"More than half of all hospitals now use teleradiology services," said Jonathan Clark, assistant professor of health policy and administration, Penn State. "Hospitals send their X-rays, CT scans, MRIs and other images to outsourcing companies who then forward the images to individual radiologists. Over the course of time, these radiologists gain a tremendous amount of experience by reading images from hundreds, or in some cases thousands, of customers."
But is the experience gained by reading images for so many hospitals the same as the experience gained by repeatedly reading images for just a few hospitals? "There is debate out there about whether or not we should be outsourcing this kind of work," Clark said. "Some say that one CT is the same as another, so it does not really matter if the CT is coming from Hospital A or Hospital B; what matters is that the person reading the image has the right training and experience. The other side of the debate says that radiological images are not commodities and that the process is more nuanced than simply exchanging bits of information over the information super highway. From this perspective a radiologist's performance will improve as he or she learns the nuances of reading images from a particular hospital."
To determine whether learning and performance improvement are customer specific, the researchers examined the experience and productivity of 97 radiologists reading more than 2.7 million images from 1,431 hospitals. "By estimating learning curves, we were able to determine the extent to which a radiologist's productivity reading an image for hospital A was a function of his or her prior experience reading for hospital A, versus his or her prior experience reading the same image for other hospitals," Clark said.
The team found that the radiologist's prior experience with an ordering customer has a greater effect on performance than his or her overall experience reading the same type of image for other customers. "Somebody might look at that finding and say either what we should have is outsourcing radiologists who read for only a few customers or we should not have outsourcing at all," Clark said, "because if you are going to focus a radiologist on one or two customers, then you might as well make the argument that you should have him or her work as an employee for that customer."
The researchers also found that the customer-specific knowledge gained by individual radiologists is aided by the variety of customers with whom a radiologist has experience. "How can it be that both specialization and variety are important?" Clark said. "Our models can only take us so far toward answering this question, but we think the implication is that there are limits to both customer specialization and customer variety, and that the optimum is to have a nice mix of both specialized experience and variety in customers."
In addition, Clark and his colleagues found that the customer specificity of learning diminished as the outsourcing company gained more experience with a particular customer. "The implication is that customer-specific capabilities may also be built at the level of the outsourcing organization as a whole and not just at the level of the individual radiologist," Clark said. "Overall, this finding suggests that while there is customer specificity to learning for outsourcing radiologists, the implication is not that we should not outsource. Rather, our findings provide guidance to outsourcers in terms of how they might think about designing the work of their radiologists to maximize productivity. For new customers, an outsourcing firm may wish to keep individual providers relatively more focused on serving a specific customer. As organizational experience accrues with a specific customer, however, such dedication becomes less critical."
MEDICA.de; Source: Penn State University