The Miss Rate of Lung Cancer in Chest Radiograph Interpretation and its Impact o
May 1, 1999
The Miss Rate of Lung Cancer in Chest Radiograph Interpretation and its Impact on Stage of Disease
abstract & commentary
Synopsis: A Netherlands study has found that about 19% of patients with lung cancer had the diagnosis missed on the original chest radiograph on which it could be seen in retrospect after the diagnosis was established at a later time. In 57% of the cases, the delay in diagnosis did not influence stage of disease. In 43%, tumors that had originally been T1 in extent became T2. In the majority of missed cases of lung cancer, superimposed structures were mainly responsible for the failure to detect cancer.
Source: Quekel LGBA, et al. Chest 1999;115:720-724.
Although lung cancer is the second most common cancer in both men and women and the leading cause of cancer death in both men and women, chest radiography is generally considered not sufficiently sensitive or specific to be used to screen high risk populations. Part of the sensitivity problem relates to the frequency with which early lung cancer lesions are missed on chest radiograph. Estimates in the literature on the proportion of missed lung cancers on chest radiographs range from 25-90%. However, these numbers emerge from studies of divergent design with distinct end points (for example, one estimated the detection of solitary pulmonary nodules rather than documented lung cancer). Based upon retrospective radiograph reviews, most estimates of the error rate for the detection of lung cancer range between 20-50%.
Quekel and colleagues set out to establish the miss rate for the detection of early lung cancer based on the chest radiograph in a large community hospital in the Netherlands. Three radiologists reviewed the chest radiographs of 259 patients with biopsy-documented non-small-cell lung cancer and nodular lesions on radiograph. By consensus, detectable radiographic abnormalities were missed on 49 (19%) of the patients. As might be expected, missed lesions tended to be smaller than recognized lesions. About three-fourths of missed lesions did not have sharp borders. Although no significant differences were noted between missed and recognized lesions in sites of the lung involved with tumor, superimposing structures were more frequently noted in radiographs with missed lesions.
Delay in diagnosis ranged from a few weeks to more than 24 months. In about 25%, the delay was less than six months, in about 18%, the delay was 6-12 months. One-third of patients had delay of 12-24 months and 25% had a two-year or greater delay. In 28 patients (57%), the missed diagnosis did not result in a change in tumor stage; 22 T1 patients stayed T1 and six T2 patients stayed T2. In 21 patients (43%), T1 patients became T2.
Commentary
In the litigious climate of medicine at the end of the second millenium, we all have experienced (and do so too frequently) being asked one’s medicolegal opinion about the consequences to patients of delays in diagnosis. Most frequently the question takes the following form: "Mr. X is now dying from progressive lung cancer, but we feel that if the radiologist had made the diagnosis earlier, he could have been cured. Would you please testify that during the period between the radiograph with the missed diagnosis and the time of the actual diagnosis that the tumor went from being curable with surgery to incurable?"
Quekel et al address this issue directly in the discussion of their paper. Given the fact that 19% of patients have missed lesions and in 43% of those (8% of the total) the tumor progresses from T1 to T2, Quekel et al estimate that those who progress go from a five-year survival rate of 61% to a five-year survival rate of 38%; a decline in five-year survival of 23%. It is remarkable that the rate of clinical progression in this series appeared to be so low. In this study, 75% of the patients with missed lesions were not diagnosed for periods of more than six months. Yet the fraction with documented disease progression is quite small.
Tumors are missed for a variety of reasons: indistinct borders, too small, masked by overlapping or superimposed structures. However, the 19% miss rate reported here in the setting of a community hospital seems quite good and seems unlikely to be improved upon with current chest radiograph technology. It is possible that application of newer technology, for example, spiral computed tomographic (CT) scanning, may improve the sensitivity of the test. One study from Japan suggested that spiral CT was a better test than plain radiography.1 However, the time and expense associated with a spiral CT make it less cost-effective than would be hoped for in a screening tool. Some less than optimal choices are still being made.
Reference
1. Kaneko M, et al. Radiology 1996;201:798-802.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content