Professional Documents
Culture Documents
Background: Team-based care is the strategy that has ment algorithm (−4.00 mm Hg). The odds ratios (95% con-
had the greatest effect on improving blood pressure (BP). fidence intervals) for controlled BP were: nurses, 1.69 (1.48-
The purpose of this systematic review was to determine 1.93); pharmacists within primary care clinics, 2.17 (1.75-
the potency of interventions for BP involving nurses or 2.68); and community pharmacists, 2.89 (1.83-4.55). Mean
pharmacists. (SD) reductions in SBP were: nursing studies,5.84(8.05)
mm Hg; pharmacists in clinics,7.76(7.81) mm Hg; and
Methods: A MEDLINE search for controlled clinical trials
community pharmacists, 9.31(5.00) mm Hg. There were
that involved a nurse or pharmacist intervention was con-
no significant differences between the nursing and phar-
ducted. Mean reductions in systolic (S) and diastolic (D)
macy studies (Pⱖ.19).
BP were determined by 2 reviewers who independently
abstracted data and classified the different intervention
Conclusions: Team-based care was associated with im-
components.
proved BP control, and individual components of the in-
Results: Thirty-seven articles met the inclusion criteria. tervention appeared to predict potency. Implementation
Education about BP medications was significantly asso- of new hypertension guidelines should consider changes
ciated with a reduction in mean BP (−8.75/−3.60 mm Hg). in health care organizational structure to include impor-
Other strategies that had large effect sizes on SBP include tant components of team-based care.
pharmacist treatment recommendations (−9.30 mm Hg),
intervention by nurses (−4.80 mm Hg), and use of a treat- Arch Intern Med. 2009;169(19):1748-1755
B
L O O D P R E S S U R E (BP) I S pose of the present study was to conduct a
poorly controlled in the systematic review of the research litera-
United States. 1-5 The 8th ture and to evaluate the potency of team-
Joint National Committee based care involving pharmacists or nurses.
on Prevention, Detection, We theorized that the effect size would be
Evaluation and Treatment of High Blood greater for nurses or pharmacists working
Pressure ( JNC-8) is currently consider- in a physician’s office or more indepen-
ing strategies to improve the implemen- dently by protocol than with more distant
tation of the guidelines and achieve higher interventions, such as recommendations
BP control rates. Investigators from the from a community pharmacist.
Stanford University/University of Califor-
nia, San Francisco, Evidence-Based Prac- METHODS
tice Center conducted an analysis of con-
trolled clinical trials examining quality We followed the same process as Walsh et al6
improvement strategies and found that the by including quasi-randomized trials, con-
only strategy that significantly improved trolled before-after studies, interrupted time-
Author Affiliations: BP involved interdisciplinary, team- series studies, patient-randomized trials, and
Department of Pharmacy based care.6 Most of the quality improve- cluster-randomized trials. Quasi-randomized
Practice and Science, College of ment interventions included multiple com- trials were defined as those that included at least
Pharmacy (Drs Carter and 2 patient cohorts identified prospectively using
ponents. These different strategies or the
Rogers), and Department of an arbitrary but nonrandom allocation proce-
Family Medicine, Roy J. and
potency of the intervention may explain dure.6 Controlled before-after studies were de-
Lucille A. Carver College of the apparent differences in effect sizes.7 fined as those with contemporaneous observa-
Medicine (Drs Carter, Daly, One strategy to improve guideline ad- tion of cohorts that differed primarily with
Zheng, and James), The herence is to use team-based care involv- respect to exposure to the intervention.6 Inter-
University of Iowa, Iowa City. ing pharmacists or nurses.8-13 The pur- rupted time series required that the study re-
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1748
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1749
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1750
tended those of the previous report6 pants in the physicians’ office.20 We tion, and potency of the interven-
that found involving pharmacists or could also have classified this study tion. For instance, Carter et al8,21,23
nurses was the most potent quality as “pharmacist in the clinic,” which conducted 3 studies in community
improvement strategy to improve BP would have reduced the OR for com- pharmacies, where the pharmacists
control. We also wanted to deter- munity pharmacy studies and in- had no prior established relation-
mine whether specific aspects of team creased the OR for studies involv- ship with the physicians and the in-
care were more potent. Our analysis ing pharmacists in clinics. Second, terventions were only 4 and 5
found that studies involving phar- we classified one study as a nursing months in length. These studies had
macists resulted in not only lower BP intervention for the OR calcula- modest ORs for controlled BP (1.56,
but also a greater OR of achieving BP tions, but the intervention in- 1.74, and 2.46). Carter et al52 re-
control compared with studies in- volved both a nurse and a commu- cently completed a randomized,
volving nurses. However, the reduc- nity pharmacist (OR, 1.79). 1 9 controlled effectiveness (ie, prag-
tions in SBP and CIs for controlled Excluding the first 2 studies and add- matic) study of a 6-month pharma-
BP overlap for the different health ing the third study to the analysis of cist intervention among 402 pa-
care providers (Figure 2). community pharmacy studies would tients from 6 family medicine clinics
We had hypothesized that stud- have resulted in an OR closer to 1.8 that was not included in this sys-
ies involving community pharma- for the community pharmacy group. tematic review because it had not
cists would be less potent than those Finally, one large study was con- been published at the time of our
involving nurses or pharmacists ducted within a managed care orga- evaluation. In that study, SBP was re-
within primary care clinics. Of in- nization that involved education by duced by 12.0 mm Hg more in the
terest, studies involving commu- a pharmacist via the Internet.35 We intervention group than the con-
nity pharmacists had the highest OR classified this study as one within trol group, and the OR for con-
(2.89). These findings may be based primary care, but the effect was not trolled BP was 3.2 (95% CI, 2.0-
on how the reviewers categorized the as great (OR,1.88) compared with 5.1). Finally, these investigators
studies. First, one study conducted studies in which the pharmacist conducted an efficacy study in which
in community pharmacies in Por- adjusted therapy either alone or in BP was controlled in 54% of pa-
tugal had an extremely high OR collaboration with physicians tients in the control group and 89%
(29.71).22 Another study in a com- (ORs, 7.38-9.98). Without that in the intervention group (OR, 7.38;
munity pharmacy had an OR of 4.29, study, the OR would have been 3.27 95% CI, 3.43-15.91).30 The main rea-
but this pharmacist worked closely for pharmacists in clinics. son for high BP control in this lat-
with 2 physicians and reviewed It may be possible to explain our ter study was attributed to asser-
medical records of study partici- findings based on the dose, dura- tive and frequent medication
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1751
B
Mckenny et al,20 1973 4.29 (0.99-18.59)
Park et al,21 1996 2.46 (0.76-7.89)
Carter et al,8 1997 1.56 (0.50-4.90)
Garcao et al,22 2002 29.71 (6.89-128.19)
Zillich et al,23 2005 1.74 (0.83-3.66)
C
Schneider et al,24 1982 3.54 (0.78-16.03)
McGhan et al,25 1983 8.4 (0.97-72.95)
Erickson et al,26 1997 1.91 (0.76-4.79)
Bogden et al,27 1998 5.05 (2.01-12.67)
Mehos et al,34 2000 2.80 (0.66-11.92)
Vivian et al,28 2002 9.98 (2.78-35.81)
Borenstein et al,29 2003 1.97 (1.12-3.47)
Green et al,35 2008 1.88 (1.36-2.60)
Carter et al,30 2008 7.38 (3.43-15.91)
Figure 2. The odds ratio (OR) (confidence interval [CI]) that systolic blood pressure is controlled in the intervention group compared with the control group. A
higher OR indicates a more effective intervention. A, Eight studies involving nurses. B, Five studies conducted in community pharmacies. C, Nine studies involving
pharmacists in primary care clinics.
intensification recommended by the tion in which the intervention is per- laboratively with physician col-
pharmacist. Therefore, the ORs for formed (home, work site, commu- leagues and/or provided more
the 5 studies by these investigators nity pharmacy, or primary care autonomous care. Pharmacists
were: community pharmacy stud- clinic), and whether the study is an within primary care clinics work
ies (BS-trained pharmacists), be- efficacy or effectiveness trial. These closely with physicians, and the ex-
tween 1.56 and 2.46; the pragmatic factors, as well as the intervention pected levels of trust and coopera-
trial of clinical pharmacists (PharmD procedures, predict the potency of tion might be higher than with com-
with residency or fellowship), 3.20; the intervention. munity pharmacists, for whom
and the efficacy trial (ideal interven- Studies involving community interaction is usually not in person
tion delivery) with clinical pharma- pharmacists largely involved mak- and occurs from distant loca-
cists (PharmD with residency), 7.38. ing recommendations to physi- tions.23,51,53 In fact, recommenda-
Therefore, when the literature in- cians by telephone or facsimile. Stud- tions to change BP medications were
volving team care is evaluated, it is ies involving pharmacists in clinics accepted 95% of the time from phar-
critical to assess the duration of the typically involved pharmacists em- macists within the same clinic30 but
intervention, the type of organiza- ployed in the clinic who worked col- only 45% to 50% when recommen-
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1752
SE
0.40
educational background, or train-
ing, but 4 studies used either RNs or 0.30
nurse practitioners.16,18,42,49 Nursing
interventions seemed more likely to 0.20
involve home visits, use of a treat-
ment algorithm, and patient engage- 0.10
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1753
This evaluation of team-based care tration Health Services Research and Evaluation, and Treatment of High Blood Pres-
Development Service, US Depart- sure; National Heart, Lung, and Blood Institute;
for hypertension found that inter- National High Blood Pressure Education Pro-
ventions involving nurses or phar- ment of Veterans Affairs (Dr Carter). gram Coordinating Committee. Seventh report of
macists are effective strategies to im- Disclaimer: The views expressed in the Joint National Committee on Prevention, De-
prove BP control. Several individual this article are those of the authors tection, Evaluation, and Treatment of High Blood
components were associated with and do not necessarily reflect the po- Pressure. Hypertension. 2003;42(6):1206-1252.
sition or policy of the Department of 15. McClellan WM, Craxton LC. Improved follow-up care
improvements in BP. Research in- of hypertensive patients by a nurse practitioner in
volving team-based care must be Veterans Affairs. a rural clinic. J Rural Health. 1985;1(2):34-41.
carefully designed, reported, and Additional Information: The 16. Curzio JL, Rubin PC, Kennedy SS, Reid JL. A com-
interpreted to include the organiza- eTable is available at http://www parison of the management of hypertensive pa-
.archinternmed.com. tients by nurse practitioners compared with con-
tional structure in which the inter- ventional hospital care. J Hum Hypertens. 1990;
vention is performed, the educa- 4(6):665-670.
tional level and training of the 17. Garcia-Peña C, Thorogood M, Armstrong B, Reyes-
REFERENCES Frausto S, Munoz O. Pragmatic randomized trial
intervention providers, and the in-
of home visits by a nurse to elderly people with
dividual components of the inter- 1. Garg JP, Elliott WJ, Folker A, Izhar M, Black HR; hypertension in Mexico. Int J Epidemiol. 2001;
vention so that similar interven- RUSH University Hypertension Service. 30(6):1485-1491.
tions can be implemented within a Resistant hypertension revisited: a comparison of 18. Hill MN, Han HR, Dennison CR, et al. Hyperten-
given health system. two university-based cohorts. Am J Hypertens. sion care and control in underserved urban Afri-
2005;18(5, pt 1):619-626. can American men: behavioral and physiologic out-
2. Berlowitz DR, Ash AS, Hickey EC, et al. Inad- comes at 36 months. Am J Hypertens. 2003;
Accepted for Publication: June 26, equate management of blood pressure in a hy- 16(11, pt 1):906-913.
2009. pertensive population. N Engl J Med. 1998;339 19. McLean DL, McAlister FA, Johnson JA, et al; SCRIP-
(27):1957-1963. HTN Investigators. A randomized trial of the effect
Correspondence: Barry L. Carter, 3. Oliveria SA, Lapuerta P, McCarthy BD, L’Italien of community pharmacist and nurse care on im-
PharmD, Department of Pharmacy GJ, Berlowitz DR, Asch SM. Physician-related bar- proving blood pressure management in patients
Practice and Science, Room 527, riers to the effective management of uncon- with diabetes mellitus: study of cardiovascular risk
College of Pharmacy, The Univer- trolled hypertension. Arch Intern Med. 2002; intervention by pharmacists-hypertension
162(4):413-420. (SCRIP-HTN). Arch Intern Med. 2008;168(21):
4. Hyman DJ, Pavlik VN. Characteristics of patients 2355-2361.
‡References 16, 17, 19, 22, 31-33, 39, with uncontrolled hypertension in the United 20. McKenney JM, Slining JM, Henderson HR, Devins
43-45, 47. States. N Engl J Med. 2001;345(7):479-486. D, Barr M. The effect of clinical pharmacy ser-
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1754
(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 19), OCT 26, 2009 WWW.ARCHINTERNMED.COM
1755