2015 Quarter 3.” (Another sentence in the same letter gives 23.8 percent for end-2011.) https://www.cms.gov/medicare/provider-enrollment-and-certification/surveycertificationgeninfo/downloads/sc-letter-16-28-partnership-update-report.pdf
7. HHS Office of Inspector General. CMS Could Improve the Data It Uses To Monitor Antipsychotic Drugs in Nursing Homes, OEI-07-19-00490, May 2021. “there were 98,227 residents age 65 and older whom nursing homes reported as having schizophrenia. Of those, 29,617 residents (30 percent) had no record of a schizophrenia diagnosis in any of their 2017 and 2018 Medicare Part A or B claims or Part C encounter data”; 71 percent of them had a Part D antipsychotic claim; “for-profit nursing homes account for 78 percent of the MDS entries that are inconsistent in this way” while 69 percent of nursing homes are for-profit; 12,091 long-stay residents (5 percent) had an antipsychotic claim but were not reported as receiving one. CMS, in the report: “Between 2011 and the second quarter of 2020, the national prevalence of antipsychotic medication use among long-stay nursing home residents was reduced by 41 percent to 14.1 percent nationwide.” OIG notes some of these residents may truly have schizophrenia and simply not be receiving services for it. https://oig.hhs.gov/oei/reports/OEI-07-19-00490.pdf
8. Centers for Medicare & Medicaid Services. QSO-23-05-NH, Adjusting Quality Measure Ratings Based on Erroneous Schizophrenia Coding, 18 January 2023. “CMS is concerned that some nursing homes have erroneously coded residents as having schizophrenia, which can mask the facilities’ true rate of antipsychotic medication use”; “When nursing home residents are given erroneous schizophrenia diagnoses, they are subject to poor care and unnecessary antipsychotic medications, both of which can be very dangerous.” Facilities found miscoding lose quality-measure stars. https://www.cms.gov/files/document/qso-23-05-nh-adjusting-quality-measure-ratings-based-erroneous-schizophrenia-coding-and-posting.pdf
9. Braun RT, et al. Association of private equity investment in US nursing homes with the quality and cost of care for long-stay residents. JAMA Health Forum 2021;2(11):e213817. “Private equity acquisition was not significantly associated with antipsychotic use (−0.2 percentage points; 95% CI, −1.7 to 1.4 percentage points; P = .83).” PMC8796926. https://pmc.ncbi.nlm.nih.gov/articles/PMC8796926/

adam