Patient Safety Tip of the Week

June 14, 2011      Failure to Follow Up



We’ve written numerous columns about important test results slipping through the cracks. Maybe some day, when Accountable Care Organizations (ACO’s) and truly integrated health systems exist and interoperability across a wide variety of electronic medical record platforms is functional, tracking and following up on test results may be easier. But in the meantime most systems have areas of significant vulnerability where patients may have tests done and no one follows up on them. Tests pending after hospitalization are one such area (see our Patient Safety Tip of the Week March 1, 2011 “Tests Pending at Discharge”). Often studies done during an inpatient hospitalization may not have results reported at the time a patient is discharged and the fact that those results are pending may never be properly communicated to the physician who will be following the patient after discharge.


But the emergency department is an even greater area of vulnerability to this problem. Think about it. Emergency physicians (and nurse practitioners and physician’s assistants) work shifts in the ED. They may hand off patients to incoming physicians and not be involved in the discharge of that patient. Additionally, some tests are not run immediately (they may only be run certain days of the week or even be sent out to other labs) so results are not available prior to discharging the patient. Also the ED physician often sees only a “wet read” on radiology or imaging studies and the official report (which may differ or contain additional findings not contained in the “wet read”) may come back the next day. And sometimes that emergency physician may never again work a shift at that particular ED. So there is plenty of opportunity for such studies to “fall through the cracks” and potentially endanger patients.


A new article by a physician/attorney (Moore 2011) puts the issue in perspective and discusses many of the legal and liability issues involved. Moore notes that the failure of “duty to follow up” is often the reason for large liability rewards. He provides actual case examples to demonstrate why it is imperative the ED physician (or group) ensure that the patient or the physician who will be following the patient, or preferably both, are notified of any abnormal test results and the need to follow up.


In one case, the ED physician initialed a radiology report showing a lung mass and ordered it faxed to the PCP’s office. However, the PCP never received the report and eventually the emergency physician group (but not the ED physician herself) was held liable when a delayed diagnosis of cancer led to a malpractice suit. In another, a PSA test was ordered in the ED and the abnormal results never followed up. That led to the emergency physician being held liable when the patient found out he had prostate cancer at a later date.


The author goes on to discuss the legal concepts of contributory negligence and comparative fault, which may vary from state to state. The bottom line is that (at least in most states) even when a patient is partially at fault for failure to follow up, a physician may be held at least partially liable if all appropriate steps were not taken to foster appropriate follow-up.


Moore’s recommendations are important.  First, hospitals (and emergency and radiology departments and laboratories) need to have protocols in place for communicating abnormal test results or report discrepancies for patients seen in the ED, and ensure that those protocols are adhered to.


He discusses the critical importance of verbally conveying to the patient the findings and why follow-up is important, including the potential consequences if he/she does not follow up. Documentation of that discussion in the medical record is critical. And, if the patient has not yet left the ED, this should be included in the discharge instructions. Moore also stresses that in some areas where access to follow-up medical care may be limited the ED physician may need to specify whom to contact if he/she is unsuccessful at getting that follow-up.


In addition, every attempt should be made to ensure that there is appropriate notification of the physician who will ultimately follow up on the abnormal test result. And you need to close the loop and ensure that such physicians have, indeed, been made aware of the result and their need to follow up with the patient. Simply faxing the report to the physician without ensuring they actually saw the report is not enough. So if you call or otherwise contact that physician, document it in the chart.


When the patient has already left the ED, every attempt should be made to contact the patient (by phone, written or electronic means – maintaining HIPAA privacy of course) and that contact should also be documented in the medical record. Sometimes certified mail must be used. He also notes that you can’t give up if the patient is not at the address or phone number given to the ED. In such cases, you need to at least check the phone book or other resources to try to locate the patient.


And one very good recommendation: avoid ordering unnecessary tests! And, if you do order them, make sure you document who is going to follow up on that result (the examples given are ordering a cholesterol or PSA for convenience of the patient).


But the ED and emergency physicians are not the only ones with potential liability. The radiologists and radiology departments also have duties to ensure that appropriate notification of patients and recommendations for follow-up have taken place. We’ve previously discussed some of the systems radiology departments have put into place (see articles listed at the end of today’s column).


Our March 9, 2010 Patient Safety Tip of the Week “Communication of Urgent or Unexpected Radiology Findings” noted a Pennsylvania Patient Safety Advisory on communication of radiograph discrepancies between radiology and emergency departments. The Pennsylvania Patient Safety Authority had received over 3000 reports of ER radiology discrepancies over a 4 year period and has some good risk reduction strategies. They recommend a system be in place to review discrepancies between the ER physician interpretation and the radiologist interpretation for all shifts. The system can be either paper-based or use notations on the PACS system. But the key thing is that it is a two-way system. The radiologist needs to see how the ER physician interpreted the study and vice versa. Discrepancies must be communicated to the ER in a timely fashion. The method of communication is most often verbal between practitioners and such communication should be documented in the patient’s medical record. And then a system must be in place for timely communication of the discrepancies and findings to the referring physician or physician who will be assuming care of the patient or the patient himself. There has been an increasing trend for radiologists to communicate directly with patients regarding the findings.


A new study (Anthony 2011) describes the policy and protocol for action on critical results put in place at the Brigham and Women’s Hospital in Boston and how compliance with the follow-up improved from 28% to over 90% with their quality improvement program. They used the Joint Commission requirement for reporting critical test results and recommendations from the American College of Radiology, the Massachusetts Hospital Association, and Massachusetts Coalition for the Prevention of Medical Errors to develop and implement their policy.


First they had to define what critical results were. They then stratified them into priority categories corresponding to the urgency with which notification was needed. “Red” alerts (eg. for ischemic bowel or intracerebral hemorrhage) required notification within 60 minutes. “Orange” alerts were for conditions which could lead to significant morbidity if not addressed within 2-3 days (eg. impending fracture or intraabdominal abscess). “Yellow” alerts were for conditions not immediately life-threatening but could lead to significant morbidity if not addressed (eg. solitary pulmonary nodule or renal mass). Notification for these was required within 3 days.


They then set up an “escalation” process for assuring timely communication of test results. This included the referring/ordering physician, covering physician, housestaff, other member of the care team, attending physician, chief of service, department chair and chief medical officer in that order. For outpatients cared for by patients outside the medical center, if the patient’s physician could not be contacted, the patient should be contacted for red or orange alerts and told to go to the medical center’s emergency department (or nearest ED) or to follow up with their physician for yellow alerts. Red and orange alerts required face-to-face or telephone contact. Yellow alerts could be via face-to-face or telephone contact or other “verifiable” method of communication. And the policy required documentation of the communication in the medical record.


About 10% of their studies had results falling into one of the 3 alert categories, though the percentage varied substantially by type of study and origin of the patient. For instance about 25% of studies done on emergency patients or patients referred for nuclear medicine tests required communication of abnormal results, whereas less than 5% of those having musculoskeletal imaging needed such alerts.


The authors point out that there is no current national standard of what results are considered “critical” or “alertable”. They note that almost 10% of their tests resulted in findings in one of the 3 alertable categories so that the communication process could be time-consuming and interruptive for both radiologists and recipients of the alerts. Thus potentially automating the process may be a desired future enhancement.


So you need to set up a system in your organization for identification of pending tests and one for notification when results are abnormal. In the ED, identification of pending tests could be as simple as using paper-based checklists to determine whether ordered tests have been resulted or not. Or you can develop more complicated rules-based algorithms to identify those tests with pending results in your electronic medical records. If your ED physicians do not return for several days (or never return) you need to have another person designated to identify test results as they come back  and close the loop with both the patients and the physicians who will be rendering subsequent care. Then follow the advice given in the above articles to make the appropriate contacts and document your efforts in the medical records.


Note also one of the problems we have noted in several hospital systems is that the database of referring physician addresses and phone numbers is often not up-to-date. So faxing a report to the fax number you have on hand is not adequate. You need some way to ensure that the physician received the full report and will take action on it.




See also our other columns on communicating significant results:










Moore GP. Liability of Emergency Physicians for Studies Ordered in the Emergency Department: Court Cases and Legal Defenses. J Emerg Med 2011; 40(2): 225-228



Anthony SG, Prevedello LM, Damiano MM, et al. Impact of a 4-year Quality Improvement Initiative to Improve Communication of Critical Imaging Test Results. Radiology 2011; 259(3): 802-807



PPSA Communication of Radiograph Discrepancies between Radiology and Emergency Departments

Pa Patient Saf Advis 2010; 7(1): 18-22
















Tip of the Week Archive


What’s New in the Patient Safety World Archive