Medication Monitoring for People with Dementia in Care Homes: Clinical Impact of Nurse-led monitoring

Our aims:

Preventable Adverse Drug Reactions (ADRs) (sometimes referred to as side-effects) account for 5-8% unplanned hospital admissions. Failure to monitor patients is a greater problem than poor prescribing.

We aimed to test whether our intervention, the nurse-led West Wales ADR (WWADR) Profile for Mental Health Medicines improved care, and patient outcomes, for people with dementia prescribed mental health medicines (antipsychotics, antidepressants, antiepileptics, benzodiazepines).

Our objectives:

1)      To explore changes in documentation for:

• Amelioration of problems identified on the WWADR Profile.
• Problems found and actions taken using the WWADR Profile.
• Medication review and medication changes.

2)      Describe the clinical impact on individual participants.

3)      Report views of care home staff.

4)      To comply with recommendations in NICE Medicines’ Optimisation guidance (NG5)

Recommendation 9:  '"Consider using a screening tool to identify potential medicines-related safety incidents".

Recommendation 27: During a structured medication review, take into account …

• How safe the medicines are, how well they work for the person, how appropriate they are
• Any monitoring that is needed

5)      To comply with recommendations in NICE guidance Managing Medicines in care homes (SC1):

Recommendation 1.8 Reviewing Medicines:

1.8.3 Health and social care practitioners should ensure that medication reviews involve the resident and/or their family members or carers and a local team of health and social care practitioners (multidisciplinary team).

1.8.5 Health and social care practitioners should discuss and review the following during a medication review:

• the resident's (and/or their family members' or carers') concerns, questions or problems with the medicines
• all prescribed, over-the-counter and complementary medicines that the resident is taking, and what these are for
• how safe the medicines are, how well they work, how appropriate they are, and whether their use is in line with national guidance
• any monitoring tests that are needed
• any problems the resident has with the medicines, such as side effects or reactions, taking the medicines themselves (for example, using an inhaler) and difficulty swallowing
• helping the resident to take or use their medicines as prescribed (medicines adherence)

Adverse drug reactions (ADRs) are responsible for:

• 5-8% unplanned UK hospital admissions
• 4% of UK hospital bed occupancy;
• 10% (68/678) hospitalisations in US Veterans’ Affairs Medical Centres;
• 3.7% hospital admissions globally;
• £1-2.5bn NHS costs and ~$30bn US expenditure annually.

Of 10 cases, 9 fatal, of substandard care reported by the Health Service Commissioner (2011), 2 relate to failure to monitor prescription medicines, including olanzapine.

However, adverse consequences of untreated conditions are equally common in primary care. Enhanced patient monitoring might prevent many of these problems, and our nurse-led interventions have improved care in adult mental health, and respiratory medicine.

Most service users are prescribed long-term medication for a range of conditions, and some receive more than 20 different preparations. There are particular concerns over associations between antipsychotics and increased overall mortality, worsening cognitive impairment, hip fracture, diabetes and stroke. Although antidepressants are generally considered safe, they are associated with serious adverse events, including: serious bleeding, paradoxical aggression, falls, and fractures.

Concerns have been raised over medicines’ management in Staffordshire, and in hospitals caring for older people in South Wales. Further inspections following Andrews & Butler’s report (2014) indicated unresolved issues, reported in a Welsh Government paper in 2014.

The Welsh Government’s Prudent Healthcare initiative calls for interventions to prevent adverse events and optimise medicines’ management and prescribing, without causing harm or increasing costs. Nurses are well-placed to promptly recognise, identify and report ADRs, but need additional support, which our Profile provides.

Our profile originated in applied pharmacology courses and was developed with practitioners and service users to improve recognition and response to ADRs. It asks nurses to monitor patients’ vital signs (heart rate, blood pressure, oxygen saturation) observe and interview patients to check for signs and symptoms of possible ADRs. It also checks that any adverse effects of medicines are not affecting intake and other aspects of health promotion. The attached guidelines offer suggestions on managing any problems identified and when to contact prescribers.

This prospective ‘before and after’ record review explored the clinical gains from nurse-led medication monitoring using the West Wales ADR Profile for Mental Health Medicines on two occasions, one month apart.

Setting

The work was undertaken November 2012 to February 2013 in three homes caring for 81 service users with cognitive deficits and challenging behaviours. 

Participants

Service users were assessed for eligibility and approached by their nurses. Inclusion criteria were: resident at the care home; diagnosed with dementia; currently taking antipsychotics and/or anti-epileptics and/or antidepressants; willing and able to give informed, signed consent themselves, or where capacity was lacking, a guardian was willing to give informed, signed consent. We excluded those: not well enough to participate, as screened by their nurses; aged <18. 

Data Collection

Service users’ records were reviewed on three occasions: 1) before introduction of the profile, 2) after the first use of the profile, 3) after the second use of the profile one month later. We extracted data from service users’ case notes and completed profiles to identify: problems recorded and actions taken following administration of the ADR Profile. We noted:

• Number and nature of problems documented as present, actioned or discussed with prescribers with and without the profile
• Follow-up actions and change in care plans
• Changes in prescribing

We did not record the involvement of other professionals.

Analysis

Problems noted and actions documented as taken were described.

Ethics

Approval for the study was granted by the SW Wales NHS Research Ethics committee (reference: 12/WA/0311, 23.10.2012) and the University College of Human and Health Science Research Ethics Committee. Service users’ or guardians’ written consent for researchers to review service users’ notes was sought by qualified nurses, who were familiar with the Mental Capacity Act (2005) and employed by the care homes. It was agreed that the project posed no physical risks. Usual standard care was delivered throughout.

Barriers

To explore potential barriers, feedback on the WWADR Profile was sought from 10 nurses in cognitive interviews. Nurses were asked to read the Profile items aloud, reflecting on the meaning and interpretation of each. Results were used to redraft the Profile.

Costs

A researcher was funded for 4 months to collect data. Additional care was standard NHS provision. 

Of 81 residents in 3 care homes, 29 met the inclusion criteria, and 11 consultees/ guardians signed consent for participation.

Problems identified

On first administration, the WWADR Profile identified previously undocumented problems for all service users: mean 12.7 [SD 4.7], 95% CI 9.6-15.9, range 8-22. Some problems, such as abnormal movements (4), postural hypotension (2), pain (2) and fluid intake were recorded and monitored for the first time.

Problems addressed

When the ADR profiles were re-administered 1 month later, all service users had had at least one problem ameliorated: mean 4.9 [SD 3.6], 95% CI 2.5-7.3, range 1-11. Examples included: access to dentist (n=5) (one service user needed a filling); constipation treated (n=2); fluid intake normalised (n=4). New care plans were in place for:

• Oral hygiene (n=2)
• Risk of deydration due to low fluid intake (n=2)
• Skin care (n=2)
• Constipation (n=1)
• Postural aids (n=1)

Other problems, such as aggression and cognitive decline, were more intractable.

Three new medicines were prescribed at follow up:

• Creams to manage psoriasis
• Terbinafine 1% mane to treat a rash
• On referral to the Parkinson's service, a new therapeutic regimen was prescribed.

Two therapeutic regimens were revised, in response to problems found:

• Increased and regularised administration of analgesic creams for arthritic pain in the knee. Abnormal posture had ameliorated at follow up
• Reduced benzodiazepine doses: sedation was no longer a problem at follow up.

Feedback from nurses

 Nurses indicated that the Profile items were user friendly and easily understood: 

• N2      It is easy and self-explanatory.

There were some concerns over the blurred boundaries between ADRs and underlying conditions.

• N4 It is important to consider whether the problem is a side effect of meds or disease deterioration.

However, problems should be addressed, regardless of their cause or any subjectivity in the assessment.

Nurses were using their intimate knowledge of the participants to complete the Profile, which might lead to inconsistencies:

• N5 Hard to assess if you don’t know the service user. Things could be missed. Changes could be gradual. Having a worsening response option is an advantage but if two different people are assessing then interpretation may be different.

The Profile took 10-25 minutes to complete for each service user. However, this was seen as a useful investment:

• N5 Time consuming but if used benefits service users

• Nurse-led medicines’ monitoring:
• improved some aspects of care, including prescribing, mouth care and pain management
• focused attention on  participants’ reports of adverse events.

The Profiles are: low cost, low risk and convenient for service users and professionals. Potential benefits include:

• cost savings
• increased quality and safety of care
• bridging communication gaps between patients and prescribers.
•  Profiles do not replace clinical knowledge and experience, but they may help by suggesting solutions to problems potentially related to prescribed medicines.
• Nurses need to find up to 30 minutes per patient to complete the Profiles. This can be divided into 10 minute slots, if preferred. Time should also be set aside to deal with any gaps in care identified.
• Nurses can action many problems themselves e.g. mouth care, fluid intake.
• Many problems identified need prescribers’ involvement and efficient communication. This should be in place before introducing the Profiles.
• Full benefit will only be realised by follow up administration 2-4 weeks later.

If implemented and actioned, the Profiles will address recommendations from the Francis’ and Andrews’ inquiries into health care failings:

• enhanced transparency and inter-disciplinary teamwork;
• easily accessible and systematic recording of routine observations;
• frequent checking to minimize medication errors;
• a “zero tolerance” approach to the improper administration of sedation and medicines;
• comprehensive education initiatives in medication management;
• monitoring food and fluid intake.

The study is described in more detail in Jordan et al (2014), where the Profile can be found as an online appendix.

Future work should more closely involve pharmacists and prescribers. We are seeking funding for a study where pharmacists will be asked to review the completed Profiles before they are passed to prescribers.