Effectiveness of Pharmacist Interventions on Cardiovascular Risk in Patients With CKD: A Subgroup Analysis of the Randomized Controlled RxEACH Trial

Original Investigation

Effectiveness of Pharmacist Interventions on Cardiovascular Risk in Patients With CKD: A Subgroup Analysis of the Randomized Controlled RxEACH Trial Yazid N. Al Hamarneh, Ross T. Tsuyuki, Charlotte A. Jones, Braden Manns, Marcello Tonelli, Nairne Scott-Douglass, Kailash Jindal, Wendy Tink, and Brenda R. Hemmelgarn Background: Affecting a substantial proportion of adults, chronic kidney disease (CKD) is considered a major risk factor for cardiovascular (CV) events. It has been reported that patients with CKD are underserved when it comes to CV risk reduction efforts. Study Design: Prespecified subgroup analysis of a randomized controlled trial. Setting & Participants: Adults with CKD and at least 1 uncontrolled CV risk factor were enrolled from 56 pharmacies across Alberta, Canada. Intervention: Patient, laboratory, and individualized CV risk assessments; treatment recommendations; prescription adaptation(s) and/or initiation as necessary; and regular monthly follow-up for 3 months. Outcomes: The primary outcome was change in estimated CV risk from baseline to 3 months after randomization. Secondary outcomes were change between baseline and 3 months after randomization in individual CV risk factors (ie, lowdensity lipoprotein cholesterol, blood pressure, and hemoglobin A1c), risk for developing endstage renal disease, and medication use and dosage; tobacco cessation 3 months after randomization for those who used tobacco at baseline; and the impact of rural versus urban residence on the difference in change in estimated CV risk.

A

Measurements: CV risk was estimated using the Framingham, UK Prospective Diabetes Study, and international risk assessment equations depending on the patients’ comorbid conditions.

Complete author and article information provided before references.

Results: 290 of the 723 participants enrolled in RxEACH had CKD. After adjusting for baseline values, the difference in change in CV risk was 20% (P < 0.001). Changes of 0.2 mmol/L in lowdensity lipoprotein cholesterol concentration (P = 0.004), 10.5 mm Hg in systolic blood pressure (P < 0.001), 0.7% in hemoglobin A1c concentration (P < 0.001), and 19.6% in smoking cessation (P = 0.04) were observed when comparing the intervention and control groups. There was a larger reduction in CV risk in patients living in rural locations versus those living in urban areas.

Am J Kidney Dis. 71(1): 42-51. Published online September 11, 2017. doi: 10.1053/ j.ajkd.2017.07.012

© 2017 by the National Kidney Foundation, Inc.

Limitations: The 3-month follow-up period can be considered relatively short. It is possible that larger reduction in CV risk could have been observed with a longer follow up period. Conclusions: This subgroup analysis demonstrated that a community pharmacy–based intervention program reduced CV risk and improved control of individual CV risk factors. This represents a promising approach to identifying and managing patients with CKD that could have important public health implications.

substantial proportion of Canadian adults (w1 in 10) are living with chronic kidney disease (CKD),1 defined as a reduction in kidney function (estimated glomerular filtration rate [eGFR] < 60 mL/min/1.73 m2) or markers of kidney damage (albuminuria with albumin excretion ≥ 3 mg/mmol or abnormalities in urine sediment or renal imaging) for more than 3 months.2 CKD is associated with a high burden of comorbid conditions and adverse outcomes,3 including increased cardiovascular (CV) risk.4 Suboptimal treatment has been reported in patients with CKD and has been associated with increased risk for progression to end-stage renal disease (ESRD).5 Identification and management of patients at early stages of CKD is therefore pivotal for slowing the progression of kidney dysfunction, preventing or delaying the development of ESRD, and reducing CV events. CV disease (CVD) is the most common cause of death in patients with CKD,6 accounting for w40% of overall 42

Correspondence to B.R. Hemmelgarn (brenda. [email protected])

deaths.2 Clinical guidelines recommend using CV risk assessment equations to guide CVD prevention and management,7 and it is now recommended that CKD should be included as part of this assessment.8,9 Despite these recommendations, risk assessment has not become part of many clinicians’ daily routine. Indeed, most patients attending physicians’ clinics report never having had a CV risk assessment.7 As primary care professionals who see patients with chronic diseases frequently,10 pharmacists are well positioned to identify patients with CKD,11 determine their CV risk, and assist in their disease management. The efficacy of pharmacists’ interventions in chronic disease has been well documented in the literature.12-20 Moreover, pharmacists in Alberta, Canada, can order and interpret laboratory tests, conduct medication management assessment, and prescribe medications. This advanced scope of practice for pharmacists, combined with the fact that >95% of patients AJKD Vol 71 | Iss 1 | January 2018

Original Investigation with CKD are cared for in primary care settings without input from a nephrologist,21 provides a unique opportunity for the use of this innovative community-based care model to aid in the identification and management of these high-risk patients. We recently published results of a multicenter randomized controlled trial18 reporting that a community pharmacy–based case finding and intervention program led to a 21% reduction in CV risk over a 3-month period when compared to usual care. In this prespecified substudy, we evaluate the effect of this intervention on estimated CV risk in the subset of patients with CKD because CVD is the most common cause of death in this patient population.6 Previous work conducted by our group showed that CVD risk factors are among the top 5 comorbid conditions in patients with CKD3 and that 95% of this population are cared for in the community without nephrologist input.21 These findings, combined with the facts that early stages of CKD are often asymptomatic11 and a large proportion of community-dwelling patients with CKD are underdiagnosed and undertreated22 highlight the care gaps in this highly vulnerable population. Such gaps indicate the need to implement and evaluate innovative models of care. Methods Overview This prespecified subgroup analysis was conducted as a part of “The Alberta Vascular Risk Reduction Community Pharmacy Project: RxEACH.”18 The CKD subgroup was identified a priori as being a high-risk group of interest. The original protocol was developed for patients with CKD only (Item S1, available as online supplementary material). This was later expanded to include other high-risk groups and additional funds were sought to undertake a larger trial, which ultimately became RxEACH.18 The study aimed to evaluate the effect of a community pharmacy–based case finding and intervention program on estimated CV risk in patients at high risk for CV events.18 RxEACH was a randomized controlled trial (with the patient as the unit of randomization) conducted in 56 community pharmacies across Alberta, Canada.18 Patient Population With respect to this subgroup, patients were eligible if they were adults (≥18 years of age) who had CKD and at least 1 uncontrolled risk factor (ie, blood pressure [≥140/ 90 or ≥130/80 mm Hg if the patient had diabetes], lowdensity lipoprotein [LDL] cholesterol concentration [>2 mmol/L], hemoglobin A1c [HbA1c] concentration [>7%], or current tobacco use [self-reported]). CKD was defined as having at least 1 of the following: 2 consecutive eGFRs both < 60 mL/min/1.73 m2 over a 3-month period and/or 2 consecutive random urine albumin-creatinine ratios (ACRs) both ≥ 3 mg/mmol over a 3-month period and/or at least 1 ACR ≥ 30 mg/mmol. AJKD Vol 71 | Iss 1 | January 2018

Patients were excluded if they were unwilling to sign the consent form/participate, unwilling or unable to participate in regular follow-up visits, or pregnant (same exclusion criteria as RxEACH).18 Recruitment Pharmacists systematically identified potential participants by focusing on target prescriptions for diabetes (eg, metformin), hypertension (eg, angiotensin-converting enzyme inhibitors), dyslipidemia (eg, statins), and previous vascular events (eg, antiplatelet agents and anticoagulants). As part of routine care, pharmacists checked the most recent laboratory test results for those patients (through the provincial electronic health record). Pharmacists then checked whether patients met the inclusion criteria (which include having at least 1 elevated risk factor). Those who met the inclusion criteria were considered eligible and were invited to participate in the study. RxEACH was approved by the research ethics boards of the University of Alberta (Pro 00041644) and University of Calgary (REB 13-0751) and registered at ClinicalTrials.gov (study number NCT01979471). Once written informed consent was obtained, patients were randomly assigned (by using a centralized secure website to ensure allocation concealment) in a 1:1 ratio to either the intervention or control group. Pharmacists used the CKD Clinical Pathway (www. CKDpathway.ca)23 to screen for CKD using eGFR and ACR. Results of tests performed in the prior 12 months were used if available and were ordered by the pharmacist if not. Test results for eGFR and ACR were entered into the CKD Clinical Pathway to confirm the presence of CKD. Participants were also asked if they had a previous diagnosis of CKD. Patients with CKD were further categorized as “known” CKD (defined as having test results showing decreased kidney function [as defined above] and a previous diagnosis of CKD as reported by the patient [confirmed by laboratory results from the provincial electronic health record] and/or pharmacist knowledge/ awareness of a previous CKD diagnosis [confirmed by the provincial electronic health record]11) or “previously unrecognized” CKD (defined as having no previous diagnosis of CKD as reported by the patient or pharmacist and no laboratory confirmation of CKD [as defined in the preceding]11). Intervention Patients randomly assigned to the intervention group received a Medication Therapy Management consultation from their pharmacist. This included: (1) patient assessments: blood pressure,24 waist circumference, weight and height measurements; (2) laboratory assessments: HbA1c and lipid profile (if not done within 3 months); and (3) individualized CV risk assessment: risk calculation, education, and discussion of this risk using an interactive online tool (https://www.epicore.ualberta.ca/rxeach/) 43

Original Investigation that explains individual CV risk and targets for intervention. Treatment recommendations and prescription adaptation(s) and/or initiation were made when necessary to meet lipid, blood pressure, and glycemic control targets and tobacco cessation, based on the most up-to-date Canadian clinical practice guidelines. Pharmacists also arranged regular communication (including any changes in the treatment regimen or updates in the laboratory results) with the patient’s treating physician(s) after each contact with the patient, and regular follow-up visits with all patients every 4 weeks for 3 months. Control Patients randomly assigned to the control group received usual pharmacist and physician care with no specific interventions. At the end of the 3-month control period, all patients were offered the intervention as outlined in the previous section. Both groups had the same assessments (including laboratory testing) at baseline and final visits. Outcomes The primary outcome was change in estimated CV risk (defined as risk for future CV events as calculated by a validated risk assessment equation) from baseline to 3 months after randomization. CV risk was calculated using an online tool. We estimated CV risk based on the Framingham risk assessment equation25 for patients who had CKD without other comorbid conditions because it has been used as the reference for CV risk engines and has been internally validated.25 However, it cannot be used in patients with previous vascular disease and it also uses diabetes as a dichotomous variable.25 If the patient had other comorbid conditions (ie, diabetes or previous vascular disease), risk was calculated using Framingham and the most appropriate risk assessment equation based on the patient’s medical history. The UK Prospective Diabetes Study risk assessment equation26 was used for those with diabetes. Because it was designed specifically for patients with diabetes, the UK Prospective Diabetes Study risk assessment equation uses HbA1c concentration as a continuous variable; it has been internally and externally validated and performed better than the Framingham risk engine in patients with diabetes,26 yet it cannot be used in patients without diabetes or patients with previous vascular events.27 The International Model to Predict Recurrent CVD risk assessment equation28 was used for patients with previous vascular disease because it is the first international comprehensive risk model to predict subsequent CV events and CV death in outpatients with established vascular disease and it has been internally validated.28 Yet it cannot be used in patients without previous vascular disease.28 If the patient had CKD and other comorbid conditions (ie, diabetes or previous vascular disease), the risk was calculated using all the respective risk assessment equations, and the risk assessment equation

44

estimating the highest risk was used. We also conducted sensitivity analysis of all risk assessment equations that we used in order to validate our primary outcome (refer to Item S2 for risk factors used in each risk assessment equation). It has been reported that 15% risk as calculated by the UK Prospective Diabetes Study risk assessment equation is equivalent to 20% CV risk as calculated by Framingham risk assessment equation.27 Secondary Outcomes Secondary outcomes were: (1) change in LDL cholesterol concentration; blood pressure, and HbA1c concentration from baseline to 3 months after randomization; (2) tobacco cessation 3 months after randomization for those who used tobacco at baseline; (3) change in 5-year predicted risk for developing ESRD from baseline to 3 months after randomization29; and (4) change in medication use and dosage from baseline to 3 months after randomization. We also assessed: (5) whether there was a difference in change in estimated CV risk for rural versus urban patients with CKD. Sample Size and Analytical Plan Using the difference between 2 independent mean values, 2-tailed t test, and the following assumptions, demographic and clinical information (age, sex, blood pressure, lipid panel, and diabetes status) from Manns et al,21 standard deviation (SD) of 4.5 from Grover et al,30 the CV risk assessment equation from D’Agostino et al25 (resulting in a starting risk level of 18.5%), 90% power, and α of 0.05, a sample size of 250 patients with CKD (125 in each group) was required to detect a 10% difference in change in CV risk between the intervention and control groups. This sample size was inflated to 288 (144 in each group) to account for possible dropouts, losses to follow-up, and withdrawals of consent. All analyses were conducted on an intention-to-treat basis. In the case of missing data, a last-observationcarried-forward approach was used. We used a generalized estimating equation in order to account for center effect when analyzing the primary outcome. The dependent variable was change in estimated CV risk from baseline to 3 months after randomization. Baseline CV risk and treatment group were considered as independent variables in the model. Secondary outcomes of change in blood pressure, LDL cholesterol concentration, HbA1c concentration, and ESRD risk were analyzed using analysis of covariance with change in the respective risk factor as dependent variable and baseline measurement and treatment group as independent variables. Tobacco cessation and other categorical variables were analyzed using χ 2 test or Fisher exact test, as appropriate. The EPICORE Centre at the University of Alberta provided the data and trial management and biostatistical support (www.epicore.ualberta.ca).

AJKD Vol 71 | Iss 1 | January 2018

Original Investigation Pharmacist Training The pharmacist training program was based on the current Canadian guidelines.2,31-35 The research team developed an online training program that was hosted online and provided at face-to-face regional meetings. The training program included modules on case finding, CV risk calculation and communication, CKD, hypertension, dyslipidemia, diabetes, tobacco cessation, diet and lifestyle management, and documentation of care plans for remuneration by Alberta Health.

Results We enrolled 723 patients in RxEACH (the first patient was enrolled in January 2014, the last patient was enrolled in June 2015, and follow-up was completed in September 2015).18 Of those, 290 patients had CKD; 147 were randomly assigned to receive the intervention, and 143 were randomly assigned to receive usual pharmacist and physician care (Fig 1). The 2 treatment groups were well balanced in baseline demographic and clinical parameters (Table 1). Mean age was 65.5 ± 12 (SD) years, 55.2% were men, 80.3% were white, 85.9% had at least a high school education, and 36.2% were working for pay/profit. Nearly a quarter (24.5%) of participants were current tobacco users (selfreported), and mean body mass index was 33.2 ± 11.8 kg/m2. The most common comorbid condition was hypertension (90%), followed by dyslipidemia (86.6%), diabetes (82.1%), and vascular disease (35.5%). In order to meet the inclusion criteria, study participants had to have at least 1 poorly controlled risk factor. Of these, 76.5% (of those with diabetes) had poor glycemic control as measured by HbA1c concentration, 70.3% had poorly controlled blood pressure, 51% had poorly controlled dyslipidemia as measured by LDL cholesterol

Table 1. Baseline Demographic and Clinical Characteristics Characteristic Age, y Male sex BMI, kg/m2 Hypertension Dyslipidemia Diabetes Previous vascular disease Stroke/TIA Myocardial infarction Acute coronary syndrome Angina Coronary revascularization Peripheral artery disease Atrial fibrillation Heart failure Systolic BP, mm Hg Diastolic BP, mm Hg Cholesterol Total, mmol/L HDL, mmol/L LDL, mmol/L eGFR, mL/min/1.73 m2 ACR, mg/mmola ACR, mg/g Social factors Married/common law At least high school Working for pay/profit White No alcohol use No specific diet Sedentary lifestyle Eligibility for study Uncontrolled HbA1c Uncontrolled BP Uncontrolled LDL cholesterol Tobacco userb

Intervention (n = 147)

Control (n = 143)

61.3 ± 12.4 162 (56.5%) 34.2 ± 14.4 134 (91.0%) 128 (87.0%) 117 (80.0%) 53 (36.0%)

61.0 ± 11.8 159 (55.6%) 34.3 ± 11.4 127 (89.0%) 123 (86.0%) 121 (85.0%) 50 (35.0%)

16 (5.5%) 13 (4.5%) 11 (3.8%)

21 (7.2%) 25 (8.6%) 13 (4.5%)

15 (5.2%) 19 (6.6%)

19 (6.6%) 15 (5.2%)

14 (4.8%)

12 (4.1%)

10 (7.0%) 4 (3.0%) 140.0 ± 21.7 79.7 ± 12.9

10 (7.0%) 9 (6.0%) 137.8 ± 23.1 79.1 ± 13.5

4.5 ± 1.3 1.2 ± 0.4 2.3 ± 1.6 65.5 ± 25.7 6 [1.9-30.9] 53.1 [16.8-273.5]

4.2 ± 1.1 1.2 ± 0.4 2.2 ± 1.2 63.9 ± 26.1 7.1 [2.8-20.2] 62.8 [24.8-178.8]

95 (65.0%) 132 (90.0%) 50 (34.0%) 115 (78.0%) 92 (62.6%) 91 (61.9%) 90 (61.2%)

102 (71.0%) 117 (82.0%) 54 (38.0%) 118 (83.0%) 87 (60.8%) 97 (67.8%) 95 (66.4%)

84/117 (71.8%) 98/121 (81.0%) 106/147 (72.1%) 98/143 (68.5%) 86/140 (61.4%) 62/133 (46.6%) 32/147 (21.8%)

39/143 (27.3%)

Note: Values for categorical variables are given as frequency or n/N (percentage); values for continuous variables, as mean ± standard deviation or median [interquartile range]. Conversion factor for cholesterol in mmol/L to mg/dL, /0.02586. Abbreviations: ACR, random urine albumin-creatinine ratio; BMI, body mass index; BP, blood pressure; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; TIA, transient ischemic attack. a mg/mmol ACR/0.113 = mg/g ACR. b Tobacco use was self-reported.

Figure 1. Study flow. Randomization refers to the original randomization in RxEACH (The Alberta Vascular Risk Reduction Community Pharmacy Project). More information about the number of the overall patients screened has been reported elsewhere.18 Abbreviation: CKD, chronic kidney disease.

AJKD Vol 71 | Iss 1 | January 2018

concentration, and 24.5% were current tobacco users (categories not mutually exclusive). More than half (59%) of the patients had known CKD, whereas 41% had previously unrecognized CKD. Of those with previously unrecognized CKD, 83% had elevated ACRs (68% had 2 ACRs between 3 and 29 mg/mmol, 16% 45

Original Investigation had 2 ACRs ≥ 30 mg/mmol, and 16% had 1 ACR ≥ 30 mg/mmol), 9% had reduced eGFRs (2 consecutive eGFRs both < 60 mL/min/1.73 m2), and 8% had abnormal results for both tests. Those with previously unrecognized CKD were referred to their treating physician and received care from their pharmacist according to the study allocation. The overall rate of missed appointments during the study was w15%. Estimated CV risk was reduced from 25.7 ± 13.2 to 20.9 ± 12.3 in the intervention group and from 28.7 ± 14.1 to 28.6 ± 15.3 in the control group over the 3-month follow-up period. When adjusted for baseline characteristics and center effect, this corresponded to a relative reduction of 20% (absolute reduction, 5.03; 95% confidence interval [CI], 3.40-6.65; P < 0.001) in estimated CV risk over a 3-month period (Fig 2). In patients with diabetes, HbA1c concentration reduction was the largest contributor to the CV risk reduction. This was followed by the reduction in systolic blood pressure, total cholesterol to high-density lipoprotein cholesterol ratio, then tobacco cessation (self-reported). In patients without diabetes, systolic blood pressure reduction was the largest contributor to the CV risk reduction, followed by reduction in total cholesterol to high-density lipoprotein cholesterol ratio and tobacco cessation (self-reported). Sensitivity analysis showed a significant CV risk reduction irrespective of the risk assessment equation used (Table 2). With regard to individual risk factors, significant improvements were observed in blood pressure, LDL cholesterol concentration, HbA1c concentration, and tobacco cessation in the intervention group when compared to the control group (Table 3). Similar results were found in the subgroup of patients with CKD defined by only eGFR < 60 mL/min/1.73 m2. Pharmacists initiated more angiotensin-converting enzyme inhibitor and statin prescriptions for patients in the intervention group compared with those in the control group (14.3% vs 1.4% [P = 0.004] and 31.5% vs 11.6% [P = 0.02], respectively). New angiotensin receptor blocker prescriptions had a nominal increase (7.3% vs

4%), but this was not statistically significant (P = 0.3). Table 4 outlines medication use and changes made by pharmacists. There were more changes in diabetes, hypertension, and dyslipidemia medications in the intervention group compared to the control group. In patients with eGFRs < 60 mL/min/1.73 m2, the 5year predicted risk for ESRD was reduced from 8.9% ± 21.8% at baseline to 3.4% ± 8.8% after 3 months in the intervention group while there was no change (5.6% ± 11.9% and 5.7% ± 12%) in the control group. This corresponded to a 27% relative reduction in predicted risk for ESRD (absolute difference, 2.38; 95% CI, −7.48 to 2.73), yet this reduction was not statistically significant (P = 0.4). We observed no difference in the magnitude of CV risk reduction in those with previously unrecognized CKD versus those with known CKD (absolute difference, 1.33; 95% CI, −2.03 to 4.69; P = 0.4). We observed an absolute difference of 3.15 (95% CI, −0.2 to 6.51; P = 0.07) in CV risk reduction when comparing rural versus urban patients. There was no difference in CV risk reduction (−0.09; 95% CI, −1.95 to 2.13; P = 0.9) between patients with CKD and those who did not have CKD in RxEACH. However, estimation of CV risk reduction for the subgroup with CKD was one of the initial questions of interest; thus, results for the CKD group specifically have been reported. There were no adverse events reported during the trial.18 Accounting for center effect did not affect the primary outcome significantly (absolute reduction in CV risk of 5.03 [95% CI, 3.62-6.44; P < 0.001] with center effect vs 5.04 [95% CI, 3.65-6.42; P < 0.001] without center effect); as such, we elected not to adjust for center effect for secondary outcomes. Discussion In this prespecified subgroup analysis of RxEACH,18 a community pharmacy–based case finding and intervention program (including prescribing and ordering laboratory tests) in patients with CKD reduced the risk for major CV events by 20% when compared to usual

31

Estimated CV risk (%)

29 27 25

Control, 3 months

Control, Baseline Intervention, Baseline

23 Intervention, 3 months

21 19 17

Intervention

Control

15 Baseline

46

3 months

Figure 2. Primary outcome, the difference in change in estimated cardiovascular (CV) risk between the intervention and control groups at 3 months, found a relative risk reduction of 20% (absolute reduction, 5.03; 95% confidence interval, 3.4-6.65; P < 0.001) in estimated CV risk over a 3-month period between the intervention and control groups, after adjusting for baseline values and the center effect. Vertical bars denote error bars.

AJKD Vol 71 | Iss 1 | January 2018

Original Investigation Table 2. Risk Difference by Study Arm Using Different Risk Assessment Equations Δ in Estimated CV Risk From BL to 3 mo After Randomization

Difference in Estimated CV Risk Reduction

Risk Assessment Equation

N

Intervention

Control

Relative

Absolute (95% CI)b

P for Absolute Differencea

Framingham UKPDS International

290 238 103

3.7 (5.4) 5.4 (8.6) 0.9 (1.5)

0.6 (4.7) 0.4 (6.4) −0.5 (2.0)

13.7% 25.5% 10.8%

−3.2 (−4.3 to −2.0) −5.4 (−7.3 to −3.6) −0.8 (−1.2 to −0.4)

<0.001 <0.001 <0.001

Abbreviations: BL, baseline; CI, confidence interval; CV, cardiovascular; UKPDS, UK Prospective Diabetes Study. a P value was calculated using Wald test. b Values from generalized estimating equations.

practice. The intervention was also associated with improvements in blood pressure, LDL cholesterol concentration, tobacco cessation (self-reported), and HbA1c concentration (in those with diabetes) and a nonsignificant reduction in 5-year predicted risk for developing ESRD. There was no statistically significant reduction in CV risk for patients with known versus previously unrecognized CKD. There was larger reduction in CV risk (of borderline statistical significance) in patients with CKD living in rural locations when compared with those living in urban areas. We found that pharmacists’ application of the CKD Targeted Screening Guidelines was successful in identifying patients with CKD. Of note, 41% of those had previously unrecognized CKD, illustrating the important role that case finding by primary care pharmacists plays in CKD care. The findings of this prespecified subgroup analysis are in line with the findings of RxEACH.18 Importantly, they can be used to raise the awareness of CVD prevention and management in community-dwelling patients with CKD, who are often underdiagnosed and undertreated.22 RxEACH-CKD adds to the body of literature of pharmacist effective interventions in improving guideline adherence,36 medication management,37 and quality of

medication dosing38 in patients with CKD in different care settings. Our findings are consistent with those of Santschi et al,14 who conducted a systematic review to assess the impact of pharmacist nonprescribing interventions on CV risk factors. They reported improvements in blood pressure, total cholesterol concentration, LDL cholesterol concentration, and smoking cessation.14 Our group has also previously reported the positive outcomes of pharmacists’ independent prescribing interventions on glycemic, blood pressure, and lipid control in patients with poorly controlled type 2 diabetes, hypertension, and dyslipidemia, respectively.17,19,20 To our knowledge, RxEACH is the first large randomized controlled trial to assess the impact of pharmacist intervention on overall CV risk among patients at high risk in a community pharmacy setting.18 The larger reduction in CV risk in patients with CKD living in rural locations compared with those living in urban areas may be related to the lack of consistent or lower levels of access to family physicians and specialists in rural areas. The 3-month follow-up period was relatively short; it is possible that the effects of the intervention could be short lived. However, it is also possible that greater

Table 3. Secondary Outcomes: Changes in Individual Risk Factors Intervention Risk Factor

BL b

Systolic BP, mm Hg (n = 283) Diastolic BP, mm Hgb (n = 283) LDL cholesterol, mmol/L (n = 257) HbA1c, % (n = 234) Tobacco cessationd

Control 3 moa

Δ From BL to 3 moa BL

3 moa

Difference Δ From BL in Change (95% CI) to 3 moa

P

140 ± 21.7 127.7 ± 15.2 12.3 ± 18.1 137.8 ± 23.1 136.7 ± 19.4 0.5 ± 13.9

10.5 (7.4-13.5) <0.001c

79.7 ± 12.9 75.5 ± 11

4.3 ± 11.8

79.1 ± 13.5 77.7 ± 12.1

1.1 ± 9.4

2.7 (0.6-4.9)c

0.01c

2.3 ± 1.6

2 ± 0.9

0.4 ± 0.9

2.2 ± 1.2

2.1 ± 1.1

0.03 ± 0.6

0.2 (0.1-0.4)c

0.004c

8.3 ± 1.7

7.7 ± 1.4 25%

0.6 ± 1.3

8.6 ± 1.9

8.6 ± 2 5.4%

0.02 ± 0.9

0.7 (0.4-0.9)c 19.6%e

<0.001c 0.04f

c

Note: Unless otherwise indicated, values are given as mean ± standard deviation. Conversion factor for cholesterol in mmol/L to mg/dL, /0.02586. Abbreviations: BL, baseline; BP, blood pressure; CI, confidence interval; HbA1c, glycated hemoglobin; LDL, low-density lipoprotein. a Three months after randomization. b BP was measured using validated automatic BP machines according to Canadian Hypertension Education Program guidelines. c Value was calculated using analysis of covariance. d Tobacco use was self-reported. e Percentage was calculated by subtracting percentage of patients who used tobacco at baseline and stopped using tobacco at 3 months in control group from percentage of patients who used tobacco at baseline and stopped using tobacco at 3 months in intervention group. f P value was calculated using Fisher exact test.

AJKD Vol 71 | Iss 1 | January 2018

47

Original Investigation Table 4. Medication Use and Changes Intervention Condition and Medication Diabetes No. of medications 0 1 2 3 4 5 No. of pts No. of medication changes/no. of ptsa No. of dose changes/no. of ptsc Hypertension No. of medications 0 1 2 3 4 5 No. of pts No. of medication changes/no. of ptsa No. of dose changes/no. of ptsc Dyslipidemia No. of medications 0 1 2 No. of pts Medication changes for no. of ptsa Dose changes for no. of ptsc

Control

BL

Follow-up

BL

Follow-up

P

6 (5.1%) 50 (42.7%) 38 (32.5%) 20 (17.1%) 3 (2.6%) 0 (0%) 117

4 (3.5%) 44 (38.3%) 44 (38.3%) 18 (15.7%) 4 (3.5%) 1 (0.9%) 115 45/115 (39.1%) 35/70 (50.0%)

7 (5.8%) 53 (43.8%) 43 (35.5%) 16 (13.2%) 2 (1.7%) 0 (0%) 121

5 (4.2%) 48 (40.3%) 48 (40.3%) 18 (15.1%) 0 (0%) 0 (0%) 119 28/119 (23.5%) 40/91 (44.0%)

0.01b 0.5b

5 (3.8%) 38 (28.8%) 48 (36.4%) 27 (20.5%) 13 (9.9%) 1 (0.8%) 132 42/132 (31.8%) 24/90 (26.7%)

8 (6.3%) 34 (26.8%) 42 (33.1%) 30 (23.6%) 13 (10.2%) 0 (0%) 127

6 (4.9%) 33 (26.6%) 40 (32.3%) 30 (24.2%) 13 (10.5%) 2 (1.6%) 124 28/124 (22.6%) 13/96 (13.5%)

0.09b 0.03b

25 (19.8%) 94 (74.6%) 7 (5.6%) 126 23/126 (18.3%) 4/103 (3.9%)

27 (22.0%) 90 (73.2%) 6 (4.9%) 123

24 (20.3%) 87 (73.7%) 7 (5.9%) 118 7/118 (5.9%) 2/111 (1.8%)

0.003b 0.4d

7 (5.2%) 40 (29.9%) 53 (39.6%) 21 (15.7%) 12 (9.0%) 1 (0.8%) 134

37 (28.9%) 84 (65.6%) 7 (5.5%) 128

Note: Unless otherwise indicated, values are given as number (percentage). Abbreviations: BL, baseline; pts, patients. a Medication change was defined as initiating a new medication or stopping a current medication. The denominator indicates the number of patients who had the condition of interest at end of follow up period. b P value was calculated using χ2 test. c Dose change was defined as change(s) in medication dose. The denominator was calculated subtracting the number of patients who had medication changes from total number of patients who had the condition of interest at end of follow-up period. d P value was calculated using Fisher exact test.

improvements leading to larger CV risk reduction could have been observed with longer follow-up, for example, in reductions in HbA1c concentration, which often take 3 months to fully manifest. Furthermore, there is no wellaccepted risk assessment equation to calculate CV risk in patients with CKD, as such, we used Framingham because it is the most commonly used risk assessment equation, although it may underestimate CV risk in such a population. Our sensitivity analysis, which showed a significant reduction in CV risk irrespective of risk assessment equation used, supports the validity of our primary outcome. This prespecified subgroup analysis was conducted as a part of RxEACH,18 as such, it may have resulted in a higher number of patients with CKD when compared to a lowerrisk population. Nevertheless, pharmacists followed the targeted screening guidelines set out in the CKD Clinical 48

Pathway23 and the international clinical practice guidelines.2,39 The use of self-report and pharmacist knowledge/awareness as some of the criteria to define CKD could introduce recall bias, which could falsely elevate the number of unrecognized CKD cases. However, pharmacists reviewed all previous laboratory tests (using the provincial electronic health record) to assess patients’ kidney status and function before categorizing them into known or unrecognized CKD. Moreover, the high proportion of unrecognized cases in the community is consistent with what has been reported in the literature.40 Furthermore, due to the nature of the intervention, blinding was not possible. Pharmacists who provided the intervention also conducted the assessment and entered the information into the study online system in which CV risk was calculated. This could have introduced bias; however, the study team AJKD Vol 71 | Iss 1 | January 2018

Original Investigation checked the entered information against source documents to ensure accuracy. We conducted a qualitative study to evaluate patients’, family physicians’, and pharmacists’ perceptions of pharmacists providing CV risk reduction services.41 Patients indicated that they appreciated being taken care of by a team of experts. They highly valued the compassion, knowledge, and empowerment they received from the pharmacist. They also highlighted easy accessibility as another enabler for pharmacists to provide such services.41 The importance of communication, ability to share patient information, trust, and better understanding of the roles, responsibilities, accountabilities, and liabilities of pharmacists within their expanded role of practice were emphasized at all levels of the health care system by the physicians and pharmacists.41 A cost-effectiveness analysis of our intervention is underway. Marra et al42 assessed the cost-effectiveness of pharmacist comprehensive care (including providing medication management, patient education, and prescribing) in patients with poorly controlled hypertension and reported improved patient outcomes and cost savings to the health care system.42 The scope of pharmacist practice in the United States and Canada is governed by state/provincial law.43,44 The majority of pharmacists across Canada can initiate a prescription for certain ambulatory conditions (eg, allergic rhinitis and sore throat) and tobacco cessation and in an emergency. They can also administer vaccines, make therapeutic changes, adjust medication dosage, and extend prescriptions for continuity of care.44 In Alberta, pharmacists can initiate prescriptions and order and interpret laboratory tests.44 In the United States, the majority of pharmacists can screen for blood pressure, cholesterol, diabetes, obesity, and tobacco use. They can also provide diet and obesity counseling, tobacco cessation interventions, and immunizations (in individuals at certain ages).45 Pharmacists can also join a “collaborative practice agreement” with a prescriber in order to broaden their scope of practice to include starting and modifying a treatment regimen, depending on the state in which they practice.43 In order to achieve the full benefit of pharmacist care as seen in RxEACH, legislative changes would be needed in the United States. Of note, the improvements in CV risk were achieved in addition to (not instead of) usual physician care. Interprofessional communication and collaboration are key in patient care. We encourage policy makers in other jurisdictions to consider broadening pharmacists’ scope of practice and pharmacists and pharmacy organizations to seize the opportunity to enhance patient care. In this prespecified subgroup analysis of patients with CKD, we demonstrated that a community pharmacy–based intervention program reduced CV risk and improved the control of individual CV risk factors. This represents a AJKD Vol 71 | Iss 1 | January 2018

promising approach for identification and management of patients with CKD, which could have major public health implications. Supplementary Material Item S1: RxEACH protocol. Item S2: Risk factors for each risk assessment equation.

Article Information Authors’ Full Names and Academic Degrees: Yazid N. Al Hamarneh, BSc(Pharm), PhD, Ross T. Tsuyuki, PharmD, MSc, Charlotte A. Jones, MD, PhD, Braden Manns, MD, MSc, Marcello Tonelli, MD, PhD, Nairne Scott-Douglass, MD, PhD, Kailash Jindal, MD, Wendy Tink, MD, and Brenda R. Hemmelgarn, MD, PhD. Authors’ Affiliations: EPICORE Centre, Department of Medicine, University of Alberta, Edmonton, Alberta (YNAH, RTT); Southern Medical Program, University of British Columbia, Kelowna, British Columbia (CAJ); Department of Medicine, University of Calgary (BM, MT, NS-D, BRH), Interdisciplinary Chronic Disease Collaboration (BM, MT, BRH); Department of Community Health Sciences, University of Calgary, Calgary (BM, MT, BRH); Department of Medicine, University of Alberta, Edmonton (KJ); and Department of Family Medicine, University of Calgary, Calgary, Alberta, Canada (WT). Address for Correspondence: Brenda R. Hemmelgarn, MD, PhD, FRCPC, Cumming School of Medicine, University of Calgary, TRW Building, 3rd Floor, 3280 Hospital Drive NW, Calgary, AB Canada T2N 4Z6. E-mail: [email protected] Authors’ Contributions: Research idea, study design, and funding obtainment: YNA, RTT, CAJ, BM, MT, NS-D, KJ, WT, BRH; data analysis/interpretation: YNA, RTT. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. Support: We acknowledge the funders of RxEACH-CKD: Alberta Health (Workforce Planning), The Cardiovascular Health and Stroke Strategic Clinical Network of Alberta Health Services, and Merck Canada (investigator-initiated funding for the educational program). None of the funders had any role in the study design; collection, analysis, and interpretation of the data; writing the report; and the decision to submit for publication. Financial Disclosure: The authors declare they have no other relevant financial interests. Acknowledgements: We acknowledge Craig Curtis and Carlee Balint (research pharmacists), Imran Hassan (statistician), Dr Navdeep Tangri (nephrologist), and Ian Creuer from the Pharmacy Department of Alberta Health Services for support. Our study could not have taken place without the dedication and caring of the RxEACH investigators, listed in descending order of recruitment: Jen Winter and Lonni Johnson (Winter’s Pharmacy, Drayton Valley); Tyler Watson and Andrew Fuller (Pharmacare Specialty Pharmacy, Edmonton); Janelle Fox (Pharmasave #325, Bonnyville); Rick Siemens (London Drugs #38, Lethbridge); Jasbir Bhui and Jasmine Basi (Medicine Shoppe #170, Edmonton); Murtaza Hassanali, Shamas Arshad, and Manpreet Mann (Shoppers Drug Mart #371, Edmonton); Theresa Lawrence, Michelle Ewen, and Maged Radwan (Rexall Pharmacy #7222 and #7266, Blairmore and Pincher Creek); Anita and Bob Brown (Shoppers Drug Mart #2401, Okotoks); Leanna St. Onge, Otti Gohrbandt, and Chelsey Collinge (Co-op Pharmacy, Rocky Mountain House); Jelena Okuka and Michelle Teasdale (Co-op Pharmacy, Lloydminster); Hyder Mohammed (Shoppers Drug Mart #2318, Lethbridge); Gehan Rizkalla (Loblaws Pharmacy #4950, 49

Original Investigation Leduc); Dixie Richardson (Safeway Pharmacy #848, Edmonton); Roberta Taylor (Roots & Berries Pharmacy, Maskwacis); Marnie Kachman and Kristy Russ (Medicine Shoppe #264, Leduc); Anita Wong (Rexall Pharmacy #9801, Edmonton); Sheilah Kostecki (Safeway Pharmacy #2730, Calgary); Terrilynn Eriksen and Sharon Beaudry (Costco Pharmacy #254, Grande Prairie); Nader Hammoud (Shoppers Drug Mart #2326, Calgary); Jim Kitagawa (Pharmasave #345, Brooks); Morenike Olaosebikan (Shoppers Drug Mart #381, Edmonton); Rosalia Yeun (Medicine Shoppe #328, Edmonton); Carlene Oleskyn and Jelena Okuka (Meridian Pharmacy, Stony Plain); Dactin and Monika Tran (Sandstone Sarcee, Calgary); Rita Lyster and Alex Lischuk (Ritas Apothecary, Barrhead); Barb Bryan and Arin Getz (Sobeys Pharmacy #1129, Calgary); Lyn Gilmore and Rick Krieser (Shoppers Drug Mart #2374, Edmonton); Nermen Kassam (Pharmacy Plus Ltd, Calgary); Azita Rezai (Shoppers Drug Mart #385, Calgary); Tony Nickonchuk and Stacy Billows (Walmart Pharmacy #1068, Peace River); Piere Danis, Brittany Dyjur, and Meghan Gainer (Safeway Pharmacy #861, Edmonton); Paulise Ly (Walmart Pharmacy #1144, Calgary); Jody Keller and Wade Mannle (Pharmasave #378, Carstairs); Uzma Saeed (Walmart Pharmacy #1071, Vegreville); Rick Mah (Sobeys Pharmacy #1139, Calgary); Ken Pitcher (Save-On Foods Pharmacy #6642, Lethbridge); Mike, Pat, and Vanda Kinshella (Value Drug Mart, Peace River); Nadine Abou-Khair and Aileen Coutts (Coop Pharmacy, Calgary); Duy Troung (Shoppers Drug Mart #2300, St. Albert); Janice Chua (Shoppers Drug Mart #324, Wetaskiwin); Brendan Ihijerika (Guardian Pharmacy, Mundare); Marlene Bykowski (Remedy Rx 222, St. Albert); Jaclyn Katelnikoff (Stafford Pharmacy, Lethbridge); Rashid Jomha (Sobeys Pharmacy #3132, Edmonton); Nick Leong (Shoppers Drug Mart #328, Edmonton); Peter Lok (Shoppers Drug Mart #357, Ponoka); Kim Lau (Remedy’s Rx #223, Calgary); Brittany Zelmer (Safeway Pharmacy #864, Edmonton); Anar Dato (Shoppers Drug Mart #2391, Calgary); Kelly Laforge (Shoppers Drug Mart #2448, Edmonton); Todd Pranchau (Shoppers Drug Mart #2450, Sylvan Lake); Folake Adeniji (Shoppers Drug Mart #310, Leduc); Brad Couldwell and Trudy Arbo (Shoppers Drug Mart #321, Calgary); Anita and Reid McDonald (Sunset Ridge Pharmacy, Calgary); and Jacki Swindelhurst (Rexall Pharmacy #7245, Drayton Valley). Peer Review: Received February 1, 2017. Evaluated by 3 external peer reviewers, with editorial input from a Statistics/Methods Editor, an Associate Editor, and the Editor-in-Chief. Accepted in revised form July 14, 2017.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

References 1. Arora P, Vasa P, Brenner D, et al. Prevalence estimates of chronic kidney disease in Canada: results of a nationally representative survey. CMAJ. 2013;185(9):E417-E423. 2. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl. 2013;1:1-150. 3. Tonelli M, Wiebe N, Guthrie B, et al. Comorbidity as a driver of adverse outcomes in people with chronic kidney disease. Kidney Int. 2015;88(4):859-866. 4. Hemmelgarn BR, Manns BJ, Lloyd A, et al. Relation between kidney function, proteinuria, and adverse outcomes. JAMA. 2010;303(5):423-429. 5. Zillich AJ, Saseen JJ, Dehart RM, et al. Caring for patients with chronic kidney disease: a joint opinion of the ambulatory care and the nephrology practice and research networks of the American College of Clinical Pharmacy. Pharmacotherapy. 2005;25(1):123-142. 6. Barylski M, Nikfar S, Mikhailidis DP, et al. Statins decrease allcause mortality only in CKD patients not requiring dialysis

50

19.

20.

21.

22. 23.

24.

therapy- a meta-analysis of 11 randomized controlled trials involving 21,295 participants. Pharmacol Res. 2013;72:35-44. Grover SA, Lowensteyn I. The challenges and benefits of cardiovascular risk assessment in clinical practice. Can J Cardiol. 2011;27(4):481-487. Khanji MY, Bicalho W, van Waardhuizen, et al. Cardiovascular risk assessment: a systematic review of guidelines. Ann Intern Med. 2016;165(10):713-722. Anderson TJ, Gregoire J, Pearson GJ, et al. 2016 Canadian Cardiovascular Society guidelines for the management of dyslipidemia for the prevention of cardiovascular disease in the adult. Can J Cardiol. 2016;32(11):1263-1282. Shiu JR, Simpson SH, Johnson JA, Tsuyuki RT. Quantifying opportunities to affect diabetes management in the community. Can Pharm J. 2006;139(3):37-38. Al Hamarneh YN, Hemmelgarn B, Curtis C, et al. Community pharmacist targeted screening for chronic kidney disease. Can Pharm J. 2016;149(1):13-17. Vivian EM. Improving blood pressure control in a pharmacistmanaged hypertension clinic. Pharmacotherapy. 2002;22(12):1533-1540. Scott DM, Boyd ST, Stephan M, et al. Outcomes of pharmacist-managed diabetes care services in a community health centre. Am J Health Syst Pharm. 2006;63(21):21162122. Santchi SR, Chiolero A, Burnand B, et al. Impact of pharmacist care in the management of cardiovascular disease risk factors. Arch Intern Med. 2011;171(16):1441-1453. Tsuyuki RT, Johnson JA, Teo KK, et al. A randomized trial of the effect of community pharmacist intervention on cholesterol risk management – the Study of Cardiovascular Risk Intervention by Pharmacists (SCRIP). Arch Intern Med. 2002;162(10): 1149-1155. McLean DL, McAlister FA, Johnson JA, et al. A randomized trial of the effect of community pharmacist and nurse care on improving blood pressure management in patients with diabetes mellitus – Study of Cardiovascular Risk Intervention by Pharmacists-Hypertension (SCRIP-HTN). Arch Intern Med. 2008;168(21):2355-2361. Al Hamarneh YN, Charrois T, Lewanczuk R, Tsuyuki RT. Pharmacist intervention for glycaemic control in the community (the RxING study). BMJ Open. 2013;3(9):e003154. Tsuyuki RT, Al Hamarneh YN, Jones CA, Hemmelgarn BR. The effectiveness of pharmacist interventions on cardiovascular risk: the multicenter randomized controlled RxEACH Trial. J Am Coll Cardiol. 2016;67(24):2846-2854. Tsuyuki RT, Houle SK, Charrois TL, et al. Randomized trial of the effect of pharmacist prescribing on improving blood pressure in the community: the Alberta clinical trial in optimizing hypertension (RxACTION). Circulation. 2015;132(2):93-100. Tsuyuki RT, Rosenthal M, Pearson GJ. A randomized trial of a community-based approach to dyslipidemia management pharmacist prescribing to achieve cholesterol targets (RxACT Study). Can Pharm J. 2016;149(5):283-292. Manns B, Tonelli M, Culleton B, et al. A cluster randomized trial of an enhanced eGFR prompt in chronic kidney disease. Clin J Am Soc Nephrol. 2012;7(4):565-572. Mitra PK, Bradley JR. Chronic kidney disease in primary care. J R Soc Med. 2007;100(1):40-45. Curtis C, Balint C, Al Hamarneh YN, et al. Online clinical pathway for managing adults with chronic kidney disease. Can Pharm J. 2015;148(5):257-262. Leung AA, Nerenberg K, Daskalopoulou SS, et al. Hypertension Canada’s 2016 Canadian Hypertension Education Program guidelines for blood pressure measurement, diagnosis,

AJKD Vol 71 | Iss 1 | January 2018

Original Investigation

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

assessment of risk, prevention, and treatment of hypertension. Can J Cardiol. 2016;32(5):569-588. D’Agostino RB, Vasan RS, Pencina MJ, et al. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008;117(6):743-753. Stevens RJ, Kothari V, Adler AI, Stratton IM, Holman RR. The UKPDS risk engine: a model for the risk of coronary heart disease in type 2 diabetes (UKPDS 56). Clin Sci. 2001;101(6):671-679. Diabetes Trials Unit: University of Oxford. UKPDS risk engine. https://www.dtu.ox.ac.uk/riskengine/FAQ.php. Accessed May 1, 2017. Wilson PWF, D’Agostino R, Bhatt DL, et al. An international model to predict recurrent cardiovascular disease. Am J Med. 2012;125(7):695-703. Tangri N, Stevens LA, Griffith J, et al. A predictive model for progression of chronic kidney disease to kidney failure. JAMA. 2011;305(15):1553-1559. The calculator can be found online at, http://kidneyfailurerisk.com/. Grover SA, Lowensteyn I, Joseph L, et al. Patient knowledge of coronary risk profile improves the effectiveness of dyslipidemia therapy. The CHECK-UP Study: a randomized controlled trial. Arch Intern Med. 2007;167(21):2296-2303. Canadian Diabetes Association. Canadian Diabetes Association 2013 clinical practice guidelines for the prevention and management of diabetes in Canada. Can J Diabetes. 2013;37(suppl 1):S1-S212. Hackam D, Quinn R, Ravani P, et al. The 2013 Canadian Hypertension Education Program recommendations for blood pressure measurement, diagnosis, assessment of risk, prevention, and treatment of hypertension. Can J Cardiol. 2013;29(5):528-542. Genest J, McPherson R, Frohlich J, et al. 2009 Canadian guidelines for the diagnosis and treatment of dyslipidemia and prevention of cardiovascular disease in the adult–2009 recommendations. Can J Cardiol. 2009;25(10):567-579. Anderson TJ, Gr egoire J, Hegele RA, et al. 2012 Update of the Canadian Cardiovascular Society guidelines for the diagnosis and treatment of dyslipidemia for the prevention of cardiovascular disease in the adult. Can J Cardiol. 2013;29(2):151-167. Tobe S, Stone J, Brouwers M, et al. Harmonization of guidelines for the prevention and treatment of cardiovascular

AJKD Vol 71 | Iss 1 | January 2018

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

disease: the C-CHANGE initiative. CMAJ. 2011;183(15): E1135-E1150. Cooney D, Moon H, Liu Y, et al. A pharmacist based intervention to improve the care of patients with CKD: a pragmatic, randomized, controlled trial. BMC Nephrol. 2015;16: 56-65. Chang AR, Evans M, Yule C, et al. Using pharmacists to improve risk stratification and management of stage 3A chronic kidney disease: a feasibility study. BMC Nephrol. 2016;17(1): 168-177. Aspinall SL, Cunningham FE, Zhao X, et al. Impact of pharmacist-managed erythropoiesis-stimulating agents clinics for patients with non-dialysis-dependent CKD. Am J Kidney Dis. 2012;60(3):371-379. KDIGO. KDIGO clinical practice guideline for lipid management in chronic kidney disease. http://www.kdigo.org/clinical_ practice_guidelines/Lipids/KDIGO%20Lipid%20Management% 20Guideline%202013.pdf. Accessed January 27, 2017. Szczech LA, Stewart RC, Su HL, et al. Primary care detection of chronic kidney disease in adults with type-2 diabetes: the ADD-CKD Study (awareness, detection and drug therapy in type 2 diabetes and chronic kidney disease). PLoS One. 2014;9(11):e110535. Donald M, King-Shier K, Tsuyuki RT, et al. Patient, family physician and community pharmacist perspectives on expanded pharmacy scope of practice: a qualitative study. CMAJ Open. 2017;5(1):E205-E212. Marra C, Johnston K, Santschi V, Tsuyuki RT. Cost-effectiveness of pharmacist care for managing hypertension in Canada. Can Pharm J. 2017;150(3):184-197. Centers for Disease Control and Prevention. Collaborative practice agreements and pharmacists’ patient care services: a resource for pharmacists. https://www.cdc.gov/dhdsp/pubs/ docs/translational_tools_pharmacists.pdf. Accessed May 1, 2017. Canadian Pharmacists Association. Pharmacists’ expanded scope of practice in Canada. https://www.pharmacists. ca/pharmacy-in-canada/scope-of-practice-canada/. Accessed May 1, 2017. American Pharmacists Association. Pharmacist scope of services. https://www.pharmacist.com/sites/default/files/files/APhA %20%20PAPCC%20Scope%20of%20Services.pdf. Accessed May 1, 2017.

51