Effectiveness and Sustainment of a Tailored Over-the-Counter Medication Safety Intervention in Community Pharmacies: A Randomized Controlled Trial

Abstract

Introduction:

The Senior Section is a continuation of a previous intervention that aims to address a gap in medication safety, specifically related to older adult selection and use of over-the-counter medications. The purpose of this paper is to describe the protocol of this study.

Methods:

This study will occur in three phases: an adaptation phase, an effectiveness phase using a randomized controlled trial, and a sustainment phase. This study will take place in conjunction with administrative leadership and pharmacy sites of a regional Midwest integrated health system. Eye tracking technology will inform the adaptation of the intervention and demonstrate effectiveness in the randomized controlled trial. Following the randomized controlled trial, the health system will implement the intervention without research team support. Fidelity and long-term effectiveness outcomes will be collected to demonstrate sustainment.

Discussion:

The potential implications of this study are a complete and sustained redesign of the pharmacy setting to include educational and directional materials on medication safety, leading to a decrease in over-the-counter medication misuse in older adults. This project could provide a road map for pharmacy organizations to tailor and adopt the Senior Section, to ultimately reduce inappropriate over-the-counter medication use in older adults.

Introduction

Over-the-counter (OTC) medications are frequently involved in major adverse drug events (ADEs). Representing a major patient safety concern, OTC use leads to 178,000 hospitalizations in the U.S. annually.¹ Older adults aged 65+ years are at most risk despite representing only 16% of the U.S. population.² These older adults account for approximately 40% of all OTC medication use, making them especially at risk for ADEs.³ The typical older adult uses an average of three OTC medications along with three prescribed medications, while a quarter of older adults use a combination of 10 or more OTC and prescription medications.⁴ Acetaminophen overdose leads to 14,000 visits to the emergency department, as well as up to half of all yearly acute liver failures.^5–7 Diphenhydramine, commonly used as a sleep aid,⁸ has clotting properties that increase an older adults risk of falls through a loss of balance and dissiness.^9,10 Every year, non-steroidal anti-inflammatory drugs (NSAIDs) use by older adults results in over 100,000 preventable ADEs.¹¹ In fact, although anticoagulants account for 10.2% of all ADEs and are a top priority drug class for the Department of Health and Human Services National Plan for ADE prevention, NSAIDs a greater percentage of ADEs at 15.4%.^12–14 Clearly, the use of OTC medications is a cause for concern, especially for higher-risk medications and older adults.

Despite the prevalence of OTC medication use and the cost of accumulated harms, this patient safety gap has remained relatively unnoticed. Most older adults are unaware of the potential OTC safety risks, including inappropriate dosages and drug-drug interactions.^4,15,16 In addition to the gap in patient knowledge, providers have a limited scope of their patient’s OTC medication use.¹⁷ This unintentional lack of communication may multiply treatments unnecessarily and lead to medication overdosing. The Centers for Medicare and Medicaid Services have stressed the importance of pharmacist review of certain prescription orders because of their wide utilization in OTCs and the possibility of duplicate treatments.^18,19

Pharmacy-Related Barriers to OTC Safety

Pharmacies can be an important source of OTCs for older adults. Pharmacists are uniquely trained to advise and counsel patients on medications, including OTCs, and they are considered trustworthy resources for such information. Indeed, 80% of Americans report that pharmacist recommendation would influence their decision to purchase or avoid an OTC medication. ²⁰ In addition, 57% of older adults taking chronic prescription and OTC medications were not using their OTC safely and required a pharmacist intervention.²¹ As a result, pharmacist availability at the point of OTC medication sale has great potential for decreasing older adult OTC misuse. ²⁰

However, pharmacist barriers and older adult barriers exist within the pharmacy environment that limit patient-safety interactions. ^22,23 First, OTC medications that pharmacists identify as risky are often physically distant from the prescription department. Such placement not only limits the pharmacist’s ability to see and proactively approach a customer who might need assistance, it also makes it more time consuming for the pharmacist to approach customers about OTC selection. Second, the distance disincentivizes the pharmacist from leaving the prescription department and aggravates the already high workload and time pressure that is well documented in the community pharmacy setting.^24,25 Finally, pharmacists report that lack of complete and relevant health information impedes their ability to make safe recommendations.^24,26 When pharmacists engage with older adults in the OTC aisles, they do not have access to their prescription profiles for ascertaining potential concomitant disease states or drug interactions. This is particularly problematic because customers frequently are unable to recollect the medications that they take.²³

Consistent with published literature,²⁷ our previous research indicated that many older adults overestimated OTC medication safety because of their OTC status or believed in their safety with other medications unless the label warned against it. Concerningly, older adults demonstrated difficulty using the labeling information to select a medication²⁸ and/or compare medications due to the overwhelming number of marketed OTC products. Even in cases when participants made direct comparisons between two medications, labels were still misinterpreted. For example, a participant stated that she could not take extra acetaminophen because of her prescriptions, but then selected a combination product containing acetaminophen.²⁹

In a pilot study with 20 older adults,³⁰ the present team found surprising levels of potential misuse: at least one indication of misuse was identified for 95% of participants (19 out of 20). In total, 87 potential instances of misuse were documented for 19 participants. This misuse rate was significantly higher than other estimates for this population.²¹

To reduce the prevalence of older adult OTC medication misuse, a participatory design-informed system-redesign intervention was developed,³¹ guided by a human factors conceptual framework to improve patient outcomes called the Systems Engineering Initiative for Patient Safety (SEIPS) 2.0.^32–35

Our Intervention: The Senior Section^™

The Senior Section^™ is an innovative redesign of a pharmacy’s OTC aisles with the objective of promoting safe OTC use by facilitating communication with a knowledgeable and accessible pharmacist. This redesigned structural layout aims to increase awareness of higher-risk OTC medication categories (i.e., pain, cough/cold, allergy, and sleep). Specifically, the Senior Section is comprised of the following features: (A) a dedicated section of well-lit shelving with a curated list of OTC medications that are safe for older adults, (B) proximity and sight line to the prescription department, (C) tools (e.g., lighted magnifying glass), (D) strategically-placed signage, and (E) shelving height to assist visually and physically impaired older adults (Figure 1).

Figure 1:

The Original Senior Section

Using a pre/post implementation design and an overall sample of 87 older adults, implementing the Senior Section in the three pilot pharmacies led to significant reductions in older adult OTC misuse and improved the quality of pharmacist-older adult encounters. ^26,36,37 When extrapolated, these results predict a 27% decrease in OTC misuse among older adults.^38,39 NSAID misuse would decrease by 55%, potentially preventing 55,000 adverse drug events each year.^11,38

The Senior Section is the first and only physical redesign intervention to show effectiveness at reducing OTC medication misuse in older adults without increasing pharmacist workload. Although successful at achieving particular outcomes, intervention implementation was limited to three pharmacies with homogenous layouts and patient populations, evidencing only a signal of effectiveness. Adopting the Senior Section across heterogeneous pharmacies and measuring outcomes with diverse patient populations was the clear next step to evolving and sustaining this medication safety strategy.

Implementing the Senior Section: Next Steps

Our next project (the topic of this paper) was designed to build on our preliminary work in service of the long-term goal of reducing OTC medication misuse and other subsequent harms in older adults, by scaling our Senior Section to a large health system serving a demographically diverse catchment area. A Senior Section that is tailored and culturally adapted to specific settings and patient populations is expected to lead to greater adoption and sustainment by pharmacists and older adults and reduce OTC medication misuse. The rationale for this study was to test Senior Section effectiveness across different pharmacy settings to facilitate its broad professional acceptance as a valuable tool for pharmacies to implement and sustain for preventing OTC medication misuse with older adults.

Methods & Design

To simultaneously test the effectiveness of the Senior Section intervention while evaluating its implementation with researcher support and even after such support has been withdrawn, a Type 2 effectiveness-implementation hybrid design ²⁶ was used. A hybrid design, which blends design components of effectiveness and implementation research, was chosen because it improves the speed of knowledge creation and increases the usefulness and policy relevance of the research being conducted.²⁶ A brief description of the three study aims, and the methods used to investigate those aims, are as follows:

Aim 1.

To adapt the Senior Section based on the characteristics of pharmacies and their patient population. Using insights gained from prior research with pharmacy staff and older adult stakeholders, the study team will collect data on older adult OTC selection activities (using eye tracking technology) and formative interviews with pharmacy staff. This information will be used to adapt the Senior Section physically and culturally to a pilot sample of pharmacies and remote dispensing sites. An Implementation Package will be produced with a focus on sustainment.

Aim 2.

To evaluate the effectiveness of the Senior Section in preventing OTC medication misuse in older adults. The Senior Section is hypothesized to reduce older adult misuse of OTC medications. In addition, the intervention is predicted to not negatively affect pharmacy staff workload. These hypotheses will be evaluated through a post-test only randomized trial of control/test pharmacy sites.

Aim 3.

To evaluate the implementation of the finalized Senior Section. The Senior Section will be implemented in all retail pharmacies affiliated with a large Midwest integrated health system, without research team support. Data will be collected and analyzed to assess fidelity of delivery, long-term effectiveness, and pharmacy staff perceptions of Senior Section sustainability.

Conceptual Model

The guiding premise for this study is that interventions benefit from ongoing optimization as they are applied in different contexts.⁴⁰ As such, the Exploration, Preparation, Implementation, Sustainment (EPIS) framework was selected as the optimal conceptual implementation framework because it allows examination of a change process at multiple levels (organization and pharmacy staff), across time, and through successive stages that build deliberately toward implementation and sustainment (Figure 2).^41,42 Consistent with the human factors engineering and systems approach used for our previous research to conceptualize and pilot-test the Senior Section,⁴³ EPIS takes into account the outer context and the inner organizational contextual factors. The outer context comprises the needs and preferences of diverse patient populations.⁴⁰ The inner context includes organizational culture, pharmacy staff characteristics, staffing policies, fidelity, and quality assurance support. These contextual factors are instrumental to the success of the intervention across the EPIS stages. Finally, EPIS promotes a cyclical approach that includes rapid-cycle testing principles and permits tracking of the intervention’s progressive development through all study aims.

Figure 2:

EPIS Framework (Exploration, Preparation, Implementation, and Sustainment)

Study Sites

An integrated Midwest healthcare system serves as the collaborative partner for this research and is committed to scaling and sustaining the Senior Section to all of its community pharmacies. This healthcare system is the 10th largest not-for-profit, integrated health system in the U.S., and serves nearly 3 million patients annually across more than 500 sites of care. This study will take place in their 63 sites, which are geographically dispersed throughout eastern Wisconsin. Both urban and rural catchment areas are represented, with significant populations of older adults with diverse socioeconomic status.

The pharmacy sites can be categorized into one of two physical layout sizes: (1) large pharmacies which have a front end that is 30 square feet or larger that can accommodate the Senior Section without any significant changes, and (2) small pharmacies with a front end under 30 square feet for which the Senior Section may require additional modification. There are also has tele-pharmacy sites, called “remote dispensing sites,” in which pharmacists and patients are not in the same place and interact using a real-time video link. In these sites, technicians and patients connect with pharmacists working in physically distant pharmacies, potentially in another city, to verify the accuracy and appropriateness of prescriptions and to counsel patients. Despite this unique communication procedure, remote dispensing pharmacies have the same physical layout as traditional pharmacies with a prescription department and a front-end OTC medication area that can be grouped into the two size classifications described above.

Aim 1: Adaptation Phase

The Adaptation Phase consists of a round of separate stakeholder meetings with pharmacy staff and older adults, initial implementation of the Senior Section, collecting older adult medication selection data, pharmacy staff interviews, and concluding with a second round of stakeholder meetings (Figure 3). As suggested when describing the various types of pharmacies, it is expected that the Senior Section will require adaptation to accommodate the differing front-end sizes of the pharmacies. Another consideration for Aim 1 is to explore how to culturally adapt the Senior Section to disadvantaged groups, guided by processes described by Wallerstein and Cabassa,^40,44–46 and recruitment efforts will help ensure a diverse group of older adult stakeholders for this purpose. As such, the planned data collection methods are designed to engage end-users, pharmacy staff and older adults, who can provide salient information about this critical adaptation. At the end of the adaptation phase, these stakeholder groups will offer feedback that will assist in developing an Implementation Package that will guide the strategy for the health system’s broad adoption of the Senior Section in Aim 3.

Figure 3:

Aim 1- Adaptation of the Senior Section to develop the Implementation Package

Recruitment and Data Collection

Stakeholder group meetings:

The Adaptation Phase will involve two stakeholder groups for pharmacy staff (i.e., pharmacists and pharmacy technicians): one group each for small and large pharmacy layouts. Stakeholders will be involved directly in tailoring and culturally adapting the Senior Section intervention. Evidence suggests that active involvement of patients and providers is well suited for the design of interventions in complex work systems that have a great diversity of end-users, and has led to a more in-depth understanding of the reality of problems faced by users and the considerations regarding the implementation and use of the intervention.^47–49 Fourteen pharmacists/technicians will be recruited from small and large pharmacy sites to participate in both pharmacy staff stakeholder groups. Representation from pharmacists/technicians who work or support a remote dispensing pharmacy also will be included. Separate stakeholder groups will be held for each pharmacy size.

In addition, a third stakeholder group will be comprised of an ethnically and racially diverse group of seven older adults who receive prescriptions at a pharmacy. Our largest underrepresented group in the catchment area is African American older adults. Therefore, the study team will plan to include at least three African American older adults in the stakeholder group, with input from a researcher with extensive experience recruiting and engaging African American stakeholders. The study population will be 7 older adults who receive prescriptions at one of the pharmacies selected for this pilot phase, and who would purchase or consider purchasing an OTC medication from one of the following categories: pain, sleep, cough/cold, allergy. The participants must be able to provide verbal consent and be willing to participate in four focus group meetings.

The goal of these stakeholder groups is to gain insight about what Senior Section components they consider essential, require adaptation, and/or need to be removed to contribute to older adults’ safe selection of OTC medications most effectively. This approach also will help optimate intervention features specific to the different pharmacy layouts and patient populations.

Older adult medication selection:

Study participants must be able to provide written informed consent, pass a cognitive screener,⁵⁰ and speak English, and must not have visual impairments that cannot be remedied through corrective lenses or that limit the participant’s ability to utilize eye-tracking glasses.

Recruitment will result in a convenience sample of older adults who frequent the pharmacy during the week of data collection, following implementation of the Senior Section. A UW researcher stationed outside each pharmacy or remote-dispensing site will describe the study to interested participants and obtain consent for participation. Eight participants will be recruited from each of the four pilot pharmacies, for a total of 32 participants.

Older adults will be provided with three symptom scenarios for which they might consider purchasing an OTC medication: sleep, allergy/cough/cold, and pain.³⁰ The older adults will then select the scenario that most resonates with them. Older adults will wear a pair of Tobii Pro Glasses 2 wearable eyetracker ⁵¹ to collect their natural gaze data as they shop for an OTC medication in the pharmacy store. The researcher will affix the eye tracking glasses to the participant and calibrate the participant’s gaze. The eye tracking glasses easily accommodate to an older adult’s vision, by utilizing snap-on corrective lenses that range from –8 to +3 diopter in 0.5 diopter steps. Gaze data will be recorded as the older adult searches for and selects an OTC medication, to compare areas of interest in the aisle. The eye tracking glasses tracks natural gaze at a rate that provides extremely accurate gaze data that can be attributed to specific objects of interest, such as OTC medication packaging (e.g., warning labels) or other available information (e.g., price tags). Once the glasses are calibrated and the recording has started, the researcher will prompt participants with the symptom scenario and will ask them to physically select the medication they would use to treat the symptom.

The researcher will remotely start and stop the eye-tracking recording using a study-provided tablet. As long as the eye tracking device is in Wi-Fi range of the tablet, a live view of the participant’s gaze is available. This is particularly advantageous because it removes an in-aisle presence and potential influence of a researcher while preserving the ability to monitor the participant, start or stop the recording at any time, and intervene if necessary. During each session, the participant’s gaze recording is collected and stored on an SD card. At the completion of each pharmacy’s participation, the researchers will upload the recordings to the Tobii Pro Lab Analyzer program.

After the participants have selected their medication, researchers will stop the eye-tracking recording and begin audio recording. Participants will be asked about how they would use this medication, given the scenario, and a semi-structured interview will be conducted to learn about the older adult’s experience with the Senior Section. A follow-up phone interview will be conducted to collect information about their medications, health literacy, health history, and demographics. Cumulatively, these data will allow us to build our understanding of which Senior Section components are most instrumental in improving older adults’ safe OTC medication selection, older adult decision making, and reading and comprehension of critical health information while they navigate the aisles of the pharmacy.

Pharmacy staff interviews:

Two pharmacy staff members will be recruited by the health system and research team members from each of the pilot site pharmacies, for a total of eight participants. Eight semi-structured interviews will be conducted to elicit descriptions of each respondent’s role and how the intervention was implemented in their pharmacy. Guided by the EPIS framework and the SEIPS 2.0 model, the study team will probe for facilitators and barriers related to each of Proctor’s implementation outcomes (acceptability, adoption, appropriateness, feasibility, fidelity).⁵²

Data Analysis

Eye-tracking data:

The Tobii Pro Lab Analyzer program will be used to characterize the viewing behavior of participants relative to the Senior Section components. First, a series of photo images will be uploaded to the Lab Analyzer to create a “code book” to define areas of interest, such as intervention features (e.g., signage) or product navigation. Older adults’ eye gaze data will then be mapped onto the predefined areas of interest in the recording. The Lab Analyzer uses identified areas of interest to calculate quantitative eye movement measures, including how long a participant gazes at a specific OTC product and the pathway of their gaze (where numbers indicate the order of the eye movement path and larger circles indicate more prolonged gazing) (Figure 4).

Figure 4:

Example of how the Tobii Pro Lab Analyzer may analyze pathway and time length of eye gaze

Areas of interest related to the Senior Section (e.g., price, active ingredient, etc.) can be aggregated together for further analysis (e.g., measuring gaze for all pain-related products). This allows the research team to aggregate the eye movements for all areas of interest related to the Senior Section over multiple points across the timeline. The grouped areas and the individual areas of interest within the Senior Section will be used to determine which Senior Section components are not useful and can be revised. This data-driven refinement will optimize the Senior Section as an intervention.

The following measures will be statistically analyzed: Percentage of participants who fixate their gaze on each area of interest in the Senior Section, total duration of all fixations within an area of interest, and fixation count (number of times that a participant fixates their gaze on an area of interest).⁵³ It is possible that additional measures will be analyzed if they are determined to be of interest to the project objectives.

Qualitative interview data:

Transcripts from pharmacist and technician interviews will be subjected to rigorous qualitative data analysis using techniques that the research team has employed in previous pharmacy medication safety qualitative research.^{48,50,54–59} Deductive content analysis will be used to discover patterns, themes, and categories in our data, guided by our theory-based approach.⁶⁰ A conceptual coding structure will be created and, as each interview transcript is added to the analysis, passages will be classified according to existing codes, and codes will be added as needed. Inductive content analysis will also be conducted to determine unanticipated work system characteristics that may inhibit the successful implementation of the Senior Section. QSR Nvivo software will be used to conduct the analysis.

Implementation Package

The information obtained during Aim 1 will be used to create a full Implementation Package to provide to the partnering health system (for use in Aim 3). The following items are anticipated to be included in the Implementation Package:

1. A description of why pharmacies should implement the Senior Section, including statistics demonstrating the significance of the problem, and the evidence that the Senior Section will reduce OTC medication misuse in their own older adult population.

2. All specifications for the finalized physical redesigned Senior Section including signage, a curated list of medications, tools that will be incorporated in each of the pharmacy layouts with key elements and specifications for what can be adapted and how such adaption will occur.

3. An online training module, adapted from our in-person training guide, that can be uploaded to the health system’s online training and continuing education platform.

4. A plan for pharmacy leadership to provide continuous quality improvement and evaluation of the Senior Section.

5. Example text that can be included in the health system’s monthly email communication highlighting the Senior Section.

Aim 2: Effectiveness Phase

The Effectiveness Phase consists of a post-test only randomized controlled trial (RCT), older adult data collection, and pharmacy staff interviews. Our guiding principle, which is consistent with the EPIS Framework, is that an intervention cannot be optimized prior to implementation, and that the most compelling evidence on the maximal benefit of any intervention can only be realized through ongoing development, evaluation, and refinement in diverse populations.⁶¹ Thus, the goal of Aim 2 is to use a post-test only RCT – randomly selected and paired test and control site pharmacies based on pharmacy and patient characteristics to optimize internal validity – to evaluate the effectiveness of the Senior Section. Randomization and post-test only control group design were chosen precisely to mitigate most threats to internal validity, and the lack of a pre-test removes the threat of testing bias (thereby enhancing external validity).⁶² The primary study outcome is the change in the proportion of older adults who misuse OTC medications between test and control arms. The secondary outcome is the change in how pharmacy staff and patients interact, in relation to OTC medication selection and use.

RCT Randomization

Randomized at the pharmacy-level, test site pharmacies (n = 10) will be matched with control site pharmacies (n = 10) based on pharmacy and patient characteristics. Control site pharmacies will not receive the Senior Section intervention. The partnering health system’s pharmacies have large and small physical layouts, with different amounts of OTC shelving space in front of their prescription department. Further, the health system has 19 remote dispensing sites that exist in both large and small physical layouts. The study team will analyze the effectiveness and implementation of both large and small, and both face-to-face and remote dispensing models.

The 10 matched site pairs will comprise five each of large and small pharmacy pairs, representing 3 pairs each of small and large traditional pharmacies and 2 pairs each of small and large remote dispensing sites. Test and control sites are matched by patient characteristics, proportion of African American patients served, and Area Deprivation Index scores. Area Deprivation Index ranks neighborhoods by socioeconomic status disadvantage based on socioeconomic indicators such as educational standing, family income, home values and poverty rate. ^54,63

Each test site will receive the Senior Section physical redesign intervention. Pharmacy staff will receive a brief face-to-face training to explain the rationale for each component of the Senior Section.

Recruitment and Data Collection

Older adult scenario-based data collection:

Recruitment will follow the same process as the pilot phase. During data collection, 12 participants will be recruited from each traditional test and control pharmacy site, and 18 participants will be recruited from each remote dispensing test and control site. This variance is intended to capture an equal distribution of participants between traditional and remote dispensing sites. This will result a total of 288 participants across 20 pharmacies.

After the older adult participant selects the OTC medication pursuant to the scenario-based data collection described in Aim 1, they will check-out the OTC medication at the pharmacy counter using a “play” credit card provided by the researcher. This will complete the simulated shopping experience. Then, a short semi-structured interview will be conducted to gather information on how the older adult intends to use the medication (e.g., dosing and duration). For test site participants only, participants will be asked questions about their understanding of the Senior Section. The in-person component of the older adult data collection is estimated to take up to 15 minutes, and the follow-up phone interview is estimated to take up to 20 minutes. All older adults will self-disclose their medications, health literacy, health history, and demographics in a brief follow-up phone call.

Pharmacist and technician interviews:

Recruitment will follow the same process as the pilot phase. The study team will recruit two pharmacy staff members from each of the test and control sites, for a total of 40 participants. Interviews will be conducted with test site pharmacy staff only, which will follow the same process as the pilot phase.

OTC Encounter Form data collection:

In an effort to capture pharmacist-patient interactions related to OTC medication selection and use, the interactions that pharmacists and technicians have with consumers (who were not selected as participants) regarding OTC medications will be documented. An OTC Encounter Form will be used to collect recommendation characteristics and outcomes that are in addition to the OTC medication misuse outcome (e.g., referral to a physician, behavioral or medication recommendations). Further, the OTC Encounter Form will be used to quantify the time required to interact with consumers at both the control and test sites. The form has been successfully utilized and was found to be an effective and pragmatic data collection tool.³⁶

Following each interaction with a consumer, the pharmacist and/or technician will document the interaction. The questions on the form include who initiated the conversation, whether the pharmacist or technician left the prescription department to help the consumer, the problem or issue, assistance that the pharmacist or technician provided, and perceived time required to help the consumer. Because all questions have fixed-choice responses, the time required to document each encounter is about one minute. Pharmacist participants in a previous iteration of the study indicated that it was a simple and quick form to complete. To further minimize the data collection burden, both test site and control site pharmacists will collect encounter data for one week, scheduled 2–4 weeks after implementation.

Data Analysis

OTC medication misuse analysis:

The OTC medication misuse outcome will be dichotomized into “safe use” or “misuse.” The same criteria used in our previous studies^35,64 will determine if a selected OTC medication is considered safe use or misuse. The two types of potential misuse that represent our primary outcomes are reported usage that exceeds product labeling guidelines and drug-disease interaction. The two types of potential misuse that represent our secondary outcomes are drug-drug interaction⁵⁵ and drug-age interaction (Table 1).⁵⁶

Table 1:

Classification of Misuse of OTC Medication

	Primary Outcomes		Secondary Outcomes
Type	Reported Usage	Drug-Disease Interaction	Drug-Drug Interaction Risk Ratingb	Drug-Age Interaction
Misuse	Dosage: Exceeds daily product labeling dosage Duration: Exceeds product labeling duration of use	Interaction listed on Beer’s Lista Condition listed on product labeling	X: Avoid combination D: Consider therapy modification C: Monitor therapy	Medication on Beer’s List
Safe Use	Dosage: Within daily product labeling dosage Duration: Within product labeling duration of use	Interaction not listed on Beer’s List Condition not listed on product labeling	B: No action needed A: No known interaction	Medication not on Beer’s List

^aAmerican Geriatrics Society 2019 Updated AGS Beers Criteria.

^bLexi-Comp I. (Lexi-Drugs^®)

To assess if the OTC medication is associated with “safe use” or “misuse,” a review committee comprised of two geriatric pharmacotherapists and one OTC medication expert will independently adjudicate the primary outcome. A review committee conducted the same evaluation in the previous iteration of the study.^36,38 The review committee will be provided with the name of the OTC medication that participants select, a description of how the participant intends to use the OTC medication from the older adult interviews, and participant health information, including demographics, and self-reported medication information and health conditions. The review committee will be blinded to the participant’s status as a test or control site participant to help guard against information bias.

Mixed-effects logistic regression models will be used to assess the impact of the intervention on safe OTC medication usage. The intervention indicator variable and pharmacy type will be included as a fixed effect, while pharmacy will be included as a random effect. The primary comparison of interest is the score test for the intervention indicator variable. Potential heterogeneity of intervention effects across pharmacy will be assessed using appropriate interaction terms, and the influence of race and Area Deprivation Index rankings will be specifically evaluated. Similar models will be used to assess the effects of the intervention on other OTC medication usage outcomes, both overall misuse and specific types of misuse.

OTC Encounter Form data analysis:

All comparisons involving the OTC Encounter Form items will evaluate for differences between the test sites and control sites. Comparisons will include the type of patients, number and types of recommendations made, drug recommendations, and non-drug recommendations (e.g., behavioral modifications or referral to physician), and length of the encounter.

All items from the form are categorical. Items for which a single response is required range from dichotomous to five categories and will be analyzed as a distinct variable. However, multi-categorical items, in which more than one response can be selected, will require individual analysis of each response category. Given the categorical nature of the items, non-parametric statistics will be used for analyses depending on the frequency distributions.

Sample Size Calculation

Sample size calculations are based on the primary outcome for Aim 2, OTC medication misuse, an OTC not contraindicated by the participant’s health conditions and used consistently with the product labeling. Our prior studies found that safe OTC medication usage by 5–26% of our study population. Following the implementation of the Senior Section, safe OTC medication usage rose to 53%. Based on the results from our previous studies, the study team conservatively assumes that the proportion of older adults who misuse OTC medications will decrease from 74% to 60%. Using the score test for the intervention indicator variable from a two-level (subjects nested within pharmacies) logistic regression model, a sample size of 12 older adults at traditional pharmacies or 18 older adults at remote dispensing sites with a total of 288 older adults per data collection phase, will provide 80% power to detect the hypothesized difference using a two-sided 5% level test.

Aim 3: Sustainment Phase

The goal of Aim 3 is for the integrated healthcare system to implement the finalized version of the Senior Section in all of their pharmacies and remote dispensing sites, without research team support. This approach is designed to create a foundation for sustaining the Senior Section after research team support is withdrawn. To demonstrate the sustainment of the Senior Section, the study team will evaluate (1) the fidelity of delivery, (2) the long-term effectiveness of the Senior Section intervention, and (3) pharmacy staff perceptions of sustainability.

Fidelity of Delivery of the Senior Section

The Senior Section is a physical redesign of the pharmacy. If key elements of the Senior Section intervention are omitted or lost over time (e.g., if signage or tools are removed or the curated list of OTC medications is revised), the expected intervention effect may decrease over time, leading to “program drift.”⁵⁷ To monitor for program drift, pharmacy leadership will take digital pictures of the Senior Section approximately 3 and 6 months after implementation during their regular visits to meet with pharmacy staff. The pictures will be analyzed to determine whether pharmacy staff are, on their own, adapting the delivery of the Senior Section. Identification of any individual adaptation may perpetuate on the need for pharmacy leadership to provide additional support to mitigate program drift.

Long-Term Effectiveness of the Senior Section

Long-term effectiveness will be evaluated by comparing the baseline older adult OTC medication selection/use and pharmacist-patient encounters collected at the six control pharmacies and four control remote dispensing centers to new data collected at these same sites 6 months after implementation. These data, and their related analysis methods, are described below.

Older adult OTC medication use data:

UW researchers will recruit a new cohort of 12 older adults at traditional pharmacies and 18 older adults at remote dispensing sites at each of the 10 control sites from the randomized controlled trial, 6 months after the healthcare system implements the finalized Senior Section. Data from this new sample will be analyzed using a pre- (control)/post- (intervention) test to determine the long-term effectiveness of the finalized Senior Section to decrease OTC medication misuse in older adults. Aim 2 outcomes represent the data for this analysis.

OTC Encounter Form data:

Pharmacy staff at the 10 control sites (both pharmacies and remote dispensing centers) will collect the same OTC Form outcomes as in Aim 2, approximately 6 months after Senior Section implementation. Again, analysis of these data will be a pre- (control)/post- (intervention) test.

Longitudinal sales data of risky OTC medications:

Every OTC medication has a unique 10-digit National Drug Code denoting the manufacturer, drug product (including strength, dosage form, and formulation), and package size. It is our expectation that the curated list of OTC medications that will be included in the Senior Section will be a segregated list from the OTC medications that are merchandized throughout all of the pharmacies or remote dispensing centers.

Interrupted time series analysis (ITSA) will be conducted to determine the effect of the Senior Section on the sales of OTC medications. ITSA was chosen for its strength as a repeated measures design – generally considered the “gold standard” quasi-experimental design in causal inference.⁵⁸ This design is robust because it will evaluate Senior Section effects on OTC medication sales each week in each pharmacy setting, while controlling for trends in the outcome and threats to internal validity,⁵⁹ and is also unobtrusive, which is a necessary component of a sustainment phase. Efforts will be made by the project team to query pharmacy staff to ensure that trend changes are not the result of concomitant occurrences other than Senior Section implementation.

Using this approach, sales data trends for the six months before the finalized Senior Section is implemented in all pharmacy sites will be compared to trends six months post implementation, equalizing the times when each pharmacy site implements the Senior Section. The focus of our analysis will be on every product containing the four active ingredients that represent the most significant ADE burden: ibuprofen, aspirin, acetaminophen, and diphenhydramine.

Pharmacy Staff Perceptions of Sustainability

The study team will administer a practical and relevant survey that measures specific intervention-related characteristics that will focus on the “fit” between the Senior Section and the pharmacy work system.⁵⁷ Relevant items from current sustainability surveys have been adapted into a brief pragmatic survey to elicit information about pharmacy staff perceptions of organizational support available to sustain the Senior Section. All pharmacy staff will be asked to complete this online survey six months after the Senior Section is implemented.

Discussion

Previous efforts to improve OTC medication safety have primarily focused on improved labeling, with the FDA making significant improvement in labeling pertaining to prolonged therapy and high dosages. However, since only about 10% of consumers adhere to labeling recommendations,¹⁹ patient safety outcomes have been mixed. This situation is exacerbated by patient characteristics such as low health literacy rates and visual impairment, making it especially difficult for older adults to read and fully understand labels. Such a culmination of factors can impede informed decisions about OTC medication selection and use. ^36,65 In addition to these factors, efforts to improve OTC medication safety in older adults have been traditionally limited to small studies that have not demonstrated generalizability or the potential for sustainment.

This project was designed to focus on achieving systems-based improvement in an underemphasized area of the medication use process. To this end, critical data will be generated to shed light on a potentially enormous, but invisible, patient safety concern – that of OTC medication misuse and resulting harms. The objective and methodology of this project was meant to build on our preliminary work, which was successful at providing an initial signal of effectiveness of our Senior Section intervention and was guided by both theory and stakeholder engagement.

The current evaluation of the Senior Section significantly expands the approach of our past research, utilizing adaptation, effectiveness, and sustainment phases to quantify intervention efficacy, along with employing a method to measure pharmacy staff perceptions about the prospects of its sustainability. In addition, this project will create a resource designed specifically to encourage and guide the Senior Section’s broader implementation throughout the healthcare system (called an Implementation Package) during the project’s final phase. It is planned that the Implementation Package will provide justification for adopting the Senior Section, a description of the intervention features that contribute to its impact, and online training, quality improvement and communication tools. These activities, as well as virtually all aspects of this project, will be informed by the valuable feedback elicited during focus groups of pharmacy staff and older adults -- the principal stakeholders of the Senior Section – since a principal objective of this intervention is to promote effective interactions between pharmacy staff and older adult patients around OTC medication safety issues.

Additional, and importantly, progress over prior Senior Section research relates to the use of Tobii Pro Eyeglass Tracking. This new method was adopted specifically to achieve the dual and concomitant purposes of better avoiding researcher bias and overseeing and detailing the OTC medication selection process. There are also, however, substantial gains in collecting data relating to medication selection, which represent an important secondary advantage from the Tobii Eyeglasses. This advanced technological resource, which will be applied to a real-world environment (rather than in a laboratory setting), will collect accurate data related to the frequency, duration, and order of participants’ gazes. Such characteristics of these naturalistic viewing behaviors will be collected not only for the OTC products that are considered and selected, but also for the intervention features (e.g., magnifying glass or signage) that are viewed or ignored. Aggregate eye movement data will allow determination of both product and intervention factors that influence OTC decision-making, offering a strong quantitative foundation for conclusions about Senior Section implementation.

Project Innovations

The ability to partner with the integrated Midwest healthcare system for the duration of this project is a considerable strength and represents a significant innovation for two reasons. First, as a health system, the pharmacies have access to the electronic health records (EHR). Although most community pharmacies that are not affiliated with health systems currently do not have access to EHR patient-level data, health systems have been early adopters for providing EHR access to their retail community pharmacies. Such access allows pharmacists to review a patient’s conditions, laboratory values, and other pertinent healthcare information prior to recommending an OTC medication, while also providing them an avenue to document and share OTC usage with prescribers. Having ready access to data regarding potential contraindications increases the likelihood of making proper OTC recommendations. Second, this healthcare system has a well-developed system of remote dispensing sites, which provides an untapped opportunity to explore how pharmacists can triage, make recommendations, and educate patients about self-care and OTC medication needs in physically distant locations. With the prospect that tele-health will be a prominent treatment mechanism for the foreseeable future, due to COVID-19, our efforts to evaluate both the role of and relational characteristics within remote dispensing pharmacy services is more important than ever.

In addition, the distinctions between our proposed research and prior OTC safety interventions are highlighted by the following project innovations:

• The Senior Section intervention is grounded in human factors engineering theories and methods. Previous interventions for OTC medications fail to address the complexity of the OTC selection process within the larger work system and have had limited benefit. Decreasing OTC medication misuse will be achieved by targeting systems barriers that were identified during through stakeholder focus group feedback. These same human factors theories and methods will be tailored to the Senior Section for this project to adapt for diverse pharmacies and patient demographics, while supporting effectiveness and sustainment.

• One significant barrier to testing an OTC medication intervention has been the lack of access to retail-focused pharmacy organizations. As a result, there has been almost no research conducted in the places where many older adults purchase OTC medications. However, this study has the strong support of the healthcare system, one of the 10 largest not-for-profit integrated health systems in the U.S. with more than 60 retail pharmacies and remote dispensing centers, at both the organization level and the pharmacy staff (pharmacists or technicians) level. They are committed to providing resources to support the implementation and sustainability of the intervention.

• Most studies to improve medication safety have been provider-centric. ^66,67 Our study engages patients as partners in reducing medication misuse. The need for patient engagement to improve medication use quality and safety is identified as critical by the National Academy of Medicine (previously Institute of Medicine) and the National Patient Safety Foundation. ^68,69

• Eye tracking, an innovative data collection method used in market research to understand human behavior in real-world settings, has not been applied to understanding medication safety in community pharmacies.

• This study will be one of the first to use an implementation framework to rigorously evaluate the implementation of a generalizable and tailored medication safety intervention in community pharmacies.

• Successful project completion is ensured through a trans-disciplinary team that leverages a resource-rich environment and strong institutional support. Cumulatively, the research team and advisors comprise an expertise in geriatrics and tele-health, medication safety, community pharmacy, human-factors engineering, statistics and mixed methods, survey research, cultural adaptation, and dissemination and implementation science.

Project Limitations

Despite this project’s many innovative characteristics, there are a number of design limitations that require consideration. First, even at the conclusion of this study, the Senior Section intervention cannot be considered generalizable to all settings in which OTC medications can be purchased. The study team acknowledges that there are differences between pharmacies within this healthcare system and other pharmacy organizations, such as national chains like Walgreens or mass merchandise stores like Walmart. However, when proven to be effective and sustainable for this healthcare system, there is greater potential for the intervention to be implemented in most pharmacies (national chain, independent, mass merchandise, and grocery stores). Also, when outcome success is documented in relation to remote dispensing sites, such findings may begin generating insights into how pharmacists can successfully interact with older adults in retail establishments that currently do not have an on-site pharmacist (e.g., convenience stores or gas stations). Second, by asking older adults to opt-in to participating in the study, findings from such a self-selected sample (even if that sample represents diverse demographic characteristics) may not be generalizable to the entire population. Third, recruiting African American older adults may be challenging, especially from outside of large urban areas. To address this possibility, a research team member who has significant experience recruiting and engaging African American stakeholders will help us identify strategies to effectively solicit and retain older adults in this demographic. Finally, remote dispensing sites represent a pool of pharmacists who interact with technicians and patients only through tele-health technology when off-site at a pharmacy. As a consequence, some pharmacists could end up providing OTC recommendations for patients at both test sites and control sites, which can potentially influence those recommendations if the site’s status is known. This possible confound should be largely mitigated, however, since pharmacists using the tele-health technology will have to be actively queried by the patient, and that interaction will be localized to a video conference screen. Given this format of operation, the pharmacist should remain blinded to whether they are interacting with a patient at a test or control site (unless the older adult mentions the Senior Section). Further, the study team plans to elicit feedback to address this concern during our pharmacist stakeholder meetings.

Conclusions

A deliberate choice was made to partner with this integrated Midwest healthcare system for this project for three primary reasons: (1) There is a trusting relationship with both their leadership and their pharmacy staff, (2) their organizational structure and value mission are stable, and (3) their pharmacies serve a diverse patient population including underserved populations, and patients living in both urban and rural areas. Embarking on a three-year effectiveness-implementation study in collaboration with this healthcare system has both a great likelihood of success and broad generalizability to patient populations with the greatest need for OTC medication safety. Completion of this study will provide evidence that potential risks of OTC medications in older adults can be diminished and will provide a road map for pharmacy organizations to tailor and adopt the Senior Section in their own settings. Ultimately, the study team believes that the successful completion of this study will ultimately reduce ADEs associated with inappropriate OTC medication use in older adults.

The Senior Section intervention, when proven effective and feasible, will lead to new permanent structures and processes. At the pharmacy level, a system redesign will immediately change the quality and availability of information for older adults as they approach the OTC aisles. The study team anticipates that this will lead to greater awareness of risk so that older adults can more easily determine if that risk pertains to their own health situation and select safer OTC medications with confidence. As the results are disseminated at the national level, there may be growing recognition by both health care professionals and patient groups of how crucial the management of OTC medications in concert with prescription medications is to prevent ADEs.

Abbreviations

OTC

Over-the-counter

RCT

Randomized controlled trial

Availability of data and materials

Not applicable