Accuracy Of Data Entry Research Articles

Automated Entry of Hospital Infection Surveillance Data

To assess the accuracy of an automated data entry system employing optical scanning technology and to provide an analysis of its costs as compared to manual data entry.The accuracy and cost of automated data entry of 100 surgical-wound infection surveillance questionnaires was compared to manual entry.The Surgical Directorate, The Royal Hospitals, Belfast, Northern Ireland.The use of optical scanning technology greatly improved the speed and accuracy of data entry. The time spent by the keyboard operator on data entry was reduced substantially. For each surgical-wound infection questionnaire automatically processed, there was a saving in clerical time equivalent to $0.63. The automated data entry process resulted in a 22-fold productivity increase compared to manual data entry with validation. After validation, an error rate of < 0.2 errors per 1,000 responses was detected in automatically entered data compared to a rate of 12.4 errors per 1,000 responses for manually entered data. The automated system, including validation, provided a seven-fold productivity increase compared to "quick-and-dirty" manual data entry without validation.Hospital information technology systems may achieve total integration of data management, but realistically this would appear to be very much in the future. Until then, in view of the accuracy and substantial savings in time and money, we recommend the use of automated data entry technology. This system would be especially useful where data are transported from outlying hospitals to a central receiving center for collation and analysis.

A pilot study of cervical screening in an inner city area--lessons for a national programme.

The objectives of this study were to examine aspects of organization of a proposed national screening programme based in general practice. The target population of women aged 25-59 years and their general practitioners (GPs), in a defined inner city area, was identified from a population register of persons eligible for free medical services; a computerized system was developed for invitations and record linkage of cytology results. Smears were examined in one laboratory and follow up of women with abnormal smears was undertaken by one gynaecologist. A random sample of non-responders was surveyed by questionnaire. Response following two invitations was only 20%. Practices with male doctors only had significantly lower response rates (P < 0.001) than those with a female doctor/nurse. A survey of non-responders showed that over 20% of addresses were incorrect and 16% of those interviewed were ineligible for smear tests. A preference for a female to undertake smears was expressed by 67%, and 77% believed that the purpose of the cervical smear was to detect cancer. An accurate population register, health promotion, support for GP practices, provision of alternative venues for smear tests, development of computer systems, accurate data entry and fail-safe follow up are aspects of a cervical screening service which must be addressed prior to setting up a national service.

An Alternative Method for Computer Entering Large Data Sets

In this article, the authors offer an easy method to enter survey data that results in less fatigue, faster input time, and sharply reduced data entry errors. They discuss its advantages over traditional spreadsheet formats and outline the process they have successfully developed and used with both telephone and written surveys. It is especially beneficial for large data sets, as the survey instrument is virtually reproduced on the computer screen for more accurate data entry.

Testing of the National Crime Information Center Missing/Unidentified Persons Computer Comparison Routine

Abstract For many years the effectiveness of the National Crime Information Center (NCIC) missing and unidentified persons matching program has been in question. The reliability of the NCIC matching system was tested using actual cases. When entered with the original information as provided by the investigating agencies, none of the four test cases were successfully matched. The NCIC system is the only source that links the states with this type of information. The use of NCIC for the entry of missing and unidentified persons information is still recommended, however, entered information should be reviewed by a qualified individual to insure accurate data entry. For accurate matching of information a proven parallel system should be used by every state.

Improving dental epidemiologic data collection with computers.

A computerized dental data recording system (DDRS) was developed for the New England Elder Dental Study to improve data quality and increase field staff efficiency. The DDRS displays video screens similar to traditional paper forms to record data on coronal and root caries, dentate and denture status, subacute bacterial endocarditis screening, gingival bleeding, calculus, and periodontal attachment level. DDRS provides facilities for date and exam-component time tracking, on-line contextual comments, random record retrieval, editing, data backup, and data output in various data formats. This study compared the DDRS with a paper-form system for data entry accuracy. Dental caries and periodontal disease measurement data from 38 subjects were recorded on paper forms and independently entered using DDRS. The DDRS identified 150 illogical data errors, 39 inconsistent data errors, 7 invalid data and 34 miscellaneous data errors. Four technicians with field experience using both paper forms and DDRS reported time savings using DDRS in the field. DDRS has the potential for additional time savings by minimizing the time for data coding, cleaning, and management. Results demonstrate that DDRS could improve the quality of oral epidemiologic data by mandating strict adherence to protocols, preventing errors, and increasing field efficiency.

WHAT COST ACCURACY?

How accurate are those numbers you rely upon so heavily? Do you have problems getting accurate data entry from your clerks? How do you determine the degree of accuracy of your data entry crew? How accurate must your figures be?

Sources of Human Error in Data Entry Tasks Using Speech Input

Where speech recognition devices are used for entering data, some form of feedback is usually provided so that recognition errors can be detected and corrected. The present experiments were designed to investigate how the accuracy of data entry is influenced by the type of feedback used and by the structure of the error correction dialogue. Experiment 1 compared the effects of visual and spoken feedback delivered either after each item or at the end of the data string. With visual feedback, most failures to achieve correct data entry occurred when subjects failed to detect recognition errors. Spoken feedback was much more closely monitored, but with word-by-word feedback, subjects had difficulty in recovering from mis-recognitions in the error correction dialogue. Experiment 2 was designed to examine possible interactions between presentation and feedback modalities when data strings were entered from memory. Although there was some evidence of an interaction, the effects were small in relation to the much greater disruption of memory caused by correction of recognition errors in midentry. The implications of these findings for the design of feedback and error correction dialogues are discussed.

DEPOT: database for electronics parts and other things

DEPOT has been developed to provide tracking for the Stanford Linear Collider (SLC) control system equipment. For each piece of equipment entered in the database, a complete service, maintenance, modification, certification, location, and (optionally) radiation exposure history can be maintained. To facilitate data entry accuracy, efficiency, and consistency, barcoding technology has been used extensively. DEPOT has been found to be an important tool in improving the reliability of the microsystems controlling SLC. It is now being adopted by other systems at the Stanford Linear Accelerator Center. >

An examination of the efficiency of some quality assurance methods commonly employed in clinical trials

The cost and efficiency of training clinical centre staff and of duplicate data entry in clinical trials is reviewed. Training is an essential component of quality assurance programmes and it is usually carried out at regular intervals in long-term clinical trials. Initial training of staff and regular retraining is important to assure standardization and it can lead to reduced trial costs. Interim training for new staff and for remedial purposes is less efficient than regularly scheduled training sessions. Regional centres for training and the use of computer aided instruction are two ways such interim training can be made more efficient and standardized. Duplicate data entry or verification can result in a substantial reduction in data entry errors depending on the nature of the data being keyed. Selective verification should especially be considered for important fields for which consistency checks cannot be performed, that are alphabetic or that are several characters in length. Quality assurance procedures should be implemented to monitor data entry accuracy in clinical trials.

Of mice and men--data capture in the clinical environment.

We have considered the problem of data capture in the critical care area. Review of previous work on the use of alternative input devices for speed and accuracy of data entry yields conflicting information but suggests that requirements depend on input task, data type and skill of the user. We have performed comparative studies of the QWERTY keybord, cursor control keys, mouse and graphics tablet for data entry in two intensive therapy unit (ITU) environments. The graphics tablet proved overall the best of the devices studied. We report the potential applications of this type of input device.

Accuracy and time requirements of a bar-code inventory system for controlled substances

The effect of a bar-code system on personnel time requirements and data-entry accuracy in an existing automated controlled substances inventory system was determined. In the previous system, technicians used a keyboard and alphanumeric codes to enter into the computer data about the physical transfer of controlled substances among hospital areas. A system for barcode data entry was adapted for use with the existing procedure. After learning to use the bar-code system, four experienced technicians entered data by the keyboard method for eight days and the bar-code method for eight days during a 32-day study period. The amount of time required to enter all transactions and the accuracy of data entry were measured. Mean data-entry times for the keyboard and bar-code methods were not significantly different, most likely because of the greater number of manipulations needed for bar-code data entry. The mean percentage error associated with the bar-code method (0.79%) was significantly less than the error associated with the keyboard method (1.53%). For this particular computer system in which bar-code data entry was adapted to existing procedures, use of bar codes to enter controlled substances inventory data was not substantially faster but was more accurate than a traditional key-board data-entry method.

Uses and limitations of microcomputers for epidemiological research in occupational health.

The rapid improvements in microcomputer systems have contributed to ease of performance of epidemiological research in occupational health. Computers now available can aid in all stages of research- from the early stages of literature review to data collection, analysis and presentation. Optical storage devices now allow large occupational health data bases to be stored on a single compact disc, with the information accessible by microcomputer. Hand-held computers and optical scanners allow paperless collection of field data. Epidemiological analysis and data presentation is easily done with software packages, some of which are in the public domain. However, there are limitations in the use of computers for any epidemiological research. Investment of time is required to learn the use of a computer. Care needs to be taken to ensure accurate data entry, and correct choice has to be made of methods for subsequent statistical analysis. Microcomputers are valuable as labour saving devices, but cannot replace proper planning, execution and appropriate data analysis in epidemiological research in occupational health.

Personal Computers and Relational Databases in an Eye Clinic and Eye Pathology Laboratory

The authors have implemented a sophisticated relational database program on a Macintosh personal computer to assist in organizing and retrieving patient information and clinical diagnoses. The program uses available software and is inexpensive, versatile, and easy to use-particular features of Macintosh computers. Routines are incorporated to encourage accurate data entry when multiple users will input and access the database. This system allows for customized built-in coding schemes, and data transfer can be formatted to interface with IBM compatibles and mainframe computer systems. The authors also review a basic approach to developing a personal database system.

Software Reviews : OVERVUE

Summary: OVERVUE is a potent information manager, well suited for checkbooks, mailing lists, bibliographies, etc. This relational system allows multi-file processing and interacts with WORD, MACWRITE, MULTIPLAN, and other programs. Several innovative features facilitate fast, accurate data entry. It is possible to create and store lengthy analytical procedures operating on numeric and alphanumeric data, as well as on times and dates. The lack of a square root function hampers its use for conventional social research analyses.

Bar codes and their potential applications in hospital pharmacy

The technology and terminology of bar codes, the Health Industry Bar Code (HIBC) Standard, issues facing the pharmaceutical industry with respect to bar codes, and potential applications of bar codes in hospital pharmacy are reviewed. Bar codes consist of a series of parallel alternating dark lines and empty spaces. Commonly used bar codes include UPC (Universal Product Code), Interleaved 2 of 5, Codabar, Code 93, Code 128, and Code 39. There are several different types of scanners that may be used for scanning bar codes. Some of the variable features are portability, contact and noncontact, autodiscrimination, and the presence of numeric keypads. The HIBC Council (HIBCC) is a steering committee charged with developing bar-code standards for health care, assigning identification numbers to manufacturers, and providing information to interested individuals. The HIBCC Board includes representatives from health-care providers, equipment manufacturers, drug manufacturers, and wholesalers. Drug manufacturers are concerned about several issues with regard to bar codes; for example, who will pay for the cost of implementation, on what level of packaging will the bar codes be required, and are there legal implications if a poor scan results in patient harm? Bar codes have already been applied in some hospitals for coding blood containers, roentgenogram jackets, medical records, and capital equipment. The potential applications in hospital pharmacy include inventory control, verifying the accuracy of dispensing to both inpatients and outpatients, and record keeping for drug product expiration or disposal. Bar codes also offer pharmacists the ability to perform in-process dispensing controls that are not practical to perform manually. Bar-code technology offers health-care personnel the opportunity to improve work efficiency and increase the accuracy of data entry into automated systems.

Rapid, accurate, three-dimensional location of multiple seeds in implant radiotherapy treatment planning

The three-dimensional (3-D) locations of the radioactive seeds in an interstitial implant can be obtained from stereo-shift or orthogonal radiographs. When an implant contains a large number of seeds, however, there are often ambiguities in identifying corresponding seed images on two radiographs. The authors' method achieves accurate spatial locations by using three radiographs whose X-ray source positions are (i) in a plane which does not intersect the implant and (ii) widely separated to achieve adequate parallax. All triangulation is done in an anatomical coordinate reference frame attached to the patient so that patient motion does not introduce errors. The authors achieve independent error-insensitive calibration of each radiograph, rapid and accurate data entry in the local coordinates of each radiograph, 3-D seed location in the patient (anatomical) reference frame, and automatic verification of each calculation.

A Comparison of Full and Reduced-Alpha Keyboards for Aircraft Data Entry

CRT's and associated data entry keyboards are being used in aircraft to replace dedicated switches and instruments for entry and display of flight relevant data. This study compared the speed and accuracy of data entry using two different types of alphanumeric keyboards designed for entering data into aircraft mission computers. A full-alpha keyboard has a separate key for each letter of the alphabet as well as each number; a reduced-alpha keyboard (e.g., telephone type) has more than one letter per key. Reduced-alpha keyboards generally weigh less and occupy less space than full-alpha keyboards, but these advantages might be nullified by operational disadvantages. Alpha entries on a full-alpha keyboard require only one stroke per letter, while alpha entries on a reduced-alpha keyboard require at least two keystrokes per letter: one to identify the correct key, and one or more to select the appropriate letter on that key. Because increasing the number of required keystrokes increases both entry time and the opportunity for error, it was hypothesized that alpha entries would take longer and be less accurate on a reduced alpha keyboard. In this study, twelve subjects (eleven of them pilots) entered flight-related information on each of two keyboards. On the reduced-alpha keyboard, numbers and letters were arranged on a 3 X 4 key matrix, with three letters and one number on each key. The full-alpha keyboard had number keys arranged in a 4 X 3 matrix and letter keys (one letter to a key) in a separate 7 X 4 matrix. The keyboards were installed in similar flight simulators and were situated forward and to the left of the operator. For both types of keyboards, entries appeared on the “scratchpad” line of a monochrome CRT just above the keyboard. For all entries on the full-alpha keyboard, this meant that the single number or letter on the selected key appeared on the CRT. For all entries on the reduced-alpha keyboard, the single number on the selected key appeared on the CRT scratchpad line. At the same time, the three letters on the selected key appeared just below the scratchpad line, and just above three unlabeled “alpha select” keys, which were physically distinct from the alphanumeric keyboard. Numeric entries were complete at this point; to finish an alpha entry, the “alpha select” key just under the appropriate letter had to be pressed. After keyboard training and practice, each subject entered flight initialization information, a flight plan, and flight plan modifications for each of two “flights” using one of the keyboards. This required a total of approximately 80 alpha and 240 numeric entries. This procedure was repeated two days later on the other keyboard for two different “flights.” Order of keyboard and “flights” was counterbalanced. Performance measures were the average entry time (computer recorded) and the number of errors for alpha and numeric characters on each type of keyboard. As expected, alpha entries were significantly slower on the reduced-alpha keyboard (x‾ = 3.84 sec) than on the full-alpha keyboard (x‾ = 2.28 sec), since they required two keystrokes. Contrary to expectation, numeric entries were also significantly slower on the reduced-alpha keyboard (x‾ = 1.70 sec) than on the full-alpha keyboard (x‾ = 1.18 sec). This result may have been due to the confusion involved in making two different types of entries (with different procedures) using the same set of keys for the reduced-alpha keyboard, while alpha and numeric entries were physically separate and identical in procedure on the full alpha keyboard. Errors were minimal for both types of entries on both types of keyboards. A total of 26 errors in numeric entries were made on each type of keyboard. There were 30 alpha entry errors on the reduced alpha keyboard and nine on the full-alpha keyboard. There was no evidence of a speed-accuracy tradeoff in entries on the two keyboards since entries were faster and as accurate on the full-alpha keyboard. Whether the entry speed advantage of the full alpha keyboard would outweigh its greater size and weight would depend on design factors and the time criticality of necessary entries for a specific application.