Abstract

The aim of this study was to describe an integrated informatics approach to aggregating and displaying clinically relevant data that can identify problems with medication adherence and facilitate patient–provider communication about strategies to improve medication use. We developed a clinical dashboard within an electronic health record (EHR) system that uses data from three sources: the medical record, pharmacy claims, and a personal health record. The data are integrated to inform clinician–patient discussions about medication adherence. Whereas prior research on assessing patterns of medication adherence focused on a single approach using the EHR, pharmacy data, or patient-entered data, we present an approach that integrates multiple electronic data sources increasingly found in practice. Medication adherence is a complex challenge that requires patient and provider team input, necessitating an integrated approach using advanced EHR, clinical decision support, and patient-controlled technologies. Future research should focus on integrated strategies to provide patients and providers with the right combination of informatics tools to help them adequately address the challenge of adherence to complex medication therapies.

Background and significance

Type 2 diabetes mellitus (T2DM) is a major public health crisis. Over 28 million adults have T2DM, and an additional 79 million are at risk of developing the disease.1 Moreover, these numbers are expected to increase rapidly.2 Direct and indirect healthcare costs attributable to T2DM exceeded US$245 billion in 2012, which accounted for almost 20% of the healthcare costs incurred by the entire US population.3 The prevalence and costs of T2DM continue to rise at nearly epidemic rates globally as well, driven by urbanization, growing increases in obesity, and aging of populations.4

A key finding of major studies investigating the quality of T2DM care being delivered is the discrepancy between available management strategies and outcomes.5–9 Many persons with T2DM do not reflect the quality of management that would be expected given their access to appropriate therapeutics. One factor that may account for this discrepancy is that patients with T2DM may not be adhering to their pharmacologic therapy. Indeed, increasing evidence indicates that patients with T2DM often show poor adherence to prescribed medication therapies.10,11

Even though persons with T2DM often show poor adherence to medication regimens, the reasons why patients do not take their medications as prescribed are poorly understood.12 Previous studies on adherence have relied on patient self-report data on medication use, or they have used more ‘objective’ measures such as gaps in prescription coverage or technologies to determine if patients are taking their medications. An example of a technology to measure adherence is the electronic medication event monitoring system (MEMS) cap, which documents when a pill bottle is opened.13–15 Such approaches have limitations, such as accuracy, reporting and response bias, and limited effect size.16

It is often assumed that the number of medications, and the complexity associated with their appropriate use, is a contributing factor.17,18 However, a poly-pharmacy regimen, while common among patients with T2DM, is not the only cause of poor adherence.11,15 A wider range of factors, including social, psychological, and economic situations, have also been implicated and should therefore be considered by clinicians when working with patients to manage medication therapy.19 Unfortunately, it is not common for primary care providers to routinely assess whether their patients are having difficulty in following a medication regimen.20 Indeed, few have assessed the role of barriers perceived by patients to medications use and how perceived barriers may be addressed by intervention.

To address medication adherence problems facing individuals with complex medication regimens, we have developed an integrated dashboard to help clinicians identify which patients with T2DM may have low adherence and face barriers that inhibit them from taking their medications as prescribed. The dashboard combines objective data on medication possession ratios with laboratory and point-of-care testing data, as well as patient-entered data on perceived barriers to adherence. By integrating information and presenting it through the electronic health record (EHR) at the point of care, we seek to better inform T2DM therapy decision-making processes and increase the communication between patients and their providers about appropriate medication use.

Materials and methods

System description

The clinical dashboard (figure 1) is a Java-based module that plugs into the Regenstrief CareWeb framework, an open-source EHR platform developed by the Regenstrief Institute Center for Biomedical Informatics. When the clinician selects a patient in CareWeb, the dashboard refreshes with content from three distinct sources (figure 2). First, the dashboard requests blood pressure, the latest low-density lipoprotein cholesterol value and latest hemoglobin A1c (HbA1c) value from the patient's medical record. Next, the dashboard calls the clinical decision support system (CDSS) for information about the patient's medication-dispensing events. Finally, the dashboard queries the patient's personal health record (PHR) for data on perceived barriers to medication adherence. The dashboard displays the data from all three sources in distinct areas on the screen.

Figure 1

Screenshot of a clinical dashboard for reviewing key medication adherence information from three integrated sources: electronic health record, pharmacy claims, and personal health record.

Figure 1

Screenshot of a clinical dashboard for reviewing key medication adherence information from three integrated sources: electronic health record, pharmacy claims, and personal health record.

Figure 2

Information flow diagram outlining how the clinical dashboard for medication adherence retrieves and integrates data from the electronic health record, pharmacy claims, and personal health record. CDSS, clinical decision support system; EHR, electronic health record; PDC, proportion of days covered; PHR, personal health record; T2DM, type 2 diabetes mellitus.

Figure 2

Information flow diagram outlining how the clinical dashboard for medication adherence retrieves and integrates data from the electronic health record, pharmacy claims, and personal health record. CDSS, clinical decision support system; EHR, electronic health record; PDC, proportion of days covered; PHR, personal health record; T2DM, type 2 diabetes mellitus.

The CDSS, an independent web service in the Regenstrief infrastructure, provides the dashboard with information on current adherence to prescribed medications for treating T2DM and cardiovascular risk factors. The CDSS gathers medication refill data from the Indiana Network of Patients Care (INPC), one of the nation's largest and most tenured health information exchange networks with more than 55 hospitals as well as data sources such as Surescripts, a pharmacy benefits manager clearinghouse, and pharmacy claims-based dispensing data from payers including Medicaid.21,22 Using pharmacy claims data available from the INPC, the CDSS calculates the proportion of days covered (PDC), which has been shown in numerous studies to accurately identify patients who fail to fill or refill their medications as directed by their physician or pharmacist for any number of reasons.23

The PHR was implemented on the Open Medical Record System (OpenMRS) platform,24,25 although the dashboard could be integrated with any PHR. OpenMRS includes a forms module that allows collection of standardized data from patients. Using the forms module, we implemented a five-point Likert-style, validated questionnaire developed by researchers at the Diabetes Translational Research Center affiliated with the Indiana University School of Medicine.26,27 The questionnaire uses 20 items to assess possible barriers to medication adherence. For example, valid responses as to why patients may not take their prescribed medications include ‘I can't afford them’ and ‘I just forget to take them.’ Individual items are grouped into categories and ranked on the basis of aggregate scores from patient responses. The three highest ranking categories are stored for retrieval by the dashboard.

System development

User-centered design principles were used to develop the system. Clinical and informatics experts outlined high-level system functions and goals. Then, with input from primary care physicians and T2DM educators, engineers and human-computer interaction specialists iteratively developed the dashboard and PHR interfaces. Multiple iterations were reviewed by the extended team of developers and practicing clinicians. For example, early iterations used a stoplight metaphor to represent medication adherence. Multiple stoplights were confusing to clinicians, who preferred to see numeric PDC values with abnormal values highlighted. The alternative approach also fits better with the general design of CareWeb and other projects, including the redesign of the Regenstrief Medical Gopher.28

System workflow

Lack of fit between a clinical informatics system and its users, as well as users' work environment, can create inefficiencies and facilitate unintended consequences, which can prevent achievement of the primary aims of the system and lead to a decrease in patient safety.29–32 We implemented the dashboard so it would integrate into existing clinical workflows used in the pilot primary care clinics. Clinicians in these clinics are accustomed to opening patient charts before entering the exam room. Since the dashboard is populated upon patient selection, the data are available for review before the clinical encounter. Furthermore, we implemented a reminder to prompt clinicians that the dashboard is available when they select a diabetic patient whose data can be viewed. We further engineered the dashboard to support a variety of patient-centered medical home models being tested throughout the health system in which we are conducting the pilot. This will enable the dashboard to be used by a variety of clinicians on medical home teams in the future, including pharmacists and social workers, who play critical roles in helping patients address barriers to medication adherence.

Study design

The dashboard and PHR were released into production at the end of 2012. In early 2013, we started enrolling providers and their patients at three community health centers in metropolitan Indianapolis into a 6-month pilot demonstration of the dashboard. We seek to enroll 20–30 clinicians and 150–200 patients to examine the efficacy of the dashboard for improving patient–provider communications about medication adherence.

The pilot will be evaluated using a pre–post quasi-experimental design involving mixed methods33,34 to assess providers' and patients' usage of the system, effects on T2DM control as well as medication adherence, and perceptions of providers and patients regarding clinician–patient discussions about medication adherence. Usage will be captured from system logs. CareWeb and the CDSS capture both logins and clicks as users navigate between the EHR, the dashboard, and other CareWeb components. Perceptions about and occurrence of adherence-related discussions will be captured in surveys of providers and patients before and after the demonstration period. In addition, we will collect demographic information about patients and providers to enable comparison across health center locations and socioeconomic data, which has been shown to affect patient access to the internet, impacting on PHR usage.35,36 Patients will further be asked to complete the survey on medication barriers at least twice during the 6-month demonstration.

Analysis will principally focus on comparing PDC rates before and after the pilot (quantitative). Satisfaction and perceived impact on provider–patient conversations will be a secondary target of the analysis (qualitative). In addition, we will further compare medication barrier responses over time (quantitative) in relation to perceptions of the dashboard (qualitative), conversations (qualitative), and diabetic control (quantitative). Individual quantitative and qualitative analyses will be integrated37 to inform our interpretation of the system's impact on process and health outcomes. Our hypothesis is that the dashboard will improve not only the volume and quality of conversations between providers and patients but also medication adherence rates and T2DM control.

Discussion

In this paper we present a novel informatics approach to address the challenge of medication adherence in patients with chronic and complex diseases such as T2DM. The goal of our approach is to improve adherence by providing personalized assessment and facilitating meaningful discussion between patient and provider using an integrated informatics strategy. It is the first approach of which we are aware that combines objective and subjective data and integrates EHR with consumer-oriented technologies to address medication adherence. In this context, our approach increases the possibility of patient engagement and meaningful clinical discussion that may improve health outcomes.

In the traditional approach, physicians assess barriers to adherence by asking patients simple, straightforward questions about whether or not they are taking prescribed medications. This approach is not only subject to bias and fear of interpersonal confrontation, but lacks accuracy compared with the use of a validated set of questions. Unplanned, conventional conversations to assess medication adherence are therefore affected by clinicians' underestimating the unknown or uncommon barriers. Asking incomplete, uninformed questions will lead to inaccurate responses from patients and consequently to inaccurate clinical decisions.

In addition, conventional methods rarely maintain the consistency of checking the adherence at each visit. This enables the provider to capture issues affecting adherence that may be situational in nature. By integrating the tool into the EHR and incorporating the system into work practice, it will be more likely to continually maintain the assessment during each clinical visit. This continuous assessment is critical for long-term chronic diseases such as T2DM.

Several studies have investigated potential barriers to medication adherence. However, these studies often analyze the result of the recruited group and apply the sum of this result to each individual. In most cases, individuals are treated on the basis of the average of other groups where individual preference and under-represented groups are neglected. Analyzing data at the individual level and applying it to the same individual will have far more impact than using generic information.38 The growing availability of EHRs, PHRs, and CDSS39–41 for use in practice will make tailored or patient-centered approaches to medication adherence more feasible.

To fully realize the potential of EHR, PHR, and CDS technologies to affect population health outcomes, informatics approaches need to be integrated. Prior studies in informatics to improve adherence have focused on single modalities to change provider or patient behavior. For example, in Vollmer et al,42 an interactive voice response system called patients who appeared to have gaps in refilling their asthma medication. The system was statistically significant in changing adherence, but the mean change was not clinically meaningful. Similarly, a recent systematic review of patient portals identified just one study that demonstrated an effect on T2DM care delivery.43 However, in the one identified study, the portal was found to be associated with a change in medication regimen but had no impact on clinical outcomes as measured by HbA1c and blood pressure.44 In this article, we outline one approach that integrates multiple informatics tools to address medication adherence in a coordinated fashion. While this one approach may not be the only method to stimulate better adherence for patients with chronic illness, we believe that future approaches in informatics need to draw upon the power of EHRs, PHRs, and CDSS to make the measurement and monitoring of medication adherence easier to perform in the context of routine clinical care.

We are currently gathering data on the approach we describe in this article. The evidence we gather will help to demonstrate whether our hunch that an integrated approach is more effective than individually targeting providers or patients is correct. Initial reactions from healthcare administrators, providers, patients, and researchers have been quite positive and reassure us that our integrated dashboard is headed in the right direction. We are eager to share the results with the biomedical informatics community.

Conclusion

A variety of electronic data sources are increasingly available in real-world clinical settings and are underutilized resources for addressing the complex challenge of measuring and monitoring medication adherence. We describe an approach for integrating multiple electronic data sources to inform patient–physician communication regarding medication use. Ultimately we hope the approach can improve clinicians' measurement and monitoring of medication adherence, leading to better outcomes for patients. While our approach is unlikely to be the only way, future research should explore integrated strategies that include both the patient and the wide array of care providers engaged in treating complex chronic diseases such as T2DM.

Acknowledgements

The authors thank Jeremy Leventhal, Jonathan Cummins, and Hui Xiao of the Regenstrief Institute Center for Biomedical Informatics for their time and talent in architecting and developing the systems described in this article.

Contributors

DGM and BED guided the design of the study. AMJ and EOP provided key support for the recruitment and enrollment of patients. BED and AMJ drafted the article. DGM and EOP provided critical revision of the article. BED finalized the article for publication. All authors reviewed and approved the final version.

Funding

This work was supported by Grant Number R34DK092769 from the National Institute for Diabetes and Digestive and Kidney Diseases. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Diabetes and Digestive and Kidney Diseases, the National Institutes of Health, the Department of Veterans Affairs, or the federal Government. Identifiable information on which this report, presentation, or other form of disclosure is based is protected by US federal law, Section 934(C) of the Public Health Service Act, 42 U.S.C. 299C-3(C). No identifiable information about any individuals or entities supplying the information or described in it may be knowingly used except in accordance with their prior consent. Any confidential identifiable information in this report or presentation that is knowingly disclosed is disclosed solely for the purpose for which it was provided.

Competing interests

None.

Ethics approval

Institutional Review Board at Indiana University.

Provenance and peer review

Not commissioned; externally peer reviewed.

References

1
Centers for Disease Control and Prevention
.
Diabetes report card 2012
. In:
US Department of Health and Human Services
, ed.
Atlanta, GA
:
Centers for Disease Control and Prevention
,
2012
:
1
4
. http://www.cdc.gov/Diabetes//pubs/reportcard.htm
2
Boyle
JP
Thompson
TJ
Gregg
EW
et al
.
Projection of the year 2050 burden of diabetes in the US adult population: dynamic modeling of incidence, mortality, and prediabetes prevalence
.
Population Health Metrics
 
2010
;
8
:
29
.
3
Dall
T
Nikolov
P
Hogan
P
.
Economic costs of diabetes in the US in 2002
.
Diabetes Care
 
2003
;
26
:
917
32
.
4
George
B
Cebioglu
M
Yeghiazaryan
K
.
Inadequate diabetic care: global figures cry for preventive measures and personalized treatment
.
EPMA J
 
2010
;
1
:
13
18
.
5
Kim
C
Williamson
DF
Herman
WH
et al
.
Referral management and the care of patients with diabetes: the Translating Research Into Action for Diabetes (TRIAD) study
.
Am J Manag Care
 
2004
;
10
(2 Pt 2)
:
137
.
6
Kim
C
Williamson
DF
Mangione
CM
et al
.
Managed Care Organization and the Quality of Diabetes Care The Translating Research Into Action for Diabetes (TRIAD) study
.
Diabetes Care
 
2004
;
27
:
1529
34
.
7
Levit
K
Smith
C
Cowan
C
et al
.
Trends in US health care spending, 2001
.
Health Aff
 
2003
;
22
:
154
64
.
8
Kerr
EA
Gerzoff
RB
Krein
SL
et al
.
Diabetes care quality in the Veterans Affairs Health Care System and commercial managed care: the TRIAD study
.
Ann Intern Med
 
2004
;
141
:
272
81
.
9
Mangione
CM
Gerzoff
RB
Williamson
DF
et al
.
The association between quality of care and the intensity of diabetes disease management programs
.
Ann Intern Med
 
2006
;
145
:
107
16
.
10
Hanlon
JT
Schmader
KE
Ruby
CM
et al
.
Suboptimal prescribing in older inpatients and outpatients
.
J Am Geriatr Soc
 
2001
;
49
:
200
9
.
11
Vik
SA
Maxwell
CJ
Hogan
DB
.
Measurement, correlates, and health outcomes of medication adherence among seniors
.
Ann Pharmacother
 
2004
;
38
:
303
12
.
12
Benson
J
Britten
N
.
Patients' decisions about whether or not to take antihypertensive drugs: qualitative study
.
BMJ
 
2002
;
325
:
873
.
13
Cramer
JA
.
Microelectronic systems for monitoring and enhancing patient compliance with medication regimens
.
Drugs
 
1995
;
49
:
321
7
.
14
Cramer
JA
.
Enhancing patient compliance in the elderly
.
Drugs Aging
 
1998
;
12
:
7
15
.
15
Cramer
JA
.
A systematic review of adherence with medications for diabetes
.
Diabetes Care
 
2004
;
27
:
1218
24
.
16
Farmer
KC
.
Methods for measuring and monitoring medication regimen adherence in clinical trials and clinical practice
.
Clin Ther
 
1999
;
21
:
1074
90
.
17
Shalansky
SJ
Levy
AR
.
Effect of number of medications on cardiovascular therapy adherence
.
Ann Pharmacother
 
2002
;
36
:
1532
9
.
18
Grant
RW
Devita
NG
Singer
DE
et al
.
Polypharmacy and medication adherence in patients with type 2 diabetes
.
Diabetes Care
 
2003
;
26
:
1408
12
.
19
Mann
DM
Ponieman
D
Leventhal
H
et al
.
Predictors of adherence to diabetes medications: the role of disease and medication beliefs
.
J Behav Med
 
2009
;
32
:
278
84
.
20
Brown
MT
Bussell
JK
.
Medication adherence: WHO cares?
Mayo Clin Proc
 
2011
;
86
:
304
14
.
21
Dixon
BE
Zafar
A
Overhage
JM
.
A framework for evaluating the costs, effort, and value of nationwide health information exchange
.
J Am Med Inform Assoc
 
2010
;
17
:
295
301
.
22
Biondich
PG
Grannis
SJ
.
The Indiana network for patient care: an integrated clinical information system informed by over thirty years of experience
.
J Public Health Manag Pract
 
2004
;
10
:
S81
6
.
23
Sikka
R
Xia
F
Aubert
RE
.
Estimating medication persistency using administrative claims data-page
.
Am J Manag Care
 
2005
;
11
:
449
57
.
24
Mamlin
BW
Biondich
PG
Wolfe
BA
et al
.
Cooking up an open source EMR for developing countries: OpenMRS–a recipe for successful collaboration
.
AMIA Annual Symposium Proceedings
;
2006
,
American Medical Informatics Association
,
2006
:
529
.
25
Wolfe
BA
Mamlin
BW
Biondich
PG
et al
.
The OpenMRS system: collaborating toward an open source EMR for developing countries
.
AMIA Annual Symposium Proceedings
;
2006
,
American Medical Informatics Association
,
2006
:
1146
.
26
Monahan
PO
Lane
KA
Hayes
RP
et al
.
Reliability and validity of an instrument for assessing patients' perceptions about medications for diabetes: the PAM-D
.
Qual Life Res
 
2009
;
18
:
941
52
.
27
Marrero
D
Monahan
P
Lane
K
et al
.
Validation of a scale to measure patient-perceived barriers to medication use
. 2006 International Society for Quality of Life Research meeting abstracts. Qual Life Res 2006;15(Suppl 1):A34–35, Abstract #1223.
28
Biondich
PG
Dixon
BE
Duke
J
et al
.
Regenstrief medical informatics: experiences with clinical decision support systems
. In:
Greenes
RA
. ed.
Clinical decision support: the road to broad adoption
 .
2nd edn.
Burlington, MA
:
Elsevier, Inc
,
2014
; In press.
29
Zayas-Caban
T
Dixon
BE
.
Considerations for the design of safe and effective consumer health IT applications in the home
.
Qual Saf Health Care
 
2010
;
19
(Suppl 3)
:
i61
7
.
30
Harrison
MI
Koppel
R
Bar-Lev
S
.
Unintended consequences of information technologies in health care—an interactive sociotechnical analysis
.
J Am Med Inform Assoc
 
2007
;
14
:
542
9
.
31
Middleton
B
Bloomrosen
M
Dente
MA
et al
.
Enhancing patient safety and quality of care by improving the usability of electronic health record systems: recommendations from AMIA
.
J Am Med Inform Assoc
 
2013
;
20
(e1)
:
e2
8
.
32
Campbell
EM
Guappone
KP
Sittig
DF
et al
.
Computerized provider order entry adoption: implications for clinical workflow
.
J Gen Intern Med
 
2009
;
24
:
21
6
.
33
Johnson
RB
Onwuegbuzie
AJ
.
Mixed methods research: a research paradigm whose time has come
.
Educ Res
 
2004
;
33
:
14
26
.
34
Creswell
J
Klassen
A
Plano Clark
V
et al
.
Best practices for mixed methods research in the health sciences
 .
Washington, DC
:
Office of Behavioral & Social Sciences Research
,
2011
.
35
Tang
PC
Ash
JS
Bates
DW
et al
.
Personal health records: definitions, benefits, and strategies for overcoming barriers to adoption
.
J Am Med Inform Assoc
 
2006
;
13
:
121
6
.
36
Yamin
CK
Emani
S
Williams
DH
et al
.
The digital divide in adoption and use of a personal health record
.
Arch Intern Med
 
2011
;
171
:
568
74
.
37
Zhang
WQ
Creswell
J
.
The use of "mixing" procedure of mixed methods in health services research.
Med Care
  2013;51:e51–7.
38
Ownby
RL
.
Development of an interactive tailored information application to improve patient medication adherence
.
AMIA Annual Symposium Proceedings
;
2005
,
American Medical Informatics Association
,
2005
:
1069
.
39
Jha
AK
Burke
MF
DesRoches
C
et al
.
Progress toward meaningful use: hospitals' adoption of electronic health records
.
Am J Manag
 
2011
;
17
(12 Spec No.)
:
SP117
24
.
40
Patel
V
Jamoom
E
Hsiao
CJ
et al
.
Variation in electronic health record adoption and readiness for meaningful use: 2008–2011
.
J Gen Intern Med
  2013;28:957–64.
41
Dixon
BE
Simonaitis
L
Goldberg
HS
et al
.
A pilot study of distributed knowledge management and clinical decision support in the cloud
.
Artif Intell Med
 
2013
;59:45–53.
42
Vollmer
WM
Feldstein
A
Smith
DH
et al
.
Use of health information technology to improve medication adherence
.
Am J Manag Care
 
2011
;
17
:
SP79
87
.
43
Ammenwerth
E
Schnell-Inderst
P
Hoerbst
A
.
The impact of electronic patient portals on patient care: a systematic review of controlled trials
.
J Med Internet Res
 
2012
;
14
:
e162
.
44
Grant
RW
Wald
JS
Schnipper
JL
et al
.
Practice-linked online personal health records for type 2 diabetes mellitus: a randomized controlled trial
.
Arch Intern Med
 
2008
;
168
:
1776
.

Comments

0 Comments