LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES
paraneoplastic neuropathies. Having a readily analyzable
detailed electronic record of all serial examinations of a patient’s
neuropathy is useful not only for treatment decisions but also for
later research needs.
These results, therefore, indicate that scanned programmed
assessment with adequate quality control tends to be more
accurate, although both approaches give accurate final results.
With respect to speed of assessment of case report forms, our
results are different from what has previously been reported.
Guerette et al5 compared the time needed and the error rate of
data scanned or manually entered by keyboard into database and
did not find a difference. These conclusions were supported by
another study done by Nies et al.12 By contrast, we found a major
savings in time (a five-fold improvement) by using scannable
records and interactive programmed analysis. Strictly speaking,
our analysis should not be compared to these earlier studies
because more than accuracy of entered data by double keyboard
entry versus scanning entry was involved in our comparison. We
used CNA to monitor disease and study criteria and concordance
of test data not compared in these earlier studies.
DISCUSSION
Scannable medical records whose information can be
automatically read, evaluated, summarized, transferred and
stored in database, are useful for a variety of medical purposes.
The approach may be used as an update of a patient’s medical,
family, drug, treatment, physical fitness, life-style choices and
risk-factor history prior to an event of clinical evaluation and
care.1 More typically it may be used to develop a database useful
in medical billing2 or for capture of data for later study of patient
cohorts, disease classification, medical events, practices,
treatments or outcomes.3-9 The approach has also been used to
study environmental and occupational health issues10 or cancer
prevalence and risk factors.11
Scannable case report forms and automatic reading,
evaluation and transfer of information to database are also being
introduced into pharmaceutical trials. At a former time, and even
now, clinical report forms were developed for a specific trial.
Data were manually entered, checked for completeness and
errors at study sites and periodically by a visiting clinical
research associate from the sponsoring company and then sent to
the pharmaceutical house for manual double entry by keyboard
and reconciliation of data and then entered into database. The use
of scannable forms and the necessary hardware and computer
software approaches have made it possible to create, modify,
maintain, archive, retrieve, or transmit clinical data intended for
submission to the Food and Drug Administration (FDA)
(Guidance for Industry Computerized Systems Used in Clinical
Trials, April 1999).
In summary, when there is a need to record, transfer and enter
into database standard characterizing and clinical data on large
numbers of patients, scannable records with interactive
programmed analysis provide improved accuracy and speed and
may have other advantages over normal double entry of data and
may be cost effective.
ACKNOWLEDGEMENTS
We thank the physicians who have participated, or are participating
in trials using CNA. The Genentech trial has been published and lists the
names of co-investigators (Apfel SC, et al13). Jane Norell, B.S. and Carol
Overland (CO) performed the timed studies and Mary Lou Hunziker
prepared the manuscript.
This work was supported in part from grants received from the
National Institute of Neurologic Disorders and Stroke (NS36797).
Financial support was also received from Genentech, Inc (San
Francisco, CA), Roche, Inc. (Basel, Switzerland), Asta Medica, Inc.
(Frankfurt, Germany), and Eli Lilly Inc. (Indianapolis, IN).
REFERENCES
In the approach we developed and describe here, we use a
standard paper form as the source document but all subsequent
steps (reading the data by scanning, transference of the facsimile
and of the data, interactive analysis, and entry into database) are
automatic, but using interactive surveillance and quality control.
We could have used a paperless computer approach for entry and
further handling of data but decided against this approach
because a paper source document was desired and because
personal computers were not readily available or acceptable to
all participating investigators. Also, the accuracy of video
without a paper record by physicians has not been established.
Does the use of scannable case report forms (and the analysis
by the computer programs described here) improve the accuracy
of the information in database? Was time saved? Is the approach
cost-effective? A difference in accuracy was observed between
scanned and manual entry of information into database. With
scanning of 500 serial CNA records, 1.2% of records were found
to have faulty data entry, which in quality control was readily
recognized as due to a line shift of the data, and then was readily
corrected. For the 25 serial CNA evaluated in timed studies, no
line shifts occurred. By comparison, when all steps of CNA data
evaluation were done by hand and the data entered by keyboard,
8 of 25 records were found to have data entry discrepancies.
When the differences between them were reconciled, the data
was completely concordant with the scanned data entry results.
1. Hagen PT, Turner D, Daniels L, Joyce D. Very large-scale
distributed scanning solution for automated entry of patient
information. Toward an electronic patient record ’98:
proceedings 1998; 1:228-232.
2. Adams WT, Snow GM, Helmick PM. Automated charge processing
streamlines data entry. Health Financ Manage 2000; 54:50-53.
3. Shiffman RN, Brandt CA, Freeman BG. Transition to a computer-
based record using scannable, structured encounter forms. Arch
Pediatr Adolesc Med 1997; 151:1247-1253.
4. Allen DM, Hey LA, Heinz TR, Golal R, Levin LS. Development
and implementation of an extremity free-tissue-transfer database.
J Reconstr Microsurg 1997; 13:475-487.
5. Guerette P, Robinson B, Moran WP, et al. Teleform scannable data
entry: an efficient method to update a community-based medical
record? Community care coordination network Database Group.
Proceedings - The Annual Symposium on Computer Applications
in Medical Care 1995;86-90.
6. Moss RL, Kolaric D, Watts A. Therapeutic agents utilized in
urban/rural prehospital care. Prehospital Disaster Med 1993;
8:161-164.
7. Xakellis GC, Gjerde CL, Emerson M. Using an optical scanner and
data base program to manage a family medicine teaching
program. Fam Med 1994; 26:421-424.
8. McKegney FP, Schwartz CE, O’Dowd MA, Salamon I, Kennedy R.
Development of an optically scanned consultation-liaison data
base. Gen Hosp Psychiatry 1990; 12:71-76.
9. Dayhoff R, Kirin G, Pollock S, Todd S. Data capture workstations,
scanned forms, and pen-based systems for clinician use. Medinfo
1995; 8(Pt 2):1679
Volume 29, No. 3 – August 2002
265