Reducing inappropriate polypharmacy in primary care through pharmacy-led interventions

Introduction

Inappropriate polypharmacy is described as when one or more medications are prescribed that are not or no longer needed, either because:

1. There is no evidence-based indication, the indication has expired, or the dose is unnecessarily high;
2. One or more medicines fail to achieve the therapeutic objectives they are intended to achieve;
3. One or a combination of several medicines cause unacceptable adverse drug reactions (ADRs), or put the patient at a high risk of such ADRs;
4. The patient is not willing or able to take one or more medicines as intended^[1]
   .

Polypharmacy is a problem that is receiving increasing attention nationally. Several documents have been published highlighting that the absolute benefit made by each additional medicine is likely to reduce when a person is taking multiple preventative medicines, and that the risk of harm is likely to be increased further as they take more medicines^[1],[2],[3]
. Patients at highest risk of inappropriate polypharmacy are those with the greatest frailty, who are taking the most medicines and are taking high-risk medicines, such as benzodiazepines^[1]
.

In 2017, following the recommendations made in the national guidance, the East Staffordshire Clinical Commissioning Group (ESCCG) Medicines Optimisation Team (MOT) identified reducing inappropriate polypharmacy as a major quality, innovation, productivity and prevention (QIPP) goal^[1],[2],[3]
. The practice-based pharmacist team set up polypharmacy clinics to review patients taking multiple medicines with the aims of tackling inappropriate polypharmacy, achieving overall cost savings and improving quality of prescribing for patients. The polypharmacy savings target was set at £50,000 for the 2017/2018 financial year.

Methods

General practice selection

The polypharmacy clinics were set up at eight GP practices and ran from April 2017 until March 2018. The service continued at six practices over the following nine months (April 2018 to December 2018). Only eight practices were originally selected, owing to limitations in available resources.

Patient selection

To be eligible for review, patients had to be registered with one of the eight practices and take ten or more prescribed medicines per day. Selection criteria also included: taking high-risk medicines (e.g. benzodiazepines); taking medicines that can cause acute kidney injury (e.g. angiotensin-converting-enzyme [ACE] inhibitors or non-steroidal anti-inflammatory drugs [NSAIDs]); and taking anticoagulants.

Housebound patients were excluded because the ESCCG team could not perform domiciliary visits. These patients were consequently referred to a pharmacist providing community services.

Staffing and training

The ESCCG MOT project allocation was 0.4 whole-time equivalent (WTE) pharmacists covering the polypharmacy clinics. This accounted for the time spent setting up and running the clinics over the duration of the project. One pharmacist was allocated as the project lead and other pharmacist team members provided support.

The MOT completed training packages from the Centre for Pharmacy Postgraduate Education on polypharmacy and consultation skills, and the pharmacists involved had post-graduate clinical diplomas^[4],[5]
.

Developing the service

The MOT reviewed various prescribing tools to help assess patients’ medicines, including the Virgin Care Count Tool, STOPP-START and 7-Steps, before selecting 7-Steps to be used in the polypharmacy clinics^[6]
,[7]
. This was chosen because it covered all of the main points needed when conducting a review, including: the aims and objectives of therapy; the essential medicine therapy; the current ADRs and risk of ADRs; cost-effectiveness; and adherence.

An ESCCG standard operating procedure (SOP) that incorporated the 7-Steps tool was created for the polypharmacy review clinics. This had to be signed by a lead GP for each practice prior to the initiation of the clinics. The SOP stated the review must be a level 3 medicine review, detail the search criteria and outline the level of support required from the practice. After the review was complete, it was recorded as a medicine review in the patient’s notes to save duplication by the GP. Owing to capacity, patients were generally only seen once.

Polypharmacy review

The appropriateness of the patient’s medicines was checked during each review, and issues with concordance (i.e. if the patient needs support in medicine taking, along with prescribing communication) and ordering quantities, as well as opportunities to deprescribe, were identified. Medicines were stopped if they were no longer required; switched if an acceptable, more cost-effective alternative was available; or added when clinically indicated (e.g. adding gastroprotection when taking an NSAID). Condition-specific national and local guidance was used as the evidence base for interventions made.

The pharmacist conducting the clinics would action any agreed interventions and the GP surgery would follow up with the patient. An EMIS template (Egton Medical Information Systems) was created in order to speed up consultations and capture all of the required information. The template prepopulates the consultation screen with the required discussion areas and allows a methodical approach, which helps to save time.

In line with legislation, interventions that involved amending a prescription-only medicine were checked with a doctor first before being implemented; therefore, the clinics did not require pharmacists to be prescribers^[8]
. Changes to pharmacy-only or general sales list medicines were completed by the pharmacist during the consultation. The pharmacist also made formulary choice brand switches that had been agreed by the CCG and South Staffordshire Area Prescribing Committee^[9]
.

Data collection and cost savings

The team invested in a web-based intervention recording tool — MedOptimise — to capture interventions and cost savings^[10]
. The following data from the reviews were recorded:

• Patient data:
  • Sex and age;
  • Number of items prescribed;
  • GP surgery.
• Medicine interventions:
  • Medicine stopped, switched, added or quantities synchronised to match with patient usage;
  • MedOptimise automatically calculated the cost differences for each change. The medicine prices in MedOptimise were derived from the NHS dictionary of medicines and devices. The prices were updated weekly.
• Administration functions:
  • This included a range of information, such as any blood test requests; changes to allergy status; any counselling given to the patient; whether the patient’s inhaler technique had been checked (if they had asthma); whether any information leaflets were given to the patient; and whether the patient was referred to another healthcare professional;
  • No costs were associated with these actions with the exception of ordering blood tests, which were agreed with the local acute trust provider and updated annually.
• GP appointment time:
  • The cost linked to an avoided GP appointment was documented. The value of a consultation was obtained from the Personal Social Services Research Unit at an average of £30 per 11.7 minute GP consultation (excluding direct care staff costs)^[11]
    . 
• Hospital admissions avoided:
  • Actual or virtual (potential).

Actual and virtual hospital admissions avoided

The ability to record avoided hospital admissions — both virtual and actual — on MedOptimise was only available after April 2018. Prevention of an actual hospital admission was recorded as when a patient was deemed as being medically unstable during a consultation and at high risk of being admitted to critical care without immediate intervention, but stable enough that they did not require emergency medical attention. For example, if a patient was found to have symptoms such as:

• Chest pain;
• Breathlessness;
• Uncontrolled epilepsy;
• Syncope;
• Uncontrolled heart rate^[10]
  .

Virtual hospital admissions were regarded as ‘potential’ hospital admissions. In these cases, patients were stable but at risk of a hospital admission in the long term if no interventions were made; these virtual admissions included scenarios where patients with diabetes were taking the wrong type of insulin, or when a patient’s methotrexate dose was not being recorded on their notes. The proposed savings for a virtual and actual hospital admission were derived from the ‘Payment by Results’ tariff costs for non-elective admissions (i.e. those that have not been arranged in advance)^[12]
. The virtual admissions also factored in annualised numbers needed to treat (NNT) data. The NNT was obtained from evidence (e.g. National Institute for Health and Care Excellence [NICE] guidance).

As the cost avoidance saving from a virtual hospital admission was much less than an actual hospital admission, there was a need to ensure the correct type of admissions was recorded into MedOptimise. Therefore, all cases of hospital admission prevention were discussed and peer reviewed at bimonthly clinical meetings attended by senior clinical pharmacists within the CCG.

Results

While the clinics were running, the MedOptimise software was updated (i.e. the functionality to record actual and virtual hospital admissions was added). Therefore, the ability to record information changed across the evaluation period, resulting in anomalies in some figures.

Overall cost savings

The gross medicine cost savings from the actioned interventions of polypharmacy reviews in the 2017/2018 financial year (i.e. between April 2017 and March 2018) totalled £50,766.63 (see Table 1). During this period, 370 patients were reviewed across the eight GP surgeries. The gross cost savings from April 2018 until December 2018 were £17,942.01. During this period, 209 patients were reviewed across six surgeries, with the other two surgeries dropping the clinics owing to lack of capacity.

NICE guidance recommends that people on multiple medicines should have a structured annual review; by running a polypharmacy clinic, pharmacists can help save time and money for GPs^[13]
. The polypharmacy review clinics saved 579 GP surgery consultations over both study periods, worth £17,370.

A total of 17 actual hospital admissions were prevented from April 2018 until December 2018, which saved £30,909 (see Table 1).

Polypharmacy review

A range of symptoms and conditions were seen in patients during the polypharmacy clinics, some of which are provided in Table 2 and Table 3 with the associated hospital admission that was prevented (as found in MedOptimise).

Medicines optimisation interventions were made during the clinics. Examples of these included:

• Reducing dosage owing to renal/hepatic impairment;
• Stopping medicines with no valid indication;
• Stepping down of medicines, such as proton pump inhibitors;
• Identifying non-adherence;
• Removing prescribed items no longer used, such as creams;
• Reviewing the quantity of catheter and/or stoma products being used;
• Discussing use of over-the-counter items;
• Drugs without recent blood test results recorded.

Interventions and recommendations were discussed with patients during reviews to ensure concordance.

Cost savings by intervention type

Savings by intervention type were extracted from MedOptimise to help plan future clinics (see Table 4).

Cost savings by age group

Savings by age group were extracted from MedOptimise (see Table 5). This analysis is useful for planning future clinics (e.g. refining target age groups to achieve the highest number of reviews and overall savings made). The majority of polypharmacy reviews were conducted with patients aged 65–74 years, but the greatest savings per review across the two study periods were made in patients aged 25–34 years.

Discussion

The overall savings in Table 1 demonstrate that the polypharmacy review clinics produced QIPP savings. In addition, the clinics identified safety issues (see Table 2 and Table 3) and prevented hospital admissions.

Greater QIPP savings were found in the  2017/2018 financial year (£50,766.63) compared with the nine-month period between April 2018 and December 2018 (£17,942.01). However, it should be noted that there were changes in the team during the first year when the only full-time pharmacist left: there was some locum cover provided for three months before a a new pharmacist started. Therefore, the number of GP surgeries running clinics was, therefore, temporarily reduced from six to two between January and March 2018 until the new pharmacist had started, completed induction and became ready to take on the clinics. This impacted on the project and the savings produced.

It is also important to note that a possible consequence of the feedback to GPs about the interventions in the first 12 months led to an improvement in cost-effective prescribing and, therefore, a reduction in potential savings from interventions seen in the subsequent 9-month period. Fewer reviews were conducted in the second study period, owing to the reduced number of surgeries running clinics, which also impacted savings made.

Table 2 and Table 3 show that the pharmacy-led clinics produced several safety interventions and reduced harm, with the effect of avoiding both potential hospital and actual admissions. The cost saving associated with stopping a hospital admission is substantial.

Table 4 highlights that the greatest cost savings were associated with deprescribing. This was one of the primary aims of the polypharmacy clinics (i.e. to reduce inappropriate polypharmacy when possible). Reducing overordering of medicines had the smallest impact on cost savings.

Table 5 shows that most patients reviewed were in the 65–74 years and 75–84 years age bands. There were fewer reviews undertaken with patients aged over 85 years, potentially because there are fewer people at that age registered with the GP surgeries or perhaps owing to frailty and being unable to attend the clinic.

Patients aged 25–34 years produced the highest savings per review. However, only two patients were reviewed in this age group and one change in an expensive medicine could inadvertently skew the figures.

The savings per review for both periods (April 2017–March 2018 and April 2018–December 2018) increased in patients aged over of 55 years. This is to be expected because as people age, they tend to develop polymorbidities and have more medicines prescribed for those associated conditions. As such, conducting polypharmacy reviews in patients aged over of 55 years will produce the highest QIPP savings.

Practical considerations

Scoping patient groups

When requested by the surgery, patients outside of the original scope of the project were booked into the clinics. The conditions for this were that the patients were taking more than four medicines or had issues with their medicines (e.g. those with long-term conditions, high frailty index score, they were referred by their GP or the patient requested a review). GPs were required to review all severely frail patients as part of their General Medical Services contract^[14]
. It was necessary to compromise and accommodate these requests to obtain good engagement and enhance working relations.

Targeting patients with a high anticholinergic burden was suggested, with the intention to reduce the risk of side effects. This was included as part of the review, but was not used in the original patient search criteria.

Booking appointments

One of the initial barriers encountered was the process of booking patients in at the practices. There was variation among the practices: some were willing to identify patients and book them in for the team, but other practices did not want any involvement or additional work. In the latter situation, the MOT had to identify and invite patients to call and book themselves into the clinic.

It is estimated that around half of patients invited to the clinics attended. Anecdotally, it was found that sending out letters of invitation, along with a leaflet explaining what the clinics did, and calling from a list of identified patients had equal success rates for patient attendance. It was necessary to monitor the clinic lists weekly to ensure appointments were full.

Timing

Patients were initially allocated 30 minutes for each appointment, but this was shortened to 20 minutes in some practices as it offered a limited number of slots. For patients taking ten or more medicines daily, this made completing the full 7-Steps review and documenting it in the time available challenging.

Pharmacists took time to review the patient notes prior to the appointment to prepare for the review. In addition, adding the interventions onto MedOptimise took further time, but gradually the pharmacists became more skilled at using the 7-Steps tool and conducting consultations within appointment times.

Pharmacist competencies

Although pharmacists followed the SOP, depending on their skill set, there may have been a different degree of emphasis placed on the review. For example, pharmacists with a community pharmacy background may be more aware of medicine costs and so could produce higher medicine cost savings, while a practice-based pharmacist may be able to identify more medicine safety issues and prevent hospital admissions. There were meetings held for peer reviews of interventions, which helped to support the consistency of the reviews, but a more robust training package will be needed for the clinics moving forward.

GP expectations and response

After each clinic, a list of actions was either produced for GPs or tasks sent using the computer systems at the surgeries for the GPs to review. The response time and how the GPs dealt with the actions varied. Some GPs preferred to action changes themselves, whereas others asked the pharmacists to carry out the changes once they had authorised them. Similarly, some surgeries wanted the pharmacists to reauthorise the medicines on repeat for another set time period once the review was completed, but some surgeries wanted them left in order to prompt the regular medicine review by the doctor.

If actions identified that blood tests or monitoring was due, some surgeries wanted the pharmacists to fill out the blood test request form themselves. The pharmacist would then give them to patients or refer them to book in with a healthcare assistant for tests to be done. It is acknowledged that although time may have been saved in reducing the number of appointments required for medicine reviews, some time was required for the doctors to review interventions and follow-up, such as checking blood test results or monitoring the effect of changes to medicine. This time requirement could be reduced by using pharmacist prescribers.

Legality

The MOT had initial concerns about who took clinical responsibility for changes made and reauthorisation. The legal position of the clinics, the type of interventions made and actions carried out as covered by the SOP were discussed with a pharmacist indemnity insurance company. The advice given from the company was:

1. Reauthorising a repeat medicine is stating a medicine regimen is clinically suitable at that moment in time. If the patient’s condition changes, the medicines must be reviewed at that point and amendments made if appropriate. Neither the pharmacist nor the CCG will not be held responsible for future unforeseen issues.
2. Reauthorisation is not the same as prescribing. The decision to prescribe a particular medicine was done by a prescriber after a clinic assessment of the patient’s condition. Reauthorisation accepts the diagnosis and treatment. The review is to ensure the medicines are still suitable on this basis and to see if any changes are required in light of any new information (e.g. compliance, patient preferences, new guidelines, cost-effective alternative medicines, blood levels, symptoms and side effects). Ultimate responsibility for the prescription lies with the clinician who signs it.
3. Responsibility for the overall care of the patient remains with the GP, who has essentially delegated the medicine review to the pharmacist (see General Medical Council guidance on delegation of care^[15]
   ). The pharmacist reviewing must work within their own area of competence and in-line with their CCG’s polypharmacy review protocol. It is recommended there is a clinician at the practice that can be contacted for queries when outside their scope.

Prior to running the clinics, the authors advise pharmacists and pharmacy teams to check their legal position with their indemnity provider and employing organisation.

Feedback

The pharmacists involved received verbal feedback from patients about the reviews. These were positive and most were appreciative of the time taken to go through their medicines. Patients felt they gained a greater understanding of what they were taking and why they were prescribed their medicines. Providing feedback forms to seek patient opinion of the review, and contacting patients who did not attend their appointments to clarify the reason for non-attendance is an action that should be considered for future clinics.

Anecdotally, the GPs were in favour of the clinics, but some did not like the additional work generated and the use of a clinic room at the practice. A feedback form for the GPs to complete that outlines their experience of the polypharmacy review clinics would be useful in future projects.

Limitations

The opportunities to deliver health promotion advice were incorporated into the review by the pharmacist. For example, patients were given information on healthy eating and management of glucose levels if they had diabetes; advice was given on acute kidney injury if the patient was taking ACE inhibitors, metformin, NSAIDs or diuretics; appropriateness of inhaler devices and usage techniques were assessed if the patient had asthma; and CCG campaigns promoted as appropriate (e.g. self care, Pharmacy First, appropriate use of antibiotics and managing medicines waste). However, the impact of such advice is difficult to capture and quantify.

There were some staff changes during the first study period, which meant that there were fewer clinics run for several months while a new member of staff was inducted.

The sample size was limited because some of the GP surgeries approached did not want to run clinics. This was either because they had their own practice pharmacist (who already ran clinics) or the GPs completed the medicine reviews during regular patient appointments and so did not see a potential benefit.

Refusal of interventions/recommendations made from the reviews was captured at the time, but not included in the savings figures. Therefore, the results only reflect actual savings made and not potential savings based on all suggested interventions.

Follow-up of non-attending patients and the outcome of interventions were not always completed owing to time constraints and the limited capacity of the CCG team.

Future developments

The clinics could be further developed by:

• Running them on a regular basis for patients with long-term conditions. This would enable issues to be followed up by the pharmacist, rather than being passed over to a practice-based healthcare professionals. This could allow for better continuity of care and build a closer relationship with the surgeries and patients;
• It would be beneficial if the MOT pharmacist was an independent prescriber with competencies in several clinical areas (e.g. diabetes, cardiovascular and respiratory), so the medicine changes could be implemented by the pharmacist;
• Specialist clinics run by pharmacists could be created for patients who require reviews relating to the long-term management of their medicines, such as reducing analgesics or reducing hypnotics;
• Developing training for the prescribers at the surgery based on the common themes that emerged following interventions;
• Upskilling of pharmacists to allow simple checks (e.g. blood pressure checks, spirometry and pulse) to be carried out, preventing the patient from having to book a separate appointment or rely on outdated results;
• Conducting a baseline survey using structured questions for patients and practice staff to obtain an objective assessment of the service. This should be done before starting clinics and at 12-weekly intervals to get a long-term view of their benefits. A similar survey should be completed for pharmacists to seek their opinion, degree of job satisfaction and potential further improvements;
• The inclusion of domiciliary medicine reviews to capture patiets who are housebound (e.g. those with a high frailty index) or live in residential care. This will help support practices to comply with the GMS contract where all patients with a severe frailty index should have a medicine review^[14]
  .

Conclusion

Switching to formulary choice medicines and reducing unnecessary repeat medicines has proven to be an effective way of reducing the prescribing budget. Conducting in-depth medicines reviews, and ensuring medicines were used appropriately and in line with evidence-based guidelines improved the quality of prescribing. Hospital admissions were prevented, creating significant savings for the CCG and improving patient outcomes. The clinics included actions to improve patients’ understanding of their medicines and empower them to self-manage their conditions.

The CCG MOT clinics were limited to once per month because they needed to be balanced against other CCG QIPP priorities. To gain additional benefits and further improve the outcomes, it is recommended the clinics are run more regularly (e.g. weekly) with pharmacist independent prescribers. This will allow for better continuity of care, the follow up of issues and a closer relationship with the surgeries and patients.

The pharmacists involved in the project felt the reviews were worthwhile; enhanced job satisfaction by providing engagement with patients and surgeries; and raised the profile of the CCG in a positive manner.

Creating such a service in a GP surgery requires advance planning and the ability to manage unforeseen issues. Pharmacists need to be able to negotiate with the surgery staff to produce a clinic that caters for polypharmacy patients, but fits in with the different surgery systems and priorities. This can be time consuming, but failure to agree, monitor and chase up appointments can lead to empty clinics and reduced benefits (e.g. patient and cost-saving benefits).

Medicine review clinics give an opportunity to reduce inappropriate polypharmacy, decrease prescribing costs and enhance quality while improving patient outcomes. This project demonstrates the benefits that pharmacists can contribute to the vision of clinical pharmacists in the ‘NHS long-term plan’. Giving adequate support and training to pharmacists can show considerable benefits to patient outcomes and may reduce GP workloads.

• This article was amended on 16 December 2019 to clarify that GPs were required to review all severely frail patients as part of the General Medical Services contract.