Thomas Jefferson University Hospital said Friday that eight patients in 2013 and 2014 tested positive for drug-resistant "superbug" bacteria after being examined with a special kind of tough-to-clean endoscope.

Yet hospital officials said they had found no proof that the devices, called duodenoscopes, had actually transmitted the infections.

Jefferson's finding came after an investigation lasting several months, with help from the U.S. Centers for Disease Control and Prevention - highlighting continued nationwide scrutiny being paid to the complex, $40,000 instruments, which are used to examine the bile and pancreatic ducts.

Hospitals in Los Angeles, Seattle, and the Chicago suburb of Park Ridge have linked duodenoscopes with recent outbreaks of superbugs, which have an estimated mortality rate of 25 percent to 50 percent. The microbes are so nicknamed because they can survive an assault from some of the most potent antibiotics in medicine's arsenal, a class of last-resort drugs called carbapenems.

The Food and Drug Administration has said the complex internal mechanisms of duodenoscopes make them liable to harbor bacteria even when hospitals follow all of the recommended steps for cleaning and disinfection. This week, the FDA issued stricter guidance on the steps that manufacturers should take to demonstrate that all medical devices can be properly cleaned.

Jefferson based its conclusion that the scopes were not to blame for its patients' infections on the results of sophisticated laboratory tests, chief medical officer Anne Docimo said Friday. The evidence suggests that the patients entered the hospital already infected from some other source, such as another health-care facility, Docimo said.

"No definitive link was found" with the scopes used at Jefferson, she said.

Still, the hospital has enhanced its regimen for cleaning the devices and will continue to watch for any signs that they pose a problem, she said.

Jefferson identified the eight cases by searching its records for any patients who tested positive for a superbug who also were treated with one of the scopes between January 2013 and June 1, 2014, she said.

Two of the patients died, but the deaths were deemed "not clearly related" to the infection, the Philadelphia Department of Public Health said last month. The city did not identify Jefferson by name. The hospital confirmed that it was the site of the infections Friday in response to a media inquiry.

In addition to finding that patients were infected with several strains, Jefferson tested the devices themselves for superbugs and found none, Docimo said. She acknowledged, however, that the scopes had been disinfected multiple times before being tested.

Biomedical engineer Lawrence Muscarella, an infection-control consultant who has studied duodenoscopes, praised Jefferson for conducting an investigation but said more work was needed to rule the devices out as the culprits.

Muscarella, who is based in Montgomeryville, said the tests used to detect superbugs on medical equipment are not perfect, sometimes indicating no bacteria are present when they are.

"I would commend Jefferson for coming forward and providing us with some important details," he said. "Further investigation would seem appropriate to see whether the instruments can be confirmed as being the source."

The devices are used in 500,000 procedures a year and are made by three manufacturers: Fujifilm, Pentax Medical, and Olympus, which has U.S. headquarters in Center Valley, Pa.

Among the models of duodenoscope that Jefferson used, according to the city health department, was one from Olympus that lacked FDA clearance. The agency said this month that it had asked the manufacturer for a formal application but decided to allow continued use of the device because it was similar to others that had been previously approved.

Lawsuits already have been filed against Olympus in the wake of infections at Ronald Reagan UCLA Medical Center, where the scopes have been blamed for contributing to two deaths.

Asked if a hospital could be held liable, lawyer Royce Smith said a plaintiff would need to show that the facility had reason to know of or suspect a pattern of such infections in patients treated with unsafe scopes.

"The first rule of medicine is 'do no harm,' " said Smith, cochair of the medical-legal committee of the Philadelphia Bar Association. "That fundamental rule was broken here, and people deserve to know why."

The bacteria can be transmitted by other means, including hand-to-hand contact in hospitals and nursing homes. While such infections are uncommon, they appear to be on the rise.

In a December 2012 city health department survey, 11 hospitals reported that in the previous year, they had at least one case - meaning a patient who was either infected with carbapenem-resistant bacteria or was "colonized" with them but did not get sick. Five of the 11 reported more than 20 cases.