sábado, 12 de enero de 2013

AHRQ WebM&M: Morbidity and Mortality Rounds on the Web

AHRQ WebM&M: Morbidity and Mortality Rounds on the Web
 Agency for Healthcare Research and Quality

U.S. Department of Health and Human Services www.hhs.gov

December Issue of AHRQ Web M&M Examines Imaging for Benign Lung Nodules

The December issue of AHRQ Web M&M features a Spotlight Case that describes a case involving a patient with a benign nodule on the lung who underwent more than 20 CT scans over an 8-year period.  While the physical harm caused by the additional CT scans is debatable, some studies have estimated that as few as 1 in 300 patients who receive scans will develop cancer.  The economic impact is more clear, with each scan costing several hundred or even thousands of dollars.   A commentary, written by Alex Balekian, M.D., of the University of Southern California Keck School of Medicine, and Michael K. Gould, M.D., of Kaiser Permanente Southern California, provide best practices on management of small and larger pulmonary nodules.  The Perspectives on Safety section features an interview with Sharon K. Inouye, M.D., at Harvard Medical School, on methods to prevent delirium in hospitalized patients.  Physicians and nurses can receive free CME, CEU, or training certification by taking the Spotlight Quiz.  Select to access AHRQ's Web M&M.
 full-text ► and MORE
 AHRQ WebM&M: Morbidity and Mortality Rounds on the Web
December 2012


Commentary by Alex A. Balekian, MD, MSHS, and Michael K. Gould, MD, MS

























































Case & Commentary—Part 1:


After moving to a new city, a 67-year-old man presented to a primary care physician for an initial visit to establish care. In discussing his past medical history, the patient described having a "spot" on his lungs that doctors had been following since 2004.
The solitary pulmonary nodule (SPN) can be defined as a single, well-circumscribed radiographic density measuring less than 3 cm, surrounded by aerated lung, and without any evidence of atelectasis, hilar enlargement, or pleural effusion.(1) Depending on the study population, between 15% and 75% of these nodules prove to be malignant.(2,3) The differential diagnosis includes malignant, infectious, inflammatory, and miscellaneous benign etiologies (Table 1).

In the past, most nodules were discovered incidentally on plain chest radiographs, as in this case. In recent years, it is more common for nodules to be identified by chest computed tomography (CT). However, nodules detected by CT often do not meet the classic definition of the SPN because either more than one nodule or an accompanying finding like atelectasis is present. Except for obviously calcified nodules (which can be attributed to old granulomatous disease), non-calcified nodules seen on chest radiography should be confirmed and better characterized by chest CT. CT characteristics associated with a benign etiology include smaller size, fat density, or a central, diffuse, or popcorn pattern of calcification. CT characteristics associated with malignancy include larger size, upper lobe location, irregular margins or spiculation, thick-walled cavitation, and hilar or mediastinal lymphadenopathy (Figures 1–3).

Patient-specific risk factors for malignancy include older age, current or former smoking status, prior history of extrathoracic malignancy, and asbestos exposure. Although most clinicians intuitively estimate the probability of malignancy, validated clinical prediction models can help refine decision-making. The Memorial Sloan-Kettering model calculates a patient's 10-year risk of developing lung cancer; although this model estimates risk before any radiographic imaging has been performed, it might be helpful in identifying patients in whom CT screening for lung cancer should be discussed.(4) Two other models—the Mayo Clinic model and the Veterans Affairs model—combine patient-level and radiographic characteristics to estimate the probability of malignancy for patients in whom a nodule has already been identified.(5,6) Based on a patient's probability for malignancy (low, intermediate, high), clinicians can use these models to choose an appropriate management strategy: monitoring with serial CT scans at designated intervals, performing functional imaging (e.g., positron emission tomography [PET]) to characterize the nodule further, or proceeding directly to biopsy (or even surgical resection) without further testing. The Mayo Clinic model has been validated in other populations and is widely used in practice to distinguish between nodules with a low risk of cancer (which can often be followed radiographically) and those with a high risk of cancer (which require tissue diagnosis and/or prompt resection).(7,8)

The most common avoidable error in the management of lung nodules is neglecting to review prior imaging studies. This can be essential in management; for example, if a solid nodule has been present for 2 or more years and has not changed, this is very strong presumptive evidence of a benign etiology, and no additional follow-up is required.(9) For sub-solid (ground glass) nodules, which are frequently premalignant or malignant but slow-growing, it may be necessary to demonstrate longer periods of stability to exclude malignancy.

The next most common error is to choose a strategy of "wait and watch" but neglect to "watch." This error ranges in severity from a minor delay in imaging (such as missing a single examination) to its most severe manifestation, in which the patient is lost to all follow-up either due to miscommunication or lapses in adherence. To date, robust electronic reminder systems for nodule evaluation have not been widely adopted, but there is a clear need for developing and testing these kinds of system-level solutions. Fortunately, the surveillance strategy is typically selected when the probability of cancer is very low or low, so this type of error infrequently results in harm.

Some might argue that it is an error to choose surveillance (as opposed to immediate intervention) for a nodule that ultimately proves to be cancerous. However, as long as this choice was the product of thoughtful deliberation that considered the risk of cancer and the benefits and harms of competing alternatives, we would characterize this as an undesirable outcome that resulted from an appropriate decision. Of note, there have been few systematic studies of outcomes following delayed diagnosis in patients with cancerous nodules, so the magnitude of harm associated with delay is highly uncertain.

Analogously, surgical resection of a benign nodule, although undesirable, may represent an appropriate course of action for a nodule that was considered likely to be malignant. In lung nodule evaluation, as in other areas of medicine, a bad outcome does not necessarily mean that an error was made.

However, it may be an error in judgment to select surveillance for a nodule that is likely to be cancerous, or to select surgical diagnosis for a nodule that is likely to be benign, highlighting the importance of accurately estimating the clinical probability of cancer. Such discordance should be infrequent, except for cases in which there are extenuating circumstances (e.g., strong patient preferences).

< Previous Next >

No hay comentarios: