Price pressures haunt digital mammography

As if haunted by some kind of Ghost of Healthcare Future, the latest crop of breast imaging papers reported at the RSNA meeting conveyed an ominous undercurrent of uncertainty. Though the technology is evolving nicely, with growing sophistication in digital modalities that could strongly support or even replace mammography, practical challenges concerning cost-effectiveness remain.

Early results of the government-funded Digital Mammographic Imaging Screening Trial of almost 50,000 women were greeted with much optimism and fanfare in 2005. In contrast, a November 2006 session exploring the data in greater detail was somewhat subdued, planting seeds of doubt about the profitability of digital imaging on a large scale.

New DMIST analyses show that digital mammography is superior for just one subgroup: pre- and perimenopausal women younger than 50 with dense breasts. Screen-film mammography was actually more sensitive in women over 65 with fatty breasts, and, although the difference was not quite statistically significant, the data could indicate a slight trend toward better performance, according to Dr. Etta Pisano, who presented the results.

"We believe improved contrast in digital mammography is likely to be the factor that allows for improved performance [in dense breasts]," said Pisano, chief of breast imaging at the University of North Carolina. "My own hypothesis is that contrast for digital mammography is optimized for dense breasts and not yet optimized for fatty breasts."

Readers' varying abilities may also help explain differences in the performance of digital versus screen-film mammography, but a full analysis of interpretation skills in the trial is not yet available.

The DMIST data were a bit of a letdown for some.

"The government just paid $26 million for this study, and what did they find out? Not much bang for the buck. They didn't get a slam dunk, a finding that digital is great for everyone," said Dr. Robert Schmidt, professor of radiology at University of Chicago Hospitals.

The new DMIST analyses led RSNA attendees to ask whether it would be necessary to keep conventional systems running for women with fatty breasts, alongside pricier new digital units with proven benefits for dense breasts. That approach may not be necessary, but Pisano advised providers to spend more time examining digital cases involving older women with fatty breasts.

Digital mammography needs to promise better performance to pay its way. The technology costs significantly more than conventional equipment, and providers are therefore looking for higher reimbursement.

Medicare has been offering differential reimbursement rates of $86 for screen-film and $135 for digital. Using these fees, DMIST researchers analyzed cost-effectiveness for women of various ages and breast densities, costs for treatment, effects on quality of life, and mortality rates. Digital mammography was cost-effective if used in a targeted fashion for younger women with dense breasts. In these women, it helped to increase the number of cancers detected at the time of screening, while cutting mortality rates.

Less pay, more work?

But the case for screening all women with digital technology was not borne out. In fact, it was more expensive and less effective than using digital mammography in a targeted way, said Dr. Anna Tosteson, a professor of medicine at Dartmouth-Hitchcock Medical Center.

"Digital screening for all women does not provide sufficient health gains to warrant increased costs," Tosteson said.

Digital exam reimbursement from Medicare would need to drop to $92 to approach cost-effectiveness for all women, according to the study. Some insurers have already eliminated the premium rate for digital screening, and it's unclear how the new DMIST results will affect this trend, not to mention the possible effect on Medicare rates.

"Cost issues can't just be ignored. If we were to do the best thing possible, I would advise women to have an MRI, a mammogram, and maybe an ultrasound too," Schmidt said.

Digital screening is also likely to create much more work for radiologists, according to a survey of Society of Breast Imaging members, who are likely to be early adopters of new technology and to have extensive experience in mammography.

About 400 of the society's 1700 members responded to the survey, and half of the respondents are using digital mammography regularly. The survey asked them to estimate how many conventional versus digital mammograms they could read in one hour with no distractions. The number of people who could read 35 or more screen-film mammograms in one hour was three times higher than the digital rate. Average reading time per study was 6.2 minutes for digital versus 1.8 minutes for screen-film.

"In the current setup, digital takes longer to read than screen-film. There is a need to work on improving speed of our interpretation systems," said Dr. Gary Whitman, medical director of the mobile mammography program at the University of Texas M.D. Anderson Cancer Center.

Meanwhile, in Europe, where breast imaging staff salaries and mammogram costs are lower, enthusiasm for digital technology appears to be growing. Benefits of digital technology such as teletransmission may justify higher costs, according to a study published in the European Journal of Radiology in January 2006. Slight reductions in digital costs could begin to allow for a turnover.

Unlike the DMIST cost-effective analysis, which was based purely on Medicare reimbursement rates, this European study stressed the need to account for the full economic impact, including differences in archiving and staffing costs.

Prof. Stefano Ciatto, lead author of the January 2006 study, presented results of a new trial involving women aged 50 to 69 participating in a mammography screening program in Florence from 2004 to 2005. Researchers compared digital versus conventional mammography in two groups of 14,385 patients each. Patients were matched by age and breast density, and the participating radiologists and radiographers were identical for both cohorts.

Full-field digital mammography had a higher recall rate (4.29% versus 3.46%) for radiological abnormalities but a lower recall rate for technical quality problems (.27% versus .50%) compared with conventional mammography. Digital picked up 105 cancers versus 84 for screen-film. Most of the additional cancers were low grade and of intermediate suspicion. Digital also identified more microcalcifications in dense breasts.

It is too soon to tell whether mortality rates will benefit from the higher number of cancers detected and whether digital mammography is efficacious in the screening setting.

"There is no doubt that digital is at least as good as screen-film. We see more, recall more, and detect more cancers, particularly calcifications. That does not necessarily mean you are saving more lives," said Ciatto, head of diagnostic imaging at the Centro per lo Studio e la Prevenzione Oncologica in Florence.

In another study, researchers analyzed data for 23,929 women who were part of the Oslo II randomized screening study, including 16,985 who underwent screen-film mammography and 6944 examined using FFDM. Of 105 cancers detected, 64 were diagnosed on screen-film (cancer detection rate, .38%) and 41 on FFDM (cancer detection rate, .59%). The FFDM recall rate was 4.2% versus 2.5% for screen-film. Researchers also analyzed radiologists' interpretations and found no significant difference in interobserver agreement between the two modalities.

Meanwhile, researchers at the University of North Norway assessed results in the Norwegian Breast Cancer Screening program for women aged 50 to 69. Almost 5000 women received digital mammography versus about 12,000 who had screen-film mammography.

"Digital mammography had a lower average radiation dose relative to screen-film mammography and tended to have higher recall and detection rates," said Solveig Hofvind, who presented results. "Digital mammography can safely be used in a population-based screening program."