Clinical Senior Lecturer in Ageing and Health
The evidence } There is evidence that medication review in older people can lead to an
improvement in the appropriateness of prescribing
} There is evidence that these changes persist (at least out to 6 months)
} There is very little evidence that this process reduces adverse drug events,
hospitalisation, or improves quality of life
} This is due to absence of evidence, not evidence of ineffectiveness
Aim of the guidelines } To provide advice and support to primary care physicians undertaking
} Not prescriptive } Aims to help navigate a complex area of practice } Not a checklist } Highlights particularly high-risk situations or combinations } Aims to help improve appropriateness of prescribing, not merely a tool to
} Provides background information to guide who might usefully undergo
Guideline development } NHS Highland and NHS Tayside guidelines
} Cross-fertilisation between the two
} Informed development of NHS Scotland guidance
} National guidance exists to al ow each NHS board to develop their own
Example: CNS medications } Benzodiazepines - discuss reducing long-term therapy with the aim of stopping } Antipsychotics in dementia - review the continued need, consider reducing the dose or
} Cholinesterase inhibitors - What is the most recent MMSE? (may not be indicated if MMSE
<11/30, contact specialist services for advice). Is the cholinesterase inhibitor stil effective/
} Review combinations of antidepressants such as tricyclic antidepressants for analgesia used in
combination with other antidepressants for depression
} Metoclopramide- review long-term use } Antihistamines for vertigo- review long-term use
Example: High risk prescribing } NSAID
} + Angiotensin Converting Enzyme Inhibitor [ACE] or Angiotensin 2
Receptor Blocker [ARB] + Diuretic [‘Triple Whammy’ combo]
Any long-acting or a long-term benzodiazepine or other hypnotic e.g. Zopiclone. It is useful to reduce the dose if
the medication cannot be stopped completely. Use the safest available hypnotic e.g. Short half life. Specific
information on the reduction of benzodiazepines and Z drugs can be found as a NHS Clinical Knowledge Summary,
http://www.cks.nhs.uk/benzodiazepine_and_z_drug_withdrawal and insomnia guidance at
Antihypertensives including diuretics. Stop or reducing the dose of Calcium Channel Blockers if they cannot be
stopped completely. This may be indicated particularly if the patient has ankle swelling resistant to diuretics. For
people on combinations of antihypertensives, consider reducing the doses or stopping some of the drugs. Ask the
patient if they are dizzy on standing up. If diuretics are for dependent ankle oedema consider other management.
Optimise antihypertensive therapy bearing in mind the fal s risk, mobility and postural hypotension.
Other CNS medication- review the need for antidepressants and antipsychotics, and anti-epileptic medication
First generation (sedating) antihistamines
Review any anticholinergic drugs for bladder spasm or other drugs with anticholinergic side effects e.g. tricyclic
Who should be reviewed? } Older, frail people are at highest risk of drug side effects
} People on a lot of medications are at higher risk of side effects
} There is however no consensus on which groups should be prioritised, or
} The guideline suggests a number of groups who might benefit from review;
groups to review wil clearly depend on time and resources available
Possible groups to review: } Care home residents (high risk, may have limited life expectancy and shifted
} Patients on more than 12 medications (high risk, most scope for
} Patients on pal iative care register (short life expectancy), or with DNAR
} Patients on virtual ward / anticipatory care plans } Patients with frequent hospitalisation } Patients who fal } Housebound patients with care management or DN involvement } As part of long-term condition review
Practical issues around review decisions } Stopping the meds that do the harm
} Deciding which meds don’t do any good
Association of medication classes with falls
} JAGS systematic review: All psychotropics:
Leipzig RM et al. JAGS 1999; 47: 30-9 and 40-50
Do antihypertensives really cause falls? } Thiazides
Sleeping meds are bad for you } Study on inpatients prescribed zolpidem (n=16300)
} Compared those who received zolpidem with those prescribed, but not
} Adjusted for comorbidity, insomnia, delirium, fal s risk, visual impairment,
length of stay, dementia, gait problems, other meds, age, sex
} Crude rate of fal s 3.04% vs 0.71% (p<0.001)
} Adjusted OR for fal s in those receiving zolpidem: 4.37 (95% 3.34 to 5.76)
Anticholinergics } MRC Cognitive function and ageing study
} 13000 people aged 65 and over; 2 yr fol ow up
} Progression of decline on MMSE; mortality
} No anticholinergics vs definite vs possible anticholinergics } Greater decline in MMSE in definite anticholinergic group:
0.33 points cf no use (95% CI 0.03 to 0.64)
} Greater mortality in definite group:
Anticholinergic side effects and falls risk
} Fal s in residential home residents (US) (n=330)
Drug burden index <1 (low) OR = 1.61 (1.17 to 2.23) Drug burden index >1 (high) OR = 1.90 (1.30 to 2.78) After adjustment for age, sex, polypharmacy, comorbidity, fal s history, cognition, walking aids, incontinence
N=3355. Mean age 81.6 yrs; 40% male. Mean fol ow up 5.2 yrs Cox regression analysis; forced entry for al covariates. Risk of death after discharge from rehab:
Unadjusted HR Adjusted HR model 1 Adjusted HR model 2 (95% CI) (95% CI) (95% CI) ARS = 0 to 1 ARS = 2 to 3 ARS = 4 to 5 ARS = 6 and over P for trend
Model 1: Age, sex, discharge Barthel, statin use, diagnosis of stroke, fracture, heart failure, antiplatelet use Model 2: As above plus DFOM scores for cognition, mental health, understanding, food intake, fluid intake, swal ow
Life expectancy and NNTs } Some older, frail people are likely to die within a few months
} Continuing secondary prevention medication in this group is probably not
} Use of NNT data from trials has been proposed as a way of identifying
medications that are unlikely to benefit those with limited life expectancy
- It is not easy to tell who has a limited life expectancy
- NNTs derived in young trial populations may not apply in older people –
may be less efficacious; but absolute risk higher, so NNT may actual y be lower
- Patient wishes are particularly important here
} The Tayside Day Hospital cohort (n=299) were frail, old (mean 79 yrs) with
multiple comorbidities (JAGS 2005; 53: 1991-5).
} Yet only 43% were dead at 3 years after being seen in Day Hospital.
} The Tayside Rehabilitation cohort (n=3180) who were discharged alive
from their first rehab admission show the fol owing survival data:
NNT: Example from HYVET trial } Data from HYVET (Hypertension in the Very Elderly Trial; NEJM
2008;358:1887-98) – mean age 84. BP at recruitment 170/90. Only 12% had previous CV disease.
Day Hospital
Patient wishes: Example } 87 year old woman with previous smal stroke and AF, on warfarin
} Now presents with several fal s over last few months
} If pt terrified of further stroke – strong indication for continuing warfarin
} If pt more concerned re: bleeding into brain from fal s, should stop warfarin
Opportunity costs } There is also a danger of focussing al the energy of a review on stopping meds
which do no good – but then not stopping those actively doing harm:
} Mrs W has an history of fal s, stroke, and was recently hospitalised with
pneumonia. She has moved into a residential home
} After review, you stop the amlodipine, perindopril and atorvastatin
} However, the zopiclone is the biggest risk factor for fal s
} And the PPI is a risk factor for a further bout of pneumonia
} Antihypertensives are only associated with a marginal increase in fal s risk
} Atorvastatin unlikely to be doing harm, even if evidence of benefit
} 6 fal s over last 2 months. Two witnessed; no prodromal symptoms, no
} History of stroke 2005, fractured NOF 2009, depression, peptic ulcer 1994,
hypertension, atrial fibril ation, knee osteoarthritis
} Mobilises with two sticks; rarely leaves the house. Home help twice daily;
} On examination: Nil of note CVS / RS / Abdo / NS
} Lying BP 162/74; standing BP 166/80. Pulse 92 irregular
A practical approach: } 1. Target the central y acting medications
} Consider using a single drug for multiple indications (e.g. mirtazepine
} Anything with anticholinergic side effects; consider if necessary or
} Consider physiotherapy (esp quads training) for joint pain
} 3. Get rid of meds that aren’t there for any good reason
} Quinine – doesn’t work in most people
} PPI’s – often no indication; they don’t cause fal s, but they may contribute
to osteoporosis, pneumonia, C diff, renal impairment, low sodium, low magnesium, microscopic colitis. .
} 4. Consider balance of risk/benefit for CVS meds
} Switch from alpha blockers to alternatives
} Switch from digoxin to betablockers
} Stop calcium channel blockers if ankle oedema – then you can stop the
} If heart failure with preserved systolic function, stop spironolactone (and
} Rationalise antianginals – e.g. stop nitrates and give betablocker instead
- Stopped (poor practice; doesn’t work)
- Stopped (oedema was due to amlodipine)
} Cocodamol - Stopped (changed to paracetamol and
} Calcium/vitamin D (1g/800IU) daily
} Alendronate 70mg weekly (T score -2.8 at hip on DXA)
} Piroxicam gel (1 applic to knees) twice daily
} Thus its not al about the number of meds; it is about the appropriateness of
} Appropriate prescribing is the target, not polypharmacy per se
The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne A COMPARISON OF RISPERIDONE AND HALOPERIDOL FOR THE PREVENTION OF RELAPSE IN PATIENTS WITH SCHIZOPHRENIA JOHN G. CSERNANSKY, M.D., RAMY MAHMOUD, M.D., M.P.H., AND RONALD BRENNER, M.D., ABSTRACT Background lifetime prevalence of 0.7 percent in the Unit-of maintenance treatment in patients with psychoticed Sta
Clinical Senior Lecturer in Ageing and Health
The evidence } There is evidence that medication review in older people can lead to an
improvement in the appropriateness of prescribing
} There is evidence that these changes persist (at least out to 6 months)
} There is very little evidence that this process reduces adverse drug events,
hospitalisation, or improves quality of life
} This is due to absence of evidence, not evidence of ineffectiveness
Aim of the guidelines } To provide advice and support to primary care physicians undertaking
} Not prescriptive } Aims to help navigate a complex area of practice } Not a checklist } Highlights particularly high-risk situations or combinations } Aims to help improve appropriateness of prescribing, not merely a tool to
} Provides background information to guide who might usefully undergo
Guideline development } NHS Highland and NHS Tayside guidelines
} Cross-fertilisation between the two
} Informed development of NHS Scotland guidance
} National guidance exists to al ow each NHS board to develop their own
Example: CNS medications } Benzodiazepines - discuss reducing long-term therapy with the aim of stopping } Antipsychotics in dementia - review the continued need, consider reducing the dose or
} Cholinesterase inhibitors - What is the most recent MMSE? (may not be indicated if MMSE
<11/30, contact specialist services for advice). Is the cholinesterase inhibitor stil effective/
} Review combinations of antidepressants such as tricyclic antidepressants for analgesia used in
combination with other antidepressants for depression
} Metoclopramide- review long-term use } Antihistamines for vertigo- review long-term use
Example: High risk prescribing