Work redesign and role restructuring in a pharmacy department with pharmacist assistants.

A new staffing model for decentralized pharmacists and support staff at a university hospital is described. A new technical support position--the pharmacist assistant--was created, and activities were reallocated among the pharmacists, pharmacist assistants, and pharmacy technicians according to the recommendations of two total quality management teams. Pharmacist assistants were to perform many of the drug distribution and record-keeping functions previously performed by pharmacists. The activities marked for reallocation accounted for about 50% of pharmacist time in the existing staffing model; they would not be performed by experienced pharmacy technicians who received brief training to be pharmacist assistants. Nine pharmacists and nine technicians tested the new staffing model on four patient care units over a four-week period. The reassignment of pharmacist tasks to pharmacist assistants theoretically gave pilot-unit pharmacists more time for providing and documenting patient-specific clinical activities and an additional 12 hours a week to participate in such global patient care activities as critical-pathway and quality improvement meetings. By the end of the study, the pharmacist assistants demonstrated that they could perform the reassigned activities; however, the pharmacists were not yet confident in the assistants' ability to do so. A pilot study of a new decentralized pharmacy staffing model demonstrated that technical support personnel can perform many distributive and record-keeping activities traditionally performed by decentralized pharmacists.

Automation support of patient-focused care.

The use of automation in hospital pharmacies is essential to support the concept of patient-focused care. Patient-focused care principles are defined, and the positive effect automation can have on supporting these principles is discussed. Automated technologies supporting drug distribution, sterile product production, prescription filling, and point of care information systems are summarized. Drug distribution technologies include distribution to and from the patient care area, distribution of medication to the patient, inventory control, controlled substance systems, and documentation of medication administration. Finally, a comprehensive automated technology system at a university teaching hospital is described.

Development of standards for providing prescription drugs by mail.

The development and implementation of practice standards for a mail-service pharmacy are described. Literature, licensing regulations, and existing national practice standards were reviewed. Selected mail-service pharmacies were surveyed on their practice standards. Patients who had received prescription drugs by mail from a university hospital pharmacy were asked their opinions of the service in a telephone survey. Based on the information gathered, a set of practice standards was drafted. The draft was considered by a panel of experts who reached a consensus on mail-service pharmacy practice standards. The standards were implemented at a university hospital's outpatient pharmacy. Standards of practice for mail-service pharmacies were developed and applied at a university hospital outpatient pharmacy. Other pharmacies may find the standards useful.

Time and cost requirements for decentralized pharmacist activities.

Time and cost requirements for pharmaceutical services in patient-care areas at a 548-bed university hospital were studied. The study was conducted in 1987 and 1988 to (1) define the clinical and distributive activities of decentralized pharmacists, (2) develop time standards for each activity, (3) determine whether the time requirements of decentralized pharmacists depend on the type of patient involved, (4) determine the actual costs of decentralized pharmacist services for various types of patient, and (5) compare costs with reimbursement for clinical pharmacy services. Time standards were established based on data from seven patient categories representing a cross section of the institution's patients. The mean frequency of each activity and the total time and cost per patient day for all activities were determined. Pharmacist time spent daily in each patient-service category ranged from 2.3 hr for low-intensity medical care to 20.8 hr for trauma-burn intensive care. Decentralized pharmacists spent approximately 50% of their time on clinical activities in all patient-service categories. The daily cost per patient day for clinical activities was lowest for low-intensity medical care and highest for adult intensive care. The institution's daily charge for clinical activities ($10/admission and $10/day) exceeded the cost of clinical services during the study period. The decentralized pharmacist time requirement per patient day, and thus the costs of delivering pharmaceutical services, varied by patient-service category. The provision of clinical services generated a profit. A pharmacy workload analysis system that can identify costs and correlate them with patient types can be valuable in hospital pharmacy management.

PharmaTrend as a management tool: implementation in a university hospital.

The implementation of ASHP's hospital pharmacy management information system, PharmaTrend, in a university hospital is described. The pharmacy department at the University of Wisconsin Hospital and Clinics (UWHC) is a decentralized operation with 63.5 pharmacist full-time equivalents (FTEs) and 81.6 supportive staff FTEs. The multipharmacy capability of PharmaTrend, which allows workload and productivity data for several pharmacies (usually satellites) to be combined into total departmental data for a given period, was used by UWHC to describe the pharmacy department's major cost centers individually. These centers were inpatient services, outpatient services, home health i.v. services, and two offsite pharmacies. Data were divided into elements collected by computer, elements not previously collected or requiring new procedures, and elements requiring manual collection. Data entry forms specific to each cost center were developed. With these forms, collected data can be categorized appropriately by technical staff members. Work hour allocation for each cost center was determined on the basis of PharmaTrend definitions and institution-specific considerations; only those hours reflecting direct fiscal burden of the department on the hospital were included. A manual process for collecting and categorizing data on hours worked and hours paid was automated and adapted for use in PharmaTrend. Finally, some preliminary experience with the system was obtained so that information could be collected for the development of policies, procedures, and training methods and on the use of forms and categorization methods. In particular, the time requirements for collecting, categorizing, and entering selected PharmaTrend data by cost center were determined. Optimal use of PharmaTrend requires creativity and a commitment of time and effort.

Evaluation of prescription charges in a health maintenance organization.

Differing patterns in drug use and charge variances among three pharmacy sites were determined for a university-based health maintenance organization (HMO). Computerized prescription claim data for the HMO inhouse pharmacy and those chain and independent pharmacies contracting with the HMO were analyzed on a personal computer using database management and statistical software. Chain and independent pharmacies were under contract to charge the average wholesale price plus a $2.00 fee per prescription. Drug-claim data were separated by pharmacy site (inhouse, chain, and independent), and the data for each site were sorted by national drug code number. High-charge and high-use drugs were identified for evaluation on a unit-charge basis. During the six-month study period, 22,393 prescriptions were filled, representing total charges of $347,063.79. The mean per-prescription charges submitted to the HMO by the chain, independent, and inhouse sites were $14.99, $16.08, and $16.41, respectively. The inhouse pharmacy had many charges in high dollar ranges. Of 10 high-charge drugs analyzed, isotretinoin 40 mg, flunisolide nasal solution, naproxen 250 mg, cyclosporine oral solution, and atenolol 50 mg were dispensed in significantly different proportions at each site. There were significant differences in unit prices of nine high-use drugs among pharmacy sites. This study indicates the importance of evaluating adverse selection among pharmacy sites and the value of auditing claims submitted.

Determining drug dispensing costs for use in cost-accounting systems.

Identification of pharmacy costs to be used in a university hospital's cost-accounting system (CAS) is described. At the University of Wisconsin Hospital and Clinics (UWHC), Madison, standard pharmacy labor times for seven categories of products were developed by determining the pharmacist and technician times for purchasing, ordering, transcribing orders, manufacturing, and distributing and administering medications; pharmacy technicians administer most of the medications to patients at UWHC. The labor cost per dose (standard time multiplied by average wage including fringe benefits) was added to drug acquisition cost, which was obtained from the hospital's computerized formulary. The direct costs associated with drug distribution were identified for use in the hospital CAS. These data can be used to compare the cost-effectiveness of various medication administration schedules; they may also be useful in productivity monitoring and flexible budgeting.

Prospective study of replacing administration sets for intravenous therapy at 48- vs 72-hour intervals. 72 hours is safe and cost-effective.

We prospectively studied the safety of replacing intravenous delivery systems, including those used in total parenteral nutrition, at 72- compared with 48-hour intervals in 487 patients. Although the prevalence of contamination of intravenous fluid was higher in administration sets replaced at 72-hour intervals (10/664, 1.5%) than in sets replaced every 48 hours (6/710, 0.8%), the difference is not statistically significant. Contamination in both groups was almost exclusively with small numbers of coagulase-negative staphylococci (range, 1 to 27 colony-forming units/mL); no contaminated infusion was associated with clinical signs of sepsis or concordant bacteremia. Contaminants were recovered less frequently from peripheral venous infusions (0.6%) than from infusions used for central venous access or hemodynamic monitoring (1.5%) or total parenteral nutrition (3.6%); infusions in an intensive care unit were more frequently contaminated (2.5%) than infusions on medical and surgical wards (0.9%). These data indicate that extrinsic contamination of intravenous fluid is a rare cause of endemic nosocomial septicemia, and for most infusions it is unnecessary to routinely replace delivery systems more frequently than every 72 hours.

Application of drug-use review principles to intravenous administration set use: a preliminary study.

A quantitative study of the use of intravenous administration sets in a 556-bed hospital was conducted as a first step in establishing a use-review program for administration sets. Central supply and pharmacy distribution points for i.v. administration sets were identified. Patient use of i.v. administration sets was recorded twice daily for seven-day periods on each of six patient-care units; data for all units were collected within a three-month period. Use of six common administration sets was studied. For each patient-care unit studied, the majority of sets used were of two of the six types; the specific set types differed by unit. More sets were used between the hours of 0801 and 2000. The number of sets per i.v. patient day (one patient receiving i.v. therapy on one day) was greatest in the intensive-care units. Through application of components of drug-use review, the numbers and types of i.v. administration sets used in various patient-care areas were identified.

Effect of use-criteria for intravenous administration sets.

Numbers of intravenous administration sets used were studied in a 480-bed hospital before and after criteria for selecting sets were implemented. Criteria for use of eight common administration sets were developed by a committee of nurses and pharmacists. Hospital drug-use review was used as a pattern. Baseline data were collected for seven-day periods on each of five patient-care units; twice daily, one pharmacist recorded the numbers of sets used and determined their compliance with the criteria, which had been established but not disseminated to nurses or implemented. After nurses were instructed on use of the criteria, data were collected for a second seven-day period on each unit. Use of each set type as a percentage of all sets and compliant use of each type as a percentage of all compliant use were calculated. Annual hospitalwide cost savings were predicted from the observed changes in use and the hospital's costs in the six-month period before the study. Significant changes occurred in overall use and in use of individual set types. For each set type, use compliant with the criteria increased. Notable changes included decreased use of volumetric chambers (burettes) and increased relative use of Y maxidrip and Y minidrip sets. Annual savings of more than +20,000 were predicted. Criteria for selection of i.v. administration sets had an effect on set use at this hospital.

Cost comparison of unit dose and traditional drug distribution in a long-term-care facility.

Unit dose and traditional drug distribution systems were compared in a 352-bed long-term-care facility by analyzing nursing time, medication-error rate, medication costs, and waste. Time spent by nurses in preparing, administering, charting, and other tasks associated with medications was measured with a stop-watch on four different nursing units during six-week periods before and after the nursing home began using unit dose drug distribution. Medication-error rate before and after implementation of the unit dose system was determined by patient profile audits and medication inventories. Medication costs consisted of patient billing costs (acquisition cost plus fee) and cost of medications destroyed. The unit dose system required a projected 1507.2 hours less nursing time per year. Mean medication-error rates were 8.53% and 0.97% for the traditional and unit dose systems, respectively. Potential annual savings because of decreased medication waste with the unit dose system were $2238.72. The net increase in cost for the unit dose system was estimated at $615.05 per year, or approximately $1.75 per patient. The unit dose system appears safer and more time-efficient than the traditional system, although its costs are higher.

Comparative analysis of drug distribution costs for controlled versus noncontrolled oral analgesics.

Total costs for controlled substance oral analgesics and non-controlled analgesics were compared for patients at a 548-bed university hospital. During 1983, all cost elements involved in drug delivery (excluding large-volume parenterals) were identified. Direct and indirect pharmacy labor costs were determined. Personnel costs were calculated from time studies of nurses (in 1979-80) and pharmacy technicians (in 1982). Other pharmacy costs, based on the hospital's 1982 data, included inventory holding costs, computer services, supplies, and drug acquisition costs. Costs were calculated for four oral analgesics--acetaminophen with codeine, aspirin with codeine, ibuprofen, and zomepirac sodium--used during a 30-day period in 1981. For all medications, total average cost per dose for 1,949,418 doses was $2.44, of which 41% was drug acquisition cost. Personnel costs for pharmacy and nursing accounted for 43% and 11%, respectively, of total costs. For 46% of 5111 oral analgesic doses, frequency of administration was at least four times daily. Average purchase cost per dose for the oral analgesics was $0.15, while total costs for the controlled and non-controlled drugs were $1.02 and $0.50, respectively. For the four oral analgesics in this study, cost was affected by dosage schedule and controlled or noncontrolled status. Calculation of the total average cost per dose is useful in projecting annual costs and in identifying areas for cost reduction.

Work analysis of an admixture service.

The use of work analysis to identify problems and measure the effect of procedural changes in a hospital pharmacy i.v. admixture service is described. Problems in the work flow in the sterile-products area (SPA) were identified by process analysis, and a work-sampling study was conducted to quantify the time spent in various activities. The number of unused and expired large-volume parenterals (LVPs) that were received by the pharmacy during the work-sampling period was determined, and the lag times between writing of LVP orders and preparation of these LVPs were measured. After reviewing the major activities in the SPA, four procedural changes (automated label typing, use of a new total parenteral nutrient (TPN) base solution, batch production and storage of antibiotic minibags in a freezer, and scheduling changes in the SPA) were implemented. Subsequently, the percentage of time spent in label typing, preparation of TPN solutions, and minibag production decreased. The small decrease in returned LVPs attributed to scheduling changes was related to the failure of decentralized pharmacists to notify SPA personnel about discontinued orders. Monitoring and enforcement of policies regarding notification of discontinued i.v. orders were recommended to management as a means of increasing work flow efficiency. Work analysis was helpful in evaluating and improving the efficiency of an i.v. admixture service. This technique can be applied to other areas of the pharmacy department as well.

Preparing to implement.

The steps involved in preparing to implement a pharmacy computer system are described. Although the order in which the various steps have to be completed varies according to the hospital specifics, important elements of the preparation process include preparation of the site, organization of personnel, development of policies and procedures, developing and ordering forms, preparation of the implementation schedule, file development, and preparation of other departments. A well-designed plan and a realistic time schedule for the preparation process facilitates the actual implementation of a pharmacy computer system.

Primary nursing and pharmacy systems: are they compatible?

The effects of primary care nursing on the provision of pharmaceutical services are discussed. The major pharmacy-related problem caused by primary care nursing is the increased complexity of communications and drug distribution. Decentralized pharmaceutical services and pharmacy-coordinated drug administration programs can be used to overcome the problem. Primary care nursing's advantages are improved control over controlled substances, quicker access to accurate patient information, more complete patient-care plans and drug-charge records, and a more efficient discharge medication program. Understanding the intent of primary care nursing is fundamental to maximizing these benefits and avoiding problems with this type of nursing service.