Efficiency of a Semiautomated Coding and Review Process for Notification of Critical Findings in Diagnostic Imaging
Abstract
OBJECTIVE. When a significant unexpected finding such as malignancy is noted on a study, the standard of care generally holds that the radiologist communicate the findings to the referring physician and document the communication in the radiology report. Despite this standard, for a variety of reasons it remains possible that the direct care provider might receive such notification but not initiate an appropriate workup. On the basis of prior root cause analysis, we developed and instituted a semiautomated process for notification of critical diagnostic imaging findings. We now report our 12-month experience with the process.
MATERIALS AND METHODS. A diagnostic code was attached to every radiology report. When a significant unexpected finding occurred, our radiologists, in addition to contacting the appropriate clinician, gave the report the designation code 8. On a weekly basis, a list of code 8 cases was passed to the cancer registrar at our institution, who tracked the cases to ensure that they were appropriately followed up.
RESULTS. In the 12-month period after initiation of this system, we performed 37,736 radiologic examinations at our institute. Of these, 395 cases were given code 8. All code 8 cases were followed up by the tumor registrar. In 35 cases, no workup was documented after 2 weeks. Of these, eight cases would have been completely lost to follow-up if this safety net had not been in place.
CONCLUSION. Failures of communication, documentation errors, and various system failures may lead to an untoward outcome for the patient. We devised a simple system to ensure that significant unexpected findings on imaging received appropriate attention. An additional level of redundancy has increased the probability of optimal patient outcome.
Introduction
Asteady stream of high-visibility medical accidents has kept patient safety in the forefront among health care issues. Although controversy exists as to the exact volume of medical errors [1], To Err is Human: Building a Safer Health System, a report by the U.S. Institute of Medicine [2], estimated that at least 44,000 and perhaps as many as 98,000 Americans die each year as a result of medical errors. The core mission of hospitals is to optimize patient outcomes by providing appropriate, timely, and safe treatment. The Joint Commission on the Accreditation of Healthcare Organizations has published national safety goals that are updated annually. The current emphasis on patient safety involves the continuous monitoring of care activities, seeking opportunities to improve high-risk processes and to respond to catastrophic or near-miss events that show the system weaknesses.
We report on a semiautomated coding and review process for notification of critical diagnostic imaging findings that was developed in response to an institutional sentinel event in which evidence of unsuspected malignancy was noted on a preoperative chest radiograph but was not appropriately acted on because of a series of communication failures. The subsequent root cause analysis—a method used to identify the reasons the system failed and the factors that contributed to create the conditions in which the error occurred [3-5]—led to the institution of mandatory radiology report coding and periodic review of reports coded as “unexpected finding, probable malignancy.” We present our 12-month experience with this communication process.
Materials and Methods
The Process
After a root cause analysis was performed [3], the department of radiology, in conjunction with the oncology service, designed a semiautomated process for coding and review of unexpected findings indicating malignancy to facilitate appropriate communication and follow-up of patients. The protocol was an extension of the routine film interpretation and report finalization. The existing voluntary coding system used codes 1-9, including codes for normal (code 1); minor abnormality, no attention required (code 2); major abnormality, no attention required (code 3); abnormality, attention required (code 4); major abnormality, physician aware (code 5); undictated examination, films not returned (code 6); unsatisfactory or incomplete examination (code 7); possible malignancy, attention required (code 8); and interesting case (code 9). Many of the codes were not directly related to patient safety, and those related to patient safety were overlapping and occasionally confusing. When coding was made a mandatory reporting function, the initial coding scheme was simplified to include only three codes related to patient safety.
The final diagnostic coding scheme was as follows: code 1, no unexpected findings requiring follow-up; code 4, detection of an unexpected acute finding (not cancer) requires follow-up; and code 8, detection of an unexpected finding indicating malignancy requires follow-up. When a significant unexpected finding such as malignancy was found, the interpreting radiologist contacted the referring clinician or appropriate member of the clinical team caring for the patient by telephone or, rarely, by secure e-mail. The radiologist documented the contact in the report. A diagnostic code 8 was dictated by the interpreting radiologist and was coded in the electronic record. An equivocal finding for which the interpreting radiologist recommended further evaluation because of a concern for malignancy, no matter how equivocal the finding, was assigned a code 8, and the referring clinician was appropriately notified. Coding was mandatory; that is, every radiology report had to be coded with one of these three codes before the report was finalized.
On a weekly basis, the cancer registrar, a nurse practitioner who maintains a database of all oncology patients in the institution, retrieved a list of reports that had been coded as possible malignancy. The registrar monitored the electronic clinical record for documentation of appropriate follow-up, including reviewing all patient notes, patient appointments, and additional radiology or pathology visits scheduled. If the registrar could not ascertain that the referring provider was aware of the unexpected finding of possible malignancy or that additional investigation had been ordered, if required, within 2 weeks after initial radiologic interpretation, the registrar contacted the ordering provider to ascertain why no additional follow-up had been arranged.
In addition, the registrar notified the hospital's tumor board, which included physicians and section chiefs from various departments including pathology, radiology, radiation oncology, nuclear medicine, general surgery, and medicine oncology; and selective involvement on an as-needed basis by physicians from the departments of ear, nose, and throat; urology; and pulmonary medicine. After the ordering provider or the tumor board or both were notified, the responsibility for tracking additional patient evaluation was delegated to the nurse practitioner in charge of active cancer cases.
Predictors of Appropriate Follow-Up
We analyzed the type of imaging examination and the referring service as predictors of documented appropriate follow-up of code 8 cases. Referring services included urgent care and emergency department, primary medicine clinic or general medicine, non-primary care surgery, and non-primary care medicine services. Further, because physicians and nurse practitioners can have primary responsibility for a patient in the primary medicine clinic, we also evaluated this variable.
Statistical significance was tested using the chisquare test.
Results
From April 2003 to March 2004, 37,736 radiology examinations were performed at our institution, and all were coded. The studies performed included conventional radiography, sonography, CT, MRI, barium examinations, excretory urography, myelography, and angiography. No mammography is performed at our institute. During this period, 395 cases were given a code 8, representing an unexpected finding of possible malignancy (Fig. 1). Of all the studies coded as unexpected possible malignancy, the electronic medical record documented that 360 cases were appropriately investigated within a 2-week period. Of these 360 cases of suspected malignancy, 163 were found to be malignant and 188 were not malignant. Of the nine cases in which the diagnosis was not known, five patients refused additional follow-up; the results of additional imaging or evaluation were pending in the remaining four patients at the time of this writing.
Thirty-five patients did not have any follow-up noted in the electronic medical record within 2 weeks of interpretation. When the tumor registrar contacted the ordering physician or nurse practitioner or the tumor board, in 27 patients physicians were aware of the unexpected finding on imaging and follow-up investigation had been initiated on 25 patients but had not been documented in the electronic medical record. One patient who had multiple medical problems died within 3 weeks after the initial investigation, and one patient did not want to follow up on his additional findings. The only common causes for delay in this group were the lack of proper contact information in four patients and the presence of multiple medical problems in three patients in whom the referring physician did not see an immediate need to follow up on the unexpected radiologic finding.
Eight patients would have been completely lost to follow-up if this semiautomated coding and review process had not been instituted. Of these eight patients, five were ultimately diagnosed with a malignancy. These eight patients represent 2.0% of all examinations receiving code 8 for suspected malignancy and 0.02% of all imaging tests performed within the department in the 12 months of this study.
The type of imaging test did not predict appropriate follow-up (p = 0.18). Patients referred from the urgent care and emergency department service were significantly more likely to have appropriate documentation of follow-up (p < 0.02). Of the patients referred from non-urgent care and emergency services, the referring services did not predict appropriate follow-up (p = 0.72). In the primary medicine service, no difference was seen between physicians and nurse practitioners in appropriate documentation of follow-up (p = 0.61).
Discussion
Although controversy about the exact size of the medical error continues, a 1999 report by the U.S. Institute of Medicine claimed that between 44,000 and 98,000 hospital deaths per year in the United States were due to medical errors [2, 6]. Medical errors have been extensively studied both here and in other countries that have reported patient harm [7-13]. When patients go to hospitals for treatment, they need to feel confident that the procedures performed will optimize their safety and outcomes. However, patient harm frequently arises from problems in the health care process. It is a human failing to attribute blame simply because there has been a serious outcome. To err is human [2, 14].
When an untoward event occurs, it is often attributed to the failure of an individual [2, 15, 16]. The medical malpractice system stresses the punishment of the individual who errs. Although the current tort system focuses on the individual who has committed the error, it does not necessarily prevent errors from recurring. The solution to error lies not in accountability or punishment of individuals, but rather in redesigning the system to reduce the risk of error and to limit and manage the consequences when errors do occur.
We had previously conducted a root cause analysis and devised a system redundancy to address a communication risk that existed at our institution [3]. From our analysis of the efficacy of this risk reduction process, the probability of a patient not receiving appropriate follow-up is a tiny fraction of all cases seen in the department as a whole. One may argue, however, that eight patients for whom appropriate follow-up was not instituted are eight patients too many. Indeed, this highlights the value of instituting a coding and review process.
The standard of care [17, 18] generally holds that notification of the referring physician by the interpreting radiologist should occur in a manner that reasonably ensures receipt of the findings. However, that manner may be insufficient to ensure appropriate follow-up clinical management. The safety net we devised includes a mandatory code embedded in the reporting process and automatic generation of a weekly list of cases coded as “unexpected finding, probable malignancy” (code 8). We leveraged resources that were available in the institution by extending the responsibilities of the existing cancer registrar, a nurse practitioner who maintains a database of all cancer patients in the institution, to include weekly checks of patients whose reports were coded 8. This patient list averaged seven or eight patients per week. The instituted mechanism provides an additional level of safety in the communication of critical radiology findings at minimal additional cost to the department and no additional burden on the interpreting radiologist.
The most surprising finding in our study was a statistically significant difference in the rate of appropriate follow-up evaluation between patients referred from urgent care and the emergency department compared with other services. That is, no patients referred from urgent care and the emergency department had a delay in documentation of appropriate management of the unexpected critical finding. For the remainder of the services, no differences in the level of follow-up care were based on the type of referring service. Furthermore, for primary care services in which care is routinely managed by physicians or by nurse practitioners, no differences in the rate of appropriate follow-up evaluation were detected between physician providers or nurse practitioner providers.
The most significant limitation of the study is the lack of baseline data on the number of patients with unexpected critical imaging findings who did not receive appropriate follow-up before institution of the semiautomated notification process, precluding evaluation of interval effect of the notification process. Nevertheless, we have shown that the notification process has contributed to improving patient care. In our university-affiliated hospital, much of the day-to-day care, particularly of inpatients, is provided by residents. The electronic medical record does not permit segregation of the ordering provider by level of training to permit evaluation of this variable. We hypothesize that the monthly rotation of residents through the clinical services may have contributed to a lapse in continuity of care. This suggests that provision of this continuity may, in some respects, lie with the radiology service [3].
The total case volume at our institution is small compared with many large academic centers, where the system we devised may not be easily implemented; however, for many practices with moderate volumes, the method described here may provide a viable additional layer of safety, particularly with the use of integrated PACS and reporting systems with searchable field functions.
Communication errors are a common source of untoward outcomes in radiology, as evidenced by malpractice claims reports [19-21]. Many communication problems that contribute to patient error are structural in nature. We have addressed one of the structural issues by leveraging the existing integrated radiology reporting system and instituting a low-intensity mechanism that adds a level of redundancy at minimal additional cost and at no additional burden to the radiologist. We believe that the further integration of PACS and reporting systems with searchable field functions will allow the use of these information systems in efforts to improve the quality of care.
Our attempt at creating an effective and infallible system for managing abnormal and critical radiologic findings combines computer technology and case-by-case tracking. This semiautomated process of coding has added an additional level of redundancy to make our environment safer for patients and is designed to reduce morbidity and mortality in patients examined in the radiology department at nominal cost.
Footnote
Address correspondence to V. R. Choksi ([email protected]).
References
1.
Hayward RA, Hofer TP. Estimating hospital deaths due to medical errors: preventability is in the eye of the reviewer. JAMA 2001; 286:415-420
2.
Institute of Medicine. To err is human: building a safer health system. Washington, DC: National Academy Press,2000
3.
Choksi V, Marn C, Piotrowski M, Bell Y, Carlos R. Illustrating the root cause analysis process: creation of safety net with a semi-automated process for notification of critical findings in diagnostic imaging. J Am Coll Radiol 2005; 2:768-776
4.
Office of Safety and Mission Assurance, Chief Engineers Office, National Aeronautics and Space Administration. Root cause analysis overview. July 2003 report. Available at: www.hq.nasa.gov/office/codeq/rca/rootcauseppt.pdf. Accessed January 2005
5.
U.S. Department of Energy. DOE guideline: root cause analysis guidance document, February 1992. Available at: www.eh.doe.gov/techstds/standard/nst1004/nst1004.pdf. Accessed January 2005
6.
Hofer T, Hayward R. Are bad outcomes from questionable clinical decisions preventable medical errors? A case of cascade iatrogenesis. Ann Intern Med 2002; 137(5 part 1):327-333
7.
Brennan TA, Leape LL, Laird NM, et al. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. N Engl J Med 1991; 324:370-376
8.
Leape LL, Brennan TA, Laird NM, et al. The nature of adverse events in hospitalized patients: results of the Harvard Medical Practice Study II. N Engl J Med 1991; 324:377-384
9.
Thomas EJ, Studdert DM, Burstin HR, et al. Incidence and types of adverse events and negligent care in Utah and Colorado. Med Care 2000; 38:261-271
10.
Donaldson L. An organization with a memory: report of an expert group on learning from adverse events in the NHS. Norwich, United Kingdom: The Stationery Office, 2000
11.
Vincent C, Neale G, Woloshynowych M. Adverse events in British hospitals: preliminary retrospective record review. BMJ 2001; 322:517-519
12.
Davis P, Lay-Yee R, Briant R, et al. Adverse events in New Zealand public hospitals: principal findings from a national survey. Wellington, New Zealand: Ministry of Health,2001
13.
Schiøler T, Lipczak H, Pedersen BL, et al. Incidence of adverse events in hospitals: a retrospective study of medical records [in Danish]. Ugeskr Laeger 2001; 163:5370-5378
14.
Stefl ME. To err is human: building a safer health system in 1999. Front Health Serv Manage 2001; 18:1-2
15.
Grasha AF. Into the abyss: seven principles for identifying the causes of and preventing human error in complex systems. Am J Health Syst Pharm 2000; 57:554-564
16.
Wu AW. Medical error: the second victim—the doctor who makes the mistake needs help too. (editorial) BMJ 2000; 320:726-727
17.
American College of Radiology. ACR practice guideline for communication of diagnostic imaging findings. Available at: www.acr.org/s_acr/bin.asp?TrackID=&SID=1&DID=12196&CID=541&VID=2&DOC=File.PDF. Accessed January 2005
18.
Kushner DC, Lucey LL. Diagnostic radiology reporting and communication: the ACR guideline. J Am Coll Radiol 2004; 2:15-21
19.
Berlin L. Communicating findings of radiologic examinations: whither goest the radiologist's duty? AJR 2002; 178:809-815
20.
Berlin L. Statute of limitations and the continuum of care doctrine. AJR 2001; 177:1011-1016
21.
Berlin L. Standards, guidelines, and roses. AJR 2003; 181:945-950
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: December 17, 2004
Accepted: February 25, 2005
First published: November 23, 2012
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.