Applied Radiology

Any missed diagnosis can be a serious medicolegal problem, and perhaps one of the most troublesome of these is a missed diagnosis of lung cancer. In a study of more than 100,000 medicolegal cases, collected from 1985 to 1995 by a large data sharing organization and encompassing all medical specialties, actions involving malignant neoplasms of the bronchus or lung were the sixth most common. Among radiologists, errors in diagnosis related to malignant neoplasms of the bronchus were the second most common cause for litigation, after legal actions arising from malignant neoplasms of the female breast. Approximately 90% of the alleged errors in diagnosis of lung cancer occurred on chest radiographs, 5% on CT scans, and 5% on other studies. Forty-five percent resulted in indemnity payments, which averaged $150,000 (Lori Bartholomew, Physician Insurers Association of America, personal communication).

In view of these statistics, several questions are relevant: 1) Why do radiologists miss lung cancer? 2) When is missed lung cancer malpractice? 3) What can be done to reduce the likelihood of missed lung cancer?

Why do radiologists miss lung cancer?

Factors that contribute to missed lung cancer can be categorized as arising from observer error, lesion characteristics, or technical considerations.1

Observer error is probably the most important factor. Kundel et al have described three types of observer error.2 A scanning error occurs if a nodule is not fixated during the 350 msec that a lesion is focused on the fovea. A recognition error results if a lesion is scanned adequately but is not detected. A decision-making error (figures 1,2) is caused by incorrect interpretation of a recognized abnormality as normal. Among 20 errors by 4 observers who interpreted 36 chest radiographs with simulated nodules, scanning errors accounted for 30%, recognition errors 25%, and decision-making errors accounted for 45% of observer errors.2

Satisfaction of search error, or "tunnel vision," is another observer error that may contribute to missed lung cancer. This error occurs if the radiologist is distracted by important but unrelated radiographic abnormalities, leading to failure to diagnose a lung cancer (figure 1). A recent study demonstrated reduced nodule detectability on radiographs with other major abnormalities as compared with normal radiographs, suggesting that "tunnel vision" is an important cause of observer error.3

Lesion characteristics contribute substantially to the likelihood of diagnosing an early lung cancer. For example, the size of a lesion is of great importance (figure 3). The smallest lesion that can be visualized on chest radiographs, even in retrospect, is 4 mm.4 Studies using simulated nodules of 1 cm on chest radiographs have demonstrated a false negative error rate ranging from 40% to 87%.1 However, in a retrospective study of overlooked lung cancer by Austin et al, 31% of overlooked lesions were greater than 2 cm in diameter.5 Thus, it is clear that lesion characteristics other than size affect detection of lung nodules.

Lesion conspicuity is another crucial factor in the detectability of lung cancer.1 Conspicuity refers to the extent to which a lesion is distinguishable from adjacent opacities that may impair its visibility (or more simply, the extent to which it "stands out"). The term encompasses both the density of the nodule itself and the surrounding structures that reduce visibility such as ribs, clavicles, or blood vessels that project adjacent to or over the nodule (figure 3). Although lesion conspicuity is a useful concept, a rigorous definition has not been agreed on.

Technical considerations also may have a role in a failure to diagnose lung cancer.1 An anteroposterior examination can be a contributing factor, particularly if portable technique is used. In the recent study by Latief et al, the mean diameter of overlooked lung cancer

on anteroposterior radiographs was 2.3 cm.6 Additionally, patient motion or a suboptimal inspiratory effort increases the likelihood that a lung lesion will be overlooked. Factors such as film contrast, density, and kvp also affect lesion detectability (figure 3). In general, a wide-latitude, low-contrast technique (130-140 kvp) is preferred because it provides more uniform film exposure than a low kvp technique.

When is missed lung cancer malpractice?

It has been argued by plaintiff attorneys that any missed lung cancer constitutes malpractice, but there is compelling evidence in the radiologic literature that overlooked lung cancer should not necessarily lead to a successful medicolegal action.7 Muhm et al used chest radio-graphs to screen for early lung cancer at 4-month intervals in a large cohort of men who were heavy smokers.8 Of the 50 peripheral lung cancers diagnosed, 45 (90%) were visible in retrospect on a radiograph taken 4 months earlier. Four cancers (8%) were visible in retrospect 2 years prior to diagnosis, and one was visible 53 months earlier. Twelve (75%) of 16 perihilar lesions also were visible on prior examinations.8 In this study, failure to diagnose lesions occurred despite a high level of suspicion on the part of at least two interpreters. This study seems to give weight to the belief that expecting radiologists to diagnose all lung cancers is unreasonable.

In general, negligence is the legal criterion that is used to decide whether malpractice has occurred in cases of missed lung cancer, or other missed diagnoses.7 Negligence can be proven if a radiologist breaches the standard of care by failing to diagnose a lung cancer, and if this failure to diagnose is a proximate cause of substantial injury to the patient. Although it is possible to breach the standard of care and not be found liable because of a lack of proximate cause or substantial injury, it is far easier to argue that the standard of care was not breached. Unfortunately, even this defense is not straightforward.

Because there is no rigorous definition as to what constitutes standard of care in the diagnosis of early lung cancer, the final decision typically relies on the ability of the experts of each of the contesting parties to persuade the judge or jury that their definition is correct. In this endeavor, the conspicuity of a lesion (as defined above) probably is the single most important factor in determining whether the standard of care has been breached. A lesion of high conspicuity (i.e., an "obvious" lesion) is far more likely to be associated with an adverse legal outcome than a lesion of lower conspicuity.7

What can be done to reduce the likelihood of missed lung cancer?

As is evident from the preceding discussion, complete elimination of the missed diagnosis of lung cancer is an unrealistic goal. If radiologists were to lower the threshold for classifying a finding on chest radiographs as suspicious, observer error would still result in failure to detect some lesions. Moreover, the resources required to evaluate every such finding would be enormous, and the prevalence of lung cancer could be quite low. Therefore, a more appropriate goal would be to substantially reduce the rate of overlooked lung cancer, especially among lesions of high conspicuity.

Several methods of error rate reduction have been suggested. In the recent study of lung cancer litigation by the Physician Insurers Association of America, an organization of liability companies owned or directed by physicians and dentists, a major factor in the occurrence of overlooked lung cancer was failure to compare the radiograph on which misdiagnosis was alleged with pertinent prior examinations.9 Similarly, in a recent study of missed bronchogenic cancer, failure to compare with one or a sequence of prior radiographs was the most common cause of interpretive error.5

Another approach to decreasing error would be to emphasize the need to avoid satisfaction of search error, as this appears to be an important contributor to diagnostic failure in the presence of unrelated major findings. Finally, the use of two independent interpreters, or double-reading, has been recommended.10 Unfortunately, the study of Muhm et al, in which two interpreters were used, showed that even this method does not guarantee a successful diagnosis.8

Missed lung cancer on CT

Missed lung cancer also can occur on CT. Two recent studies have investigated this problem in 23 patients.11,12 Many of the overlooked lesions were very small (less than 4 mm). However, some larger lesions also were missed, particularly those in an endobronchial location (figure 4). Several lawsuits have arisen as a result of overlooked lung cancer on CT.

Conclusion

Missed lung cancer remains a serious medicolegal issue despite widespread awareness by radiologists of the problem. The reasons for errors are many, but frequently they are due to observer performance and the specific characteristics of the undetected lesion. Missed lung cancer is often unavoidable and does not always result in litigation, but overlooked lesions of high conspicuity are more likely to be associated with an unfavorable legal outcome. No approach can be expected to eliminate the occurrence of missed lung cancer entirely, but rigorous comparison of the current examination with one or more prior radiographs appears to be advisable. AR

References

1. Brogdon BG, Kelsey CA, Moseley RD: Factors affecting perception of pulmonary lesions. Radiol Clin North Am 21:633-654, 1983.

2. Kundel HL, Nodine CF, Carmody D: Visual scanning, pattern recognition, and decision making in pulmonary nodule detection. Invest Radiol 13:175-181, 1978.

3. Samuel S, Kundel HI, Nodine CF, Toto LC: Mechanism of satisfaction of search: Eye position recordings in the reading of chest radiographs. Radiology 194:895-892, 1995.

4. Breckenridge JW, Bird GC: Errors of omission in pulmonary nodule detection. Appl Radiol 6:51-54, 1977.

5. Austin JHM, Romney BM, Goldsmith LS: Missed bronchogenic carcinoma: Radiographic findings in 27 patients with a potentially resectable lesion evident in retrospect. Radiology 182:115-122, 1992.

6. Latief K, White CS, Protopapas Z, et al: Search for a primary lung neoplasm in patients with brain metastasis: Is the chest radiograph sufficient? AJR 168:1339-1344, 1997.

7. Potchen EJ, Bisesi MA: When is it malpractice to miss lung cancer on chest radiographs? Radiology 175:29-32, 1990.

8. Muhm JR, Miller WE, Fontana RS, et al: Lung cancer detected during a screening program using four-month chest radiographs. Radiology 148:609-615, 1983.

9. Physicians Insurers Association of America, Lung Cancer Study, 1992.

10. Forrest JV, Friedman PJ: Radiologic errors in patients with lung cancer. West J Med 134:485-490,1981.

11. White CS, Romney BM, Mason AC, et al: Primary carcinoma of the lung overlooked at CT: Analysis of findings in 14 patients. Radiology 199:109-116, 1996.

12. Gurney JW: Missed lung cancer at CT: Imaging findings in nine patients. Radiology 199:117-122, 1996.

Dr. White and Dr. Meyer are in the Department of Diagnostic Radiology at the University of Maryland Medical Center in Baltimore, MD.