The Impact of a Pharmacotherapy Consultation on the Cost and Outcome of Medical Therapy John Jameson, PharmD; Glenn VanNoord, MD; and Karen Vanderwoud, PharmD fy Rapids and Grand Rapids, Michigan Uckground- One important task for physicians is to op­ ter the consultation, the number of drugs, the number timize their patients’ medication regimen. Involvement o f doses, and the 6-month drug costs all decreased in of clinical pharmacists who have specific training in drug the consult group and increased in the control group; regimen design has been associated with improved pa­ the net difference was 1.1 drugs (P = .0 0 4 ), 2.15 doses tient outcomes for specific medical conditions, eg, hyper­ per day (P = .0 0 7 ), $586 per year (P = .008). The side ef­ tension and anticoagulation. This prospective, randomized fects score improved by 1.8 points more in the consult trial investigated whether a single consultation by a clinical group compared with the control group (P=NS). Simi­ pharmacist with high-risk patients and their primary physi­ larly, the prescribing convenience score in the consult cians would result in improved prescribing outcomes. group improved by 1.4 points more than that o f the control group (P=NS). Methods. Patients at risk for medication-related prob­ lems were identified and randomized to receive a phar­ Conclusions. This study demonstrates several important macotherapy consultation (consult group) or usual benefits o f integration o f a clinical pharmacist into a pri­ medical care (control group). Outcomes, including the mary care setting, including improvement in cost and number o f drugs, number o f doses per day, cost of simplification of the medication regimen with no reduc­ medications, and patient reports of adverse effects, were tion in quality o f care. recorded at baseline and at 6 months following the in- j tervention. Key words. Pharmacists; clinical pharmacist; drug thera­ Results. Fifty-six subjects were evaluable: 29 in the con­ py; family practice; cost. trol group, and 27 in the consult group. Six months af­ (/ Fam Pract 1995; 41:469-472) Among the most critical tasks of primary care physicians is venient. Both drug-related hospital admissions and com­ I to optimize their patients’ drug regimens. An optimal pliance are associated with the number o f medications drug regimen is one that produces the desired benefits prescribed.1-2 while minimizing the number o f drugs, doses per day, The clinical pharmacist is a potentially valuable re­ source in this process. Clinical pharmacists are pharma­ cost, and adverse effects. To improve compliance, patients cists who have obtained specific training in providing pa­ should also perceive that their medication regimen is con- tient care in the area of pharmacotherapy. The impact of a clinical pharmacist’s services on out­ __________________ |_ comes for specific diseases has been evaluated in a con­ Submitted, revised, July 5, 1995. trolled fashion. At least three separate studies on the con­ j This paper was presented at the Society o f Teachers o f Family Medicine Annual trol o f hypertension have shown improved blood pressure Meeting, Atlanta, Georgia, May 3, 1994, and at the Michigan Family Practice Research Day, East Lansing, Michigan, May 19, 1994. control with the use of a pharmacotherapy consulta­ tion.3-5 Anticoagulation therapy managed by clinical j From the College o f Pharmacy, Ferris State University, Big Rapids, (J-J-, K- V.), and J Grand Rapids Family Practice Residency, Grand Rapids (]■]-, G.V.), Michigan. pharmacists has also demonstrated favorable results.6 i Requests for reprints should be addressed to Glenn VanNoord, MD, Grand Rapids Studies have demonstrated that collaboration between a j Family Practice Residency, Saint Mary's Health Services, 200 Jefferson SE, Grand Rapids, MI 49503. clinical pharmacist and the treating physician benefits pa- © 1995 Appleton & Lange ISSN 0094-3S09 The Journal of Family Practice, Vol. 41, No. 5(Nov), 1995 469 Pharmacotherapy Consultation Jameson, VanNoord, and Vanden tients with multiple medical problems; however, these cused interview based on the current regimen and the studies have not been conducted in a randomized, con­ side-effects questionnaire described above. The pharma­ trolled fashion.5’7- '0 cist evaluated the medication regimen for the following In this randomized clinical trial conducted in a pri­ drug-related problems: drug interactions, unneeded mary care practice, we tested the hypothesis that a brief drugs, adverse drug reactions, therapeutic duplication in-office pharmacotherapy consultation involving a clini­ suboptimal drug selection, inappropriate dosage intend cal pharmacist, at-risk patients, and treating physicians and cost. The pharmacist then met with the treating phy­ would be associated with improved outcomes including: sician to discuss his findings. A new regimen was devel­ decreased number o f medications; decreased cost o f med­ oped by a collaborative dialogue between the physician ications; decreased number o f doses per day; improve­ and the pharmacist. Finally, the pharmacist conducted a ment in reported adverse effects; and improvement in brief educational session with the patient to explain any patient understanding and compliance with their medica­ changes in the regimen and to improve the patient’s un­ tion regimens. Further, it was hypothesized that these derstanding o f their drug therapy. improvements would persist for 6 months. One month after the intervention, the pharmacist contacted the patient by telephone (5 to 10 minutes) to reinforce the treatment plan. Six months after the inter­ Methods vention, the five outcomes were again measured. Physicians and other patient caregivers were specifi­ Description of the Study cally excluded from knowledge that patient outcomes From a primary care patient population, we identified were being studied. They were aware only that the office’s patients at risk for medication-related problems using the clinical pharmacist was performing pharmacotherapy con­ instrument validated by Koecheler et ah11 We randomly sultations on selected patients. Similarly, we did not in­ assigned patients to intervention and control groups and terfere in normal physician-patient assignments. As a re­ measured the outcomes for each group at baseline and sult, some physicians cared for patients in both groups. after 6 months. We measured the following outcomes by means o f chart review and questionnaires*: (1) the num­ ber of chronic prescription medications in the regimen; Patient Selection (2) the number o f individual doses per day; (3) monthly cost o f prescription drugs, based on “maximum allowable All patients seen in the Family Health Center of the Grand cost” for Medicaid reimbursement; (4) patient self- Rapids Family Practice Residency were placed on a se­ reports of compliance and of the drug regimen conve­ quentially numbered list each month. Charts were se­ nience using a scale developed by the authors*; and (5) lected for review using a computer program designed to side effects and problems, based on self-reports using a randomly select from the list. These randomly selected scale developed by the authors.* charts were reviewed by a research assistant for the pres­ ence o f two or more risk factors for adverse consequences o f medication therapy as defined by Koecheler et al.11 Intervention Patients were enrolled from June 1991 to December Each patient in the intervention group was given a 45- to 1992. The risk factors included: (1) five or more medica­ 60-minute pharmacotherapy consultation. The consulta­ tions in current regimen; (2) 12 or more daily doses; (3) tion was provided by a clinical pharmacist with a post­ four or more medication changes in the last 12 months; baccalaureate PharmD degree and experience in ambula­ (4) more than three concurrent disease states; (5) docu­ tory care. The goals o f the consultation were to simplify mentation of medication noncompliance in the medical the regimen, improve the effectiveness o f the regimen, record; and (6) drugs that require therapeutic monitor­ and decrease adverse effects. A secondary goal was to ing, eg, digoxin, theophylline. decrease cost if this could be accomplished without ad­ Patients were not eligible for enrollment if they: (1) versely affecting the first three goals. had evidence in the medical record o f active alcohol or Patients were asked to bring in all their medicines in illicit drug abuse; (2) were unwilling or unable to return a “ brown bag.” After performing a chart review, the phar­ for a pharmacotherapy consultation; (3) had their regi­ macist conducted a medication history. This was a fo- men primarily managed by an outside consultant; (4) were terminally ill; or (5) were less than 18 years of age. Eligible patients were randomly assigned to interven­ *Copies o f the questionnaires and scales used in this study are available from the authors. tion or control group using a random number table. The Journal o f Family Practice, Vol. 41, No. 5(Nov), 1995470 pharmacotherapy Consultation Jameson, VanNoord, and Vanderwoud Statistical Analysis Table 1. Presence of Screening Risk Factors at Baseline Demographic data were analyzed with Student’s ttest for Consult Control continuous data and Fisher’s exact test for categorical Group, % Group, % Risk Factor (n= 27) (n= 29) data. Baseline outcome variables and within-group 5 or more long-term medications 89 90 changes from baseline were compared using Fisher’s exact test for categorical data and the Mann-Whitney U test for 12 or more doses per day 26 28 continuous data. This comparison o f within-group 4 or more medication changes in 59 52 changes using the Mann-Whitney U test was necessary, the past year rather than the usual two-way ANOVA because the fre­ quency distribution was skewed with outliers. The only More than 3 chronic diseases 70 76 two-way ANOVA available for non-normal data is the Documented noncompliance 18 21 Friedman’s test for matched samples, which did not apply to our data.