Christie JD, Rosen IM, Bellini LM, Inglesby TV, Lindsay J, Alper A, Asch DA. Prescription Drug Use and Self-prescription Among Resident Physicians. JAMA. 1998;280(14):1253-1255. doi:10.1001/jama.280.14.1253
From the Department of Medicine (Drs Christie, Rosen, Bellini, and Asch), the Leonard Davis Institute of Health Economics (Dr Asch), and the Center for Bioethics (Dr Asch), University of Pennsylvania, and the Veterans Affairs Medical Center (Drs Christie, Rosen, Bellini, and Asch), Philadelphia, Pa; Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Md (Dr Inglesby); Department of Medicine, Stanford University School of Medicine, Stanford, Calif (Dr Lindsay); and the Department of Medicine, Tulane University School of Medicine, New Orleans, La (Dr Alper).
Context.— Self-prescription is common among practicing physicians, but little
is known about the practice among resident physicians.
Objective.— To determine prescription drug use and self-prescription among US resident
Design and Setting.— Anonymous mail survey of all resident physicians in 4 US categorical
internal medicine training programs in February 1997.
Main Outcome Measures.— Self-reported use of health care services and prescription medications
and how they were obtained.
Results.— A total of 316 (83%) of 381 residents responded; 244 residents (78%)
reported using at least 1 prescription medicine and 162 residents (52%) reported
self-prescribing medications. Twenty-five percent of all medications and 42%
of self-prescribed medications were obtained from a sample cabinet; 7% of
all medications and 11% of self-prescribed medications were obtained directly
from a pharmaceutical company representative.
Conclusions.— Self-prescription is common among resident physicians. Although self-prescription
is difficult to evaluate, the source of these medications and the lack of
oversight of medication use raise questions about the practice.
UNLIKE THE REST of the population, when physicians become ill, they
can prescribe medicines for themselves. Medical knowledge and access to prescription
medications increase the potential for self-treatment. Although many warn
of the loss of objectivity that can accompany self-prescription, previous
studies suggest that self-prescription is common among practicing physicians.1- 4
Almost nothing is known about self-prescription and self-care among
resident physicians. They have long and unpredictable work hours that make
scheduling conventional care difficult. They work within a hospital and so
special issues of privacy arise. Residents have easy access to prescription
medications through on-site pharmacies and the presence of sample medications.
These factors might make self-prescription and self-care particularly common
among resident physicians. The purpose of this study was to evaluate self-prescription
and self-care practices among resident physicians in 4 US internal medicine
residency training programs.
The subjects were all resident physicians at 4 categorical internal
medicine training programs in the United States: Johns Hopkins University,
Baltimore, Md; Stanford University, Stanford, Calif; Tulane University, New
Orleans, La; and the University of Pennsylvania, Philadelphia. In February
1997 we mailed each intern or resident an 8-page questionnaire asking about
their use of health care services, including the use of prescription pharmaceuticals
and how they obtained them. All respondents were anonymous. We provided residents
with postcards to mail separately from the completed instrument to indicate
they had responded.5 We mailed nonresponders
up to 3 copies of the instrument at 3-week intervals.
To determine prescription drug use, we asked residents, "Since beginning
your PGY [postgraduate year]–1 year, have you taken any prescription
medications, whether prescribed to you or not (eg, antibiotics, contraceptives,
antihypertensives, prescription-dose antiulcer drugs, prescription-strength
topical steroids and/or prescription inhalers)?" For each medication, we asked
residents to indicate the medication's duration, frequency, and indication.
In asking about who prescribed each medication, we were careful to distinguish
between recommending a medication and the various activities related to obtaining
it. For example, a physician caring for a medical resident might order or
recommend a medication, yet that resident might write the prescription, call
it in to a pharmacy, or obtain samples directly from a sample cabinet. Because
we were interested in learning about prescription medications that residents
decided to take on their own, we first asked, "Who provides you with care
for the condition indicated?" and we provided possible responses of (1) no
one (self); (2) fellow resident; (3) personal physician who sees me in an
office or a clinic; (4) other physician who gives me advice outside an office
visit or clinic; or (5) other. We then asked, "How did you obtain the medication?"
and provided possible responses of (1) clinician other than self wrote or
called in prescription to pharmacy or provided a sample; (2) self wrote or
called in prescription to pharmacy; (3) sample obtained personally from sample
closet; (4) sample obtained directly from pharmaceutical company representative;
or (5) other. Finally, to determine the prescriber of each medicine, we asked,
"Regardless of how this medication was obtained, who prescribed the medication?"
with possible responses of (1) individual listed in the first question; (2)
self; or (3) other.
We also asked residents to respond to several statements about self-prescription
and self-care using 5-point scales from strongly disagree to strongly agree.
We also invited any other comments. All responses were read by 3 physicians
to categorize medications and indications and to verify that the medication
was not otherwise available without a prescription at that dose.
The protocol was approved by the human subjects committees at all 4
A total of 389 instruments were mailed; 8 were undeliverable because
of bad addresses and 316 completed instruments were returned for a response
rate of 83%. The response rates across the 4 institutions ranged from 81%
to 84%. Three subjects were excluded from further analysis, 2 because they
indicated they were PGY-4 without providing an explanation and 1 who reported
being a fellow, leaving a total of 313 subjects.
A total of 244 residents (78%) reported using at least 1 prescription
drug since beginning their PGY-1 year, and in aggregate, they reported taking
605 prescription medications. Among PGY-1, PGY-2, and PGY-3 residents, 69%,
84%, and 84% reported using prescription medications (P = .01 by χ2) and the mean number of such medications
used by these residents was 1.6, 2.1, and 2.3, respectively (P = .01 by analysis of variance). Table 1 shows the classes of drugs used by residents, their frequency,
and examples of their indications. Table
2 shows the prescribers of these medications. About half of these
medications were self-prescribed and another 6% were prescribed by a fellow
A total of 162 residents (52%) reported self-prescribing medications
since beginning PGY-1. Among PGY-1, PGY-2, and PGY-3 residents, 39%, 55%,
and 67% reported self-prescribing medications (P<.001
by χ2), and the mean number of self-prescribed medications
for these residents was 0.7, 1.1, and 1.2, respectively (P = .003 by analysis of variance). A similar proportion of residents
self-prescribed across the 4 programs (P>.3 by χ2). Table 2 shows the most
common classes of self-prescribed medications and examples of their indications.
Table 3 shows the sources
of the prescription drugs. The most common source for self-prescribed medications
was a sample cabinet. This mechanism was popular even for medications prescribed
by others. In addition, a substantial minority of all prescription medications
were provided to residents directly from pharmaceutical company representatives.
A total of 152 residents (49%) indicated that either they had no primary
care physician (n = 116) or they were their own primary care physician (n
= 36). These residents were no more likely to self-prescribe than the 157
residents who indicated they had a primary care physician. Residents who reported
self-prescribing were more likely to agree with the statement, "It is reasonable
for physicians to prescribe medications for themselves if they are knowledgeable
about the clinical condition they are treating" (3.6 vs 3.3 on a 5-point scale; P = .004) and to disagree with the statement, "Physicians
should not treat their own medical problems or conditions" (2.9 vs 3.3 on
a 5-point scale; P<.001). Residents who self-prescribed
were also more likely to agree that "Many of the attending physicians I know
treat their own medical problems or conditions" (3.5 vs 3.3 on a 5-point scale; P = .02). Many residents provided handwritten comments
such as the following: "In general, I disagree with self-diagnosis and prescription,
but it is tempting for minor complaints, especially when busy." "I believe
physicians can perform maintenance therapy on themselves—eg, medical
residents treating their own allergies, hypertension, hypercholesterolemia—but
must be aware of problems that they cannot self-treat or do not have the training
to treat." "Many problems can be addressed by oneself. Of course I did not
used to think that. I draw the line at antibiotics and narcotics—but
there is nothing wrong with a little H2 [histamine2]
blocker or NSAIDs [nonsteroidal anti-inflammatory drugs]."
Self-prescription among internal medicine residents is common. We are
unaware of prior studies of resident physician self-prescribing practices.
In a study of 247 established physicians in the United Kingdom, 84% of all
medications taken during a 5-year period were self-prescribed.6
In another study of 306 physicians in Rhode Island, 61% reported self-prescribing
during a 3-year period.4 These and other findings7,8 support the notion that self-prescribing
is common among physicians. Our study demonstrates that this practice is also
common among residents. The similar frequency of self-prescription among internal
medicine residents across the 4 programs strongly suggests that self-prescription
is a general phenomenon among internal medicine residents rather than a reflection
of the culture of one particular institution.
The most common source of self-prescribed medicines was the sample closet
(26% of all medications and 42% of all self-prescribed medications). This
finding is consistent with a recent study in which 51 of 53 physicians, residents,
nurses, and staff in a family practice group reported taking pharmaceutical
samples for personal and family use.7 Additionally,
we found that 7% of all medications and 10% of those self-prescribed were
obtained directly from pharmaceutical company representatives. We are unaware
of other systematic studies of this practice.
The most commonly self-prescribed medications used by residents in our
study were antibiotics, allergy medicines, and contraceptives. Previous studies
suggest antibiotics are the most commonly self-prescribed medicines among
established US and United Kingdom physicians.2,4,6
Of our 313 resident physician respondents, only 7 (2%) reported the use and
self-prescription of psychotropic medicines. In contrast, McCauliffe and colleagues8 reported that 25% of 342 practicing physicians from
New England surveyed had treated themselves with a psychotropic drug within
the previous 12 months. This difference might be explained by differences
in disease incidence between our resident physicians and more senior physicians,
differences in comfort with prescribing these medicines, differences in access
to these medications, or differences in reporting.
This study is subject to several limitations. First, we surveyed residents
at only 4 US internal medicine training programs, all based at university
teaching hospitals. Second, our results are based on self-reports in the setting
of an anonymous mail survey. Our instrument was pilot tested to judge completeness,
readability, and accuracy, but, in the end, the validity of these results
depends on respondents' comprehension, recall, and honesty. Recall bias would
likely result in underreporting the number of self-prescribed medications.
Similarly, despite anonymity, residents might be reluctant to report use or
prescription of certain classes of medications.
There are many ways to view these results. One view recognizes self-prescription
among residents as a reaction to the intense time pressures residents face
and their difficulties receiving medical care. Many of the residents we surveyed
remarked that their busy and unpredictable schedules made it difficult to
arrange care for acute conditions and arrange appointments for disease prevention
or management of more chronic conditions. However, internal medicine residents
are not alone in being busy. Residents differ because in addition to being
busy, they also have knowledge of health conditions and easy access to medications.
In addition, their close working relationships with other physicians may create
barriers to privacy and make it difficult to identify physicians with whom
they can comfortably assume patient roles.
A second way to view these results is to ask whether they reveal problems
in professional behavior and, if so, what practical solutions such problems
might suggest. Our results probably reflect a wide range of circumstances;
it may be difficult to identify clear boundaries that separate inappropriate
self-prescription from more acceptable examples. Distinctions might be made
according to the type of medication; for example, perhaps it is acceptable
for a physician to self-prescribe prescription-dose antiulcer drugs but not
antidepressants, particularly since antiulcer drugs, in lower doses, are available
without prescription. Distinctions might be made according to the indication
for the medication; for example, perhaps it is inappropriate for a physician
to self-prescribe β-blockers for treatment of hypertension but acceptable
if the same drug is used occasionally to reduce the symptoms of stage fright
when giving public presentations. Distinctions might be made according to
the monitoring a drug requires; for example, self-prescription of lipid-lowering
drugs requires laboratory monitoring of serum lipids and other biochemical
parameters and so invokes another element of self-treatment that the self-prescription
of a nonsedating antihistamine does not. Distinctions might be made according
to past medical care as well; for example, perhaps it is inappropriate for
a physician to self-initiate inhaled bronchodilators to treat asthma but more
acceptable if the physician continues the same therapy prescribed in the past
by another physician no longer seen. Although it is easy to identify extremes
of behavior as either inappropriate or acceptable, many examples of self-prescription
probably fall in between and, moreover, might be judged by any of these different
Although the personal use of pharmaceutical samples raises ethical concerns,
the difficulty presented by self-prescription is more a question of clinical
judgment. Based on our survey, we cannot evaluate the clinical outcomes attained
by residents who self-prescribe. Although we believe that many examples of
self-prescription are tolerable, it remains reasonable to question behavior
that lacks the objectivity and professional distance crucial to the assurance
of quality in any medical encounter.
If self-prescription is a problem, this study suggests that the problem
starts early in physicians' careers and may remain hidden unless specifically
addressed. For these reasons, we believe that medical schools, residency programs,
and professional societies should discuss the issue of self-prescription openly.
These discussions may help resident and practicing physicians evaluate their
own self-treatment practices more closely and may lead to more practical standards
for appropriate behavior. These discussions are important given that the common
practice of obtaining these medications from a sample closet or directly from
a pharmaceutical company representative challenges professional standards
of behavior. Although not all cases of self-prescription demand the same moral
or policy response, the profession of medicine must remain vigilant about
this practice to ensure that convenience and easy access do not undermine
professionalism in the prescription and use of medications.