IMPACT OF DRUG USE EVALUATION UPON AMBULATORY PHARMACY PRACTICE Edward P. Armstrong and Allan K. Terry

OBJECTIVE: To review the expansion of ambulatory drug use evaluation (DUE). A description of ambulatory DUE characteristics and methodology is included. In addition, DUE computer usage, documentation concerns, and future research issues are addressed.

A MEDLINE search was used to identify pertinent literature, including reviews.

DATA SOURCES:

STUDY SELECTION: Primary and secondary literature describing ambulatory DUE was selected. Articles describing inpatient DUE were excluded. DATA EXTRACTION: Rigorous studies evaluating current ambulatory DUE programs are limited, but the available literature and a description of existing program characteristics are included. All studies available at the time of publication were reviewed.

Ambulatory DUE can provide useful information to assist in providing pharmaceutical care. The Omnibus Budget Reconciliation Act of 1990 has prompted an expansion of DUE programs, and most of the formalized programs are retrospective in design. Prospective programs provide online, patient-specific drug use assessment whenever new prescriptions are entered into pointof-service databases.

DATA SYNTHESIS:

Although more well-designed evaluations of existing ambulatory DUE programs are needed, initial results indicate these programs may be extremely useful in identifying significant medication therapy problems and improving patients' drug therapy.

CONCLUSIONS:

Ann Pharmacother 1992;26: 1546-53.

has become a major influence in ambulatory care. I Professional roles are changing and many pharmacists are using this opportunity to enhance the pharmaceutical care they provide. DUE activities are expanding through government programs as well as private-sector initiatives,' and the growth of managed care has increased the emphasis upon such evaluations.Y Most DUE programs emphasize improved quality of patient care," Two other important goals are the education of medical and pharmacy providers and increased cost-effectiveness of drug therapy selections," This article explores the growth of the DUE process within the ambulatory care setting and its potential development. The types of DUE, methodology, computer applications, documentation, and future research issues are reviewed. We examined DUE as part of the pharmacist's DRUG USE EV ALUATION (DUE)

EDW ARD P. ARMSTRONG, Phann.D., is an Associate Professor, Department of Pharmacy Practice, College of Pharmacy, University of Arizona, Tucson, AZ 85721; and ALLAN K. TERRY, M.S., is the Chief of Pharmacy Service. William Keller Anny Community Hospital, West Point, NY. Reprints: Edward P. Armstrong, Phann.D.

1546 •

The Annals ofPharmacotherapy •

growing professional responsibility for patient-oriented pharmaceutical care. Although past pharmacist involvement has often been limited to contacting physicians when drug therapy problems were noted during dispensing procedures,? new DUE systems will help pharmacists identify and deal with these problems. The success of such programs depends upon the active participation of pharmacists and other healthcare professionals," The term "drug use evaluation" is increasingly preferred in discussions of multidisciplinary programs that examine the use of drugs, evaluate that use for actual or potential problems, provide corrective actions, and institute followup measures.' The terms "drug use review" or "drug utilization review" (DUR), found in most previously written drug use literature, are somewhat ambiguous in that a review does not necessarily assess effectiveness of remedial actions or quality of outcomes realized.i"!' Despite some of these differences, the terms are frequently used interchangeably," The Omnibus Budget Reconciliation Act of 1990 (OBRA 90) has caused important reevaluation in community pharmacy and medical practice. A great deal has been published and discussed about the impact of this legislation on drug manufacturer rebates and discount pricing. Other extremely important components of this legislation are the mandates for patient counseling and for retrospective and prospective drug use evaluation. This legislation, although focused on the Medicaid population, is expected to alter the care provided to all patients. Reimbursement to pharmacies for outpatient drug costs covered under Title XIX (Medicaid) will be contingent upon comprehensive DUE programs being implemented within all participating states by January 1, 1993.13 This requirement is one of many formal steps taken in recent years by third-party payers (governmental and private) to assure that the medical care for which they pay is of high quality, is cost-effective, rational, and unlikely to have adverse consequences.

Review ofthe Ambulatory Drug Use Evaluation Process RETROSPECTIVE DRUG USE EVALUATION

Retrospective DUE has been the most frequent mainstay in the formal evaluation of ambulatory drug use." In many settings, this has not been very sophisticated. Some programs review randomly selected patient profiles on a monthly basis; others review drug regimens of all patients receiving specific therapies.v-" Some plans focus upon ar-

1992 December,Volume 26

eas with well-established problems (i.e., all patients receiving more than eight weeks' therapy of full-dose histamine, H 2 blockersj."-" Others may scrutinize therapies for patients at risk of developing dependence on classes of medications such as the opioids or benzodiazepines. Although computers were introduced into pharmacy largely to aid billing and dispensing activities, there is now an expanded focus on other uses of the information in their databases. These data can be employed on a clinical basis to aid in reviewing and improving patients' drug therapy regimens. More sophisticated retrospective DUE programs have used tapes from pharmacy claims processors that summarize health plan or payer pharmacy data." These tapes can be crossmatched with medical diagnosis tapes and then used to create patient profiles listing both diagnostic information and complete medication regimens (regardless of pharmacy) covered by the payer." Other systems have developed relational databases, incorporating laboratory determinations with diagnostic and drug histories. It must be kept in mind, however, that the diagnostic data are not always accurate. Use of these data has led to the development of different computer software programs that can detect potential problems with patients' drug therapy regimens. The advantages and disadvantages of different DUE types are summarized in Table 1. Retrospective evaluations allow flexible scheduling of medication profile reviews, and these programs can identify patterns and trends in medication use by medical and pharmacy providers. Retrospective systems have lower direct expenses and their drug use criteria are easily modified. More sophisticated systems may add computerized weighting to focus on the most problem-prone patients. Retrospective DUE programs have several important limitations to their use of information. Their primary shortcoming is the difficulty in changing therapy for specific patients during courses of treatment. For example, during a retrospective review, analysis of a patient profile may indicate that a pediatric patient has received an overly broadspectrum, expensive antibiotic to treat her otitis media in-

fection. Retrospectively, it is difficult (if not impossible) to contact the prescriber before the patient finishes the tenday course of antibiotic. It can also be difficult to review retrospectively clinical data (e.g., to determine whether the patient has had an unusual culture and sensitivity) which would modify appropriate therapy. Despite their limited impact upon short-course therapy, retrospective programs do maintain their educational focus and make ongoing therapeutic interventions. Currently, there are several important limitations to the analysis of these DUE programs. Software systems from different vendors use proprietary information that has not been rigorously evaluated. No comparisons have been published of the abilities of alternative software systems to identify drug therapy problems using the same data tapes. Specific-vendor drug use criteria have been kept confidential; however, OBRA 90 requires that criteria be made available by January 1, 1993, for programs used for Medicaid recipients. The current primary retrospective DUE vendors are Health Information Designs, Inc. (Arlington, VA), Q-A, Inc. (West Des Moines, IA), and First Health Services Corp. (Glen Allen, VA). Universities have also become involved in developing their own systems, such as the Center for Health Systems Research and Analysis at the University of Wisconsin-Madison and the Department of Preventive Medicine and Biometrics of the University of Colorado Health Science Center. With the tremendous interest in ambulatory DUE, it is expected that the number of vendors will increase. CONCURRENT DRUG USE EV ALUAnON

Concurrent DUE is conducted once drug therapy has been started. It allows changes to be made within a reasonable tirneframe, thereby benefitting individual patients and prescribers." DUE programs that review the appropriateness of prescription refills in conjunction with new prescription orders may be called concurrent. In contrast to OBRA 90, concurrent DUE has also been defmed by pharmacy claims processors as the assessment of drug therapy by the pharmacist at the time of dispensing." Computer

Table 1. Drug-Use Evaluation Programs TYPE

WHEN PERFORMED

ADVANTAGES

Retrospective

afIer dispensing drug

Concurreni

during drug therapy

Prospective

prior to dispensing drug

offers time and personnel flexibility identifies patterns and trends is least cosily allows sampled screening of data can incorporate diagnosis-related data can use computer weighting to focus on the most problem-prone patients allows timely changes in therapy (at point of service makes it possible for users to obtain missing data makes the pharmacist responsible for intervention can identify and correct problems before starting therapy can prevent adverse drug reactions. interactions, or therapeutic duplications can reduce unnecessary drug costs criteria apply to all patients patients makes it possible for users to obtain missing data

The Annals ofPharmacotherapy

DISADVANT AGES

benefits and interventions are ofIen delayed obtaining clinical data. including accurate diagnosis. is difficult changing therapy during course of treatment is difficult occurs when therapy has already begun often lacks diagnostic data reluctance to act reduces effectiveness

is most costly requires immediate action and online access to patient information action is delayed if unable to contact prescriber override of warning messages possible reluctance to act reduces effectiveness

•

1992 December. Volume 26

•

1547

software programs employed by many community pharmacies to detect drug- drug interactions, dosing ranges beyond accepted guidelines, and overuse are valuable concurrent DUE interventions." Detection of therapeutic duplications within similar therapeutic classes usually has been considered a concurrent intervention. Long-term care facilities offer an excellent opportunity to implement concurrent DUE. Pharmacists concentrating on ambulatory care are often responsible for the pharmaceutical care of these patients. A published three-part series has reviewed the role of DUE in skilled-nursing facilities.24-26 Quality assurance and assessment committees are responsible for the quality of care issues within most nursing facilities. Many consultant pharmacists who provide monthly drug regimen reviews (DRRs) are expanding their services to include DUE activities because judgments made during DRRs provide an excellent background for a DUE program in that facility. Data collection for concurrent DUEs can be done simultaneously with monthly DRRs. DRRs involve individual consultations between pharmacists and prescribing physicians about individual patients' medication problems. In contrast, DUE identifies locationspecific drug therapy problems and incorporates an "interdisciplinary team approach" to create interventions for identified problems. 27 PROSPECTIVE DRUG USE EVALUATION

Prospective DUEs are conducted prior to dispensing new patient prescriptions. Prospective involvement is most costly in terms of professional time needed for pharmacy and medical providers to collect appropriate information and to discuss important concerns. Such evaluation allows therapy changes to prevent adverse outcomes or unnecessary drug costs. Optimal performance of prospective DUE activities requires immediate access to complete information about individual patients. With the OBRA 90 mandate for state-level prospective DUE programs for Medicaid patients, a tremendous emphasis has been placed on improving different software programs. It is difficult to assess these programs because they are currently undergoing significant modifications and upgrades. Currently, PCS Health Systems, Inc. (Scottsdale, AZ) has the greatest experience with their Quantum Alert prospective system, but other companies are also creating programs. Quantum Alert (through the pharmacy claims being processed by PCS, Inc.) detects inappropriate drug doses, drug-drug interactions, therapeutic duplications, drug-age conflicts, excessive use, early refills, and drug-pregnancy advisories." Because the system functions through the pharmacy claims processor, the Quantum Alert system operates without concern for where the patient receives his prescription. Excessive use is currently the major medication problem detected with this system." Some claims processors say they provide prospective DUE when all they actually do is screen (or block payments) for medications that are refilled too early. This lack of sophistication leads to tremendous confusion in discussions of different prospective DUE systems and their costs and benefits. For optimal patient benefit, a combination of the above three methods is recommended, particularly retrospective and prospective DUE. Prospective DUE may be especially beneficial for detecting and intervening on specific types of 1548 •

The Annals ofPharmacotherapy •

concerns such as inappropriate doses, excessive use, and medication-pregnancy conflicts, and retrospective DUE programs detect a broad range of problems, determine patterns and trends, and provide educational information. DRUG USE EVALUATION METHODOLOGY

The well-known Joint Commission on Accreditation of Healthcare Organizations' Examples ofDrug Usage Evaluation describes DUE as a ten-step formal process," In applying DUE to ambulatory practice, this process can be modified in the following manner. Step 1. Assign Responsibility for the Drug Use Evaluation Program. Identify committee membership and determine specific physicians and pharmacists who will be involved in the program. OBRA 90 mandates the creation of state DUR boards to formulate criteria and evaluate Medicaid patient therapies. These boards must comprise at least one-third, but not more than 51 percent, licensed and actively practicing physicians, and at least one-third licensed and actively practicing pharmacists." Step 2. Examine Previous Drug Use and Identify Problem Areas. Retrospective prevalence data can be helpful during this assessment. Such well-known problems as prolonged use of H2 blockers at full doses, use of newer antidepressants, use of more expensive nonsteroidal antiinflammatory drugs, or prescribing of broad-spectrum antibiotics may help identify initial areas of focus. Guidelines for selecting drugs to study have been developed by the American Society of Hospital Pharmacists." a modified version of which appears in Table 2. Although these guidelines are helpful, no extensive research has been performed to determine the actual frequency of their use within varying healthcare settings. Step 3. Develop (or Revise) Drug Use Evaluation Criteria. Criteria represent "predetermined elements against which aspects of the quality of medical services may be compared.'?' Pharmacists and physicians develop criteria using the available literature and their clinical experience. Criteria may be developed by members of a DUE committee, then submitted for practitioner input, subsequently approved by a DUE committee, and finally distributed to those whose behavior will be evaluated. OBRA 90 requires that drug criteria be based upon peer-reviewed medical literature and three specific sources: United States Pharmacopeial Drug Information, American Medical Association Drug Evaluations, and the American Hospital Formulary Service Drug Information:" Criteria developed externally for use in a particular practice setting will require modification and periodic updating. Some ambulatory DUE software vendors have kept specific drug criteria as propriTable 2. Rationale for DUE Study Drug Selection" Use likely to result in interactions or adverse reactions Use common in high-risk patients High level of use High cost Currently undergoing evaluation for formulary changes Specific manner of use required for maximum effectiveness Discomfort or toxicity possible at normal dosages Required by administrative policy DUE = drug use evaluation. •Adapted from Reference 30.

1992 December, Volume 26

Drug Use Evaluation

etary information, limiting the educational component of their DUE programs, but this practice will change when OBRA 90 is implemented. Specific drug criteria have not routinely been given to all providers in most ambulatory DUE programs. These programs have typically focused on large patient databases involving hundreds or thousands of medical providers, and it is a huge task to notify all potential providers about specific criteria for all ongoing DUE studies. This is further complicated when multiple DUEs are running concurrently and criteria are frequently updated or modified. Noting the educational focus of OBRA 90, it seems there will be a greater emphasis on the distribution of criteria. In 1974, Knapp et aI. published an excellent overview of the development and application of criteria to ambulatory DUE programs, referring to systematic development as the critical component in a quality DUE program." Screening criteria are those used to perform an initial assessment of large amounts of data (e.g., all prescriptions in the database for a specific drug) to find any significant problems not meeting predetermined criteria. Prescriptions that do not meet initial screening criteria require further evaluation. In 1991, Knapp reviewed systematic development of criteria in an ambulatory DUE program." Step 4. Data Collection and Analysis. In the ambulatory care setting, this step is largely performed by computer systems to identify specific patient profiles for review by a DUE committee or other peer review organization. It is important to ensure that data within the computer system are properly formatted and that a mechanism exists to identify individual physicians and pharmacies. Currently, there is considerable controversy about how individual physicians should be identified. Although the practice is criticized, many programs identify physicians by their Drug Enforcement Administration numbers, and will probably continue to do so until an improved identification system is developed. Step 5. Develop a Plan of Action Appropriate to the Specific Situation. Extensive research has assessed the effectiveness of various methods of influencing physicians' prescribing patterns.v-" Soumerai et aI. reviewed 44 drug use studies and identified several types of interventions used to bring physician prescribing behaviors within established norms, including distribution of educational materials (newsletters, journal articles); feedback on prescribing activities (physician-specific or institutional); group educational meetings (inservices, seminars, conferences, continuing education programs); one-on-one counseling sessions (using physician-counselors or clinical pharmacists); and a wide range of punitive measures (withholding payment for services rendered, restricting future practice privileges)." Measures used have generally depended upon practice settings involved, resources available, the effectiveness of past interventions, and administrative decisions." Ideally, interventions should be reevaluated periodically for effectiveness in alleviating drug use problems. This will identify the most effective interventions within a particular setting and with individual prescribers. In those cases where noncompliance persists, it may be wise to examine both the feasibility of criteria changes and punitive measures. Freidson found that programs successful in changing physician behaviors have clearly defmed educational goals

and behavioral objectives, as well as active physician involvement." Soumerai and Avorn recommended the use of "academic detailing" (methods similar to those successfully employed by drug manufacturers' detail people) to alter prescribing patterns. Such tactics include efforts aimed at local opinion leaders; follow-up visits with positive reinforcement for improved compliance; to-the-point graphic educational materials; and repetition of key messages." DUE interventions can be expected to work most effectively when physicians are employees of provider organizations or at fmancial risk for the behavior of their peers. With retrospective DUE in the ambulatory care setting, the primary intervention has been letters written to the physicians and pharmacists involved in patient care." Most retrospective programs use a basic form letter with the specific concerns about the patient's profile listed. Programs differ in the specific information to be included in these letters. Most programs include a peer review mechanism whereby pharmacists and physicians review patient profiles generated by a computer system. Commonly, the peer review process identifies not only the reason why the computer flagged a profile but also other potential problems in the medical regimen, and the process can then be used to determine which items to include in the letters written to medical and pharmacy providers. Contrarily, the retrospective DUE programs that do not include this peer review analysis will distribute letters based solely on those profiles failing to meet specific drug use criteria. These programs save the cost of the specific pharmacist and physician reviewers serving on a peer review committee, but they miss the expertise of such analysis of the entire patient profile. No studies evaluating the effectiveness of these two approaches have been published. Prospective DUE programs do not generally intervene with letters when DUE action is required. Typically, a pharmacist in an ambulatory setting will be notified onscreen that some type of problem exists with a patient's drug regimen. Actions here often involve contacting prescribers about specific concerns and resolving issues before dispensing prescriptions. Widespread implementation of prospective DUE programs will produce important changes in the practice of ambulatory pharmaceutical care. State pharmacy practice acts, documentation of patient assessments and interventions, and current emphasis upon patient counseling will be affected. Step 6. Assessment of Actions and Documenting Improvement. Brodie et aI. suggest that any comprehensive DUE program evaluation should include quality of care changes (evidenced by reductions in adverse drug reactions or the incidence of use or misuse of specific drugs); changes in medication inventories and associated costs attributable to the ongoing DUE program; changes in healthcare providers' and patients' attitudes toward the DUE process; and a calculation of total costs involved in the program." The extent to which current ambulatory DUE programs conduct such overall evaluations is unclear and so is a potential area of research. The cost-effectiveness of ongoing DUE programs is an essential component of their overall evaluation. Whereas both indirect and direct program costs can be identified rather easily, effectiveness or benefit is more difficult to determine. Most programs emphasize comparisons involving original patient profiles, interventions, and current pa-

The Annals ofPharmacotherapy •

1992 December, Volume 26 •

1549

tient profiles (usually three to nine months later)." A common design flaw has been the lack of a control group. Patients on chronic full-dose H 2 blockers may have the drugs stopped after the initial profile has been examined and acted upon, but the drugs may have been discontinued regardless of whether the prescribing physicians received intervention letters. This creates a situation in which the patients' therapies were improved, but it is not completely clear that the DUE letters were responsible for the change. The impact of the educational process within DUE programs is also difficult to determine. If the DUE program educates providers to improve their prescribing practices, the benefits likely will not be confined only to patients whose therapies are reviewed. It is anticipated that the improved prescribing will carry over to other patients seen by those providers. This will expand the benefits of the DUE program beyond individual patients. Confmement of DUE outcome assessments to only those patients receiving interventions would underestimate the effectiveness of any evaluation program. Hospitalizations, readmissions, or lengths of stay reduced and total drug expenditures lowered may also indicate program effectiveness. Other factors, such as improvements in quality of life, general well-being, or patient satisfaction have been described as important measures of healthcare effectiveness.t-" but few published studies have tried to characterize the value of DUE programs based on these factors.? Step 7. Reporting Relevant Information to Appropriate Committees for Further Action. The number of committees will vary with the type of ambulatory DUE program. Within managed care settings, pharmacy and therapeutics, quality assurance, and executive committees are most often involved. Within state Medicaid programs, peer review committees, DUR boards, and the state Medicaid agencies will interact.

ComputerApplications for Drug UseEvaluation Computerization of ambulatory care pharmacy operations is essential for efficient use of time and for participation in many prescription drug reimbursement plans." Computers save valuable professional time and enhance efforts to control healthcare costs. Documented uses of computers include maintaining patient profiles, checking for drug interactions, identifying most- and least-prescribed drugs, generating prescription labels, performing businessrelated functions, and monitoring other data." Helling et al. evaluated four different approaches to DUE, including a "computer-assisted medical record audit." The authors recommended this system for settings in which healthcare providers have access to both medical records and prescription files." Kawahara and Jordan published an interesting account of the insertion of informational texts into computerized pathways for physician order entry. This effort was designed to alter prescribing by offering therapeutic alternatives of lower cost or demonstrated greater efficacy to physicians using the organization's computer monitors." This "prospective intervention" holds great potential for future applications, but has numerous limitations. Unless such orders are entered by the actual prescribers, no changes in prescriber behavior can be expected to occur. 1550 •

The Annals ofPharmacotherapy •

More recently, a microcomputer database software program was used to identify areas appropriate to drug use evaluation through broad-based data collection. The authors used this system to print detailed, prescriber-specific reports for DUE committee evaluation and follow-up." New prospective ambulatory DUE activities most often occur in conjunction with online electronic submission of payment claims. Not only are denials of payment for claims nearly eliminated by this process, but pharmacists are also encouraged to intervene to resolve identified drug-related problems." Adequate performance of such procedures can best occur with point-of-service (POS) computer terminals which provide ready access to all pertinent patient-specific information. pas computer systems have been widely used in community pharmacies since 1987. Because online programs require complete, accurate data when service is rendered, prescribers may be prompted to furnish all pertinent information to pharmacists involved in a patient's care. Optimal assessments of outcome can best be performed where ambulatory and inpatient care settings can access one another's patient data files.

Computerized Prospective Drug UseEvaluation There are new prospective DUE products that provide online, patient-specific utilization reviews and clinical monitoring whenever prescription drug claims are filed. A complete medication profile, including refill information, is maintained for each patient enrolled in these programs. These real-time, pas systems incorporate patient-specific data into an easily accessible central database, regardless of the number of pharmacies or prescribers involved. The software prospectively checks patient eligibility and screens for both overuse and potential clinical problems. Criteria are usually established by the claims processor's professional staff. When warning messages appear in response to claim submissions under these programs, pharmacists can intervene or override the system, but they must intervene in nearly all cases of overuse warning messages. The incidence of system overrides is certainly worthy of detailed re-

search. Identification of drug-related problems can have a positive impact upon patients' quality of life. The National Council for Prescription Drug Programs (NCPDP) has established a DUR work group to develop industry standards for the transmission of warning messages from claims processors to pharmacies and response messages from pharmacies back to claims processors." There are currently no published data indicating the impact of these systems on overall drug or healthcare costs, and such data will be critical to assess these programs adequately and to develop methods to improve the systems and aid interactions with pharmacy providers. Many software products are available for use when conducting ambulatory DUE programs. Although most of these systems have focused on retrospective claims analysis, increasing emphasis on prospective DUE is prompting many companies to develop prospective programs. Unfortunately, no comparisons of different retrospective DUE systems or different prospective DUE programs have been published. Prospective DUE programs have certainly not been problem-free. In most cases no provision has been made to

1992 December, Volume 26

Drug Use Evaluation

compensate pharmacists who perform interventions that result in no prescription being dispensed, and this lack of economic incentive can reduce the utility of such systems. Costs involved in system maintenance and training of pharmacy personnel must be considered negative factors, particularly when several noninterfacing systems are used. Another important issue is that of patient safety versus confidentiality. There is considerable concern about the inadvertent relaying of patient information and about access to confidential data by multiple parties. Resolution of unfavorable trends in physician prescribing, pharmaceutical care provided, or patient noncompliance may require action by the client-payer for pharmaceutical benefits, and these payers may have little influence over nonconforming providers, short of excluding them from the program. Such acts could negatively impact patients cared for by excluded providers. Drug Use Evaluation in the Provision ofPharmaceutical Care

Pharmaceutical care has been defined as pharmacists sharing in the responsibility for drug therapy outcomes." This requires the profession to refocus upon appropriate drug therapies and relevant drug therapy monitoring. Computerized DUE programs (retrospective, concurrent, or prospective) are not a substitute for the pharmacist talking to the patient, but they can be effective tools for enhancing pharmaceutical care. Retrospective DUE systems frequently detect therapy problems that could have been alleviated by ambulatory care pharmacists." Further, newer prospective systems may involve a large number of pharmacy practitioners not currently taking part in these types of decisions. Employers and associations of pharmacists must provide instruction so that pharmacists can fully use this technology. The central focus must be upon improved patient care. Some prospective DUE software companies are developing algorithms for pharmacists to follow when different types of drug therapy problems are noted, including documentation of important questions pharmacists have asked, when prescribers were contacted, and actions taken if prescribers could not readily be contacted. It will become very important to assess the time and effort devoted by pharmacists on behalf of their patients and to develop adequate payments for these critical cognitive services.

Rather than solely using predetermined NCPDP pharmacy response codes, free text documentation would be beneficial, allowing pharmacists to fully document professional judgments and actions. Future Research Issues

Because ambulatory DUE has expanded dramatically and involves many pharmacy and medical practitioners, a number of important questions need to be answered. Evaluations of different programs must compare not only specific drug criteria but also the different DUE processes used by various organizations. In addition, identical DUE criteria may provide different outcomes in populations containing mostly young, healthy patients versus elderly or indigent patients. Research should be conducted to determine the impact of established DUE programs on patients' quality of life, hospitalization rates, laboratory tests, and emergency office visits. These will be complicated studies, but the results will be vital for assessing the role of DUE and developing methodologies for systems improvement. Research is also needed to determine the best mix of DUErelated interventions (i.e., letters, telephone calls, face-toface visits, or newsletters) to deliver optimal patient outcomes. Use of confidential data must be refmed. A quality retrospective, concurrent, or prospective DUE program must have access to confidential information, but this information must be protected and maintained under the auspices of professional quality assurance programs. We must devise methods of restricting access to online data (involving not only prescription but also diagnostic information) to those pharmacists and physicians providing patient care. Most community pharmacies have inhouse computer systems with limited types of DUE capabilities. These systems should be evaluated to determine what types of problems are being found, what types of interventions are being used, and their relative successes in improving patient outcomes. It is unclear how different online prospective DUE programs will interface with one another (e.g., a national claims processor prospective system and an inhouse pharmacy system). Further, individual pharmacy practitioners will need proper training to work with a variety of DUE systems, and research is needed to determine the most effective educational methods for practitioners to properly interface with these systems.

Documentation

Another essential component in the provision of pharmaceutical care is proper documentation. 50 This is necessary for an effective ambulatory DUE program and is especially important for a successful prospective DUE process. Providing pharmaceutical care, paying for cognitive services, and demonstrating action and outcomes from DUE programs will not occur until there is adequate pharmacy documentation." Historically, medicine and nursing have been more successful than pharmacy in properly documenting their assessments and actions. Employers and payers must emphasize the importance of documentation. Retrospective systems have traditionally relied upon paper documentation, but it would seem natural for the newer prospective systems to use computerized documentation technology.

Summary

New DUE technology has significantly expanded pharmacists' opportunities to improve patients' drug therapies. Retrospective and prospective systems seem to have the greatest utility for most practitioners, but the roles of these different systems need to be clarified to promote optimal use. DUE programs can help ambulatory pharmacists provide the highest quality of pharmaceutical care. ~ References I. Sturek JK. Point-of-service computer system and drug-use evaluation: implications for pharmacy practice in ambulatory care. Am J Hasp

Pharm 1989;46:517-20. 2. Curtiss FR. Managed health care. Am J Hasp Pharm 1989;46:742-63. 3. Cotterell CC, Dombroske L, Fischermann EA. Comprehensive drug-

The Annals ofPharmacotherapy

•

1992 December, Volume 26 •

1551

use evaluation program in a health maintenance organization. Am J HospPharm 1991;48:1712-7. 4. Summers KH. Managed health's impact on pharmacy practice. US Pharm 1991;16:H-16-36. 5. Brodie DC, Smith WE. Constructing a conceptual model of drug utilization review. Hospitals 1976;50:143-9. 6. Rucker TO. The need for drug utilization review. Am] Hosp Pharm 1970;27:654-8. 7. Campbell WH, Johnson RE, Christensen DB. A procedural and conceptual analysis of drug use review. Drugs Health Care 1975;2:211-30. 8. Laventurier M. Utilization and peer review by pharmacists. J Am Pharm Assoc 1972;NS12:166-70. 9. Myers C. Keeping up-to-date with Joint Commission requirements: the case of drug-use evaluation. Am] Hosp Pharm 1988;45:64,69. 10. Hlynka IN, Danforth DH. Drug usage review. Part one: designing the BC Pharmacare program. Can Pharm J 1981;11:421-3. II. Stolar MH. Conceptual framework for drug usage review, medical audit and other patient care review procedures. Am] Hosp Pharm 1977;34: 139-45. 12. Fish CA, Ascione FJ. Drug usage evaluation (DUE) activity in PSAOs. Am Pharm 1988;NS28:651-4. 13. Omnibus Budget Reconciliation Act (OBRA) of 1990. Conference Report 101-964 to accompany HR 5835. Washington, DC: US Government Printing Office, 1990:160-71. 14. Huntsinger L. Drug use review of Oregon. Pharm Times 1990;56:39-41. 15. Brandenburg JA. Kentucky Medicaid DUR: planning for the future. Pharm Times 1992;58:67-70. 16. Goldman MP. Ciprofloxacin drug utilization review and prospective drug use evaluation. DICP Ann Pharmacother 1990;24:82-6. 17. Ulaszek KM, Seabloom KD, Sampliner RD, Jones WN. Appropriateness of long-term cimetidine prescribing. Drug Intell Clin Pharm 1984;18:623-5. 18. Mead RA, McGhan WF. Use of histaminej-receptor blocking agents and sucralfate in a health maintenance organization following continued clinical pharmacist intervention. Drug Intell Clin Pharm 1988;22:

466-9. 19. Martin S, Menighan TE. APhA announces national DUR program. Am Pharm 199O;NS30:525-7. 20. Strom BL, Morse ML. Use of computerized databases to survey drug utilization in relation to diagnoses. Acta Med Scand 1988;721 (suppi):13-20. 21. Miller OW, Knapp DA. Drug use review in the community pharmacy. Am Druggist 1990;7:44,46-7. 22. Levine GM, Deardeuff JC. NCPDP's concurrent drug utilization review standard: a blueprint for improving the management of pharmaceutical care. Drug Benefit Trends 1992;4:3-8. 23. Poirier TI, Giudici R. Evaluation of drug interaction microcomputer software: comparative study. Hosp Pharm 1991;26:30-3,37. 24. Hernandez P, Davidson HE, Bauwens SF, Chamberlain TM. Druguse evaluation, part I: concepts and methods for evaluating the use of drugs in organized health care environments. Consult Pharm 1991;6: 643-8. 25. Davidson HE, Chamberlain TM, Hernandez P, Bauwens SF. Druguse evaluation, part II: drug-use analysis in pharmaceutical care and quality assurance. Consult Pharm 1991;6:836-43. 26. Davidson HE, Gutshall EL, Chamberlain TM, Bauwens SF. Druguse evaluation, part III: implementing a drug-use evaluation plan. Consult Pharm 1991;6:968-75. 27. Feinberg JL. Meeting the mandate for quality assurance through druguse evaluation: drug-use evaluation procedures create a tearn spirit that enhances the consultant pharmacist's drug-regimen review efforts. Consult Pharm 1991;6:611-20. 28. PCS Quantum Alert. Scottsdale, AZ: PCS Health Systems, Inc., May 1991. 29. Examples of drug usage evaluation. Chicago, IL: Joint Commission on Accreditation of Healthcare Organizations, 1990. 30. American Society of Hospital Pharmacists. ASHP guidelines on the pharmacist's role in drug-use evaluation. Am] Hosp Pharm 1988; 45:385-7. 31. National professional standards review council: defmitions. Washington, DC: PSRO Council, November 26, 1973. 32. Knapp DA, Knapp DE, Brandon BM, West S. Development and application of criteria in drug use review programs. Am] Hosp Pharm 1974;31:648-56.

1552

•

The Annals ofPharmacotherapy

•

33. Knapp DA. Development of criteria for drug utilization review. Clin Pharmacol Ther 1991;50(part 2):600-2. 34. Raisch OW. A model of methods for influencing prescribing: Part I. A review of prescribing models, persuasion theories, and administrative and educational methods. DICP Ann Pharmacother 1990;24:417-21. 35. Raisch OW. A model of methods for influencing prescribing: Part II. A review of educational methods, theories of human inference, and delineation of the model. DICP Ann Pharmacother 1990;24:537-42. 36. Soumerai SB, McLaughlin TJ, Avorn J.lmproving drug prescribing in primary care: a critical analysis of the experimental literature. M i1bankQ 1989;67:268-317. 37. Schaffner W, Ray WA, Federspiel CF, Miller WO.lrnproving antibiotic prescribing in office practice: a controlled trial of three educational methods. ]AMA 1983;250:1728-32. 38. Freidson E. Doctoring together: a study of professional social control. Chicago, IL: University of Chicago Press, 1975. 39. Soumerai SB, Avorn J. Principles of educational outreach ("academic detailing") to improve clinical decision making. ]AMA 1990;263:549-6. 40. Groves RE. Therapeutic drug-use review for the Florida Medicaid program.Am] HospPharm 1985;42:316-9. 41. Brodie DC, Smith WE, Hlynka IN. Model for drug usage review in a hospital. Am] Hosp Pharm 1977;34:251-4. 42. Shearer SW, Burns WR, Sarlin MD. Cost impact of drug utilization evaluation and management in a private medical practice with an HMO component. Med Interface 1991;4:24-33. 43. Craig RP. On-line adjudication of third party prescription claims: a technological revolution. Calif Pharm 1989;10:45-7. 44. Fassett WE, Christensen DB. Computer applications in pharmacy. Philadelphia: Lea & Febiger, 1986. 45. Helling OK, Hepler CD, Herman RA. Comparison of computer-assisted medical record audit with other drug use review methods. Am ] HospPharm 1979;36:1665-71. 46. Kawahara NE, Jordan FM. Influencing prescribing behavior by adapting computerized order-entry pathways. Am] Hosp Pharm 1989;46: 1798-1801. 47. Loertscher MV, Groneman L.lmproving drug usage evaluation with microcomputer database software. Hosp Form 1990;25:752-6. 48. Hepler CD, Strand LM. Opportunities and responsibilities in pharmaceutical care. Am] Hosp Pharm 1990;47:533-43. 49. Armstrong EP. DUE software highlights therapeutic issues. Am Pharm 1992;NS32:806-10. 50. Hatoum HT, Vlasses PH. Patient outcome and the future practice of pharmacy. DlCP Ann Pharmacother 1991;25:208-10. 51. Angaran OM. Quality assurance to quality improvement: measuring and monitoring pharmaceutical care. Am] Hosp Pharm 1991;48:1901-7.

EXTRACfO

EI objetivo primario de este articulo es revisar la expansi6n de la evaluacion de uso de medicamentos (EUM) a nivel ambulatorio. Se induye una descripci6n de las caracterfsticas y metodologfa de una EUM a nivel ambulatorio. En adici6n, se rnencionan el uso de computadora en EUM, preocupaciones en cuanto a la documentaci6n y asuntos para futura investigaci6n. OBJETlVO:

Se realiz6 una biisqueda computarizada utilizando MEDLINE para identificar literatura pertinente, induyendo articulos de revisi6n. FUENTES DE INFORMACION:

Se seleccion6 literatura primaria y secundaria que describfan programas de EUM a nivel ambulatorio. Artfculos describiendo la EUM a nivel de pacientes hospitalizados fueron exclufdos,

SELECCION DE FUENTES DE INFORMACION:

METODO DE EXTRACCION DE INFORMACION: Estudios rigurosos que evahian programas de EUM a nivel ambulatorio estan limitados. Se induye la literatura disponible y la descripci6n de las caracterfsticas de programas existentes. Todos los estudios disponibles al momenta de publicaci6n fueron revisados.

La EUM a nivel ambulatorio puede proveer informaci6n util para ayudar en la provisi6n de cuidado farmaceutico, EI Omnibus Budget Reconciliation Act de 1990 (OBRA 90) ha motivado una expansi6n de programas. La rnayorfa de los programas de EUM a nivel

SINTESIS:

1992 December, Volume 26

Drug Use Evaluation

ambulatorioson retrospectivos. Los programasprospectivos proveen una evaluaci6ndel uso de farmacoscomputarizada y especificapara cada pacientecada vez que las nuevas recetas son dispensadas. CONCLUStONES: Aunque se necesitanevaluacionesde programasde EUM a nivel ambulatoriomejor disefiadas, los resultadosiniciales indicanque estos programas podrian ser extremadamenteutilespara identificarproblemassignificativos en la terapia con farmacosy para mejorar la fannacoterapiaen los pacientes. GISELLE C. RIVERA

RESUME

OBJECTIF: Revoir l'expansion recente qu'ont connue des programmesde revue d'utilisation des medicaments(RUM). Une descriptiondes caracteristiques et de la methodologiedes RUM est aussi inclue. De plus, des sujetsconnexes aux RUM tels que l'utilisation d'ordinateurs, les considerations documentaires et les futurs domaines de recherche sont aussi abordes. REVUE DE LfITERATURE: Une revue de la Iitterature II partir de 1abanque de donnees MEDLINE a permisd'identifier les publications pertinentes y compris les articlesde revue. SELECI10N DES I!.'TUDES: Seules les publications primaireset secondaires decrivantdes RUM effectueeschez des patientsambulantsont ete

selectionnees, Les publications decrivantdes RUM chez des patients hospitalises ont ete exclues, SELEeI10N DE L'INFORMATION: II existe Ires peu d'etudes decrivantdes RUM chez des patients ambulants.Celles qui utilisentune methodologie rigoureusesont encore plus rares. Les auteursont inclus une description des caracteristiques des differentsprogrammeset une liste de la litterature disponible. Toutes les etudes recenseesjusqu'au moment de la preparation de cette publication ont aussi incluses. RESUME: Les RUM permettentd' obtenir de I'informationIres utile II la distribution des soins phannaceutiques.La loi omnibusde 1990(OBRA 90) a declenchel'expansion de ce genre de programmes. La plupartdes RUM sont effectueesde facon retrospective. Les programmes prospectifspermettentquant II eux d'evaluer l'utilisation des medicamentsau fur et II mesure que les nouvellesprescriptions sont inscritesII la base de donnees. CONCLUSIONS: Bien qu'une methodologieplus rigoureuse soit souhaitabledans Ie futur, les resultatsinitiauxobtenus II partir des RUM effectueeschez des patientsambulants indiquentque de tels programmespeuvent eire extremernentutilespour identifierles problemes importantsrelies II la medicationet ameliorerla therapie medicamenteuse,

The Annals ofPharmacotherapy •

SUZANNE LAPLANfE

1992 December, Volume 26 •

1553