Current perspectives on pharmacist home visits: do we keep reinventing the wheel?

The scope of clinical pharmacy services available in outpatient settings, including home care, continues to expand. This review sought to identify the evidence to support pharmacist provision of clinical pharmacy services in a home care setting. Seventy-five reports were identified in the literature that provided evaluation and description of clinical pharmacy home visit services available around the world. Based on results from randomized controlled trials, pharmacist home visit interventions can improve patient medication adherence and knowledge, but have little impact on health care resource utilization. Other literature reported benefits of a pharmacist home visit service such as patient satisfaction, improved medication appropriateness, increased persistence with warfarin therapy, and increased medication discrepancy resolution. Current perspectives to consider in establishing or evaluating clinical pharmacy services offered in a home care setting include: staff competency, ideal target patient population, staff safety, use of technology, collaborative relationships with other health care providers, activities performed during a home visit, and pharmacist autonomy.

Introduction

Over the past several decades, the scope of clinical pharmacy services has expanded both in terms of skills and areas in which services are offered. Traditionally, the availability of clinical pharmacy services has been in the purview of hospitals where increased clinical pharmacy services has been associated with reduced length of stay and mortality.1 Recognition of the value of the role of the pharmacist has resulted in expansion of clinical services into outpatient settings, including patient homes. For example, the Home Medicines Review (HMR) program that was established in Australia in 2001 provides funding for pharmacists to visit patients at home to assess their medication regimens.2 In Canada, provincial governments are compensating pharmacists for providing medication reviews (MRs) for non-hospitalized patients3 and also authorizing pharmacists to prescribe.4

While there is evidence to suggest that pharmacist prescribing activities can improve patient outcomes in outpatient settings,5–7 the evidence to support the benefit of MRs in outpatient settings is equivocal. Holland et al conducted a systematic review and meta-analysis to evaluate the impact of pharmacist-led MR in older adults and reported that there was no effect on reducing mortality or hospital admissions, but that the intervention may reduce the number of prescribed drugs and improve drug knowledge and adherence.8 More recently, an evaluation of the MR service available in the province of British Columbia, Canada, reported that there had been little impact on prescription drug use in the province as a result of this program.9 In contrast, a systematic review and meta-analysis that evaluated medication reconciliation programs at hospital transitions and included pre- and post-discharge pharmacist visits reported significantly reduced adverse drug event (ADE)-related hospital revisits attributable to the interventions, which included pharmacist home visits (HVs).10 A recent randomized-controlled trial (RCT) determined that an extended intervention that included both a pharmacist-led pre-hospital discharge MR and post-discharge follow-up significantly reduced readmissions within 30 or 180 days compared with usual care; however, the MR alone did not.11

These studies were not focused solely on clinical pharmacy services in home care and so applicability to this setting is limited. A review of clinical pharmacy services offered in the home concluded that more rigorous evaluation is needed to support the value of these services and highlighted that questions remain about optimal practice models and target patient populations.12 In our health authority, home care clinical pharmacy services have matured to the point where they are an established component of home care in locations where they are available, with ongoing requests for more. The maturation of these services has seen the pharmacist involved in increasingly more aspects of home care services, beyond what was initially supported by evidence.13 Determining best practices for clinical pharmacy services offered in the home, as well as other settings, is important to guide practice that will ensure maximum patient benefit. Furthermore, changes in technology, patient and provider experience, safety, and expectations for pharmacy services are possible influencers of how services are delivered or valued.

The purpose of this review is to identify outcomes associated with clinical pharmacy services provided in the home, as well as to describe current perspectives of practice described in the literature.

Literature search

Two separate literature searches were undertaken to identify articles published for the time period from January 2007 to December 2017. This time frame was chosen to follow up on a previous review published in 2008.12 Using the key terms “Pharmacist” and “home visit”, EMBASE, Medline, OVID, CINAHL, Biomedical Reference Collection, EBMR, and Google Scholar were searched. The search was limited to the citations published in the English language and involved human subjects. Additionally, the gray literature and reference lists of articles found were searched for additional records. One hundred and fifty-six unique records were found, of which 54 were excluded as they were conference abstracts or the full article access was not possible. In addition, a further 27 were excluded as they did not describe pharmacists doing HVs in a unique study published in 2007 or onward, leaving 75 articles that were included in this review.

Different programs and authors use different terminology to refer to similar concepts. We will be referring to medication reconciliation (MRec) as the act of comparing all medication lists in order to reconcile and create a master list of what the patient should be taking. MR refers to the act of compiling a list of medications the patient is taking and assessing the appropriateness of each medication and the regimen as a whole. MRec may be included in the process of MR. We will refer to medication, therapy, or drug-related problems as drug-related problems (DRPs).

Evaluation of pharmacist home visit initiatives

Pharmacist HV initiatives in 11 countries were found described in the literature: Australia, Brazil, Canada, Japan, Jordan, the Netherlands, New Zealand, Singapore, Thailand, the UK, and the USA.

Randomized-controlled trials

The nine RCTs and two cost-effectiveness analyses of pharmacist HV initiatives are outlined in Table 1.14–25 In general, the programs included those older than 60 years and who were expected to be at increased risk of medication misadventure. Five studied patients being discharged from hospital14,16,18,19,22 and four recruited from outpatient settings.15,20,23,24 Souter et al recruited from both an inpatient and outpatient setting.25 Additional eligibility criteria used to define the target study population included number of medications (≥2 to ≥5); functional decline, frailty, or disease-specific (CHF/stroke). Six of the studies described the qualifications of the pharmacists conducting the intervention, indicating training or experience beyond an entry to practice degree.14–16,18,19,23

Table 1

Randomized controlled trials (RCTs) of pharmacist home visit services

Study details	Patient characteristics	Mean/median age	Patient numbers and gender	Pharmacist characteristics	Intervention	Evaluation period	Outcomes
Holland et al14 2007 UK	Age >18 yearsDischarge home after an emergency hospitalization for HF≥ 2 medications on discharge	76.9	149/14436.5% Female	Postgraduate qualification or recent CE in therapeutics7 hours training in HF 4 hours communication training (1/2 group)17 pharmacists	RX provided copy of discharge letterHV × 2 (within 2 and 8 weeks of discharge)Med ReviewHF EducationPCPAdherence	6 months	↔ emergency hospital admissions↔ mortality
HeartMed
Lenaghan et al15 2007 UK	Age >80 years≥4 medications≥1 medication risk factor	84.3	68/6665.6% Female	Postgraduate qualificationExperience with home-based medication review	RX received current medication and medical history	6 months	↔ non-elective hospital admissions↔ care home admissions
POLYMED	Registered with GP practice			1 pharmacist	Med ReviewEducationPCPAdherenceRemove meds		↔ mortality↓ medications prescribed↔ quality of life
Pacini et al16 2007 UKCost-effectiveness analysis of HOMER17	Age >80 yearsDischarge home from hospital≥2 medications	85.4	415/41462.4% Female	Postgraduate qualification orRecent CE in therapeutics2-day training course on prescribing in the elderly22 pharmacists	HV × 2 within 2 months of dischargeMed ReviewEducationRemove medsPCPAdherence (CommRX)	6 months	Low probability the intervention was cost-effective
Triller and Hamilton18 2007 USA	Age ≥21 yearsDiagnosis of HFReferred from hospital for home care nursing	79.7	77/7772% Female	Doctor of Pharmacy degreeResidency in home care20+ years of clinical experience1 pharmacist	HV within 1 week, plus7–10 and 18–21 days laterMed Review (HF and non-HF)Education	6 months	↔ all-cause hospitalization↔ HF hospitalization↔ all-cause mortality
Vuong et al19 2008 Australia	Age ≥55 yearsDischarge home from hospitalHospital admit for medication misadventure/misuse≥3 medicationsMedication regimen change during hospitalizationNewly trained on use of applianceRequired medication monitoring with 7 days of dischargeDexterity, vision, hearing, or other impairment that may impact medication takingChronic conditionLanguage difficultiesImminent loss of independence, housebound, or living alone	71.8	152/16447.4% Female	Bachelor of PharmacyPostgraduate diploma in clinical pharmacy2 pharmacists	HV within 5 days of dischargeMed ReviewAdherenceRemove meds	8–12 weeks	↑ self-perceived medication understanding↑ medication knowledge score↑ self-reported adherence
Hogg et al20 2009 CanadaGray et al 201021Cost-effectiveness analysis	Age ≥50 yearsOn GP practice rosterRisk of functional decline, physical deterioration, or emergency servicesGood candidate for additional medical resources	71.3	64/5657.5% Female	1 pharmacist	Anticipatory and Preventive Team Care (APTCare)Addition of RX and 3 NP to practiceMed ReviewPCP	12–18 months	↑ quality of care (QOC) for chronic disease management↑ QOC for disease preventionNot cost-effective
Barker et al22 2011 Australia	≥4 medicationsMeets Framingham criteria for HFHospital stay >48 hours	72.5	64/5654% Female	Hospital based pharmacists	HV within 96 hours of discharge, at 1 and 6 monthsAdherenceEducationRemove medsCommRX	6 months	↔ mortality↔ hospitalizations↑ all-cause and HF hospital days↓ non-HF hospital days↔ Quality of life
Elliott et al23 2012 Australia	≥2 medicationsReferred to Aged CareAssessment Team (ACAT)	84	40/4063.8% Female	≥5 years clinical pharmacy experience including subacute aged care and hospital outreach medication managementNot accredited to conduct HMR3 pharmacists	HV within 28 days of referralMed ReviewDiscuss with geriatricianPCPRemove meds	6 weeks up to20 weeks	↑ HV within 28 days of referral (100% vs 35%)↑ medication regimen changes↑ use of adherence aid↑ reported easier to manage medications
Basheti et al24 2016 Jordan	Age >18 yearsOutpatient clinic rosterAt least one of:≥5 medications≥12 doses/dayDischarge from hospital (within 4 weeks)Medication regimen changes in previous 3 monthsSymptoms suggestive of ADRSubtherapeutic response to medication	60.6	48/4970% Female	1 pharmacist	Med review (HV or clinic) for all patientsIntervention received:EducationPCP	3 months	↑ DRP resolution↓ DRP↑ medication adherence↔ quality of life
Souter et al25 2017 Scotland	Stroke diagnosisDischarged home from hospital or attends outpatient neurology clinic	73	18/1740% Female		Med review in hospital or clinicEducationHV at 1,3 and 6 months after hospital discharge or clinic visitPCPCommRX	6 months	5.8 DRPs/patient19/23 recommendations accepted by GP

Abbreviations: Adherence, adherence assessment and/or aids provided; ADR, adverse drug reaction; CE, continuing education; CommRX, communicate with community pharmacy; DRP, drug-related problem; GP, general practitioner; HF, heart failure; HMR, Home Medicines Review; HV, home visit; MD, medical doctor; Med Review, assessment of medication regimen for the purpose of identifying and resolving drug-related problems; RX, pharmacist; NP, nurse practitioner; PCP, contact primary care physician to resolve DRPs; Remove meds, removal of discontinued or expired medications.

Two studies reported reduced health care utilization attributable to the pharmacist HV intervention: reduced prescribed medications15 and reduced non-heart failure hospital days.22 The cost-effectiveness analyses of the Anticipatory and Preventative Team Care (APTcare) trial and the HOMER trial (published in 2005) did not support cost-effectiveness of the pharmacist HVs.16,17,21 The HOMER trial intervention group experienced significantly increased emergency readmissions.17 APTcare, a multidisciplinary collaboration focused on patients with chronic disease, despite providing increased quality of care was not cost-effective, this may have been reflective of it being in the implementation stage, rather than established.16,17

The clinical outcomes reported from these RCTs indicated that these programs can improve medication understanding, knowledge, and adherence and result in increased resolution of DRPs. No benefit on quality of life was reported.

While not all of the RCTs evaluated economic outcomes, it is hard to explain the limited impact of the pharmacist interventions on health care costs. The interventions undertaken in these trails all appeared to involve pharmacists conducting MR for the purpose of identifying DRPs with subsequent communication to a physician. What is unclear is the depth of the medication regimen assessment, for example, were the recommendations in line with evidence to support reducing morbidity and mortality? Also, the acceptance of recommendations made by the pharmacist was not always reported. Moreover, the extent of access the pharmacists had to medical and laboratory information was sometimes limited and may not have allowed for a comprehensive MR.14–16,19 Maturation of clinical pharmacy services may have occurred over the course of years these studies were undertaken, and later studies appeared to involve pharmacists doing more detailed reviews with greater prescriber collaboration,24,25 but did not evaluate economic outcomes. The most recent trial to evaluate economic outcomes was conducted by Barker et al; however, the usual care group received an extensive intervention which may have limited the impact of the study intervention.22

Other activities performed by the pharmacists included removing expired or discontinued medications,15,16,18,22,23 education,14–16,18,22,24,25 and adherence assessment.14–16,18,20,22 While these activities alone or together may be of benefit to patients, unless the medication regimen is optimized to ensure maximal efficacy and minimal harm, they might not be enough to significantly improve patient outcomes.

Comparison studies

The 17 studies using a comparison design for evaluation of a pharmacist HV intervention are outlined in Table 2.26–42 The majority of these studies evaluated the pharmacist HV intervention in a population of those recently discharged (acute care or skilled nursing facility). 27–31,33,34,36–38,40,41,42 Some specified patients from a primary care setting,35,39 while two studies sourced patients from administrative claims data.26,32

Table 2

Comparison studies of pharmacist home visit services

Study	Study intervention population				Pharmacist HV intervention	Comparison strategy	Comparison number	Evaluation period
	Characteristics	Age	% Female	Number
Roughead et al26 2009 Australia	Veterans or war widowsAge ≥65 yearsDispensed beta-blocker for heart failure	81.6	30	273	HMR	Matched controls 20:1 from administrative claims	5444	1 year post
Hugtenburg et al27 2009 The Netherlands	Registered at 1 of 37 study pharmaciesDischarged home from hospital≥5 prescribed medications	69.7	51.2	336	HV within 1 week of dischargeMRecMedication overview for patient and GPEducationCheck of home suppliesSynchronized medication dispensing	Delivery of medication at discharge, usual care according to DutchPharmacy Standard	379	6–9 months
Setter et al28 2009 USA	Age ≥50 yearsDischarged from hospital to home care≥1 selected diagnosis	74.9	57	110	Medication discrepancies identified and documented (all patients).Intervention received: MRec facilitated	Eligible patients assigned to a geographically separate nursing team	110	8 weeks
Flanagan et al29 2010 Canada	Age ≥65 years≥6 medicationsDischarge home	80	60.2	836	Medication ManagementProgram (MMP)HV within 1 week of dischargeMed ReviewEducationPCPAdherenceRemove meds	Before and after	836	1 year
Castelino et al30 2010 Australia	Age ≥65 yearsReferred for HMR	76.1	55	372	HMR	Before and after	372	n/a
Castelino et al3 2010 Australia	Criteria examples:≥5 medications>12 doses/daySignificant medication changes in previous 3 monthsMedicine with narrow therapeutic indexDischarged from an institution within previous 4 weeks	75.3	55	270	HMR	Before and after	270	n/a
Roughead et al32 2011 Australia	Veterans or war widowsAge ≥65 yearsWarfarin use	81.6	36	816	HMR	Matched controls 20:1 from administrative claims	16,320	>1 year
Stafford et al33 2011 Australia	Age >18 yearsDischarged from hospitalWarfarin	67.7	38	129	HMR2–3 HVs within2–8 days of discharge	Usual care	139	90 days
Stafford et al34 2012 Australia	Age >18 yearsDischarged from hospitalWarfarin	67.7	40	129	HMR2–3 HVs within 8–10 days of dischargePharmacists had completed post-discharge service (PDS) training	Usual care	139	90 days
Desborough et al35 2012 UK	Age >65 yearsOn GP rosterLiving at homeDifficulty managing medications	n/a	n/a	117	Med ReviewPCPAdherence	Before and afterCost-consequence analysis	92(16 lost to follow-up, 9 died)	6 months
Reidt et al36 2014 USA	Discharged from acute careReferred for home care nursing≥9 medications	60	67	153	HV within 1 week of home care admissionMed ReviewMRecPCP	Matched controls 1:1–3	380	30 days
Kogut et al37 2014 USA	Discharged from hospitalAge ≥50 yearsChronic medical condition	n/a	46.7 (entire cohort)	20	HV within 14 days of dischargeMed ReviewDemonstration of an electronic personal health record (ePHR) system	Those who declined the ePHR system	10	≥30 days
Hanna et al38 2016 Australia	Patients discharged from hospital at high risk of medication misadventure, eg, history of non-adherence, chronic disease, language/cultural barriers, ≥4 medications/day	72.1	51	398	HOMR (Hospital Outreach Medication Review) service provided by a Health Authority pharmacistMed ReviewPCP	Eligible patients who declined the service	118	12 months
Hamano et al39 2015 Japan	Age ≥65 yearsPrescribed medications by one of five primary care clinicsReceived HV by MD	84.7	59.3	182	Consults for patients about effects of drugs and monitor adherence	Patients who did not receive a pharmacistHV	248	Cross sectional8 months
Reidt et al40 2016 USA	Discharged home from skilled nursing facility	70.8	57	87	Pre-discharge Med ReviewEducationAdherenceHV (or by phone) one week after discharge	Usual care of geriatrician and nurse practitionerGroup assignment depending upon discharge day	189	30 days
Shcherbakov and Tereso41 2016 USA	Medicare advantage enrolleesDischarged from acute care or skilled nursing facility	78.4	43.6	156	MRecEducationPCP	Those who declined the intervention or did not respond to phone call	89	30 days
Cheen et al42 2017 Singapore	Age ≥60 years>5 medications≥2 unplanned admissions within previous 3 monthsDischarge from acute care	73.6	50.5	99	Care coordinator and RX HVHome Based Medication Review (HBMR):Med ReviewEducationPCP	Care coordinator HV	402	6 months

Abbreviations: Adherence, adherence assessment and/or aids provided; GP, general practitioner; HMR, Home Medicines Review; HV, home visit; IV, intravenous; MD, medical doctor; Med Review, assessment of medication regimen for the purpose of identifying and resolving drug-related problems; MRec, medication reconciliation; RX, pharmacist; PCP, contact primary care physician to resolve DRPs; Remove meds, removal of discontinued or expired medications; n/a, not available.

The largest proportion of studies were evaluations of the HMR in Australia.26,30,31–34 Through the HMR program, accredited pharmacists are funded to provide home-based MR services for community-based patients at risk of medication misadventure.2 In addition to conducting a comprehensive MR, pharmacists provide patients with education, assess and aid in adherence and removal of old medications. A report documenting findings and recommendations must be sent to the patient’s physician and community pharmacy.

Most of the other studies were evaluations of pharmacists conducting an HV intervention similar to the HMR,29,40 except that not all reported pharmacists removed expired or discontinued medications.26,35,38,40,42 Some authors described a MRec intervention rather than a MR.27,28,41

The outcomes evaluated and reported in these studies are outlined in Table 3. In contrast to the RCT data, more of these studies reported reduced health care costs. The exception to this was Hanna et al, who reported an overall increase in hospital admissions; however, when they broke the study population down by age, there was a benefit of reduced hospitalizations among those aged 51–65 years.38 Improvements in clinical and humanistic outcomes were also reported in these studies. The difference in impact of the pharmacist interventions on health care costs reported in these studies, compared with the RCTs, may be attributable to study design. The patients and settings were similar, as well as the extent of pharmacist training, to those described in the RCTs. The evaluation time periods in the RCTs were at least 6 months or longer, whereas these studies reported economic benefits over 30 days 36,40 and at 6 months.35,42

Table 3

Outcomes reported from non-randomized comparison studies

Economic	↓ Emergency department visits36,40,42
	↓ Hospitalization26,29,32a,35,42
	↑ Hospitalization38
	↑ Medication costs39b
	↓ Hospital and medication costs35
Clinical	↓ Drug Burden Index (DBI)30
	↑ Medication appropriateness31
	↑ Medication discrepancy resolution28
	↑ Oral anticoagulation knowledge (OAK)34c
	↓ Major and minor hemorrhagic events33
	↑ Warfarin persistence33
	↑ Medication adherence35
	↑ DRP identification36
Humanistic	↑ Satisfaction27,37d,41d

Notes:

For the time period 2–6 months after RX intervention. No difference for <2 months, 6–12 months; ↑hospitalization >12 months.

Significantly higher costs of potentially inappropriate medications (PIMs) compared to those who received a home visit from a nurse and no pharmacist home visit.

Significantly higher than usual care at 8 days post-intervention, but not at 90 days.

For intervention group only.

Abbreviations: DRP, drug-related problem; RX, pharmacist.

Program reports

Table 4 outlines the 23 articles describing evaluations of clinical pharmacy home care services in which no comparison group was used.43–65 The post-hospital discharge patient population was the most represented in these articles.44,46,48,52–54,59,60,63,64 Other authors describe programs established in community settings43,47,49,56,58,61,65 and clinics.45,50,55,57,62 Other patient characteristics included being elderly, presence of a chronic disease, or number of medications.

Table 4

Program evaluations of clinical pharmacy home visit services

Study	Patient characteristics	Service description	Evaluation details
Moultry andPoon43 2008 USA	Isolated, elderly needing assistance managing medicationsReferred by community agencyN=30 recipients over 1 year	HV 15–60 minutesMed ReviewMRecEducation emergency preparednessDisaster proof medication storageRemove medsDocumentation: action plan with DRPs to take to MDReferral to other services as needed	30-item survey completed by 18/3096% felt knowledgeable about medications after HV73% felt HV would reduce visits to MD94% satisfied/somewhat satisfied100% would recommend program to others
MacAulay et al44 2008 Canada	Discharged from hospital to home care and one of the following:Age ≥ 80 years≥ 5 medicationsUse of high risk medicationChronic conditionSuboptimal adherenceBenefit from medication educationMedication changes during hospitalizationUnresolved DRPs at discharge N=27Average age=81.1 years67% Female	HV on average 11.7 days after hospital dischargeFollow-up HV or by telephoneMed ReviewAdherenceEducationPCP	98 DRPs: 3.6 DRPs/patient↓ DRPs from visit #1 to visit #2116 recommendations: 4.3 recommendations/patientRecommendation significance17% very significant71% significant11% somewhat significantSatisfaction survey (n=16)Overall satisfaction 9.9/10Importance of HV 9.8/10Usefulness of HV 9.5/10
Stell et al45 2008 Australia	Outpatient Disease Management UnitReferrals to RX from unit coordinator for those who may benefit from RX review eg, patients taking multiple medications they organized themselves, new patients, available patientsN=24 patients received HVAverage age =79 years42% Female	Med ReviewPCP	20 medication recommendations17 medication issues identified for further clinician reviewN=34 MD responsesPerceived medication list more accurate when completed by RX↓ Time for other clinicians to obtain medication list5.4 minutes/patient not seen by RX vs 1.8 minutes/patient seen by RX
Flanagan et al46 2010 Canada	Age ≥65 years≥6 MedicationsDischarge home N=110Average age =84 years56% Female	Medication Management Program (MMP)HV within 1 week of dischargeMed ReviewEducationAdherencePCPRemove meds	259 DRPs: median 2 DRPs/patient135 Medication discrepancies: median 1 discrepancy/patient
Eichenberger et al47 2011 Switzerland	Medication history available at community pharmacy (n=79 pharmacies) with fifth year pharmacy master student internsN=54 Diabetic and age ≥60 yearsAverage age =71.4 years37% FemaleN=22 Transplant patient and age ≥18 yearsAverage age =52.6 years50% Female	76 HVs by studentsMed ReviewAdherenceRecommendations summarized for supervising RX who could decide on intervention	7.4 DRPs/patient identified vs 3.6 DRPs/patient if HV not conductedExperience of an ADE19 (86.4%) of transplant patients26 (48.1%) of diabetes patients
Hussainy et al48 2011 Australia	Patients referred to palliative care (medication screening by pharmacist)Patients discharged home from hospital: hospital visit prior to discharge and HV 7–10 days thereafter or HV if from a different hospital	HMREnsuring medication accessTeam member educationConsult and collaboration with teamLiaison with other health providers (continuity of care)Implementation through education of symptom management protocol (education)	N=422N=52 HVaverage 54.4 minutesn=113 DRP interventionsn=120 recommendationsSurvey n=20/32 (63%) response100% role was helpful90% improved medication knowledge60% changed practice95% more likely to discuss medication issues with the pharmacist
Castelino et al49 2011 Australia	Age ≥ 65 yearsHMR conducted by seven accredited pharmacistsN=224Average age =74.6 years53% Female	HMR	1110 DRPs: Average 4.9 DRPs/patient1114 recommendations to GP964 recommendations required evidence support; 94% evidence based
Willis et al50 2011 USA	Age ≥65 yearsRegistered at primary careN=118	Undergraduate pharmacy students performing HV, n=75Activities:Best possible medication historyFalls risk evaluationBlood pressure checkReviewed by pharmacist afterwards	57 (48%) patients had a change in therapy102 (86%) prescribed a falls risk medication
White and Klinner51 2012 Australia	Chinese and Vietnamese immigrants eligible for, but who have not received HMRN=17 (6 Chinese, 11 Vietnamese)	No RX HV or intervention	Two focus groups to assess perceptions of HMR among immigrantsHad not heard of HMR, but welcomed itConcern that HMR would upset MD or lack of cooperationConcerns and confusion about medicinesRX role is medicine supplyGP role is medication decisionsNeither GP nor RX helpful in responding to detailed medication questionsDifference between ethnicities in trust for MDLanguage barrier for accessing medication information
Novak et al52 2012 USA	Medicare patients recently discharged from acute or subacute careHigh risk for readmission, eg, multiple chronic conditions, multiple medications, multiple hospitalizations in the previous 12 months	Pharmacist Care Manager (PCM)HV 2–3 hours followed by at least weekly telephone callsMed ReviewMRecAdherenceEducationAssessment of falls risk, cognition, mental health, nutrition and caregiver needsPCP	30% reduction in readmissionsPCM job satisfaction
Kwint et al53 2012 The Netherlands	Age ≥65 years≥5 oral medicationsDischarge from hospitalUse one of 10 community pharmaciesN=155Median age =76 years54% Female	HV conducted by trained community pharmacistsMed Review adjusted and completed by two independent reviewers pharmacists. Reviewer pharmacists prioritized DRPs and sent back to pharmacists to discuss with MD within 4 weeks	DRPs1565 (10/patient) DRPs based on pre-visit review415 DRPs identified through HV905 (58%) DRPs resulted in a recommendation264/905 (29%) recommendations implementedDRPs identified during HV more likely to have a higher priority and recommendations implemented
Flanagan et al54 2013 Canada	Age ≥65 years≥6 medicationsDischarge from hospital103/175 (58.9%) respondentsAverage age = 79.1 years54.4% Female	Medication Management Program (MMP)HV within 1 week of dischargeMed ReviewEducationAdherenceRemove medsPCP	Satisfaction survey (telephone)High level of satisfactionPharmacists easy to understandAppreciation for resources pharmacist providedRecommendations to have more pharmacist home visits and offer phone visits
Martins et al55 2013 Brazil	Patients with hypertension referred for pharmaceutical careAge 30–74 yearsPlus 2 of the following:Blood pressure ≥140×90 mmHgUsing ≥3 medicationsRegimen changed ≥ twice in previous yearComorbidityNon-complianceN=14Average age = 61.6 years85.7% Female	6 HV × 1 hour/HV, average 30 days between visitsMed ReviewBlood pressure measurementCardiovascular risk assessmentAdherencePCP	142 DRPs (mean=10.1/patient) identified66/135 (48.8%) pharmaceutical interventions implemented:– pharmacological intervention to optimize treatment: n=27– preventive pharmacological intervention: n=23– non-pharmacological intervention: n=16Cardiovascular risk↓ n=3 patients↑ n=1 patient↔ n=9
Moultry et al56 2015 USA	African-American patientsAge ≥65 years≥1 anti-hypertensiveLiving independentlyN=306Average age = 74 years83% Female	Managing Your blood pressure (MY BP) program2 HV with RX, 1 hour each at baseline and 6 monthsBiweekly telephone calls by pharmacy studentMed ReviewMedication record and action planPCP	At 6 months↓ SBP (mean 140 vs 137 mmHg)↔ DBP90% using home BP machine↓ nonadherence↑ hypertension knowledge
Poon et al57 2015 USA	Home-Base Primary CareVeterans’ AffairsPatients likely to benefit from a HVN=49Average age =81 years12% Female	Drug Regimen Review (DRR) initially and quarterly via chart review by a pharmacistAddition of HV by pharmacy residents and students (accompanying nurse practitioners)	53 DRR and 56 HV133 recommendations→93(70%) accepted44(33%) from DRR→27 accepted89(67%) from HV→66 accepted↑DRPs identified and recommendations accepted with HV vs DRR
Onda et al58 2015 Japan	Age ≥65 yearsHad received a pharmacist HVN=4243Average age =82.7 years73% Female	Survey to pharmacists who did HVs to identify prevalence of adverse drug events (ADEs) and PIM (potentially inappropriate medication)	2053 (48.4%) prescribed a PIM165/2053 (8%) suspected PIM-induced ADETop 5 PIMs: H2 blockers, short-acting benzodiazepines, chronic stimulant laxative use, long-acting benzodiazepine, digoxin Top 5 medications associated with ADEs: anticholinergic antihistamines, ultra-long-acting benzodiazepines, sulpiride, short-acting benzodiazepines, digoxin
Kalista et al59 2015 USA	Recently discharged from hospital to Visiting Nurse Service (VNS) with a primary diagnosis of heart failureN=10Average age =81.3 years60% Female	HV by pharmacist/pharmacy resident within 1 week of VNS admission and two telephone calls (at weeks 1 and 4)Med ReviewAdherenceEducation	At 28 days:↑ Adherence2 patients readmitted vs 38% readmission rate for VNS heart failure patients1 patient died
Hanna et al60 2015 Australia	Patients discharged from hospital at high risk of medication misadventureN=487Average age =72.8 years50.3% Female	HOMR (Hospital Outreach Medication Review) service provided by a Health Authority pharmacistMed ReviewAdherenceEducationPCP	N=217 (45%) patient questionnaire responseHV worthwhile↑ Medication knowledge and understanding of how medications helped medical conditions↑ Confidence and ↓ confusion about medicationsPharmacist was helpful and suggestions would help them take medications properlyN=105/487 (21.6%) MD questionnaire response96% (n=101) agreed with recommendations92% would adopt some or all of recommendations81% (n=85) review provided greater understanding of patients’ medication management abilities
Ahn et al61 2015 Australia	Patients who had received HMRN=15	HMR	Semi-structured interviewsParticipants had limited understanding of HMRBenefits: ↑ knowledge, holistic review, medication improvement, ↑ health seeking behavior, strengthened self-management, encouraged others to have HMRDifficulties: limited information and engagement from pharmacist; delays in process; limited GP follow-up and support for program
Reidt et al62 2016 USA	Ambulatory care clinic patientsTransportation barriers to clinic attendanceUnwilling to bring medications to clinicConcerns about environmental factors affecting medication useN=53 patients (74 HV)55% age ≥65 years57% Female	Home-based Medication TherapyManagement (MTM)HV 30–60 minutesMed ReviewEducationAdherencePCP	62% referrals from internal medicine clinic51% referrals from MDTop referral reasons: 17% eachNonadherenceTransportation barriersMedication reconciliation with public health nurseMedian 3 DRPs/patient40% compliance related
Bailey et al63 2016	≥2 Chronic conditions≥2 Hospitalizations or 1 hospitalization and ≥2 emergency department visits in previous 6 monthsTarget condition driving diagnosis for index hospitalizationMedicaid/Medicare enrolleeAge ≥18 years≥6 medications or 1 high-risk medicationN=374	Pre-hospital discharge:Med ReviewEducationMedication listSafeMed: Pharmacy technician conducted post-discharge HV, within 72 hours, and follow-up by telephone callsAssist with MRec and Med ReviewReinforce EducationPharmacist: resolve DRPs through targeted MTM via telephone or clinic visit	1264 DRPs: Average 3.4 DRPs/patient642 DRPs resolved50.8% of pharmacist recommendations acceptedEstimated cost-avoidance =US$370,681Cost-avoidance/DRP identified =US$293.30
Surbhi et al64 2016 USA
Walus et al65 2017 Canada	Patient referrals sourced from:Home care intakesPatients waiting in acute care for home care serviceDirect referralsN=122 (135 referrals)Average age =71 years63.1% Female	HV or telephone appointment with pharmacistN=40 comprehensive Med ReviewN=95 targeted Med Review or educationDocumentation and communication in chart, phone calls, fax.	271 DRPs identified: average 2.1/referral250 recommendations36/81 (44%) accepted by prescriber37/4336/40 pharmacist19/36 patientAverage of 1.5 clinical pharmacy key performance indicators (cpKPIs) identified/referral: DRP resolution, education, development of pharmaceutical care plan

Abbreviations: Adherence, adherence assessment and/or aids provided; DRP, drug-related problem; GP, general practitioner; HV, home visit; MD, medical doctor; Med Review, assessment of medication regimen for the purpose of identifying and resolving drug-related problems; MRec, medication Reconciliation; RX, pharmacist; PCP, contact primary care physician to resolve DRPs; Remove meds, removal of discontinued or expired medications; HMR, Home Medicines Review; ADE, adverse drug event.

The majority of these articles describe a program in which a HV was conducted to undertake a MR.43–50,52–65 The HVs were typically conducted by a pharmacist, with some authors describing the use of pharmacy technicians,63,64 pharmacy students,50 pharmacy residents/students accompanying a nurse practitioner57 or a pharmacy resident, or a pharmacist.59 Onda et al do not specifically describe an intervention; however, a pharmacist-conducted MR is assumed.58 These authors sent a survey to pharmacists who conducted HVs, the purpose of which was to determine the prevalence of ADEs and potentially inappropriate medication use among the population.

The impact of the HV programs described in these reports were mainly related to identification of DRPs, recommendations made, or medication changes that occurred as a result of the pharmacist’s actions.44–50,53,55,62–65 A variety of other impacts were also reported: satisfaction,43,44,48,52,54,60 time reduction for other disciplines,45 ADE identification,47,58 perceptions of program,51 experience,61 adherence,56,59 clinical parameters,55,56 and knowledge.48,56,60,61 Three programs reported economic outcomes including reduced readmission rate52,59 and cost-avoidance.63

National surveys

Five nationwide surveys evaluating pharmacist HV services were identified.66–70 An evaluation of general practitioner (GP) engagement in HMR in Australia received 376 (33%) respondents, of which 180 had participated in HMR.66 The authors reported that of those who had participated in the HMR, over half did not provide written feedback on the HMR report to the pharmacist or discuss it with the pharmacist. Further, only 10.6% provided the pharmacists with patient information such as recent laboratory results and 6.7% accepted the pharmacist’s recommendations, yet over half agreed or strongly agreed that the HMR benefits their patients.

A Canadian survey received 17 responses from pharmacists who provided HVs.67 Services provided include: medication reconciliation, adherence assessment, education for patients/caregivers and health professionals, chronic disease monitoring, and assessing acute health concerns. None of the programs was government funded, and three of the pharmacists reported charging a private home care agency for their services. Facilitators for HVs identified in the survey were referrals from physicians and support from management. The barriers cited by respondents were insufficient remuneration and lack of time for completing visits.

A similar survey undertaken among British pharmacists received 247 respondents (81.5% response rate).68 The authors reported that 74% of respondents had specific but undefined training, and 81% of the services were funded through Primary Care Trusts. HV services operational beyond a year were those that included social services, GPs, and community nurses in the service protocol of operations and those that received more of their referrals from GPs (90% vs 50%).

Patient preference for medication therapy management was evaluated in Thailand.69 Based on the 265 respondents, the authors reported that patients valued this service and preferred pharmacist visits to occur in the pharmacy rather than their home and that the preferred visit length was 20 minutes rather than 1 hour.

In the Netherlands, an evaluation of implementing a HV service to patients after hospital discharge was undertaken using a focus group (22 pharmacists) to identify barriers and facilitators, followed by a survey (20 pharmacist respondents) to score the relevance and feasibility of items identified during the focus groups.70 The pharmacists included in this evaluation conducted on average 5.4 HVs/year. The authors reported that both the need for reimbursement and the readiness of community pharmacy to adapt daily routines to implement such a service as two barriers to implementation.

Current perspectives

In addition to the aforementioned reports, 18 articles describing clinical pharmacy services in a home care setting were identified in the literature.71–88 In these reports, the pharmacist HV intervention was not evaluated. The following section highlights some current perspectives based on these articles, together with those articles previously described that provided an evaluation of clinical pharmacy HV services.

Competency

Training and qualifications for pharmacists, pharmacy residents and students, and pharmacy technicians involved in HV programs varied. The HMR program in Australia requires pharmacists to be accredited.2 In some initiatives training was provided to pharmacists, pharmacy students, or pharmacy technicians who would be providing the service.14–16,33,35,41,42,50,51,53 In other reports, background education or experience of the pharmacists was mentio ned.14–16,18,19,22,23,38,42,45,60,72,73 No comparison was done at the level of qualifications, experience, or training to outcomes. In our health authority, the pharmacists working in a home care setting as part of the Medication Management Program (MMP) must have completed an Accredited Canadian Pharmacy Residency or equivalent in order to be hired. They receive orientation on conducting HVs and documentation thereafter.

Use of pharmacy students, residents, and pharmacy technicians highlights the use of resources to both provide learning opportunities and also extend the scope of clinical pharmacy services.

Competency of personnel to provide the service influences the extent to which DRPs and issues preventing patients from achieving optimal health can be identified and resolved. It includes clinical knowledge about disease states and drug therapy and the ability to communicate to extract and provide information.

Patients

The most commonly studied patient population was patients who had recently been discharged from hospital.14,16,18,19,22,24,25,27–29,33,34,37,38,41,42,44,46,48,52,54,59,60,62,63,74–76 Heart failure was the most commonly mentioned diagnosis.14,16,18,22,26,59,74,77 While HVs can be more convenient, not all patients may want or need a HV MR.69 Furthermore, they may have preferences for how long it should take.69 Several authors commented on the length of time spent at a HV, ranging from 15 minutes to 2 hours.14,19,29,37,38,43,47,48,55,56,59,74,78 In addition to HV time, travel time must be considered and these together can prevent HVs from being a broadly available service and highlight the need to restrict the service to those for whom it is necessary.

Several authors reported an increased identification of DRPs as a result of a HV compared with medication list review47 or chart review53,57 and that the DRPs identified during a HV may be more likely to result in a medication change.24,28,64 Patients included in these studies were those who might be expected to have many medications: diabetes,47 transplant,47 older patients,57 and older patients discharged from hospital.53 Poon et al identified veterans who were likely to benefit from a HV service; however, they do not further articulate this criteria.57 Age was often a consideration in the articles included in this review and may impact the outcome; although this was reported by Hanna et al, the numbers in each age group were too small to make conclusions about the impact of age.38 Vuong et al described inclusion criteria indicative of frailty in their study that selected individuals beyond age, number of medications, and discharge from hospital.19 Frailty may be a criteria to use in deciding for whom outpatient clinical pharmacy services be delivered, including HVs, as medications can impact both physical and cognitive functioning.89

Safety

Safety for pharmacists conducting HVs was discussed in five articles.36,60,76,79,80 Safety strategies reported include: conducting HVs in pairs;76 texting to inform of arrival and departure times76 calling patients not previously met prior to arrival;80 and wearing a uniform or badge.80 Pre-screening of patients with a safety risk assessment was described, with those patients believed to be a safety risk to staff ineligible for a HV.60 Similarly, in our health authority, a pre-visit telephone risk assessment screen is conducted, with follow-up items to be assessed during the HV. Depending upon the risk identified and whether or not it can be mitigated for the HV, either staff do not conduct the HV or conduct it with a security personnel.

As patients for whom HVs are provided are typically more frail, staff safety may be overlooked in HV initiatives. However, the safety of the neighborhood, the residence, the presence of pets and other inhabitants, as well as patient/caregiver/cohabitant illness and recreational drug use must also be considered.

Technology

Ten articles discussed the use of technology to aid in pharmacist HVs. 27,36,37,40,50,62,74,76,81,83 The majority described using an electronic medical record (EMR) as a way for the pharmacist to get information about the patient’s medical conditions and/or communicate with the primary care provider.27,36,40,50,62,74,76,81 Access to medical records, whether EMR or not, is essential to aid a pharmacist to better assess a medication regimen.12 The use of an electronic personal health record (ePHR) that allows patients or caregivers to maintain medical information and a medication list and exchange this information with health professionals was reported to result in identification of DRPs in significantly more patients during a HV compared to patients who did not use the ePHR.37 Use of a clinical information system to assess patient genomics and support a pharmacist’s assessment of drug interactions among home care clients resulted in significantly reduced re-hospitalizations compared to those whose drug interactions were assessed using clinical judgment and a drug information resource.82 This RCT was not an evaluation of a pharmacist HV service; some HVs were provided, but illustrates a resource that could be used to enhance MR services provided in the home.

Besides the ePHR system, all the technology described was for use by pharmacists prior to and/or after a HV, and the need for Internet connectivity in patient homes was not discussed. The ePHR system would necessitate patient access to the Internet. Pharmacist access to the Internet at patient homes is an important aspect to consider in expanding the use of technology for HV clinical pharmacy services.

Collaboration

The majority of HV programs described in the literature involved pharmacists providing the service and connecting with other health care professionals, such as physicians in order to communicate the findings from their assessment and make suggestions for changes. The reported physician acceptance of recommendations varied from 18% to 95%.14,18,23,24,28,29,42,44,65 The extent to which communication with prescribers occurred or the suggestions for change that were implemented was not always detailed. Furthermore, pharmacist and physician collaboration may not happen, even if it was the expectation of a program.66 Authors of an evaluation of pharmacist recommendation implementation and the extent of collaboration between pharmacists and GPs reported on average 50% (range 17%–86%) of pharmacist recommendations were implemented in the 12 RCTs included in the review.90 Implementation rate was higher with increased presence of elements reflective of collaboration, such as pharmacist with clinical experience; patient’s regular pharmacist providing the intervention; sharing of medical records; patient interview by a pharmacist; referral by GP; case conference; formulation of an action plan; and follow-up on actions.

As many of the programs described and evaluated in this review were new initiatives, the time needed for relationship building for collaborative practice with other health care professionals may not have been sufficient to be able to effect changes to patients’ medications and consequently health outcomes. Strategies to leverage existing relationships or create the opportunity for relationship building described in the studies include involving community pharmacists in providing HV programs,27,47,53,67 inserting a pharmacist as part of a multidisciplinary team,23,42,48,77,81,82 or adding the HV component to an existing clinical pharmacy service.57

The extent of collaboration can also depend upon the setting from which HV services are offered. Settings identified in the HV literature include: dispensing pharmacy,24,44,53,58,69,71,76,77,78,84 home care,28,29,36,46,54,59,65,81,83 chronic disease management or specialty service,23,25,43,45,48,54,62,75,79,82 institutional transition service,22,37,38,40,42,44,63,64,73,74 health care agency,41,52,57,70,72 and primary care.15,20,24,35,39,50,55 Pharmacists working in health authority or multidisciplinary teams may have more opportunity to establish collaborative relationships; however, collaborative partnerships can also be established in community settings. A downside of HV services being offered from a community pharmacy can be limited time to conduct HVs and lack of funding.65,66,68

Several authors described pharmacists providing HV services with other health care providers: paramedics,74 nurses,28,42 social workers,84 multidisciplinary teams,36,48,68,77,81 and with a nurse practitioner and primary care physician.21,40,85 Co-visiting patients with other health care providers is not only an opportunity to strengthen the team relationships but can enhance collaboration at the point of patient care through the opportunity for complementary skill sets. For example, a pharmacist working in a palliative care team reportedly increased medication-related knowledge of team members and patients.48

Another important aspect of relationship and collaboration is referral. Receiving referrals from a physician may not only impact the longevity of a HV program,68 but also may result in more collaboration for making medication changes through case conferences.15,53,71 However, receiving referrals for a pharmacist HV intervention may not occur, despite being recommended.23

HV activities

MR and MRec were the two most commonly reported HV activities, with education, adherence assessment, and removal of medications no longer used occurring often. Other activities reported less frequently were: pharmacist performing physical assessments;36,50 chronic disease monitoring;20,25,50,55,56,67 education for lifestyle changes;87 falls assessment;50,52 and assessment of cognition,52 mental health,52nutrition,52 and caregiver needs.52 A HV is an ideal opportunity to assess many aspects of a patient’s health status, balancing that with what is the best use of a pharmacist during the HVs needs to be considered.

Autonomy

The impact of pharmacists being able to enact their medication recommendations was not reported; rather pharmacists relied on prescriber acceptance of their recommendations. For example, unlike hospital settings where anticoagulation protocols have been established to allow pharmacists to dose adjust warfarin, HMR pharmacists discussed warfarin dosing changes with a physician.73 Prescribing authority for pharmacists is likely to impact this. Matthies describes his role conducting HVs to patients discharged from an emergency department and his ability to initiate or alter patients’ medications.88 His collaboration with a primary care physician and health authority position allows him access to both EMRs, as well as a collaborative working environment. Collaborative working relationships with other health care professionals and access to information necessary to properly assess drug therapy should not be considered less important if pharmacists have prescribing authority.

Limitations

It is likely that there are more home care clinical pharmacy services occurring than have been reported in the literature and identified for this review. Surveys done in Canada and the UK illustrate the breadth of services available in these jurisdictions; however, individual reports of all services included in the surveys were not found. Furthermore, it is likely that not all publications were found as the two separate literature searches conducted had only 22 citations in common.11,14,15,22,26,27,29,32,39–42,50,59,60,66,67,73,75–78 In addition, one evaluation of the MMP, that exists in our health authority, failed to show up in either search.46 No comparison to inpatient clinical pharmacy literature was conducted to evaluate whether elements that contributed to positive outcomes in the inpatient setting can or do exist in the HV clinical pharmacy services literature.

Conclusion

Pharmacist HV services are available in many countries throughout the world. Unlike literature from inpatient settings, the outcomes reported are equivocal, particularly related to the impact of a pharmacist HV intervention on subsequent health care costs. Mirroring the conclusions of a previous review of clinical pharmacy services in the home, further refinement of how pharmacist HV services should exist is needed, including the patient population ideally served by a HV and a practice model that best contributes to collaborative practice.12 Other important elements to consider in both establishing and evaluating a HV program, and which may be applicable to other settings in which clinical pharmacy services are offered, were identified. These include: staff competency, use of technology, staff safety, activities to be performed during a clinical pharmacy intervention (eg, HV), and pharmacist autonomy. Consideration of these elements could help to generate further substantiation of the role of pharmacists providing clinical services in a home care setting.