Volume 85, Issue 7 p. 1396-1406
SYSTEMATIC REVIEW AND META-ANALYSIS
Free Access

Availability and readability of patient education materials for deprescribing: An environmental scan

Michael Anthony Fajardo

Michael Anthony Fajardo

Sydney School of Public Health, ASK-GP Centre of Research Excellence, The University of Sydney, NSW, Australia

Search for more papers by this author
Kristie Rebecca Weir

Kristie Rebecca Weir

Sydney School of Public Health, ASK-GP Centre of Research Excellence, The University of Sydney, NSW, Australia

Wiser Health Care, The University of Sydney, Sydney, NSW, Australia

Sydney Health Literacy Lab, Sydney School of Public Health, The University of Sydney, NSW, Australia

Search for more papers by this author
Carissa Bonner

Carissa Bonner

Sydney School of Public Health, ASK-GP Centre of Research Excellence, The University of Sydney, NSW, Australia

Wiser Health Care, The University of Sydney, Sydney, NSW, Australia

Sydney Health Literacy Lab, Sydney School of Public Health, The University of Sydney, NSW, Australia

Search for more papers by this author
Danijela Gnjidic

Danijela Gnjidic

School of Pharmacy and Charles Perkins Centre, The University of Sydney, Sydney, NSW, Australia

Search for more papers by this author
Jesse Jansen

Corresponding Author

Jesse Jansen

Sydney School of Public Health, ASK-GP Centre of Research Excellence, The University of Sydney, NSW, Australia

Wiser Health Care, The University of Sydney, Sydney, NSW, Australia

Sydney Health Literacy Lab, Sydney School of Public Health, The University of Sydney, NSW, Australia

Correspondence

Dr Jesse Jansen, The University of Sydney, Sydney School of Public Health, ASK-GP Centre of Research Excellence, Edward Ford Building (A27) Room 126A, NSW 2006, Australia.

Email: [email protected]

Search for more papers by this author
First published: 08 March 2019
Citations: 45
The authors confirm that the PI for this paper is Michael A. Fajardo.

Abstract

Aims

To identify and evaluate content and readability of freely available online deprescribing patient education materials (PEMs).

Methods

Systematic review of PEMs using MEDLINE, Embase, CINAHL, PsycINFO and The Cochrane Library of Systematic Reviews from inception to 25 September 2017 to identify PEMs. Additionally, deprescribing researchers and health professionals were surveyed to identify additional materials. Known repositories of materials were searched followed by a systematic Google search (22–28 January 2018). Materials were evaluated using an approach informed by the Patient Education Material Assessment Tool and the International Patient Decision Aids Standards Inventory. Readability of text-based materials was assessed using the US-based Gunning–Fog Index and Flesch–Kincaid Grade level.

Results

Forty-eight PEMs were identified. PEMs addressing deprescribing of medications for symptom control (81%) were most common. Preventative medications were rarely addressed and material (39%) focused on older people. Only 37% of PEMs provided information about both potential benefits (e.g. reducing risk of side effects) and harms (e.g. withdrawal symptoms, increased risk of disease) of deprescribing, while 40% focussed on benefits only. Readability indices indicated an average minimum reading level of Grade 12. Option Grids and Decision Aids (mean reading level below Grade 10) were most suitable for people with average literacy levels.

Conclusions

Over 1/3 of deprescribing PEMs present potential benefits and harms of deprescribing indicating most of the freely available materials are not balanced. Most PEMs are pitched above average reading levels making them inaccessible for low health literacy populations.

What is already known about this subject

  • Systematic reviews show that reducing or stopping specific medicines (deprescribing) may decrease adverse events and improve quality of life
  • Patient involvement in deprescribing process is critical and patient education materials are 1 mechanism to encourage and facilitate involvement
  • It is unknown whether patients and clinicians have access to suitable patient education materials for deprescribing

What this study adds

  • We identified 48 online patient education materials for deprescribing available to patients, ranging widely in aim and focus
  • The education materials had above average reading levels and were therefore not suitable for lower health literacy patients
  • Few materials presented balanced information about potential benefits and harms of deprescribing and thus were considered not helpful for informed choice

1 INTRODUCTION

Deprescribing is the process of reducing or discontinuing medication (whether prescribed or over the counter) to improve the outcomes of patients for whom the actual or potential harms outweigh the benefits with respect to the patients values, preferences, life expectancy, care goals and level of functioning.1 Older adults and individuals with multiple comorbidities are key populations where deprescribing is important, as polypharmacy and treatment burden are common in these groups.2, 3 Deprescribing might be appropriate when medication: (i) benefits are unclear or symptoms have resolved; (ii) is unlikely to benefit patient over their remaining lifespan; or (iii) results in unacceptable treatment burden.1 Evidence from nonrandomised studies has suggested that deprescribing can reduce medication burden, adverse events and improve the appropriate use of medicines.4, 5

Deprescribing should be considered as part of high-quality prescribing, and decisions to reduce or stop medications require the same principles of patient involvement and informed choice as decisions to start medications.1 At the same time, stopping is much more difficult than starting. Qualitative studies suggest that the concept of stopping medication is often counterintuitive and people may value their medicines despite evidence that they might not be of benefit and even harmful.6-8 Identified patient barriers to deprescribing include that perception medicines are highly important, focussing on benefits over harms of medications and fear around the uncertainty of not knowing what will happen when medicines are stopped.9 Educating patients and engaging them in the deprescribing process is therefore critical.1 Patient education materials positively influence knowledge and satisfaction10 and can reinforce verbal communication between health care providers and patients.11, 12 An educational booklet based on social constructivist learning and self-efficacy theory13 has been demonstrated to empower patients to engage in shared deprescribing decisions.14 When patients are engaged and better informed they tend to choose more conservative options (e.g. fewer medicines), further facilitating deprescribing.15 Educating patients on the steps involved and how to identify drug-induced harm and adverse effects that may occur when stopping or reducing medications is key to successful and safe deprescribing.1

Educational materials could support patients' understanding of and involvement in the deprescribing process in different ways. Patient decision aids are seen as the gold standard for increasing patient understanding and involvement in decisions options, as there is a strong research base supporting their use and internationally recognised standards (the International Patient Decision Aid Standards) to guide development and evaluation of quality.15-19 Other commonly used patient education methods include fact sheets and question prompt lists (defined as structured list of questions for the patient, or their caregiver to ask their doctor or another health professional).20

Appropriate reading levels are 1 factor that can impact the effectiveness of patient education materials.11 We know that health literacy plays an important role in patient understanding and health outcomes. The majority of people in the general population (e.g. Australia: 59%21; Europe: 47%22; Canada: 60%23) have inadequate health literacy to access, comprehend and act on reliable health information and the proportions are even higher in older people.24 This is associated with poor self-management, less access to the health system, increased chronic disease, reduced adherence to medication and increased medication related harm.25 It is therefore important to specifically address the needs of patients with low health literacy.26, 27 Previous work in CVD and chronic kidney disease concluded that patient education materials are generally too difficult to read for patients with average literacy levels.28, 29 This suggests that many patients lack the skills required to read, understand, and engage with educational materials.30 Readability however is a limited measure to assess whether a material can be understood as they tend to focus only on word count, sentence length and syllables.31 Explicit consideration of how a material can be understood and how information can be turned into action are other factors to consider.32 Presentation of content is another aspect, for example, the gold standard for presenting numerical information is to use visual aids such as an Icon Array (e.g. 2% = 2 coloured dots out of 100 black dots).33 The Patient Education Materials Evaluation (PEMAT) tool was 1 attempt to address the limitations of readability measures.32 Patient-centred, balanced education materials at an appropriate level for the reader should facilitate optimal understanding of the deprescribing process, support shared decision making, and ultimately improve health outcomes.

At present, peer-reviewed patient education materials about deprescribing are limited but show potential. A scoping review on different types of deprescribing interventions for benzodiazepines and Z drugs identified pharmacological approaches as the most prominent with very few educational interventions suggesting they are not routinely used in deprescribing trials.34 A clinical trial of an educational intervention for patients that was shown to effectively elicit a shared decision making process for deprescribing medication14 has also demonstrated in a larger trial that a pharmacist-led educational intervention increased discontinuation of prescription for inappropriate medications after 6 months.35 Knowing which patient materials are effective is important but knowing what is available to patients is equally important, as these materials may be used in practice regardless of underlying evidence.

We aimed to identify and evaluate the content presentation and readability of online patient education deprescribing materials for people aged 18 years or older. Specifically, the review aimed to summarise available deprescribing education materials, information communicated about the deprescribing process and the readability of materials.

2 METHODS

We adapted the methodology of this environmental scan from previous research in the patient communication field.36-39 This methodology starts with a traditional systematic review and is complemented with systematic Internet searches and questionnaires to identify materials that would not appear in a traditional systematic review.

2.1 Patient education material inclusion and exclusion criteria

Materials were included if they were: (i) focused on providing information to patients about deprescribing 1 or more currently prescribed medications/supporting informed choice for those aged 18 years or older; (ii) written in English; and (iii) able to be printed or read online (e.g. multimedia, interactive websites). Where there were multiple mediums for the same material (e.g. PDF version and online interactive version), each medium was counted uniquely only if the content differed. Materials were excluded if they were: (i) focused on a medical device rather than prescribed medication; (ii) only provide an online forum or information on what to do (e.g. how-to instructions) rather than information to support a decision; (iii) not solely focused on deprescribing, (e.g. an advance care planning information booklet which might deprescribing in a nonsubstantive manner); (iv) not freely available to the public; or (v) aimed at health professionals/clinicians or parents, family or carers.

2.2 Patient education materials identification

Four main sources for eligible patient education materials for deprescribing were sequentially searched: known patient resource repositories, peer-reviewed literature, key informants, and an Internet search (Google Australia). The results from these sources were then collated with duplicates removed and reassessed for final inclusion. The following section describes the approach to searching each source.

2.2.1 Known repository search

During September 2017, 2 independent reviewers (M.F. and K.W.) searched known repositories of patient resources identified in previous research (see Table 1).28 Keywords used included: stop, cessation, coming off and deprescribe/deprescribing.

Table 1. List of known online repositories for patient education materials
Organisation Website
Primary Health Tasmania https://www.primaryhealthtas.com.au/resources/deprescribing
Deprescribing Network http://deprescribing.org
Canadian Deprescribing Networka https://www.deprescribingnetwork.ca/
National Prescribing Network https://www.nps.org.au/
The Decision Aid Library Inventory (DALI) – Ottawa Research Institute https://decisionaid.ohri.ca/AZlist.html
Option grids http://optiongrid.org/
Agency for Healthcare Research and Quality http://www.effectivehealthcare.ahrq.gov/tools-and-resources/patient-decision-aids/
NHS (accessible) http://sdm.rightcare.nhs.uk/shared-decision-making-sheets/
NICE Decision Aids https://www.nice.org.uk/about/what-we-do/our-programmes/nice-guidance/nice-guidelines/shared-decision-making
Mayo Clinic decision aids http://www.mayoclinic.org/
MAGIC SHARE-IT Public Guidelines/Decision Aids https://www.magicapp.org/app#/guidelines
Decision Boxes at Laval University http://www.decisionbox.ulaval.ca/
Annalisa Decision Aids at Sydney University http://healthedecisions.org.au/team/
CeMPED Decision Aids at Sydney University http://www.psych.usyd.edu.au/cemped/com_decision_aids.shtml
Health fact boxes at the Harding Centre for Risk Literacy https://www.harding-center.mpg.de/en/health-information/fact-boxes
Cochrane DAs for Muskuloskeletal group http://musculoskeletal.cochrane.org/decision-aids
Patient DA site (mostly NHS, OG, M) http://patient.info/decision-aids
NHS (restricted access) http://sdm.rightcare.nhs.uk/pda/
Annalisa DAs at Norway (restricted access) https://mybetterdecisions.org/
  • a This website is also linked from the Deprescribing Network website.

2.2.2 Peer-reviewed literature

MEDLINE, Embase, CINAHL, PsycINFO and the Cochrane Library of Systematic Reviews were searched. The search adapted previous strategies related to deprescribing, patient education materials and medication,15, 40, 41 with support and advice from an academic librarian (see Supporting information Appendix S1). Two authors (M.F. and K.W.) independently screened all titles and abstracts. Inclusion criteria for the systematic review were: (i) systematic review; (ii) adults currently on a medication; (iii) about deprescribing; (iv) educational intervention with patient education materials (e.g. decision aid, pamphlet, audiovisual material), and exclusion criterion was: not aimed at health professionals or parents. The same 2 authors independently screened full text articles for inclusion or exclusion where discrepancies were resolved by discussion. Consultation with a third author (J.J.) was sought if a consensus was not reached. All potentially-relevant papers excluded from the review at this stage are listed as excluded studies (see Supporting information Appendix S2 Figure A). The studies these systematic reviews included were searched for patient education materials (see Supporting information Appendix S2 Figure B). Authors were contacted if a paper did not provide access to the patient education material.

2.2.3 Key Informant input on deprescribing materials and search terms

Organisations and professionals in the fields of deprescribing health communication/decision making such as the Australian and Canadian Deprescribing Networks and the International Shared Decision Making Network were contacted to obtain patient education materials for deprescribing (see Supporting information Appendix S3 for the full list). An online survey was adapted from previous research,36 which incorporated a passive snowball approach (participants were asked to forward the survey link to others) to maximise reach (see Supporting information Appendix S3 for email correspondence and Qualtrics survey). The survey also asked participants what search terms they would use if they were to conduct a Google search for deprescribing patient education materials. The survey was sent in November 2017 and closed in January 2018 and was approved by the University of Sydney Human Research Ethics Committee (ref: 2017/806).

2.2.4 Google search

The results of the Key Informant survey, peer-reviewed literature search and known repositories informed the search terms for a Google search to identify other patient education materials for deprescribing that may not have been captured in the previous steps. Two independent reviewers (M.F. and K.W.) conducted the search on 22 and 30 January 2018 respectively. Search terms were divided into 2 themes with 5 terms each: Deprescribing (Coming Off Medication, Deprescribing, Discontinuing Medication, Stopping Medication, Tapering) and Patient education materials (Decision Aid, Decision Support, Fact Sheet, Patient Information, Question Prompt). One term from each theme were combined for a Google search resulting in 25 searches. As per previous research, URLs for the first 50 search results (not including advertisements) were exported using the SEOQuake plugin for Mozilla Firefox as searches after the first 50 were unlikely to meet critiera.28 The web browser cache was cleared before each search to minimise Google Search optimisation. This resulted in 2 independent pools of 1250 URLs where agreement per search strategy ranged from 22 to 50% (see Supporting information Appendix S2 Table A). The 2 pools were combined, and duplicate URLs were removed for the 2 reviewers (M.F. and K.W.) to independently assess webpages to determine whether inclusion criteria were met. Disagreement between authors was resolved by discussion and a third author (J.J.) was consulted where necessary. Reasons for exclusion are provided in Supporting information Appendix S3 (Figure C).

Results from each of the 4 sources were combined and then duplicates were removed. This list was reassessed by 2 independent reviewers (M.F. and K.W.) against inclusion/exclusion criteria before finalising the list, with reasons for exclusion provided in Figure 1.

Details are in the caption following the image
Study diagram for final search results

2.3 Patient education deprescribing materials assessment

Two reviewers (M.F. and K.W.) used a predefined data extraction form (see Supporting information Appendix S4) to rate included materials. Due to the diversity in patient education materials, this predefined data extraction form was adapted from the International Patient Decision Aid Standards Inventory16 and the Patient Education Materials Assessment Tool.32 Basic descriptive information was collected on material type, medication and health issue. Material types included:
  • Fact sheet and brochure/booklet/handbook: materials that presented information about a given topic; fact sheets are shorter in length (usually 1 page) compared to brochures/booklets/handbooks
  • Question prompters: provided specific questions to ask health professionals
  • Blogs: written by credible source(s) expressing personal/professional experiences and opinions
  • Decision Aids or Option Grids: explicitly identify a health care decision and present positive and negative features of each option available. Note: at the time of this research, Option Grids were freely available, but this is no longer the case.
  • Self-directed knowledge tests: provided information and a self-test quiz
  • Patient focused component of clinical algorithms: included a patient focussed risk index and deprescribing planner tool
  • Video/documentary
Health issues were classified using the chapter headings of the International Classification of Primary Care, second edition, which classifies patient data and clinical activity in the domains of Family/General Practice and Primary Care.42 The extraction form also included whether the material provided information on the benefits (e.g. reduced chance of side-effects) and/or harms (e.g. withdrawal symptoms, disease relapse) of deprescribing. Reviewers used a random sample of nine materials to calibrate the form, discuss discrepancies and finalise the form. Initial interrater agreement for the dichotomous evaluation items was 85.0% (Cohen's κ = .34). To remedy disagreement, discrepancies were reconciled via discussion between the independent reviewers to be used as the final results. Materials with at least 100 words were assessed in terms of their readability using the Flesch–Kincaid Grade Level Score and the Gunning Fog Index.43, 44 Readability indices, while limited, were used as a pragmatic tool to obtain a very rough indication about appropriateness for low health literacy population which can be compared across different types of materials. The Gunning Fog Index uses average words per sentence and complex words to calculate the final score whereas the Flesch–Kincaid Grade Level score uses average words per sentence and syllables to calculate the final score. Text analysed for readability scores had bullet points removed where bullet point text that listed single words were put into a single sentence with each word was separated by a comma. Bullet point text that did not list single words had a period entered at the end of the bullet point. Drug names were included. Titles of the text, hyperlinks and reference list text were excluded. Headings and subheadings were appended with a period if appropriate punctuation was absent (e.g. a question or exclamation mark). The final score for both readability measures can be interpreted as the minimum years of US formal education needed to read the material. Based on Morony and colleagues,29 average readers were classified as reading at an US 8th Grade Level or higher, and low health literacy readers were classified as reading at least at an US 5th Grade level.

2.4 Statistical analysis

Descriptive analysis of data were performed using Microsoft Excel and SPSS Version 24. Comparisons between dichotomous categorical variables were assessed using χ2 tests.

3 RESULTS

3.1 Final search results

Forty-eight materials were identified in this study (see Figure 1, or Supporting information Appendix S5 for the list of URLs). Known repository searches identified 20 potential patient education materials. The systematic review of systematic reviews identified 10 studies mentioning an education material. None of the included primary studies within these systematic reviews provided (links to) this education material. Hyperlinks to 3 of the 10 materials mentioned were provided by authors. The key informant survey (25 responses) identified a further 4 materials and the National Prescribing Network (Australia) provided data from their consumer scan which provided an extra eight materials. The peer review process also identified 1 additional material. Thirty-two percent of key informant responses identified www.deprescribing.org as a key resource. This website is maintained by research teams at Bruyère Research Institute (Ottawa) and Université de Montréal. The Google Search identified a total of 84 patient education materials with 33 materials already identified by the previous methods, thus providing an additional 51 materials. The final list was rechecked for final eligibility with reasons for exclusion listed in Figure 1.

3.2 Descriptive data

The characteristics of the included materials is summarised in Table 2 and Supporting information Appendix S5 provides the list of materials and URLs. Seventeen (35%) of the materials identified had a geriatric focus. One video was a 75-minute documentary about deprescribing that was not included in the following descriptive statistics. Two types of materials were identified, focusing on: (i) deprescribing specific medication classes; and (ii) medication reviews (i.e. deprescribing in the context of polypharmacy). All materials focusing on deprescribing specific medication classes in older people focused on medication that has been identified as potentially inappropriate medication in older adults according to Beers' criteria.45 The most common terms used for deprescribing were stopping (26%), reducing (19%), coming off (7%) and tapering (7%). The term deprescribing was used in 4% of materials. The most common rationales presented for deprescribing were long-term use of medication can cause side effects (50%) and the medication may no longer be needed (26%). Fifty-three percent of materials referred to peer-reviewed evidence. Thirty-eight percent provided information about what might happen if you do nothing (i.e. continue with medication). All materials suggested talking to a GP/Health Professional about deprescribing, other provided options include: lifestyle changes as alternative to medication (e.g. exercise, stop smoking; 11%) and tapering as an alternative to stopping (10%). Forty percent of materials provided information about both the potential benefits and potential harms of deprescribing, 36% provided only benefits, 13% provided only harms and 11% provided neither (see Figure 2). Twenty-five percent of materials highlighted the role of patient preferences, goals and/or priorities in the deprescribing process. Materials that addressed patient preferences presented both benefits and harms of deprescribing more often (75%) compared to those that did not address patient preferences (29%; χ2 = 7.998, P = .005).

Table 2. Characteristics of the included materials (n = 46)
n (%)
Material Type
Fact Sheet 21 (43)
Handbook/Booklet/Brochure 8 (17)
Self-Directed Knowledge Test 5 (10)
Blog 4 (8)
Decision Aid/Option Grida 4 (8)
Patient-focussed component of clinical algorithms 2 (4)
Question Prompter 2 (4)
Video/Documentary 2 (4)
Developer
Non-for-profit organisation 34 (71)
Individual 12 (25)
Government 2 (4)
Year of publication
2010 1 (2)
2011 1 (2)
2012 2 (4)
2013 2 (4)
2014 5 (10)
2015 2 (4)
2016 10 (21)
2017 4 (8)
2018 1 (2)
Not stated 20 (42)
Medicationb
Psychotropic drugsc 17 (35)
Deprescribing in context polypharmacy 12 (25)
Benzodiazepine receptor agonists (includes Z-Drugs and sleeping tablets) 7 (15)
Proton pump inhibitors 6 (13)
Opioids 4 (8)
Antihistamines 1 (2)
Sulfonylurea 1 (2)
Rationale for deprescribingd
Long term use can cause side effects 33 (70)
Medication is no longer needed 16 (34)
Reduce medication burden 6 (13)
Potential harms outweigh benefits 4 (8)
Maintain or improve quality of life 2 (4)
Not addressed by material 2 (4)
Health issuee
Psychological (P) 23 (49)
General (A) 13 (28)
Digestive (D) 6 (13)
Neurological (N) 4 (8)
Respiratory (R) 1 (2)
Metabolic, endocrine, nutrition (T) 1 (2)
  • a Option Grids were initially free materials, although they no longer are at point of publication.
  • b No materials identified were preventive medication.
  • c These include antidepressants, antipsychotics and mood stabilisers.
  • d Percentages will not add to 100 as 1 material may describe multiple aims of deprescribing.
  • e Parenthesised letters indicate International Classification of Primary Care, second edition Classification coding.
Details are in the caption following the image
Mention of potential benefits (e.g. reducing risk of side effects), potential harms (e.g. withdrawal symptoms, increased risk of disease), and both potential benefits and harms of deprescribing comparing: (A) Materials focusing on deprescribing of specific medication (all medication combined) vs 1 or more medications in the context of polypharmacy and medication reviews; (B) separation of materials for medication classes

For all text-based materials with at least 100 words of text (i.e. excluding videos and Medstopper) the average Gunning–Fog index was 11.9 (standard deviation = 1.97, range: 7.83–15.50) and the average Flesch–Kincaid Grade level was 10.0 (standard deviation = 1.78, range: 6.65–13.34). Supporting information Appendix S5 contains the readability scores for all materials and a graphical representation (Figure A; adapted from Morony and colleagues, 2015) of the different material types and their respective Gunning–Fog and Flesch–Kincaid Grade level scores.

4 DISCUSSION

4.1 Principal findings

This systematic review identified 48 patient education materials to support deprescribing 1 or more medications. Of importance, only 4 Decision Aids/Options Grids, which are reputable types of materials that increase patient knowledge of a healthcare decision,15 were identified. This is further supported by the finding that only 37% of identified materials present balanced content on potential benefits (e.g. improved well-being) and harms (e.g. withdrawal effects, return of symptoms or disease) of deprescribing. This implies that most patient education materials are not balanced in their presentation of deprescribing which may introduce bias and is not conducive to informed consent.6 The rationale for deprescribing differed across materials, in line with the different ways in which medication can be inappropriate,46 including materials focussing on deprescribing medications with high risk of harms or medications that are no longer needed or deprescribing to reduce medication burden and improve quality of life. This heterogeneity makes it difficult to compare the different materials. On one hand, one could argue that all materials should contain balanced information about potential benefits and harms of deprescribing in order to support an informed choice6 and to educate patients about what side effects to report on.1 On the other hand, a more directive approach with a strong focus on benefits relative to harms might be warranted in some cases when the evidence for deprescribing is strong (e.g. opioids). Identifying the key information to include and how to effectively present this information, particularly with ethical deliberation around the concept of balance of evidence and other aspects to improve understanding and safe and effective deprescribing, should be the focus of future research.

Materials for specific drug classes all focused on disease or symptom management, primarily deprescribing psychotropic drugs; this is not surprising and there is clear evidence of the potential harms of inappropriate use of these medicines and the potential benefits of deprescribing47 in particular in vulnerable groups such as the elderly. Importantly, medications for primary prevention did not feature in any of the materials. There is increasing attention for the potential harms of inappropriate use of preventive medication for example statins and antihypertensives in asymptomatic older frail people and the need for shared decision making about deprescribing so this is an area where more materials should be developed.48, 49

Readability indices suggest that all of the materials demand literacy levels above that of the average patient and are unsuitable for patients with lower health literacy.50 Since more than a third of the materials are aimed at older patients, this is of particular concern as both lower health literacy and cognitive decline are more common in this group.25 Of all the materials, Decision Aids and Option Grids were easiest to read although still pitched well above the average patient's reading level. Blogs are consistently in the higher ends of the readability indices suggesting that these materials may not be suitable for low health literacy patients. Fact sheets, handbooks/brochures/booklets are more dispersed across readability indices potentially highlighting the variability in the consideration applied to these materials in terms of their accessibility to low health literacy population. Compared to print chronic kidney disease materials and drug safety information for patients, both of which averaged a ninth-grade reading level,29, 51 presently identified deprescribing patient education materials were much more difficult to read with at least an average 10th-grade reading level. However, due to the limitation of readability indices which consider a limited set of parameters and are not always indicative of ease of comprehension,31 readability indices form only a fraction of the picture so should be interpreted cautiously, particularly given that different readability measures can provide very different results.52

Despite wide acknowledgement that incorporating patient preferences and values are essential to achieve successful deprescribing, they are rarely addressed in the materials we identified. This is concerning, as, in many cases, the need for deprescribing is largely determined by patient goals, complications may arise from withdrawal, and there is uncertainty around the benefits and harms of discontinuing medication in the long term.1, 6, 53 This means that whether or not deprescribing is the best option depends on each individual patient's context and preferences. This requires a patient-centred approach to the deprescribing process, and educational materials can play an important role in empowering patients to be involved and even drive deprescribing discussions.14 Value clarification methods,54, 55 questions within materials asking about what matters most to the patient, or tables of pros and cons of each option in terms of physical, social and/or psychological impact are some examples that may help address patient preferences and values in their healthcare decision. Wording of these questions should be considered to help reduce the cognitive impact on the patient, thus in turn improving comprehension.56 Of note, the materials that present both benefits and harms of engaging in the deprescribing process are more likely to be the ones that address patient preferences, values or goals in some manner.

4.2 Strengths and limitations

A strength of this study lies in the novelty of an environmental scan to allow us to describe the real-time landscape of patient education deprescribing materials as a traditional systematic review of the academic literature would have been an ineffective approach in this context.36-39 An environmental scan can be used to supplement systematic reviews that focus solely on peer-reviewed literature.34 The snowball approach improved chances that we reached as many important key informants in this space,57 but does not guarantee all important contacts were reached. Despite this reach, our study is limited in looking at English only materials, thus potentially excluding other language materials of relevance to international patient populations. Additionally, unlike a traditional peer-reviewed literature search, the results of a Google search are dynamic and thus unlikely to be replicable; therefore, the Google search results would only reflect the landscape for a short period. However, we have tried to mitigate this issue by including more stable sources: known repositories, systematic reviews in the academic literature and a Key Informant survey. Moreover, as the focus of this study was on online materials, materials that may be provided in the healthcare setting may not have been captured thus our findings may not be representative of what consumers actually use. A next step would be to explore consumer information seeking behaviours pertinent to the context of deprescribing. Another limitation of this study lies with our operationalisation of readability because text alone accounts for only a few of the elements that determine understandability. The selection of an online readability tool impacts the readability measures as a different online readability tool can produce different readability indices. A more thorough analysis of specific types of educational materials looking at other elements such intended purpose of the material, style, organisation, use of visual aids and actionability is needed to fully understand the effectiveness of these materials and address the limitations of readability indices, e.g. by using the Suitability Assessment of Materials50 and PEMAT32 and IPDAS criteria for decision aids. However, the aim of our study was to provide an initial overview of different types of available online patient education materials and due to the diversity of materials there was not 1 single validated measure to compare across this diversity. These limitations do not change our overall conclusion that much work remains to be done in developing appropriate materials to educate patients about the deprescribing process and empower them to be involved.

4.3 Practice implications

There is a pressing need to develop materials that provide balanced information about both benefits and harms of deprescribing, so that patients can make informed choices and know which warning signs and symptoms to look out for when medications are being discontinued. Materials also need to address the health literacy needs of the average patient by aiming at improving comprehension and understanding. Additional tailoring might be required to meet the needs of older patients and those with lower health literacy, research methodologies such as thinking aloud58 and qualitative interviews or focus groups may help assess consumer comprehension and identify how materials can be improved. Journal editors should encourage authors to make evaluated patient education materials available online so that patients, researchers and clinicians are able to freely access them or repost them in a central repository. Given that the most common (32%) Key Informant survey response was www.deprescribing.org, this website appears to be the most appropriate location to collate materials (see policy for including resources: https://deprescribing.org/wp-content/uploads/2018/09/Website-linking-and-resource-policy_August2018.pdf). This endeavour, however, comes with its own practical and financial challenges as well as how to determine the quality of a hosted material. Resources should also be listed on professional societies and national health institutes/organisations (see Table 3). The context in which the material is used should also be considered. For example materials intended to be used in the consultation room together with a health professional could embed additional elements to empower patient involvement such as teach-back prompts.59 Teach-back (also referred to as talk-back) is the process of asking the patient to explain to the health professional what they think something means in their own words, which can help clarify patient understanding.59

Table 3. Recommendations
How to comprehensively develop and disseminate patient education materials for deprescribing
• Refer to the IPDAS criteria for guidance on decision aid development including recommendations on balanced presentation of benefits and harms, risk communication, values elicitation and how to address lower health literacy: http://ipdas.ohri.ca
• Refer to the criteria for development of easy to understand materials for people with lower health literacy to ensure readability is pitched at an accessible level: https://www.cdc.gov/healthliteracy/pdf/Simply_Put.pdf
• Materials evaluated in research studies should be made easily accessible to other researchers at point of publication
• Disseminate materials through a central, international source: e.g. www.deprescribing.org (owned by the Bruyère research institute and University of Montreal) as well as professional societies and national health institutes/organisations (e.g. US National Institute of health, Australian National Health and medication research council)

5 CONCLUSION

Currently available patient education materials for deprescribing are generally not balanced in the information they provide in relation to evidence and guidance to support deprescribing in practice and may difficult to read for the average person let alone low health literacy populations. Future developers of deprescribing patient education materials need to address these issues to help patients make informed choices and safely discontinue their medication.

M.A.F. was supported by a NHMRC funded Centre for Research Excellence called Ask, Share Know: Rapid Evidence for General Practice Decisions. K.R.W. was supported by a Sydney Medical School ECR PhD scholarship. C.B. was supported by a National Heart Foundation of Australia (Vanguard Grant 101326) and Royal Australian College of General Practitioners and Therapeutic Guidelines (TGL/RACGP Research Grant TGL16b). D.G. was supported by the Australian National Health and Medical Research Council Dementia Leadership Fellowship (1136849). J.J. was supported by an Early Career fellowship from the National Health and Medical Research Council (NHMRC; No. 1037028).

ACKNOWLEDGEMENTS

We would like to acknowledge Ms Jayne O'Hare, Academic Liaison Librarian of the University of Sydney for her support in refining the strategy used for the peer-review literature search.

    COMPETING INTERESTS

    The authors have no competing interests to declare.

    CONTRIBUTORS

    M.A.F. contributed to the methodology, study selection, data extraction, patient education material evaluation, data analysis and drafting the manuscript. K.R.W. contributed to the study selection, data extraction, patient education material evaluation and revising the manuscript. C.B. contributed to the conceptualisation, methodology, data interpretation and revising the manuscript. D.G. contributed to the conceptualisation, methodology, data interpretation and revising the manuscript. J.J. contributed to the conceptualisation, methodology, data interpretation, drafting and revising the manuscript.