Interface Development Pharmacist
CPPE Local Tutor
It is difficult to eliminate the risk of an adverse incident such as a dispensing error happening, and even more difficult to assess and manage the risk of any resulting harm from that incident. We can build in systems that try to manage out the risk – if no incident occurs, no adverse outcomes will be seen. When medicines management is taken as a whole the challenge is to manage the risks that we as healthcare professionals introduce, both by our choice of medication and how we monitor its use. There is a fine balance between the benefits achieved with medicines and the harm that they might cause.
Reducing risk from inappropriate prescribing
Prescribing is a skill that is not taught formally to doctors. It is not surprising therefore that the quality of prescribing is variable and that risk is introduced as soon as a prescription is written. The individual prescription may be the perfect choice for the condition being treated but there are too many variables to be considered for all prescribing to be safe for all patients. Pharmacist input is essential but we need to go further than the traditional role, where interventions are made when there is an obvious risk to the patient such as an overdose. The most difficult prescriptions to query will be those that can be dispensed without question but which may still be inappropriate for the patient.
A simple example would be the prescription for an elderly patient to take diclofenac as a regular painkiller. If we have not assessed the renal function for that patient we are taking a risk and the patient is being exposed to possible harm. A recent paper looked at prescribing errors in a psychiatric hospital and found that 12.5% of errors were not in the traditional area of prescription writing.(1) Furthermore the majority of errors were in nonpsychiatric medication, implying that prescribing outside of one’s specialty adds extra risk. Perhaps the most worrying aspect of this paper is that 42% of drugs prescribed erroneously had already been administered before the error was detected! Another factor contributing to the error rate was patient throughput, with the more acute, short-stay unit having a higher error rate. This study did not look at the impact of prescribing on the patient when corrections had been made so prescribing was then presumed to be correct, but is it risk free?
Assessing the risk posed by correctly prescribed medications
Assessing patient risk from correctly prescribed medication is a new area for most of us and yet for years papers have been published looking at the relationship between drugs and hospital admissions. When we are carrying out our final discharge medication check should we not be assessing the risk of re-admission? This may be due to a variety of factors including treatment failure, adverse drug-related events, drug interactions, noncompliance and poorly monitored therapy. One paper from the Netherlands reports that 24% of admissions within the study (n=106) had a serious drug reaction and 12% were admitted because of a drug reaction.(2)
There are some simple indicators of risk including previous fall, gastrointestinal bleed or haematuria and the use of three or more drugs. In our A&E departments, admission units and orthopaedic trauma wards should we not, therefore, carry out a comprehensive risk assessment on all fallers taking three or more drugs? Falls are a serious risk to life but you will not find them listed in the side-effects for any drug listed in the British National Formulary. This is because falls are the outcome of both side-effects and therapeutic effects. The risk–benefit ratio may well indicate that by treating a relatively minor ailment or by our selection of drug we have introduced a life-threatening risk for the patient. Falls not only have serious physical implications for patients, but are also associated with loss of confidence and poor quality of life. The UK Department of Health announced in April that consideration would be given to changing diazepam to a schedule 2 controlled drug. If this results in a reduction in benzodiazepine prescribing in the elderly it can only be a good thing. So many elderly patients are now taking benzodiazepines habitually that our next challenge is to wean them off. The risk to patients does not just stem from CNS sedation, but also results from poor coordination, slowed response times, impaired attention and poor balance control. Why are both hospitals and community continuing to prescribe this group of drugs outside of the recommended guidelines within the BNF, and what can pharmacists do to reduce the risks created by such prescribing?
Targeting resources to maximise risk reduction
Limited time and resources are a major problem but there are studies that may help us target resources. We have already stated that previous fallers need to be assessed, but which other groups should be on our list? If readmission rates are our criteria then chronic obstructive pulmonary disease (COPD) and coronary heart failure (CHF) closely followed by asthma, stroke and diabetes are the best diseases to focus on.(3) The research so far does not tell us how well concentrated pharmaceutical care planning and risk reduction strategies work but preventing just a few readmissions is of major benefit to the patient and of course reduces cost to the acute sector.
We could also target particular drugs, or groups of drugs in our risk reduction strategy. An Addenbrookes hospital survey of acute admissions showed that the most frequently implicated drugs are cardiovascular, CNS and nonsteroidal anti-inflammatory drugs (NSAIDs).(4) The paper also looks at causality and severity with adverse drug reactions (ADRs) being top of both lists. Therapy failures were close behind but drug overdoses constituted a large group in the severity table.
Preventing the risk of readmission
The idea of preventing the risk of readmission by carrying out medication review and care planning is admirable, but what about the majority of patients who do not get readmitted, but whose quality of life is impaired by the drugs prescribed?
If we look at all the adverse events to patients discharged from hospitals an American paper shows that drugs are again a key risk factor.(5) A variety of events were recorded in the period immediately after discharge but of the 19% who had an event (n=400) 66% of these were drug related. Going home from hospital is therefore a high-risk period. Examples include the patient who was discharged with spironolactone as an additional drug. With no electrolyte monitoring serum potassium rose to 7.5 and the patient developed symptoms of extreme weakness and anorexia. This is a known risk but a carefully written care plan could have prevented this incident. Another patient developed b-blocker-induced respiratory problems after discharge postmyocardial infarction despite not having shown signs during the hospital stay, once again a predictable risk that only came to light at the follow-up appointment when the drug was stopped. He had not been warned and presumably had also not read the patient information leaflet supplied, but could this event have been predicted and prevented?
Some recent papers have also looked at drugs that are simply too risky to use in elderly patients.(6,7) They include predictable drugs, such as benzodiazepines, sedating antihistamines and tricyclic antidepressants but also dipyridamole, doxepin and long-acting sulphonylureas for type 2 diabetes. We need to remember that as people get older their renal and hepatic functions change so that drug levels increase. Dose adjustments should always be considered to counter unexplained ADRs for patients that may have been taking a drug for some years.
Single assessment tool
A major new UK National Health Service initiative is the introduction of a single assessment tool. An assessment at initial contact will highlight the need for a specialist pharmacist review and detailed risk assessment. Trusts will need to ensure that resources are made available to meet the demand as, without dedicated staff and time, proper detailed assessments and follow-up care planning cannot be done. The Department of Health website contains guidance on the process but there are no validated pharmacy risk assessment tools currently available. Work commissioned in London has put together a simple set of questions for use by nonpharmacists but the specialist assessment tool is a more difficult prospect.(8) Other work going on around the country includes the use of a comprehensive risk assessment tool with associated care plans with one model having been in use in Colchester for over 4 years now.
Moving risk down the ladder
We have looked at a variety of process-based risks and how we can manage out their occurrence. We must also recognise that the use of medicines is inherently a risky activity so we also need to manage these risks. It is imperative that the events described in readmission and postdischarge studies are minimised by a degree of anticipation and good quality care planning.
We need to move the risk to a more favourable place on the risk ladder. Medicines management objectives need to go beyond the concept of the right drug at the right dose and the right time for the right patient. To achieve this, pharmacists must be able to assess patients, balance risk against clinical benefit, and plan to maximise the latter while minimising the former.
- C Haw, J Stubbs. Prescribing errors at a psychiatric Hospital. Pharmacy in Practice 2003;13:64-6.
- C Manesse, et al. Contribution of adverse drug reactions to hospital admission of older patients. Age Ageing 2000;29:35-9.
- G Westert, et al. An international study of hospital readmissions and related utilisation in Europe and the USA. Health Policy 2002;61:269-78.
- N Bhalla, et al. The incidence and nature of drug-related admissions to hospital. Pharm J 2003;270:583-6.
- Forster A, et al. The incidence and severity of adverse events affecting patients after discharge from hospital. Ann Int Med 2003;138:161-7.
- Pitkala K, et al. Inappropriate drug prescribing in home-dwelling elderly patients. Arch Int Med 2002;162:1707-12.
- Zhan C, et al. Potentially inappropriate medication use in the community dwelling elderly. JAMA 2001;286:2823-9.
- Available from URL: www.rpsgb.org.uk/nhsplan/pdfs/oldnsfmedasstoolforms.pdf