Study design

This is a single-arm, non-randomised feasibility study using a pre-post evaluation design. Any medication changes related to PIM and other medication issues before and after medication review will be compared. An explanatory sequential mixed method approach will be used for data collection and analysis, where the qualitative findings (e.g., interviews) will be used to explain the quantitative findings (e.g., recruitment rate, implementer survey results) [29]. This study consists of two parts: (1) implementation phase of the intervention (6 months) and an (2) evaluation phase (3 months). An outline of the feasibility study including the intervention and evaluation phases is shown in Fig. 1. Ethics approval was obtained from National Healthcare Group Domain Specific Review Board (NHG DSRB) domain F (Ref. no: 2022/00491). The reporting of this protocol is guided by the standard protocol items: recommendations for interventional trials (SPIRIT) [30] (see Additional file 1: SPIRIT 2013 checklist). The Clinical Trial registration for this study is NCT05756478.

Fig. 1

Overview of the Pro-M study design

Setting

The study will be conducted at the geriatric medicine (GRM) outpatient clinic of two public acute hospitals in Singapore. In Singapore, 80% of the primary care is provided by the private sector but the opposite is observed for secondary care (e.g., outpatient settings) [31, 32]. As our study target older adults, GRM outpatient clinics providing care to older patients with multiple chronic conditions are the ideal location to trial the intervention. Furthermore, the intervention was co-created with the stakeholders from both study sites and that medication review for older adults with multimorbidity is not a routine practice. The clinic physicians are invited to refer eligible patients to the study, while the pharmacists conducting medication reviews are part of the study teams at each hospital site.

Screening and recruitment of patients

Eligible patients will be pre-screened and invited to join the study by their attending GRM physicians through convenience sampling. The inclusion criteria are as follow: aged 65 and above, current patient of GRM outpatient clinic, and taking five or more medications daily. On the other hand, patients will be excluded if they are below 65 years old, currently receiving other types of pharmacist services (e.g., medication therapy management), are unable to understand and communicate in English, Chinese, or Malay and if patient or caregiver decline to participate in the study. Participation in this study is fully voluntary, and if they agree to participate, written informed consent will be taken face-to-face by a study team member at each hospital. There is no consensus on the optimal sample size for pilot or feasibility studies and the size is dependent on the objective of the study [33]. Hence, we did not perform a sample size calculation. Instead, we consulted and discussed with stakeholders at both sites and agreed to recruit 30 patients per site for this study, which will provide sufficient insights into the implementation process.

Intervention specification

The prototype for the intervention evolved from multiple scoping reviews and modified Delphi studies and was finalised through a co-creation exercise with stakeholders consisting of geriatricians and pharmacists at both study sites during phase 1. In summary, various theoretical frameworks and taxonomy, such as Theoretical Domains Framework (TDF), Behaviour Change Wheel (BCW), and Behaviour Change Techniques (BCTs), were employed at different time points to develop the intervention [34,35,36]. The detailed account of the intervention development process has been reported in another publication [23]. The main components for this intervention are: (1) pharmacist-facilitated medication review with feedback and recommendation to physicians and (2) physicians communicating medication changes made by other prescribers when needed. Table 1 shows the process of operationalisation of the BCTs in the Pro-M study.

Table 1 Operationalisation of BCTs in Pro-M study

As our intervention was developed with the considerations on integrating medication review into routine outpatient care, variations in the delivery format between the sites were accommodated, due to the operational context and preference in practice at each site. For instance, there are two modes of recruiting eligible patients at both sites. One site has the option of calling patients beforehand to introduce the study, whereas the other site will mainly recruit patients on-the-spot during the day of appointment. Patients who consent to participate in the study will undergo a one-time medication review with a pharmacist. Medication review will be conducted either in-person or through a tele-med consultation, using preferred prescribing tools of choice (e.g., Beers criteria, STOPP/START) by the site pharmacists. Findings of any PIMs and other issues will be highlighted to the physicians so that they could take this information into consideration when making prescribing decisions during their consultations with patients. The changes would then be documented and communicated to the patients and other relevant prescribers via discussions or a memo. The feasibility of implementing medication reviews in routine outpatient practice among the stakeholders (patients, physicians, pharmacists) will be evaluated through surveys. In addition, in-depth interviews will also be conducted with selected implementers (physicians and pharmacists) to understand additional barriers experienced in the implementation process.

Outcome measuresPrimary outcome and data collection

The primary focus of this study is to assess the feasibility of implementing the physician-pharmacist collaborative care intervention from the stakeholders’ perspectives, which will be reported qualitatively. Surveys will be conducted to explore stakeholders’ attitudes and experiences during the intervention. In-depth interviews will follow with selected implementers to elicit insights as well as qualitative interpretations to the survey findings.

A short 11-item patient survey (see Additional file 2: Pro-M patient survey) will be administered after the intervention to measure patients’ attitudes toward acceptance and appropriateness of the intervention. They will also be asked if they will be willing to pay for medication review as part of their routine care in the future. On the other hand, physicians and pharmacists involved in the intervention will be invited to participate in the evaluation phase. To achieve this goal, a 25-item implementer survey was developed using Proctor’s implementation framework (Penetration, Appropriateness, Acceptability, Fidelity, Feasibility, and Sustainability) [28] and each item will be measured using a 5-point Likert scale (see Additional file 3: Pro-M implementer survey).

In addition, some quantitative data, such as time taken to conduct the medication review, number of discussions between physicians and pharmacists, number of communication efforts by the physicians, and the number of agreements between the physicians and the pharmacists on PIM and other medication issues identified, will be collected alongside the intervention. This quantitative information will be triangulated with the survey and interview results to provide a fuller picture on the feasibility of the intervention.

Secondary outcomes and data collection

To provide information for sample size calculation for the next phase scale-up study, the average number of PIMs per patient before and after medication review and the prevalence of PIMs among patients will be collected. For the purpose of this study, prevalence is defined as having at least one PIM prescribed. Medication data on identified PIM and other medication issues will be collected over two time points: once during the medication review and a retrospective data collection from patient’s last GRM appointment. The retrospective data will form the comparison and represent ‘usual care’ where medication review is only done in an ad-hoc and informal basis, if any. This comparison will help to ascertain if medication reviews impact physicians’ prescribing decisions in any way.

In addition, a basic cost analysis will be conducted to compare the cost of manpower for medication review and the cost of PIMs or other medications issues identified before and after the medication reviews. Manpower cost to conduct medication reviews will be estimated by the time needed to a conduct a medication review. Cost of PIMs and other medication issues identified and resolved will be collected to estimate the monthly cost savings from discontinued medications and any substituted medications, if there are [21]. The unit cost of medications will be calculated using the private costs of medications before government subsidies. The cost information collected will provide insights on the longer-term sustainability of implementing medication review in routine care for older adults with multimorbidity.

Data analysisQuantitative data

Descriptive statistics will be used to report patients’ characteristics, the prevalence of PIMs and other medication issues identified, and results from the patient and implementer surveys. We plan to observe the direction of impact of medication review on PIMs and other medication issues identified. In addition, a paired t-test for the difference in two means will be used to determine any changes in PIMs before and after medication reviews, which could be used to estimate the effect size for sample size calculation during the next phase scale-up study.

Qualitative data

The survey findings will be supplemented with findings from the semi-structured in-depth interviews. The interviews will only be on a selected pool of implementers (physicians and pharmacists) from both hospitals and will be audio-recorded and transcribed for analysis. The transcripts will be coded using a hybrid approach of inductive and deductive coding. The codes will be examined and analysed for key themes.

Data Management and protection

Informed consent will be taken from all participants involved either during the intervention or the evaluation phase of this study. All participants (patients, physicians, and pharmacists) will be given a unique subject identifier (ID), and the data collected will be anonymised. Informed consent forms, completed surveys, and audio recordings and transcripts from interviews will be kept in a locked cabinet, within an access-restricted office or stored on a secured network at the research sites. The findings from this study will be disseminated through platforms such as conferences and journal publications.