Calclum and phosphorus compatlblllty/Computer appllcallons
`
`birth weight infants: the absorption of calcium and fat. Pedi(cid:173)
`atrics. 1976; 57:16-25.
`9. Rowe JC, Wood DH, Rowe DW et al. Nutritional hypophos(cid:173)
`phatemic rickets in a premature infant fed breast-milk. N Engl
`J Med. 1979; 300:293-6,
`10. Kulkarni PB, Hall RT, Rhodes PG et al. Rickets in very low
`birth weight infants. J Pediatr. 1980; 96:249-52.
`11. Day GM, Chance GW, Radde IC et al. Growth and mineral
`metabolism in very low birth weight infants. II. Effects of cal(cid:173)
`cium supplementation on growth and divalent cations. Pediatr
`Res. 1975; 9:568-75. ·
`
`12. Knight PJ, Buchanan S, Clatworthy W Jr .. Calcium and phos(cid:173)
`phate requirements of preterm infants who require prolonged
`hyperalimentation. JAMA. 1980; 243:1244-6.
`13. Schuetz DH, King JC. Compatibility and stability of electro(cid:173)
`lytes, vitamins, and antibiotics in combination with 8% amino
`acid solution. Am J Hosp Pharm. 1978; 35:33-44.
`14. Weast R, ed. Handbook of chemistry and physics. Cleveland:
`CRC Press; 1976:B-100.
`15. Henry RS, Jurgens RW Jr., Sturgeon Retal. Compatibility of
`calcium chloride and calcium gluconate with sodium phosphate
`in a mixed TPN solution. Am J Hosp Pharm. 1980; 37:673-4.
`
`Review Article
`
`Am J Hosp Pharm. 1982; 39:53-70
`
`Review of corr1puter applications in insti'tutional pharmacy-197 5-
`1981
`
`Ken W. Burleson
`
`A literature review of computer applications in institutional pharmacy, covering papers pub(cid:173)
`lished from 1975 to 1981, is presented.
`Articles are categorized as computer concepts, applications to administrative functions, con(cid:173)
`trollcd substances, drug distribution systems (including on-line and off-line services, intrave(cid:173)
`nous admixture services, and ambulatory services), drug information, clinical services (includ(cid:173)
`ing drug-use review, drug interactions and therapeutic incompatibility surveillance, and phar(cid:173)
`macokinetics), and pharmacy-related applications developed by nonpharmacists.
`Before 1975, computer applications in institutional pharmacy reported in the literature were
`largely single-use applications. After 1975, many reports described the integration of individu(cid:173)
`al applications into sophisticated systems that supported many functions. There is still a need
`for good cost justification studies of computerization in pharmacy.
`
`Index terms: Administration; Automation, data processing, computers; Controlled sub(cid:173)
`stances; Drug distribution systems; Drug information; Drug interactions; Incompatibilities;
`Pharmacy, institutional
`
`In the past 20 years electronic data processing (EDP) in
`hospital pharmacy has grown from applications that im(cid:173)
`proved accounting procedures to sophisticated multifunc(cid:173)
`tional, integrated systems for institutional drug control and
`clinical pharmacy support. Early innovators were hospital
`pharmacists who applied computers to accounting and
`billing functions. However, as pharmacists became aware
`of the benefits of automation and as they gained expertise
`in the field, applications of EDP became varied. Innovative
`approaches to pharmacy practice have been instituted that,
`
`Ken W. Burleson is Director, Pharmacy Services, Catawba County Memorial
`Hospital, Hickory, NC 20601.
`The assistance of the librarians of the Area Health Education Center
`(AHEC), Catawba County Memorial Hospital, in performing the literature
`searches; Ethel Hall for secretaria! help; and Margery Adams, M.S,N., for
`critica! review is acknowledged.
`
`Copyright © 1982, American Society of Hospital Pharmacists, Inc, All rights
`reserve d.
`
`without automation, would be too time-consuming, too
`costly, or too difficult to implement. Automated drug control
`systems, medication delivery systems, and support of clinical
`services are examples of these applications. Interest in au(cid:173)
`tomation for pharmacy practice has stimulated vendors of
`commercial systems to develop hardware and software
`packages designed for pharmacy.
`The expanded use of electronic data processing in phar(cid:173)
`macy practice has been due to both the development of more
`sophisticated hardware and software during the past 20 years
`and the experiences of individual practitioners in applying
`EDP to various segments of practice. The literature has
`contributed substantially to the increased awareness of the
`individual pharmacist of the benefits of automation. In 1975,
`Knight and Conrad1 published an extensive review of
`pharmacy applications of electronk data processing made
`to that time. This article reviews those applications made
`from 1975 to the present.
`
`0002-9289/82/01o1-0053$04.50
`
`Vol 39 Jan 1982 American Journal of Hospital Pharmacy 53
`
`Wockhardt 1012
`
`

`
`Computer appllcatlons
`
`Computer Concepts
`Data processing concepts and technology are areas in
`which few pharmacists have had formal training. They
`should become familiar with fundamentals of systems
`analysis, design, and computer technology prior to involve(cid:173)
`ment with application development. A number of articles
`have described the basic concepts of data processing for the
`pharmacist.
`Nelson2 presented an introduction to computer hardware
`that could be used in a pharmacy system. He described the
`various hardware components, including the central pro(cid:173)
`cessing unit (CPU), input devices, and data storage devices.
`Advantages and disadvantages of different components were
`presented. Downtime, system security, and vendar systems
`evaluations were discussed. He also presented a dictionary
`of computer terminology. Mehl3 outlined the mínimum re(cid:173)
`quirements for a pharmacy data processing system. He
`compared advantages of centralized versus decentralized
`systems, and examined the methods of data entry and types
`of drug coding for developing a data base. He also empha(cid:173)
`sized the need for arder verification to prevent errors.
`Computer hardware configurations range from the large
`mainframe computer to the minicomputer to the most recent
`development in the field, the microprocessor. Given the
`premise that the pharmacist has a choice in the selection of
`hardware, an understanding of the advantages and disad(cid:173)
`vantages of each can be essential to the development of a
`successful application. Knowles4 explained the differences
`between mainframe systems and minfoomputers. Lauer et
`al. 5 explained the apparent and subtle differences between
`a mainframe computer shared with other users (shared
`system) anda stand-alone dedicated minicomputer system
`for pharmacy, particularly in regard tó ambulatory phar(cid:173)
`macy practice. Advantages and disadvantages of each con(cid:173)
`figuration were presented. Data storage was found to be the
`most serious drawback to a stand-alone system, a disad(cid:173)
`vantage that could be minimized with a properly designed
`system. The authors concluded that in most situations for
`pharmacy practice, either configuration could provide ad(cid:173)
`equate support.
`Another configuration is a pharmacy application devel(cid:173)
`oped as a part of a total hospital information system (HIS).
`Ballet al.6 examined the past, present, and future develop(cid:173)
`ments in hospital data processing systems, from stand-alone
`pharmacy systems to large hospital information systems.
`The authors predicted that in the immediate future, many
`physicians would have terminals in their offices interfaced
`with hospital information systems. Mecklenburg7 described
`the expanding applications of hospital information systems,
`including pharmacy and other clinical applications.
`An important concept for the pharmacist to understand
`befare developing an application is the methods used to
`justify the need for and cost of a computerized system.,Al(cid:173)
`though many articles have been written describing the varied
`applications of computerization, few articles have described
`controlled documented studies to justify the cost and eval(cid:173)
`uate the effects of a computerized system. The fact that
`evaluation of systems and intensive cost-benefit studies have
`
`54 American Journal of Hospital Pharmacy Vol 39 Jan 1982
`
`not been accomplished may be a majar reason why there has
`not been a greater acceptance of pharmacy systems by hos(cid:173)
`pital administrators or pharmacists. In 1975, Gouveia8 re(cid:173)
`viewed the few studies to date that had attempted to analyse
`the effects of computerization on hospital costs, medication
`errors, and patient care. He found that the few studies
`published actually raised more questions than they an(cid:173)
`swered. He emphasized the need for research to establish
`adequate cost-benefit ratios to justify computerization to
`hospital administrators, patients, and third party payers.
`Since that time, other studies have been published de(cid:173)
`scribing the steps in volved with analyzing and justifying the
`need for and cost of computerization. Freibrun9 analyzed a
`traditional pharmacy system in a 360-bed hospital. He
`identified procedures needing improvements and examined
`alternate manual and automated approaches for change;
`areas in which automation would offer potential savings;
`developed a rating scale for vendors; and described steps
`involved in successful operations analysis in a pharmacy.
`Kay et al. 10 described the method used to analyse a hospital
`pharmacy's need for automation and identify the various
`areas where automation would benefit both pharmacy and
`the hospital. The impact of the proposed system upon other
`areas of the hospital were also listed. Cost justification for
`a dedicated mini-computer was developed. The authors were
`successful in justifying automation of the pharmacy de(cid:173)
`partment, based upon a potential cost savings and an im(cid:173)
`proved. medication delivery system.
`Neal11 developed an in-depth cost proposal to justify to
`hospital administration the computerization of a hospital
`pharmacy. He identified seven areas of tangible cost savings
`(reduction in costs, elimination of salaries paid, increased
`revenues) and two areas of intangible savings (reduced
`overhead, labor reallocation). He was able to project tangible
`dallar savings to each of these areas. He found that auto(cid:173)
`mation of the department would result in a substantial cost
`savings. He also described the various steps in developing
`and presenting the analysis to hospital administration.
`Gray12 evaluated the cost of computerization of an i.v. ad(cid:173)
`mixture service in a hospital pharmacy. Staffing analysis and
`life cycle cost projection were determined. Workloads and
`staffing patterns both with and without the computer were
`calculated. The basis of the study was to determine the
`amount of money that could be invested in a computer sys(cid:173)
`tem as justified by staffing reductions and other savings. The
`author's conclusion was that the computer was éost effective
`and, therefore, should be purchased.
`Lauer13 gave an overview of the need for automation in
`pharmacy practice and the benefits to be realized from
`computerization. Three areas of savings as a result of auto(cid:173)
`mation-time, space, and personnel costs-justified the cost
`of computerization.
`Two authors have described the methods of dealing with
`vendors of computer systems. Olsen et al.14 described the
`method used to select an upgraded computer system that
`wouldl support an automated clinical department in the
`hospital. Although the article dealt specifically with an au(cid:173)
`tomated laboratory system, the auth6rs presented a general
`
`Wockhardt 1012
`
`

`
`discussion of vendors, vendor selection, systems require(cid:173)
`ments, terminal requirements, and hardware and software
`that could be used by the pharmacist in selecting a pharmacy
`system. Cutely15 presented an extensive list of questions a
`pharmacist should ask a hardware or software vendor of a
`commercial system when considering the purchase of such
`a system.
`Before beginning development of any automated appli(cid:173)
`cations, the pharmacist must develop a data base, or drug
`file, listing all the drugs to be found in the pharmacy, along
`with any information pertinent to the description of each
`drug. The method in which a data base is developed can
`mean the difference between a flexible system with the
`ability for expanded applications anda rigid, single appli(cid:173)
`cation system. Hanson et al. 16 have described the develop(cid:173)
`ment of a master drug file that was developed by examining
`an existing computerized drug data file to determine which
`existing fields of information should be retained for the new
`data base. The new data base was developed to support in(cid:173)
`creasingly sophisticated pharmacy applications. Twenty(cid:173)
`seven data fields for the master drug file were identified.
`Programs were written to permit entry and maintenance of
`the file by using punched cards input toan offline computer.
`The authors envisioned using online entry of data through
`a cathode ray tube (CRT) in the future. Programs were
`written that permitted machine verification of data ac(cid:173)
`cording to predefined specifications. Any errors detected
`were rejected for correction. This editing feature resulted
`in a high degree of accuracy of the stored data. Strand et al.17
`developed a master drug file after comparing commercially
`available data bases which they found to be deficient.
`Twenty-seven different data fields were identified and in(cid:173)
`formation for each drug entered to a coding form which was
`used for data entry. Entry was online via a cathode ray tube.
`The data base was used to support certain administrative
`and drug distribution programs for both inpatient and
`outpatient services. The authors also examined the cost of
`the development of the file. They found that more than 900
`hours were involved with the development, ata total salary
`cost of $8451. This calculated to $7.43 per line ítem in the
`data base. Although this cost was twice that of a commer(cid:173)
`cially available data base, the authors thought that the ad(cid:173)
`ditional data fields that were available to them justified the
`cost.
`The American Society of Hospital Pharmacists provides
`a comptiter-generated, machine-readable data base called
`Drug Products Information File (DPIF)ª for use by phar(cid:173)
`macists in computerized systems. Frankenfeld18 examined
`the problems associated with the National Drug Code (NDC)
`system as a pharmacy data base and the potential for in(cid:173)
`terfacing it to DPIF. He explained the advantages of cross(cid:173)
`referencing the information in the two files.
`
`Administrative Applications
`
`Because of the computer's inherent ability to quickly
`tabulate numerical data, and to store, retrieve, and compile
`statistical information, certain administrative functions are
`
`Computcr appllcatlons
`
`ideally suited to automation. Among these functions are
`patient billing and accounting, drug use review, and inven(cid:173)
`tory control. A number of articles have described applica(cid:173)
`tions of EDP in these areas.
`Silverman 19 described the administrative functions that
`could be automated using a dedicated minicomputer.
`Among these applications were personnel management,
`patient billing and accounting, and inventory control.
`Wuest and Schaengold2º described an automated ac(cid:173)
`counting system shared by two hospital pharmacies, using
`a time-shared computer system. Data were entered from
`dispensing records showing all transactions for each phar(cid:173)
`macy. The system generated a monthly report of expenses
`for chargeable patient drugs and nonchargeable floor stock.
`Drug use statistics from this report were used for purchasing
`and inventory control. The system also printed a formulary
`for each hospital.
`In a hospital without data processing capability, Elliott21
`contracted to use the computer services of a local drug
`wholesaler to develop an inventory and purchasing system.
`The wholesaler's programs for inventory control were
`modified to adapt to the special needs of the hospital. All
`drug issues to stock from inventory were manually recorded
`on an inventory master list by a clerk and this was sent to the
`wholesaler for keypunching into the system. A weekly
`computer-generated report summarized the use of each item,
`and this list served as a stock status report and purchase list.
`Each item that had reached a predetermined order point was
`flagged. Items supplied by the wholesaler were automatically
`shipped and entered into the computerized inventory. Or(cid:173)
`ders to direct vendors were completed by the pharmacist,
`working from the report. The system also generated a hos(cid:173)
`pital formulary by use of therapeutic category coding. Both
`an alphabetical listing and a listing by therapeutic categories
`were available.
`Pickup et al.22 utilized the Massachusetts General Hos(cid:173)
`pital Utility Multi-Programming System (MUMPS) to de(cid:173)
`velop programs to control ward stock levels and contribute
`to workload analysis in a quality control section of a hospital
`pharmacy. The system was programmed to determine each
`hospital ward's minimum stock levels, based upon historical
`demand and the ability of the pharmacy to respond to the
`needs ofthe various wards. Results showed that the system
`could reduce the inventory of drugs on the nursing units,
`thereby effecting a cost savings, without any deterioration
`of service or inconvenience to the nursing units. The system
`also handled data concerning raw materials in the quality
`control section. It was determined that a time savings could
`be realized by automating sorne of the reports in this area.
`Automated patient billing has been a feature of hospital
`computer systems for many years. This is one of the earlier
`applications to which EDP was applied in pharmacy. Tru(cid:173)
`deau23 modified an existing tinie-shared payroll and ac(cid:173)
`counting system to provide drug use review and patient drug
`billing. The time saved from these applications was used to
`1
`permit the pharmacy to complete a hospital-wide traditional
`unit <lose system. The author did not,use the computer di(cid:173)
`rectly to support the unit <lose system. Priest24 used com-
`
`Vol 39
`
`Jan 1982 American Journal of Hospital Pharmacy 55
`
`Wockhardt 1012
`
`

`
`Computer appllcatlons
`
`puter-printed gummed labels to be attached to intravenous
`fluids and other pharmacy patient charge items that were
`kept as floor stock on the nursing units. These gummed
`stickers functioned as a charge voucher as well as a stock
`replenishment mechanism. The author concluded that the
`system aided in the capture of more charges, while simul(cid:173)
`taneously saving personnel time.
`Fish25 described.a computerized patient billing system
`that was based on a combination of a percentage markup of
`drug costs plus adose fee. Seven different dose fees (factors)
`were used, depending upon the type.of drug product ad(cid:173)
`ministered to the patient; e.g., oral unit dose, injectable unit
`dose, and i.v. additives. Manual patient profiles were used
`for a unit dose medication system, and cumulative charges
`for each patient were maintained on these profiles. At the
`time of patient discharge, the profile for the patient was
`inactivated and all drug charges were added. A pharmacy
`technician also éntered the patient number, computer drug
`code, and dose factor for all drugs. The profile was then sent
`to the pharmacy pricing area, where the charges were entered
`in to the computer vía a cathode ray tu be. A final patient bill
`was produced as a result of data entry. The system offered
`advantages of an accurate, itemized statement; charges were
`equitable, based upon the type of drug administered; and
`the system produced useful statistical reports. The time
`required to enter charges manually into the system was a
`drawback, and the author proposed an automated unit dose
`system that would eliminate much of the manual data
`entry.
`Gurtel et al,26 projected drug use review statistics for a
`pharmacy and therapeutics committee to use in determining
`the benefit of adding a new drug to the formulary of an am(cid:173)
`bulatory patient care clinic. A computer-supported ambu(cid:173)
`latory pharmacy system was used to determine if a new drug,
`ticrynafen, a diuretic with. uricosuric properties, would
`benefit patients in the clinic. The computerized patient
`profile was used to determine the number of patients taking
`a diuretic who were also taking a uricosuric agent, to deter(cid:173)
`mine how many patients could benefit from the new drug
`whiéh offered both therapeutic actions. Computer analysis
`revealed that only 8% of the patients on a diuretic were si(cid:173)
`multaneously receiving a uricosuric agent. In view of the cost
`of the new drug and the limited application, as shown by
`computer analysis, the pharmacy and therapeutics com(cid:173)
`mittee chose to add the drug only on a restricted formulary
`status. The drug was eventually prescribed for four patients.
`Later, the drug was recalled from the market because of its
`adverse reactions. The computer was used to search the
`patient profiles for those patients receiving the drug at the
`time of recall, so that their physicians could contact them
`and make appropriate changes in therapy. The authors
`concluded that the use of the computerized patient infor(cid:173)
`mation system enabled the pharmacy and therapeutics
`committee to prevent the potential exposure of more than
`160 patients to the ad verse effects of the drug.
`Not all administrative applications of data processing
`must be developed on a computer. Word processing equip(cid:173)
`ment is similar to a computer, with the exception that word
`
`56 American Journal of Hospital Pharmacy Vol 39 Jan 1982
`
`processing equipment ordinarily has no built-in logic. The
`system is used for storage of small amounts of data and re(cid:173)
`trieval and printing of this information on a repetí ti ve basis.
`Letcher27 compared three different commercial brands of
`word processing equipment to various applications in hos(cid:173)
`pital pharmacy. The applications studied were label pro(cid:173)
`duction; storage of personnel information; scheduling of
`repetitive tasks; and composition of narrative information,
`such as drug bulletins and procedure manuals. The evalua(cid:173)
`tion included keyboard design, disk storage capabilities,
`software, print format, and security of data. One of the three
`systems was clearly superior to the other two because of its
`flexibility o.f applications. The author summarized the re(cid:173)
`sults of the evaluations of each machine.
`
`Controlled Substances Applications
`
`Controlled substances are defined as those drugs which
`have the potential and liability for abuse; i.e., narcotics and
`barbiturates. Federal and state laws require that practi(cid:173)
`tioners who dispense these medications maintain records of
`disposition for all drugs under this regulation. Because ofthe
`large number of drugs in this category, proper recordkeeping
`has been time-consuming for both the pharmacist and
`nursing personnel. Automation of this segment of practice
`can reduce time involvement for both the pharmacist and
`nurse, while maintaining accurate control and accountability
`records as required by law.
`Petoletti28 used an off-line system in an outpatient clinic
`to monitor for potential abuse of controlled. drugs by pa(cid:173)
`tients. Dispensing data were entered on a source document
`for each prescription dispensed. These data were key(cid:173)
`punched weekly, and reports were generated that notified
`the pharmacist of those patients receiving excessive supplies
`of controlled drugs.
`McDanieF9 modified an existing patient accounting sys(cid:173)
`tem to develop an automated recordkeeping system. All
`controlled drugs were assigned specific service corles within
`a designated group of service numbers. A separate file was
`set up in the computer for this group designation. As charges
`were posted to the patient's account, records of controlled
`drugs dispensed were created. Daily and cumulative monthly
`reports were printed, which showed distribution of con(cid:173)
`trolled drugs. Nazzaro30 developed a program on an off-line
`computer to provide accurate records for controlled sub(cid:173)
`sta.nces accountability, while reducing manual transcriptions
`involved with record maintenance. The system was used for
`both inpatient ward stock and outpatient prescriptions in
`a military hospital. All prescription transactions were re(cid:173)
`corded manually on punched cards, and included patient
`number or hospital ward corle, physician's identification
`code, drug corle, and quantity of drug dispensed. All data
`were later keypunched and entered into the computer.
`Various records were generated by the system, including
`perpetual records of each drug by patient or ward, monthly
`inventory of all controlled substances, and ward monitoring
`lists of excess stock of controlled substances. The system
`could also search for prescriptions by patient or prescriber.
`
`Wockhardt 1012
`
`

`
`Shaver et al.31 described a system for an inpatient and out(cid:173)
`patient military hospital pharmacy which used limited
`computer hardware. The system was run on a remote
`mainframe computer through a telephone hookup. A phar(cid:173)
`macist or technician entered all transactions daily via a
`cathode ray tube. After data entry was complete, all trans(cid:173)
`actions were verified before the update program was run, to
`ensure accuracy of information. The system generated a
`number of reports, including transactions by drug and cur(cid:173)
`rent inventory balances. It also permitted patient drug use
`screening and physician prescribing screening to monitor
`for potential drug abuse. The system was capable of tying
`in terminals at other military hospitals.
`Finally, Dickinson32 reported how the Drug Enforcement
`Agency used computers to map entire states to show drug
`distribution, to pinpoint areas ofpotential drug abuse. Data
`are obtained from two sources-the Automated Reporting
`and Consolidated Order System (ARCOS), which shows
`drug distribution from manufacturers and wholesalers to
`pharmacies; and the Drug Abuse Warning Network
`(DA WN), which tracks drugs from selected hospital emer(cid:173)
`gency rooms. Ali data were entered into the computer and
`analyzed. The resultant output was distributed to DEA field
`offices and other local. and state law enforcement officials
`for follow-up.
`
`Drug Distribution System Applications
`
`Traditional manual drug distribution systems are time
`consuming, involve much clerical work for both pharmacist
`and nurse, and tend to be error-prone. Automation of the
`medication cycle can provide substantial benefits to the
`pharmacist, nurse, and patient by reducing the amount of
`clerical work involved with maintaining a medication system,
`reducing errors, improving administrative control, and
`freeing the pharmacist for more clinical involvement. Much
`work has been done in automating various segments of the
`medication cycle. Many pharmacists have automated one
`or more procedures involved in medication delivery. How(cid:173)
`ever, prior to 1975, only a few systems had integrated the
`various components ofthe cycle into a completely automated
`medication delivery system. Since that time, several articles
`have described the development and application of total
`systems for automated medication delivery. This increased
`development has been due, in part, to the reduced cost of
`hardware necessary to support an automated medication
`delivery system and to the entry of vendors that provide
`hardware and software packages. Yet, the use ofEDP in the
`medication cycle is not extensively employed by hospitals.
`Stolar, 33 in a 1978 national survey of hospital pharmacies,
`found that of the 738 reporting hospitals, only 13% of the
`large hospitals and 5% of the small hospitals used computer
`systems in the drug dispensing process.
`The importance ofEDP in hospital drug delivery syste~s
`and the role of the pharmacist in implementing its use has
`been recognized by the American Society of Hospital
`Pharmacists (ASHP). The ASHP Statement on Hospital
`Drug Control Systems34 states, "The pharmacist should
`utilize EDP to decrease the many traditional paper-handling
`
`Computer appllcatlons
`
`chores so that his clinical role may be effectively expanded
`and his talents utilized properly." This position statement
`also outlined the many applications of EDP to pharmacy
`practice and the role of the pharmacist in systems develop(cid:173)
`ment.
`Off-line Medication Systems. Prior to 1975, many
`automated medication distribution systems were developed
`utilizing off-line systems, usually by modifying an existing
`batch process accounting system to provide pharmacy ap(cid:173)
`plications. In recent years, more and more pharmacists have
`had access to on-line computer systems, and only a few au(cid:173)
`thors have described the development of systems using
`off-line computer hardware.
`Swift 3~ described a semi-automated approach to a unit
`dose syste:in, in which punched cards were used to provide
`information for medication cart filling. Each drug order was
`transferred in writing by a pharmacy technician to a
`punched card, called the master dose card. After verification
`of the transcription by a pharmacist, the data were key(cid:173)
`punched onto cards. The cards for each patient's medica(cid:173)
`tions were then placed in the proper medication cart drawer.
`These cards were color-coded by drug type and alphabetical
`name to simplify cart filling. The technicians filled a 24-hour
`supply of medication from the information on the cards. The
`daily charge data were then entered manually on each card
`for the amount of drug dispensed. After the pharmacist
`checked the medications in the cart, using the master dose
`cards, the cards were sent to the data processing center for
`daily charging. The cards were then returned to pharmacy
`for subsequent use in the medication system. The system was
`cumbersome, since the cards were often misplaced and daily
`maintenance of the patient records involved a substantial
`amount ofpaper handling. The cards were not used to gen(cid:173)
`erate a patient profile; therefore, monitoring of patient
`medication records was not possible.
`Gilbert et al.36 described a batch mode order entry pro(cid:173)
`cedure on an off-line computer. The pharmacist reduced all
`drug orders to numeric codes on punched cards. Once each
`eight hours, the coded orders were batch keypunched. The
`computer printed a drug distribution log for unit dose cart
`filling, a cumulative patient profile, anda daily medication
`charting document for nursing. Automated patient charging
`and census control were features ofthe system. The author
`overcame the lapse between profile printings by sending
`doses ofmedications for new orders to cover the interim until
`the next cart exchange.
`On-line Medication Systems. On-line systems allow the
`operator to interact directly with the computer, allowing
`immediate access to data stored, thus overcoming the time
`lapse in information processing which occurs with off-line
`systems. Thus, on-line systems lend themselves to more
`flexible programming. For this reason, on-line medication
`systems have usually involved more sophisticated applica(cid:173)
`tions. A number of such systems have been successfully
`implemented, both as dedicated pharmacy modules andas
`subsystems of larger hospital. information systems.
`A series of articles has described the medication distri(cid:173)
`bution system at The Johns Hopkins Hospital. Two gener-
`
`Vol 39 Jan 1982 American Journal of Hospital Pharmacy 57
`
`Wockhardt 1012
`
`

`
`Computer appllcallons
`
`ations of the system, the first of which was implemented in
`1970, have been developed using a large computer. Derewicz
`and Simborg, in two separate communications,37,38 described
`the second generation of this system, which used cathode ray
`tubes with light-pen entry capability. Each decentralized
`satellite pharmacy had a computer terminal Medication
`orders were entered into the CRTs. Drug orders were auto(cid:173)
`matically checked against patient allergies and diagnosis for
`incompatibility. 'For each hour that medications were due,
`the computer generated a single unit dose envelope. Phar(cid:173)
`macy technicians placed the medications in the envelopes
`and these were delivered to the nursing units. The envelopes
`w