From hospital to community pharmacy – development of a pragmatic in-hospital service

Brühwiler, Lea Domenica. From hospital to community pharmacy – development of a pragmatic in-hospital service. 2017, Doctoral Thesis, University of Basel, Faculty of Science.

Available under License CC BY (Attribution).


Official URL: http://edoc.unibas.ch/diss/DissB_12641

Downloads: Statistics Overview


In Swiss ambulatory care, a patient’s medication is usually prescribed by the family general practitioner (GP) and dispensed by the community pharmacy. In the canton of Aargau (AG), GP and pharmacy density is lower than the Swiss average, and self-dispensing by physicians is not allowed. This canton is of interest in this thesis, as most of the projects were performed there.
Inpatient care is provided by 267 hospitals in Switzerland. They are differentiated according to size, location and teaching responsibilities. The Cantonal Hospital of Baden (Kantonsspital Baden AG, KSB) is located in AG and is the study site of most of the projects presented in this thesis. Upon admission of a patient to hospital, a best possible medication history has to be taken. Information may be obtained from many different sources such as GPs, the patients themselves or the patients’ community pharmacies. At admission, a significant portion of medication lists are incomplete or contain mistakes. Medication reconciliation (MedRec) is a structured but time-consuming approach to obtain the correct information. Upon hospital discharge, the medication list has to be updated once again. MedRec helps to identify intentional medication changes and to define a good, reliable discharge medication list. This approach may take some time, and its quality is often lacking due to the spontaneous nature of many discharges.
The ward resident physician is responsible for discharge documents and patient counselling. The discharge summary is the most complete document, and usually contains information about medication, clinical situations and follow-up. The summaries are often sent directly to the patient’s GP by postal or electronic delivery. GPs in Switzerland and abroad complain about the quality of these summaries. Also, the late transfer of summaries is a problem. In the KSB, only a short discharge summary is given to the patients. The patients could transfer it to their GP or community pharmacy, if they wanted to.
The hospital provides the discharge prescription to obtain new medication supply, and sometimes a medication chart (MC) to the patient. In the canton AG, discharge medication can only be obtained in a public community pharmacy with a prescription. That is why this system was studied in this thesis. The medication charts given to patients are often designed by health care professionals and usually have a tabular design with some surrounding information. It is known that comprehensibility is crucial to benefit of these MCs. Patients with low health literacy are especially susceptible to misunderstanding. Furthermore, it is important that the documentation meets the patients’ needs with respect to content and design. A first step should be to gain an overview of existing charts to see all possibilities of design and content. However, there is no systematic comparison of existing MCs.
Counselling upon discharge may be of low quality and impaired by time constraints. It is clear that counselling and its effect depend to a high degree on the counsellor’s personal priorities, experience, and also the patient’s need and understanding. Patients and relatives complain about the low amounts of information obtained at discharge.
When filling the discharge prescription, community pharmacists reconcile the prescription with the patient’s history in the pharmacy software. Drug related problems (DRPs), that affect many patients, may be detected this way. Through talking with the patient or their relatives, pharmacists also detect handling difficulties, inadequate package sizes, or nonadherence. Clarifications to solve DPRs are often needed, but, as physicians are often difficult to contact, community pharmacists may lack information to care appropriately for the discharged patient. To bridge this gap, different strategies have been evaluated. The hospital may provide better information, e.g. on a handover form or on the prescription. Furthermore, a liaison pharmacist could provide information on request from the pharmacies. A third option is to counsel patients extensively while they are still in hospital. With improving their own knowledge, they may later be a valuable information source.
Irrespective of the methodology chosen to optimise discharge, some important aspects have to be kept in mind. Firstly, good discharge processes are a combination of discharge coordination, information content and information transfer. All three aspects have to be addressed when discharge processes are to be changed successfully. Secondly, many studies use extensive resources for their optimisation strategies. As resources are often limited in health care and evidence on cost-effectiveness is rare, their later implementation is often impaired. Thirdly, also regarding future implementation of a service, the difference between explanatory and pragmatic approaches should be distinguished. They use different study designs, different structures and provide different results. Pragmatic trials use existing processes, resources and experiences, to estimate the effect of a service in daily practice.
Goal of this thesis
The overall goal of this thesis was to develop a service to optimise care of discharged patients by a pragmatic in-hospital service. Three different aims helped to reach the goal.
1. The first aim was to assess the views of discharged patients (Project A1). Views should be obtained about the challenges they face upon discharge. In particular, the comprehensibility of MCs should be evaluated. Furthermore, experiences of discharge counselling and supply problems should be evaluated (A2.1 and A2.2). It was an aim to ask patients if they see any optimisation strategies for discharge problems (A2.3).
2. The second aim targeted in Project B was to assess the community pharmacist’s views of hospital discharge. In this population as well, problems and possible optimisation strategies should be assessed.
3. With all of this background information, we aimed to develop a service to optimise hospital discharge (Project C). The most important aspect was information transfer from the hospital to the community pharmacies.
Projects with results
In Project A1, different MCs from hospitals, pharmacies or projects were compared. All contained brand name, strength, dosage form and a dosing scheme. In many plans, the first column contained the name of the active ingredient. However, of the 45 patients from internal medicine, surgical and dialysis ward of the KSB who were interviewed, mostly preferred brand names in first position. There was a trend that “eMediplan” was the patients’ favourite MC, but the “AMTS-Apothekenplan” was judged as the clearest MC. Also, health care professionals preferred the “eMediplan”. Patients were then asked to interpret standard dosing instructions in a MC. The abbreviation „Mo“ for the german word for morning (“Morgen”) was misinterpreted by 24.4% as Monday. 55.6% interpreted the abbreviation „Na“ (night, german = “Nacht”) correctly as before going to bed, while 24.4% would take the medication during the night or in the afternoon instead (“Nachmittag”). Electronic patient records in hospitals may generate abbreviated dosing instructions. The maximum daily dose for the dosing instructions „3x/d 1 tablet“ was correctly interpreted by 82.2% of all participants. 42.2% understood correctly the dosing instructions „max. 2 tablets max. 4x/24h“. Of 45 interviewees, 36 interpreted the expression „on empty stomach“ (the german word means the same as sober) as medication intake without food.
In Project A2, telephone interviews were conducted with 100 patients from the surgical and internal medicine wards at the same study site. Patients were called between the 2nd and 6th day after discharge to ask about their medication knowledge (Project A2.1). A combination of oral and written instruction was the most preferred method of delivery (69% of all patients), but only 55% received it that way. According to five physicians, to whom these results were presented, and who were interviewed, all patients should have received oral and written instructions. However, the patients had overall good knowledge about medication indications and the latest changes. It should also be taken into account that they reported this knowledge themselves. Asked about when they filled their discharge prescription (A2.2), 75 patients had filled the prescription within two days of discharge, and 73 had obtained all medications. There were some patients experiencing supply problems, such as unavailable medication. But of these 14 patients, there were only four patients with therapy gaps. Patients discharged from internal medicine wards or with polymedication experienced most supply problems. Interviewed physicians stated that therapy gaps seemed unexpectedly low, although the proportion of patients experiencing supply problems was higher than acceptable. Patients were further asked how hospital discharge could be optimised (A2.3). Most patients (88%) were satisfied with the general discharge process, although there was room for optimisation. Asked if communication between hospital and the community pharmacy could be a strategy, 21% agreed, but other ideas, such as bridging supply, were also suggested. The five physicians were undecided about the advantages of improved information transfer.
The Project B aimed to assess the community pharmacists’ views about hospital discharge. A mixed method approach was chosen, with a focus group of six pharmacists and a nationwide online-questionnaire sent to 1348 Swiss pharmacies. All pharmacists reported a general lack of information. Medication changes, allergies, specifications for “off-label” medication use or contact information were reported as often unavailable. This led, presumably often, to therapy gaps. Focus group participants reported extensive workload with discharge prescriptions in order to enable good and continuous patient treatment. In the focus group and the questionnaire, pharmacists emphasised the importance of more extensive information transfer. This applied especially to medication changes, unclear prescriptions, and information about a patient's care. They stated that information should be delivered in a structured way, but no clear preference for one particular transfer method was found.
The aim of Project C was to develop a pragmatic in-hospital service to optimise discharge. Within Project C1, the study design should be tested and the success of a later intervention study should be estimated. The service in the study should target information transfer from the hospital to community pharmacies, and should use the usual prescription as transfer method. The aim should be to reduce the community pharmacies’ workload and enhance patient safety. Based on a model for evaluation of complex interventions, important uncertainties and criteria were sampled which could influence quality, feasibility and efficiency of the study. The uncertainties were then assessed with the help of different piloting procedures. In the hospital, patient screening was tested with different inclusion and exclusion criteria, which were continuously specified according to the previous findings. With three screenings, good knowledge about the eligible population was gained. The recruitment tests revealed that many patients were missed, and the procedure was therefore adapted. Collaboration with the community pharmacies and their data recording proved to be feasible after some adaptations. For the community pharmacists, time constraints were a major barrier in filling out the case report form, but the pharmacies found the research question interesting.
Based on the previous findings, a pragmatic in-hospital service was studied in Project C2, a randomised controlled trial (RCT). The service was tested on adult internal medicine patients who were discharged to home. They were included if the patients gave informed consent and if their pharmacy agreed to participate. Patients were randomised and control group patients received usual care. In the intervention group, the prepared prescription was checked by a clinical pharmacist. Flaws were discussed with the physician and corrected or specified on the usual prescription. When the patient filled their prescription in the pharmacy, the staff documented the pharmaceutical interventions (PIs), the established contacts, the time needed to fill the prescription and their satisfaction level. In each group, 76 patients were included in the final evaluation and their characteristics did not differ significantly. In an adjusted Poisson regression analysis, the intervention group had a relative risk of 0.78 (95% CI 0.62-0.99, p=0.04) for the number of PIs increasing by one, compared to the control group. The comparison of the PIs showed that the pattern was different between the groups. There were less clinically significant PIs performed, but more economically significant ones in the intervention group. The number of contacts with hospital physicians by the community pharmacies was lower. The time that was needed to fill the prescription was 10 minutes in both groups and was therefore not influenced by the service. However, the community pharmacy staff was statistically significantly more satisfied with the quality of the prescriptions. The pragmatic service in the hospital took 6 minutes per patient.
A qualitative study (Project C3) was conducted after Project C2 to learn from experiences and to complement the quantitative outcomes. Five involved resident physicians and five community pharmacists from the RCT were interviewed about their general impression, the methodology and effects. Also wishes for further services were evaluated. It proved that both professional groups were positive about the involvement of the hospital pharmacy in the discharge process. The interviewees stated that patient safety was increased. Physicians were aware of the problems at transitions of care and some reported having changed their behaviour. A topic most prominently discussed during the interviews were medication changes. Both groups stated that it was appropriate to communicate through the prescription. However, pharmacists reported that the standardised addition of a medication chart would be enormously helpful. For the future, both groups would benefit from a continuation of the studied service. It would ensure that every prescription is checked by two persons and it would enhance patient safety. Residents appreciated the wider presence of the pharmacy staff on the ward, and community pharmacists reported a desire for any kind of collaboration.
To conclude, this thesis showed the following:
Patients’ views of hospital discharge
- MCs differ significantly in their design and content. The preference for the best chart differed between the hospital staff and the patients, indicating that people that design charts should be aware of this.
- MCs do not meet the patients’ needs in all aspects, and patients wish for other information items, or the same items in another order, for example brand names.
- Dosing instructions, which are commonly used in the hospitals and are therefore also prescribed at discharge, were not well understood. This may impair patient outcomes. All written dosing instructions should be accompanied by proper counselling.
- Discharged patients get less instruction than they expect. Counselling at hospital discharge was unsatisfying according to the patients, and a quarter stated that they have neither been counselled orally, nor in written form. A combination of oral and written instruction was most preferred by the patients.
- Patients reported rather good knowledge on their medication, which is in contrast to the unsatisfying counselling. However, there was no control on the correctness of the patients’ responses.
- Discharge prescriptions are filled later than expected, and a relevant portion of patients had not filled their prescriptions until the 2nd day post discharge. Therapy gaps were infrequent, but can be cumbersome and should be prevented.
- Patients were satisfied with the general discharge process. They suggested that a bridging supply would be helpful. Transfer of information to their community pharmacy was not clearly welcomed by the interviewed patients and physicians, although literature shows promising approaches.
Community pharmacists’ views of discharge
- Swiss community pharmacists rarely received sufficient information along with discharge prescriptions. They complained that many information items are unavailable, although useful. Community pharmacists estimate also that through the lacking information, they are faced with extensive workload and patients experience therapy gaps.
- To transfer information from hospital to the pharmacy, pharmacists would prefer a structured method of transfer. But no clear trend for electronic over paper based transfer was found.
- Not only patient-specific information transfer, but also general collaboration was very welcome. Community pharmacists stated a wish for more exchange with the hospital personnel, either shared courses, or practical information as contact information, about compounding or guidelines used in the hospital.
Optimising discharge by a pragmatic in-hospital service
- An optimisation of hospital discharge should be tested by a RCT. The previous feasibility testing helped to identify and assess uncertainties and criteria, which may possibly influence the study success.
- The primary outcome, the total number of PIs performed in the pharmacy, was reduced in the intervention group compared to the controls. This was also true particularly for the clinically significant ones. There were more PIs with economic significance, indicating that costs could be lowered by the pharmacies of patients who underwent the service.
- In the community pharmacy, the time needed for prescription filling was not influenced, but pharmacy staff were significantly more satisfied with the prescription quality.
- The pragmatic in-hospital prescription check and the transfer of information to the community pharmacy proved to be feasible and resource-saving. This would help later implementation.
- The physicians were highly aware of the problems at transition of care and of medication changes. They reported that their behaviour had changed. The physicians appreciated involvement of hospital pharmacists in patient care.
- Community pharmacists appreciated the hospital’s efforts to optimise patient discharge. Their processes did not change. The service should be continued in the pharmacists’ opinion. In addition to the prescription, the pharmacists desire the medication chart for appropriate pharmaceutical care.
Advisors:Taxis, Katja and Hersberger, Kurt E.
Faculties and Departments:05 Faculty of Science > Departement Pharmazeutische Wissenschaften > Ehemalige Einheiten Pharmazie > Pharmaceutical Care (Hersberger)
UniBasel Contributors:Hersberger, Kurt E.
Item Type:Thesis
Thesis Subtype:Doctoral Thesis
Thesis no:12641
Thesis status:Complete
Number of Pages:1 Online-Ressource (242 Seiten)
Identification Number:
edoc DOI:
Last Modified:08 Feb 2020 14:53
Deposited On:18 Jun 2018 09:15

Repository Staff Only: item control page