Chart Reviews In Emergency Medicine Research: Where Are The Methods?☆,☆☆,★
Section snippets
INTRODUCTION
Medical record reviews are used frequently to generate original research in emergency medicine. However, although medical records contain important clinical information, they are not produced for research purposes.1 There are no universally-accepted criteria for a "well-conducted" chart review, but eight methodologic strategies (Figure 1) may enhance the validity, reproducibility, and overall quality of data collected from clinical records.1, 2
MATERIALS AND METHODS
Original research articles published between January 1, 1989, and December 31, 1993, were retrieved from the American Journal of Emergency Medicine, Annals of Emergency Medicine, and Journal of Emergency Medicine. Case reports, letters, editorials, subject reviews, metaanalyses, special theme papers, and symposium proceedings were excluded. Those of the original research articles that relied solely or mostly on data from medical records to answer the questions posed by the study were judged to
RESULTS
During the 5-year period ending December 31, 1993, approximately 3,000 articles of all types were published in the three peer-reviewed emergency medicine journals. Of these, 986 were original research articles, among which chart reviews supplied most or all of the important data in 244 (25%; 95% CI, 22% to 28%). The number of chart review articles each year varied from 38 (16%) in 1991 to 61 (25%) in 1990. The majority (65.6%) of chart review articles appeared in Annals of Emergency Medicine.
DISCUSSION
Between 1989 and 1993, one fourth of all scientific investigations published in emergency medicine journals were medical chart reviews. However, as the present investigation indicates, most chart reviews lack sound methods. Frequently, information is collected by poorly trained, unmonitored personnel, who may be aware of the study hypotheses and group or treatment assignments. Most chart reviews do not report the reproducibility of their data. Chart reviews in emergency medicine research may
References (17)
- et al.
Observer variation in the classification of the information from medical records
J Chron Dis
(1979) - et al.
Assessing the reliability of epidemiologic data obtained from medical records
J Chron Dis
(1984) - et al.
High agreement but low Kappa: I. The problem of two paradoxes
J Clin Epidemiol
(1990) - et al.
A bibliography of publications on observer variability (final installment)
J Clin Epidemiol
(1992) - et al.
An analysis of the quality of medical record reviews in general medicine journals
Clin Res
(1992) - et al.
The epidemiology of cancer therapy: III. The management of imperfect data
Arch Intern Med
(1969) The reliability of clinical methods, data and judgements I and II
N Engl J Med
(1975)
Cited by (0)
- ☆
From the Colorado Emergency Medicine Research Center, Division of Emergency Medicine*, and the Division of Internal Medicine‡, University of Colorado Health Sciences Center, Denver, Colorado.
- ☆☆
Address for reprints: Steven R Lowenstein, MD, MPH, Division of Emergency Medicine, University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Box B211, Denver, Colorado 80262, 303-270-8711, Fax 303-270-5614
- ★
Reprint no. 47/1/69724