The activist group Public Citizen asked the secretaries of defense and health and human services on Tuesday to investigate the deaths of two people who received platelet transfusions contaminated with bacteria at the National Institutes of Health’s research hospital this summer.

The transfusions caused infection, shock and multiple-organ failure in cancer patients with compromised immune systems. One died a month later, the other six weeks later, according to a letter sent by Sidney M. Wolfe, the physician who heads Public Citizen’s Health Research Group.

“My conclusion . . . is that the deaths from overwhelming sepsis . . . were entirely preventable had proper blood bank procedures been followed,” he wrote.

NIH officials acknowledged the deaths publicly after Wolfe delivered the letter and made copies available to the news media. NIH spokesman John Burklow said the institution “is deeply saddened by the deaths of two patients who were participants in clinical research” there.

Wolfe said a physician who helped care for one of the patients approached Public Citizen with details of the cases. He described the whistleblowing physician only as “Dr. X” and declined to put a Washington Post reporter in touch with him.

Platelets are small, cell-like structures that are crucial to blood clotting. They circulate in the bloodstream and form the first plugs in damaged or leaky veins and arteries. Platelets are formed in the bone marrow, which is often temporarily damaged by chemotherapy drugs. That causes the number of platelets in circulation to fall, sometimes to dangerous levels. People who have suffered massive trauma, cancer patients and sufferers of a few autoimmune disorders occasionally need platelets to prevent fatal bleeding.

This blood product, however, is in chronically low supply. Unlike blood, which is stored just above freezing, platelets require room temperature to be preserved. As a consequence, their shelf life is less than a week and they are unusually prone to bacterial contamination.

According to Wolfe’s account, the patients received the platelets on July 25 and within hours were gravely ill. Dr. X told Public Citizen he suspected the platelets were to blame and sent the container they were in to be tested. The product was grossly contaminated with bacteria in the Morganella genus. Cultures of the two patients’ blood contained the same strain of bacteria, indicating they were each infected with the same batch of platelets.

The Clinical Center had gotten the material from the National Naval Medical Center, across the street, Wolfe wrote. But Sandy Dean, a spokeswoman for that hospital, said Wolfe’s assertion was wrong. She said the platelets came from Walter Reed Army Medical Center in the District. Asked whether the blood product had been identified as contaminated there, she said that “there was an incident at Walter Reed Army Medical Center.”

Walter Reed and the Naval Medical Center have now merged. The combined institution, called the Walter Reed National Military Medical Center, is on the former Naval Medical Center’s campus in Bethesda.

Burklow, the NIH spokesman, said the platelets “were labeled as suitable for transfusion” when they were given to the patients.

But the hospital “subsequently learned that the platelets were in fact contaminated. The patients and their families were informed and every effort was made to treat their infections,” said Burklow, adding that “we are doing everything we can to make sure this never happens again.”

He declined to answer questions about the sequence of events.

In his letter, which was sent to Defense Secretary Leon E. Panetta and HHS Secretary Kathleen Sebelius, Wolfe said Dr. X was invited to a “root cause analysis” meeting, in which the events of the accident were to be discussed.

An NIH lawyer told those in attendance that what they said “would remain within the room.” Dr. X reportedly said he would not comply with that condition and left.