By 2030, the number of individuals with heart failure (HF) will approach 8 million (Benjamin et al., 2018). HF is the most common cause of hospitalization in patients older than 65 (Lewis et al., 2017). HF costs are estimated to drastically rise to $69.7 billion by 2030, and approximately 50% of individuals die within 5 years of receiving a HF diagnosis (Benjamin et al., 2018). Medications are crucial in HF treatment to alleviate symptoms, prolong life, and reduce hospital admissions (van der Wal & Jaarsma, 2008). Despite the importance of medications, only 50% of individuals with HF adhere to their medication regimens, leading to poor patient outcomes and hospital admissions (Zhang & Baik, 2014).
Medication adherence intervention studies have been conducted with only a small improvement in medication adherence (Conn, Ruppar, & Chase, 2016; Russell, Ruppar, & Matteson, 2011). A small effect on medication adherence rates has been noted when focusing only on the individual when attempting to improve medication adherence. There is a need to consider the environment in which one exists when formulating interventions (Conn, Ruppar, Enriquez, & Cooper, 2016). The SystemCHANGE™ intervention focuses on changing the individual's environment by incorporating medication taking into existing routines with support from individuals who impact these routines using small experiments with feedback (Alemi & Neuhauser, 2005; Moore et al., 2006). This intervention has been found effective in improving eating behaviors (Alemi & Neuhauser, 2005) and increasing exercise (Moore et al., 2006). Testing this intervention in medication adherence research is in its early stages, but it has been used in pilot studies with individuals with inflammatory bowel disease and kidney transplant recipients (Matteson-Kome, Winn, Bechtold, Bragg, & Russell, 2014; Russell et al., 2011).
An accurate approach to measure medication adherence is necessary to quantify intervention effectiveness. Medication adherence can be measured using objective measures (e.g., pill counts, electronic measuring devices, refill reports, supervised dosing, blood drug levels) or subjective measures (e.g., self-report, provider assessments) (Bosworth, Oddone, & Weinberger, 2006). A common objective measurement in the adult population is electronic medication monitoring, which does not rely on recall or self-report. Electronic medication monitoring is a computerized monitoring system in which a chip is located inside the pill bottle and provides data of the date and time the pill bottle is opened (Park, Howie-Esquivel, & Dracup, 2015).
The purpose of the current study was to examine the feasibility and acceptability of a SystemCHANGE™ intervention in improving medication adherence in older adults with HF. This article focuses on challenges with using the medication event monitoring caps.
Sample and Setting
Institutional Review Board approval was obtained prior to the start of the study from the University of Missouri-Kansas City and Belleville Illinois Community. A convenience sampling approach was used to select eligible participants who met inclusion criteria: (a) age 65 or older; (b) HF diagnosis confirmed by an ejection fraction <40%; (c) prescribed diuretics; (d) self-administering medications; (e) able to open an electronic cap; (f) able to speak, hear, and understand English; (g) not hospitalized; and (h) no cognitive impairment. Cognitive impairment was determined by a score ≥4 on the 6-item Mini-Mental State Exam (Tombaugh & McIntyre, 1992).
There was a 2-month screening phase in which adherence to a diuretic was measured using the Medication Event Monitoring System (MEMS®; access http://www.AARDEXgroup.com), which is an electronic package that automatically records reliable dosing events. When participants returned the MEMS cap, an adherence report was generated that included an adherence rate, frequent days missed, and drug holidays. Participants with an adherence rate ≥88% during the screening phase exited the study. The first month of adherence data were excluded when calculating medication adherence to reduce the Hawthorne effect. An adherence rate of 88% was selected because negative health outcomes can occur if adherence rates fall below 88% in the HF population based on a study measuring HF medication adherence using MEMS (Wu et al., 2009).
The SystemCHANGE™ intervention focuses on changing the individual's environment to change behavior using small experiments with feedback (Alemi & Neuhauser, 2005; Moore et al., 2006). The Socioecological Model and Plan-Do-Check Act model were the frameworks for the intervention. The Socioecological Model assumes that changes in the environment will produce changes in the individual (Bronfenbrenner, 1977).
The first step of the SystemCHANGE™ intervention comprised an initial home visit in which the MEMS report from screening was reviewed, important individuals involved in medication taking were identified, environmental routines were listed, cycles of routines were explored, and solutions to improve medication taking were determined collaboratively by the principal investigator (PI; A.M.A.) and participants. During the second step, the two personal systems–focused solutions were discussed with the important individuals for medication taking if applicable and solutions were incorporated into individuals' existing routines. Step three included data collection using the MEMS, which participants continued to use, and evaluation of the solutions. After 1 month, the PI telephoned participants to discuss the previous month's MEMS report. The discussion focused on whether the personal system solutions were working or if new solutions needed to be implemented. Participants continued to implement the solutions and use the MEMS and MEMS diary. After 2 months, the PI again telephoned participants to discuss how the personal system solutions were working and whether any changes needed to be made.
The attention-control intervention was delivered by the PI every month using two educational brochures developed by the American Heart Association that summarized HF and signs to report. At the end of the 2-month intervention, the PI summarized the education received.
After completion of the intervention phase, a 1-month maintenance phase occurred. Participants continued to use the MEMS and MEMS diary and were called at the end of 1 month to return the MEMS via mail.
Medication adherence was measured electronically using MEMS caps, and adherence rates were calculated using the percent of prescribed doses taken on time. Participants were asked to complete a MEMS diary where they recorded any accidental openings, openings to remove pills to take at a later time, or openings to refill. Participants also recorded extra doses of diuretics taken or dates hospitalized in the diary; these data were not monitored. MEMS cap data were corrected using the MEMS diary. Participants using pill boxes were instructed to use Tic Tacs® as a reminder to open their MEMS, which has been found effective in previous studies (Russell et al., 2011). Participants were instructed on how to use the caps and diary via an instructional sheet and were asked during every study interaction if they were having any issues or had any questions about the MEMS caps. Medication adherence was measured continuously during the intervention and maintenance phases for the intervention and attention-control groups.
Feasibility and acceptability were assessed, and recruitment, retention, and attrition rates were tracked throughout the study. Protocol fidelity was tracked through documenting on protocol checklists and making field notes on all participant interactions. Every time a survey was completed by participants, the time it took to complete the survey and any difficulties with any questions were documented. Challenges and inconveniences were noted on the field notes from participant interactions. During the 1-week telephone call, participants were asked if they had any questions about the MEMS caps. At the end of the three phases of the study, an open-ended feasibility and acceptability questionnaire was administered to participants asking about time, inconveniences, and suggestions for improvements.
Descriptive statistics were calculated for demographics, adherence scores, and time duration of surveys and visits. Data were analyzed using SPSS version 25.
Thirty participants were enrolled in the study. Nineteen participants completed the screening phase. Sixteen (53%) participants had medication adherence rates >88%; therefore, they exited the study. Two of three nonadherent participants agreed to continue into the intervention phase; only the participant in the attention-control group completed the study. For the 30 participants entering the screening phase, average age was 68.1 years (range = 54 to 90 years, SD = 8.6 years), average number of current prescription medications was 10.8 (range = 4 to 22, SD = 4.5), and average ejection fraction was 33.5% (range = 10% to 65%, SD = 15.3%). Sixty-six percent of the sample was male and African American. All participants had a cognitive score of 6 of 6 on the Mini-Mental State Exam, meaning no cognitive impairment was present.
Due to slow recruitment, the age limitation was changed to 50 or older and the ejection fraction criteria were removed to enroll more participants. The demographic survey, which was completed at the beginning of screening, took an average of 2.2 minutes (range = 1 to 6 minutes, SD = 1 minute) to complete and participants found no questions to be challenging. The duration of the initial visit prior to the screening phase of the study took an average of 21.8 minutes (range = 11 to 40 minutes, SD = 7.7 minutes). The duration of the home visit at the beginning of the intervention phase was 47 minutes with the SystemCHANGE™ participant and 40 minutes with the attention-control participant. The monthly telephone calls with the attention-control participant totaled 45 minutes. The protocol checklist was followed for the attention-control and SystemCHANGE™ participants.
One acceptability aim addressed challenges with using the MEMS cap. Three participants were confused by the number on top of the MEMS cap and why it reset to zero every night. Three participants sent their caps back before the 2-month screening period ended, one at 2 weeks and two at 7 weeks. One participant continued to use his pillbox and opened the MEMS cap at night to refill a daily pill box.
The one participant in the attention-control group who completed the entire study provided feedback using the acceptability questionnaire. The participant replied “no” when asked about inconveniences of the survey or education intervention and believed the time in the study was “okay.” The participant thought the MEMS cap was practical to use, had a positive effect on medication taking, and said it was a “good reminder” and that he “liked the number on top of cap, let you know if you took your medications or not,” and wanted to continue to use the MEMS caps. The participant's only suggestion for improving the study was measuring more than one medication. The participant responded, “I do not know, couldn't tell you” when asked if he received the SystemCHANGE™ or attention-control intervention.
The purpose of this mixed methods study was to examine the feasibility and acceptability of a SystemCHANGE™ intervention on improving medication adherence in a highly prevalent population of older adults with HF who have unique challenges to medication adherence. This study is distinctive in that it included only nonadherent participants in the intervention phase to avoid the ceiling effect (Conn, Ruppar, Enriquez, et al., 2016). Previous studies have suggested that less adherent participants do not complete studies, which could explain why 11 participants did not complete screening (Bouvy et al., 2003). It is possible the same barriers influencing medication adherence may also hinder study retention.
Multiple recruitment and retention barriers in research studies have been noted in older adults, including poor cognition and health status (McDougall, Simpson, & Friend, 2015). HF is a progressive, debilitating disease that decreases one's health status and leads to cognitive impairment due to a poor cardiac function, which leads to lower brain perfusion (Gottesman et al., 2010). The SystemCHANGE™ intervention is focused on individuals administering their own medications. Participants must be able to follow directions; therefore, adequate cognition is important. One adherence intervention study including older adults with HF reported many participants being too confused to continue in the study (Varma, McElnay, Hughes, Passmore, & Varma, 1999). It may be beneficial to include cognitive screening at multiple time points throughout the study, not just before enrollment, to ensure only cognitively intact participants continue with the protocol. Another strategy could include involving important individuals who assist with medication taking, perhaps targeting them with the SystemCHANGE™ intervention.
Another important consideration is technology use in the older adult population, as medication adherence was measured in this study using MEMS caps. Barriers to technology include personal attitudes and acceptance of the technology. Although the MEMS cap is perceived as being easy to use (Russell et al., 2009), it can be intimidating to some individuals. Ways to increase acceptance of using technology include having a thorough discussion and training of use and explaining the benefits of the technology (Fletcher & Jensen, 2015). The instruction could be delivered by an older adult peer who has had success using the MEMS cap.
An additional barrier to technology use in older adults is related to the underlying chronic disease producing changes in cognitive, motor, or sensory function; therefore, technology needs to be adapted for this population (Fletcher & Jensen, 2015). One study reported that older adult participants had problems physically opening the cap and that the MEMS cap only measured one medication, not accounting for poly-pharmacy commonly seen in older adults (Chambers, O'Carroll, Dennis, Sudlow, & Johnston, 2015).
A lesson learned in the current feasibility study was that participants had challenges with the MEMS caps. For example, two participants did not use the MEMS correctly in conjunction with a pillbox and lacked understanding of the changing number on top of the MEMS cap. Future studies could include further assessment of how older adults use their pillboxes. Participants could show the researcher their medications, where they are stored, and how they administer the medications using the pillbox with the MEMS cap. Participants could conduct a return demonstration using the MEMS cap after thorough education, which would also reinforce how the numbers change on the top of the cap. Written instructions in addition to verbal instructions could be provided on the use of MEMS caps for those with decreased memory (Lewis et al., 2017).
The final lesson learned was during the screening phase of the study. Participants were screened for 2 months using the MEMS cap, and only those with adherence rates <88% moved into the intervention phase. If this cut point was changed from 88% to 95%, more participants could benefit from the intervention and perhaps outcomes could be further improved. Another consideration is whether it was necessary to exclude the first 1 month of the 2-month screening due to Hawthorne effect, which has been documented to last 30 days (De Geest et al., 2006). Although there were some decreases noted in adherence rates between the first and second month of screening, most participants were below the 88% cut point at the end of the first month of screening. Perhaps moving nonadherent participants into the intervention phase at the 1-month point could prevent participants from not completing the screening phase because they would be receiving more contact from the PI. Moving into the intervention phase sooner, as opposed to dropping from the study, could lead to an improvement in their medication taking and HF outcomes.
Limitations of the current study include recruiting from one site and having a small sample size, thereby decreasing generalizability, even though the goal of this pilot study was testing feasibility and acceptability. Additional limitations were a low consent rate and not obtaining demographics on those who did not consent, which would have allowed for comparison of demographics with those who did consent to participate.
Additional pilot testing is needed with this population prior to conducting a fully powered study. Using multiple sites can help with recruitment and generalizability. Further knowledge can be gained through interviewing participants who dropped from the study to determine study barriers. Cognitive status must also be considered and assessed throughout the study process and education should be adapted accordingly. Education should also focus on addressing MEMS challenges and technology acceptance. Changes could be made in future studies regarding the adherence cut point and shortening the screening phase.
Feasibility and acceptability of the SystemCHANGE™ intervention on improving medication adherence in older adults with HF were explored. The study findings suggest the need for additional revisions to the research protocol. Lessons learned from this study can guide future pilot studies. Further qualitative work could lead to greater understanding of barriers and facilitators to participation in a research study of this nature using electronic medication monitoring technology for measuring medication adherence in the older adult population.
- Alemi, F. & Neuhauser, D. (2005). A thinking person's weight loss and exercise program. Victoria, BC: Trafford.
- Benjamin, E.J., Virani, S.S., Callaway, C.W., Chamberlain, A.M., Chang, A.R., Cheng, S. & Muntner, P. (2018). Heart disease and stroke statistics—2018 update: A report from the American Heart Association. Circulation, 137(12), e67–e492. doi:10.1161/CIR.0000000000000558 [CrossRef]
- Bosworth, H.B., Oddone, E.Z. & Weinberger, M. (Eds.). (2006). Patient treatment adherence: Concepts, interventions, and measurement. Mahwah, NJ: Lawrence Erlbaum Associates.
- Bouvy, M.L., Heerdink, E.R., Urquhart, J., Grobbee, D.E., Hoes, A.W. & Leufkens, H.G.M. (2003). Effect of a pharmacist-led intervention on diuretic compliance in heart failure patients: A randomized controlled study. Journal of Cardiac Failure, 9, 404–411. doi:10.1054/S1071-9164(03)00130-1 [CrossRef]
- Bronfenbrenner, U. (1977). Toward an experimental ecology of human development. American Psychologist, 32, 513–531. doi:10.1037/0003-066X.32.7.513 [CrossRef]
- Chambers, J.A., O'Carroll, R.E., Dennis, M., Sudlow, C. & Johnston, M. (2015). Personal experiences of electronic measurement of medication adherence in elderly stroke survivors. Drugs & Therapy Perspectives, 31, 167–174. doi:10.1007/s40267-015-0200-6 [CrossRef]
- Conn, V.S., Ruppar, T.M. & Chase, J.D. (2016). Blood pressure outcomes of medication adherence interventions: Systematic review and meta-analysis. Journal of Behavioral Medicine, 39, 1065–1075. doi:10.1007/s10865-016-9730-1 [CrossRef]
- Conn, V.S., Ruppar, T.M., Enriquez, M. & Cooper, P. (2016). Medication adherence interventions that target subjects with adherence problems: Systematic review and meta-analysis. Research in Social & Administrative Pharmacy, 12, 218–246. doi:10.1016/j.sapharm.2015.06.001 [CrossRef]
- De Geest, S., Schäfer-Keller, P., Denhaerynck, K., Thannberger, N., Köfer, S., Bock, A. & Steiger, J. (2006). Supporting medication adherence in renal transplantation (SMART): A pilot RCT to improve adherence to immunosuppressive regimens. Clinical Transplantation, 20, 359–368. doi:10.1111/j.1399-0012.2006.00493.x [CrossRef]
- Fletcher, J. & Jensen, R. (2015). Mobile health: Barriers to mobile phone use in the aging population. On-Line Journal of Nursing Informatics, 19(3). Retrieved from https://www.himss.org/mobile-health-barriers-mobile-phone-use-aging-population
- Gottesman, R.F., Grega, M.A., Bailey, M.M., Zeger, S.L., Baumgartner, W.A., McKhann, G.M. & Selnes, O.A. (2010). Association between hypotension, low ejection fraction and cognitive performance in cardiac patients. Behavioural Neurology, 22, 63–71. doi:10.1155/2010/725353 [CrossRef]
- Lewis, S.M., Bucher, L., Heitkemper, M.M., Harding, M., Kwong, J. & Roberts, D. (2017). Medical-surgical nursing: Assessment and management of clinical problems (10th ed.). St. Louis, MO: Mosby.
- Matteson-Kome, M.L., Winn, J., Bechtold, M.L., Bragg, J.D. & Russell, C.L. (2014). Improving maintenance medication adherence in adult inflammatory bowel disease patients: A pilot study. Health Psychology Research, 2(1), 1389. doi:10.4081/hpr.2014.1389 [CrossRef]
- McDougall, G.J. Jr.. , Simpson, G. & Friend, M.L. (2015). Strategies for research recruitment and retention of older adults of racial and ethnic minorities. Journal of Gerontological Nursing, 41(5), 14–23. doi:10.3928/00989134-20150325-01 [CrossRef]
- Moore, S.M., Charvat, J.M., Gordon, N.H., Pashkow, F., Ribisl, P., Roberts, B.L. & Rocco, M. (2006). Effects of a CHANGE intervention to increase exercise maintenance following cardiac events. Annals of Behavioral Medicine, 31, 53–62. doi:10.1207/s15324796abm3101_9 [CrossRef]
- Park, L.G., Howie-Esquivel, J. & Dracup, K. (2015). Electronic measurement of medication adherence. Western Journal of Nursing Research, 37, 28–49. doi:10.1177/0193945914524492 [CrossRef]
- Russell, C., Conn, V., Ashbaugh, C., Madsen, R., Wakefield, M., Webb, A. & Peace, L. (2011). Taking immunosuppressive medications effectively (TIMELink): A pilot randomized controlled trial in adult kidney transplant recipients. Clinical Transplantation, 25, 864–870. doi:10.1111/j.1399-0012.2010.01358.x [CrossRef]
- Russell, C.L., Owens, S., Hamburger, K.Q., Thompson, D.A., Leach, R.R., Cetingok, M. & Wakefield, M.R. (2009). Medication adherence and older renal transplant patients' perceptions of electronic medication monitoring. Journal of Gerontological Nursing, 35(10), 17–21. doi:10.3928/00989134-20090903-06 [CrossRef]
- Russell, C.L., Ruppar, T.M. & Matteson, M. (2011). Improving medication adherence: Moving from intention and motivation to a personal systems approach. Nursing Clinics of North America, 46, 271–281. doi:10.1016/j.cnur.2011.05.004 [CrossRef]
- Tombaugh, T.N. & McIntyre, N.J. (1992). The mini-mental state examination: A comprehensive review. Journal of the American Geriatrics Society, 40, 922–935. doi:10.1111/j.1532-5415.1992.tb01992.x [CrossRef]
- van der Wal, M.H.L. & Jaarsma, T. (2008). Adherence in heart failure in the elderly: Problem and possible solutions. International Journal of Cardiology, 125, 203–208. doi:10.1016/j.ijcard.2007.10.011 [CrossRef]
- Varma, S., McElnay, J.C., Hughes, C.M., Passmore, A.P. & Varma, M. (1999). Pharmaceutical care of patients with congestive heart failure: Interventions and outcomes. Pharmacotherapy, 19, 860–869. doi:10.1592/phco.19.10.860.31565 [CrossRef]
- Wu, J.-R., Moser, D.K., De Jong, M.J., Rayens, M.K., Chung, M.L., Riegel, B. & Lennie, T.A. (2009). Defining an evidence-based cutpoint for medication adherence in heart failure. American Heart Journal, 157, 285–291. doi:10.1016/j.ahj.2008.10.001 [CrossRef]
- Zhang, Y. & Baik, S.H. (2014). Race/ethnicity, disability, and medication adherence among Medicare beneficiaries with heart failure. Journal of General Internal Medicine, 29, 602–607. doi:10.1007/s11606-013-2692-x [CrossRef]