Electronic adherence monitoring (EAM) is increasingly used to improve adherence. However, there is limited evidence on the effect of EAM in across chronic conditions and on patient acceptability. We aimed to assess the effect of EAM on adherence and clinical outcomes, across all ages and all chronic conditions, and examine acceptability in this systematic review and meta-analysis.
A systematic search of Ovid MEDLINE, EMBASE, Social Work Abstracts, PsycINFO, International Pharmaceutical Abstracts and CINAHL databases was performed from database inception to December 31, 2020. Randomised controlled trials (RCTs) that evaluated the effect of EAM on medication adherence as part of an adherence intervention in chronic conditions were included. Study characteristics, differences in adherence and clinical outcomes between intervention and control were extracted from each study. Estimates were pooled using random-effects meta-analysis, and presented as mean differences, standardised mean differences (SMD) or risk ratios depending on the data. Differences by study-level characteristics were estimated using subgroup meta-analysis to identify intervention characteristics associated with improved adherence. Effects on adherence and clinical outcomes which could not be meta-analysed, and patient acceptability, were synthesised narratively. The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guideline was followed, and Risk of bias (RoB) assessed using the Cochrane Collaboration’s RoB tool for RCTs. The review is registered with PROSPERO CRD42017084231.
Our search identified 365 studies, of which 47 studies involving 6194 patients were included. Data from 27 studies (n = 2584) were extracted for the adherence outcome. The intervention group (n = 1267) had significantly better adherence compared to control (n = 1317), (SMD = 0.93, CI:0.69 to 1.17, p<0.0001) with high heterogeneity across studies (I2 = 86%). There was a significant difference in effect according to intervention complexity (p = 0.01); EAM only improved adherence when used with a reminder and/or health provider support. Clinical outcomes were measured in 38/47 (81%) of studies; of these data from 14 studies were included in a meta-analysis of clinical outcomes for HIV, hypertension and asthma. In total, 13/47 (28%) studies assessed acceptability; patient perceptions were mixed.
Patients receiving an EAM intervention had significantly better adherence than those who did not, but improved adherence did not consistently translate into clinical benefits. Acceptability data were mixed. Further research measuring effects on clinical outcomes and patient acceptability are needed.
Citation: Chan AHY, Foot H, Pearce CJ, Horne R, Foster JM, Harrison J (2022) Effect of electronic adherence monitoring on adherence and outcomes in chronic conditions: A systematic review and meta-analysis. PLoS ONE 17(3): e0265715. https://doi.org/10.1371/journal.pone.0265715
Editor: Ivan D. Florez, Universidad de Antioquia, COLOMBIA
Received: March 18, 2021; Accepted: March 7, 2022; Published: March 21, 2022
Copyright: © 2022 Chan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: AC provides freelance consultancy to UCL-Business spin-out company Spoonful of Sugar Ltd., and Janssen-Cilag, and has received research grants from A+ charitable trust (Auckland District Health Board), Maurice and Phyllis Paykel trust, Universitas 21, New Zealand Pharmacy Education Research Fund, Auckland Academic Health Alliance, Asthma UK, University of Auckland, Health Research Council, and Oakley Mental Health Foundation, outside the submitted work, is the holder of the Robert Irwin Postdoctoral Fellowship; HF provides freelance consulting UCL-Business spin-out company Spoonful of Sugar Ltd.; RH has received research grants from NIHR, and Asthma UK, and consultancy fees from Medical Innovation Academic Consortium (CASMI), AbbVie, Amgen, Biogen, Idec, Gilead Sciences, GlaxoSmithKline, Janssen, Pfizer, Roche, Shire Pharmaceuticals, MSD, Astellas, AstraZeneca, DRSU, Novartis, Universitatsklinikum Hamburg-Eppendorf, and Teva Pharmaceuticals, and is the Founder and Director of UCL-Business spin-out company Spoonful of Sugar Ltd., outside the submitted work. All other authors declare no financial relationships with any organisations that might have an interest in the submitted work in the previous three years and no other relationships or activities that could appear to have influenced the submitted work. This does not alter our adherence to all PLOS ONE policies on sharing data and materials.
Poor medication adherence costs the US health system between $100 and $300 billion of avoidable health care costs annually, and is associated with increased morbidity and mortality [1, 2]. Despite the large body of literature on adherence, medication adherence remains suboptimal . Interventions to improve adherence have had only modest impacts on adherence, and have uncertain long-term sustainability due to the short trial durations and need for intensive resources [1, 3]. Digital solutions can address some of these concerns by potentially improving intervention sustainability through automation and reduce resources for implementation . Exploring new ways of delivering healthcare is essential with the COVID-19 pandemic and increased pressures on health systems [5–7].
Electronic adherence monitoring (EAM)  use electronic devices that record medication-taking, usually the time and date of each dose. These medication monitors are increasingly used as part of strategies to improve adherence. EAM is seen as the gold standard of adherence measurement due to their objectivity and data recording accuracy [4, 8], and can be used to improve adherence through direct patient reminders for medication-taking , and/or by facilitating adherence feedback to the patient [10, 11]. Previous reviews have looked at the effect of certain features of EAM and associated electronic devices such as reminders [7, 12–15], medication packaging , or adherence feedback , on adherence. The reviews generally report a positive effect on adherence [7, 12–15] however no reviews have examined EAM specifically, or across all chronic conditions rather than specific conditions [14, 15]. Christensen et al. for example conducted a systematic review of studies on EAM for oral antihypertensive medicines, and found that most reported average adherence rates above 80%, though adherence did vary from 0 to 101% . The authors did not perform a meta-analysis. In a systematic review and meta-analysis by Yaegashi et al., adherence as measured by EAM was reported to be 71% for antipsychotics in schizophrenia . Lee et al. conducted a meta-analysis of RCTs of EAM in children with asthma and reported that the EAM group was 1.50 times more likely to adhere to inhalers compared with the control group . However, these reviews have not included clinical outcome data , or where there is outcome data, the review has not been systematic  or did not include a meta-analysis , or focused on specific populations or medication [17, 18].
Given the costs of poor adherence, and the increasing investment into EAM to improve adherence, there is a need for a high quality systematic review and meta-analysis. The findings will inform public health decision-making and future strategies to improve adherence and outcomes across chronic conditions. This systematic review examines the effect of EAM across all chronic conditions on adherence and clinical outcomes.
Materials and methods
This systematic review was conducted based on Guidelines of the Cochrane Collaboration as described in the Cochrane Handbook of Systematic Reviews of Interventions, version 6.0 (updated July 2019)  and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). The review is registered with PROSPERO CRD42017084231.
Search strategy and selection criteria
We did a systematic search of the literature using Ovid Medline, EMBASE, Social Work Abstracts, PsycINFO, International Pharmaceutical Abstracts, EBM Reviews–Cochrane Central Register of Controlled Trials and CINAHL from database inception to June 1, 2020. Indexing terms based on electronic monitoring, adherence, and intervention were used to develop the search strategy; full details are in S1 Appendix. No language or participant type limit was used. This search strategy was supplemented by a manual search of the reference lists of the identified studies to find other relevant studies. All titles and abstracts were screened separately by 2 authors independently (AC first, with a second screen by JH/CP/HF). Full texts were obtained for eligible studies or abstracts that did not have sufficient information for review. Studies that did not meet the inclusion criteria or had reasons for exclusion were not reviewed further, and reasons for exclusion documented.
Inclusion criteria were: a) the intervention evaluated EAM as an intervention to improve medication adherence; b) participants were individuals with chronic conditions, defined as a long-term, persistent health condition lasting 3 months or more ; c) one of the outcome measures was medication adherence, though this did not need to be measured by electronic monitoring, and clinical outcomes did not need to be assessed; and d) the study was a RCT or a controlled clinical trial (to ensure the highest quality of evidence was included). Studies were still eligible for inclusion even if only one group had adherence monitored electronically and the other did not. Cluster trials were eligible for inclusion. EAM was defined as any mechanism or device that measures and records adherence electronically, regardless of whether or not the EAM devices had a reminder function. All studies meeting the inclusion criteria were included regardless of how adherence was measured (via EAM or not), the adherence measurement (objective or self-report), definition of adherence (taking, timing adherence, or difference between two time-points) or analysis method (e.g. mean ± SD, odds ratios). Adherence had to be measured for all participants at the individual level and not at a group level e.g. adherence of individual patients not adherence of patients within one pharmacy site. Studies that used a downstream measure to approximate adherence e.g. adherence knowledge as a proxy for adherence, or used electronic monitoring for adherence measurement only, rather than to improve adherence, were excluded. Studies using a within-subjects design or historical controls were excluded due to the risk of bias arising from factors other than the intervention itself. Studies using contemporary controls were included.
AC and HF extracted the following data for each study: general study information (author, year of publication); study design; study population (age, sex, health condition); study duration; type of EAM used; description of the intervention and control conditions including details on intervention complexity (i.e. how many components were included in the intervention in addition to EAM e.g. whether the intervention used EAM alone, EAM + EAM reminder, or EAM + health professional input, or all of the aforementioned components); method of adherence feedback; timing of the adherence feedback to the individual (immediate or delayed); presence of participant blinding to adherence monitoring function of EAM; how adherence was measured; outcome measures–effect on adherence, clinical outcomes and other findings; and any data on patient perceptions of EAM.
Studies were classified based the chronic condition of the participants in the study, and on how EAM were used in the adherence intervention–either direct-to-patient to improve adherence (e.g. via a reminder or visual feedback), or through an indirect provider-to-patient interaction (e.g. adherence feedback by the health provider), or both. The effect on clinical outcomes, where reported, was classified as “significantly improved”, “trend towards improvement but not significant”, “no effect”, or “worsened”. Patient perceptions of the EAM intervention were categorised as perceptions of EAM or of the adherence intervention.
GRADE was used to rate the quality of evidence according to risk of bias, consistency, directness, precision and reporting bias . The risk of bias (RoB) in each included study was assessed independently, using the Cochrane Collaboration’s RoB tool for RCTs , by AC and HF/CP. A funnel plot was used to evaluate the effect of publication bias.
Data were pooled from studies which reported medication adherence of participants in the intervention and control groups. The primary outcome measure was the difference in medication adherence between intervention and control groups, expressed as the mean difference (MD) with 95% confidence intervals (CIs), and derived using random-effects models to account for both within-study and between-study variance (tau-squared [τ2]).SMD was used to account for different measures of adherence reporting. The SMD expresses the intervention effect in standard units rather than the original units of measurement and shows the difference in mean effects between the intervention and control groups divided by the pooled standard deviation of participants’ outcomes . A positive SMD (i.e. greater than 0) indicates better adherence in the intervention group compared to control. To accommodate for differences between studies in adherence measures, adherence definitions, and analysis methods, the generic inverse variance outcome type was used. All estimates are presented as SMDs. As medication adherence differs significantly among different health conditions, the adherence outcome was analysed by chronic disease. There were two secondary outcomes: the difference in medication adherence between intervention and control groups in studies that a) measured taking adherence (i.e. studies that measured the percentage of prescribed doses taken, regardless of timing ) and b) studies that used an objective measure of adherence.
We contacted authors for studies that did not have data or could not be converted into the required format for meta-analysis (i.e. SMD and standard deviation). The primary outcome measure was chosen for studies that had multiple medications or dosing regimens or adherence measures (e.g. timing and taking adherence); or reported multiple intervention or control groups. If outcomes were reported at multiple time points, we extracted these and included the latest reported time point. We excluded post-intervention follow-up data. If multiple measures of adherence were used, we included the most objective measure in the review. Reporting in the study of one or more of the outcomes listed here was not an inclusion criterion for this review. Intention-to-treat (ITT) or ’full analysis set’ analyses were used where these were reported. For studies that did not report data in a form that allowed meta-analysis, this data were reported narratively (e.g. as medians and interquartile ranges for each group).
Studies that included all as primary or secondary outcomes, decisions were made in the following order: the most recent (or end) time-point in the intervention period; the group with the largest number of participants; the dosing regimen with the least daily doses; taking adherence; and the intervention group that most closely represented EAM alone and the control group that most closely represented usual care.
The Ԛ test  and the I2 index were used to identify and quantify study heterogeneity respectively. Cochrane RevMan Software version 5.4  was used for all statistical analyses, and p-values <0.05 denoted statistical significance.
Study-level characteristics: Subgroup analysis.
We conducted pre-specified subgroup analyses to investigate the effect of the following study-level characteristics on adherence: 1) age; 2) healthcare setting; 3) intervention complexity (i.e. EAM alone, EAM + EAM reminder, EAM + health professional input or all aforementioned components); 4) method of adherence feedback; 5) timing of adherence feedback to the participant (immediate or delayed); 6) study duration; and 7) participant blinding to the EAM adherence monitoring function.
Effect on clinical outcomes. Based on the heterogeneity of the different disease measures, we conducted meta-analyses only when this was meaningful, that is, when treatments, participants, and the underlying clinical question were similar enough for pooling to make sense, for example, where studies used similar outcome measures. We therefore performed a meta-analysis by grouping together similar measure types according to the chronic disease. For studies that did not report data in a form that allowed
meta-analysis, this data were reported narratively (e.g. as medians and interquartile ranges for each group).
GRADE was used to rate the quality of evidence according to risk of bias, consistency, directness, precision and reporting bias . The risk of bias (RoB) in each included study was assessed independently, using the Cochrane Collaboration’s RoB tool for RCTs , by AC and HF/CP. Funnel plots were used to evaluate the effect of publication bias.
Data were pooled from studies which reported the clinical outcome of interest in the intervention and control groups. Continuous data (data that can take any numerical value) was analysed as mean differences (MDs) using a random-effects model and 95% confidence intervals (CIs) if the measures used in the studies were reported on the same scale. If data were reported using different measures or scales, SMDs were used to account for the different methods of measurement (e.g. different asthma control questionnaires. If both change from baseline and endpoint scores were available for continuous data, change from baseline scores were used. For data reported as rates or proportions, this was analysed as risk ratios using a random-effects model and by inverse variance. If a study reported outcomes at multiple time points, we used the measure taken at the last follow-up. Intention-to-treat (ITT) or ’full analysis set’ analyses were used where these were reported.
Patient acceptability of the EAM intervention. Don patient acceptability were synthesised narratively.
Our search identified 565 records, of which 365 were screened after duplicates were removed. 66 full-text articles were assessed for eligibility and 47 studies involving 6194 patients met the inclusion criteria for inclusion in this systematic review (Fig 1). Table 1 describes the main characteristics of the studies. Study population size ranged from 6  to 784  participants (mean = 128, median = 80). Most (n = 41, 87%) were in adults with only 6 studies in children. The most common conditions were in asthma (n = 10, 21%) [11, 27–35], or human immunodeficiency virus (HIV) (n = 9, 19%) [36–44], or hypertension (n = 6, 13%) [14, 45–49].
The most common EAM device type was an electronic cap fitted onto an oral medication bottle (the ‘Medication Event Monitoring System (MEMS)‘ (n = 14, 30%) [26, 37, 39, 40, 42, 43, 47, 48, 50–55] or similar (n = 10, 21%) [38, 41, 46, 49, 56–61]. The four (9%) other studies of oral medicines used electronic medication blister cards [14, 62–64]. Some used EAM devices that fitted to a specific medication formulation such as inhalers (n = 12, 26%) [11, 27–35, 65, 66] or eyedrops (n = 2, 4%) [67, 68]. Five (11%) [36, 44, 45, 69, 70] used an integrated medication management system (MMS) which included recording of dosing times and symptoms, reminders about lifestyle and /or medication-taking, and information about disease control.
Most studies, except two early studies [46, 63], used electronic monitoring to measure adherence, though this was frequently used with other measures such as serum medication levels [39, 40, 59, 63], self-report [14, 38, 39, 41, 43, 49, 51–55, 64–66, 71], adherence questionnaire [36, 45, 51, 61], pill count [38, 48, 51, 59, 69, 70], canister weight [65, 66],or prescription refill data [52, 55].
Of the 47 included articles, 27 (57%) studies provided sufficient data for the primary outcome meta-analysis, through the published manuscript or author contact. Fourteen (30%) authors were uncontactable, three (6%) studies did not report on adherence differences in both control and intervention groups and two (4%) authors could not provide further data. From these 27 included studies, 25 were eligible for the secondary outcome analysis of studies measuring taking adherence, and 24 of studies using objective adherence measures.
Effect of EAM on medication adherence
The primary outcome analysis of pooled data from 27 studies (n = 2584) showed that the intervention group (n = 1267) had significantly better adherence than control (n = 1317), (MD = 0.93, CI: 0.69 to 1.17, p = <0.0001). Statistically significant heterogeneity was present (Q = 187.65, p = <0.0001) and of substantial degree (I2 = 86%). The forest plot for all studies is shown in Fig 2.
SE, standard error; CI, confidence intervals for effect size.
The secondary outcome analysis of the 25 studies (n = 1127 in the intervention, n = 1175 in the control) that measured taking adherence showed a positive effect size (MD = 0.95, CI: 0.69 to 1.22, p = <0.0001). Simlarly, analysis of the 24 studies using objective measures of adherence (n = 1131 intervention, n = 1164 control group) also showed a statistically significant positive effect size (MD = 1.02, CI: 0.76 to 1.28, p = <0.0001).
Study-level characteristics: Subgroup analysis.
Separate subgroup analyses are shown in Table 2. All subgroups had positive effect sizes, with no significant differences among subgroups except for the “complexity of intervention” variable (p = 0.01). EAM-only interventions did not improve adherence (SMD = 0.24, CI:-0.35 to 0.84) as much compared to interventions where the EAM was used with a reminder and / or health professional input (SMD ranged from 0.73 (EAM + health professional input) to 1.51 (EAM + EAM reminder) (CI range: -0.54–2.22).
Effect on clinical outcomes
Table 3 shows the effects of EAM on clinical outcomes across the 47 included studies summarised narratively. Nine studies (19%) did not assess clinical effect [26, 38, 54–57, 65, 66, 68]–most these (7/9) were of a shorter study duration (six months or less). There were 38 (81%) studies that reported clinical outcomes; ten (26%) reported statistically significant improvements [27, 28, 30, 32, 37, 39, 46, 47, 52, 62].
Due to the wide heterogeneity of the types of measures used to assess clinical outcomes, meta-analyses could only be conducted by disease group, where there were three or more studies reporting on clinical effect in a similar way. Across the 38 studies that reported on clinical outcomes, meta-analyses could be conducted using the following outcome measures–viral load for HIV, blood pressure for hypertension; and asthma control measures for asthma. For studies in HIV, 5 [36, 37, 72–74] of the 9 HIV studies reported on proportion of patients with undetectable viral load as an outcome;. In hypertension, 4 [49, 75–77] of the 7 studies reported on mean change in blood pressure (systolic and diastolic); and in asthma, 5 of the 9 studies reported on change in asthma control [28, 32, 34, 78, 79], in a way that could be included in the meta-analysis. Table 4 shows the evidence profile for these three clinical outcomes, and for adherence.
There were 3 studies that reported outcomes in transplant patients, however only 2 reported numbers of patients with rejection [51, 58] while the other reported on 5-year event-free survival rates . The remaining 11 studies reporting on outcomes were in a range of health conditions: heart failure (2/10); diabetes (2/10); glaucoma; bipolar disorder; percutaneous coronary intervention; psychosis; COPD; coronary artery disease; and schizophrenia, all reporting on outcomes using different measures (e.g. symptom scores or hospitalisation rates) and therefore could not be synthesised via meta-analysis.
Fig 3 shows the effect of EAM on clinical outcomes in HIV, hypertension and asthma. For all outcomes, the analysis crossed the boundary of now effect, but showed a non-significant effect favouring the EAM group. For HIV, those receiving the intervention had a 1.08 (95% CI 0.91–1.29, p = 0.39) chance of having an undetectable viral load. In hypertension, the EAM group achieved a lower systolic and diastolic blood pressure by 2mmHg, though the 95% CI was wide. For asthma, a SMD of 0.09 (95% CI, -0.07–0.24, p = 0.43) was seen, indicating a small but positive improvement in asthma control favouring the intervention, though this was not significant. Table 3 describes narratively the outcomes for other health conditions, which report a range of effect from significant improvements in outcome to no effect. There were 15 studies that reported no effect on outcomes; these had improvements in adherence from 2% [43, 48] to 34% , and had a lack of blinding or real-time feedback, with most (11/15 studies) not feeding back adherence in real-time to the participant.
Forest plot of effect of the electronic adherence monitoring intervention compared to control on clinical outcomes by a) viral load for Human Immunodeficiency Virus; b) blood pressure for hypertension; c) asthma control for asthma.
Not all studies reporting clinical benefit had improvements in adherence. Four studies [45, 58, 61, 64] had no significant effect on adherence but reported clinical benefit. Three [36, 41, 69] had improvements in some but not all clinical parameters (e.g. reduced viral load, but no effect on CD4 count ), and two [44, 63] had a negative effect on clinical outcomes but a positive  or no effect  on adherence. In a study of patients with bipolar affective disorder, the EAM intervention group reported higher rates of anxiety, depression and somatism than the controls . Wu et al. reported worsened quality of life in patients with HIV who received the EAM intervention  than those who did not, but no differences in disease control.
Assessment of risk of bias
Most studies had at least one domain rated as having a high risk of bias (n = 40, 85%, S1 and S2 Files). There was a high performance bias in most studies (n = 35, 74%). This is expected as blinding of patients and health professionals is difficult due to the nature of the EAM intervention. Most studies (n = 38, 81%) had low detection bias as adherence and clinical outcome measures were objectively measured. Whist the risk of selective reporting bias was low within each study, overall, there was moderate selective reporting bias as 20 of the 47 studies did not report data in a way that could be meta-analysed.
Assessment of publication bias across studies
The funnel plot (S3 File) indicated acceptable plot symmetry for both adherence and clinical outcomes data, suggesting limited publication bias, though the large amount of study heterogeneity needs to be considered.
Quality of evidence
Table 4 shows the GRADE ratings for adherence and clinical outcomes for the 3 conditions where this could be meta-analysed.
Risk of bias for adherence (S1 and S2 Files) was rated as serious for the adherence outcome due to the high number of studies with unclear allocation concealment and the risk of performance bias, as a results of the inability to blind the intervention to the participants and outcome assessors in most studies, which could affect outcome reporting. For clinical outcomes however, the nature of the outcomes (viral load and blood pressure) are objective measures that are unlikely affected by knowledge of group allocation; as such evidence was not downgraded for HIV and hypertension but was so for asthma. Inconsistency was downgraded and rated as serious for adherence and blood pressure, given the high heterogeneity of the studies.
Indirectness, imprecision and publication bias were not downgraded for adherence, but evidence was downgraded for risk of bias and inconsistency. Quality was upgraded given the strong association and large effect size. This gave an overall rating of moderate certainty in the effect estimate for adherence. For clinical outcomes, overall certainty in the evidence was deemed low, mainly as a result of inconsistency and imprecision as a result of the small sample sizes per condition.
Patient acceptability perceptions of the EAM intervention
Fourteen studies evaluated patient acceptability of the EAM intervention (see S4 File). Nine assessed usability of the device [11, 30, 37, 38, 44, 49, 67, 69, 70] though one study did not report the results ; four focused on the acceptability of the adherence feedback and interaction with health providers [47, 53, 54, 56]; and one study focused on both the device and the health provider interaction .
Patient perceptions of EAM.
Of the nine studies that reported patient perceptions of the device, perceptions were negative in nearly half [37, 44, 69, 70]. In the Velligan et al. study, participants preferred getting medication support from staff rather than from EAM , and had negative feedback, primarily about the device’s reminder beeps. Similarly, Wu et al. reported that the EAM device was too large and too loud, leading to unwelcome questions and possible revelation of the patient’s HIV status . In contrast, the EAM device used by Frick et al., which included a button to silence the alarm, received positive feedback, with 99% stating they would use the vial again and 97% finding the alarm helpful . Other studies reported mixed results, with some participants enjoying EAM as they felt it helped them stay on schedule with their doses, whilst others “hated the device” and felt their lives were regulated by the EAM device, and found EAM to be a nuisance , unnecessary and unattractive [37, 80].
Patient perceptions of the adherence intervention.
Patient perceptions of the adherence intervention were more positive than for the device . Patients and health providers found the adherence review and discussions the most beneficial parts of the intervention and “looked forward” to receiving their adherence data [47, 53, 56, 62]. The adherence feedback did not make patients feel uncomfortable  and was not perceived to be intrusive , or a burden .
Although most studies did not aim to evaluate patient perceptions, four studies attributed recruitment issues, patient drop-outs and non-participation to issues with device acceptability by patients (S4 File) [14, 37, 45, 51]. Training was also identified as a factor to consider for intervention acceptability .
This is the first systematic review and meta-analysis of EAM and the effect on medication adherence and clinical outcomes, across all chronic conditions. To our knowledge, this systematic review and meta-analysis is the largest in this field, comprising 47 RCTs in the systematic review and 27 studies in the meta-analysis of adherence. Whilst there have been systematic reviews across conditions, these have been for multiple types of interventions rather than specifically examining the effect of EAM. Patients receiving an EAM intervention had significantly better adherence compared to those who did not, with a large magnitude of effect (SMD = 0.93). SMD measures effect when studies report efficacy as a continuous measurement, with zero meaning the intervention and control groups have equivalent effects, with SMD increasing as the difference between the intervention and control group increases. An SMD over 0.8 is considered a large effect . Putting this into perspective, when SMD = 0, the probability that the intervention outperforms control is 0.5 (no better than chance); and when SMD = 1, the probability increases to 0.76. In this review, SMD = 0.93, meaning for individuals who receive EAM, there is a approximately a 0.7 probability that their adherence will improve than if they didn’t receive the EAM. As our review found however, this may not consistently translate to clinical benefits, as this appears to vary depending on the population and chronic condition. This highlights the potential of EAM to improve medication adherence in patients with chronic conditions. Similar effect sizes were seen in studies measuring only taking adherence, and in studies using objective adherence measures. We found that the effect of EAM appeared particularly large for asthma and HIV, similar to findings from other reviews, though the number of studies per condition are small, which limits our confidence in the findings. Lee et al. reported that in children with asthma, those receiving an EAM intervention were 1.5 times more likely to adhere than those in the control group . Similarly Christensen et al. noted high adherence rates reported across included studies for HIV populations receiving EAM , though the authors did not conduct a meta-analysis. Our findings are similar to previous reviews of the effect of reminders or adherence feedback on adherence, but the magnitude of effect in prior reviews was smaller and non-significant  or not quantified if findings in studies that only reported results narratively [7, 9, 10, 12, 13]. Previous reviews may have only been limited to one health condition [17, 18], or evaluated only one aspect of EAM (e.g. reminders, adherence feedback, or the packaging), which may explain our larger magnitude of effect. We found improvements in clinical outcomes in HIV, hypertension and in asthma, but none of these reached statistical significance due to the small number of studies that were able to be included.
There are several implications for physicians, researchers, and payers. First, our review found that intervention complexity was important for intervention effectiveness. Studies using an EAM device by itself, without reminders or health provider input, did not improve adherence. This aligns with previous literature showing that complex interventions–i.e. those involving more than one intervention element–are more effective . A systematic review of electronic packaging interventions on adherence, including both RCTs and non-RCTs, found that complex interventions with EAM were the most effective for improving adherence . Second, the delivery format of the EAM intervention did not appear to greatly influence the magnitude of effect. Intervention effectiveness was not influenced by how adherence feedback was provided or the timing of the feedback, nor by the age of the participants, with EAM being effective in both children and adults.
Third, although there were significant improvements in adherence, few RCTs reported corresponding benefits in clinical outcomes. Those that did show clinical improvements reported a greater magnitude of increase in adherence. It is possible that a minimum threshold of percentage adherence change is needed before any clinical change can be achieved, however the threshold is unknown for most conditions and depends on the medication pharmacology . Whilst there are many studies demonstrating the association between adherence and clinical outcomes [86, 87], it is not known whether the relationship is a linear, exponential or logarithmic one, and the relationship is likely to be affected by the disease, medication and patient . We found that, on average, only half of those interventions that improve adherence translate to corresponding improvements in clinical outcomes.
There are several limitations to consider. The impact on adherence was pooled from 27 studies; over 40% of the studies (20/47) did not report adherence data in way that could be meta-analysed. The impact on clinical outcomes is also less clear. Several studies did not measure clinical outcome data or where data were measured, the outcomes were not relevant to the condition. For example, Elixhauser et al. used a general psychological symptom questionnaire  to assess the effect of an intervention to improve lithium adherence, rather than a validated mania scale, which would have better reflect lithium adherence and response . Even in conditions where disease control can be easily measured and validated disease control questionnaires exist, such as in asthma, heart failure, or diabetes, there was large heterogeneity in the measures used and outcomes studied that precluded inclusion in a meta-analysis [11, 14, 27–31, 45–48, 52, 53, 61, 69]. This lack of standardisation in outcome measures across the same disease state makes inter-study analyses and comparisons difficult. In this review, we performed a meta-analysis to synthesise clinical outcomes but only where meta-analysis was meaningful. As outcome measures were highly varied, we opted only group together measures for the same condition where these made clinical sense. The limitation of this approach is the small number of studies that were able to be included, which reduced our confidence in the findings. Additionally, for the studies that could not be included in the meta-analysis, this could only be described narratively and whether studies reported statistically significant benefits. This has limitations as it does not provide information about overall effect size. Questions also remain about the sustainability of the intervention effects–only 21 of the 38 studies that assessed clinical outcomes were of 6 months duration or longer. Whether these benefits are maintained in the long-term for chronic conditions is not known as initial intervention effects may wane over time. Whilst most studies assessed adherence well, by triangulating data from more than one objective adherence measure to evaluate adherence, the measurement of clinical outcomes is less consistent. Future research should use validated markers of disease control that can be used in different research studies and clinical settings. As adherence is only a mediator of therapeutic outcomes, adherence studies should ideally always include a measure of clinical outcomes as an endpoint, as achieving adherence is meaningless if patients are not getting any clinical benefits. This is seen in the studies that reported clinical benefits, but no effect on adherence. Of note, a limitation for our review is how quality of evidence was assessed. We used the Cochrane Risk of Bias 1.0 tool for evidence certainty grading; however, we note that random sequence generation and allocation concealment were often rated as ‘unclear risk of bias’ due to absence of sufficient detail in reported studies. If the Cochrane Risk of Bias 2.0 tool were used, this may potentially downgrade the evidence certainty to ‘very serious’ risk of bias.
Our results emphasise the need to consider the patient in healthcare interventions. Less than a third of the studies reported on patient acceptability, yet findings show that patient perceptions of the devices were often negative. In contrast, patient perceptions about the adherence interventions were positive, particularly about the provision of adherence feedback and the opportunity to interact with a health provider. Issues relating to the loudness of the reminder and device size are common themes that future interventions involving EAM should consider. Devices where patients are able to silence reminders and / or personalise the reminder setting may be more acceptable . Training resources need to be considered, particularly as technologies change and more EAM devices become available. Our findings highlight the importance of including feasibility and patient acceptability measures in future research.
By combining data from RCTs, our systematic review and meta-analysis found that EAM can have a significant effect on medication adherence in chronic conditions. How these adherence improvements translate into clinical benefit is less clear. A quarter of the studies reporting adherence improvements had corresponding clinical benefits. The lack of standardised outcome measures that reliably and accurately reflect disease control prevents us from definitively answering how EAM affects clinical outcomes. Future research should measure clinical outcomes using standardised and validated tools; be of adequate study duration to assess the sustainability of improvements in adherence and disease control in the medium- to long-term; and importantly, evaluate the patient acceptability of the EAM intervention.
S1 File. Summary risk of bias graph (n = 27) for adherence outcome using Cochrane Collaboration’s tool for assessing risk of bias for randomised controlled trials.
Studies are categorised as ‘Low risk’ of bias (green), ‘High risk’ of bias (red) or ‘Unclear risk’ of bias (yellow).
S2 File. Cochrane collaboration’s tool for assessing risk of bias for randomised controlled trials for adherence outcome.
Studies are categorised as ‘Low risk’ of bias (+), ‘High risk’ of bias (-) or ‘Unclear risk’ of bias (?).
S4 File. Summary of patient acceptability data on interventions.
AC was supported by a Lottery Health Doctoral scholarship when this work was started, and is currently supported by the Robert Irwin Postdoctoral Fellowship.
Previous presentation at meeting: Presented as an oral presentation at the 31st Conference of the European Health Psychology Society and British Psychological Society Division of Health Psychology Conference, 23rd– 27th August 2017.
- 1. Nieuwlaat R, Wilczynski N, Navarro T, Hobson N, Jeffery R, Keepanasseril A, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2014;2014(11):CD000011. pmid:25412402
- 2. Iuga AO, McGuire MJ. Adherence and health care costs. Risk Management and Healthcare Policy. 2014;7:35–44. pmid:24591853
- 3. Linn AJ, Vervloet M, van Dijk L, Smit EG, Van Weert JCM. Effects of eHealth Interventions on Medication Adherence: A Systematic Review of the Literature. Journal of Medical Internet Research. 2011;13(4):e103. pmid:22138112
- 4. Chan AHY, Reddel HK, Apter A, Eakin M, Riekert K, Foster JM. Adherence Monitoring and E-Health: How Clinicians and Researchers Can Use Technology to Promote Inhaler Adherence for Asthma. Journal of Allergy and Clinical Immunology: In Practice. 2013;1(5):446. pmid:24565615
- 5. de Jongh T, Gurol‐Urganci I, Vodopivec‐Jamsek V, Car J, Atun R. Mobile phone messaging for facilitating self‐management of long‐term illnesses. Cochrane Database of Systematic Reviews. 2012;(12). pmid:23235644
- 6. Mahtani KR, Heneghan CJ, Glasziou PP, Perera R. Reminder packaging for improving adherence to self‐administered long‐term medications. Cochrane Database of Systematic Reviews. 2011;(9). pmid:21901694
- 7. Vervloet M, Linn AJ, van Weert JC, de Bakker DH, Bouvy ML, van Dijk L. The effectiveness of interventions using electronic reminders to improve adherence to chronic medication: a systematic review of the literature. J Am Med Inform Assoc. 2012;19(5):696–704. pmid:22534082
- 8. Foster JM, Smith L, Usherwood T, Sawyer SM, Rand CS, Reddel HK. The reliability and patient acceptability of the SmartTrack device: a new electronic monitor and reminder device for metered dose inhalers. Journal of Asthma. 2012;49(6):657–62. pmid:22741746
- 9. Kini V, Ho PM. Interventions to Improve Medication Adherence: A Review. JAMA: Journal of the American Medical Association. 2018;320(23):2461–73. pmid:30561486
- 10. van Heuckelum M, van den Ende CHM, Houterman AEJ, Heemskerk CPM, van Dulmen S, van den Bemt BJF. The effect of electronic monitoring feedback on medication adherence and clinical outcomes: A systematic review. PLoS ONE [Electronic Resource]. 2017;12(10):e0185453-e. pmid:28991903
- 11. Foster JM, Usherwood T, Smith L, Sawyer SM, Xuan W, Rand CS, et al. Inhaler reminders improve adherence with controller treatment in primary care patients with asthma. Journal of Allergy and Clinical Immunology. 2014;134(6):1260. pmid:25062783
- 12. Tran N, Coffman JM, Sumino K, Cabana MD. Patient reminder systems and asthma medication adherence: a systematic review. J Asthma. 2014;51(5):536–43. pmid:24506699
- 13. Checchi KD, Huybrechts KF, Avorn J, Kesselheim AS. Electronic medication packaging devices and medication adherence: A systematic review. JAMA—Journal of the American Medical Association. 2014;312(12):1237. pmid:25247520
- 14. Christensen A, Christrup LL, Fabricius PE, Chrostowska M, Wronka M, Narkiewicz K, et al. The impact of an electronic monitoring and reminder device on patient compliance with antihypertensive therapy: a randomized controlled trial. Journal of hypertension. 2010;28(1):194. pmid:19770778
- 15. Wise J, Operario D. Use of electronic reminder devices to improve adherence to antiretroviral therapy: a systematic review. AIDS Patient Care & STDs. 2008;22(6):495. pmid:18462071
- 16. Christensen A, Osterberg LG, Hansen EH. Electronic monitoring of patient adherence to oral antihypertensive medical treatment: a systematic review. Journal of Hypertension. 2009;27(8):1540–51. pmid:19474761
- 17. Yaegashi H, Kirino S, Remington G, Misawa F, Takeuchi H. Adherence to oral antipsychotics measured by electronic adherence monitoring in schizophrenia: a systematic review and meta-analysis. CNS drugs. 2020;34(6):579–98. pmid:32219681
- 18. Lee JR, Leo S, Liao S, Ng WR, Tay TYN, Wang Y, et al. Electronic adherence monitoring devices for children with asthma: A systematic review and meta-analysis of randomised controlled trials. International Journal of Nursing Studies. 2021;122:104037. pmid:34391027
- 19. Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al. Cochrane Handbook for Systematic Reviews of Interventions version 6.1 (updated September 2020): Cochrane; 2020. Available from: www.training.cochrane.org/handbook.
- 20. Bernell S, Howard SW. Use Your Words Carefully: What Is a Chronic Disease? Frontiers in Public Health. 2016;4:159–. pmid:27532034
- 21. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924–6. pmid:18436948
- 22. Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. pmid:22008217
- 23. Demonceau J, Ruppar T, Kristanto P, Hughes DA, Fargher E, Kardas P, et al. Identification and assessment of adherence-enhancing interventions in studies assessing medication adherence through electronically compiled drug dosing histories: a systematic literature review and meta-analysis. Drugs. 2013;73(6):545–62. pmid:23588595
- 24. Hedges LV, Olkin I. Statistical methods for meta-analysis. San Diego, CA: Academic Press. San Diego, CA: Academic Press; 1985.
- 25. Review Manager (RevMan). Review Manager (RevMan). 5.4 ed: The Cochrane Collaboration; 2020.
- 26. Matteson-Kome ML, Winn J, Bechtold ML, Bragg JD, Russell CL. Improving Maintenance Medication Adherence in Adult Inflammatory Bowel Disease Patients: A Pilot Study. Health Psychology Research. 2014;2(1):1389. pmid:26973930
- 27. Burgess SW, Sly PD, Devadason SG. Providing feedback on adherence increases use of preventive medication by asthmatic children. Journal of Asthma. 2010;47(2):198. pmid:20170329
- 28. Chan AH, Stewart AW, Harrison J, Camargo CA Jr., Black PN, Mitchell EA. The effect of an electronic monitoring device with audiovisual reminder function on adherence to inhaled corticosteroids and school attendance in children with asthma: a randomised controlled trial. The Lancet Respiratory Medicine. 2015.
- 29. Charles T, Quinn D, Weatherall M, Aldington S, Beasley R, Holt S. An audiovisual reminder function improves adherence with inhaled corticosteroid therapy in asthma. Journal of Allergy and Clinical Immunology. 2007;119(4):811. pmid:17320942
- 30. Duncan CL, Hogan MB, Tien KJ, Graves MM, Chorney JM, Zettler MD, et al. Efficacy of a parent-youth teamwork intervention to promote adherence in pediatric asthma. Journal of pediatric psychology. 2013;38(6):617. pmid:23248342
- 31. Onyirimba F, Apter A, Reisine S, Litt M, McCusker C, Connors M, et al. Direct clinician-to-patient feedback discussion of inhaled steroid use: Its effect on adherence. Annals of Allergy, Asthma and Immunology. 2003;90(4):411. pmid:12722963
- 32. Morton RW, Elphick HE, Rigby AS, Daw WJ, King DA, Smith LJ, et al. STAAR: a randomised controlled trial of electronic adherence monitoring with reminder alarms and feedback to improve clinical outcomes for children with asthma. Thorax. 2017;72(4):347–54. pmid:27815524
- 33. Sulaiman I, Greene G, MacHale E, Seheult J, Mokoka M, D’Arcy S, et al. A randomised clinical trial of feedback on inhaler adherence and technique in patients with severe uncontrolled asthma. European Respiratory Journal. 2018;51(1):1701126. pmid:29301919
- 34. Vasbinder EC, Goossens LMA, Rutten-Van Molken MPMH, De Winter BCM, Van Dijk L, Vulto AG, et al. E-Monitoring of Asthma Therapy to Improve Compliance in children (e-MATIC): A randomised controlled trial. European Respiratory Journal. 2016;48(3):758–67. pmid:27230437
- 35. Gregoriano C, Dieterle T, Breitenstein AL, Dürr S, Baum A, Giezendanner S, et al. Does a tailored intervention to promote adherence in patients with chronic lung disease affect exacerbations? A randomized controlled trial. Respir Res. 2019;20(1):273. pmid:31796013
- 36. Andrade ASA, McGruder HF, Wu AW, Celano SA, Skolasky RL, Selnes OA, et al. A Programmable Prompting Device Improves Adherence to Highly Active Antiretroviral Therapy in HIV-Infected Subjects with Memory Impairment. Clinical Infectious Diseases. 2005;41(6):875. pmid:16107989
- 37. de Bruin M, Hospers HJ, van Breukelen GJP, Kok G, Koevoets WM, Prins JM. Electronic monitoring-based counseling to enhance adherence among HIV-infected patients: a randomized controlled trial. Health Psychology. 2010;29(4):421. pmid:20658830
- 38. Frick PA, Lavreys L, Mandaliya K, Kreiss JK. Impact of an alarm device on medication compliance in women in Mombasa, Kenya. International Journal of STD & AIDS. 2001;12(5):329.
- 39. Rosen MI, Dieckhaus K, McMahon TJ, Valdes B, Petry NM, Cramer J, et al. Improved adherence with contingency management. AIDS Patient Care & STDs. 2007;21(1):30. pmid:17263651
- 40. Rigsby MO, Rosen MI, Beauvais JE, Cramer JA, Rainey PM, O’Malley SS, et al. Cue-dose Training with Monetary Reinforcement: Pilot Study of an Antiretroviral Adherence Intervention. Journal of General Internal Medicine. 2000;15(12):841. pmid:11119180
- 41. Sabin LL, DeSilva MB, Hamer DH, Xu K, Zhang J, Li T, et al. Using electronic drug monitor feedback to improve adherence to antiretroviral therapy among HIV-positive patients in China. AIDS and Behavior. 2010;14(3):580. pmid:19771504
- 42. Smith SR, Rublein JC, Marcus C, Brock TP, Chesney MA. A medication self-management program to improve adherence to HIV therapy regimens. Patient education and counseling. 2003;50(2):187. pmid:12781934
- 43. Wilson IB, Barton Laws M, Safren SA, Lee Y, Lu M, Coady W, et al. Provider-focused intervention increases adherence-related dialogue but does not improve antiretroviral therapy adherence in persons with HIV. Journal of acquired immune deficiency syndromes. 2010;53(3):338. pmid:20048680
- 44. Wu AW, Snyder CF, Huang I, Skolasky R Jr., McGruder HF, Celano SA, et al. A Randomized Trial of the Impact of a Programmable Medication Reminder Device on Quality of Life in Patients with HIV. AIDS Patient Care and STDs. 2006;20(11):773. pmid:17134351
- 45. Erickson SR, Ascione FJ, Kirking DM. Utility of electronic device for medication management in patients with hypertension: a pilot study. Journal of the American Pharmacists Association: JAPhA. 2005;45(1):88. pmid:15730122
- 46. McKenney JM, Munroe WP, Wright JT Jr. Impact of an electronic medication compliance aid on long-term blood pressure control. Journal of clinical pharmacology. 1992;32(3):277. pmid:1564133
- 47. Ruppar TM. Randomized pilot study of a behavioral feedback intervention to improve medication adherence in older adults with hypertension. Journal of Cardiovascular Nursing. 2010;25(6):470. pmid:20856132
- 48. van Onzenoort HA, Verberk WJ, Kroon AA, Kessels AG, Neef C, van der Kuy PH, et al. Electronic monitoring of adherence, treatment of hypertension, and blood pressure control. American Journal of Hypertension. 2012;25(1):54. pmid:21993365
- 49. Mehta SJ, Volpp KG, Troxel AB, Day SC, Lim R, Marcus N, et al. Electronic Pill Bottles or Bidirectional Text Messaging to Improve Hypertension Medication Adherence (Way 2 Text): a Randomized Clinical Trial. Journal of General Internal Medicine. 2019;34(11):2397–404. pmid:31396815
- 50. Forni Ogna V, Menetrey I, Muller O, Tousset E, Guihard L, Fontana P, et al. Effect of long-term adherence to clopidogrel on the VASP-PRI after elective coronary stent implantation: a randomized controlled study. British Journal of Clinical Pharmacology. 2016;82(6):1486–97. pmid:27447737
- 51. Joost R, Dorje F, Schwitulla J, Eckardt KU, Hugo C. Intensified pharmaceutical care is improving immunosuppressive medication adherence in kidney transplant recipients during the first post-transplant year: a quasi-experimental study. Nephrology Dialysis Transplantation. 2014;29(8):1597. pmid:24914089
- 52. Murray MD, Young J, Hoke S, Tu W, Weiner M, Morrow D, et al. Pharmacist intervention to improve medication adherence in heart failure: a randomized trial. Ann Intern Med. 2007;146(10):714–25. pmid:17502632
- 53. Rosen MI, Rigsby MO, Salahi JT, Ryan CE, Cramer JA. Electronic monitoring and counseling to improve medication adherence. Behaviour Research & Therapy. 2004;42(4):409. pmid:14998735
- 54. Russell C, Conn V, Ashbaugh C, Madsen R, Wakefield M, Webb A, et al. Taking immunosuppressive medications effectively (TIMELink): a pilot randomized controlled trial in adult kidney transplant recipients. Clinical transplantation. 2011;25(6):864. pmid:21077956
- 55. Yeh EA, Grover S, Powell V, Noguera A, Edwards K, Slater R, et al. Impact of an electronic monitoring device and behavioral feedback on adherence to multiple sclerosis therapies in youth: Results of a randomized trial. Neurology Conference: 69th American Academy of Neurology Annual Meeting, AAN. 2017;88(16 Supplement 1). pmid:28393317
- 56. De Geest S, Schafer-Keller P, Denhaerynck K, Thannberger N, Kofer S, Bock A, et al. Supporting medication adherence in renal transplantation (SMART): a pilot RCT to improve adherence to immunosuppressive regimens. Clinical transplantation. 2006;20(3):359. pmid:16824155
- 57. Hardstaff R, Green K, Talbot D. Measurement of compliance posttransplantation—The results of a 12-month study using electronic monitoring. Transplantation proceedings. 2003;35(2):796. pmid:12644142
- 58. Henriksson J, Tyden G, Hoijer J, Wadstrom J. A Prospective Randomized Trial on the Effect of Using an Electronic Monitoring Drug Dispensing Device to Improve Adherence and Compliance. Transplantation. 2016;100(1):203–9. pmid:26588006
- 59. Kozuki Y, Schepp KG. Visual-feedback therapy for antipsychotic medication adherence. International clinical psychopharmacology. 2006;21(1):57. pmid:16317318
- 60. Reddy A, Huseman TL, Canamucio A, Marcus SC, Asch DA, Volpp K, et al. Patient and Partner Feedback Reports to Improve Statin Medication Adherence: A Randomized Control Trial. Journal of General Internal Medicine. 2016:1–6. pmid:27612487
- 61. Sutton S, Kinmonth A, Hardeman W, Hughes D, Boase S, Prevost AT, et al. Does electronic monitoring influence adherence to medication? Randomized controlled trial of measurement reactivity. Annals of Behavioral Medicine. 2014;48(3):293. pmid:24573909
- 62. Brath H, Morak J, Kastenbauer T, Modre-Osprian R, Strohner-Kastenbauer H, Schwarz M, et al. Mobile health (mHealth) based medication adherence measurement—a pilot trial using electronic blisters in diabetes patients. Br J Clin Pharmacol. 2013;76 Suppl 1:47–55. pmid:24007452
- 63. Elixhauser A, Eisen SA, Romeis JC, Homan SM. The effects of monitoring and feedback on compliance. Medical care. 1990;28(10):882. pmid:2232919
- 64. Dobbels F, De Bleser L, Berben L, Kristanto P, Dupont L, Nevens F, et al. Efficacy of a medication adherence enhancing intervention in transplantation: The MAESTRO-Tx trial. Journal of Heart & Lung Transplantation. 2017;06:06. https://dx.doi.org/10.1016/j.healun.2017.01.007.
- 65. Nides MA, Tashkin DP, Simmons MS, Wise RA, Li VC, Rand CS. Improving inhaler adherence in a clinical trial through the use of the nebulizer chronolog. Chest. 1993;104(2):501–7. pmid:8339641
- 66. Tashkin DP, Rand C, Nides M, Simmons M, Wise R, Coulson AH, et al. A nebulizer chronolog to monitor compliance with inhaler use. American Journal of Medicine. 1991;91(4A):33S. pmid:1951401
- 67. Okeke CO, Quigley HA, Jampel HD, Ying Gs, Plyler RJ, Jiang Y, et al. Interventions Improve Poor Adherence with Once Daily Glaucoma Medications in Electronically Monitored Patients. Ophthalmology. 2009;116(12):2286. pmid:19815286
- 68. Hermann MM, Papaconstantinou D, Muether PS, Georgopoulos G, Diestelhorst M. Adherence with brimonidine in patients with glaucoma aware and not aware of electronic monitoring. Acta Opthalmologica. 2011;89(4):e300. pmid:21106046
- 69. Artinian NT, Harden JK, Kronenberg MW, Vander Wal JS, Daher E, Stephens Q, et al. Pilot study of a Web-based compliance monitoring device for patients with congestive heart failure. Heart and Lung: Journal of Acute and Critical Care. 2003;32(4):226. pmid:12891162
- 70. Velligan D, Mintz J, Maples N, Li X, Gajewski S, Carr H, et al. A randomized trial comparing in person and electronic interventions for improving adherence to oral medications in schizophrenia. Schizophrenia bulletin. 2013;39(5):999. pmid:23086987
- 71. Okeke CO, Quigley HA, Jampel HD, Ying Gs, Plyler RJ, Jiang Y, et al. Adherence with Topical Glaucoma Medication Monitored Electronically. The Travatan Dosing Aid Study. Ophthalmology. 2009;116(2):191. pmid:19084273
- 72. Rosen MI, Dieckhaus K, McMahon TJ, Valdes B, Petry NM, Cramer J, et al. Improved adherence with contingency management. AIDS Patient Care and STDS. 2007;21(1):30–40. pmid:17263651
- 73. Sabin LL, DeSilva MB, Hamer DH, Xu K, Zhang J, Li T, et al. Using electronic drug monitor feedback to improve adherence to antiretroviral therapy among HIV-positive patients in China. AIDS and Behavior. 2010;14(3):580–9. pmid:19771504
- 74. Smith SR, Rublein JC, Marcus C, Brock TP, Chesney MA. A medication self-management program to improve adherence to HIV therapy regimens. Patient Educ Couns. 2003;50(2):187–99. pmid:12781934
- 75. Erickson SR, Ascione FJ, Kirking DM. Utility of electronic device for medication management in patients with hypertension: a pilot study. Journal of the American Pharmaceutical Association. 2005;45(1):88–91. pmid:15730122
- 76. McKenney JM, Munroe WP, Wright JT Jr, Impact of an electronic medication compliance aid on long-term blood pressure control. Journal of Clinical Pharmacology. 1992;32(3):277–83. pmid:1564133
- 77. van Onzenoort HA, Verberk WJ, Kroon AA, Kessels AG, Neef C, van der Kuy PH, et al. Electronic monitoring of adherence, treatment of hypertension, and blood pressure control. American Journal of Hypertension. 2012;25(1):54–9. pmid:21993365
- 78. Duncan CL, Hogan MB, Tien KJ, Graves MM, Chorney JM, Zettler MD, et al. Efficacy of a parent-youth teamwork intervention to promote adherence in pediatric asthma. Journal of Pediatric Psychology. 2013;38(6):617–28. pmid:23248342
- 79. Foster JM, Usherwood T, Smith L, Sawyer SM, Xuan W, Rand CS, et al. Inhaler reminders improve adherence with controller treatment in primary care patients with asthma. J Allergy Clin Immunol. 2014;134:1260–68. pmid:25062783
- 80. Foster JM, Reddel HK, Usherwood T, Sawyer SM, Smith L. Patient-perceived acceptability and behaviour change benefits of inhaler reminders and adherence feedback: a qualitative study. Respiratory medicine. 2017;129:39–45. pmid:28732834
- 81. Faraone SV. Interpreting estimates of treatment effects: implications for managed care. P & T: a peer-reviewed journal for formulary management. 2008;33(12):700–11. pmid:19750051
- 82. Van Heuckelum M, Van Den Ende E, Houterman A, Heemskerk C, Van Dulmen S, Van Den Bemt B. Electronic monitoring feedback to improve medication adherence and clinical outcomes. International Journal of Clinical Pharmacy. 2016;38 (6):558. http://dx.doi.org/10.1007/s11096-015-0240-y.
- 83. McDonald HP, Garg AX, Haynes RB. Interventions to enhance patient adherence to medication prescriptions: scientific review. JAMA: Journal of the American Medical Association. 2002;288(22):2868–79. pmid:12472329
- 84. Checchi KD, Huybrechts KF, Avorn J, Kesselheim AS. Electronic medication packaging devices and medication adherence: a systematic review. JAMA: Journal of the American Medical Association. 2014;312(12):1237–47. pmid:25247520
- 85. Assawasuwannakit P, Braund R, Duffull SB. Quantification of the Forgiveness of Drugs to Imperfect Adherence. CPT: pharmacometrics & systems pharmacology. 2015;4(3):e00004–e. pmid:26225235
- 86. DiMatteo MR, Giordani PJ, Lepper HS, Croghan TW. Patient adherence and medical treatment outcomes: a meta-analysis. Med Care. 2002;40(9):794–811. pmid:12218770
- 87. Horwitz RI, Horwitz SM. Adherence to treatment and health outcomes. Archives of Internal Medicine. 1993;153(16):1863–8. pmid:8250647
- 88. Kellner R. A symptom questionnaire. J Clin Psychiatry. 1987;48(7):268–74. pmid:3597327