Aim
To examine whether interventions made by community pharmacists in a community hospital were appropriate and to compare the effectiveness of an external peer review with an expert panel as a quality assurance mechanism.
Design
An 8-month prospective study in which interventions were documented on a specially designed and piloted form. An external peer group (trust principal clinical pharmacist and health authority pharmaceutical adviser) and an expert panel (clinical pharmacologist, senior hospital pharmacist and experienced community pharmacist) then assessed these interventions.
Subjects and settings
Patients on a general ward and an elderly assessment ward in a community hospital. Participating community pharmacists were based near the hospital.
Outcome measures
Interventions were considered in terms of safety, quality of life, value for money and adherence, and graded with regard to their significance.
Results
There was no difference between the expert panel and the peer review group in the overall grading of interventions, although the peer group interventions related significantly more to safety. The study assessed the interventions made by the community pharmacists and showed that those made were appropriate, worthwhile and had a positive impact on patient outcomes of care.
Conclusions
Community pharmacists with appropriate training and support can provide a clinical service in community hospitals.
This study was designed to determine quantitatively whether community pharmacists could provide a clinical pharmacy service in a community hospital setting. The principle was to improve communication across the interface for the benefit of patients. As general practitioners (GPs) provide medical services for the hospital, it seemed logical for community pharmacists to provide some hospital pharmaceutical services locally.
The study ran for eight months on a general ward in a community hospital, which was part of an acute trust. The hospital site was chosen because it provided a wide variety of cases and was near the community pharmacy involved in the study. Beds on the wards (n=55) were managed by GPs on a day-to-day basis.
We have previously set up and conducted a qualitative study that demonstrated the feasibility of community pharmacists assisting patients with their medicines on discharge from this community hospital.1
That study showed that there is an appreciation of this service among patients and health professionals. The service complemented the work of the other health professionals, and promoted teamwork locally.
Berwick infirmary is a community hospital and is part of an acute trust which provides a wide range of pharmaceutical services to most of the population of Northumberland. It is situated 63 miles from the pharmacy department in the district general hospital which provides its drug supplies.
Studies conducted in both hospital and community settings have indicated the need for patients to receive information on medicines they receive on discharge, as medicine-related problems which occur on discharge from hospital are well documented.2-6
Medicine problems have been reported as being responsible for 10-30 per cent of admissions of elderly and medical patients to hospital.5
Giving patients information on discharge can therefore have benefits regardless of the patient’s level of adherence. Problems which may occur include: duplication of treatment, reverting to previous medicines or taking inappropriate combinations of drugs.
Any of these issues may lead to failure of treatment, drug-related morbidity and, possibly, readmission to hospital. Studies have suggested that there is discontinuity between drugs prescribed on discharge and those being taken two to three weeks later in up to 90 per cent of discharged patients.4,5
Research has shown that counselling a patient is likely to be most useful when it lasts at least 15 minutes, and is accompanied by written information personalised to the patient.7
This project was developed to quantify the benefits of community pharmacists providing a clinical pharmacy service to community hospital, to examine the need for external peer review as a quality assurance mechanism, and to examine readmission rates, patient mortality, patient attitudes, information sheet usefulness and health care staff attitudes.
This paper will concentrate on the peer review model examined. Future papers will provide results to the other research questions above.
Aim The aims of the study were to examine (i) whether interventions made by community pharmacists in a community hospital were appropriate and (ii) the effectiveness of an external peer review compared with an expert panel as a quality assurance mechanism.
Methods
Medical, nursing and administrative staff from the infirmary were informed of the project, and were actively involved in establishing the operational protocols necessary to ensure successful completion of the work.
The two participating community pharmacists (who had previously worked in an NHS hospital) received additional training. This consisted of an orientation programme organised by the trust principal pharmacist together with ward visits and an introduction to the systems in place at the base hospital. The pharmacists were accompanied on visits to the base hospital so that the types of intervention that occurred could be described. In addition, the pharmacists were provided with mentoring on the infirmary wards before the study began in order to familiarise them with the working practices at the base hospital and the community hospital. The accompanied visits at the community hospital helped to establish the community pharmacists as part of the secondary care team. Ongoing support was provided to the community pharmacists, usually by telephone by the principal pharmacist. Both the community pharmacists were experienced pharmacists with sound clinical experience. Training issues centred more around encouraging the challenging of prescribers (as is routine in secondary care).
Patients were included in the study if they were taking one or more medicines and were able to give informed consent. Some patients were excluded if they were discharged at short notice.
All interventions were recorded using a record sheet developed from a template used at the district general hospital. This modified data collection sheet was piloted and amended accordingly prior to the active phase of the study. Data collected included current drug therapy, relevant medical history, basic demographic details, details of the intervention, and outcome of the intervention (accepted or rejected by doctor).
The expert panel consisted of a clinical pharmacologist, a senior clinical pharmacist and an experienced community pharmacist. The panel members were well regarded locally and had an understanding of the local situation. The peer review group included the trust principal clinical pharmacist and the health authority pharmaceutical adviser.
The expert panel was asked to consider the interventions in terms of safety, value for money, quality of life, and adherence. Interventions were then graded: very significant (A), significant (B) or not significant (C). The panel was also instructed to identify those interventions which could be considered deleterious to patient well-being.
Funding was obtained from the NHS “Seizing the opportunities” development fund. Money received was matched by Northumberland health authority. Before commencing the study, ethical committee approval was obtained. In addition the trust drug and therapeutics committee approved the work in order to ensure that issues of liability and indemnity were considered and clarified. Consent to take part in the study was obtained by the pharmacist at the time of intervention. The participating pharmacists were registered under the Data Protection Act.
Results
The mean number of drugs prescribed per patient was six (range 2?16). Fifty-seven interventions, relating to 44 patients, were made during the study period. The prescriber accepted 54 (95 per cent) of these. Gender was recorded in 31 interventions (54 per cent); of this group 21 (68 per cent) were female and 10 (32 per cent) were male.
The overall grading of the interventions by the expert panel and the peer group are indicated in Table 1 (P>0.05, c2 test). In 74 per cent of the interventions (42/57) there was agreement between the panel and the peer group. Both the panel and the peer group had exactly the same response for 47 per cent of interventions (27/57). The expert panel graded 25 per cent of the interventions (14/57) higher than the peer group. Correspondingly, the peer group graded 28 per cent of interventions (16/57) higher than the panel. In 15/57 cases (26 per cent) there was agreement that the intervention should have been made, but there was a marginal difference in the grading of significance (A or B). Neither the expert panel nor the peer review group considered any of the interventions which were made by the community pharmacists to be deleterious to the patients’ health.
Although there was no significant difference in the overall grading of the interventions between the expert panel and peer group, the peers had significantly more interventions related to safety (p<0.005, c2 test). There was no significant difference in the value for money interventions between the expert panel and the peer group.
| Table 1: Comparison of overall gradings given to the interventions by the expert panel and the peer review comparisons | ||
|---|---|---|
Significance ofintervention | Panel | Peer |
A | 11 | 14 |
B | 35 | 33 |
C | 11 | 10 |
Total | 57 | 57 |
Intervention times The average time for the expert panel to reach agreement on the significance of interventions was 2.7min (range 1?12min) (Table 2). The total time taken to review the 57 interventions was 156min.
The expert panel reached consensus usually within a few minutes in all categories of intervention. However, in one case considerable debate took place on the significance of the intervention.
| Table 2: Modal time to reach agreement on each of the significance gradings | ||
|---|---|---|
Significance of | Mode time (min) | Range (min) |
A | 2.5 | 1?3 |
B | 2 | 1?12 |
C | 1 | 1?4 |
Significance of interventions The significance of interventions was given a weighting by awarding a value of 2 to category A interventions, 1 to category B interventions and 0 to category C interventions (Table 3). Improvements to safety and quality of life are the most common benefits of the pharmacists’ interventions studied.
| Table 3: Significance of interventions | ||||
|---|---|---|---|---|
Significance of intervention | Intervention | |||
| Safety | Value for money | Quality of life | Adherence |
Peer review group |
|
|
|
|
A | 24 | 6 | 18 | 4 |
B | 23 | 13 | 15 | 9 |
Total | 47 | 19 | 33 | 13 |
Expert panel |
|
|
|
|
A | 14 | 2 | 18 | 0 |
B | 22 | 10 | 21 | 5 |
Total | 36 | 12 | 39 | 5 |
Interventions rejected by doctor Examination of the interventions which were rejected by the prescriber (n=3, 5 per cent) (Table 4), shows that the panel considered the interventions to be significant and should have been accepted. This suggests that presentation of the information to the prescriber may have been unclear or ambiguous, rather than a problem with the information itself. The peer group graded only one of these rejected interventions as significant.
| Table 4: Interventions rejected by doctor | ||
|---|---|---|
Patient number | Overall score | |
| Panel | Peer |
1 | B | C |
4 | B | C |
155 | B | B |
Discussion
The study enabled community pharmacists to be seen by medical staff, including consultants, and nursing staff on the wards of community hospitals carrying out clinical duties. From this point of view it has been a resounding success as a public relations exercise: the clinical value of community pharmacists has been seen, rather than the perceived commercial aspects of their business.
The service was also well accepted by the hospital pharmacists in the acute trust. Working across the pharmaceutical primary/secondary care interface has been facilitated, as has working between the nursing, medical and pharmacy staff on the wards.
Although there was no significant difference in the overall grading of the interventions between the expert panel and peer review group, the peer group had significantly more interventions related to safety. No significant difference was shown in the value for money interventions between the expert panel and the peer group. This is of particular interest as pharmacists are often perceived as providing information on or identifying the most inexpensive interventions.
If this type of service provision is to be extended, a number of factors must be considered:
- The community pharmacists should have appropriate clinical experience
- Appropriate training, orientation and support must be provided
- Professional indemnity issues working across the interface must be addressed
- Accountability issues need to be clarified
- Documentation and communication of the interventions made should be agreed
- Clear standards of practice must be identified
In these times of constrained resources, this service has the potential to reduce hospital readmission and drug-related morbidity, thus saving NHS resources. In addition, the current and predicted workforce shortages in the profession, particularly in the hospital sector, require us to consider alternative models of service delivery and to recognise that pharmaceutical resources need to be appropriately targeted.
Conclusions
This study has shown that community pharmacists with appropriate training and support, can provide a clinical service in community hospitals. The study has quantitatively assessed the interventions made by the community pharmacists and shown that the interventions that were made were appropriate, worthwhile and had a positive impact on patient outcomes of care. The beneficial nature of this work is confirmed by the fact that none of the interventions made by the pharmacists were considered deleterious to the patient’s health (by either the expert panel or the peer review group).
We have demonstrated that an expert panel is not necessary as a quality assurance mechanism for this type of service. We can also report anecdotally that the views of the two peer reviewers did not differ significantly, nor did it take long to reach a consensus in both cases. This suggests that an experienced clinical pharmacist working within an acute setting could monitor a service with no patient detriment.
ACKNOWLEDGMENTS We would like to thank the nursing, medical and clerical staff in Berwick infirmary (especially Tweed and Dewar wards), Wansbeck general hospital, and the Primary Care Audit Group for their co-operation and help in carrying out this study.
Sheila Woolfrey is principal pharmacist (clinical services) at Wansbeck general hospital. Muhammad Asghar is pharmacist consultant in public health at Northumberland health authority. Stephen Gray is a community pharmacist in Tweedmouth, Northumberland. Andrew Gray is a community pharmacist in Berwick, Northumberland. Correspondence to Dr Woolfrey at Pharmacy Department, Wansbeck General Hospital, Woodhorn Lane, Ashington, Northumberland NE63 9JJ
References
1. Woolfrey S, Asghar MN, Gray S, Gray A. Use community staff in hospitals. Pharmacy in Practice 1999;9:155-60.
2. Binyon D, Radley AS. GP and community pharmacist attitudes to receiving a hospital pharmacy discharge plan. Procedings of the Progress in Practice Symposium. November, 1993. United Kingdom Clinical Pharmacy Association.
3. Cantrill JA, Clark CM. Discharge counselling by pharmacists: the need and the reality. Hosp Pharm Practice 1992;2:429-33.
4. Cochrane RA, Mandal AR, Ledger-Scott M, Walker R. Changes in drug treatment after discharge from hospital in geriatric patients. BMJ 1992;305:694-6.
5. Coombes J, Horne R. A checklist for medication discharge planning. Pharm J 1994; 253:161-3.
6. Slater S, West P, Mottram DR. A study into the correspondence between hospitals and GPs on patient discharge. Pharm J 1993;251(Suppl): R40.
7. Raynor DK. Patient compliance: the pharmacist’s role. Int J Pharm Pract 1992;1: 126-35.
