Medicare's national coverage determinations in diagnostic radiology: examining evidence and setting limits.

OBJECTIVE: To systematically review and summarize the Center for Medicare and Medicaid Services (CMS) national coverage determination (NCDs) pertaining diagnostic imaging technologies from 1999 through 2010. METHODS: All NCDs pertaining to diagnostic imaging were identified from the Tufts Medical Center NCD database. The variables under study included the quality of the clinical evidence and the final coverage determination. The types of restrictions were categorized. We also categorized the final decisions as "positive coverage" or "no positive/no change in coverage" and assessed the correlation between positive coverage and other variables using Fisher exact test. RESULTS: Twenty-two of 152 (15%) NCDs pertained to diagnostic imaging technologies. The supporting evidence was judge to be good, fair, and poor in 5, 6, and 11 cases, respectively. Eleven technologies (50%) were covered with conditions, four (18%) deferred the coverage decision to local level, and two (9%) were completely not covered. In five instances there was no change to the prior coverage status. Of the 11 decisions resulting in positive coverage, 8 (73%) restricted use to specific population subgroups, 5 (46%) applied restrictions related to treatment, 4 were covered with evidence development, and 2 were restricted to care in specific settings. A significantly higher rate of positive coverage decisions was achieved if the available evidence was good (100% 5/5) or fair (83% 5/6) compared to technologies with poor evidence (10% 1/10) (P < .01). CONCLUSION: CMS has demonstrated a propensity to limit the use of advanced diagnostic imaging to scenarios in which appropriateness is supported by adequate evidence of clinical utility and improved outcomes with the quality of evidence being a significant factor on final decisions. Understanding the need for high-quality evidence and the types of limitations placed on coverage allows for appropriate planning for the incorporation of diagnostic imaging technologies into clinical practice.

Comparative study with a moving heart phantom of the impact of temporal resolution on image quality with two multidetector electrocardiography-gated computed tomography units.

OBJECTIVE: The objective of this study was to compare the temporal resolution-related image quality of electrocardiography-gated images acquired with two multidetector computed tomography (CT) units with a moving heart phantom, at similar fixed heart rates, using half-scan and multisector acquisition modes. METHODS: An adjustable moving heart phantom (Limbsandthings, Horfield, Bristol, UK) was used. Specific heart rates (47, 55, 64, 66, 69, and 73 beats per minute [bpm]) were chosen. On a General Electric CT unit (LightSpeed Plus; General Electric Medical Systems, Milwaukee, WI), retrospective half-scan and multisector mode protocols were performed. On a Siemens CT unit (Somatom Volume Zoom; Siemens, Forchheim, Germany), a retrospective half-scan mode was performed at 47, 55, and 64 bpm, and a two-sector mode was performed at 66, 69, and 73 bpm. Reformatted maximum intensity projection images were qualitatively compared and related to their temporal resolution. RESULTS: Half-scan mode protocols provided similar good results with both CT units up to 55 bpm. The two-sector mode improved image quality compared with the half-scan mode. High temporal resolution with the multisector mode provided the best results. CONCLUSION: For coronary artery imaging, acquisition protocols that provide the highest temporal resolution are mandatory. The multisector mode is one technique that allows high temporal resolution but may be clinically inappropriate at heart rates below 65 bpm or when heart rate variation is observed during scan time.