Phase 1: recruitment of pharmacy sites and patient sampling.

All 34 community pharmacies from a multiple-pharmacy chain in one geographical area of South-East Wales were invited to participate. Five pharmacies consented to take part, with one participating pharmacist in each site. Recruitment from a single pharmacy chain ensured uniformity in standard operating procedures for reordering repeat prescriptions, providing consistency in capturing adherence data. Prescription dispensing ranged from 6,000–20,000 items per month and all participating pharmacies also provided a broad range of additional services. The patient recruitment target was 100 (20 from each study site). As this was a pilot study, a power calculation was not deemed necessary [33], therefore sample size was informed by other feasibility studies for pharmacy-based medication-related interventions [34–36].

Eligible patients were over 18 years old, with an existing diagnosis of hypertension, prescribed at least one antihypertensive medication for a minimum of three months, and using one of the study pharmacies to pick up their prescriptions. Individuals were not eligible if they had a medication compliance aid prepared by the pharmacy or a visual impairment that prevented them from viewing a computer or electronic tablet screen.

Hypertensive patients were recruited between June and August 2021 by the participating pharmacy either by: (1) spontaneous recruitment, through identification of suitable patients when the individual was collecting a prescription or (2) direct recruitment, where eligible individuals were identified through pharmacy records and contacted directly, via telephone, mailshot or using an informational leaflet attached to their medication bag.

Phase 2: recruitment of pharmacists and patients.

All pharmacists employed by the pharmacy chain in Phase 1 were invited to participate via email. Recruitment from the same chain provided consistency between phases. The recruitment target of five participating pharmacies was deemed appropriate, to enable a variety of locations and business sizes, following guidelines for qualitative aspects of feasibility testing [37]. Patient recruitment conducted between November 2021 and January 2022 followed the same approach as Phase 1.

Retention was determined by the number of participants providing data at each timepoint.

Illness perceptions and medication beliefs.

Outcome measures were recorded at baseline (T0), immediately post-intervention (T1) and at three-month follow-up (T3), for illness beliefs, medication adherence and BP, to investigate the preliminary direction of effect and further inform the design of a future evaluative trial. The appropriateness of outcome measures was assessed through the percentage of self-report questionnaires with missing data. Illness and medication beliefs were captured using the Brief Illness Perception Questionnaire (B-IPQ) [38] modified for hypertension and the Beliefs about Medicines Questionnaire-Specific Subscale (BMQ-S) [39] adapted for anti-hypertensive medication.

Medication adherence.

The implementation phase of medication adherence was studied, i.e., the extent to which the patient’s medication-taking corresponds to the agreed dosing regimen [40], using both self-report questionnaires and pharmacy data. The MARS-5 questionnaire [41] and a three-item measure of recent adherence capturing medication taken over the previous two weeks were used. The date of antihypertensive medication dispensing and the date the prescription was collected from the pharmacy, where data were available, were used to calculate the Medication Possession Ratio (MPR) for the three months pre-intervention and post-intervention. MPR was calculated using the formula: MPR = number of days’ supply in the refill interval/number of days in the refill interval.

Blood pressure.

BP was measured three times in the seated position, with the average of the second and third reading used for analysis, according to BIHS guidelines [42]. Measurements were taken at baseline (T0) and three-month follow-up (T3) using a semi-automated oscillometric blood pressure device (Omron M3™). Depending on patient preference, a home BP reading taken within one week of the consultation could also be used in preference to an in-consultation measurement. Demographic information including age, gender, education, number of antihypertensive medications and number of other long-term medications prescribed was recorded.

Community pharmacy (CP) and patient recruitment.

The recruitment target for CP and patient participants was met, with all pharmacies recruiting at least one patient per week throughout the study period. One of the pharmacies chose to contact eligible patients via mail as part of their direct recruitment. Recruitment letters were sent to the first 65 eligible participants identified from the Pharmacy Patient Medication Record, from which four participants responded and were recruited to the study. It is not known how many individuals were invited spontaneously, or how many of the informational leaflets were given to patients, therefore it is not possible to calculate a response rate for these recruitment methods. In total, 88 individuals expressed an interest in taking part and 69 (78%) provided their informed consent. Reasons for non-participation were mainly related to availability to attend an appointment within the weekday opening hours of the pharmacy (Monday to Friday between 9am to 6 or 7 pm). Recruitment varied from eight patients in one pharmacy to 17 (median = 14).

Patient participants’ age ranged from 49 to 84 years (mean 68.9, SD = 6.85). Thirty-eight (55.1%) were male, and 68/69 (98.6%) were white, with one participant reporting mixed ethnicity. Almost half (n = 33; 47.8%) were educated to the end of compulsory education (16-years old in the UK). The mean number of years since their hypertension diagnosis was 14.67 (range 0.5–45 years; SD = 9.22), with most prescribed one or two antihypertensive medications (range 1–5; median = 2). The majority of patients (64; 92.8%) were also prescribed medications for other long-term conditions (range 0–11; median = 3). All pharmacies were located in geographical areas ranked in the most deprived 60% of areas in Wales, i.e., within the lowest three deprivation quintiles [45]. All patients attending consultation one provided both baseline (T0) and post-intervention (T1) data. Of these patients, 45 (65%) provided adherence self-report data two-weeks post-intervention (T2). In total 78% provided three-month follow-up (T3) data; fifty-one (74%) visited the pharmacy for this and three (4%) provided data remotely, meaning BP data was not available for these (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Patient participant flow through Phase 1.

https://doi.org/10.1371/journal.pone.0339871.g003

Pharmacist recruitment.

Six pharmacist participants engaged with Phase 2 of the research across five pharmacy branches. Three pharmacists from Phase 1 declined to participate in Phase 2, and a further four were recruited. Since two of the pharmacists worked in the same branch, the intervention was delivered across five branches in Phase 2. Reasons provided for those three branches that did not continue to Phase 2 related to workload (e.g., prioritising delivery of remunerated services) and lack of staff continuity (e.g., pharmacists leaving the pharmacy).

Patient recruitment.

Three patients were recruited by two pharmacists in two different branches between November 2021 and March 2022. The remaining four pharmacists did not recruit any participants; therefore, the recruitment target was not met. Self-report adherence data at two-weeks post-intervention were returned for two patients, and no data were returned for the three-month follow-up consultation.

Theme 1: affective attitude.

Patients described being interested in the intervention and those with maximal adherence were reassured that their medication-taking had produced the desired effect. Pharmacists were enthusiastic towards the intervention, with those in Phase 1 finding recruitment straightforward and good acceptance to participate. While Phase 2 pharmacists were keen to deliver Hi-BP consultations, there were frustrations around time-sensitive and remunerated tasks taking priority.

Theme 2: intervention coherence.

Patients felt Hi-BP helped them make sense of their hypertension citing it as particularly helpful for newly diagnosed patients or those with a longstanding diagnosis to maintain the perceived importance of treatment. Pharmacists quickly understood the aim and need for Hi-BP, providing an easy ‘take-home message’ for patients, illustrating that hypertension was a condition to take seriously but to reassure them that it could be effectively managed.

Theme 3: perceived effectiveness.

Patients felt that Hi-BP strengthened their understanding of hypertension and provided a clear rationale for antihypertensive treatment, and reassurance regarding the effectiveness of the medication. They also believed that this visual prompt would encourage other people to make positive changes to address their hypertension. Pharmacists could see the potential for intervention effectiveness, particularly with newly diagnosed individuals, older people, or those taking multiple antihypertensives. It was suggested that the visuals would provide an appealing, novel way of re-engaging patients’ interest in their condition.

Theme 4: fit within community pharmacy.

Patients believed that the CP was a good place to host the intervention since the healthcare setting provided credibility and good access. The attributes of the person delivering the Hi-BP consultation were considered key to its success. Patient opinion was divided on whether a healthcare background was necessary, with some suggesting that an in-depth understanding of the intervention and effective communication skills were sufficient. It was suggested that suitably trained healthcare staff could deliver the intervention.

Pharmacists saw medicines adherence support as a core part of their responsibilities and highlighted the evolving role of community pharmacists in the UK towards being more clinically focused. However, they described taking an ad-hoc approach to supporting patients with medicines management since the decommissioning of a remunerated service in 2020 in Wales, which led to a perception of reduced emphasis for this type of activity in a CP-setting. Some pharmacists suggested that chronic medicine management was more suited to General Practice (GP)-based pharmacists than CP.

Theme 5: Self-efficacy: can patients access the intervention and pharmacists deliver it?.

Patients felt confident that the intervention setting and content of the intervention itself were appropriate to achieve the intended goal. Accessibility was key, in terms of ease of access to the CP setting, which was considered to be more accessible than a GP surgery, highlighting the importance of feeling relaxed in a familiar setting during the consultation. Patients believed that the utility of visuals and understandable language was a strength of the intervention content which would also be suitable for those with lower educational levels. The in-person facilitated conversation was welcomed over a self-directed digital application or remote consultation.

Pharmacists also thought that the CP setting was a convenient location to deliver Hi-BP as CP is now seen as a first port of call for many patients, with increasing volume of people approaching them for advice. Pharmacists also believed the content was accessible for patients, providing agency to tailor the conversation to the patient’s needs. Participating pharmacists felt confident that their understanding of hypertension was sufficient to facilitate the intervention, with some choosing to refresh their knowledge of clinical guidelines beforehand.

Theme 6: barriers and opportunity costs: what has to be given up to deliver Hi-BP?.

Patients did not offer any direct barriers or opportunity costs when asked. Whilst they acknowledged that the consultation was longer than previously experienced, they welcomed the opportunity to discuss and understand their hypertension. In contrast, lack of time and competing priorities were mentioned by all pharmacists as a potential barrier to delivering the intervention. Whilst they described wanting to provide new pharmacy services such as this, the daily balance between this and prescription dispensing made it a challenge, and priority often given to remunerated activities and managing acute conditions. Their workload was perceived to have increased in recent years, with changes to how and where patients access primary care being hastened by the COVID-19 pandemic. Pharmacists noted that the nature of the skill mix in the pharmacy team also impacted their workload, making it difficult to delegate tasks in order for them to engage with new patient-facing services such as this. Two pharmacists who had consented to take part in Phase 2 subsequently realised these challenges which led to no attempt being to recruit patients. For indicative quotes, see Supplementary Material S2 File.

In summary, Phase 1 demonstrates the acceptability of Hi-BP as a researcher-facilitated intervention, however, Phase 2 highlighted issues at the time with the feasibility of implementation, as a pharmacist-facilitated intervention, in this setting. Table 1 summarises the feasibility and acceptability indicators of Hi-BP and its delivery in a community pharmacy-based setting across Phases 1 and 2 of the study.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Feasibility and acceptability indicators for Hi-BP delivered in a community pharmacy setting.

https://doi.org/10.1371/journal.pone.0339871.t001

Illness and medication beliefs outcomes.

There was a statistically significant difference in scores across the three timepoints for the total B-IPQ score (X²(2)=10.83, n = 54, p = 0.004). Median scores decreased between baseline (32.50, interquartile range [IQR] 24.75–39.00) and T1 (28.50, IQR 24.00–35.50), increasing slightly at T3 (30.00, IQR 25.25–35.00), indicating that participants had a less threatening view of their hypertension post-intervention. There were also significant positive differences for three of the individual B-IPQ constructs: personal control (X²(2)= 6.16, n = 54, p = 0.046), treatment control (X² (2)=9.04, n = 54, p = 0.011), and illness coherence (X² (2)=32.35, n = 54, p=<0.001), suggesting that participants had greater illness coherence and a more positive perception of personal control over their hypertension and the utility of their treatment for controlling their hypertension post-intervention.

There was a significant positive difference in BMQ necessity subscale scores across the three timepoints (X²(2)=11.71, n = 54, p = 0.003). BMQ concern subscale scores also showed significant differences over time (X²(2)=11.71, n = 54, p = 0.003), with concern scores lowering immediately post-intervention and increasing slghtly at three-month follow-up. The necessity-concerns differential (NCD) was calculated over the three timepoints, showing a statistically significant increase (X²(2)=9.95, n = 54, p = 0.007). At baseline, 14.5% of participants (10/69) had an NCD of less than zero, indicating that they had greater concerns about their antihypertensives than belief in the necessity of the treatment. At T1 this percentage had dropped to just 1.4%, with only one participant scoring less than zero. At T3 no one had an NCD less than zero, indicating that by three-month follow-up the perceived necessity for antihypertensives was greater than the concerns about taking the medication for all participants.

Adherence outcomes.

There was no statistically significant change in MARS-5 scores across the three timepoints (X² (2)= 2.4, n = 45, p = 0.299). The median score at all three timepoints was 25 (out of a maximum of 25), demonstrating that self-reported adherence was consistently high, however, the range of scores indicates there was some improvement over time (baseline Md 25, range 9−25, IQR 24−25; T2 Md 25, range 18−25, IQR 24−25; T3 Md 25, range 20−25, IQR 25−25). A similar pattern was seen with the recent adherence scores, with no statistically significant difference reported across the three timepoints (X²(2)=5.89, n = 45, p = 0.053). As with the MARS-5 scores, the range of scores indicates there was a move towards greater adherence over time, with 60% (27/45) reporting 100% adherence at baseline, 73.3% (33/45) 100% adherence two weeks post-intervention and 80% (36/45) at three-month follow-up. Sixty-eight participants had valid data for medication dispensing for the three months pre-intervention and post-intervention. No statistically significant difference was found pre- and post-intervention (Z = −0.76, n = 68, p = 0.455). Medication collection data also showed no significant change in median scores (Z = −1.10, n = 45, p = 0.271).

Blood pressure measurements.

There was a statistically significant change in systolic blood pressure (SBP) between baseline and three-month follow-up (Z = −3.91, n = 51, p=<0.001). The median SBP reduced from 148 mmHg (IQR 135 mmHg to 160 mmHg) at baseline, to 133 mmHg (IQR 126 mmHg to 148 mmHg) at three-month follow-up. A similar result was observed for diastolic blood pressure (DBP), with a statistically significant change between timepoints (Z = −2.98, n = 51, p = 0.030), with the median DBP 82 mmHg at baseline (IQR 77 mmHg to 92 mmHg), reducing to 79 mmHg at three-month follow-up (IQR 72 mmHg to 87 mmHg).

For further detail, see Supplementary Material S1 File.