Impact statements

  • Pharmacists working in general practice can offer a range of interventions to facilitate transitions of care.

  • Although the evidence suggests that pharmacists can have a positive impact on facilitating transitions of care, limited research evidence is available from outside the USA.

Introduction

Research suggests that patients will have, on average, 4.4 drug changes upon hospital discharge [1] and 50% of adult patients experience medication errors or unintentional medication discrepancies during this transition [2]. This is a well-known risk factor for medication-related harm, a significant burden on healthcare systems globally and identified as a national and global priority area for quality improvement [3, 4]. It is estimated that 74% of potentially avoidable costs are attributed to medication prescribed at hospital discharge [5, 6] and medication being one of the three main causes of potentially preventable 30-day hospital re-admissions [7]. Hospital discharge is often a very disruptive time for patients with many factors contributing to poor understanding and adherence of medications [8, 9], increasing the risk to patient safety.

Much consideration in terms of research, policy and practice has been invested to achieve safer transition of care such as implementing digital transfer of information [3, 10, 11], and recognising the need for timely intervention pathways post-discharge [12]. Nonetheless, medication management during transitions remains a significant problem, especially for the most vulnerable patients [13, 14].

Pharmacists in collaboration with patients, other healthcare professionals and/or carers [15], can play an important role to support safe and effective medicines-related continuity of care [15, 16]. Many studies have evaluated the impact of pharmacist interventions such as medicines reconciliation [16,17,18], and medication reviews [19], in hospital and community pharmacy settings. Many systematic reviews and meta-analyses to date have reported limitations with poor and inconsistent descriptions of pharmacist interventions in such studies [17,18,19,20,21]. Some systematic reviews on medicines reconciliation in the primary care sector do not differentiate between the role of the primary care pharmacist (PCP) in a general practice setting or in a community pharmacy setting [17]. This nuance of where particular packages of care are delivered and by whom they are delivered are important pieces of contextual information about intervention implementation and delivery and crucial to better understanding of how, and why an intervention works or not [22].

Clinical pharmacy practice is evolving at pace around the world [23, 24]. Primary care is experiencing unprecedented pressures in patient health-seeking behaviours but also in managing a backlog following the COVID-19 pandemic [25, 26]. There is an accelerated training and mobilisation of pharmacists into general practice to increase capacity in this sector in the UK [27,28,29]. Healthcare systems continue to identify roles and responsibilities for these pharmacists and their involvement in hospital discharge care is highly likely [28, 29].

This scoping review sought to explore the role of PCPs in the transfer of care of patients from hospital back home to ensure ongoing policy, practice and research is informed by current evidence. PCPs for the purpose of this review are defined as pharmacists working in a general practice setting, not in community pharmacy or hospital settings.

Aim

To map the nature and extent of current evidence on the role of PCPs in facilitating patient discharge from secondary to primary care.

Objectives

  • To identify literature that investigates the role of primary care pharmacists in patients’ post-hospital discharge care.

  • To identify what study designs and types of interventions carried out by PCPs and what outcomes have been researched in post-discharge care.

  • To synthesise research evidence and identify gaps in literature to inform future practice and research.

Method

A scoping review was identified as the most appropriate methodology to map the extent and nature of research undertaken on the role of PCPs post-hospital discharge and identify knowledge gaps to inform future research [30, 31].

This scoping review was conducted in accordance with the Joanna Briggs Institute (JBI) methodology for scoping reviews [32] and a pre-defined published protocol [33].

Eligibility criteria

This scoping review aimed to include all published primary studies using both observational (e.g., case control, cohort) and experimental (e.g., randomised controlled clinical trials, quasi-experimental) methodologies. Articles were excluded if they primarily involved interventions for hospital in-patients, outpatient clinics, or home medication reviews. Studies involving paediatric and oncology patients were also excluded. Conference abstracts, protocols and case reports were excluded as these were deemed to provide limited evaluative benefit. The eligibility of studies was guided by the Population, Concept and Context (PCC) mnemonic as recommended by the JBI guidelines for scoping reviews [32].

Participants

Adult patients (aged 18 or above) recently discharged from hospital and had an intervention by a PCP, regardless of the outcome assessed, the profile of patients included or the clinical diagnosis on admission.

Concept

Articles must report interventions led by PCP for patients recently discharged from hospital.

Context

This review aimed to summarise PCP interventions carried out in a primary care/ general practice setting.

Search strategy

A comprehensive literature search was undertaken (Supplementary File 1) with the support of a medical research librarian and followed a three-step search strategy as per JBI guidelines [34]. Firstly, an initial limited search on the topic was undertaken in MEDLINE, Cochrane Library and Cumulated Index to Nursing and Allied Health Literature (CINAHL Plus). The text words contained in the titles and abstracts and index terms of relevant articles were used to develop a full search strategy. Secondly, a full systemised search strategy was conducted using MEDLINE, EMBASE, PubMed, Cochrane central register of controlled trials (CENTRAL), Web of Science and National Institute for Health and Care Excellence (NICE) Evidence from their inception until March 2022. Several websites for relevant professional organisations were also searched for grey literature relevant to the topic. These included the Royal Pharmaceutical Society, General Pharmaceutical Council, Royal College of General Practitioners, Department of Health, the UK Faculty of Public Health, and the NICE websites. Finally, reference lists of included full texts were searched for relevant articles. No date limitations were set, however studies published only in the English Language were included. The search terms used have been detailed in supplementary file 2.

All identified articles were collated into an online research tool (Rayyan) [35] and duplicates were removed. Titles and abstracts were screened for eligibility by two independent reviewers (FY and MAH) before full-text screening was undertaken. Cases of disagreement were resolved via discussion or obtaining the full text. If it was still unclear if eligibility criteria were met, a third independent reviewer (HN) was contacted. For full texts that could not be retrieved, corresponding authors were emailed to request the full text of the article. If after this stage, the full text could not be retrieved then these articles were excluded. This was to ensure that only full papers were included for a comprehensive review. Reference lists of full texts included were then reviewed by one reviewer (FY) for further relevant articles.

Data charting

Data were extracted and charted using standardised forms that were piloted on the initial 16 included articles and reviewed by the research team. The data extracted were mapped to answer the key objectives of this scoping review. During the data charting of relevant studies, an inductive content analysis approach was followed to collate the results [36]. Data extracted included: author, year, country, study design, population, concept, context, aims, methodology, outcomes and key findings, financial impacts reported, errors reported, collaborations reported, barriers/facilitators reported, and research gaps identified.

No quality or risk of bias assessment was performed as scoping reviews traditionally do not seek to assess the quality of evidence unlike systematic reviews but rather map what research has been undertaken [37].

Data synthesis

A descriptive numerical and categorical analysis approach was undertaken to examine the extent, nature and distribution of papers included in the review. Key concepts relating to the review question(s) were collated in tabular format to identify themes and synthesise the findings. To classify and summarise the type of evidence available in this field and identify further research recommendations, the PAGER (Patterns, Advances, Gaps, Evidence for practice and Research recommendations) framework [38] was subsequently followed to enhance consistency and methodological rigour.

Results

As shown in the PRISMA-ScR flow diagram (Fig. 1), the search retrieved 2271 publications. After removing duplicates (n = 915), titles and abstracts of 1764 articles were screened resulting in 42 full text articles being retrieved and reviewed for eligibility against inclusion and exclusion criteria. A grey literature search was undertaken as per the protocol and followed the same principles for screening. This yielded 11 articles for full-text screening and a citation-search of included full texts identified a further 8 studies, of which only 3 articles were suitable for inclusion. A total of 20 articles were subsequently included for the purpose of this review.

Fig. 1
figure 1

Preferred reporting items for systematic reviews and meta-analyses extension for scoping reviews (PRISMA-ScR) flow diagram

Full size image

Although no date limitation was set, all 20 included studies were published within the last 10 years (between 2013 and 2021) with half of the studies (n = 10) [39,40,41,42,43,44,45,46,47,48] published in the last 2 years (after 2020). The majority of studies (n = 17) [39, 40, 42, 44, 46,47,48,49,50,51,52,53,54,55,56,57,58] were conducted in the United States of America (USA) and one each in the UK [45], Canada [41] and Australia [43].

Of the 20 included studies, half (n = 10, 50%) used observational methods [39,40,41,42, 44, 45, 48, 50, 52, 54], 40% (n = 8) were quasi-experimental studies [46, 47, 51, 53, 55,56,57,58] and only two were randomised controlled trials [43, 49]. Most studies (60%, n = 12) were retrospective in nature [40, 42, 44, 45, 47, 48, 50,51,52,53,54, 56]. Overall, the studies were mainly conducted at a single site (n = 12) [39, 40, 42, 44, 46,47,48,49,50,51, 56, 58].

Of the 20 included studies (see Table 1), 33.3% (n = 7) [42,43,44, 47, 50, 52, 53] specified a patient population discharged on five or more regular medicines and only 14.3% (n = 3) specified those on high-risk drugs [44, 50, 53]. Two studies used a Care Assessment Need (CAN) score to assess patients who were at high risk and targeted them for the intervention [40, 55]. Seven studies [40, 41, 47, 49, 56,57,58] included patients aged 60 years or older and 14.3% of studies (n = 3) [43, 50, 58] specified those cohorts who had a primary discharge diagnosis of congestive heart failure or exacerbation of the chronic obstructive pulmonary disease, as patients with these conditions have been identified as those who experience high readmission rates [43] (Table 2).

Table 1 Characteristics of included studies
Full size table
Table 2 Summary of concepts (nature of interventions) patterning chart
Full size table

Almost all studies showed that medication review (n = 18) [39,40,41,42,43,44, 46, 48,49,50,51,52,53,54,55,56,57,58] and medication reconciliation (n = 16) [39, 40, 42,43,44,45,46,47,48,49,50,51,52,53,54,55,56] formed the main interventions made by a primary care pharmacist post-hospital discharge. Emerging roles of primary care pharmacists such as liaising with other health care professional [39, 41,42,43,44, 47, 49, 50, 52, 56], medicines optimisation [39, 48, 51, 56] and review of laboratory monitoring [39, 48], referrals [40, 48] and follow-up [48] were also evident.

Healthcare utilisation and readmission rates

The most frequently reported outcome was readmission rates, nine studies [40, 43, 47, 48, 53, 54, 56,57,58] reported readmission rates within a pre-specified time frame as their primary outcome; four studies [41, 43, 46, 53] reported readmission rates as a secondary outcome and two studies [52, 53] reported the risk of readmissions. Other studies reported on either acute care visits [50], emergency department visits or a combination to assess the impact of their intervention [41, 43, 47, 49, 51, 58]. Two studies [48, 56] reported on the time to readmissions following the intervention.

Not all studies reported the statistical significance of the outcomes, however, for those that did (n = 10) [41, 43, 44, 47, 48, 50, 54, 56,57,58], the results were inconsistent. Several studies showed a statistically significant reduction in ED presentation incidence [43] or combined readmission and ED presentation incidence [43, 47, 52, 54, 58]. In particular, the reduction in readmission rates was significant when completing the intervention, including a significantly longer time to readmission (18 ± 9 days compared with 12 ± 9 days with usual care; P = 0.015) [56]. Several other studies showed a non-statistically significant reduction in re-admission risk [52] or readmission rates post-hospital discharge [40, 43, 46, 48, 51, 56,57,58]. Furthermore, there was no statistically significant difference in outcomes when the pharmacist intervention is delivered face-to-face or over the telephone [57].

The key outcomes with statistically significant results [41, 43, 44, 47, 48, 50, 54, 56,57,58], were mapped to the intervention to review whether there was a common theme amongst them (Table 3). Overall, only ten of the 20 studies reported statistical significance in outcomes and all of those included medication review as part of the intervention. The medication review with a pharmacist varied between face-to-face or telephone medication review and all were conducted within 2 weeks of discharge from hospital varying between 2 and 14 days. The main outcomes where the significant impact was seen were categorised as a reduction in healthcare utilisation, reduction in hospital readmission rates, a longer time to readmission, improved clinical outcomes and beneficial economic impacts.

Table 3 Mapping of statistically significant outcomes against the PCP interventions
Full size table

Economic implications

From the 20 studies, eight studies (40%) reported on the financial impact of the intervention [39, 43, 44, 50, 51, 54, 56, 57]. However, only four studies (20%) evaluated the actual cost-savings [43, 44, 54, 56]. Evidence showed that there was a significantly lower total cost of care after pharmacist-clinician collaborative visits whilst also improving clinical outcomes [44]. One study showed an estimated incremental net benefit of $5054 per patient and after sensitivity analysis translated a benefit–cost ratio of 28:1 [43]. Kilcup et al. [54] estimated cost savings per 100 patients to be $35,000, equivalent to $1,500,000 in savings annually.

Collaborations with other healthcare professionals

Seventeen studies reported collaborations of primary care pharmacists with other healthcare professionals [39,40,41,42,43,44, 47,48,49,50,51,52,53, 55,56,57,58]. Contact was made with the hospital inpatient team or hospital pharmacists if the information was missing or required clarification. Often, where further support or supervision was required or the task was outside the pharmacists’ scope of practice, patient’s primary care physician was contacted through verbal, written or electronic approaches [42, 52, 53, 55]. In some studies, a physician appointment immediately followed an appointment with the PCP [43, 44, 48, 51,52,53]. A team approach was often advocated in medicines-management during transitions of care [42, 48, 50, 56, 58]. In addition to collaboration between pharmacist and physicians, communication with community-based health coaches [47], care co-ordinators [40, 58] and other healthcare providers [39, 42, 43, 51, 55, 58, 59] were also reported (Fig. 2).

Fig. 2
figure 2

Barriers and facilitators to PCP interventions

Full size image

Barriers and facilitators to PCP interventions

The main barriers identified in the studies were time demands in an already busy healthcare environment and lack of resources to provide the timely intervention. Most studies reported time to intervention of between three to seven days [39, 51, 52, 54, 56]. The time to deliver the intervention (clinic medication review with the pharmacist and documentation) substantially varied between studies and took between 45 and 90 min [39,40,41, 55]. Other key barriers reported were: acceptability of the intervention by the patients [49, 50, 56], healthcare professionals’ acceptability of recommendations [50, 53] and lack of awareness of the role or benefits of pharmacists [50, 53]. The potential financial impact on patients [54, 57] and lack of adequate payment mechanisms/incentives within organisations [50, 56], were also highlighted as possible barriers to interventions. In addition, systems and organisational barriers were reported which can be confounded by lack of accessibility to multiple electronic systems and accurate medication lists at discharge [40, 51].

Facilitators for the intervention were the value that the pharmacist skills and roles can bring and the collaboration with other healthcare professionals [40, 41]. This was recognised for complex cases during a multi-disciplinary team meeting and having a shared pharmaceutical care plan [41]. This shared care plan would enable continuity and easier access across sectors in addition to running the clinic from an established multi-disciplinary clinic to align care goals and also reduce costs [56]. One study by Slazak et al. [46] reported that having a registered nurse contact patients and complete the initial medicines reconciliation and then scheduling a follow-up visit with the pharmacist was a facilitator in the process. Furthermore, the flexibility to provide patients with a choice between face-to-face and telephone appointment was seen as a facilitator, although face-to-face visits assisted communications during transitions of care [40].

Discussion

This is the first review, to the authors’ knowledge, that maps the extent and nature of the role of PCPs in a general practice setting, specifically post-hospital discharge. The contributions of pharmacists in the transition from secondary to primary care are heterogenous and outcome success rates have been variable. Evidence has shown that both medication review and medication reconciliation feature frequently as part of the PCP intervention which corresponds with evidence about pharmacists’ roles more widely [16,17,18, 24, 63]. The PCP role is evolving to include other interventions such as: monitoring of laboratory results; management of long-term conditions and generating referrals to other health care professionals. There is growing evidence about the value of multi-disciplinary team approaches and collaborations with other practitioners to enable better patient care [25, 27] and the continued need for improved communication between secondary and primary care during transition processes [66]. We have identified the emerging of the role of the PCP to date, however there are many areas that require further research as outlined in our PAGER framework (Table 4).

Table 4 PAGER framework analysis
Full size table

Whilst it is known generally that pharmacists can have positive impacts on patients’ understanding of medications [60, 61], patient satisfaction [26] and clinical outcomes [20], few studies have looked at this in relation to PCPs role post-hospital discharge [48] and this would complement current research. Furthermore, the role of PCP involvement in motivational interviewing and adherence strategies [50] to reduce medication-related problems post-hospital discharge would be useful, especially in those cohorts who may experience poor health literacy or barriers to medication adherence. Few studies have evaluated economic impacts [43, 44, 54, 56] of PCPs post-hospital discharge, however these have reported beneficial impacts at reducing costs indirectly as some health systems have applied payment penalties for high 30-day readmission rates [50, 51, 57, 62]. These indirect cost implications and benefits of PCP involvement post-hospital discharge and in longer-term management should be further explored relevant to the respective healthcare system or certain cohorts of patients that would benefit (i.e. particular health conditions, high-risk medicines or those with specific needs like compliance aids). The lack of controlled studies and consistency in outcomes researched highlights the need for further high-quality research in this field to develop robust transition of care pathways relevant to local healthcare systems. Additionally, studies on a larger scale in different geographical locations to support wider generalisability and transferability of findings would be valuable. The identified barriers and facilitators are those commonly reported when considering the implementation of clinical services by pharmacists in primary care, including barriers to integration and lack of interoperability of digital systems [63]. Further systematic investigation of factors influencing practice would be helpful, particularly using lenses of behavioural and implementation science. This would enable designing or optimising future models of care or interventions that address the barriers and capitalise on the facilitators.

Limitations

The scoping review intended to map the evidence in this field, therefore does not allow for meta-analysis or critical appraisal of the effectiveness of the interventions studied. During the scoping review, conference abstracts and studies relating to home-visits were excluded as per the protocol which may have limited our findings. As this review was primarily conducted to inform further research relevant to the UK population, a grey literature search of only UK professional organisations and websites was a limitation as we did not search for grey literature in other countries. Furthermore, the grey literature found articles that were not eligible due to the strict nature of the inclusion and exclusion criteria in the protocol relating to publication type and primary studies of an experimental or observational nature. This may have limited any audit, evaluation studies or case reports that have been carried out in clinical practice.

Conclusion

The scoping review allowed valuable mapping of the extent of research carried out on the role of primary care pharmacists during post-hospital discharge. It is apparent that there is an evolving scope of practice which could prove valuable to a primary care based multi-disciplinary team with positive effects on patient and healthcare utilisation outcomes. Our findings highlight the gaps in evidence to date to help inform future priorities and directions of research in this area, identifying that more rigorous research is needed to establish effectiveness and generalisability of primary care pharmacist interventions.