Category: Medication and dementia care

A quarter to half of long-term care residents carry bacteria in the urine without infection. Confusion alone is explicitly not an indication to treat. Here is the evidence, and a correction.

This piece exists partly to correct something CareGuard used to publish. The correction is in the box below, and the evidence behind it is worth reading whichever way you came to the page.

The belief

Almost everyone who works in long-term care has been taught some version of it: an older resident who becomes suddenly confused probably has a urinary tract infection, so send a urine sample, and start an antibiotic while you wait. It is well meant, it is nearly universal, and it is the single most common route to an antibiotic in a nursing home.

Why it does not work

It fails on the arithmetic before it fails on anything else. If a quarter to half of the residents in a building would culture positive today with no infection at all, then a positive culture in a confused resident is not evidence of much. It is the base rate.

That is not a contrarian reading of the literature. It is the guideline position, stated as a recommendation, with the exact clinical scenarios named — confusion, and a fall.

What to do instead

• Assess for other causes. That is the guideline’s own phrase. Pain, constipation, dehydration, a new medication, a missed dose, a change in the room, a change in staff.
• Look for localizing urinary symptoms or systemic signs. Fever, hemodynamic instability, dysuria, suprapubic or flank pain. Those change the picture; a culture on its own does not.
• Observe carefully rather than treat reflexively. Also the guideline’s phrase.
• Count what is happening in your building. A facility that cannot say how many of its urine cultures last quarter were ordered on a behavior change alone is not in a position to know whether it has this problem.

Why a patient safety organization cares

Because this is a medication-burden problem wearing an infection-control costume. Every unnecessary course is an anticholinergic or an interaction or a Clostridioides difficile risk added to a resident who is already on nine drugs, and it is the same argument as the rest of the deprescribing work. DWARAA · Deprescribing in long-term care.

It is also a fall and behavior problem, because a resident whose confusion is attributed to a urine culture is a resident whose actual cause went unlooked-for. What restraint-free care actually looks like.

---

What the evidence actually shows

Bacteria in the urine without symptoms — asymptomatic bacteriuria — is not an early urinary tract infection. In long-term care it is a common, stable, colonized state. The 2019 Infectious Diseases Society of America guideline puts prevalence in long-term care residents at 25% to 50% of women and 15% to 50% of men.

At that prevalence a positive urine culture in a resident who is confused tells you very little. Half the building would culture positive on a given day.

The guideline is explicit, and it is explicit about the cases facilities actually face

• Do not screen or treat in long-term care residents. “In older persons resident in long-term care facilities, we recommend against screening for or treating ASB.”
• Confusion alone is not an indication. “In older patients with functional and/or cognitive impairment with bacteriuria and delirium (acute mental status change, confusion) and without local genitourinary symptoms or other systemic signs of infection (eg, fever or hemodynamic instability), we recommend assessment for other causes and careful observation rather than antimicrobial treatment.”
• A fall alone is not an indication either. The same recommendation is made, in the same terms, for a resident with bacteriuria who has fallen without local genitourinary symptoms or systemic signs of infection.

Treating it does not help, and it does harm

The randomized evidence goes back forty years and is consistent. A trial of 50 institutionalized women (mean age 83) followed for a year found no difference in genitourinary morbidity or mortality between treated and untreated residents, while antimicrobial therapy was associated with more reinfection (1.67 versus 0.87 per patient-year), far more adverse drug effects (0.51 versus 0.046 per patient-year) and increasingly resistant organisms.

A controlled trial in older ambulatory women found that antibiotics reduced positive cultures but symptoms of urinary tract infection were more common in the treated group, concluding that “treatment of asymptomatic bacteriuria in older women is contraindicated.” A cohort of 1,491 women with a nested trial found bacteriuria was not an independent risk factor for mortality (relative risk 1.1) and that treating it did not lower the mortality rate.

The IDSA guideline summarizes the harms: “There are high-quality data to suggest that adverse effects are particularly common following the use of antimicrobials in this population, including CDI and isolation of organisms with increased antimicrobial resistance.”

Does confusion indicate infection at all?

A systematic review found only five primary studies on the question, none randomized. Among people with delirium, urinary tract infection rates ranged from 25.9% to 32%, against 13% in those without delirium; among people with a urinary tract infection, delirium rates ranged from 30% to 35% against 7.7% to 8%. The review’s conclusion was that “all of them had significant methodological flaws that likely led to biased results. Therefore, it is difficult to ascertain the degree to which urinary tract infections cause delirium.” An association exists; a causal, treat-on-suspicion rule does not follow from it.

What this looks like in a real building

A prospective study of 444 urinary tract infection episodes across 134 nursing homes found confusion was a reported diagnostic criterion in 28% of cases, that 21% of diagnoses were based on erroneous criteria, and that on reclassification against guidelines only 10% of diagnoses were consistent — with 21% turning out to be asymptomatic bacteriuria. Empiric antibiotic therapy matched the diagnosis in 16% of cases.

That is what a safety review is looking for: not whether urinary tract infections are being caught, but whether urine is being cultured reflexively and antibiotics started on a criterion the guideline explicitly rejects.

References

1. Nicolle LE, Gupta K, Bradley SF, et al. 2019. Clinical practice guideline for the management of asymptomatic bacteriuria: 2019 update by the Infectious Diseases Society of America. Clinical Infectious Diseases 68(10):e83-e110. [clinical practice guideline] · PMID 30895288 · doi:10.1093/cid/ciy1121
2. Nicolle LE, Mayhew WJ, Bryan L 1987. Prospective randomized comparison of therapy and no therapy for asymptomatic bacteriuria in institutionalized elderly women. American Journal of Medicine 83(1):27-33. [randomized controlled trial, 50 residents] · PMID 3300325 · doi:10.1016/0002-9343(87)90493-1
3. Abrutyn E, Berlin J, Mossey J, et al. 1996. Does treatment of asymptomatic bacteriuria in older ambulatory women reduce subsequent symptoms of urinary tract infection?. Journal of the American Geriatrics Society 44(3):293-5. [controlled clinical trial] · PMID 8600199 · doi:10.1111/j.1532-5415.1996.tb00917.x
4. Abrutyn E, Mossey J, Berlin JA, et al. 1994. Does asymptomatic bacteriuria predict mortality and does antimicrobial treatment reduce mortality in elderly ambulatory women?. Annals of Internal Medicine 120(10):827-33. [cohort plus controlled trial] · PMID 7818631 · doi:10.7326/0003-4819-120-10-199405150-00003
5. Balogun SA, Philbrick JT 2013. Delirium, a symptom of UTI in the elderly: fact or fable? A systematic review. Canadian Geriatrics Journal 17(1):22-6. [systematic review, 5 studies, none randomized] · PMID 24596591 · doi:10.5770/cgj.17.90
6. Lemoine L, Dupont C, Capron A, et al. 2018. Prospective evaluation of the management of urinary tract infections in 134 French nursing homes. Médecine et Maladies Infectieuses 48(5):359-364. [prospective observational study, 444 episodes] · PMID 29747905 · doi:10.1016/j.medmal.2018.04.387
7. Wu JH, Langford BJ, Daneman N, Friedrich JO, Garber G 2019. Antimicrobial stewardship programs in long-term care settings: a meta-analysis and systematic review. Journal of the American Geriatrics Society 67(2):392-399. [systematic review and meta-analysis, 18 studies] · PMID 30517765 · doi:10.1111/jgs.15675

The 2025 Cochrane review pooled 104 trials and 68,964 residents. Exercise works while it is sustained; tailored, staff-engaged programs work; medication review alone does not.

Falls are the area of long-term care with the largest current randomized evidence base, which means there is very little excuse for guessing. There is also a widely repeated claim in this field that the evidence does not support, and it is one CareGuard has an interest in — so it gets stated plainly below.

Falls are the outcome that reorganizes a resident’s life

More than 25% of adults aged 65 and over fall each year, and falls are the leading cause of injury-related death in that age group. In long-term care the population is older, frailer and more cognitively impaired than the community samples those figures come from — one analysis of 282,518 assessments covering 75,132 residents found unstable health, unsteady gait, wandering and moderate to severe dependency in daily activities were the strongest risk factors, with cognitive impairment carrying a hazard ratio between 1.22 and 1.37.

What a facility can actually change

The short version, before the numbers: sustained exercise, a tailored program that staff are part of, vitamin D where levels are low, hip protectors for fracture risk, and the environmental work the safety review walks. The fall prevention items in the review.

And the claim that does not hold

That reviewing and reducing medication, on its own, reduces falls. It is intuitive, it is what a deprescribing-minded organization would like to be true, and three independent syntheses say it is not. The full numbers are below. The practical consequence is that medication review belongs inside a multifactorial program rather than instead of one — which is also what the strongest positive finding in the whole review points to.

The resourcing question nobody wants

A systematic review of complex falls-prevention interventions in residential aged care found no overall effect across seven pooled studies (mean difference −1.29, 95% CI −3.01 to 0.43) — but a sensitivity analysis found that interventions “delivered with additional resources at multiple levels had a significant positive effect in reducing fall rates” (mean difference −2.26, 95% CI −3.72 to −0.80).

That is the finding an off-site owner should read twice, because it is the one that costs money and is invisible from a monthly report. The visibility gap.

---

What the evidence says reduces falls in a care facility

This is one of the few areas of long-term care with a large, current randomized evidence base, and it is worth reading it accurately rather than generically. The 2025 Cochrane review of falls prevention in care facilities pooled 104 randomized trials and 68,964 residents.

• Exercise. Active exercise probably reduces the rate of falls, rate ratio 0.68 (95% CI 0.51 to 0.91; 14 trials, 2,215 participants, moderate-certainty evidence). The same review reports high-certainty evidence that if the exercise is not sustained there is no effect (RaR 1.02, 95% CI 0.78 to 1.32). It is a program, not a course.
• Multifactorial programs, done properly. Multifactorial interventions overall showed little or no effect on fall rate (RaR 0.87, 95% CI 0.68 to 1.12). But those that were tailored to the individual resident and engaged staff probably produce a large reduction: RaR 0.61 (95% CI 0.54 to 0.69; 7 trials, 3,553 participants, moderate certainty). The difference between those two numbers is the difference between having a falls policy and running one.
• Vitamin D probably reduces the rate of falls, RaR 0.63 (95% CI 0.46 to 0.86; 5 trials, 4,603 participants) — but the review notes the population in those trials had low vitamin D levels.
• Hip protectors in nursing and residential care produce a small reduction in hip fracture risk, RR 0.82 (95% CI 0.67 to 1.00), an absolute effect of about 11 fewer hip fractures per 1,000 people (14 studies, 11,808 participants, moderate quality).

Which drug classes the association actually implicates

The association between medications and falls is nonetheless large and well characterized. Three companion meta-analyses covering 131, 248 and 281 studies report adjusted odds ratios for a fall of 2.02 for SSRIs (95% CI 1.85 to 2.20), 1.81 for long-acting benzodiazepines (1.05 to 3.16), 1.60 for opioids (1.35 to 1.91), 1.57 for antidepressants overall (1.43 to 1.74), 1.55 for antiepileptics (1.25 to 1.92), 1.54 for antipsychotics (1.28 to 1.85), 1.42 for benzodiazepines overall (1.22 to 1.65), 1.36 for loop diuretics (1.17 to 1.57), and 1.75 for polypharmacy itself (1.27 to 2.41).

In one Japanese cohort of 459 residents (mean age 87) with 645 falls over a year, medication was judged to have influenced around three-quarters of all falls, and more than 80% of those involved psychotropic drugs.

References

1. Dyer SM, et al. 2025. Interventions for preventing falls in older people in care facilities. Cochrane Database of Systematic Reviews. [Cochrane review, 104 RCTs, 68,964 residents] · PMID 40832852 · doi:10.1002/14651858.CD016064
2. Cameron ID, et al. 2018. Interventions for preventing falls in older people in care facilities and hospitals. Cochrane Database of Systematic Reviews (9):CD005465. [Cochrane review, 95 trials, 138,164 participants] · PMID 30191554 · doi:10.1002/14651858.CD005465.pub4
3. Colón-Emeric CS, McDermott CL, Lee DS, Berry SD 2024. Risk assessment and prevention of falls in older community-dwelling adults: a review. JAMA 331(16):1397-1406. [review of meta-analyses; community-dwelling, not long-term care] · PMID 38536167 · doi:10.1001/jama.2024.1416
4. Francis-Coad J, et al. 2018. Effectiveness of complex falls prevention interventions in residential aged care settings: a systematic review. JBI Database of Systematic Reviews and Implementation Reports 16(4):973-1002. [systematic review and meta-analysis] · PMID 29634516 · doi:10.11124/JBISRIR-2017-003485
5. Seppala LJ, et al. 2022. Medication reviews and deprescribing as a single intervention in falls prevention: a systematic review and meta-analysis. Age and Ageing 51(9):afac191. [systematic review and meta-analysis, 49 RCTs] · PMID 36153749 · doi:10.1093/ageing/afac191
6. Lee J, et al. 2021. Deprescribing fall-risk increasing drugs (FRIDs) for the prevention of falls and fall-related complications: a systematic review and meta-analysis. BMJ Open 11(2):e035978. [systematic review and meta-analysis, 5 RCTs] · PMID 33568364 · doi:10.1136/bmjopen-2019-035978
7. de Vries M, et al. 2018. Fall-risk-increasing drugs: a systematic review and meta-analysis. I. Cardiovascular drugs. Journal of the American Medical Directors Association 19(4):371.e1-371.e9. [meta-analysis, 131 studies] · PMID 29396189 · doi:10.1016/j.jamda.2017.12.013
8. Seppala LJ, et al. 2018. Fall-risk-increasing drugs: a systematic review and meta-analysis. II. Psychotropics. Journal of the American Medical Directors Association 19(4):371.e11-371.e17. [meta-analysis, 248 studies] · PMID 29402652 · doi:10.1016/j.jamda.2017.12.098
9. Seppala LJ, et al. 2018. Fall-risk-increasing drugs: a systematic review and meta-analysis. III. Others. Journal of the American Medical Directors Association 19(4):372.e1-372.e8. [meta-analysis, 281 studies] · PMID 29402646 · doi:10.1016/j.jamda.2017.12.099
10. Oya N, et al. 2022. Over half of falls were associated with psychotropic medication use in four nursing homes in Japan. International Journal of Environmental Research and Public Health 19(5):3123. [retrospective cohort, 459 residents] · PMID 35270813 · doi:10.3390/ijerph19053123
11. Santesso N, et al. 2014. Hip protectors for preventing hip fractures in older people. Cochrane Database of Systematic Reviews (3):CD001255. [Cochrane review, 19 studies] · PMID 24687239 · doi:10.1002/14651858.CD001255.pub5
12. Cheung G, et al. 2024. Falls risk in long-term care residents with cognitive impairment: effects of COVID-19 pandemic. Journal of the American Medical Directors Association 25(1):177-182. [observational, 75,132 residents] · PMID 38104633 · doi:10.1016/j.jamda.2023.11.006
13. Montero-Odasso M, et al. 2022. World guidelines for falls prevention and management for older adults: a global initiative. Age and Ageing 51(9). [consensus guideline] · PMID 36178003 · doi:10.1093/ageing/afac205

Statin use is associated with new-onset type 2 diabetes. It is on the FDA label. What that means for an older adult, and what it does not mean.

This is a documented, labeled association, and it is a trade-off rather than a prohibition. Both halves of that sentence matter.

What is established

• The FDA added a class-wide safety label change on February 28, 2012 warning of increased HbA1c and fasting glucose with statin use.
• Sattar et al., The Lancet, 2010 — a collaborative meta-analysis of randomized statin trials found roughly a 9% increase in incident diabetes.
• The Lancet Diabetes & Endocrinology, 2024 — an individual-participant-data meta-analysis confirming new-onset diabetes and worsening glycemia.

Where the risk concentrates

It is higher with intensive dosing, and it concentrates in people already near the diabetes threshold — which in a long-term care population is a large share of residents. Someone with a fasting glucose in the prediabetic range on a high-intensity statin is the person this actually describes.

What it is not

It is not a reason to stop a statin. The American College of Cardiology’s position is that the cardiovascular benefit still outweighs this risk for patients at meaningful cardiovascular risk. Presenting a labeled side effect as a reason to discontinue a drug that is preventing strokes would be a worse error than not mentioning it at all.

What it is

A fact that belongs in the conversation, particularly in an older adult where the calculation has moved. The questions worth asking a prescriber are whether the intensity is still right, whether the time to benefit still fits, and whether glycemia is being monitored on the assumption that this can happen rather than on the assumption that it will not.

Time to benefit is the question the STOPPFrail criteria exist for, and it is a legitimate one in a resident with limited life expectancy. Deprescribing in long-term care.

---

The evidence, with the numbers

The 2010 collaborative meta-analysis pooled 13 randomized statin trials and 91,140 participants, of whom 4,278 developed diabetes over a mean of four years. Statin therapy was associated with a 9% increased risk of incident diabetes, odds ratio 1.09 (95% CI 1.02 to 1.17), with little heterogeneity between trials. In absolute terms: treating 255 patients (95% CI 150 to 852) for four years produced one extra case of diabetes. Meta-regression found the risk was highest in trials with older participants — which is the population this site is about.

The 2024 individual-participant-data meta-analysis from the Cholesterol Treatment Trialists’ Collaboration is the current and more precise source. Across 19 statin-versus-placebo trials (123,940 participants) and four more-versus-less-intensive trials (30,724):

• Low or moderate intensity statin: a 10% proportional increase in new-onset diabetes (1.3% per year versus 1.2% per year; rate ratio 1.10, 95% CI 1.04 to 1.16).
• High intensity statin: a 36% proportional increase (4.8% per year versus 3.5% per year; rate ratio 1.36, 95% CI 1.25 to 1.48).
• About 62% of new-onset diabetes cases were in participants already in the top quarter of the baseline glycemia distribution. The effect is concentrated in people who were already close to the threshold.
• In people who already had diabetes, worsening glycemia: rate ratio 1.10 (1.06 to 1.14) at low or moderate intensity, 1.24 (1.06 to 1.44) at high intensity.

What is on the label

The current FDA-approved labeling carries it as a warning and precaution: “Increases in HbA1c and fasting serum glucose levels have been reported with statins,” with the advice to “optimize lifestyle measures, including regular exercise, maintaining a healthy body weight, and making healthy food choices.” This is a labeled, acknowledged effect, not a contested claim.

References

1. Sattar N, Preiss D, Murray HM, et al. 2010. Statins and risk of incident diabetes: a collaborative meta-analysis of randomised statin trials. Lancet 375(9716):735-42. [meta-analysis, 13 trials, 91,140 participants] · PMID 20167359 · doi:10.1016/S0140-6736(09)61965-6
2. Cholesterol Treatment Trialists’ (CTT) Collaboration 2024. Effects of statin therapy on diagnoses of new-onset diabetes and worsening glycaemia in large-scale randomised blinded statin trials: an individual participant data meta-analysis. Lancet Diabetes & Endocrinology 12(5):306-319. [individual-participant-data meta-analysis] · PMID 38554713 · doi:10.1016/S2213-8587(24)00040-8
3. US Food and Drug Administration 2024. LIPITOR (atorvastatin calcium) prescribing information, section 5.4: increases in HbA1c and fasting serum glucose levels. FDA-approved label, effective 2024-04-15. [FDA-approved drug label] · Source
4. Curtin D, Gallagher P, O’Mahony D 2021. Deprescribing in older people approaching end-of-life: development and validation of STOPPFrail version 2. Age and Ageing 50(2):465-471. [Delphi-validated criteria] · PMID 32997135 · doi:10.1093/ageing/afaa159

Removing restraints without putting anything in their place is not restraint-free care. The 49 items that have to come first.

“We do not use restraints” is a claim about an outcome. Restraint-free care is a claim about 49 items of work that produce it.

The two kinds, and which one hides

Physical restraint is visible: anything a resident cannot remove or leave, including some bed rails and some chairs. Chemical restraint is not visible, because it appears in the chart as a medication — a sedative or antipsychotic given to manage behavior rather than to treat a diagnosed condition.

The second is the one a walkthrough misses and a medication review finds.

Behavior is communication

A resident with dementia who has lost the other channels will tell you something is wrong by behaving. The CareGuard review names the usual causes explicitly, and the list is short and physical: pain, discomfort, boredom, confusion, and underlying conditions including urinary tract infection, dehydration, infection and dementia progression.

One of those causes is worth handling carefully, because getting it wrong produces its own harm. New confusion is a reason to assess. It is not, on its own, a reason to culture urine and start an antibiotic — infectious-disease guidance recommends against exactly that, and a quarter to half of long-term care residents carry bacteria in the urine without infection. The evidence on that.

What has to happen before a restraint is even discussed

• Assessment — physical, mental and emotional; root cause of the behavior; history reviewed for underlying conditions; input from family on past triggers; regular reassessment.
• Eight non-pharmacological interventions — routine, safe physical activity, meaningful engagement, a calm environment, redirection, frequent hydration and bathroom breaks, comfort items, seating that supports safe mobility.
• De-escalation — staff trained in verbal de-escalation, calm communication, time to process, simple language and non-verbal cues, offering choices, involving familiar people.
• Pain and comfort — regular pain assessment especially for non-verbal residents, and non-drug relief.
• Environmental modification — hazards removed, lighting and temperature right, familiar objects, free movement within secure areas.
• Passive monitoring — bed or chair alarms used as a fall-prevention measure without restraint, increased observation, regular check-ins, early intervention.
• Care planning — an individualized plan built with providers and family, reviewed in team meetings, and known to all staff.
• Medication review — regular review to avoid oversedation, consideration of reducing or deprescribing unnecessary psychotropics, side-effect monitoring.
• Trigger work — identifying and minimizing environmental and emotional triggers, quiet retreat space, consistent familiar staff.
• Family involvement — participation in behavior management, education on the harms of restraints, open communication about strategy.
• Legal and ethical compliance — staff trained on state and federal rules against unnecessary restraints, facility policy followed, restraints used only as a last resort in emergencies with documentation and oversight, and residents and families informed of their rights.

All 49 items, in full.

The honest version of the position

The last item is the one that keeps this from being a slogan. The position is not that a restraint is never used under any circumstance. It is that using one is an emergency event, documented as such, with oversight — not a standing arrangement that quietly becomes how a resident lives.

---

CareGuard’s position on restraints · DWARAA.

---

The evidence on antipsychotics in dementia

The mortality signal, and where the label came from

A 2005 meta-analysis of 15 randomized placebo-controlled trials (3,353 patients on drug, 1,757 on placebo) found death in 3.5% of drug-treated patients versus 2.3% on placebo, odds ratio 1.54 (95% CI 1.06 to 2.23). The FDA’s own analysis of 17 placebo-controlled trials, stated in the current approved labeling, puts it this way: “a risk of death in drug-treated patients of between 1.6 to 1.7 times the risk of death in placebo-treated patients. Over the course of a typical 10-week controlled trial, the rate of death in drug-treated patients was about 4.5%, compared to a rate of about 2.6% in the placebo group.” The boxed warning followed for atypical antipsychotics in 2005 and was extended to conventional antipsychotics in 2008.

The label also records cerebrovascular adverse reactions, including fatalities, at a significantly higher incidence than placebo in elderly patients with dementia-related psychosis.

And the efficacy side

CATIE-AD randomized 421 outpatients across 42 sites for up to 36 weeks and found no significant difference between treatments in time to discontinuation for any reason. Discontinuation for intolerability ran at 24% (olanzapine), 16% (quetiapine), 18% (risperidone) and 5% (placebo). Its conclusion is one sentence: “Adverse effects offset advantages in the efficacy of atypical antipsychotic drugs.”

Withdrawal

The DART-AD discontinuation trial in 165 UK care-facility residents found 12-month survival of 70% (95% CI 58 to 80) in those continuing antipsychotics versus 77% (64 to 85) on placebo, hazard ratio 0.58 (0.35 to 0.95). At 24 months the gap was 46% versus 71%; at 36 months, 30% versus 59%.

A Cochrane review of nine randomized trials (606 participants, seven in nursing homes) concluded that “many older people with Alzheimer’s dementia and NPS can be withdrawn from chronic antipsychotic medication without detrimental effects on their behaviour” — with an important exception: two studies of people whose agitation or psychosis had previously responded well to antipsychotic treatment found an increased risk of relapse after discontinuation. Withdrawal is not uniformly safe and is not a policy to apply across a building without individual assessment.

What happened when a country tried it at scale

Antipsychotic use in US nursing homes reached 23.9% of residents by 2011. Following the National Partnership to Improve Dementia Care launched in 2012, use had declined 40.1% to 14.3% by the second quarter of 2019. Notably, sedative-hypnotic use did not rise to compensate — it declined in tandem. The evaluation’s own conclusion about what made it work: “Adequate staffing, particularly of registered nurses, is key.”

References

1. Schneider LS, Dagerman KS, Insel P 2005. Risk of death with atypical antipsychotic drug treatment for dementia: meta-analysis of randomized placebo-controlled trials. JAMA 294(15):1934-43. [meta-analysis, 15 trials] · PMID 16234500 · doi:10.1001/jama.294.15.1934
2. US Food and Drug Administration 2026. RISPERDAL (risperidone) prescribing information, boxed warning and section 5.1. FDA-approved label, effective 2026-05-28. [FDA-approved drug label] · Source
3. Schneider LS, et al. 2006. Effectiveness of atypical antipsychotic drugs in patients with Alzheimer’s disease (CATIE-AD). New England Journal of Medicine 355(15):1525-38. [randomized controlled trial, 421 outpatients] · PMID 17035647 · doi:10.1056/NEJMoa061240
4. Ballard C, et al. 2009. The dementia antipsychotic withdrawal trial (DART-AD): long-term follow-up of a randomised placebo-controlled trial. Lancet Neurology 8(2):151-7. [randomized discontinuation trial, 165 care-facility residents] · PMID 19138567 · doi:10.1016/S1474-4422(08)70295-3
5. Declercq T, et al. 2013. Withdrawal versus continuation of chronic antipsychotic drugs for behavioural and psychological symptoms in older people with dementia. Cochrane Database of Systematic Reviews (3):CD007726. [Cochrane review, 9 RCTs] · PMID 23543555 · doi:10.1002/14651858.CD007726.pub2
6. Jeste DV, et al. 2008. ACNP white paper: update on use of antipsychotic drugs in elderly persons with dementia. Neuropsychopharmacology 33(5):957-70. [expert white paper] · PMID 17637610 · doi:10.1038/sj.npp.1301492
7. Crystal S, et al. 2020. National Partnership to Improve Dementia Care in Nursing Homes campaign: state and facility strategies, impact, and antipsychotic reduction outcomes. Innovation in Aging 4(3):igaa018. [national mixed-methods evaluation] · PMID 32699827 · doi:10.1093/geroni/igaa018
8. Harris DA, et al. 2022. Potential unintended consequences of antipsychotic reduction in Ontario nursing homes. Journal of the American Medical Directors Association 23(6):1066-1072.e7. [repeated cross-sectional, 2010-2019] · PMID 35143749 · doi:10.1016/j.jamda.2021.12.043

Deprescribing is a formal discipline with named tools — Beers, STOPP/START v3, STOPPFrail — not an improvisation. What it is and how a facility supports it.

A long medication list in a nursing home is rarely anyone’s decision. It is the sum of several people’s reasonable decisions, none of whom saw the whole list.

How lists get long

A resident arrives from a hospital admission, which arrived from home, which accumulated from a primary care relationship and two specialists. Each prescriber added something defensible. Nobody subtracted, because subtracting requires knowing why something was started, and that reason is often no longer in any record anyone can see.

Then the cascade: a drug produces a symptom, the symptom is read as a new condition, the new condition is treated, and the second drug produces a symptom of its own. It is visible only if you read the list in date order as one document.

Deprescribing is a discipline, not a lapse

It has named tools and a literature.

• AGS Beers Criteria — the American Geriatrics Society’s expert-consensus list of medications that are often inappropriate in older adults.
• STOPP/START criteria, version 3 (2023) — covering both potentially inappropriate medications and potential prescribing omissions, including falls risk.
• STOPPFrail — for limited life expectancy, where a preventive drug’s time to benefit may exceed the time available to receive it.

The two harms worth naming

Falls from antihypertensives. An over-tight dose produces orthostatic hypotension, which produces dizziness and syncope on standing, which produces falls, fractures and head injuries. Beers and STOPP/START both single out centrally acting antihypertensives and alpha-blockers as not recommended for routine hypertension treatment in older adults. More on that pathway.

Statins and glycemia. The FDA added a class-wide label change on February 28, 2012 warning of increased HbA1c and fasting glucose. It is a trade-off, not a prohibition. More on that trade-off.

What a facility contributes

Deprescribing is a prescriber’s decision, but it fails without the facility.

• A baseline at intake, so later change is measurable rather than remembered. Ongoing monitoring.
• Observation that gets written down. A resident who is steadier on their feet three weeks after a dose reduction is evidence, if anyone recorded that they were unsteady before.
• Something in place of the drug. Structure, activity, social engagement, an environment that is not producing agitation. The eight interventions.
• A care plan that is actually revised. Adjusting care plans.

---

DWARAA is CareGuard’s deprescribing and dementia care program.

---

The evidence base

How much medication is actually on the list

The SHELTER study surveyed 4,023 residents across 57 nursing homes in eight European countries and found polypharmacy (5 to 9 drugs) in 49.7% of residents and excessive polypharmacy (10 or more) in 24.3%. A separate cohort of 3,234 nursing home residents found essentially the same split (50% and 24%) and reported that residents on polypharmacy and on excessive polypharmacy had significantly greater one-year decline in cognitive performance score (beta 0.10, 95% CI 0.01 to 0.20 and beta 0.13, 95% CI 0.01 to 0.24 respectively), with no significant difference in activities of daily living.

A systematic review of 13 studies covering 140 nursing homes and more than 6,900 residents found potentially inappropriate prescriptions in 67.8% to 87.7% of residents by STOPP criteria, and potential prescribing omissions in 39.5% to 99.7% by START criteria. The width of those ranges is itself a finding: practice varies enormously between buildings.

The tools, with their current versions

• AGS Beers Criteria, 2023 update. Intended, in the panel’s own words, “to be applied to adults 65 years old and older in all ambulatory, acute, and institutionalized settings of care, except hospice and end-of-life care settings.”
• STOPP/START version 3 (2023). Validated by a Delphi panel of 11 academic physicians from eight European countries over four rounds, from a literature review covering April 2014 to March 2022. It contains 190 criteria in total — 133 STOPP and 57 START — a 66.7% increase on version 2 (2015). Earlier versions were published in 2008 and 2015; a published correction accompanies version 3.
• STOPPFrail version 2 (2021). For older people approaching end of life: a method for identifying who is approaching end of life, plus 25 deprescribing criteria, with new guidance on antihypertensives, anti-anginal medication and vitamin D.

Does deprescribing change outcomes?

This is where an honest page differs from a brochure. The 2016 systematic review and meta-analysis of 132 papers and 34,143 participants found deprescribing polypharmacy significantly reduced mortality in non-randomized studies (OR 0.32, 95% CI 0.17 to 0.60) but not in randomized studies (OR 0.82, 95% CI 0.61 to 1.11). The 2024 update, covering 259 studies, found no significant mortality reduction in either randomized (OR 0.96, 95% CI 0.84 to 1.09) or non-randomized studies (OR 0.70, 95% CI 0.36 to 1.38).

The signal that does survive is about how it is done. In the 2016 analysis, patient-specific deprescribing interventions reduced mortality (OR 0.62, 95% CI 0.43 to 0.88) while generalized educational programs did not (OR 1.21, 95% CI 0.86 to 1.69). The 2024 update found the same direction for patient-specific interventions (OR 0.79, 95% CI 0.63 to 0.99).

That is the case for doing this resident by resident with a prescriber, and against doing it as a facility-wide policy campaign. It is also why deprescribing sits inside a program here rather than standing on its own.

References

1. Onder G, et al. 2012. Polypharmacy in nursing home in Europe: results from the SHELTER study. Journals of Gerontology Series A 67(6):698-704. [cross-sectional, 4,023 residents, 8 countries] · PMID 22219520 · doi:10.1093/gerona/glr233
2. Vetrano DL, et al. 2018. Association of polypharmacy with 1-year trajectories of cognitive and physical function in nursing home residents. Journal of the American Medical Directors Association 19(8):710-713. [longitudinal cohort, 3,234 residents] · PMID 29861194 · doi:10.1016/j.jamda.2018.04.008
3. Díaz Planelles I, et al. 2023. Prevalence of potentially inappropriate prescriptions according to the new STOPP/START criteria in nursing homes: a systematic review. Healthcare (Basel) 11(3):422. [systematic review, 13 studies, 140 nursing homes] · PMID 36766997 · doi:10.3390/healthcare11030422
4. American Geriatrics Society Beers Criteria Update Expert Panel 2023. American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. Journal of the American Geriatrics Society 71(7):2052-2081. [clinical practice guideline] · PMID 37139824 · doi:10.1111/jgs.18372
5. O’Mahony D, Cherubini A, Guiteras AR, et al. 2023. STOPP/START criteria for potentially inappropriate prescribing in older people: version 3. European Geriatric Medicine 14(4):625-632. [Delphi-validated criteria] · PMID 37256475 · doi:10.1007/s41999-023-00777-y
6. O’Mahony D, et al. 2023. Correction: STOPP/START criteria for potentially inappropriate prescribing in older people: version 3. European Geriatric Medicine 14(4):633. [published correction] · PMID 37326916 · doi:10.1007/s41999-023-00812-y
7. Curtin D, Gallagher P, O’Mahony D 2021. Deprescribing in older people approaching end-of-life: development and validation of STOPPFrail version 2. Age and Ageing 50(2):465-471. [Delphi-validated criteria] · PMID 32997135 · doi:10.1093/ageing/afaa159
8. Page AT, et al. 2016. The feasibility and effect of deprescribing in older adults on mortality and health: a systematic review and meta-analysis. British Journal of Clinical Pharmacology 82(3):583-623. [systematic review and meta-analysis, 34,143 participants] · PMID 27077231 · doi:10.1111/bcp.12975
9. Quek HW, Page A, Lee K, et al. 2024. The effect of deprescribing interventions on mortality and health outcomes in older people: an updated systematic review and meta-analysis. British Journal of Clinical Pharmacology 90(10):2409-2482. [systematic review and meta-analysis, 259 studies] · PMID 39164070 · doi:10.1111/bcp.16200

The pathway from an over-tight antihypertensive dose to a fractured hip runs through orthostatic hypotension. What it is, and what a facility should be watching.

The route from a blood pressure prescription to a broken hip is short, well described, and almost never written in a fall report.

The pathway

1. A dose is set to a target that suits a younger patient.
2. Standing produces a drop in blood pressure the body no longer compensates for quickly — orthostatic hypotension.
3. Dizziness, unsteadiness, sometimes syncope, in the seconds after standing.
4. A fall — typically getting out of bed at night, or up from a chair.
5. A fracture or a head injury, and the morbidity and mortality that follow it, none of which has anything to do with blood pressure.

What orthostatic hypotension is, precisely

A drop of at least 20 mmHg systolic or 10 mmHg diastolic within three minutes of standing. It affects roughly one in five community-dwelling older adults, and medications are the leading contributing factor. It is measurable in a corridor with a cuff and a chair, and it is measured far less often than it is discussed.

The drugs the criteria single out

The AGS Beers Criteria and STOPP/START version 3 (2023) both single out centrally acting antihypertensives and alpha-blockers as not recommended for routine treatment of hypertension in older adults. That is not a fringe position; it is expert consensus in the standard geriatric prescribing tools.

What a facility should be doing

• Measuring lying and standing pressures on residents who have fallen, and on residents whose medication changed.
• Recording the time of falls. A cluster at night, or within an hour of a morning dose, is the finding.
• Reading fall reports as medication data. A fall report that names the hallway and not the drug list has answered the wrong question.
• Doing the environmental work too — lighting, flooring, clutter, bed height, night supervision. The fall prevention items.

The measurement side

Orthostatic drop and autonomic function are clinical measurements, not facility observations, which is where the pairing with Measura sits. The Measura pairing.

---

The evidence on orthostatic hypotension

A 2020 systematic review and meta-analysis pooled 20 community studies (24,967 people aged 60 and over) and six long-term-care studies (2,694 residents). Pooled prevalence was 22.2% in community-dwelling older people (95% CI 17 to 28) and 23.9% in long-term care settings (95% CI 18.2 to 30.1) — roughly one resident in four.

A 2024 systematic review of 184 studies, which informed the World Falls Guidelines, found orthostatic hypotension associated with falls at an unadjusted odds ratio of 1.39 (95% CI 1.18 to 1.64) over a 12-month reporting interval — smaller than stroke (1.90) or peripheral arterial disease (1.82), and still substantial across a whole building.

On mortality, be careful with the number

A meta-analysis of 13 prospective studies and 121,913 patients found orthostatic hypotension associated with all-cause death at RR 1.50 (95% CI 1.24 to 1.81), incident coronary heart disease RR 1.41, heart failure RR 2.25 and stroke RR 1.64. In the subgroup aged 65 and over the mortality association was not statistically significant — 1.26 (95% CI 0.99 to 1.62). The headline figure is therefore not the one to quote at a nursing-home population, and it is not quoted here.

Antihypertensive intensity and serious fall injury

A competing-risk analysis of 4,961 community-living adults over 70 with hypertension found adjusted hazard ratios for serious fall injury of 1.40 (95% CI 1.03 to 1.90) with moderate-intensity antihypertensive treatment and 1.28 (95% CI 0.91 to 1.80) with high-intensity treatment, compared with non-users. Among the 503 participants who had already had a fall injury the hazard ratios rose to 2.17 (0.98 to 4.80) and 2.31 (1.01 to 5.29). A previous fall is the signal that changes the calculation.

References

1. Saedon NI, Tan MP, Frith J 2020. The prevalence of orthostatic hypotension: a systematic review and meta-analysis. Journals of Gerontology Series A 75(1):117-122. [systematic review and meta-analysis] · PMID 30169579 · doi:10.1093/gerona/gly188
2. Bourke R, et al. 2024. Cardiovascular disorders and falls among older adults: a systematic review and meta-analysis. Journals of Gerontology Series A 79(2). [systematic review and meta-analysis, 184 studies] · PMID 37738307 · doi:10.1093/gerona/glad221
3. Ricci F, et al. 2015. Cardiovascular morbidity and mortality related to orthostatic hypotension: a meta-analysis of prospective observational studies. European Heart Journal 36(25):1609-17. [meta-analysis, 121,913 patients] · PMID 25852216 · doi:10.1093/eurheartj/ehv093
4. Tinetti ME, et al. 2014. Antihypertensive medications and serious fall injuries in a nationally representative sample of older adults. JAMA Internal Medicine 174(4):588-95. [competing-risk cohort, 4,961 adults over 70] · PMID 24567036 · doi:10.1001/jamainternmed.2013.14764