Impact statements

  • Ambiguous terminologies for the process of medication review combined with a lack of agreed definitions limit the ability to compare medication reviews between studies.

  • The findings of this review will be used to underpin the development of a taxonomy to achieve international agreement on how medication review interventions are described.

Introduction

Medication review (MR) is an integral part of the practice of many different healthcare professionals and is defined as a systematic assessment of a patient’s medicines with the goal of optimising medicine use and improving health outcomes [1]. According to the available information, Pharmaceutical Care Network Europe (PCNE) classifies the medication review into three levels: type 1 includes the medication history, type 2a includes medication history and patient interview, and type 2b includes medication history and clinical data [2]. Type 3, the most advanced level of review, includes medication history, patient interview, and clinical data. The goals of MR include minimising drug-related problems (DRPs) and recommending actions [1, 3]. Activities resulting from MR include medication discontinuation, addition, changing, dose increase, dose decrease and monitoring [4, 5]. So, whilst an international definition of the term ‘medication review’ has been developed, combined with the recognition that a number of activities result from it, there are no standardised, agreed terms or definitions to describe these ensuing activities.

For example, whilst one study may describe different MR activities as discontinue a drug, add a drug, change dose, change drug, change dosage form, and change timing [4], another may use cessation of drug, addition of drug, dose increase, dose decrease, replacement of drug, drug formulation change, schedule change and monitoring [5]. The terms treatment stop, treatment start, dose adaptation, therapy switch, and change timing of administration are also used to describe similar MR activities [6]. Researchers are increasingly interested in comprehending processes within interventions to enable them to better understand the mechanism of impact (‘how the provided intervention occurred’) [7]. This is achieved by looking at fidelity (‘whether the intervention delivered as intended’); dose (‘how much of the intervention is delivered to each participant’); and reach (‘the proportion of individuals who should have received the intervention’) [7]. Therefore, a standard definition of MR activities enables the terms fidelity, dose and reach to be compared accurately. However, the existence of locally developed MR activity classifications makes it difficult to compare the different process evaluations from various countries and settings. Furthermore, when different terms are used, it is frequently without an underpinning definition, the assumption being that the reader will automatically understand the activity from the term used. Also, this issue creates difficulties in finding the medication review activities in the evidence search and synthesis.

Whilst the term ‘discontinue’ may seem intuitive, it may be used to describe either active discontinuation of ongoing therapy or removal of a medicine from a medication list because it has not been supplied for an extended duration. This is different from actively stopping a medicine after discussion with a patient and agreeing on this course of action. For process evaluation purposes, it is potentially important that active discontinuation of ongoing therapy is differentiated from discontinuation largely for administrative purposes, namely, to update records. The first activity is far more likely to have an impact on the patient than the latter.

Frequently, authors further classify MR activities with overarching terms such as clinical, technical or education [8]. Similarly, these overarching terms have no agreed, standard definitions.

The lack of standardised terms and definitions for overarching and individual MR activities creates unnecessary difficulties for researchers wishing to compare their process data with similar interventions and studies. The findings from this study will be used to develop a standardised, consensus taxonomy for overarching and individual MR activities.

Aim

The aim of this systematic review was to identify overarching and individual activity terms and definitions reported within studies, where MR was the main intervention, to inform the development of an international taxonomy.

Method

This systematic review was reported according to Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA) [9]. The systematic review protocol was developed and registered on the international database of Prospectively Registered Systematic Reviews (PROSPERO), with the registration number CRD 42,020,215,992. This systematic review was guided by Cochrane Effective Practice and Organisation of Care (EPOC). A scoping review had been performed before conducting this systematic review to identify search terms and data extraction tools, select the quality assessment tool, and review 10% of preliminary searches in order to establish inclusion and exclusion criteria; the full scoping review is attached in supplementary file 1. The search strategy is provided in supplementary file 2.

Search keywords were determined, with synonyms for medication review and for stop, start, monitor, decrease and increase identified.

Regarding participants in the included studies, no restrictions were given to age, gender, or medical condition of the patients or to healthcare professionals (HCPs). All healthcare settings were considered.

Inclusion criteria

Papers meeting all of the following criteria were included:

  • Studies that focused primarily on describing the nature of medication review interventions.

  • Studies that describe medication review interventions: clinical, technical and education interventions.

  • Medication reviews based on patients with multi-morbidity.

  • Medication reviews conducted by any healthcare professional.

  • Studies written in English.

Exclusion criteria

Any paper having any of the following exclusion criteria was not included:

  • Studies that focused on medication reviews within a single disease state such as those focusing on medication review of cardiovascular medications

  • Conference proceedings

  • Systematic reviews

  • Abstracts

  • Audits

  • Posters

  • Papers without empirical data such as protocols and editorials

  • Pilot studies

  • Grey literature studies

Databases searched were Embase (Excerpta Medica Database) (Ovid); Medline (Ovid); AMED (Alternative Medicine database) (EBSCO); PsycInfo (Psychological Information Database) (EBSCO); and CINAHL Complete (Cumulative Index to Nursing and Allied Health Literature) (EBSCO). Bibliographies of included studies were also searched. The search terms were carried out using Medical Subject Heading (MeSH) and other appropriate text words.

Search results for databases were exported into the reference manager Endnote × 9.3.3, where any duplication was identified and removed. Then, the remaining titles were exported to Microsoft Excel for titles, abstract and full-text screening. Search results were limited by two filters: publication type (journal) and the English language. An updated search was conducted in October 2021 to ensure that recent data had been included in this review.

All titles and abstracts were screened independently by two reviewers to check their eligibility against inclusion and exclusion criteria. Screening decisions were then compared between the two reviewers, and any discrepancy resolved by a discussion. A third independent reviewer was not required.

The number of screened titles, abstracts and full paper texts identified at each stage was recorded in a PRISMA diagram [9]. Inter-rater agreement was measured through Cohen’s Kappa coefficient for every stage of screening [10].

Quality assessment

An independent, duplicate quality assessment of each study was undertaken (by MA and DW). All the included studies were assessed using the Mixed Method Appraisal Tool (MMAT) which enables appraisal of different types of study designs [11].

Data extraction

In line with Cochrane Effective Practice and Organisation of Care (EPOC) [12], the following data were extracted from papers and abstracts by two independent researchers.

General characteristics of the included studies:

  • Title, author(s)

  • Year of study publication

  • Study objective

  • Study design

  • Study setting

  • Country of the study

    Characteristics of the participants:

  • Target population (e.g., HCPs or patients)

  • Participants’ sample size

Description of the main findings:

  • All terms used to describe medication review interventions conducted by HCPs in different settings

  • Description of medication review interventions

Data from eligible studies were then checked by two independent reviewers (MA and DW) to verify the accuracy and completeness of the included data. Any disagreement was resolved by a consensus between the reviewers.

Data analysis

Medication review activities’ data were themed by the research team and collated to inform the development of a medication review activities’ taxonomy. All the medication review activities were extracted. The results were then analysed by narrative synthesis.

Results

Paper selection and description

The systematic search identified 17,117 citations, of which 21 met the inclusion criteria. The process of literature identification is summarised in a PRISMA flow diagram provided in Fig. 1. The level of agreement between independent reviewers at title, abstract and full text screening was 56.7%, 89.7% and 78%, respectively, with Kappa 0.18, 0.70 and 0.55, respectively.

Fig. 1
figure 1

PRISMA flow diagram for literature review process [13]

Table 1 summarises the characteristics of the 21 papers that were included. Most of the studies were located in the UK, representing 28.5% of the total studies included (n = 6) [9, 14,15,16,17,18]; 23.8% from the Netherlands (n = 5) [5, 14, 18,19,20]; 19% from Australia (n = 4) [4, 21,22,23]; and smaller numbers (9.5%) from Sweden (n = 2) [24, 25]; 4.8% from Norway (n = 1) [26]; 4.8% from Belgium (n = 1) [5]; 4.8% from Canada (n = 1) [27]; and 4.8% from Jordan (n = 1) [28]. No specific study design was designated as this review was looking for papers for which MR activities and overarching activities had been described and defined. Accordingly, 38% of the papers were quantitative, non-randomised trials (n = 8) [5, 8, 14, 15, 20, 25, 29, 30]; 38% were randomised controlled trials (RCTs) (n = 8) [5, 16,17,18,19, 22, 23, 28]; and 24% were quantitative descriptions of medication review services (n = 5) [4, 21, 24, 26, 27]. The objectives of the studies included in this review varied and none of them described the nature of healthcare professionals’ activities, with the exception of two studies. Alldred et al. described them as stop medicine/stop drug, start medicine, dose increase/alter dose, dose reduction/alter dose, switch medicine/switch drug, alter formulation, alter timing, and test to monitor condition/test to monitor medicine [28] Petty et al. described them as stop drug, start drug, alter formulation dose, timing, switch drug, test request [15]. Papers ranged in publication date from 2002 to 2019.

Table 1 Summary of general characteristics of included studies

Quality assessment

The quality assessment used the Mixed Method Appraisal Tool (MMAT) [11] for quantitative randomised controlled trials, quantitative non-randomised controlled trials, and quantitative descriptive studies. These are shown in supplementary file 3. Thirteen papers were allocated a score of 6 (62%) [5, 5, 8, 14, 16, 18,19,20, 22, 23, 25, 28, 29]; five papers a score of 7 (23.8%) [4, 21, 24, 26, 27]; and three papers a score of 5 (14.20%) [15, 23, 30].

Descriptions and definitions of medication review interventions for clinical, technical and education interventions

Table 2 summarises terms used to describe the overarching activities which are ‘clinical’, ‘technical’, and ‘education’ interventions in the included studies.

Table 2 Summary of description of medication review activities for clinical, technical and education interventions

Eleven papers (53.3%) used term ‘clinical’ [5, 8, 14,15,16, 20, 21, 24, 27,28,29] and only one of them provided a definition for this term [15].

Four papers (19%) used the term ‘technical’ or ‘clerical’ [8, 15, 16, 29]and one of them provided a definition for this term [15].

Ten papers (47.6%) used the term ‘education’ or synonyms [4, 5, 8, 14, 15, 20, 22, 23, 28, 30] and none provided a definition for this term.

Table 2 shows that the authors agreed in their use of the ‘clinical’ term, whereas they described the technical intervention using two different terms, namely, ‘technical’ and ‘clerical.’ Furthermore, the education intervention was described in four different ways, namely, ‘educate,’ ‘advise,’ ‘aid,’ and ‘counsel.’

The authors used various terms to describe MR activities, as shown in Table 3. The term ‘stop’ has been described using four different words: ‘stop,’ ‘discontinue,’ ‘cease,’ and ‘withdraw.’ Three different words were used to describe the ‘start’ term: ‘start,’ ‘add,’ and ‘initiate.’ The term ‘increase’has been described by seven different words, namely, ‘increase’, ‘alter’, ‘adjust’, ‘change’, ‘adapt’, ‘rise’, and ‘titrate’. Dose decrease was described using seven different words: ‘reduce,’ ‘alter,’ ‘adjust,’ ‘change,’ ‘adapt,’ ‘decrease,’ and ‘titrate.’

Table 3 Summary of the terms used in previous studies to describe MR activity

Table 3 demonstrates that medication changes, formulation changes, and timing changes are described in different ways. Words used to describe medication changes included ‘switch,’ ‘change’, ‘replace,’ and ‘alter’. The words ‘switch’, ‘change,’ and ‘alter’ were used to describe formulation changes. The words ‘change’, ‘synchronise,’ ‘alter,’ and ‘adjust’ were used to describe timing changes.

Definitions of medication review terms

In seven of the included studies (33.3%) [4, 14, 15, 18, 23, 26, 27], the authors provided a description of some of the terms used for medication review as summarised in Table 4. The ‘Stop’ term was defined once (4.7%) [27], the ‘start’ term was defined four times (19%) [4, 15, 18, 27], and the ‘monitor’ term was defined three times (14.2%) [14, 23, 26].

Table 4 Summary of the definitions used in previous studies to define MR activity

Table 4 shows that some authors provided definitions for some of the terms such as ‘discontinue’, ‘add’ and ‘monitor’. However, the defined terms vary and were described by the authors without consensus.

Discussion

Main findings

This is the first systematic review to examine and propose standardisation of the way MR interventions are described. It highlights an absence of consistent terms used to define and describe such interventions. Terms used to describe instructions such as ‘alter’, ‘adapt’, ‘monitor’ lack sufficient detail to enable learning from and comparison of studies. There was limited uniformity with respect to overarching terms such as ‘clinical’ and ‘technical’, with the same activity being frequently classified differently between studies. There is a need for a consistent approach when describing medication review activities in studies.

Strength and limitations

The search strategy employed was restricted to articles written in English. This risked rejecting potentially useful studies written in another language, reporting different HCPs’ activities.

The inclusion criteria used in this systematic review, together with the wide range of databases searched, have ensured that all the related studies from the largest possible number of articles initially identified were selected. However, having a greater number of initial studies to screen can result in less agreement. The large initial pool used in this systematic review may explain the lack of agreement between reviewers in the first stage of screening.

An extensive search was undertaken to gather all possible evidence regarding medication review activities, definitions and descriptions over time because there was no publication year restriction. The study designs ranged from randomised controlled trials to service evaluations in terms of quality. The Mixed Method Appraisal Tool (MMAT) showed that most studies reviewed were of high quality.

Main discussion

The Mixed Method Appraisal Tool (MMAT) was chosen to review the included studies because no specific type of study was required, and it was expected that a variety of study types would be discovered when conducting the systematic review protocol.

It is worth noting at this point that, although they are not terms used to describe and define activities undertaken as a result of the medication review process, inconsistency was apparent to the authors in the use of the words ‘drug’, ‘medicine’, ‘medication’ and ‘treatment’. A ‘drug’ is defined as any chemical substance that affects the body's physiological process [31]. The term ‘treatment’ may refer to the use of an agent, procedure or regimen, such as surgery, drug, or exercise, in an attempt to cure or alleviate a medical issue. ‘Medicine' is defined as a formulated drug with a specific dose and dosage form used to prevent, diagnose, control, and treat disease [31]. It would seem that the latter is more appropriate. Consensus is required but it is outside the scope of this study.

The term ‘deprescribing’ first emerged in 2003 and there is a growing body of literature using it [32]. However, this systematic review showed that, despite being consistent with deprescribing, all papers researched used alternative terms under the overarching term ‘deprescribing’. This may be due to it being a relatively new term but also because it better describes the purpose rather than the activity—it is not a specific instruction. It can mean discontinuation, dose reduction or, potentially, dose increase if another medicine is up titrated to enable a second medicine to be stopped altogether. As such, it can be considered to be a generic term. However, there is growing literature on ‘deprescribing’ as an activity itself [33]. The medication review is a holistic process; the process of deprescribing includes reviewing all medications; identifying those that should be discontinued, substituted, or reduced; co-planning a deprescribing regimen with the patient; and regularly monitoring and supporting the patient [34].

The actions of stopping medication, changing dose, changing medication, formulation and monitoring are classed as clinical interventions. A clinical intervention is described as an activity performed to improve, maintain or assess the health of a person in a clinical situation. Altering the dose timing and changing from a branded to a generic version of a medication was defined as a technical intervention by some authors and as a clinical intervention by others. The classification disparity between authors of these two activities presents difficulties for researchers when comparing process evaluation studies. For example, Heselmans et al. [6] classified the medication timing change as a technical intervention [5], whilst the same activity was classified as a clinical intervention by Modig et al. [24]. Likewise, the medication formulation change is classified as a clinical intervention by Alldred et al. [29] and as a technical intervention by Chau et al. [14]. This inconsistency between authors in the use of overarching intervention categories indicates that there is a need to create a taxonomy to classify each of the conducted activities under the appropriate category which will enable meaningful comparison between process evaluations and help identify the mechanism of impact.

Various terms were used by different authors to indicate the same activity. For example, the terms ‘increase dose’ and ‘alter dose’ were used to describe one activity [29]. Another author applied the term ‘change dose’ to denote the dosage interval or frequency [18].

The act of stopping medication was described with different words in the included studies. The words ‘discontinue’, ‘cease’ and ‘stop’ suggest a permanent change whereas ‘suspend’ suggests a temporary change to medication.

The act of starting medication was described as ‘start’ and ‘restart’. It is crucial to separate the term ‘restart’ from ‘start’ because the former describes a medication deprescribing attempt that needs to be reversed.

The act of increasing the medication dose was described as ‘change dose’, ‘alter dose’, or ‘adjust dose’. Such terms do not specify whether the healthcare professional needs to increase the dose to improve medication effectiveness or decrease the medication dose to reduce side effects. Such terms give no indication of the precise instruction to be carried out or the reason for it. These vague descriptions and instructions give no indication as to the problem regarding the medication or the exact activity to be carried out by the practitioner. Despite some authors providing definitions for medication review terms used in their papers, these terms varied between authors. Such ambiguities would be eliminated by the development of a specific and inclusive consensus taxonomy of standardised medication review terms.

Conclusion

This systematic review identified gaps in previous literature regarding the terms used by healthcare professionals with respect to the medication review process. Most of the studies included had wide and varied term definitions which demonstrated that there is no consensus terminology used by all researchers. The findings from this review support the need to develop a consensus between all healthcare professionals and researchers to use unique and specific terminology that facilitates process evaluation comparison, consequently providing a better understanding of the mechanism of impact.

As an outcome of this systematic review, the next step should aim to develop a consensus among a group of experts in the field of medication review to explore their decisions and agree the specific terms through a convenient and transparent consensus methodology. This proposal will be implemented in a future study by the authors.