The truth about doctors' handwriting: a prospective study
Donald M Berwick, student, David E Winickoff, president and chief executive
officer
Institute for Healthcare Improvement
Boston, MA 02215,
USA
Abstract
Objective--To determine whether doctors have
worse handwriting than other health professionals.
Design--Comparison of handwriting samples collected
prospectively in a standardised 10 seconds' task.
Setting--Courses on quality improvement.
Subjects--209 health care professionals attending the
courses, including 82 doctors.
Main outcome measures--Legibility rated on a
four-point scale by four raters.
Results--The handwriting of doctors was no less legible than that of non-doctors. Significantly lower legibility than average was associated with being an executive and being male. Overall legibility scores were normally distributed, with median legibility equivalent to a rating between "fair" and "good."
Conclusion--This study fails to support the conventional wisdom that doctors' handwriting is worse than others'. Illegible writing is, however, an important cause of waste and hazard in medical care, but efforts to improve the safety and efficiency of written communication must approach the problem systemically--and assume that the problems are in inherent in average human writing--rather than treating doctors as if they were a special subpopulation.
Introduction
The assertion that doctors have bad handwriting holds an
honoured place in traditional lore. According to conventional wisdom,
doctors write in a code--a self righteous chicken scratch that is
decipherable only by experienced pharmacists and, with luck, by each
other. The question of doctors' handwriting, of course, has a serious
side with far reaching implications concerning the quality and safety
of health care. Some studies have found doctors' medical records and
prescriptions illegible, wasteful, and
dangerous, (1-7) but we found no evidence on whether
poor handwriting is indeed more of a problem among doctors than among
other adults. We gathered data on handwriting under controlled
circumstances to determine if, among professionals in health care,
being a doctor is associated with poorer handwriting.
Methods and results
The subjects were a mix of healthcare executives, clinicians, and
managers who were attending three different courses on quality
improvement in health care. During a brief break period we asked
participants while seated at a table to write on a blank piece of paper
the following sentence: "Quality Improvement is the best thing since
sliced bread." The participants were told to begin writing at the
word "go" and were interrupted and told to stop after 10 seconds.
Four volunteers, all non-clinicians, independently rated the legibility
of each writing sample on a four-point scale: poor, fair, good,
excellent. Raters were asked to use legibility as the rating
criterion.
A total of 209 writing specimens was obtained. In these, 50 subjects
had managed to write the complete challenge sentence, while 159 had
not. Of the samples, 82 came from doctors, 32 from chief executives and
chief operating officers (of whom 12 were doctors as well), 131 from
men, and 139 from people aged 40 or older.
The rating scheme achieved a high level of inter-rater reliability, with pairwise correlation coefficients ranging from 0.60 to 0.76. Fig 1 shows four samples of writing on which all four raters agreed completely at each of the four levels of legibility. To calculate a summary rating, we simply added the four individual ratings for each sample and subtracted three points from the total, thus yielding a final legibility score between 1 (all four ratings "poor") and 13 (all four ratings "excellent"). The resulting range of legibility scores was approximately normal, with a mean rating of 7.15, a median of 7, and a standard deviation of 3.14.
The differences between means of the various groups were compared by the t test. By this test, the handwriting of doctors was no less legible than that of non-doctors. On average, the doctors scored 0.79 points less than non-doctors, an insignificant difference (P=0.074). In comparison, chief executives and chief operating officers averaged 2.87 points lower than non-executives, men averaged 2.25 points lower than women, and older subjects averaged 1.48 points lower than younger ones, all of which differences were statistically significant (table 1). Completing the challenge sentence (presumably an indication of faster writing) was also associated with a lower score.
| Comparison of mean legibility scores by subject characteristic |
|||||
| Characteristic | Score of 1st group | Score of 2nd group | 95% Confidence interval for difference between means | P | |
| Doctors (n = 82) |
6.67 | 7.46 | -1.80 | -1.65 to 0.058 | 0.074 |
| Executives (n = 32) |
4.72 | 7.59 | -6.18 | -1.94 to -3.80 | \h0.0001 |
| Men (n = 131) |
6.30 | 8.58 | -5.56 | -1.47 to -3.09 | \h0.0001 |
| Age \g39 (n = 139) |
6.65 | 8.13 | -3.33 | -0.60 to -2.35 | 0.0011 |
| Completed sample (n = 51) |
5.65 | 7.63 | -3.78 | -0.94 to -3.03 | 0.0003 |
Discussion
In a simple, time limited handwriting task that may mimic the
type of time pressure under which busy people in medical care work,
doctors as a group did not write less legibly than other subjects. Why
do so many believe that doctors have especially bad handwriting when in
fact their writing may be quite average under the circumstances? One
possible explanation is that average is, in absolute terms, "bad,"
and that the poor handwriting of doctors is riskier than the poor
handwriting of non-doctors. The mean score in our study was 7.1 out of
a possible 13 points, indicating that handwriting among the average
subject, whether a doctor or not, lies somewhere between the two middle
samples in fig 1. In medical circumstances, where the stakes are high,
writing of this mediocre quality may be unacceptable.
If average handwriting is not acceptably safe or clear in medical care then we must seek ways either to "error proof" written communication or to reduce reliance on it. In the short run increasing individual's awareness and motivation may produce some minor gains. In the longer run, however, it may be more helpful to regard legibility as a systemic problem, not a personal one. Better physical designs, for example, might make it easier to write legibly. However, when the stakes are highest, safety may require not improving writing, but replacing it.
We have shown, in a study with an artificial task and high inter-rater reliability, that doctors have handwriting no worse than that of a comparison group of other healthcare personnel, and much better than that of healthcare executives. These findings in no way contradict lore and literature about the costs and hazards of poor writing in prescriptions and medical records, but they do refocus the problem. Illegibility in medical care may have less to do with "bad" (that is, exceptionally bad) handwriting among doctors than with handwriting in general as a form of communication. More help may be found in computerisation (8) and other systemic innovations than in pointing the finger at a profession whose members, on the whole, write with an average hand.
References 2 A study of physicians' handwriting as a time-waster.
JAMA 1979; 3 Halper NA, Colucci RD, Alicea M, Greenstein R. Intensive care
order writing practices in the USA: results in a nationwide study.
Crit Care Med 1991; 4 Brahams D. Uninsured pharmacists and illegible prescriptions.
Lancet 1989;i:510.
5 Sanders S, Shaw AB. Drug name confusion. Lancet
1991;1: 56-7.
6 Mullan K. Importance of legible prescriptions. J Royal
Coll Gen Pract 1989; 7 Fenelly WA. Can you read your own handwriting? N York J
Dentistry 1975; 8 Institute of Medicine. The computer-based medical
record: an essential technology for health care. Washington, DC:
National Academy Press, 1991.
1. White KB, Beary JF. Illegible handwritten medical records.
New Engl J Med 1986;
Current contents | Classified ads | Archive and search | Local editions | Advice to authors
Reprints | Subscriptions | Feedback | Home