Documente Academic
Documente Profesional
Documente Cultură
Background: Collecting data on medical errors is es- ported an actual major error (resulting in disability or
sential for improving patient safety, but factors affecting death). Moreover, 16.9% acknowledged not reporting an
error reporting by physicians are poorly understood. actual minor error, and 3.8% acknowledged not report-
ing an actual major error. Only 54.8% of respondents
Methods: Survey of faculty and resident physicians in knew how to report errors, and only 39.5% knew what
the midwest, mid-Atlantic, and northeast regions of the kind of errors to report. Multivariate analyses of an-
United States to investigate reporting of actual errors, like- swers to hypothetical vignettes showed that willingness
lihood of reporting hypothetical errors, attitudes to- to report was positively associated with believing that re-
ward reporting errors, and demographic factors. porting improves the quality of care, knowing how to re-
port errors, believing in forgiveness, and being a faculty
Results: Responses were received from 338 partici-
physician (vs a resident).
pants (response rate, 74.0%). Most respondents agreed
that reporting errors improves the quality of care for fu-
Conclusion: Most faculty and resident physicians are in-
ture patients (84.3%) and would likely report a hypo-
thetical error resulting in minor (73%) or major (92%) clined to report harm-causing hypothetical errors, but
harm to a patient. However, only 17.8% of respondents only a minority have actually reported an error.
had reported an actual minor error (resulting in pro-
longed treatment or discomfort), and only 3.8% had re- Arch Intern Med. 2008;168(1):40-46
T
O IMPROVE PATIENT SAFETY, in 2000.16-18 Obstacles to reporting are di-
it is necessary to under- verse and daunting,19,20 and underreporting
stand the frequency, seri- of errors is believed to be pervasive.9,21 Be-
Author Affiliations: Division of ousness, and causes of cause error surveillance systems generally
General Internal Medicine, medical errors. 1 - 8 Such rely on self-reporting,22 it is important to un-
Department of Internal Medicine knowledge is acquired by the analysis of derstand the factors that continue to impede
(Drs Kaldjian, Forman-Hoffman, data collected through error-reporting sys- reporting, as well as those that may facili-
and Rosenthal and Ms Jones) tate it. To improve our knowledge about er-
and Program in Biomedical
tems. These systems may be internal or ex-
ternal to health care institutions and may ror reporting, we conducted a survey of phy-
Ethics and Medical Humanities
be voluntary or mandatory.9 Whatever sicians in teaching hospitals. We chose this
(Dr Kaldjian), University of
Iowa Roy J. and Lucille A. their particular design, at the ground level settingbecauseoftheinfluenceteachinghos-
Carver College of Medicine, and of clinical practice, such reporting relies pitals may have on trainees’ attitudes toward
Center for Research in the on a professional culture in which physi- error reporting and patient safety.
Implementation of Innovative cians and other health care providers view
Strategies in Practice, Iowa City error disclosure as an integral part of learn- METHODS
Veterans Affairs Medical Center
(Drs Kaldjian, Forman-Hoffman,
ing and quality improvement.4,10-12 It is also
and Rosenthal and Ms Jones), important to consider how reporting er- PARTICIPANTS
Iowa City; Departments of rors to institutions to enhance patient
Internal Medicine, Hospital of safety may differ from disclosing errors to Faculty and resident physicians were drawn
St Raphael and Yale University patients as part of direct patient care.13 from 3 medical centers located in the mid-
School of Medicine, New Haven, Despite recent professional and legisla- west, mid-Atlantic, and northeast regions of the
Connecticut (Dr Wu); and tive efforts to encourage error reporting,9,14 United States. Surveys were completed from
Departments of Humanities and including the passage of the Patient Safety June 1 through September 30, 2004, with the
Pediatrics, Penn State College of exception of 32 resident physicians and 51 fac-
Medicine and Hershey Medical
and Quality Improvement Act of 2005,15 it ulty physicians at 1 study site who completed
Center, Hershey, Pennsylvania is not clear whether error reporting has be- surveys from January 1 through March 31,
(Dr Levi). Ms Jones is now with come more common since the Institute of 2005. Two of the 3 states in which the study
the Virginia Department of Medicine published its seminal report, To was conducted enacted mandatory error re-
Social Services, Richmond. Err Is Human: Building a Safer Health System, porting legislation in 2002; the third state had
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
40
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
41
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
42
% Who Agreed
institution and only 39.5% knew what kind of errors to be significant (P ⬍ .10) in 1 or more of the 3 models,
should be reported. Many respondents believed that it as shown in Table 5. Respondents were more likely to
is hard to be certain about the true causes of adverse report errors if they knew how to report them or be-
events, but few believed that reporting errors was not lieved that error reporting improves the quality of care,
worth their time. The anticipation of feedback would make and they were less likely to report errors if they believed
47.9% of the respondents more likely to report errors, that reporting them was not worth their time. Faculty
and 57.7% acknowledged concerns about professional dis- physicians were more likely than residents to report er-
cipline when thinking about disclosing errors in gen- rors, and respondents for whom forgiveness was impor-
eral. Resident physicians were less knowledgeable than tant were also more likely to report errors. Neither sex
faculty about how to report errors and what errors to re- nor exposure to malpractice was associated with differ-
port, and they were generally more concerned about pro- ences in the likelihood of reporting any of the hypotheti-
fessional discipline when disclosing errors. cal errors, and only 1 variable dropped out of 1 model
after adjusting for site differences (an association be-
VARIABLES ASSOCIATED tween specialty and reporting an error resulting in no
WITH REPORTING HYPOTHETICAL ERRORS harm).
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
43
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
44
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
45
(REPRINTED) ARCH INTERN MED/ VOL 168 (NO. 1), JAN 14, 2008 WWW.ARCHINTERNMED.COM
46