OIG Says Clarksburg VAMC Protocol Lapses Partly to Blame

U.S. Sen. Joe Manchin, D-WV, left, and VA Secretary Denis McDonough visited the Clarksburg, WV, VAMC in May. Photo from newsroom section of Manchin’s official Senate website

CLARKSBURG, WV — In May, Reta Mays, a former nurse’s assistant at VA’s Louis A. Johnson Medical Center in Clarksburg was sentenced to seven consecutive life sentences—one each for the patients that she had pleaded guilty to murdering. She also pleaded guilty to assault with intent to murder in the death of an eighth veteran under her care. 

While the decision to murder seven people was wholly hers, according to VA’s Office of the Inspector General, Mays’ ability to do so for at least a year before the hospital recognized the pattern was the result of “serious, pervasive, and deep-rooted clinical and administrative failures.”

The OIG report released last month details months of interviews with Clarksburg staff, as well as an analysis of patient records going back to 2015. 

Mays was hired at the Clarksburg VAMC in 2015 and eventually assigned to the overnight shift. As part of her job as a nurse’s assistant, she was authorized to check patients’ vitals, such as glucose levels, although she was not authorized to administer medication.

The first murder to which she admitted occurred in July 2017 when she administered a lethal dose of insulin to Navy veteran Robert Edge Sr., 82. The next murder occurred in January 2018. Two more occurred in March 2018, two in April or May 2018 and one in June 2018.

Even though she administered another overdose in June 2018, that veteran did not die until July, and it could not be proven that insulin was the direct cause of death. Mays has yet to explain her motive for the murders.

According to the OIG report, Clarksburg officials missed multiple opportunities to prevent the string of murders, possibly before they even began.

Prior to being hired by the hospital, Mays was employed as a corrections officer at a regional jail in West Virginia from 2005 to 2012, then at a residential facility from October 2012 to July 2015. OIG was unable to verify that any hiring manager ever contacted those two previous employers to determine Mays’ suitability to work at the VAMC. If they had, they would have learned that, while employed as a corrections officer, Mays was the subject of excessive force allegations.

“Ms. Mays’ conduct while employed by the jail should, at minimum, have prompted additional evaluation, and could have provided a basis for dismissal during her probationary period that ended September 2016,” investigators stated. 

The hospital had another opportunity to complete a security check on Mays when she received the Secretary’s Award for Excellence in Nursing in September 2017. The award came with a $500 check and, because the award was signed by the VA secretary, it required a background investigation of the recipient. According to the report, this was an opportunity for the hospital to notice that a background check had never been originally performed on Mays and to rectify the issue. Instead, the employee tasked with that job signed off on the award, erroneously noting that the background check had been completed upon Mays’ hire.

The OIG thoroughly analyzed the hospital records of the patients under Mays’ care. In addition to the eight patients she admitted to injecting with insulin, facility leaders had identified a ninth during the original investigation into the murders, and OIG investigators identified a 10th during their evaluation. These patients all had “profound and concerning hypoglycemic events.” Of these 10 incidents, eight had no reasonable mitigating factors that would explain the event. 

Despite the unusual nature of the events, however, seven of the patients did not undergo an evaluation to determine the cause of their hypoglycemia. 

“The first step in determining the cause of hypoglycemia is a clinical assessment to identify medical explanations, such as a contributing illness or medication,” the report states. “If the clinical assessment does not reveal the reason for the hypoglycemia, excess insulin is a likely cause and the source of the insulin may either be insulin administered to the patient or produced within the body, most commonly from an insulin-secreting tumor.” 

Because the likelihood of such a tumor is rare, the investigation would have swiftly landed on administered insulin as a likely cause, it adds. 

The report found a number of reasons for why these assessments were not performed. For one, many of the staff were misinformed about hypoglycemia and what such tests would be able to determine the cause of the events. While the facility did not have an endocrinologist on staff, the Clarksburg staff had the opportunity to consult with one in the VA Pittsburgh Healthcare System but never opted to do so.

Pattern Not Apparent

Also, because different staff members were on call when the incidents occurred, the pattern was not immediately apparent. 

“Hypoglycemia caused intentionally is very rare and thus requires a high level of suspicion on the part of the clinician,” the report noted. Still, on two occasions, providers suspected nefarious intent but did not act to confirm those suspicions or follow through with higher-ups. 

A nocturnist who evaluated one of the patients reported suspicions to investigators of surreptitious insulin administration but did not pursue a hypoglycemia workup, notify leaders or alert the clinician who assumed care the following day.

When unexplained hypoglycemia incidents occurred two days in a row in spring 2018, another staff member told OIG investigators of suspicions that “something nefarious could be occurring.” Records show that the staff member ordered insulin and C-peptide levels for the second patient, but, upon questioning, could not recall the tests and “did not appear to understand the utility of testing to help determine if there was surreptitious insulin administered.” 

No documentation was found showing that the staff member ever followed up on the test results, although the staff member reported suspicions to the nurse manager and the Medical Director for Inpatient Services the day the event occurred. Thinking that a nurse may have administered insulin meant for another patient, an assistant nurse manager reviewed the assignment sheet, searching for a patient with diabetes for whom the insulin might have been mean. When none was found, the facility did not conduct additional follow-up. 

According to the report, the employees could not explain why they did not conduct official follow-ups on these events. Some told investigators that they had discussions with their colleagues but that they were informal and focused “on the curiosity of the events rather than concern of wrongdoing.” 

Failures also occurred at the facility level, with the hospital neglecting to conduct rigorous reviews of mortality data to identify outliers. In the two years prior to Mays’ hiring at Clarksburg, there were two spikes in the unadjusted mortality rate, with the highest spike being 4.5%. During the three years Mays was working at the hospital, there were nine spikes, with the highest topping 6%. 

“While mortality spikes do not automatically imply quality-of-care deficits or unlawful activities, they do reflect a change from the facility’s normal mortality pattern,” the investigators stated. “However, responsible staff did not conduct additional reviews related to the mortality rate spikes.”

It was not until late spring 2018 that the pattern of hypoglycemia-related deaths was recognized. The same staff member who had earlier suspected nefarious intent and reported that to the nurse manager treated another patient with an unexplained hypoglycemic attack. This time, the staff member met with the associate chief of staff and discussed the three events of which they were aware and suspicions of criminal activity. A review was ordered—one that would eventually reveal the larger pattern and Mays’ culpability.

The OIG report details a number of other weaknesses in hospital oversight, as well as weaknesses in the facility’s medication management and security. For example, a nursing assistant, Mays was not permitted to access the medication room. Yet, investigators found that all staff on the ward—including nursing assistants—were able to access the room and that some medication carts were left unattended and unlocked. 

These weaknesses were responsible for Mays’ ability to commit her crimes undetected for so long, and proper oversight could have prevented the deaths of some, if not all, of her victims, the report concludes. It includes a number of recommendations to prevent an incident like this from happening again. 

As of early May, the government has reached 10 settlements with families of patients who died under suspicious circumstances at the Clarksburg VAMC, including the veterans Mays confessed to injecting.