Doctors' handwriting

Doctors' handwriting, also called physicians' handwriting, refers to the handwriting of medical doctors on prescriptions, orders, and medical records, and to the widespread belief that it is unusually hard to read. Illegible handwriting in health care is a documented source of medication errors and lost information, and it has been the basis of at least one US negligence verdict.^[1:1] Whether doctors actually write worse than other people is less settled: controlled studies from the 1990s and 2000s reached conflicting conclusions.^[2:1][3:1]

The problem has been addressed mainly by replacing handwriting rather than improving it. Health authorities and researchers have promoted electronic prescribing and computerized physician order entry (CPOE) as ways to remove handwriting from the process, though these systems introduce errors of their own.^[4:1][5]

Research on legibility

Several small studies have tested the stereotype by comparing doctors' handwriting with that of other workers. Their results disagree, and the differences in method may explain why.

In the first of these, Donald Berwick and David Winickoff collected samples from people attending courses on quality improvement. They imposed a 10-second limit to mimic the time pressure of clinical work. They found that doctors' writing was no less legible than others', and that lower legibility was instead associated with being an executive and with being male.^[2:1] The authors still called illegible writing a cause of waste and hazard in medical care and recommended system-level fixes rather than singling out doctors.^[2]

Ronan Lyons and colleagues took a different approach in 1998, asking staff at a Welsh health district to write as neatly as they could and scoring the results with character-recognition software instead of human raters. Doctors had a median of seven letter errors, compared with three for nurses and other health professionals and four for administrative staff. There was no significant difference for numerals. The authors concluded that doctors write worse than other professions even when trying to be neat.^[3:1]

A 2006 study by K. A. Schneider and colleagues compared doctors with accountants, attorneys, automobile technicians, builders, engineers, and scientists, and found no significant difference between occupations. The authors agreed with Berwick and Winickoff, while arguing that doctors should still be held to a higher standard because health care depends on accurate written communication.^[6:1]

The disagreement has not been resolved. The studies were small, drew participants from different settings, and measured legibility differently: Lyons et al. asked for neat writing of isolated characters scored by machine, while the other two studies used timed sentences judged by people. None measured handwriting as it appears in real prescriptions or clinical notes. A larger study comparing actual clinical documents written by doctors and other staff, using the same scoring method, could settle the question.

Illegibility in clinical records

Separate from the comparison with other professions, studies of real medical records have documented how often clinical handwriting cannot be read. A 2002 audit of 117 case notes at a Spanish general hospital found that 18 (15%) were so illegible that their meaning was unclear; records from surgical departments were worse than those from medical departments.^[7:1] The authors argued that illegible notes make information inaccessible to auditors, researchers, and other clinicians, and that if clinicians could not be persuaded to write legibly, the solution was an accelerated switch to computer-based records.^[7]

Consequences

Medication errors

Illegible prescriptions can lead pharmacists to dispense the wrong drug or the wrong dose. In a Texas case reported in 2000, cardiologist Ramachandra Kolluru of Odessa prescribed 20 mg of Isordil (isosorbide dinitrate, an angina drug) every six hours. The pharmacist read the prescription as Plendil (felodipine, a blood pressure drug) and dispensed it at the same dose, well above its maximum daily dose. The patient, 42-year-old Ramon Vasquez, had a heart attack a day after starting the drug and died several days later. A jury found the doctor and the pharmacy equally liable, each ordered to pay $225,000. The case was described as probably the first in the United States in which a physician was found negligent solely because of poor handwriting.^[1:1]

The "7,000 deaths" figure

A frequently repeated claim holds that doctors' handwriting kills more than 7,000 people a year in the United States. Time magazine made this claim in January 2007, attributing it to a 2006 Institute of Medicine report.^[8] Writing in The Straight Dope, Cecil Adams traced the figure to a 1998 letter in The Lancet on US deaths from medication errors, and noted that it covers all medication-related errors of any kind, not only those caused by handwriting.^[9:1][10] The number of deaths attributable specifically to illegible handwriting has not been reliably measured.

Legal and technological responses

Some jurisdictions have written legibility into law. Since 2003, Florida has required that written prescriptions be "legibly printed or typed so as to be capable of being understood by the pharmacist", and that the quantity be written in both words and numbers and the month written out in letters.^[11]

The main response has been to move prescribing onto computers. In a 1998 study at a large teaching hospital, David Bates and colleagues found that computerized physician order entry cut serious medication errors that were not caught before reaching the patient by 55%.^[12:1] The Institute of Medicine's report Preventing Medication Errors, released in 2006, recommended that all prescriptions be written electronically by 2010, noting that electronic prescribing is safer partly because it eliminates handwriting.^[4:1]

Electronic systems remove the handwriting problem but do not remove error. A 2005 study of one widely used CPOE system identified 22 kinds of error risk caused by the software itself, such as confusing displays and inflexible order formats.^[5] A 2011 study of 3,850 computer-generated outpatient prescriptions found that 11.7% contained errors, most often missing information, with error rates ranging from 5.1% to 37.5% depending on the system.^[13:1]

Analysis

This section is an editorial judgment by Agpedia contributors, based on the sources cited above.

The research suggests that the stereotype of the uniquely illegible doctor is weakly supported, but that the harm from illegible clinical writing is real. That harm depends less on whether doctors write worse than accountants than on the setting: a doctor's handwriting is read by pharmacists, nurses, and other clinicians who act on it, often under time pressure, and a misread word can mean a wrong drug. Treating illegibility as an individual failing—an approach Berwick and Winickoff argued against—does less to protect patients than designing systems that do not depend on reading anyone's handwriting.

Illegible records also limit patients' ability to understand and take part in their own care. A patient who cannot read their own prescription cannot easily check it, question it, or notice a mistake. Typed prescriptions and records make that checking possible, though only if patients are given access to them.