The Pharmaceutical Journal Vol 265 No 7120 p656-660
October 28, 2000 Original Papers

Providing pharmaceutical care using a systematic approach

By Janet Krska, PhD, MRPharmS, MCPP, John A Cromarty, MSc, MRPharmS, Fiona Arris, MSc, MRPharmS, Deborah Jamieson, BSc, MRPharmS, and Denise Hansford, PhD, MRPharmS

AIM • To investigate the feasibility of providing pharmaceutical care to individual patients in primary care.

DESIGN • Structured delivery of pharmaceutical care involving the assessment of 332 patients for pharmaceutical care issues, formulation of a pharmaceutical care plan and the implementation and monitoring of these plans in 168 patients. Patients included in the study were aged 65 years or over, were taking four or more medicines regularly and had at least two chronic diseases.

SETTING • Six randomly selected general practices in Grampian.

OUTCOME MEASURES • Epidemiological data on drug use, patterns of disease and pharmaceutical risk factors; pharmaceutical care issues (PCIs) identified in all patients; issues resolved as a result of the pharmaceutical care process in a randomly selected representative sample.
RESULTS • 237 patients (71.4%) had at least one pharmaceutical risk factor predisposing them to either increased toxicity or reduced efficacy. All patients had at least two PCIs. The number of PCIs was positively correlated with the number of medicines being taken (p<0.001), number of chronic diseases (p<0.001) and the number of pharmaceutical risk factors (p<0.005). Most PCIs were identified from the prescription (52.4%), with 18.2% from the medical record and 29.4% from the patient. Implementation of pharmaceutical care plans in 168 patients resulted in the resolution of 79% of PCIs, particularly those relating to monitoring, dose discrepancies, repeat prescriptions not required, potential adverse drug reactions and the need for education.

CONCLUSION
The study demonstrated the feasibility of delivering pharmaceutical care using a systematic approach involving access to medical records and the patient. The process enabled the identification of large numbers of PCIs, most of which were resolved by pharmacist input.

There have been an increasing number of reports of pharmacists working closely with general practitioners (GPs) to improve prescribing and to provide direct services to patients.1–7 Scottish GPs considered pharmacist review of individual patients’ medication as being desirable8 and there is evidence that this particular activity is developing rapidly.9–11 Review of medication is an integral part of the process of pharmaceutical care,12 requiring a systematic approach, involving documentation.
The Scottish Office Department of Health Clinical Resource and Audit Group recently issued guidelines12 which provide, among other things, a clear, systematic approach to the pharmaceutical care of individual patients in primary care. Key definitions taken from the CRAG guidelines are shown in the “Definitions” panel.
The three stages of the process involved in delivering pharmaceutical care to individual patients described in the guidelines are:

• Assessing patients for pharmaceutical care issues
• Formulating a pharmaceutical care plan
• Implementing and monitoring the pharmaceutical care plan

This paper describes a study undertaken in Grampian in which clinically trained pharmacists provided pharmaceutical care as described in the guidelines, and it illustrates the type of activities involved. The work focused on the elderly taking multiple medicines, since it has already been shown that patients in these categories are most likely to have pharmaceutical care issues, such as potential or actual medication related problems.2,3,5,13

Method

Ethical approval was obtained for the study. Six general practices were selected at random from the 90 in Grampian after exclusion of those with fewer than 500 patients aged 65 or over and stratification for deprivation and fundholding status. This ensured that two major factors likely to influence prescribing were controlled for and that the practices selected would have sufficient suitable patients. Within each practice, patients aged 65 years or over who were regularly requesting at least four medicines via the computerised repeat prescribing system and who were suffering from at least two chronic diseases were identified. Patients suffering from dementia were excluded as were any whom the GPs felt could not cope with the study. A maximum of 70 patients from each practice, selected at random from those identified, were invited to participate by letter. Verbal consent was obtained by telephone.

Assessing patients for PCIs Information was obtained from the practice computer records about prescribed medicines, disease states and monitoring parameters and an assessment of pharmaceutical risk factors was made. The patients were then visited at home and interviewed to obtain further information about their use of medicines, both prescribed and purchased, any problems in obtaining or using these, and their responses to the medicines, both efficacious and toxic.
A specially developed checklist of common side effects was also used to assist in making sure all relevant symptoms were covered. The information was compiled on specially designed forms into a patient medication profile and pharmaceutical care issues (PCIs) identified as part of the documentation process. The two pharmacists providing the pharmaceutical care regularly discussed the issues identified between themselves and with academic pharmacists so as to enable peer review.
The PCIs were classified using a modification of a method previously developed for this purpose.2 This classification included definitions and examples for each type of PCI to assist in ensuring consistency between users. The point at which PCIs were identified was noted as deriving from the prescription, from medical note review or from interview.

Formulating a pharmaceutical care plan For each of the PCIs identified, a desired output was documented, along with a proposed action. The list of PCIs, outputs and actions formed a pharmaceutical care plan. This was then discussed with the patient’s GP and also put into the patient’s medical records. GPs were asked whether they agreed both with the PCIs documented and with the actions recommended. Where care plans included a large number of actions these were prioritised so that in some cases the care plan could take several weeks to implement.

Implementing and monitoring the pharmaceutical care plan The actions needed to address each of the PCIs were classified according to (i) who had been involved in the action and (ii) the purpose of the action. The pharmacist then implemented all the actions with which the GPs had agreed. In cases where implementation required a further visit to the patient’s home, this was undertaken by the pharmacist.
The patients were followed up after three months to determine whether PCIs were still outstanding, whether there were any new PCIs and the reasons for PCIs not being resolved. The point at which PCIs were resolved was also noted.

Data analysis Data were analysed using the Statistical Package for the Social

Sciences (SPSS) version 6.0 and Microsoft Excel version 6.0. Medicines used were classified using the British National Formulary. Pearson’s correlation coefficient was determined to investigate relationships between continuous variables. Age, number of medicines used, number of disease states present and number of pharmaceutical risk factors were categorised to enable investigation of relationships between these factors and the presence of different types of PCI. Any probability values greater than 95 per cent were regarded as statistically significant.

Results
There were 381 patients from the six practices who initially agreed to participate, but 49 (12.9 per cent) subsequently dropped out because of ill health, hospital admission or holiday at the time of interview. Therefore, 332 patients were assessed for PCIs, and a care plan was formulated and implemented for 168 of these, selected at random and representative of the whole cohort in terms of the number and class of medicines being taken.

The data from the whole cohort provided information about the epidemiology of care issues present in a random selection of elderly patients on multiple therapy, whereas data from those whose care plans were implemented provides information about the actions required to provide pharmaceutical care to such patients. Other outcome measures are reported elsewhere.14
Assessment of PCIs The 332 patients were taking a total of 2,506 prescribed medicines, with an average of 7.5 per patient. In addition, they were taking 145 non-prescription products, with an average of 0.4 per patient. A total of 196 patients (59 per cent) were taking a different number of medicines from those listed on the practice computer: 103 were taking fewer and 93 more than listed. The most common medicines being taken, including non-prescription medicines, are shown in Table 1. The commonest disease states present are listed in Table 2. There was a significant correlation between the number of medicines taken and the number of chronic diseases (r=0.364, p<0.001). Most patients (237; 71.4 per cent) had at least one pharmaceutical risk factor predisposing them to either increased toxicity or reduced efficacy (Table 3).
A total of 2,586 PCIs were identified by the two pharmacists providing the pharmaceutical care for the 332 patients (average 7.7; range 2–21). There were no significant differences in the number of PCIs found in patients from the six different practices. The number of PCIs was positively correlated with the number of medicines being taken (r=0.552, p<0.001), number of chronic diseases (r=0.304, p<0.001) and, to a lesser extent, number of pharmaceutical risk factors (r=0.170, p<0.005). The majority of the PCIs (1,355; 52.4 per cent) were identified from the prescription, others from medical records (467; 18.2 per cent) and from patient interview (760; 29.4 per cent). The frequency of different types of PCIs identified is shown in Table 4.

Age and the number of chronic diseases were not related to the frequency with which different types of PCI were found.

Strong positive relationships were found between the number of medicines taken and the presence of care issues relating to potentially ineffective therapy, no indication for medicine use and repeat prescription no longer required (p<0.0001 for each type of PCI, c2). Suspected adverse drug reactions were also more likely to be present in patients with both high numbers of medicines and pharmaceutical risk factors (p<0.05, c2). Increasing risk factors also increased the likelihood of care issues relating to potentially ineffective therapy, compliance and the need for education (p0.05, c2).
Some PCIs involved no specific drugs (189, 7.3 per cent of all PCIs), most of the remainder involved one drug (1,538, 59.5 per cent) or two to four drugs (731, 28.3 per cent). Issues involving more than four drugs accounted for 120 issues (4.6 per cent), most of which (63) concerned medicines being obtained on repeat prescription which were not being taken, the need for education (21) or suspected/actual compliance issues (17).
Information obtained by questioning the patients showed that a total of 72 patients (21.7 per cent) had regular help with their medicines, but 39 of these (54 per cent) required only help in collecting them. This assistance was provided mostly by a member of the family (24 cases, 61 per cent) or by home care assistants, friends or neighbours (nine cases, 23 per cent) but health care professionals were involved in six cases. In total, 33 patients (10 per cent of the total study population) needed help in taking their medicines correctly. In four cases district nurses provided help and two of these filled Dosette boxes regularly. Members of the family helped in 26 cases, and home carers or friends were involved in the remaining three cases.

Formulation of pharmaceutical care plans In the 168 patients for whom a care plan was drawn up and implemented, there were a total of 1,206 PCIs. Of these, 265 were identified from prescription records and resolved on accessing medical notes. A further 98 PCIs, which had also been identified from records, were resolved by patient interview. GPs agreed with 1,155 of the PCIs identified (95.8 per cent) and with the recommendations made to resolve them in 1,053 (87.3 per cent).

The types of recommendations made by pharmacists to resolve the PCIs are listed in Table 5. These differed depending on the drugs involved and the type of PCI; for example, monitoring was recommended for 71 diuretics to resolve the issue of a potential adverse drug reaction. Changing medicines was recommended for different kinds of PCI, particularly for potentially ineffective therapy (34), potential adverse reactions (13) and was common for diuretics (17) and analgesics (15). Most of the occasions when the recommendation was to stop a drug related to use with no indication (29), inappropriate duration of therapy (15) and duplication of therapy (7). Ulcer-healing drugs (16) and diuretics (15) were the drugs most commonly recommended for discontinuation. Adding drugs was recommended for 21 cases of potentially ineffective therapy and 34 untreated indications. Advice to patients on dose (20), timing (13) and administration (13) was needed to resolve issues relating to the need for education, many of which concerned analgesics (23) and inhaled therapy (35). A total of 951 recommendations requiring action were made on care plans.

Implementation and monitoring of pharmaceutical care plans After obtaining agreement from GPs and patients, 977 separate actions were taken by pharmacists to implement the recommendations required to resolve the outstanding 843 PCIs. Some PCIs required action which involved only patients (200 PCIs; 23.7 per cent), such as the provision of advice. Most issues (588; 69.7 per cent) required communication with a health care professional and 54 (6.4 per cent) involved communication with both. The need for contact with health care professionals differed for different types of PCI. The most frequent reason for such contact involved the requirement to undertake monitoring for potential adverse reactions (151) and the need to discuss potentially ineffective therapy (109). A significant number of PCIs did not require contact with a GP, since arrangements for monitoring or changes to prescription records were done through other practice staff. However, the majority of patients (149; 89 per cent) had at least one PCI which required contact with a GP.
At three-month follow-up, 950 (78.8 per cent) of the original 1,206 care issues had been resolved and 22 (1.8 per cent) were partially resolved. A further 26 (2.2 per cent) had resolved spontaneously. Almost all issues relating to monitoring (94.6 per cent), dose discrepancies (96.4 per cent) and repeat prescriptions not required (96.4 per cent) had been resolved. A large proportion of potential adverse reactions (81 per cent) and education issues (80.7 per cent) were also resolved (Table 6).
The most frequent reason for care issues not being resolved (56 occasions) was when the GP agreed to implement the action, mostly monitoring, stopping or changing therapy, but this had not been carried out by the time of the three-month follow-up. In 44 cases, the GP did not agree with the proposed recommendation and in a further 16 they felt that the issues did not require action at that time. Most of these arose because there were large numbers of issues in individual patients requiring a prolonged period of implementation. In some cases consultant advice was felt to be needed or the changes were viewed as unnecessary or of uncertain benefit.
Unresolved care issues involving therapy changes constituted a high proportion of all issues relating to potentially ineffective therapy (28, 20 per cent), use with no indication (20, 34 per cent) and untreated indication (10, 15 per cent). The patient did not agree to implement the action in 39 cases or forgot in 20 cases, with a further 16 cases in which the change was implemented but was not successful. Other situations in which issues were not resolved involved delays in hospital appointments and reception staff changing computer records without authorisation.

Discussion
This work illustrates the feasibility of providing pharmaceutical care in a general practice setting in line with recommended guidelines.12 There was a significant reduction in the frequency of PCIs following the formulation and implementation of pharmaceutical care plans.
Pharmaceutical care issues identified, and their classification, were subject to internal peer review to ensure consistency between pharmacists. The high level of GP agreement with the PCIs identified is an indication of their validity. The types of PCI encountered in this cohort of elderly patients have been found in previous studies, both those involving the elderly3,6,15,16 and those covering a wider population.2,9,17
The need for monitoring was a major PCI, particularly for diuretics and angiotensin converting enzyme (ACE) inhibitors. Where there was no evidence of monitoring for long-term therapy within the previous 12 months, this was classed as a potential adverse drug reaction. A large number of PCIs fell into this category and were easily resolvable. The lack of adequate monitoring for both these drug classes (diuretics and ACE inhibitors) has been highlighted previously.17–19 These two drug classes were also implicated in most of the PCIs classed as potentially ineffective therapy and diuretics also featured strongly in the category of drug use with no indication. Many elderly people are treated with diuretics for gravitational oedema and/or breathlessness from various causes. These drugs have been identified as being a frequent contributory factor to hospital admission.20,21 This study highlighted the importance of reviewing elderly patients’ treatment with diuretics.
Bronchodilators and inhaled cortico-steroids were the most common drugs associated with a need for education, another area where the majority of PCIs were resolved by the pharmacist. Many issues relating to compliance, untreated indication and potentially ineffective therapy were also resolved by pharmacist intervention. The resolution of the PCIs therefore had the potential to prevent serious consequences.
The resolution of issues involving differences between dose prescribed and used and medicines no longer required is also not surprising, and confirms previous work showing that most of these problems are easily resolved by pharmacists.6,22,23 The number of patients who received assistance with taking their medicines correctly was high (33), but in only nine cases did the pharmacist consider that this would be facilitated by the use of a compliance aid. These devices are not without problems, including the time taken to fill them. Moreover, they are inappropriate for some medicines.
PCIs were found in all patients from all practices, regardless of the degree of deprivation among the practice lists and whether the practices were fundholding. Overall, the percentage of issues with which GPs agreed was high as was the percentage of recommendations accepted. Although almost all patients had PCIs requiring discussion with a GP, a substantial number of PCIs were resolvable without such contact. Thus, delegation of authority to pharmacists for requesting monitoring and changing computer records would further reduce the workload created by the delivery of pharmaceutical care. Further delegation to the pharmacist for implementing actions could have increased the number of PCIs resolved, since 53 agreed actions were not subsequently implemented by GPs.
The results also confirm previous work which found that most issues were identifiable from prescription records2 and patient interview. This shows, therefore, that many issues could be identified from community pharmacies by recourse to patient medication records (PMRs) and the patients. While there is a large number of PCIs identifiable by this means, access to medical records is necessary to identify and resolve PCIs.

Much other work has focused on review of prescription data only, with no patient interview.4,6,23,24 This is clearly a valuable means of identifying and resolving problems. However, the patient is a vital component in the medication review process. This study has demonstrated that further PCIs were identified from patient interview and that patient agreement with proposed actions was not always obtained. Given the accepted importance of patient involvement in therapy decisions,25 the inclusion of a patient interview in any medication review process, as described in the guidelines,12 is recommended. In the present study this was carried out in the patient’s own home, but similar interviews could be undertaken in the medical practice or the community pharmacy, given appropriate facilities and access to information from medical records. A community pharmacy-based service would also facilitate more frequent, regular contact, which could have again reduced the number of PCIs that were unresolved because of patients forgetting to change their medicines use. A few of the recommended changes were unsuccessful, further illustrating the need for long-term pharmacist input to enable some PCIs to be resolved.
In addition, some patients had multiple issues requiring prolonged implementation plans, resulting in some issues not being resolved by the time of the three-month follow-up. The time taken to provide this pharmaceutical care service was not measured in the present study, but would be important in planning similar services.

Acknowledgments This work was funded by Grampian Healthcare NHS Trust. We are grateful to the six participating general medical practices and their staff and to all the patients involved.

References

1. Fairbrother J, Mottram DR, Williamson PM. The doctor-pharmacist interface: a preliminary evaluation of domiciliary visits by a community pharmacist. J Soc Admin Pharm 1993; 10:85-91.
2. McGuire AJ, Silburn JN, Radley AS, Dodd TRP, Cromarty JA. Pharmaceutical care planning in a community setting. Pharm J 1996;257(Suppl):R12.
3. Beech E, Brackley K. Medicines management: (1) Domiciliary based medication review for the elderly. Pharm J 1996; 256:620-2.
4. Sykes D, Westwood P, Gilleghan J. Development of a review programme for repeat prescription medicines. Pharm J 1996; 256:458-60.
5. Naylor DM, Oxley DV. Assessing the need for a domiciliary pharmaceutical service for elderly patients using a coding system to record and quantify data. Pharm J 1997;258:479-84.
6. Goldstein R, Hulme H, Willits J. Reviewing repeat prescribing — general practitioners and community pharmacists working together. Int J Pharm Pract 1998;6:60-6.
7. Martin RM, Lunec SG, Rink E. UK postal survey of pharmacists working with general practices on prescribing issues: characteristics, roles and working arrangements. Int J Pharm Pract 1998;6:133-9.
8. Weir LFC, Cromarty JA, Krska J. Prescribing support from pharmacists to general practitioners in Scotland. Pharm J 1997;259(Suppl):R27.
9. Mackie CA, Lawson DH, Campbell A, Maclaren AG, Waigh R. A randomised controlled trial of medication review in patients receiving polypharmacy in general practice. Pharm J 1999;263(Suppl):R7.
10. Bell HM, Smyth D, Adair CG, Maguire TA, Turner K, Fitzpatrick C et al. Primary care pharmacists’ influence on rational prescribing. Pharm J 1999;263(Suppl):R58.
11. Sodha M, Dhillon S, Rajyaguru R, Shah R, Watman G. Evaluation of the role and effectiveness of pharmacists working with general practitioners. Pharm J 1999;263(Suppl):R39.
12. Scottish Office Department of Health Clinical Resource and Audit Group. Clinical pharmacy practice in primary care. Edinburgh: Scottish Office Department of Health;1999.
13. Krska J. Duffus PRS. Pharmaceutical needs assessment in general practice. Int J Pharm Pract 2000 (in press).
14. Krska J, Cromarty JA, Arris F, Jamieson D, Hansford D, Duffus PRS et al. Pharmaceutical medication review in patients over 65: a randomised, controlled trial in primary care. Age Ageing 2000 (in press).
15. Wright DJ, Chrystyn H. Residential and nursing homes: potential benefits of a community pharmacy therapeutic and biochemical monitoring service. Pharm J 1994;253(Suppl): R18.
16. Dowell J, Cantrill J, Bates F. Can community pharmacists undertake effective medication reviews in institutionalised elderly patients? Pharm J 1996;257(Suppl):R6.
17. Lau SF, Braybrook S, Robinson A, Biallas, MC, Mayne NL, John DN. Pharmacist led repeat prescribing review: medicines management in the community pharmacy. Pharm J 1998;261(Suppl):R46.
18. Kalra PA, Kumwenda M, MacDowall P, Roland MO. Questionnaire study and audit use of angiotensin converting enzyme inhibitors and monitoring in general practice: the need for guidelines to prevent renal failure. BMJ 1999;318:234-7.
19. Rhodes KA. Prescription of diuretic drugs and monitoring of long term use in one general practice. Br J Gen Pract 1992; 42:68-70.
20. Pongwecharak J. Monitoring medication related problems in the elderly and strategies for their reduction (PhD thesis). Aberdeen: Robert Gordon University; 1998.
21. Cunningham G, Dodds TRP, Grant DJ, McMurdo MET, Richards RME. Drug related problems in elderly patients admitted to Tayside hospitals: methods for prevention and subsequent reassessment. Age Ageing 1997;26:375-82.
22. Read RW, Krska J. Targeted medication review — patients with chronic pain in the community. Int J Pharm Pract 1998; 6:216-22.
23. Hawksworth GM, Corlett AJ, Wright DJ, Chrystyn H. Clinical pharmacy interventions by community pharmacists during the dispensing process. Br J Clin Pharm 1999;47:695-700.
24. Furniss L, Craig SKL, Scobie S, Cooke J, Burns A. Medication reviews in nursing homes: documenting and classifying the activities of a pharmacist. Pharm J 1998;261:320-3.
25. Royal Pharmaceutical Society of Great Britian. From compliance to concordance: achieving shared goals in medicine taking. London: The Society; 1997.

At the time of writing Dr Krska was a reader, Professor Cromarty was the national specialist in clinical pharmacy for Scotland, and Mrs Arris, Mrs Jamieson and Dr Hansford were research pharmacists at the school of pharmacy, the Robert Gordon University, Aberdeen. Correspondence to Dr Krska at College of Pharmacy practice, Barclays Venture Centre, University of Warwick Science Park, Sir William Lyons Road, Coventry CV4 7EZ (e-mail jkrska@collpharm.org.uk)