- Email: [email protected]
Evaluation of Pharmacist Care for Hypertension in the Veterans Affairs Patient-centered Medical Home: A Retrospective Case-control Study
Accepted Manuscript Evaluation of Pharmacist Care for Hypertension in the Veterans Affairs Patient Centered Medical Home: A Retrospective Case-Control Study Alan J. Zillich, PharmD, Heather A. Jaynes, RN, MS, Susan D. Bex, PharmD, Amy S. Boldt, PharmD, Cassandra M. Walston, PharmD, Darin C. Ramsey, PharmD, Jason M. Sutherland, PhD, Dawn M. Bravata, MD PII:
S0002-9343(14)01208-X
DOI:
10.1016/j.amjmed.2014.11.027
Reference:
AJM 12802
To appear in:
The American Journal of Medicine
Received Date: 17 September 2014 Revised Date:
7 October 2014
Accepted Date: 11 November 2014
Please cite this article as: Zillich AJ, Jaynes HA, Bex SD, Boldt AS, Walston CM, Ramsey DC, Sutherland JM, Bravata DM, Evaluation of Pharmacist Care for Hypertension in the Veterans Affairs Patient Centered Medical Home: A Retrospective Case-Control Study, The American Journal of Medicine (2015), doi: 10.1016/j.amjmed.2014.11.027. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Evaluation of Pharmacist Care for Hypertension in the Veterans Affairs Patient Centered Medical Home: A Retrospective Case-Control Study
RI PT
Alan J. Zillich, PharmD,a,b Heather A. Jaynes, RN, MS,a Susan D. Bex, PharmD,c Amy S. Boldt, PharmD,c Cassandra M. Walston, PharmD,c Darin C. Ramsey, PharmD,c, d Jason M. Sutherland, PhD, Dawn M. Bravata, MD b,f
a
SC
e
Department of Pharmacy Practice, Purdue University College of Pharmacy, Purdue Pharmacy
M AN U
Programs, 5th/3rd Bank Faculty Office Building, 640 Eskenazi Ave, Indianapolis, IN 46202 b
Roudebush VA Medical Center, Health Services Research and Development, Center for Health
Information and Communication Excellence, 1481 West 10th Street, Indianapolis, IN Roudebush VA Medical Center, Department of Pharmacy Services, 1481 West 10th Street,
Indianapolis, IN d
TE D
c
Department of Pharmacy Practice, Butler University College of Pharmacy & Health Sciences, 4600
School of Population and Public Health, Centre for Health Services and Policy Research, University
AC C
e
EP
Sunset Ave, Indianapolis, IN 46208
of British Columbia 201-2206 East Mall Vancouver, British Columbia V6T 1Z3, Canada f
Department of Internal Medicine, Indiana University School of Medicine, 340 West 10th Street #620
Indianapolis, IN 46202 Correspondence:
Alan J. Zillich, PharmD
ACCEPTED MANUSCRIPT 2
Purdue Pharmacy Programs 5th/3rd Bank Building
RI PT
640 Eskenazi Ave Indianapolis, IN 46202 Phone: 317-880-5430
SC
Fax: 317-880-0568
M AN U
Email: [email protected]
Key Words:
Pharmaceutical Care, Pharmacy, Care Management, Disease Management, Hypertension, Adherence,
Running Head:
TE D
Outcomes
AC C
EP
Hypertension Care Management
Conflicts of Interest: All authors confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome. All authors participated in the writing of the manuscript, and have seen and approved the submitted version. Further, each author was involved in the conception and design of the study, and/or the analysis of the data.
ACCEPTED MANUSCRIPT 3
Funding: A portion of Dr. Zillich’s time is supported by a Research Career Development Award from the Veterans Affairs Health Services Research and Development (#RCD 06-304-1). The
RI PT
funding source had no role in any of the following: study design; in the collection, analysis, and
AC C
EP
TE D
M AN U
SC
interpretation of data; in writing the report; and in the decision to submit the article for publication.
ACCEPTED MANUSCRIPT 4
Abstract
RI PT
Purpose: To evaluate a pharmacist hypertension care management program within the patient centered medical home.
Methods: This was a retrospective case-control study. Cases included all hypertension patients
SC
referred to the care management program while controls included hypertension patients who were not referred to the program during the same one-year period. Each case was matched to a maximum of
M AN U
three controls based on: primary care physician, age±5 years, gender, diagnoses of diabetes and kidney disease, baseline systolic blood pressure ±10 mmHg, and number of unique antihypertensive medications. Pharmacists provided a hypertension care management program under an approved scope of practice that allowed pharmacists to meet individually with patients, adjust medications, and
TE D
provide patient education. Primary outcomes were systolic blood pressure and diastolic blood pressure at 6- and 12-months. Multivariate regression models compared each blood pressure endpoint between cases and controls adjusting for age, co-morbidities, baseline blood pressure and baseline
EP
number of blood pressure medications.
Results: 573 patients were referred to the hypertension program; 86% (465/543) had at least one
AC C
matched control and were included as cases in the analyses; 3:1 matching was achieved in 90%(418/465) of cases. At baseline, cases and controls did not differ with respect to age, gender, race, or co-morbidity; baseline blood pressure was higher (139.9/80.0 vs. 136.7/78.2 mmHg, p≤0.0002) in the cases compared to controls. Multivariate regression modeling identified significantly lower systolic blood pressure for the cases compared with controls at both 6 and 12 months (6-month RR=9.7, 95% CI:2.7-35.3; 12-month RR=20.3, 95% CI:4.1-99.2; p<0.01 for both
ACCEPTED MANUSCRIPT 5
comparisons). Diastolic blood pressure was significantly lower at 12-months (RR=2.9, 95% CI:1.27.1; p<0.01) but not at 6-months (RR=1.0, 95% CI:0.31-3.4; p=0.9) for the cases compared to the
RI PT
controls. Conclusions: Patients referred to the pharmacist hypertension care management program had a significant improvement in most blood pressure outcomes and may be an effective method of
AC C
EP
TE D
M AN U
SC
improving blood pressure control among patients in a medical home model of primary care.
ACCEPTED MANUSCRIPT 6
Introduction An interdisciplinary approach is one strategy that significantly and consistently improves blood
RI PT
pressure control.[1-5] Interdisciplinary teams typically include non-physician providers such as
pharmacists and nurses who focus on hypertension care. One meta-analysis on the potency of team care interventions found that interventions involving clinical pharmacists resulted in lower blood
SC
pressure and greater likelihood of achieving blood pressure control.[1] The potency of team care programs has been attributed not only to the specific intervention(s) employed but also the duration of
M AN U
the program, the type of organization, and the design of the trial to measure effectiveness of the program. In another systematic review on the effect of pharmacists as team members, ChisholmBurns found that pharmacist participation significantly improved outcomes for hypertension.[6] In the evolving patient-centered medical home model, the pharmacist’s role as part of the
TE D
interdisciplinary team has been described.[7] Within the patient-centered medical home, referral to a clinical pharmacist for hypertension management is associated with improved patient outcomes and provides primary care clinicians with additional time to focus on other patient priorities.[8] Adoption
EP
and implementation of a patient-centered medical home (or Patient Aligned Care Team-PACT) model
AC C
is ongoing at Veteran Affairs (VA) Medical Centers. The model establishes a group of practitioners, including physicians, nurses, pharmacists, social workers, and additional specialties to partner with patients for a holistic and efficient primary care experience. This interdisciplinary model unites the healthcare team and improves outcomes by focusing on patient-centered care, but the effect of the pharmacist within the VA patient aligned care team model is largely unknown. The purpose of the present study was to evaluate the effectiveness of pharmacists as part of the patient aligned care team for veterans with hypertension.
ACCEPTED MANUSCRIPT 7
Methods
RI PT
Study Setting This study was conducted at an urban Midwest VA Medical Center (VAMC) that provides inpatient and outpatient health care services to almost 200,000 veterans. Within primary care, services are provided across multiple clinics. Within each clinic, PACT teamlets provide direct patient care; each
SC
teamlet consisting of the patient, their clinician (usually a physician but in some cases an advanced
M AN U
practice nurse), the nurse case manager (RN), and the health technician (LPN). Clinical pharmacists, social workers, health psychologists and support staff are designated as members of several teamlets (and are shared across teamlets within a given clinic). For this study, a care management program was evaluated for patients with hypertension provided by clinical pharmacists in four primary care clinics
TE D
and their associated PACT teamlets. Study Design
This was a retrospective case-control study using electronic medical record data to evaluate the
EP
effectiveness of the pharmacist care management program. The study was approved by the Indiana University/Purdue University-Indianapolis Institutional Review Boards and the VA Research and
AC C
Development Committee. Study Population
Patients with hypertension who were referred to the care management program during a 1-year evaluation were included as possible cases. Controls included hypertension patients, in the same primary care clinics, who were not referred to the program during the same one-year period. All patients were followed for 12 months.
ACCEPTED MANUSCRIPT 8
Care Management Program The hypertension care management program was provided by six clinical pharmacists within four
RI PT
primary care clinics throughout the duration of the evaluation period. Two clinics had one assigned pharmacist and the other two clinics had two pharmacists, based on the larger patient volume of those clinics. All pharmacists had completed a Doctor of Pharmacy (PharmD) program and at least one
SC
year of clinical pharmacy residency training. Several pharmacists had also obtained Board
Certification in Pharmacotherapy (BCPS). The pharmacists operated under a scope of practice
M AN U
allowing them to meet individually with patients, provide patient education, and initiate, change, and discontinue medications.
An electronic form was used by primary care providers to refer patients with hypertension to the pharmacists. During the referral process, many primary care providers would “curbside” the
TE D
pharmacists to discuss initial therapy options. Following referral to the program, patients were scheduled for an initial face-to-face visit with the pharmacist. Typically, the initial visit was scheduled to occur within 30 days of referral and occurred on the same day as the patient’s regular
EP
primary care clinic visit. The pharmacists met with the patient for 15-30 minutes and measured blood
AC C
pressure using standard, calibrated devices which were available to all providers in the clinic. Blood pressure measurements were conducted using the American Heart Association guidelines.[9]
As part of the pharmacists’ scope of practice, treatment decisions were based on national clinical practice guidelines from the JNC 7 and the VA/Department of Defense (which were the current guidelines at the time of this evaluation) using a stepwise approach to medication management.[10, 11] Thiazide diuretics were preferred first line antihypertensive medications with considerations for
ACCEPTED MANUSCRIPT 9
the use of other medications based on co-morbid disease states. In general, medication dosages were maximized before additional medications were added to the regimen. Blood pressure goals were
RI PT
established according to guidelines; patients with diabetes and chronic kidney disease had a systolic blood pressure goal less than 130 mm Hg and a diastolic blood pressure goal less than 80 mm Hg. All other patients had a systolic blood pressure goal less than 140 mm Hg and diastolic blood
SC
pressure goal less than 90 mm Hg.
M AN U
Patients were discharged from the program once blood pressure goals were attained at either a pharmacist visit or during a primary care provider visit. However, if the patients’ blood pressure became uncontrolled at any subsequent primary care visits, the provider could re-refer the patient back to the pharmacist for additional management. This referral, discharge back to primary care and re-referral process was designed to maximize the pharmacists’ panel size with those patients who
TE D
needed intensive treatment while maintaining the primary care providers’ role in the overall care of their patients. Patients could also be discharged from the program after two consecutive no-show
EP
visits.
AC C
Data Collection, Outcomes and Analyses
Retrospective data was extracted from the VA electronic medical record. To be included in the cohort, all unique patients with a diagnosis of hypertension (ICD-9 code: 401.1X-404.1X) and one or more primary care visits during a 1-year care management evaluation period were identified. Patients who died during the evaluation period were excluded (Figure 1). For all eligible patients, data were obtained for 1-year before the intervention, the 1-year period during the intervention evaluation and 1-year after the intervention period. The following data elements were extracted during the 1 year
ACCEPTED MANUSCRIPT 10
period prior to the evaluation: gender, race, co-morbid conditions based on outpatient ICD-9 codes (including those for diabetes and chronic kidney disease), primary care clinic assignment, primary
RI PT
care provider assignment, and the number of unique blood pressure medications. Additional data elements were extracted for the 1 year evaluation period and the 1 year follow-up period: dates of all primary care visits, blood pressure values (both systolic and diastolic) at each primary care visit, as
SC
well as blood pressure medication data (drug name, quantity, days supply, and dates of refills).
All patients cared for in the pharmacist hypertension care management program were considered to
M AN U
be potential cases. All remaining patients in the cohort were identified as potential controls and available for matching (Figure 1). Each case was matched to a maximum of three controls based on: primary care provider and clinic, age ± 5 years, gender, diagnoses of diabetes and chronic kidney disease, baseline systolic blood pressure ±10 mmHg defined as the average of the previous 3 blood
TE D
pressures, and number of unique blood pressure medications at baseline. All criteria had to be met in order to be considered a match. All cases with one or more matched controls were included in the analyses.
EP
For cases, baseline was defined as the date of referral to care management program. For controls,
AC C
patients were allocated to a baseline date that coincided with the proportion of matched cases with a baseline date during each month of the 1-year intervention period. For example, if 10% of the cases had a baseline date in June, then all 10% of those matched controls were assigned a baseline date in the month of June. All cases and matched controls were followed for 12 months from baseline. The primary outcomes were differences in systolic and diastolic blood pressure at 6- and 12-month follow-up in the cases compared to the controls. For each time point, the average of all primary care blood pressure values during that time period was used for the outcome assessment. Secondary
ACCEPTED MANUSCRIPT 11
outcomes included the number of patients reaching national guideline-define blood pressure goals (both systolic and diastolic blood pressure) at 6-and 12-months.
RI PT
Additional analyses explored the number and type of medication changes, defined as the addition of a new blood pressure medication or increasing the dose of an existing blood pressure medication, occurring in the cases compared to the controls during the 12-months of follow-up. To explore
SC
medication adherence at 12-months, the medication possession ratio (MPR) was calculated based on the method described by Steiner and colleagues as a continuous, multiple-interval measure of
M AN U
medication availability (CMA).[12] To account for medication switches and concurrent therapy, an average of all blood pressure drugs’ MPRs within a therapeutic class was computed to produce one averaged MPR for each class and an overall MPR for all blood pressure medications; the average MPR was then dichotomized with poor adherence defined as <0.8 and good adherence defined as
TE D
>0.8.[13]
Descriptive statistics were calculated to characterize the baseline demographics of both groups. Multivariate linear regression models compared each blood pressure endpoint at 6- and 12-months
EP
between the cases and controls adjusting for age, co-morbidities, baseline blood pressure and baseline
AC C
number of blood pressure medications. Logistic regression models, using similar covariates, evaluated achievement of blood pressure goals. Given that missing data for the blood pressure outcomes measurement was not random and dependent on factors such as clinic follow-up, a noncompletely-at-random (NMAR) multiple imputation method was used for the blood pressure values.[14, 15] Data were analyzed using SAS and R, with all p-values less than 0.05 considered to be statistically significant.
ACCEPTED MANUSCRIPT 12
Results Eighty six percent (465/543) of program participants had at least one matched control and were
RI PT
included in the analyses as cases (Figure 1). Three-to-one matching was achieved in 90% (418/465) and 18 (4%) patients had a 2-1 match and 29 (6%) patients had a 1-1 match. Cases and matched
controls all came from the same matched physicians and clinics and did not differ with respect to age,
SC
race, gender, or co-morbidities (Table 1). Baseline blood pressure was higher (139.9/80.0 vs.
136.7/78.2 mmHg; p≤0.0002) and baseline blood pressure control lower (35% vs. 49%, p<0.0001) in
M AN U
the cases compared to controls. Among cases, systolic blood pressure decreased from baseline by -4.0 and -7.1 mmHg at 6- and 12-months while systolic blood pressure decreased among controls by -1.6 and -2.6 mmHg at 6- and 12-months. For diastolic blood pressure, the cases decreased from baseline by -2.5 and -3.2 mmHg at 6- and 12-months while in the controls, diastolic blood pressure decreased
TE D
from baseline by -1.1 and -1.2 mmHg at 6- and 12-months.
Controlling for baseline covariates, multivariate regression modeling identified significantly lower systolic blood pressure for the cases compared with controls at both 6- and 12-months (6 month
EP
RR=9.7, 95% CI:2.7-35.3; 12 month RR=20.3, 95% CI:4.1-99.2; p<0.01 for both comparisons).
AC C
Diastolic blood pressure was significantly lower at 12-months (RR=2.9, 95% CI:1.2-7.1; p<0.01) but not at 6-months (RR=1.0, 95% CI:0.31-3.4; p=0.9) for the cases compared to the controls. Blood pressure control was better among cases than controls at 12-months (64% vs. 60%; OR=1.3, 95% CI:1.1-1.6; p<0.01) but not at 6-months (57% vs. 57%; OR=1.1, 95% CI:0.9-1.3; p=0.28). The MPR for all blood pressure medications and the proportion of patients with an MPR >80% did not differ between cases and controls at 12 months (Table 2). There were more medications added or
ACCEPTED MANUSCRIPT 13
doses increased in the cases compared to the controls for alpha-1 blockers, alpha-2 agonists, calcium channel blockers, and non-loop diuretics (Table 2).
RI PT
Discussion
Findings from this and other studies support the expanded role of the pharmacists as part of the
patient-centered medical home to improve the hypertension and other cardiovascular outcomes of
SC
patients.[8, 16] Indeed, in this VA patient-centered medical home model, pharmacists are able to
M AN U
prescribe and discontinue hypertension medications without direct oversight from the primary care physician, which has been identified as a driver of better blood pressure outcomes.[2] The results of this case-control study add to evidence from two recently published systematic reviews of team-based care for hypertension. In one review of 14 studies, pharmacist participation in hypertension teambased care decreased systolic blood pressure by 7.8 mmHg (95% CI, -9.7 to -5.8) and diastolic blood
TE D
pressure by 2.9 mmHg (95% CI, -3.8 to -2.0) [6] while the other review of 13 studies found that pharmacists’ participation lowered median systolic blood pressure by 5.0 mmHg and diastolic blood pressure by 1.7 mmHg.[2] Results from our study found similar, clinically meaningful results;
EP
lowering systolic blood pressure by 4.0 and 7.1 mmHg and diastolic blood pressure was by 2.5 and
AC C
3.2 mmHg at 6 and 12 months, respectively. This current study provides data on potential mechanisms through which these advantageous blood pressure outcomes are achieved. Our data suggest that the effect of the pharmacists’ blood pressure improvement is not due to improvement in medication adherence. Overall medication adherence at 12-months, as measured by the MPR, was 0.71 in both the cases and controls (p=0.89). Also, the proportion of patients with good antihypertensive medication adherence (MPR >0.8) was 35% for the cases and 32% for the controls (p=0.69). However, the number of changes for several classes of
ACCEPTED MANUSCRIPT 14
blood pressure medications, inclusive of newly added blood pressure medications or increases in existing blood pressure medication doses, occurred significantly more often in the cases compared to
RI PT
the controls. For example, the use of calcium channel blockers increased from 24% at baseline to 37% during follow-up in the cases compared to an increase from 27% to 30% in the controls
(p=0.005 for the follow-up comparison). These data reflect, to some extent, the primary role of the
SC
pharmacist in this patient-centered medical home model for hypertension care. Pharmacists were tasked with changing the medication regimen to maximize effectiveness, balancing the efficacy and
M AN U
safety of the medications with the preferences of the patient. While specific data on medication adverse effects was not collected, it is possible that the hypertension drug regimens initiated by the pharmacists were more potent while avoiding or mitigating adverse effects. Future studies of pharmacist care models should capture data on outcomes related to identification and mitigation of
TE D
adverse effects.
In addition, pharmacists in the patient-centered medical home model serve as “physician-extenders”, allowing the team to provide effective and efficient care while improving the Veterans’ access to
EP
primary care providers. Our team evaluated a similar patient-centered medical home model within VA for dyslipidemia and found that time to achieve LDL cholesterol control was 80% faster in the
AC C
pharmacist care group compared to a control group (p=0.002).[8] Although time to hypertension control was not examined in the current study, the results suggest that referral of patients to a pharmacist for management of hypertension is an efficient allocation of primary care provider resources, allowing the physician additional time to focus on other, more complex, priorities for the patient. Future studies should examine the time and cost savings associated with this patient-centered medical home model.
ACCEPTED MANUSCRIPT 15
The study findings should be considered with several limitations. First, the intermediate endpoint of blood pressure was the primary outcome for this study. Major endpoints such as cardiovascular
RI PT
related morbidity and mortality were not examined in this internally funded, single-site study. Similarly, other patient-reported endpoints such as medication adverse events and patient satisfaction were not collected. Nevertheless, the magnitude of decrease in systolic blood pressure is clinically
SC
meaningful, where evidence from clinical trials suggests that these blood pressure changes would result in lower morbidity and mortality.[11] Additionally, the results found significantly more
M AN U
changes in blood pressure medications for the cases compared to controls but no significant improvements in medication adherence. It is possible that improvements in medication adherence could have occurred but clinical nuances could not be detected using refill claims records and the MPR. In fact, pharmacists could have increased the dose of certain blood pressure medications by verbally informing the patient to double the dose and updating the medication list in the chart but
TE D
without ordering a new prescription for filling or instruct patients to stop other medications that were already dispensed. Therefore, measurement of adherence using the MPR may not be sensitive to
EP
small improvements. Finally, while our retrospective case-control study was robust relative to prepost study designs, the matching process for controls was incomplete. A stronger design such as a
work.
AC C
prospective case-control or randomized study would be ideal but was not possible for this unfunded
Conclusion
Patients referred to the pharmacist hypertension care management program had a significant improvement in blood pressure. This program may be an effective method of improving blood pressure control among Veterans in the patient-centered medical home model.
ACCEPTED MANUSCRIPT 16
RI PT
Acknowledgements Prior Presentations: This work was presented in abstract, poster, and podium format at the American College of Clinical Pharmacy Annual Meeting, Hollywood, FL October 20th-24th, 2012 and at the VA
AC C
EP
TE D
M AN U
SC
Health Services Research and Development Annual Meeting, Washington, DC July 16th-19th, 2012.
ACCEPTED MANUSCRIPT 17
References
6. 7. 8. 9. 10.
11.
12. 13.
14. 15. 16.
RI PT
SC
5.
M AN U
4.
TE D
3.
EP
2.
Carter, B.L., M. Rogers, J. Daly, S. Zheng, and P.A. James, The potency of team-based care interventions for hypertension: a meta-analysis. Arch Intern Med, 2009. 169(19): p. 1748-55. Proia, K.K., A.B. Thota, G.J. Njie, et al., Team-based care and improved blood pressure control: A community guide systematic review. American Journal of Preventive Medicine, 2014. 47(1): p. 86-99. Santschi, V., A. Chiolero, B. Burnand, A.L. Colosimo, and G. Paradis, Impact of pharmacist care in the management of cardiovascular disease risk factors: a systematic review and meta-analysis of randomized trials. Arch Intern Med, 2011. 171(16): p. 1441-53. Walsh, J.M., K.M. McDonald, K.G. Shojania, et al., Quality improvement strategies for hypertension management: a systematic review. Med Care, 2006. 44(7): p. 646-57. Walsh, J.M., V. Sundaram, K. McDonald, D.K. Owens, and M.K. Goldstein, Implementing effective hypertension quality improvement strategies: barriers and potential solutions. J Clin Hypertens, 2008. 10(4): p. 311-6. Chisholm-Burns, M.A., J. Kim Lee, C.A. Spivey, et al., US pharmacists' effect as team members on patient care: systematic review and meta-analyses. Med Care, 2010. 48(10): p. 923-33. Smith, M., D.W. Bates, T. Bodenheimer, and P.D. Cleary, Why pharmacists belong in the medical home. Health Aff, 2010. 29(5): p. 906-13. Smith, M.C., A.S. Boldt, C.M. Walston, and A.J. Zillich, Effectiveness of a Pharmacy Care Management Program for Veterans with Dyslipidemia. Pharmacotherapy, 2013. 33(7): p. 736-743. American Heart Association. Heart Disease and Stroke Statistics-2004 Update. Dallas, TX: American Heart Association; 2004. VA/DoD Clinical Practice Guideline for the Management of Hypertension in Primary Care. 2004 [cited 2010 June 25]; Available from: http://www.healthquality.va.gov/Hypertension_Clinical_Practice_Guideline.asp. Chobanian, A.V., G.L. Bakris, H.R. Black, et al., Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension, 2003. 42(6): p. 1206-52. Steiner, J.F. and A.V. Prochazka, The assessment of refill compliance using pharmacy records: methods, validity, and applications. J Clin Epidemiol, 1997. 50(1): p. 105-16. Peterson, A.M., D.P. Nau, J.A. Cramer, J. Benner, F. Gwadry-Sridhar, and M. Nichol, A checklist for medication compliance and persistence studies using retrospective databases. Value Health, 2007. 10(1): p. 3-12. Graham, J.W., Missing data analysis: making it work in the real world. Annu Rev Psychol, 2009. 60: p. 549-76. He, Y., Missing data analysis using multiple imputation: getting to the heart of the matter. Circ Cardiovasc Qual Outcomes, 2010. 3(1): p. 98-105. Bex, S.D., A.S. Boldt, S.B. Needham, et al., Effectiveness of a Hypertension Care Management Program Provided by Clinical Pharmacists for Veterans. Pharmacotherapy, 2011. 31(1): p. 31-38.
AC C
1.
ACCEPTED MANUSCRIPT 18
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1. Study Cohort Flow Diagram
ACCEPTED MANUSCRIPT
Table 1. Baseline Characteristics of the Cases and Controls
SC
RI PT
Controls (n=1,268) 1,268 (100) 63.2 + 11.5 451 (35) 175 (14) 642 (51) 433 (34) 32 (3) 2.5 + 1.2
136.7 + 10.6 78.2 + 8.7 617 (49)
b
EP
TE D
M AN U
Cases Characteristics (n=465) Male, No (%) 465 (100) Age, mean (SD), y 62.6 + 11.9 White, No (%) 148 (31) Black, No (%) 67 (15) Other/Unknown, No (%) 250 (54) Diabetes, No (%) 160 (34) Kidney Disease, No (%) 15 (3) Number of Co-morbidities, mean (SD)a 2.6 + 1.2 Baseline Systolic Blood Pressure, mean (SD), mm Hgb 139.9 + 11.0 Baseline Diastolic Blood Pressure, mean (SD), mm Hgb 80.0 + 9.3 b Baseline Blood Pressure Control, No (%) 164 (35) Baseline Number of Blood Pressure medications, mean (SD)b 3.3 + 1.6 Baseline Medication Classes Alpha 1 Blockers, No (%) 14 (3) Alpha 2 Agonists, No (%) 96 (21) ACE Inhibitors/Angiotensin Receptor Blockers, No (%) 294 (63) Beta Blockers, No (%) 169 (36) Calcium Channel Blockers, No (%) 110 (24) Thiazide Diuretics, No (%) 184 (40) Other (Non-Loop) Diuretics, No (%) 61 (13) Others, No (%) 10 (2) a Calculated from the Elixhauser co-morbidity index
2.9 + 1.4 38 (3) 282 (22)
814 (64) 499 (39) 340 (27) 528 (42) 193 (15) 23 (2)
AC C
Statistically significant difference between cases and controls using t-tests or chi-square as appropriate (p < 0.05)
ACCEPTED MANUSCRIPT
Table 2. Medication Changes and Medication Adherence Among Cases and Controls during the Follow-up Period
b
AC C
EP
TE D
M AN U
SC
RI PT
Characteristics Cases Controls p-valuea Medication Possession Ratio at 12 months, mean (SD) All Blood Pressure medications 0.71 (0.19) 0.71 (0.20) 0.89 Alpha 1 Blockers 0.69 (0.30) 0.67 (0.25) 0.78 Alpha 2 Agonists 0.70 (0.21) 0.68 (0.22) 0.66 ACE Inhibitors/Angiotensin Receptor Blockers 0.74 (0.22) 0.71 (0.21) 0.16 Beta Blockers 0.70 (0.22) 0.72 (0.20) 0.46 Calcium Channel Blockers 0.73 (0.22) 0.72 (0.21) 0.56 Thiazide Diuretics 0.71 (0.22) 0.71 (0.21) 0.86 Other (Non-Loop) Diuretics 0.64 (0.23) 0.72 (0.23) 0.07 Others 0.60 (0.17) 0.70 (0.21) 0.31 Proportion of Patients with Medication Possession Ratio >80% at 12 months, No (%) All Blood Pressure Medications 133 (35) 243 (32) 0.69 Alpha 1 Blockers 5 (29) 9 (50) 0.31 Alpha 2 Agonists 16 (30) 39 (33) 0.43 ACE Inhibitors/Angiotensin Receptor Blockers 90 (43) 127 (32) 0.03 Beta Blockers 54 (34) 88 (32) 0.26 Calcium Channel Blockers 67 (41) 100 (36) 0.13 Thiazide Diuretics 56 (37) 85 (34) 0.62 Other (Non-Loop) Diuretics 14 (24) 36 (38) 0.09 Others 1 (11) 2 (22) >0.99 Medication Changes Occurring During 12 month follow-up, No (%)b Alpha 1 Blockers 184 (40) 391 (31) 0.001 Alpha 2 Agonists 276 (59) 643 (51) 0.001 ACE Inhibitors/Angiotensin Receptor Blockers 198 (43) 575 (45) 0.31 Beta Blockers 193 (42) 478 (38) 0.15 Calcium Channel Blockers 172 (37) 379 (30) 0.005 Thiazide Diuretics 186 (40) 510 (40) 0.93 Other (Non-Loop) Diuretics 135 (29) 270 (21) 0.001 Others 26 (6) 44 (3) 0.05 a p-values calculated using t-tests or chi-square as appropriate
Medication Changes refers to any addition of a new blood pressure medication or an increase in the dose of an existing blood pressure medication
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Clinical Significance: •
Systolic blood pressure was lower for cases compared with controls at 6 and 12-
RI PT
months (6-month RR=9.7, 95% CI:2.7-35.3; 12-month RR=20.3, 95% CI:4.1-99.2; p<0.01) •
Diastolic blood pressure was lower at 12-months (RR=2.9, 95% CI:1.2-7.1; p<0.01)
Pharmacist hypertension care management may be an effective method of improving
EP
TE D
M AN U
blood pressure in a medical home model of primary care
AC C
•
SC
but not at 6-months (RR=1.0, 95% CI:0.31-3.4; p=0.9) for cases compared to controls
S0002-9343(14)01208-X
DOI:
10.1016/j.amjmed.2014.11.027
Reference:
AJM 12802
To appear in:
The American Journal of Medicine
Received Date: 17 September 2014 Revised Date:
7 October 2014
Accepted Date: 11 November 2014
Please cite this article as: Zillich AJ, Jaynes HA, Bex SD, Boldt AS, Walston CM, Ramsey DC, Sutherland JM, Bravata DM, Evaluation of Pharmacist Care for Hypertension in the Veterans Affairs Patient Centered Medical Home: A Retrospective Case-Control Study, The American Journal of Medicine (2015), doi: 10.1016/j.amjmed.2014.11.027. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Evaluation of Pharmacist Care for Hypertension in the Veterans Affairs Patient Centered Medical Home: A Retrospective Case-Control Study
RI PT
Alan J. Zillich, PharmD,a,b Heather A. Jaynes, RN, MS,a Susan D. Bex, PharmD,c Amy S. Boldt, PharmD,c Cassandra M. Walston, PharmD,c Darin C. Ramsey, PharmD,c, d Jason M. Sutherland, PhD, Dawn M. Bravata, MD b,f
a
SC
e
Department of Pharmacy Practice, Purdue University College of Pharmacy, Purdue Pharmacy
M AN U
Programs, 5th/3rd Bank Faculty Office Building, 640 Eskenazi Ave, Indianapolis, IN 46202 b
Roudebush VA Medical Center, Health Services Research and Development, Center for Health
Information and Communication Excellence, 1481 West 10th Street, Indianapolis, IN Roudebush VA Medical Center, Department of Pharmacy Services, 1481 West 10th Street,
Indianapolis, IN d
TE D
c
Department of Pharmacy Practice, Butler University College of Pharmacy & Health Sciences, 4600
School of Population and Public Health, Centre for Health Services and Policy Research, University
AC C
e
EP
Sunset Ave, Indianapolis, IN 46208
of British Columbia 201-2206 East Mall Vancouver, British Columbia V6T 1Z3, Canada f
Department of Internal Medicine, Indiana University School of Medicine, 340 West 10th Street #620
Indianapolis, IN 46202 Correspondence:
Alan J. Zillich, PharmD
ACCEPTED MANUSCRIPT 2
Purdue Pharmacy Programs 5th/3rd Bank Building
RI PT
640 Eskenazi Ave Indianapolis, IN 46202 Phone: 317-880-5430
SC
Fax: 317-880-0568
M AN U
Email: [email protected]
Key Words:
Pharmaceutical Care, Pharmacy, Care Management, Disease Management, Hypertension, Adherence,
Running Head:
TE D
Outcomes
AC C
EP
Hypertension Care Management
Conflicts of Interest: All authors confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome. All authors participated in the writing of the manuscript, and have seen and approved the submitted version. Further, each author was involved in the conception and design of the study, and/or the analysis of the data.
ACCEPTED MANUSCRIPT 3
Funding: A portion of Dr. Zillich’s time is supported by a Research Career Development Award from the Veterans Affairs Health Services Research and Development (#RCD 06-304-1). The
RI PT
funding source had no role in any of the following: study design; in the collection, analysis, and
AC C
EP
TE D
M AN U
SC
interpretation of data; in writing the report; and in the decision to submit the article for publication.
ACCEPTED MANUSCRIPT 4
Abstract
RI PT
Purpose: To evaluate a pharmacist hypertension care management program within the patient centered medical home.
Methods: This was a retrospective case-control study. Cases included all hypertension patients
SC
referred to the care management program while controls included hypertension patients who were not referred to the program during the same one-year period. Each case was matched to a maximum of
M AN U
three controls based on: primary care physician, age±5 years, gender, diagnoses of diabetes and kidney disease, baseline systolic blood pressure ±10 mmHg, and number of unique antihypertensive medications. Pharmacists provided a hypertension care management program under an approved scope of practice that allowed pharmacists to meet individually with patients, adjust medications, and
TE D
provide patient education. Primary outcomes were systolic blood pressure and diastolic blood pressure at 6- and 12-months. Multivariate regression models compared each blood pressure endpoint between cases and controls adjusting for age, co-morbidities, baseline blood pressure and baseline
EP
number of blood pressure medications.
Results: 573 patients were referred to the hypertension program; 86% (465/543) had at least one
AC C
matched control and were included as cases in the analyses; 3:1 matching was achieved in 90%(418/465) of cases. At baseline, cases and controls did not differ with respect to age, gender, race, or co-morbidity; baseline blood pressure was higher (139.9/80.0 vs. 136.7/78.2 mmHg, p≤0.0002) in the cases compared to controls. Multivariate regression modeling identified significantly lower systolic blood pressure for the cases compared with controls at both 6 and 12 months (6-month RR=9.7, 95% CI:2.7-35.3; 12-month RR=20.3, 95% CI:4.1-99.2; p<0.01 for both
ACCEPTED MANUSCRIPT 5
comparisons). Diastolic blood pressure was significantly lower at 12-months (RR=2.9, 95% CI:1.27.1; p<0.01) but not at 6-months (RR=1.0, 95% CI:0.31-3.4; p=0.9) for the cases compared to the
RI PT
controls. Conclusions: Patients referred to the pharmacist hypertension care management program had a significant improvement in most blood pressure outcomes and may be an effective method of
AC C
EP
TE D
M AN U
SC
improving blood pressure control among patients in a medical home model of primary care.
ACCEPTED MANUSCRIPT 6
Introduction An interdisciplinary approach is one strategy that significantly and consistently improves blood
RI PT
pressure control.[1-5] Interdisciplinary teams typically include non-physician providers such as
pharmacists and nurses who focus on hypertension care. One meta-analysis on the potency of team care interventions found that interventions involving clinical pharmacists resulted in lower blood
SC
pressure and greater likelihood of achieving blood pressure control.[1] The potency of team care programs has been attributed not only to the specific intervention(s) employed but also the duration of
M AN U
the program, the type of organization, and the design of the trial to measure effectiveness of the program. In another systematic review on the effect of pharmacists as team members, ChisholmBurns found that pharmacist participation significantly improved outcomes for hypertension.[6] In the evolving patient-centered medical home model, the pharmacist’s role as part of the
TE D
interdisciplinary team has been described.[7] Within the patient-centered medical home, referral to a clinical pharmacist for hypertension management is associated with improved patient outcomes and provides primary care clinicians with additional time to focus on other patient priorities.[8] Adoption
EP
and implementation of a patient-centered medical home (or Patient Aligned Care Team-PACT) model
AC C
is ongoing at Veteran Affairs (VA) Medical Centers. The model establishes a group of practitioners, including physicians, nurses, pharmacists, social workers, and additional specialties to partner with patients for a holistic and efficient primary care experience. This interdisciplinary model unites the healthcare team and improves outcomes by focusing on patient-centered care, but the effect of the pharmacist within the VA patient aligned care team model is largely unknown. The purpose of the present study was to evaluate the effectiveness of pharmacists as part of the patient aligned care team for veterans with hypertension.
ACCEPTED MANUSCRIPT 7
Methods
RI PT
Study Setting This study was conducted at an urban Midwest VA Medical Center (VAMC) that provides inpatient and outpatient health care services to almost 200,000 veterans. Within primary care, services are provided across multiple clinics. Within each clinic, PACT teamlets provide direct patient care; each
SC
teamlet consisting of the patient, their clinician (usually a physician but in some cases an advanced
M AN U
practice nurse), the nurse case manager (RN), and the health technician (LPN). Clinical pharmacists, social workers, health psychologists and support staff are designated as members of several teamlets (and are shared across teamlets within a given clinic). For this study, a care management program was evaluated for patients with hypertension provided by clinical pharmacists in four primary care clinics
TE D
and their associated PACT teamlets. Study Design
This was a retrospective case-control study using electronic medical record data to evaluate the
EP
effectiveness of the pharmacist care management program. The study was approved by the Indiana University/Purdue University-Indianapolis Institutional Review Boards and the VA Research and
AC C
Development Committee. Study Population
Patients with hypertension who were referred to the care management program during a 1-year evaluation were included as possible cases. Controls included hypertension patients, in the same primary care clinics, who were not referred to the program during the same one-year period. All patients were followed for 12 months.
ACCEPTED MANUSCRIPT 8
Care Management Program The hypertension care management program was provided by six clinical pharmacists within four
RI PT
primary care clinics throughout the duration of the evaluation period. Two clinics had one assigned pharmacist and the other two clinics had two pharmacists, based on the larger patient volume of those clinics. All pharmacists had completed a Doctor of Pharmacy (PharmD) program and at least one
SC
year of clinical pharmacy residency training. Several pharmacists had also obtained Board
Certification in Pharmacotherapy (BCPS). The pharmacists operated under a scope of practice
M AN U
allowing them to meet individually with patients, provide patient education, and initiate, change, and discontinue medications.
An electronic form was used by primary care providers to refer patients with hypertension to the pharmacists. During the referral process, many primary care providers would “curbside” the
TE D
pharmacists to discuss initial therapy options. Following referral to the program, patients were scheduled for an initial face-to-face visit with the pharmacist. Typically, the initial visit was scheduled to occur within 30 days of referral and occurred on the same day as the patient’s regular
EP
primary care clinic visit. The pharmacists met with the patient for 15-30 minutes and measured blood
AC C
pressure using standard, calibrated devices which were available to all providers in the clinic. Blood pressure measurements were conducted using the American Heart Association guidelines.[9]
As part of the pharmacists’ scope of practice, treatment decisions were based on national clinical practice guidelines from the JNC 7 and the VA/Department of Defense (which were the current guidelines at the time of this evaluation) using a stepwise approach to medication management.[10, 11] Thiazide diuretics were preferred first line antihypertensive medications with considerations for
ACCEPTED MANUSCRIPT 9
the use of other medications based on co-morbid disease states. In general, medication dosages were maximized before additional medications were added to the regimen. Blood pressure goals were
RI PT
established according to guidelines; patients with diabetes and chronic kidney disease had a systolic blood pressure goal less than 130 mm Hg and a diastolic blood pressure goal less than 80 mm Hg. All other patients had a systolic blood pressure goal less than 140 mm Hg and diastolic blood
SC
pressure goal less than 90 mm Hg.
M AN U
Patients were discharged from the program once blood pressure goals were attained at either a pharmacist visit or during a primary care provider visit. However, if the patients’ blood pressure became uncontrolled at any subsequent primary care visits, the provider could re-refer the patient back to the pharmacist for additional management. This referral, discharge back to primary care and re-referral process was designed to maximize the pharmacists’ panel size with those patients who
TE D
needed intensive treatment while maintaining the primary care providers’ role in the overall care of their patients. Patients could also be discharged from the program after two consecutive no-show
EP
visits.
AC C
Data Collection, Outcomes and Analyses
Retrospective data was extracted from the VA electronic medical record. To be included in the cohort, all unique patients with a diagnosis of hypertension (ICD-9 code: 401.1X-404.1X) and one or more primary care visits during a 1-year care management evaluation period were identified. Patients who died during the evaluation period were excluded (Figure 1). For all eligible patients, data were obtained for 1-year before the intervention, the 1-year period during the intervention evaluation and 1-year after the intervention period. The following data elements were extracted during the 1 year
ACCEPTED MANUSCRIPT 10
period prior to the evaluation: gender, race, co-morbid conditions based on outpatient ICD-9 codes (including those for diabetes and chronic kidney disease), primary care clinic assignment, primary
RI PT
care provider assignment, and the number of unique blood pressure medications. Additional data elements were extracted for the 1 year evaluation period and the 1 year follow-up period: dates of all primary care visits, blood pressure values (both systolic and diastolic) at each primary care visit, as
SC
well as blood pressure medication data (drug name, quantity, days supply, and dates of refills).
All patients cared for in the pharmacist hypertension care management program were considered to
M AN U
be potential cases. All remaining patients in the cohort were identified as potential controls and available for matching (Figure 1). Each case was matched to a maximum of three controls based on: primary care provider and clinic, age ± 5 years, gender, diagnoses of diabetes and chronic kidney disease, baseline systolic blood pressure ±10 mmHg defined as the average of the previous 3 blood
TE D
pressures, and number of unique blood pressure medications at baseline. All criteria had to be met in order to be considered a match. All cases with one or more matched controls were included in the analyses.
EP
For cases, baseline was defined as the date of referral to care management program. For controls,
AC C
patients were allocated to a baseline date that coincided with the proportion of matched cases with a baseline date during each month of the 1-year intervention period. For example, if 10% of the cases had a baseline date in June, then all 10% of those matched controls were assigned a baseline date in the month of June. All cases and matched controls were followed for 12 months from baseline. The primary outcomes were differences in systolic and diastolic blood pressure at 6- and 12-month follow-up in the cases compared to the controls. For each time point, the average of all primary care blood pressure values during that time period was used for the outcome assessment. Secondary
ACCEPTED MANUSCRIPT 11
outcomes included the number of patients reaching national guideline-define blood pressure goals (both systolic and diastolic blood pressure) at 6-and 12-months.
RI PT
Additional analyses explored the number and type of medication changes, defined as the addition of a new blood pressure medication or increasing the dose of an existing blood pressure medication, occurring in the cases compared to the controls during the 12-months of follow-up. To explore
SC
medication adherence at 12-months, the medication possession ratio (MPR) was calculated based on the method described by Steiner and colleagues as a continuous, multiple-interval measure of
M AN U
medication availability (CMA).[12] To account for medication switches and concurrent therapy, an average of all blood pressure drugs’ MPRs within a therapeutic class was computed to produce one averaged MPR for each class and an overall MPR for all blood pressure medications; the average MPR was then dichotomized with poor adherence defined as <0.8 and good adherence defined as
TE D
>0.8.[13]
Descriptive statistics were calculated to characterize the baseline demographics of both groups. Multivariate linear regression models compared each blood pressure endpoint at 6- and 12-months
EP
between the cases and controls adjusting for age, co-morbidities, baseline blood pressure and baseline
AC C
number of blood pressure medications. Logistic regression models, using similar covariates, evaluated achievement of blood pressure goals. Given that missing data for the blood pressure outcomes measurement was not random and dependent on factors such as clinic follow-up, a noncompletely-at-random (NMAR) multiple imputation method was used for the blood pressure values.[14, 15] Data were analyzed using SAS and R, with all p-values less than 0.05 considered to be statistically significant.
ACCEPTED MANUSCRIPT 12
Results Eighty six percent (465/543) of program participants had at least one matched control and were
RI PT
included in the analyses as cases (Figure 1). Three-to-one matching was achieved in 90% (418/465) and 18 (4%) patients had a 2-1 match and 29 (6%) patients had a 1-1 match. Cases and matched
controls all came from the same matched physicians and clinics and did not differ with respect to age,
SC
race, gender, or co-morbidities (Table 1). Baseline blood pressure was higher (139.9/80.0 vs.
136.7/78.2 mmHg; p≤0.0002) and baseline blood pressure control lower (35% vs. 49%, p<0.0001) in
M AN U
the cases compared to controls. Among cases, systolic blood pressure decreased from baseline by -4.0 and -7.1 mmHg at 6- and 12-months while systolic blood pressure decreased among controls by -1.6 and -2.6 mmHg at 6- and 12-months. For diastolic blood pressure, the cases decreased from baseline by -2.5 and -3.2 mmHg at 6- and 12-months while in the controls, diastolic blood pressure decreased
TE D
from baseline by -1.1 and -1.2 mmHg at 6- and 12-months.
Controlling for baseline covariates, multivariate regression modeling identified significantly lower systolic blood pressure for the cases compared with controls at both 6- and 12-months (6 month
EP
RR=9.7, 95% CI:2.7-35.3; 12 month RR=20.3, 95% CI:4.1-99.2; p<0.01 for both comparisons).
AC C
Diastolic blood pressure was significantly lower at 12-months (RR=2.9, 95% CI:1.2-7.1; p<0.01) but not at 6-months (RR=1.0, 95% CI:0.31-3.4; p=0.9) for the cases compared to the controls. Blood pressure control was better among cases than controls at 12-months (64% vs. 60%; OR=1.3, 95% CI:1.1-1.6; p<0.01) but not at 6-months (57% vs. 57%; OR=1.1, 95% CI:0.9-1.3; p=0.28). The MPR for all blood pressure medications and the proportion of patients with an MPR >80% did not differ between cases and controls at 12 months (Table 2). There were more medications added or
ACCEPTED MANUSCRIPT 13
doses increased in the cases compared to the controls for alpha-1 blockers, alpha-2 agonists, calcium channel blockers, and non-loop diuretics (Table 2).
RI PT
Discussion
Findings from this and other studies support the expanded role of the pharmacists as part of the
patient-centered medical home to improve the hypertension and other cardiovascular outcomes of
SC
patients.[8, 16] Indeed, in this VA patient-centered medical home model, pharmacists are able to
M AN U
prescribe and discontinue hypertension medications without direct oversight from the primary care physician, which has been identified as a driver of better blood pressure outcomes.[2] The results of this case-control study add to evidence from two recently published systematic reviews of team-based care for hypertension. In one review of 14 studies, pharmacist participation in hypertension teambased care decreased systolic blood pressure by 7.8 mmHg (95% CI, -9.7 to -5.8) and diastolic blood
TE D
pressure by 2.9 mmHg (95% CI, -3.8 to -2.0) [6] while the other review of 13 studies found that pharmacists’ participation lowered median systolic blood pressure by 5.0 mmHg and diastolic blood pressure by 1.7 mmHg.[2] Results from our study found similar, clinically meaningful results;
EP
lowering systolic blood pressure by 4.0 and 7.1 mmHg and diastolic blood pressure was by 2.5 and
AC C
3.2 mmHg at 6 and 12 months, respectively. This current study provides data on potential mechanisms through which these advantageous blood pressure outcomes are achieved. Our data suggest that the effect of the pharmacists’ blood pressure improvement is not due to improvement in medication adherence. Overall medication adherence at 12-months, as measured by the MPR, was 0.71 in both the cases and controls (p=0.89). Also, the proportion of patients with good antihypertensive medication adherence (MPR >0.8) was 35% for the cases and 32% for the controls (p=0.69). However, the number of changes for several classes of
ACCEPTED MANUSCRIPT 14
blood pressure medications, inclusive of newly added blood pressure medications or increases in existing blood pressure medication doses, occurred significantly more often in the cases compared to
RI PT
the controls. For example, the use of calcium channel blockers increased from 24% at baseline to 37% during follow-up in the cases compared to an increase from 27% to 30% in the controls
(p=0.005 for the follow-up comparison). These data reflect, to some extent, the primary role of the
SC
pharmacist in this patient-centered medical home model for hypertension care. Pharmacists were tasked with changing the medication regimen to maximize effectiveness, balancing the efficacy and
M AN U
safety of the medications with the preferences of the patient. While specific data on medication adverse effects was not collected, it is possible that the hypertension drug regimens initiated by the pharmacists were more potent while avoiding or mitigating adverse effects. Future studies of pharmacist care models should capture data on outcomes related to identification and mitigation of
TE D
adverse effects.
In addition, pharmacists in the patient-centered medical home model serve as “physician-extenders”, allowing the team to provide effective and efficient care while improving the Veterans’ access to
EP
primary care providers. Our team evaluated a similar patient-centered medical home model within VA for dyslipidemia and found that time to achieve LDL cholesterol control was 80% faster in the
AC C
pharmacist care group compared to a control group (p=0.002).[8] Although time to hypertension control was not examined in the current study, the results suggest that referral of patients to a pharmacist for management of hypertension is an efficient allocation of primary care provider resources, allowing the physician additional time to focus on other, more complex, priorities for the patient. Future studies should examine the time and cost savings associated with this patient-centered medical home model.
ACCEPTED MANUSCRIPT 15
The study findings should be considered with several limitations. First, the intermediate endpoint of blood pressure was the primary outcome for this study. Major endpoints such as cardiovascular
RI PT
related morbidity and mortality were not examined in this internally funded, single-site study. Similarly, other patient-reported endpoints such as medication adverse events and patient satisfaction were not collected. Nevertheless, the magnitude of decrease in systolic blood pressure is clinically
SC
meaningful, where evidence from clinical trials suggests that these blood pressure changes would result in lower morbidity and mortality.[11] Additionally, the results found significantly more
M AN U
changes in blood pressure medications for the cases compared to controls but no significant improvements in medication adherence. It is possible that improvements in medication adherence could have occurred but clinical nuances could not be detected using refill claims records and the MPR. In fact, pharmacists could have increased the dose of certain blood pressure medications by verbally informing the patient to double the dose and updating the medication list in the chart but
TE D
without ordering a new prescription for filling or instruct patients to stop other medications that were already dispensed. Therefore, measurement of adherence using the MPR may not be sensitive to
EP
small improvements. Finally, while our retrospective case-control study was robust relative to prepost study designs, the matching process for controls was incomplete. A stronger design such as a
work.
AC C
prospective case-control or randomized study would be ideal but was not possible for this unfunded
Conclusion
Patients referred to the pharmacist hypertension care management program had a significant improvement in blood pressure. This program may be an effective method of improving blood pressure control among Veterans in the patient-centered medical home model.
ACCEPTED MANUSCRIPT 16
RI PT
Acknowledgements Prior Presentations: This work was presented in abstract, poster, and podium format at the American College of Clinical Pharmacy Annual Meeting, Hollywood, FL October 20th-24th, 2012 and at the VA
AC C
EP
TE D
M AN U
SC
Health Services Research and Development Annual Meeting, Washington, DC July 16th-19th, 2012.
ACCEPTED MANUSCRIPT 17
References
6. 7. 8. 9. 10.
11.
12. 13.
14. 15. 16.
RI PT
SC
5.
M AN U
4.
TE D
3.
EP
2.
Carter, B.L., M. Rogers, J. Daly, S. Zheng, and P.A. James, The potency of team-based care interventions for hypertension: a meta-analysis. Arch Intern Med, 2009. 169(19): p. 1748-55. Proia, K.K., A.B. Thota, G.J. Njie, et al., Team-based care and improved blood pressure control: A community guide systematic review. American Journal of Preventive Medicine, 2014. 47(1): p. 86-99. Santschi, V., A. Chiolero, B. Burnand, A.L. Colosimo, and G. Paradis, Impact of pharmacist care in the management of cardiovascular disease risk factors: a systematic review and meta-analysis of randomized trials. Arch Intern Med, 2011. 171(16): p. 1441-53. Walsh, J.M., K.M. McDonald, K.G. Shojania, et al., Quality improvement strategies for hypertension management: a systematic review. Med Care, 2006. 44(7): p. 646-57. Walsh, J.M., V. Sundaram, K. McDonald, D.K. Owens, and M.K. Goldstein, Implementing effective hypertension quality improvement strategies: barriers and potential solutions. J Clin Hypertens, 2008. 10(4): p. 311-6. Chisholm-Burns, M.A., J. Kim Lee, C.A. Spivey, et al., US pharmacists' effect as team members on patient care: systematic review and meta-analyses. Med Care, 2010. 48(10): p. 923-33. Smith, M., D.W. Bates, T. Bodenheimer, and P.D. Cleary, Why pharmacists belong in the medical home. Health Aff, 2010. 29(5): p. 906-13. Smith, M.C., A.S. Boldt, C.M. Walston, and A.J. Zillich, Effectiveness of a Pharmacy Care Management Program for Veterans with Dyslipidemia. Pharmacotherapy, 2013. 33(7): p. 736-743. American Heart Association. Heart Disease and Stroke Statistics-2004 Update. Dallas, TX: American Heart Association; 2004. VA/DoD Clinical Practice Guideline for the Management of Hypertension in Primary Care. 2004 [cited 2010 June 25]; Available from: http://www.healthquality.va.gov/Hypertension_Clinical_Practice_Guideline.asp. Chobanian, A.V., G.L. Bakris, H.R. Black, et al., Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension, 2003. 42(6): p. 1206-52. Steiner, J.F. and A.V. Prochazka, The assessment of refill compliance using pharmacy records: methods, validity, and applications. J Clin Epidemiol, 1997. 50(1): p. 105-16. Peterson, A.M., D.P. Nau, J.A. Cramer, J. Benner, F. Gwadry-Sridhar, and M. Nichol, A checklist for medication compliance and persistence studies using retrospective databases. Value Health, 2007. 10(1): p. 3-12. Graham, J.W., Missing data analysis: making it work in the real world. Annu Rev Psychol, 2009. 60: p. 549-76. He, Y., Missing data analysis using multiple imputation: getting to the heart of the matter. Circ Cardiovasc Qual Outcomes, 2010. 3(1): p. 98-105. Bex, S.D., A.S. Boldt, S.B. Needham, et al., Effectiveness of a Hypertension Care Management Program Provided by Clinical Pharmacists for Veterans. Pharmacotherapy, 2011. 31(1): p. 31-38.
AC C
1.
ACCEPTED MANUSCRIPT 18
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1. Study Cohort Flow Diagram
ACCEPTED MANUSCRIPT
Table 1. Baseline Characteristics of the Cases and Controls
SC
RI PT
Controls (n=1,268) 1,268 (100) 63.2 + 11.5 451 (35) 175 (14) 642 (51) 433 (34) 32 (3) 2.5 + 1.2
136.7 + 10.6 78.2 + 8.7 617 (49)
b
EP
TE D
M AN U
Cases Characteristics (n=465) Male, No (%) 465 (100) Age, mean (SD), y 62.6 + 11.9 White, No (%) 148 (31) Black, No (%) 67 (15) Other/Unknown, No (%) 250 (54) Diabetes, No (%) 160 (34) Kidney Disease, No (%) 15 (3) Number of Co-morbidities, mean (SD)a 2.6 + 1.2 Baseline Systolic Blood Pressure, mean (SD), mm Hgb 139.9 + 11.0 Baseline Diastolic Blood Pressure, mean (SD), mm Hgb 80.0 + 9.3 b Baseline Blood Pressure Control, No (%) 164 (35) Baseline Number of Blood Pressure medications, mean (SD)b 3.3 + 1.6 Baseline Medication Classes Alpha 1 Blockers, No (%) 14 (3) Alpha 2 Agonists, No (%) 96 (21) ACE Inhibitors/Angiotensin Receptor Blockers, No (%) 294 (63) Beta Blockers, No (%) 169 (36) Calcium Channel Blockers, No (%) 110 (24) Thiazide Diuretics, No (%) 184 (40) Other (Non-Loop) Diuretics, No (%) 61 (13) Others, No (%) 10 (2) a Calculated from the Elixhauser co-morbidity index
2.9 + 1.4 38 (3) 282 (22)
814 (64) 499 (39) 340 (27) 528 (42) 193 (15) 23 (2)
AC C
Statistically significant difference between cases and controls using t-tests or chi-square as appropriate (p < 0.05)
ACCEPTED MANUSCRIPT
Table 2. Medication Changes and Medication Adherence Among Cases and Controls during the Follow-up Period
b
AC C
EP
TE D
M AN U
SC
RI PT
Characteristics Cases Controls p-valuea Medication Possession Ratio at 12 months, mean (SD) All Blood Pressure medications 0.71 (0.19) 0.71 (0.20) 0.89 Alpha 1 Blockers 0.69 (0.30) 0.67 (0.25) 0.78 Alpha 2 Agonists 0.70 (0.21) 0.68 (0.22) 0.66 ACE Inhibitors/Angiotensin Receptor Blockers 0.74 (0.22) 0.71 (0.21) 0.16 Beta Blockers 0.70 (0.22) 0.72 (0.20) 0.46 Calcium Channel Blockers 0.73 (0.22) 0.72 (0.21) 0.56 Thiazide Diuretics 0.71 (0.22) 0.71 (0.21) 0.86 Other (Non-Loop) Diuretics 0.64 (0.23) 0.72 (0.23) 0.07 Others 0.60 (0.17) 0.70 (0.21) 0.31 Proportion of Patients with Medication Possession Ratio >80% at 12 months, No (%) All Blood Pressure Medications 133 (35) 243 (32) 0.69 Alpha 1 Blockers 5 (29) 9 (50) 0.31 Alpha 2 Agonists 16 (30) 39 (33) 0.43 ACE Inhibitors/Angiotensin Receptor Blockers 90 (43) 127 (32) 0.03 Beta Blockers 54 (34) 88 (32) 0.26 Calcium Channel Blockers 67 (41) 100 (36) 0.13 Thiazide Diuretics 56 (37) 85 (34) 0.62 Other (Non-Loop) Diuretics 14 (24) 36 (38) 0.09 Others 1 (11) 2 (22) >0.99 Medication Changes Occurring During 12 month follow-up, No (%)b Alpha 1 Blockers 184 (40) 391 (31) 0.001 Alpha 2 Agonists 276 (59) 643 (51) 0.001 ACE Inhibitors/Angiotensin Receptor Blockers 198 (43) 575 (45) 0.31 Beta Blockers 193 (42) 478 (38) 0.15 Calcium Channel Blockers 172 (37) 379 (30) 0.005 Thiazide Diuretics 186 (40) 510 (40) 0.93 Other (Non-Loop) Diuretics 135 (29) 270 (21) 0.001 Others 26 (6) 44 (3) 0.05 a p-values calculated using t-tests or chi-square as appropriate
Medication Changes refers to any addition of a new blood pressure medication or an increase in the dose of an existing blood pressure medication
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Clinical Significance: •
Systolic blood pressure was lower for cases compared with controls at 6 and 12-
RI PT
months (6-month RR=9.7, 95% CI:2.7-35.3; 12-month RR=20.3, 95% CI:4.1-99.2; p<0.01) •
Diastolic blood pressure was lower at 12-months (RR=2.9, 95% CI:1.2-7.1; p<0.01)
Pharmacist hypertension care management may be an effective method of improving
EP
TE D
M AN U
blood pressure in a medical home model of primary care
AC C
•
SC
but not at 6-months (RR=1.0, 95% CI:0.31-3.4; p=0.9) for cases compared to controls