Team-based care: A clinical pharmacist and family physicians

Introduction

In the province of British Columbia (BC), the Primary Care Charter recognized the importance of primary care for health care system sustainability. ¹ Subsequent primary care redesign has occurred, with team-based care (TBC) being an important component. TBC refers to patient care provided through interprofessional collaboration to accomplish shared goals across multiple settings.^2,3

In Abbotsford, BC (population ~150,000), a TBC collaboration between Fraser Health (FH), the regional health authority, and local family physicians (physicians) was initiated to provide an in-home clinical pharmacist service to frail elderly. ⁴ In-home pharmacy services allow for real-world assessment of medication use and minimize patient travel and transport of medications to appointments. It also helps to address space limitations in physician clinics.

Pharmacists working in primary and home care have been reported throughout the world, with equivocal outcomes.^5-7 Impact on physician workload is also unclear.^8-10 A hospital-based clinical pharmacy intervention that was extended to include post discharge care resulted in significantly reduced readmissions. ¹¹ This demonstrates there may be benefit to extending clinical pharmacy services beyond acute care and implementing clinical pharmacists in outpatient care. With increased focus on primary care, sharing models of incorporating pharmacists into primary care can guide others embarking on such efforts. The purpose of this article is to describe the implementation and evaluation of a clinical pharmacy TBC initiative in a primary care, home care setting.

Description of practice

The implementation of a TBC clinical pharmacy service occurred cyclically, as outlined in Table 1, with learnings applied within each cycle and to subsequent cycles. Establishing the service provided by a single pharmacist occurred with 1 physician practice at a time; this is referred to as cycle implementation. Subsequent practice cycle implementation started only once the previous practice was in a maintenance phase. Four 3-month implementation cycles involving 6 physicians were completed between November 2016 and June 2018.

Table 1.

Team-based care (TBC) initiative cycle framework

Implementation phase (3 months per cycle)
	Establishing service   1. Identify interested family physician (physician) and/or practice   2. Establish Memorandum of Understanding (MOU) between physician/practice and health authority (FH)    a. Covers pharmacist confidentiality and access to practice electronic medical record (EMR)    b. Outlines cycle timeline dates and day of the week reserved for TBC initiative between physician(s) and pharmacist   3. Enable remote access to practice EMR for pharmacist   4. Physician(s) and FH define patient referral criteria   5. Physician(s) identifies patients for referral to pharmacist
	Service delivery   6. Patient home visits and the physician-pharmacist case conferences booked by practice medical office assistant and communicated to pharmacist   7. Pharmacist reviews FH and physician EMR to prepare for visit   8. Pharmacist completes in-home medication assessment   9. Physician and pharmacist in-person case conference within 1 week of pharmacist home visit to discuss findings and adjustments to the therapeutic plan  10. Documentation of medication assessments and case conferences by FH pharmacist in physician and FH EMR  11. Ad hoc collaborations with other health care professionals as needed (home health, other FH services, other physicians, community pharmacist)  12. Follow-up and ongoing physician-pharmacist discussions as needed
Maintenance phase
	Physician(s) continues to refer patients to pharmacist and case discussions occur as needed

Identification of physicians occurred in collaboration with the Abbotsford Division of Family Practice (Division). The 35 Divisions representing physicians throughout BC are an essential connection for the province’s health authorities. ¹²

Initial referral criteria included patient age >70 years, Clinical Frailty Scale 5 to 8, taking >9 medications, receiving home health (HH) service(s) and patient willingness. ¹³ After the first cycle, the referral criteria evolved, through physician collaboration, to include less frail, younger patients, those not receiving HH services and with less emphasis on the number of medications.

Patient home visits and case conferences were scheduled on the same day each week and were booked by the physician’s medical office assistant. During the visits, which typically lasted 1 to 1½ hours, the pharmacist interviewed the patient and/or caregiver, sometimes facilitated through an interpreter, to assess the current drug regimen. This assessment included reviewing current health issues and symptoms, conducting basic physical assessment, discussing current medication management and adherence, collating a medication list and providing education.

In the third and fourth cycles, the patient’s community pharmacist was included in the case conference when possible and provided with relevant documentation.

Following each 3-month implementation phase, the connection with all physicians has continued as a maintenance phase with ongoing referrals and formal and informal case conferences.

Service evaluation

Evaluation of this collaboration included patient and provider experience and use of acute care and emergency department services.

Data collected prospectively by the pharmacist in all cycles included patient age, gender and number of medications. Pharmacist-identified need for medication change(s) and physician agreement to change were also collected in cycles 3 and 4.

To gauge patient experience at the end of each initial home visit, the pharmacist asked each patient or caregiver, “Has this visit helped you make more sense of your medications?” Patient satisfaction was also collected through a telephone survey administered by a co-op student, to a sampling of patients or caregivers. They were asked to rate the TBC experience from 1 (worst) to 10 (best), whether they would recommend this service to family and friends and whether it allowed them to confidently stay at home. Patient demographic, clinical and survey data were reported as means, frequencies and ranges.

After each implementation cycle, a representative from both the Division and Fraser Health met with the physician(s) to solicit their feedback, without the pharmacist present, on the perceived value of the TBC pharmacist service for the physicians and patients. Physician feedback was reviewed cyclically to identify potential improvements.

The number of emergency department (ED) visits, hospital admissions and bed-days were collected for those patients who accessed acute care resources. These data were analyzed using an interrupted time series (ITS) in which patients were stratified into 4 groups according to their length of time in the TBC initiative: 3, 6, 9 and 12 months. Impact on the health care system was measured by changes to the level of acute care avoidance, namely, the number of ED visits and hospital bed-days before versus after the introduction of the pharmacist to the interdisciplinary care team. ITS is a quantitative approach to evaluate health care initiatives with a well-defined pre-post design and involves fitting a linear regression on the acute care utilization during the “pre” period when the pharmacist was not involved.^14-16 The fitted regression model was projected into the “post” period, representing the likely hypothetical acute care usage if the pharmacist services did not exist. The projected utilization was subtracted from the actual acute care utilization during the “post” period to derive acute care avoidance. Avoidance per patient was calculated by dividing the total avoidance with the number of patients who had either at least 1 ED visit or 1 hospital bed-day, respectively. This method is preferable to simply comparing actual pre– and post–home visit utilization, because it helps to account for unpredictability of patient clinical course over time.

Results

The TBC pharmacist initiative was provided to 141 patients; 86 (61%) were female, and ages ranged from 57 to 102 years (mean = 79). The number of medications per patient ranged from 4 to 26 (mean = 12). In cycles 3 and 4, the need for medication change was identified in 94% of patients, with 99% physician agreement.

Most patients (138, 97%) endorsed that the pharmacist home visit had helped them make more sense of their medications. Additionally, the 12 survey respondents rated their overall satisfaction at 8 or higher on a scale of 1 to 10 (mean = 9.5). All reported that the pharmacist service helped them to stay confidently at home and would recommend this service to others.

Approximately half of the patients had been connected with the pharmacist for at least 1 year, with the rest distributed among 3, 6 and 9 months (Table 2). Patients were projected to use more acute care services if the pharmacist service did not exist. Patients were projected to use more acute care services if the pharmacist service did not exist. The actual use of acute care services was lower, with 39 fewer ED visits and 96 fewer hospital bed-days, and the estimated acute care avoidance was 60 ED visits (p < 0.001) and 167 bed-days (p < 0.001). Based on 2017/2018 BC Ministry of Health average costs of $288 per ED visit and $950 per hospital bed-day, the estimated cost mitigation of this service is $17,000 and $159,000, respectively. The degree of impact declined for the 1-year cohort, as seen by the decline in acute care avoidance per patient.

Table 2.

Impact on hospital utilization

		Hospitalizations						Emergency department (ED) visits
Time (months)	No. of patients	No. of patients hospitalized pre	No. of days pre	No. of projected days post	No. of actual days post	No. of days avoided	Average number of days avoided per patient	No. of patients who visited ED pre	No. of visits pre	No. of projected visits post	No. of actual visits post	No. of visits avoided	Average number of visits avoided per patient
12	67	44	406	436	369	67	1.5	54	86	98	69	29	0.5
9	21	12	92	118	71	47	3.9	16	23	28	16	12	0.8
6	28	14	69	79	50	29	2.1	20	24	27	17	10	0.5
3	25	9	37	42	18	24	2.7	15	16	17	8	9	0.6
Totals	141	79	604	675	508	167*	N/A	105	149	170	110	60*	N/A

N/A, not applicable.

^*p < 0.001.

The physicians reported that the TBC collaboration benefited complex patients with chronic illnesses, including those with cognitive decline, taking many medications and where there were concerns about medication adherence. The pharmacist’s “home view” and the impact of it on medications assisted with care plan development. The TBC pharmacist was referred to as the “eyes and ears,” and her holistic approach was acknowledged. The physicians reported that TBC improved the quality of care provided and enhanced their relationship with patients, who often reported its value to them. The physicians appreciated that the pharmacist documented in their electronic medical records (EMRs), with 1 physician commenting it was “very helpful for patient encounters to keep on track.” The physicians commented that case conferences were an opportunity for collaborative learning, through which they were made aware of current tools and evidence that could be applied to other patients, as well as providing opportunity to “bounce concepts and thoughts by.” Being able to bill for case conferences and include the community pharmacist was valued by the physicians.

Feedback from physicians led to refinements in documentation and case conference approach and use of consistent qualified interpreters.

Discussion

This clinical pharmacist primary care/home care program received positive feedback from patients and physicians, with the physicians noting a positive impact on patient care and improving their professional satisfaction. In addition, the analysis of resource utilization suggested reduced hospital days and ED visits after the pharmacist service.

This experience adds to the literature evaluating clinical services provided by pharmacists and illustrates value in providing care in a home care setting with an intentional connection to a patient’s family physician. Similar to previous studies, this evaluation does not definitively prove a causal link between the clinical pharmacy service and subsequent reduced health care resource utilization. ⁶ Given the heterogeneous nature of the services provided, it is difficult to draw firm conclusions about outpatient clinical pharmacy services. However, this evaluation did provide information on a process that resulted in both physician and patient acceptance of the service. Recognizing that the cost analysis factors only acute care avoidance, a full financial impact assessment would ideally encompass the cost of other community services, including physician workload.

Despite an overall positive response, the small number of patients surveyed limits the generalizability of the results, as does the short time frame of the acute care use analysis and the fact that a sole pharmacist provided the service. As the initiative evolved with cycle feedback and the evaluation was retrospective, there is limited information about patient characteristics to report for others to try to replicate. During the 4 implementation cycles, it became apparent that the collaborative process through which patients were identified versus the actual “type” of patients identified was more valuable. Consequently, only a few common elements to describe patients were recorded. In practice, physician panels are not typically homogeneous, and therefore the ability to adapt to the patient needs of individual practices is key to establishing pharmacy services. Furthermore, as this was a clinical initiative, information was collected for the purpose of providing and improving clinical care and processes rather than solely for evaluation. The quantity and quality of data collected and the evaluation methods reflect the tension between diligence in evaluating new initiatives and resources required for more substantive evaluation. Given that the evaluation was not a randomized controlled trial, a causal inference between the service and the outcomes is not possible. Although a randomized controlled trial could have provided more compelling evidence, ITS has been suggested as an alternative when randomization is not an option. ¹⁷

This initiative was aimed at creating a team in primary care to provide patient care. Key in this TBC initiative is the collaboration undertaken at each stage. This occurred on a macro level between FH and the Division working together to identify physicians and evaluating the initiative. This initial collaboration was key in allowing subsequent collaboration to occur.

On a health care professional level, collaboration occurred between the pharmacist and physicians that resulted in a high degree of physician agreement to make medication changes. With the exception of consistent involvement of the community pharmacist, this initiative incorporated 7 out of 8 pharmacist-physician collaboration elements associated with medication reviews that have been described. ¹⁸ In addition to elements previously described (patient interview, physician invitation to patients, case conferencing and action planning with physician, patient follow-up), the initiative also included medical record sharing, with the pharmacist being granted full access, and pharmacist competence. The involved pharmacist completed her accredited Canadian pharmacy residency (ACPR) and has almost 2 decades of experience working specifically with the older adult population in acute care, nursing home, clinic and home care settings.

The physician invitation to patients to participate, which was facilitated by the physician medical office assistant scheduling the appointments, further expanded the collaborative scope to include the assistant in the TBC initiative, reduced the pharmacist’s administrative load and provided credibility for the pharmacist with patients. Similar administrative support within FH will be required for the TBC initiative to expand to more physician practices. An unexpected collaboration developed with the interpreter who adapted her choice of words and dialect as needed and became key to achieving an effective collaboration between patient/caregiver and pharmacist. For the pharmacist, who works as the sole pharmacist at her site of practice, collaborating as part of a team has led to professional satisfaction; this is an important consideration for further recruitment and retention in such positions. Including the community pharmacist in case conferences (cycles 3 and 4) allowed an opportunity for them to collaborate and share patient information they have through their relationship with patients/caregivers that may bear on the therapeutic plan. It provided them the opportunity to better understand the rationale for medication changes and contribute to future monitoring.

Conclusion

While in-home clinical pharmacy services have existed in Fraser Health for over a decade, operationalizing them as a primary care TBC initiative was unique and well received. Physicians reported being better able to care for their patients; patients endorsed the value of the pharmacist, often reporting the benefit to their physician. Usage of ED visits and acute care bed-days was lower after the pharmacist home visit, suggesting both a cost-effective health care model and better community patient care. Further assessment of the impact, over a longer time scale, on the use of community resources, including physician workload as well as overall costs of TBC, would provide more substantive support for development of other such services. ■