The impact of a person-centred community pharmacy mental health medication support service on consumer outcomes

Abstract

Background: Mental illness is a worldwide health priority. As medication is commonly used to treat mental illness, community pharmacy staff is well placed to assist consumers.

Aim: To evaluate the effectiveness of a multifaceted, community pharmacy medication support service for mental health consumers.

Method: Pharmacists and pharmacy support staff in three Australian states were trained to deliver a flexible, goal-oriented medication support service for adults with mental illness over 3–6 months. Consumer-related outcome measures included perceptions of illness and health-related quality of life, medication beliefs, treatment satisfaction and medication adherence.

Results: Fifty-five of 100 trained pharmacies completed the intervention with 295 of the 418 recruited consumers (70.6% completion rate); 51.2% of consumers received two or more follow-ups. Significant improvements were reported by consumers for overall perceptions of illness (p < 0.001), the mental health domain of quality of life (p < 0.001), concerns about medication (p = 0.001) and global satisfaction with medication (p < 0.001). Consumers also reported an increase in medication adherence (p = 0.005).

Conclusions: A community pharmacy mental health medication support service that is goal-oriented, flexible and individualised, improved consumer outcomes across various measures. While further research into the cost-effectiveness and sustainability of such a service is warranted, this intervention could easily be adapted to other contexts.

Keywords:

• community pharmacy
• mental health
• adherence
• intervention
• Australia
• patient-centred care

Introduction

Mental illness continues to be a topic of global discourse and research (Vigo et al., 2016; Whiteford et al., 2013) with increasing rates of antidepressant and antipsychotic prescribing. Alongside this is the ongoing focus to improve our understanding of consumer adherence to these medications (García et al., 2016; Woodward et al., 2016), or the effectiveness of interventions in this area (Al-Jumah & Qureshi, 2012; Chong et al., 2011; Thompson & McCabe, 2012). Mental health consumers have underscored the importance of health professionals assessing the appropriateness of medication before being blamed for non-adherence (Happell et al., 2004). Thus, it is not surprising that there is greater focus on consumer participation in healthcare decision-making (Chong et al., 2013).

The push towards greater consumer self-management and engagement within the primary care setting has opened up opportunities for expanding health professional roles or service delivery. This is especially evident within the pharmacy profession, where there is an increasing body of research on the delivery of professional pharmacy services (Bond, 2015). To encompass the breadth of such services, Moullin et al. (2013) explained that the action/s of a health practitioner need to optimize the process of care, with the aim to improve health outcomes and the value of healthcare. Furthermore, services should be provided in a person-centred way (McMillan et al., 2014; Mirzaei et al., 2013) by facilitating active consumer engagement to meet individuals’ healthcare needs.

Mental health consumers frequently access community pharmacies to obtain prescription medication; yet, there is limited reporting of mental health services within this setting (Hattingh et al., 2016b). Of the available studies involving pharmacists, many are from the USA (Adler et al., 2004; Bultman & Svarstad, 2002; Capoccia et al., 2004; Finley et al., 2003; Rickles et al., 2005) and focus on one or more of the following components: consumers starting medication, interventions comprising education and follow-ups, and medication adherence as the primary outcome measure (Hattingh et al., 2016b). The concern with targeting mental health consumers initiating therapy (Capoccia et al., 2004; Finley et al., 2003; Rickles et al., 2005) is the omission of a significant proportion of individuals, i.e. those who ceased taking their medication. During their Australian study, Crockett et al. (2006) had to extend their eligibility criteria from those starting a new medication for depression to include consumers already established on such therapy. The authors suggested that future research was necessary, involving enhanced mental health training and increased collaboration with other health professionals and services (Crockett et al., 2006). Reviewing the above study designs through a patient-centred lens (Morgan & Yoder, 2012), consumer follow-ups were less flexible (Adler et al., 2004; Capoccia et al., 2004; Finley et al., 2003; Rickles et al., 2005) and treatment goals focused on the mental illness rather than holistic patient-centred care. Given these gaps, research on the impact of a patient-centred, community pharmacy-based mental health support service was needed.

A three-stage exploration into how community pharmacies could support mental health consumers was funded by the Australian Government (Wheeler et al., 2015). A significant component of this research project was the development of a multi-faceted medication support service (the intervention) informed by consumers’ lived experiences of mental illness. The experiences, expectations and needs of mental health consumers within the context of community pharmacy were explored (Stage One) (Hattingh et al., 2015; Knox et al., 2015; Mey et al., 2013). Stage One findings informed development and evaluation of a comprehensive online mental health educational package for pharmacy staff (Stage Two) (Wheeler et al., 2013). Following this, a conceptual schema was undertaken which considered a range of factors, including professional pharmacy services (Table 1), organisational culture and evaluative tools to inform the design, implementation and evaluation of the intervention (Scahill et al., 2015). The aim of this study was to evaluate the effectiveness of the intervention in improving outcomes for consumers with common mental illnesses (Stage Three). Other results, including service implementation, goal setting and consumer expectations, will be reported separately (Hattingh et al., 2016a).

Table 1. Key professional services available in Australian community pharmacies (The 6CPA, 2015).

Medication adherence programmes
Clinical Interventions	Identifying, resolving and documenting medication-related issues in the pharmacy (e.g. when dispensing and supplying medication).
Compliance Aids (Dose Administration Aids)	The provision of a well-sealed, tamper-evident device that allows individual medicine doses to be organised according to prescriber instructions.
Staged Supply	The provision of Pharmaceutical Benefits Schedule (subsidised) medication at specified time-frames as directed by the prescriber, e.g. daily, weekly etc.
Medication management programmes
MedsCheck	Discussion of medications and health goals in the community pharmacy. There is also a similar service for Type 2 Diabetes (Diabetes MedsCheck). This service is currently capped by the Australian Government to 10 services per month, per pharmacy.
Home Medication Review (HMR)	A specially trained (accredited) pharmacist undertakes a home visit to identify and resolve any medication-related problems. A written report with medication recommendations is provided to the person’s general practitioner. Consumers must meet specific eligibility requirements to obtain the service.
Residential Medication Management Review	Similar to a HMR but in a residential-aged care facility.
Other specific programs or services
Return Unwanted Medication (RUM)	Consumers can return any medication no longer used for safe disposal.
Opioid Substitution Program	The provision of a range of opioid replacement therapy to consumers with opioid dependence.
Needle and Syringe Program	The provision of injecting equipment and their safe disposal.
Immunisations	Trained pharmacists can supply and inject certain vaccinations (dependent on state and territory legislation).

Methods

Study design

This experimental impact study evaluated the effectiveness of the intervention for mental health consumers, in community pharmacies in three Australian states: New South Wales, Queensland and Western Australia. States were chosen for their geographical diversity. The intervention was implemented between October 2013 and November 2014. Two assessment time-points were utilised: when each consumer participant joined (baseline, T1) and completed (T2) the intervention. Given the patient-centred and flexible nature of the intervention, T2 was completed at varying time-points by participants. To achieve 80% power, the projected sample size of completed consumers (n > 200) was considered to be of sufficient size to detect a moderate change (0.5) in paired results over time, if it existed (Cohen, 1988), e.g. in illness perceptions from T1 to T2. University ethics approval was obtained (PHM/13/11/HREC).

Initially, community pharmacies with staff that had completed the Stage Two online training (n = 229) were invited to participate in delivering the intervention; additional recruitment was also undertaken, e.g. conference promotion. Pharmacies were screened to ensure management approved participation, adequate staffing levels and private counselling areas for confidential discussions (e.g. consistent with delivery of certain professional pharmacy services in Australia such as MedsCheck; Table 1). A range of consumer recruitment materials, including postcards and information sheets, were provided to pharmacies, requesting each to recruit 10 mental health consumers. Additional information sheets and flyers advertising the intervention to local health professionals, e.g. general practitioners, social workers and services, e.g. mental health services, were provided to pharmacy staff. Financial reimbursement for the time needed to recruit and implement the intervention was offered to participating pharmacies for each consumer; between AU $50 and $100 per completed consumer depending on recruitment numbers. Pharmacies that recruited five or more consumers entered an additional prize draw (AU$1000 worth of registered training).

Consumers were eligible for the intervention if > 18 years old, lived in the community rather than an inpatient facility and were prescribed medication for a common mental illness (primarily anxiety or depression). Participants with other mental health conditions were not excluded if they had a medication-related problem; however, the General Practitioner (GP) was required to be the primary prescriber. Identification of consumers with medication-related problems was via pharmacy staff [e.g. when starting a new mental health medication, or if they needed medication management support such as the provision of a compliance aid or medication review (The Pharmacy Guild of Australia & Australian Government Department of Health, 2014b); Table 1], via a GP, carer or family member or self-identified via promotional activities. Pharmacy staff were provided with some examples of potential medication-related issues in the training, e.g. not regularly picking up repeat medication supply, thinking about stopping an antidepressant after 2 months because feeling better, etc. Written consent was obtained from each consumer after explaining the service and the project. The intervention was provided at no charge to consumers.

Study intervention and outcome measures

The primary focus of the intervention was to identify, prevent and manage medication-related problems and promote mental well-being; pharmacy staff worked with consumers and other healthcare team members to develop a tailored, goal-oriented support plan. A significant number of studies have incorporated goal-setting as part of the pharmacist intervention for various other chronic diseases, such as asthma (Emmerton et al., 2012) and diabetes (Krass et al., 2007). The development of this patient-centred intervention was also guided by two literature reviews (Hattingh et al., 2016b; Scahill et al., 2015), findings from Australian pharmacy trials (Feyer et al., 2010; Healthcare Management Advisors, 2010), and consultations with clinicians, other researchers and consumers and carers with a lived experience of mental illness. These consultations involved a Community Pharmacy Working Group comprising of key stakeholders, e.g. consumer and carer consultants, psychologist, etc., and a Project Reference Group of stakeholders with a special interest in mental health, e.g. GPs, pharmacists.

The Stage Three intervention (detailed in Table 2) involved a six-step process between 3 and 6 months’ duration, with steps 3 (Initial Health Review) and 6 (Final Health Review and Evaluation) the focus of this paper. Tools used to measure the impact of the intervention on consumer outcomes and medication management included assessment of illness perceptions, health-related quality of life, beliefs and satisfaction with medication and medication adherence (Table 2). When reviewing the tools to be included, the research team considered the objectives of the intervention, data collection burden, duplication of questions and the practicalities of using instruments that provided the necessary information while not being too time consuming to complete. Validated and reliable tools were used where possible. Consumer evaluation of the intervention was obtained using a 10 question (5-point Likert Scale) survey and three open-ended questions, e.g. please describe, in your own words, what you liked most about your experience of this intervention?

Table 2. Intervention steps.

Steps	Tools used, information collected or tasks undertaken
Step 1	Project information, consumer recruitment and consent.
Step 2 (My Health Status)	Demographics, details of support network, frequency of pharmacy visits.Project information was provided to the consumers’ healthcare team, e.g. general practitioner, in the form of a letter.
Step 3 (Initial Health Review)	Baseline data (T1): illness perceptions, quality of life, medication beliefs and satisfaction, medication adherence and community pharmacy expectations (using the Australian Community Pharmacy Service Charter [The Pharmacy Guild of Australia & Australian Government Department of Health, 2014a]).
	Brief Illness Perception Questionnaire (BIPQ) (Broadbent et al., 2006)	8 questions (11-point scale) assessing dimensions of illness perception. The overall score represents the degree to which illness is perceived as benign or threatening. A higher score reflects a more threatening view of illness.
	Short Form 12 Health Survey v1 (Ware et al., 1996)	12 questions in which physical and mental health composite scores can be computed using the scores of the 12 questions (range from 0 to 100). Zero indicates the lowest level of health.
	Beliefs about Medication Questionnaire (BMQ) (Horne et al., 1999)	Two 5-item subscales (5-point scale) assessing consumer beliefs about the necessity of taking medication to manage their illness, and their concerns about adverse outcomes. Higher scores indicate stronger beliefs.
	Treatment satisfaction Questionnaire for Medication (TSQM^©) (Atkinson et al., 2004)	14 statements (7-point scale) which form four sub-scales (Effectiveness, Side effects, Convenience and Global Satisfaction). Lower scores indicate low satisfaction.
	Morisky Medication Adherence Scale (MMAS-8^©) (Krousel-Wood et al., 2009; Morisky et al., 2008; Morisky & DiMatteo, 2011)	8 questions with yes/no answers. An overall score can be computed to assess high, medium or low adherence. Consumer participants responded to this question for all of their medications.
Step 4 (Medication Support Plan)	A partnership between pharmacist and consumer to identify challenges, set goals and associated strategies.
Step 5 (Ongoing Support and Maintenance)	Provision of medication management services to the consumer over a suggested period of 3–6 months (tailored to each individual in terms of frequency and mode of follow-up contact, e.g. face-to-face). Documentation of consumer progress and review of goal strategies.
Step 6 (Final Health Review & Evaluation)	Post-intervention data (T2): similar to Step 3 to allow for a comparison of responses.

Pharmacies received consumer data collection folders to record collaborative goal setting, follow-ups and consumer outcomes. Interventions were tailored to each consumer, with variable follow-up contacts. This allowed the consumer to connect with a health professional (i.e. the pharmacist) in-between GP visits, which, in Australia, generally depends on repeat prescription allowances for subsidised medication. For example, a consumer may receive a prescription for an antidepressant lasting up to 6 months, with repeat medication supply at 1-month intervals (Australian Government Department of Health, 2016). Pharmacy staff were initially trained during face-to-face workshops by one of three pairs of trainers (each pair consisting of a pharmacist and consumer or carer who may have had a lived mental health experience). The training aimed to provide pharmacy staff with the knowledge, skills and abilities to confidently engage consumers from the point of recruitment, to supporting them with medication management throughout the duration of (and beyond) the research. The workshop was designed around the medication support intervention, working through the background, objectives and benefits of the intervention, to identifying stages of readiness for change and progressing through working with consumers at each stage. Training processes used in the interactive training included PowerPoint^® presentations, discussions, activities, demonstrations, role-plays and case studies. Pharmacy staff were then supported throughout the intervention by one pair in a mentoring role (who were trained and undertook fortnightly debriefing sessions). Mentors utilised Kram’s (1983) four-phase model of mentoring to support pharmacy staff, for example with consumer recruitment, data collection and health professional collaboration. At the end of the intervention, the research team collected the completed data collection records. Participant data were de-identified at the point of data entry.

Data analysis

Data management and analysis were performed using SPSS v21 (IBM, Armonk, NY) and Stata 12.1 (StataCorp, College Station, TX) software. Descriptive qualitative analysis and statistics were used to report survey data, and changes were tested in questionnaire data between T1 and T2. Values were compared between participants who completed the intervention (completers) and non-completers. The overall score of the Brief Illness Perception Questionnaire (BIPQ) was calculated as per the method of Broadbent et al. (2002, 2006) and Short Form-12 Health Survey (SF-12v1) overall scores (Physical health and Mental health) were calculated using a Stata syntax based on US population means and SDs. The overall scores of the Beliefs about Medication Questionnaire (BMQ; Necessity score and Concerns score) were calculated as described in Neame & Hammond (2005). Responses to the Treatment Satisfaction Questionnaire for Medication (TSQM^©) were scored into four sub-scales (Effectiveness, Side effects, Convenience, Global) as per Atkinson et al. (2004). The overall adherence category (low/medium/high) was calculated as per the method of Morisky et al. (1986), where the question How often do you have difficulty remembering to take all your medicine? was re-coded as the question Do you ever have difficulty remembering… (no/rarely = 0; yes = 1). Hypothesis testing (null hypothesis: no difference in responses at two time-points) was completed using chi-squared and Fisher’s exact tests on categorical variables, and paired t-tests on ordinal (e.g. Likert-scale) and continuous variables. Multiplicity (problem of multiple comparisons) was considered when interpreting the results, and a lower p value (p < 0.01) was used to declare a statistically significant difference.

Results

Sample characteristics

Of the pharmacies approached to participate in Stage Three (from a pool of 229 pharmacies whose staff had participated in the online mental health education in Stage Two), 100 pharmacies were recruited with approval from pharmacy owners (n = 49 Queensland; n = 14 New South Wales; n = 37 Western Australia). Overall, 142 pharmacists and 21 support staff from the 100 pharmacies were trained to deliver the intervention; 55 pharmacies completed the intervention with one or more consumers, with 45 pharmacies considered to be non-completers. The majority of non-completer pharmacies withdrew from the study (n = 24), nine did not recruit anyone, and the remainder recruited but did not deliver or complete the intervention with consumer participants. Completer pharmacies tended to have more pharmacists present and fewer support staff than non-completers, and offered a wider range of professional services, e.g. opioid substitution (Table 1).

Overall, 418 of the 570 consumers that expressed interest in participation were recruited (73.3% response rate). Most lived in Western Australia (n = 258; 61.7%), followed by Queensland (n = 137; 32.8%) and northern New South Wales (n = 23; 5.5%). The majority were female and of Australian (Caucasian) background; fewer than half were in paid employment (Table 3). Of the 418 recruited consumers, 295 (70.6%) completed the intervention. The only significant difference in socio-demographic characteristics identified between completers and non-completers was that completers were significantly higher users of community pharmacy (p < 0.001).

Table 3. Consumer participant characteristics.

	Total, n = 418	Completers, n = 295 (%)	Non-completers, n = 123 (%)	p value
Gender	400	293	107
Female	246 (61.5)	178 (60.7)	68 (63.6)	0.610
Male	154 (38.5)	115 (39.3)	39 (36.4)
Age (years)	376	277	99
up to 25	13 (3.5)	7 (2.5)	6 (6.1)	0.060
26-40	91 (24.2)	60 (21.7)	31 (31.3)
41-65	204 (54.3)	159 (57.4)	45 (45.4)
66 or over	68 (18.1)	51 (18.4)	17 (17.2)
Language	394	289	105
English	379 (96.2)	276 (95.5)	103 (98.1)	0.372
Other	15 (3.8)	13 (4.5)	2 (1.9)
Cultural background	397	290	107
Australian (White/Caucasian)	356 (89.7)	258 (89.0)	98 (91.6)	0.577
Other	41 (10.3)	32 (11.0)	9 (8.4)
Employment status	392	286	106
Employed^a	168 (42.9)	117 (40.9)	51 (48.1)	0.419
Unemployed	48 (12.2)	41 (14.4)	7 (6.6)
Retired	67 (17.1)	51 (17.8)	16 (15.1)
Disability support pension	63 (16.1)	48 (16.8)	15 (14.2)
Other^b	46 (11.7)	29 (10.1)	17 (16.0)
Mental health prescriber	388	283	105
General practitioner	338 (87.1)	246 (86.9)	92 (87.6)	0.489
Psychiatrist	46 (11.9)	35 (12.4)	11 (10.5)
Other	4(1.0)	2 (0.7)	2 (1.9)
Have a carer^c	397	291	106
Yes	30 (7.6)	22 (7.6)	8 (7.5)	0.997
See another health worker	355	255	100
No	202 (56.9)	143 (56.1)	59 (59.0)
Psychologist	89 (25.1)	64 (25.1)	25 (25.0)	0.864
Social worker	13 (3.7)	9 (3.5)	4 (4.0)
Occupational therapist	9 (2.5)	6 (2.3)	3 (3.0)
Other	42 (11.8)	33 (12.9)	9 (9.0)
Community pharmacy visits	394	288	106
Daily	54 (13.7)	50 (17.4)	4 (3.8)	<0.001
Weekly	85 (21.6)	70 (24.3)	15 (14.1)
2–3 times/month	106 (26.9)	77 (26.7)	29 (27.4)
Monthly	123 (31.2)	75 (26.0)	48 (45.3)
Other (every few months or other)	26 (6.6)	16 (5.6)	10 (9.4)

^aIncludes full-time, casual or part-time paid employment; ^bIncludes student, volunteer, homemaker or other; ^cIncludes relative, friend, paid employee or other. p < 0.01 indicates significant difference.

Initial health review and consumer follow-up

Consumers described in-detail how much their mental illness impacted on their life, both physically, e.g. limited energy, pain and insomnia, and emotionally, e.g. stress. For some consumers, it was difficult to see “when” their particular situation would resolve. While consumers believed medication to be necessary, there were also concerns associated with long-term use with respect to dependence or side effects. With respect to goal setting, consumers predominantly sought support with symptom and medication management, e.g. medication information and assistance with adherence, or were lifestyle related, e.g. exercise, relaxation and smoking cessation. Medication adherence was self-reported as variable and influenced by a range of factors, such as dissatisfaction with treatment choice, the inconvenience associated with medication taking or symptom resolution.

The number of follow-ups recorded between individual consumers and pharmacy staff ranged between 0 and 23 per consumer. The majority of consumers (213/416; 51.2%) had two or more follow-up contacts (mean = 1.9; median = 2.0), which were mostly face-to-face (640/778; 82.3%), followed by phone calls (130/778; 16.7%). Other forms of communication were rarely used, e.g. email (8/778; 1.0%). On the basis of two follow-ups, an overall mean time of 110 minutes was spent working through the entire intervention; further information on the time involved for each step is outlined in Table 4. The intervention was delivered over a mean of 145.9 days (SD = 72.4, range 13–359 days) per consumer.

Table 4. Time spent working through steps of the medication support service.

Step number^a
Time (minutes)	2, n = 255	3, n = 268	4, n = 239	5, n = 236	6, n = 207	Total
Mean (SD)	11.7 (9.3)	23.8 (12.1)	21.1 (14.3)	14.3 (12.2)	24.7 (14.0)	95.6
Median	10	20	20	10	20	–
Range	1–60	4–75	1–90	1–75	1–120	–

^aStep 1 was recruitment and consent and a time was not recorded as it was often undertaken over more than one conversation/visit to allow for provision of study information and answering questions.

Consumer outcomes

Health-related outcomes (BIPQ and SF-12)

At T1, there was no difference for overall illness perceptions (p = 0.248) and the SF-12 physical health composite score (p = 0.724) between completers and non-completers. The only significant difference identified at T1 was that non-completers had a poorer overall SF-12 mental health score than completers (p = 0.034). The overall score at T2 (representing the degree to which illness is perceived as threatening or benign), reduced significantly from a mean of 44.8 to 38.8 (p < 0.001), indicating that consumers perceived their illness to be less threatening (Table 5).

Table 5. Outcomes measured at baseline (T1) and at the completion of each intervention (T2).

	T1 score mean (SD)	T2 score mean (SD)	p value
Health-related outcomes
Brief Illness Perception Questionnaire (BIPQ)
How much does your illness affect your life? (n = 289)	6.3	5.6	<0.001
How long do you think your illness will continue?^a (n = 288)	7.8	7.4	0.002
How much control do you have over your illness?^b (n = 287)	5.2	6.1	<0.001
How much do you think your treatment can help your illness?^b (n = 284)	7.2	7.5	0.008
How much do you experience symptoms from your illness? (n = 290)	6.4	5.7	<0.001
How concerned are you about your illness? (n = 287)	6.7	5.7	<0.001
How well do you feel you understand your illness?^b (n = 289)	7.0	7.7	<0.001
How much does your illness affect you emotionally? (n = 287)	7.0	5.9	<0.001
Total score (n = 270)	44.8 (12.2)	38.8 (12.9)	<0.001
Short Form 12 Health Survey v1 (SF-12)
SF-12 physical domain (n = 267)	39.2 (7.3)	39.0 (7.3)	0.423
SF-12 mental domain (n = 267)	43.3 (10.1)	46.1 (10.3)	<0.001
Medication-related outcomes
Beliefs about Medication Questionnaire (BMQ)^‡
My health at present depends on my medicines (n = 287)	4.1	4.0	0.074
Having to take medicines worries me (n = 284)	2.8	2.6	0.011
My life would be impossible without my medicines (n = 284)	3.8	3.7	0.242
Without my medicines I would be very ill (n = 287)	3.8	3.8	0.326
I sometimes worry about long-term effects of my medicines (n = 287)	3.3	3.2	0.068
My medicines are a mystery to me (n = 288)	2.4	2.1	0.001
My health in the future will depend on my medicines (n = 281)	3.8	3.7	0.269
My medicines disrupt my life (n = 286)	2.4	2.4	0.685
I sometimes worry about becoming too dependent on my medicines (n = 289)	3.0	2.8	0.002
My medicines protect me from becoming worse (n = 289)	4.1	4.1	0.903
Necessity sub-scale (n = 271)	19.7 (3.8)	19.4 (4.1)	=0.276
Concerns sub-scale (n = 275)	13.9 (4.6)	13.1 (4.2)	=0.001
Treatment Satisfaction Questionnaire for Medication (TSQM^©)^§
Effectiveness (n = 276)	74.8 (17.3)	78.8 (16.0)	<0.001
Side effects (n = 88)	55.6 (14.4)	57.1 (15.3)	0.681
Convenience (n = 285)	89.8 (14.1)	90.9 (12.9)	0.186
Global satisfaction (n = 286)	80.9 (17.0)	84.2 (16.5)	<0.001
Morisky Medication Adherence Scale (MMAS-8^©)	n (%)	n (%)
High adherence	18 (51.4)	17 (48.6)	=0.005
Medium adherence	106 (42.4)	144 (57.6)
Low adherence	158 (56.6)	121 (43.4)

^aRated from a very short time to forever; ^bRated in reverse so lower score indicates less threat;

^‡Higher scores indicate stronger beliefs, rated from 1 = strongly disagree to 7 = strongly agree, paired t-test; ^§Lower scores indicate lower levels of satisfaction; p < 0.01 indicates significant change.

n relates to the number of participants reporting at T1 and T2.

Overall, 54.3% of completers (159/293) rated their health (SF-12) as good/very good/excellent at T1, which increased to 62.8% at T2 (182/290). There was no change in the SF-12 physical health composite score between T1 and T2 (p = 0.423), yet significant improvement was shown in the mental health composite score (p < 0.001; Table 5). The four SF-12 questions related to mental health wellbeing (feeling calm and peaceful, having a lot of energy, feeling downhearted and blue, and physical and mental health interfering with social activities) all changed in a positive direction, with the largest improvement seen for the first question (T1 = 106/293; 36.2% and T2 = 156/294; 53.1%).

Medication-related outcomes (BMQ, TSQM^© and MMAS-8^©)

There were no differences between completers and non-completers for T1 scores relating to the following beliefs around medication: necessity (p = 0.074); concerns (p = 0.341); effectiveness (p = 0.125); side effects (p = 0.769) or convenience (p = 0.836). This result was also seen for overall medication satisfaction (p = 0.145) and medication adherence (p = 0.085).

Consumers were significantly less concerned about adverse medication consequences (p = 0.001; Table 5) after the intervention (T2). There was no change in consumer beliefs about the necessity of medication to manage their mental illness (p = 0.276); consumers held strong beliefs about the important role of medication at T1. In terms of medication satisfaction, a significant improvement was seen for medication effectiveness and global satisfaction at T2 (p < 0.001). Scores did not change for the other two TSQM^© sub-scales, side effects and convenience. Medication adherence improved for seven of the eight MMAS-8^© questions; there was reduced adherence in relation to the question asking about medication taking “the day before” (Figure 1). Overall, there was a reduction in the number of consumers with low adherence and a corresponding increase in those reporting medium adherence (p = 0.005) at T2.

Figure 1. Morisky Medication Adherence Scale (MMAS-8)^© (Krousel-Wood et al., 2009; Morisky et al., 2008; Morisky & DiMatteo, 2011). Use of the ©MMAS is protected by US copyright laws. Permission for use is required. A license agreement is available from: Donald E. Morisky, ScD, ScM, MSPH, Professor, Department of Community Health Sciences, UCLA School of Public Health, 650 Charles E. Young Drive South, Los Angeles, CA 90095-1772, [email protected].

Consumer evaluation of the intervention

The medication support service received positive consumer feedback. The majority of completers were satisfied with the pharmacy service (278/283; 98.2%), and reported that the service motivated them (236/283; 83.4%), gave them confidence to deal with their mental health (222/284; 78.2%) and improved their medication taking (176/282; 62.4%). Consumers described a number of service-specific aspects that they “liked most”, which primarily related to adherence support strategies and pharmacist follow-ups:

My medication regime is now simplified for me to take and manage scripts etc., so much easier to remember doses. (Female, Queensland)

Dedicated one-on-one time with the pharmacist where information sharing was of great benefit to help with my medication list and symptom control. (Female, Western Australia)

Discussion

This study evaluated the effectiveness of a community pharmacy medication support service for consumers with common mental illnesses, primarily depression or anxiety. Statistically significant improvements for a range of variables were identified, including perceptions of illness, mental-health related quality of life, medication satisfaction and adherence. Consumers valued the service, and were more motivated and confident to manage their mental illness. This research demonstrates that an individualised, flexible and goal-oriented medication support service can make a positive difference towards managing mental health in the community setting. This adds weight to the role of community pharmacy staff in delivering professional services for chronic conditions, including those which are enduring and complex, such as mental illness.

The strengths of this study include the large sample size and varied constituents (pharmacies, staff and consumers), and ongoing discussions with a Reference Group to enhance study validity and reliability. There were two differences identified between completer and non-completer consumers at T1: the latter group self-reported poorer mental health and fewer pharmacy visits. Why this particular group of participants did not complete the intervention is unknown and justifies further research given that they possibly had more to gain in terms of their mental wellbeing from participation. Findings were grounded in consumer and carer experiences; mental health consumer and carer consultants were involved throughout, from conception to analysis. As this study focused on consumers living in the community with common mental illnesses such as depression and anxiety and prescribed treatment at the primary care level, the conclusions drawn require confirmation with other samples before being generalised to consumers with other mental health conditions and/or different treatment arrangements. While an Australian-based study, the intervention itself is adaptable and translatable to other countries, particularly where community-based pharmacies provide professional services beyond medication supply. A possible selection bias was mental health consumers self-identifying for study participation and potentially providing socially desirable responses to pharmacists. The researchers did not routinely collect information on diagnosis and duration of illness, or co-morbid diagnoses and treatment as this was beyond the scope of the study. Medication adherence rates were not confirmed by prescription refill data, and an exploration of longer-term outcomes, including duration of effect, was beyond the remit of this study. Last, this study was designed as a feasibility study; a randomised controlled trial (RCT) to assess the effectiveness (including cost) of the intervention compared to usual care is needed.

The study findings must be considered in relation to other pharmacy intervention studies (Brook et al., 2003; Crockett et al., 2006; Rickles et al., 2005; Rubio-Valera et al., 2013). While studies differ in terms of design, intervention and sample size, the overarching goals are generally similar, i.e. to evaluate the impact of pharmacist-specific training on mental health consumer outcomes. The demonstrated improvement/s in health-related quality of life are aligned to those in Rubio-Valera’s RCT (2013). Other researchers consistently report non-significant or marginal improvements in depressive symptoms or psychological wellbeing for intervention versus control groups (Adler et al., 2004; Capoccia et al., 2004; Crockett et al., 2006; Finley et al., 2003; Rickles et al., 2005; Rubio-Valera et al., 2013). While our study was not an RCT, consumers reported feeling more settled and at peace, alongside significant improvements in their mental health. Given high rates of co-morbidities, the physical health of mental health consumers is important; while our intervention focused on the whole person, consumer goals did not always incorporate lifestyle changes. Pharmacists and other healthcare professionals need to ensure that physical functioning is discussed as an essential component of good mental health (Young et al., 2017). Last, significantly more positive attitudes were seen regarding medication use, which is consistent with Rickles et al. (2005) and Brook et al. (2003). This is a particularly important result, as it would be expected that a more positive attitude would be aligned with improved medication adherence.

This study strengthens suggestions that pharmacist interventions have a positive impact on adherence to antidepressant medication (Al-Jumah & Qureshi, 2012; Bell et al., 2005; Chong et al., 2011; Rubio-Valera et al., 2011). How this study differs from many of those included in these systematic reviews was the inclusion of consumers with conditions beyond depression (i.e. anxiety disorders), a large number of pharmacists across a variety of settings trained to improve their confidence and skills in working with mental health consumers, and involved a multi-faceted, flexible intervention within an Australian setting. Furthermore, the intervention was consumer tailored with respect to goals, support and follow-up frequency, in contrast to other studies (Brook et al., 2003; Rickles et al., 2005; Rubio-Valera et al., 2013). It was not possible to determine if specific symptoms of mental illness improved alongside increased medication adherence rates, a similar problem reported elsewhere (Al-Jumah & Qureshi, 2012; Rubio-Valera et al., 2011). This study did not prove an association between adherence and clinical outcomes, yet statistically significant improvements were seen for both variables. It is unknown why there was an increase in reports of missed medication the day prior to assessment at the end of the intervention. What is known is that medication adherence is influenced by a variety of factors. For example, consumers may be more susceptible to non-adherence because of related costs, or stop taking medication because they feel better. Ultimately, this study demonstrates that medication adherence is an ongoing battle for mental health consumers. While our consumers were more satisfied with their treatment, the intervention itself did not improve their perceptions of the side effect profile or dosing regimen of their medication. Ultimately, healthcare professionals need to discuss individual consumer’s medication management on a regular basis, including the impact of side effects, simplify dosing regimens and emphasise the importance of adherence even when feeling better. The need to have systematic, short, follow-up discussions with mental health consumers was underscored by French–Canadian pharmacists, acknowledging it as a shortcoming in their practice (Guillaumie et al., 2015).

There are practice, policy and research, implications arising from this study. This study serves as a useful model for pharmacy staff training, service development and delivery, and consumer-driven research. The intervention was based on a person-centred care approach: it was individualised, holistic, respectful and empowering (Morgan & Yoder, 2012), attributes that should apply to all healthcare consultations. Thus, the intervention has the potential to be adapted and utilised for a range of other chronic conditions. The benefits for consumers were noteworthy, providing evidence of the significant potential and value of community pharmacy as an accessible and safe healthcare space providing information and ongoing support. Aligning with other consumer satisfaction surveys (Naik Panvelkar et al., 2009), this pharmacy-based intervention was well received by mental health consumers. With the continuing focus by Governments worldwide and healthcare sectors locally to improve mental health services, community pharmacy staff can, with support and training, contribute towards improved outcomes for these consumers. A large-scale implementation study to explore and promote effective rollout, uptake and sustainability in community pharmacies is warranted (Blalock et al., 2013; Moullin et al., 2016). Such a study should include a methodologically sound assessment of the cost-effectiveness of the pharmacist intervention; poor or variable study designs has made this assessment problematic in recent times (Elliott et al., 2014).

Conclusions

Trained community pharmacy staff can effectively provide a medication support service to mental health consumers with depression or anxiety. This study not only demonstrated that a goal-oriented, flexible and individualised support service improves consumer outcomes, but confirmed that such a service is positively received. With the ongoing struggle towards improving mental healthcare and access to services, it would be remiss of the pharmacy profession and healthcare policy makers not to consider such positive outcomes.

Declaration of interest

This study was funded by the Department of Health, Australian Government as part of the Fifth Community Pharmacy Agreement Research and Development Programme managed by the Pharmacy Guild of Australia. The financial assistance provided must not be taken as endorsement of the contents of this paper.

Acknowledgements

We thank our study participants and assistance from the mentors, Project Advisory Panel and Reference Group. The authors acknowledge all other members of the Mental Health Project team: Andrew Davey, Bradley McConachie, Rhonda Knights, Amary Mey, Kathy Knox and Jasmina Fezjic.