Abstract
To systematically summarize evidence from multiple systematic reviews (SRs) examining interventions addressing medication nonadherence and to discern differences in effectiveness by intervention, patient, and study characteristics.
MEDLINE, the Cochrane Database of Systematic Reviews, and the Database of Abstracts of Reviews of Effects were searched for papers published from January 2004 to February 2017. English-language SRs examining benefits of medication adherence interventions were eligible. Inclusion was limited to adult patients prescribed medication for 1 of the following disease conditions: diabetes and prediabetes, heart conditions, hypertension and prehypertension, stroke, and cognitive impairment. Non–disease-specific SRs that considered medication adherence interventions for older adults, adults with chronic illness, and adults with known medication adherence problems were also included. Two researchers independently screened titles, abstracts, and full-text articles. They then extracted key variables from eligible SRs, reconciling discrepancies via discussion. A MeaSurement Tool to Assess systematic Reviews (AMSTAR) was used to assess SRs; those with scores below 8 were excluded. Conclusions regarding intervention effectiveness were extracted. Grades of Recommendation, Assessment, Development and Evaluation (GRADE) methodology was applied to assess evidence quality.
Of 390 SRs, 25 met the inclusion criteria and assessed adherence as a primary outcome. Intervention types most consistently found to be effective were dose simplification, patient education, electronic reminders to patients, and reduced patient cost sharing or incentives. Of 50 conclusions drawn by the SRs, the underlying evidence was low or very low quality for 45 SRs.
Despite an abundance of primary studies and despite only examining high-quality SRs, the vast majority of primary studies supporting SR authors’ conclusions were of low or very low quality. Nonetheless, health system leaders seeking to improve medication adherence should prioritize interventions that have been studied and found to be effective at improving patient adherence, including dose simplification, education, reminders, and financial incentives.
This overview identified 25 high-quality systematic reviews (SRs) of interventions addressing medication adherence published since 2004.
Despite an abundance of primary studies and despite examining high-quality SRs, the vast majority of primary studies supporting SR authors’ conclusions were of low or very low quality according to the accepted methodology for evidence grading.
SR authors most consistently found the following intervention types to be effective: dose simplification, patient education, electronic reminders, and reduced patient costs or incentives.
Medication adherence, or the extent to which patients take medications as prescribed, is remarkably low both in the United States and worldwide. Nearly 1 in 3 new U.S. prescriptions are never filled. Furthermore, many patients who fill their prescriptions do not take them as prescribed or stop taking them altogether. Taken together, these deficiencies result in patients adhering to about 50% of prescribed medications.1,2
Poor medication adherence is known to substantially increase mortality, morbidity, and healthcare costs.3-5 However, reasons for nonadherence are varied and include cost to patients, lack of information regarding disease progression, fear of drug adverse effects, low health literacy, and lack of social support.6 Given the diverse etiologies of poor medication adherence, interventions to improve adherence are correspondingly complex, varied, and abundant. Developing and testing the impact of such interventions on both adherence rates and clinical outcomes have been of intense interest, as evidenced by hundreds of systematic reviews on interventions to improve medication adherence.
Elucidating the central findings of this wide literature can be unwieldy not only because of its volume but also because findings may differ based on study setting and populations, intervention characteristics, outcomes measured, analytic methods, sample sizes, and even differing interpretations. Systematic reviews (SRs) have gained acceptance as a robust methodology to efficiently distill and summarize prior findings. However, because SRs themselves may be subject to these same concerns, especially in areas with several SRs, the Cochrane Collaboration and other thought leaders have encouraged the use of systematic overviews of SRs.7-10 This approach allows researchers to capitalize both on the accepted methodology of systematically evaluating literature and a huge body of secondary literature. With hundreds of existing SRs on medication adherence already published, we applied this systematic overview methodology not only to summarize existing literature but also to synthesize findings and bring new insights from across a vast literature.
Methods
Reporting for this SR is based on the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and other available methodological guidance for overviews of SRs.7,10-12 The study protocol is posted on the International Prospective Register of Systematic Reviews (PROSPERO) registry (CRD42017053814).
Data sources and searches
To identify all relevant peer-reviewed SRs, researchers performed a search of MEDLINE,the Cochrane Database of Systematic Reviews, and Database of Abstracts of Reviews of Effects (DARE) in February 2017. Two trained researchers developed search terms related to medication adherence interventions. The MEDLINE search used the following search terms: (“review”[ti] OR “systematic review”[tiab] OR “health technology assessment”[tiab] OR “metaanalysis”[tiab] OR “meta-analysis”[tiab] OR “review”[Publication Type] medication[ti]) AND (adherence [ti] or compliance[ti] or persistence [ti]). The Cochrane Database of Systematic Reviews and DARE were searched using the following terms: medication AND adherence. The search was limited to English language publications from January 2004 to February 2017. Finally, SRs incidentally identified during the search process (e.g., references of included studies) were considered.
Selection of systematic reviews
Eligible studies were English language SRs (with or without meta-analyses) that examined the effect of interventions designed to improve patients’ adherence to medications, defined as strategies aimed at improving accordance with the prescribed interval and dose of a dosing regimen. Inclusion was limited to adult patients prescribed medication for 1 of the following disease conditions: diabetes and prediabetes, cardiac conditions, hypertension and prehypertension, stroke, and cognitive impairment. Researchers excluded SRs explicitly focusing only on disease states unlikely to contribute to hospitalization (e.g., osteoporosis) and on disease states that may have had several different factors contributing to nonadherence (e.g., psychiatric disease and acquired immunodeficiency syndrome). Also included were non–disease-specific SRs that considered medication adherence interventions for older adults, adults with chronic illness, and adults with known medication adherence problems. Those SRs that focused exclusively on interventions implemented in low- to middle-income countries were excluded due to differences in care practices.
Two investigators independently screened titles and abstracts per the prespecified inclusion and exclusion criteria. Next, full-text publications were retrieved and independently examined by 2 researchers to determine eligibility. Discrepancies at both the title and abstract and full-text screening levels were resolved by discussion.
Evaluation of quality
We assessed the methodological quality of each relevant SR using the validated A MeaSurement Tool to Assess systematic Reviews (AMSTAR) instrument.13 The tool contains 11 requisite items that are rated as present or absent, such that each SR may receive a score ranging from 0 to 11. Two reviewers independently applied the instrument, and discrepancies were reconciled through discussion. SRs with an AMSTAR score of 8 or greater, which is an accepted cutoff used in prior work,10 were included in data extraction and synthesis.
Data extraction
For each included study, 2 researchers independently extracted data related to the key characteristics of the SRs, including dates of literature search, number and design of included primary studies, intervention types, patient populations, setting, phases of adherence, medication adherence measures, presence of meta-analytic techniques, and reported pooled effect sizes. Additionally, information concerning the reviews’ major conclusions regarding intervention effectiveness was extracted. The extracted data were compared, and discrepancies were reconciled through discussion.
Quality of evidence
We documented the quality of evidence for each comparison within each SR by applying GRADE (Grades of Recommendation, Assessment, Development and Evaluation) methodology.14 Objective criteria were used to assign a level of evidence based on the following GRADE domains: study design, study quality, consistency, directness, and other modifying factors, including precision and strength of effect estimates. The researchers did not assess the quality of the individual studies within the SRs but reported the risk of bias of studies as documented in the SRs. If information regarding a domain was missing or unavailable, researchers did not up- or downgrade the level of evidence; as such, GRADE scores represent a maximal estimate of the potential evidence quality. One researcher assessed the GRADE level of evidence for each SR’s conclusions.
Synthesis
To begin, researchers examined each SR’s major conclusions regarding the effectiveness of intervention strategies and classified conclusions into 4 distinct categories, as follows: (1) a positive association between intervention strategy and medication adherence, (2) a negative association between intervention strategy and medication adherence, (3) a null association between intervention strategy and medication adherence, or (4) preclusion from drawing conclusions due to limited or low-quality studies. To compare findings between SRs, researchers aggregated conclusion classification data by intervention category, incorporating information on patient condition and GRADE level of evidence. Definitions of intervention categories may be found in the appendix. For major intervention categories in which different SRs yielded different conclusions, the overlap of included primary studies was compared using citation matrices. The study was funded by the American Society of Health-System Pharmacists (ASHP) Research and Education Foundation. ASHP did not participate in the literature search, assessment of eligibility, data extraction and analysis, or interpretation of results.
Results
Study selection
After removing 41 duplicate references, the search identified 390 records. Upon screening titles and abstracts, 95 citations were selected for full-text review. Seventy SRs met all eligibility criteria and were subject to methodological quality evaluation; of these, 25 articles received an AMSTAR score of 8 or higher and were included in the data synthesis.15-39Figure 1 reflects the PRISMA flow diagram, detailing the primary reasons for exclusion.
PRISMA flow diagram.
Study characteristics
The general characteristics of the included SRs, including year of publication, literature search dates, patient population, setting, intervention type, presence of meta-analytic techniques, number and design of included studies, and methods for measuring medication adherence, may be found in the online-only data in Supplement Table 1 All SRs were published between 2004 and 2016, with 20 (80%) SRs published since 2012.15-20, 22-24,26–30,32,35,37-39 Twenty-one (84%) SRs were published in peer-reviewed journals,16-20,22-24,26-33,35-38 whereas 4 (16%) SRs were published as Cochrane SRs.15,25,34,39 All 25 SRs performed systematic searches and assessments of the literature. Fourteen (56%) SRs also incorporated meta-analytic techniques.15-18,20,22,23,26,30,34-36,38 Sixteen (64%) SRs only included randomized controlled trials (RCTs),15,16,18,20,24,25,28-34,38,39 whereas 9 (36%) SRs included both RCTs and observational studies.17,19,22,23,26,27,35-37 The number of included studies in each SR ranged from 3 to 182, with a mean of 36 SRs (standard deviation, +38 SRs). The patient populations and settings to which authors restricted their SRs were similarly diverse. Patient types included non–disease-specific patients (60%), 16-19,27,29-34,36-39 patients with cardiovascular disease (8%),16,20 patients taking statins (4%),28 patients with hypertension or prehypertension (8%),25,35 patients with diabetes or prediabetes (4%),23 poststroke patients (4%),22 patients with hyperlipidemia (4%),15 cognitively impaired older adults (4%),24 and patients experiencing medication adherence problems (4%).26 The majority of SRs did not restrict included studies to any particular setting (72%); 16-20,22-24,26,28-30,32,34-36,39 however, 3 (12%) SRs limited studies to ambulatory settings,15,25,38 2 (8%) SRs limited studies to community settings,27,33 1 (4%) SR excluded hospital settings,31 and another (4%) SRs excluded institutional settings.37 Most SRs did not exclude primary studies based on phase of adherence; however, the included studies almost exclusively measured medication adherence during the implementation phase (from initiation to last dose).40
The types of interventions on which SRs focused were varied and included nonrestricted intervention types (48%),15,16,22,24-26,28,31,35,37-39 special medication packaging (16%),19,27,33,34 dose simplification (8%),17,20 cognitive behavior change techniques (4%),18 electronic reminders (8%),29,30 text messaging reminders (4%),16 monitoring and messaging interventions (4%),23 psychosocial or educational interventions (4%),36 and technology-focused interventions (4%).32 While many of the systematic reviews included interventions that were pharmacist led, none of the systematic reviews focused exclusively on this type of intervention. All SRs assessed medication adherence as a primary study outcome. Most of the SRs did not restrict the method of measuring medication adherence (96%) 7,10,13,15-20,22-37,39,41; however, 1 SR included only studies that measured adherence through electronically compiled drug dosing histories.38 Twenty-one (84%) SRs included at least 1 primary study that used electronic monitoring to measure adherence,7,10,13,15,16,18-20,22-32,35,36,38,39,41 21 (84%) SRs included at least 1 primary study that used pill count,7,10,13,15,16,18-20,22-26,28-36, 39,41 21 (84%) SRs included at least 1 primary study that used patient self-report,7,10,13,15,16,18,20,22,23,25-36,39,41 17 (68%) SRs included at least 1 primary study that used pharmacy refill data,7,10,13,15-18, 22-26,28,30,32,33,35,36,39,41 1 (4%) SR included at least 1 primary study that used clinical outcomes,22 and 2 (8%) SRs included at least 1 primary study that used drug blood levels.25,32
Major conclusions of SRs
Major study conclusions for each SR, including pooled effect estimates (if applicable) are presented in the online-only data in Supplementary Table 2. Figure 2 presents an aggregate of study conclusions by type of intervention strategy, patient condition, and GRADE level of evidence. A summary of pooled effect sizes and pooled absolute mean differences by intervention category may be found in Figure 3.
Conclusions of systematic reviews that examined effectiveness of intervention components on medication adherence.
Reported pooled effect sizes and reported pooled absolute mean differences in adherence rates, by intervention category
Medication adherence interventions for all types
Among 8 SRs assessing the overall impact of medication adherence interventions (i.e., did not address specific components or types of interventions), half concluded that interventions were effective; however, most of these SRs were based on low- or very low-quality evidence. The 4 SRs reporting positive effects focused on heart failure patients, hypertensive patients, patients with poor adherence, and patients of any disease type.16,26,35,38 The remaining 4 reviews that focused generally on medication adherence interventions either concluded that interventions did not improve adherence or that they were precluded from drawing conclusions due to limitations in study number or quality.22,24,31,39
Dose simplification
Dose simplification was the most common strategy assessed by SRs. Four of 5 SRs examining dose simplification reported that it improved medication adherence; however, most of these SRs based conclusions on low- or very low-quality evidence.17,20,25,28 These 4 SRs included patients taking medication for cardiovascular disease, hypertension, high cholesterol, and osteoporosis. The fifth SR was more general. It focused on patients of any disease type in the ambulatory setting and reported a null association between dose simplification and adherence.38
Patient education
Patient education was another commonly studied intervention component. Three out of 4 SRs found patient education interventions to be effective in improving adherence among patients taking statins, patients with chronic illness, and patients of any disease type.28,37,38 An SR focused exclusively on hypertensive patients reported patient education alone to be largely unsuccessful.25
Electronic reminders
Four SRs assessed electronic reminders,16,29,30,38 (e.g., mobile text messaging, pagers, and pill bottles with alarm features), 3 of which reported positive findings.16,29,30 These 3 focused on patients with chronic diseases. A fourth SR which included only RCTs where adherence was assessed by electronic medication event monitoring methods reported nonsignificant results in medication adherence; this conclusion was based on very low-quality evidence. Citation matrix analysis showed only a modest overlap of primary studies which was due to differences in inclusion criteria among SRs, such as methods used to measure medication adherence and intervention components. The citation matrix for SRs assessing electronic reminders may be found in online-only data in Supplementary Table 3.
Incentives or reduced out-of-pocket spending
Three SRs assessed interventions incorporating incentives or reduced out-of-pocket spending.28,37,38 Two of these SRs, both of which assessed policy interventions to reduce patient spending, reported positive effects on medication adherence among patients taking statins or medication for a chronic illness.28,37 The third SR, which included patients with any disease condition, reported that rewards were not successful strategies for improving adherence (pooled mean difference, 3.3%; p = 0.44).38
Multicomponent approaches
The use of multicomponent approaches was assessed in 3 SRs. The quality of evidence underlying all 3 SRs’ conclusions was very low, but the results were generally positive.25,28,35 Two SRs, 1 focused on hypertensive patients and the other on patients taking statins, employed meta-analytic techniques to affirm the positive effects of multicomponent interventions.28,35 The third SR using nonquantitative synthesis reported that interventions with more than 1 component appeared promising but that conclusions were limited by a lack of high-quality RCTs.25
Monitoring and feedback
Three SRs examined monitoring and feedback interventions. The evidence underlying all 3 SRs’ assessments was of very low quality, and the results were mixed.23,35,38 An SR focused on patients of any disease type found electronic monitoring and feedback of adherence information to be effective (pooled mean difference, 19.8%; 95% confidence interval [CI], 10.7-28.9%) but that electronic monitoring of disease condition showed no statistically significant improvements in adherence (pooled mean difference not reported, p = 0.22).38 An SR of hypertensive patients reported adherence feedback to be a promising strategy, but this did not reach significance in a moderator analysis.35 A third SR, which focused on patients with type 2 diabetes, reported statistically null findings.23
Habit analysis
Two SRs assessed the effectiveness of habit analysis.26,35 This strategy includes interventions linking medication administration to existing habitual behaviors or routine events. Both SRs, published by the same author, reported habit analysis to be a promising intervention; however, only 1 SR reported statistically significant results.26 In this SR, the authors focused on patients with medication adherence problems and reported significant differences in effect size when comparing interventions with and without habit analysis (0.574 versus 0.222, p = 0.007).26 The second SR, which made the same comparison among hypertensive patients, reported the use of habit analysis to be a promising strategy, although this difference did not achieve statistical significance (0.412 versus 0.290, p = 0.298).35
Special medication packaging
Five SRs examined special medication packaging. All of these SRs based their assessments on very low-quality evidence, and the results were mixed.19,27,33-35 Three SRs focused on patients with any disease condition; none of these SRs could draw conclusions due to limitations in study quality or number.27,33,34 A fourth SR reported the use of special medication packaging to be a promising strategy among hypertensive patients; however, this intervention component did not reach significance in a moderator analysis.35 A fifth SR published by the Cochrane Collaborative focused on patients of any disease condition and found that interventions using special medication packaging improved adherence when measured as a percentage of pills taken (pooled mean difference, 11%; 95% CI, 6-17%) but reported the difference in proportion of self-reported adherent patients between intervention and control groups to be nonsignificant (pooled odds ratio, 0.89; 95% CI, 0.56-1.40).34 To better understand the different conclusions of the SRs, researchers used a citation matrix to compare included primary studies between SRs. Limited overlap existed due to differences in inclusion criteria, such as intervention components, patient disease condition, setting, and primary study design. The special medication packaging citation matrix may be found in the online-only data supplement.
Reminders and prompting mechanisms
More broadly, interventions incorporating reminders and prompting mechanisms of any type were assessed by 3 SRs, and the results were mixed.23,26,28 Reminder types in this category were not limited to electronic reminders. Other SRs included studies evaluating phone calls, postal mailings, and in-person follow-up. One SR restricted analysis to patients with medication adherence problems and determined that interventions incorporating prompts had significantly larger effect sizes than did those interventions lacking prompts (0.50 versus 0.23, p = 0.03).26 The other 2 SRs, which focused on patients taking statins and those with high cholesterol, reported pooled effect measures that did not reach statistical significance.23,28
Discussion
In summary, we identified 25 high-quality SRs published between 2004 and 2016 assessing interventions to improve medication adherence. Both this large number of high-quality SRs and a mean of 36 primary studies per SR reflect substantial interest in medication adherence. Furthermore, the publication of 20 (80%) of the high-quality SRs eligible for overview inclusion after 2012 is likely attributable to increasing interest, although increasing compliance with AMSTAR guidelines could be another potential explanatory factor.
In reviewing the types of interventions studied, the intervention with the most consistently positive findings was dose simplification. Four of 5 SRs concluded that it improved medication adherence, including 1 SR that based its conclusion on moderate-quality evidence. The next most promising intervention type was patient education, which 3 of 4 SRs found to be effective. Similarly, 3 of 4 SRs on electronic reminders found positive results, including 1 SR which based its conclusion on moderate-quality evidence. Finally, the category of financial incentives and reduced out-of-pocket spending was reported to be effective in 2 of 3 SRs, including 1 SR that based its assessments on moderate-quality evidence. Although there were other interventions that multiple SRs concluded were effective, these intervention types were found to be effective less commonly and consistently, and these conclusions were only supported by low- or very low-quality evidence.
We are unaware of any similar overview of SRs with which to compare these results. However, perhaps the most apt comparison is the most recent Cochrane SR on this subject.39 Nieuwlaat et al, whose Cochrane review was included in this overview, focused on 17 RCTs that all measured both adherence and downstream clinical outcomes and that had a low risk of bias. They found that despite the use of intensive, multicomponent interventions, only 5 (29%) of these RCTs demonstrated improvements in both adherence and clinical outcomes. The authors thus concluded that “current methods of improving medication adherence for chronic health problems are mostly complex and not very effective.”
In contrast to the findings of the Cochrane SR, multiple other systematic reviewers consistently made broader and more positive conclusions, at least with respect to certain intervention types. As such, the Cochrane review findings differ somewhat from our own. Some of this difference is attributable to decision points in SRs where arbitrary decisions become necessary. In this case, Cochrane reviewers made the reasonable decision only to include primary studies that measured a clinical outcome, whether or not the study was powered on this endpoint. Therefore, the decision of primary study authors regarding whether to include a secondary clinical endpoint without any power analysis could determine whether that study was included in the Cochrane review. We do not dispute this choice, but rather, we note that other decisions would also have been reasonable. For example, we would not disagree with other reviewers who did not require measurement of a clinical outcome or who only accepted studies powered on a clinical outcome.
By comparing the different decisions and the resultant different conclusions of systematic reviews initially focused on the same research questions, our overview highlighted the importance of the many arbitrary but legitimate decisions made in high-quality systematic reviews. In some cases, such choices (e.g., which patient populations, medical conditions, or settings to include) substantially affect a review’s final results and conclusions. Indeed, our citation matrices demonstrate that even reviews examining the same intervention types frequently consider different studies, which is usually driven by different study goals and thus by different study inclusion criteria. Finally, in the cases where multiple SRs converge on the same conclusions, despite their use of independent pathways to arrive at these conclusions, the systematic overview methodology reveals stronger evidence that these findings are based on a true effect. As such, we believe the strongest new evidence generated by our overview is that certain intervention types were consistently found to be effective, no matter the legitimate but arbitrary decisions made by the authors of high-quality SRs and no matter the primary studies included.
Because some of our included SRs used more than only the highest-quality RCTs to draw conclusions, we would advise that our findings should be matched to situations where decisions must be made using the best available evidence. Indeed, in navigating real-world complexities, providers and organizational leaders must often make decisions with limited choices and insufficient evidence. Based on our findings, we would advise providers and organizational leaders requiring decisions on these topics to prioritize interventions using dose simplification, patient education, electronic reminders, and multicomponent approaches. Similarly, we would encourage investigators to focus future research on these most promising areas, prioritizing rigorous methodology.
One strength of secondary assessments of adherence literature that applied to both this study and its underlying SRs is that the outcome of adherence can be compared relatively easily across most study types. Indeed, each of the 25 included SRs used adherence as the primary outcome. Furthermore, several SRs used Cohen’s d or other measures of effect size to rank the relative effect of different interventions across primary studies. Even though this may be better than trying to measure disease burden or progression in areas that lack a standardized instrument, there are still several ways of measuring medication adherence, as follows: pill count, self-report, pharmacy refill data, electronic monitoring, drug blood levels, and observation by healthcare professionals. We appreciate that different measurement methods can affect results, and this is 1a limitation of our study. Again, however, the overview approach also allowed for a comparison of results across SRs that addressed adherence outcomes differently.
Other limitations include the depth and quality of existing evidence, which are central limitations of all reviews of existing literature. Because of the hundreds of existing SRs on medication adherence interventions, the issue of the depth of existing evidence was addressed as well as could be expected for any domain. To address SR quality concerns, we focused on high-quality SRs. However, 1 central finding of our overview concerned the lack of high-quality primary studies, which was best exemplified by the most stringent SR finding only 17 high-quality RCTs assessing both medication adherence and disease outcomes. Nonetheless, our overview approach also included SRs with more liberal inclusion criteria, from which SR authors drew stronger conclusions. Indeed, synthesizing the findings of these different perspectives is one of the strengths of a systematic overview. Finally, we recognize that overview methodology necessarily misses some of the nuances appreciable in SRs and primary research and that overviews exacerbate existing time lags between the publication of primary studies and new conclusions.
Conclusion
We found 25 recent, high-quality SRs of interventions addressing medication adherence. Although the vast majority of the primary studies supporting SR authors’ conclusions were of low or very low quality, SRs most consistently found dose simplification, patient education, electronic reminders, and reduced patient cost sharing or incentives to be effective.
Disclosures
This research was supported by the American Society of Health-System Pharmacists (ASHP) Research and Education Foundation and the National Institute on Aging of the National Institutes of Health under awards K23AG049181 and R01AG058911. The authors have declared no potential conflicts of interest.
References
Appendix—Definitions of medication adherence intervention categories
I. Patient Education (General)a
Strategies focusing on in-person patient education delivered either individually or as a part of a group of community; may incorporate verbal edification and print or audio-visual materials.
II. Special Medication Packaging
Strategies that focus on altering the physical medication packaging, such as blister packs, pill boxes, or single-use containers.34
III. Dose Simplification
Strategies that improve the ease of use of dosing regimens, such as reducing the frequency of a dosing regimen (e.g., twice a day to once a day) or reducing the number of pills taken by combining medications into 1 pill (“polypill”).17,28
IV. Monitoring and Feedbacka
Interventions incorporating a summary of medication adherence or clinical indicators measured for a specific period and reported to either a patient, clinician, or institution.
V. Technology Mediated
Interventions incorporating (1) information and communication technology (e.g., computers, telephones, videos, cell phones, pagers, emails, short message services [SMSs], the and Internet), or (2) electronic medical devices (e.g., electronic drug monitors, pillboxes with an alarm system, home blood pressure monitors, or telehealth devices).32
VI. Text Messaging
Intervention using mobile telephone text messaging to promote medication adherence.16
VII. Case Managementa
Interventions coordinating the assessment, treatment, and referrals by a person or multidisciplinary team in collaboration with, or supplementary to, the primary care provider.
VIII. Electronic Reminders to Patients
Interventions designed to provide technical devices to remind the patients when it is time to take their medications (e.g., mobile phone text messages, pager, and electronic monitor with an alarm/beeper).38
IX. Patient Reminders/Prompting Mechanisms (Any Type)a
Efforts directed by a healthcare entity or health professional toward patients encouraging them to adhere to a prescribed medication regimen. Types of reminders in this category include not only electronic reminders but also phone calls, postal mailings, and in-person follow-up.
X. Patient Incentives/Reduced Out-of-Pocket Spendinga
Interventions aimed at reducing patients’ out-of-pocket expenses for prescription medications through either provision of positive or negative financial incentives, reduced medication copayments, or improved prescription drug coverage.37
XI. Self-Management
Strategies including the distribution of materials or access to a resource that enhances a patient’s ability to manage their own condition, the communication of useful clinical data to the patient, or follow-up phone calls from the provider to the patient, with recommended adjustments to care.
XII. Habit Analysis
Interventions examining a patient’s daily routines and that link medication administration to existing habitual behaviors or routine events.26,35
XIII. Motivational Strategies
Interventions aimed at increasing a patient’s desire or willingness to adhere to a prescribed medication regimen. Examples are broad and may include financial incentives, reminder aids, including diaries, and follow-up appointments.25
XIV. Multicomponent Approaches
Strategies encompassing 1 or more interventions in combination.25
XV. Psychosocial Interventions
Interventions focusing on patients’ feelings and emotions or social relationships and social support. Examples include family counseling, group meetings with peers, and stress management techniques.38
XVI. Cognitive Behavior Change Techniques
Any intervention using some form of psychological technique to change a patient’s adherence behavior, as well as their thoughts, feelings, confidence, or motivation towards adherence.18
XVII. Intensified Patient Care
Patient-focused interventions incorporating an amplified level of care typically beyond what a single clinician could provide. Examples of such interventions include electronic reminders, pharmacist-led interventions, and healthcare professional education to help patients better remember to take their medications.15
a = Definition adapted from Shojania et al.42
Comments