CEO SUMMARY: What does a lab do when it discovers that it has reported inaccurate test results? In 2004, a turnaround team arrived at the laboratory of Maryland General Hospital in Baltimore to deal with the consequences of a failed infectious disease testing program. For about two years, the lab had reported inaccurate HCV and HIV results. One member of that turnaround team shares lessons learned that pathologists can use to develop effective contingency plans for their own labs.
NEWS THAT THE NATION’S LARGEST LABORATORY reported erroneous Vitamin D results to as many as hundreds of thousands of patients during 2007 and 2008 became national news last week.
For pathologists and lab managers, it is a reminder that every laboratory should have a contingency plan that addresses the consequences of having reported inaccurate results, particularly when large numbers of patients are involved.
Richard Ouellette is one laboratory professional who has direct experience in dealing with a laboratory’s internal failures and the unfavorable national news coverage that results. In 2004, he was among the first lab experts retained to correct the deep-rooted, systemic problems in the laboratory at Baltimore’s 242-bed Maryland General Hospital (MGH).
Maryland state health officials had determined that over the 14 months beginning in June 2002, HIV and HCV testing done at the hospital’s laboratory had produced unreliable results and the Baltimore Sun immediately made this into a national story. Public health officials estimated that, as a result of flawed testing during that time, at least 4,500 individuals tested for HIV and HCV had been given potentially inaccurate results. (See TDRs, April 26 and May 17, 2004.)
In the wake of this lab testing scandal, MGH President and CEO Timothy D. Miller resigned. The hospital’s Director of Pathology and Laboratory, Philip J. Whalen, M.D., also resigned. James Stewart, Administrative Director of Laboratory Services during the time of the lab’s systemic failures, was placed on leave and resigned several months later.
Role As Interim Lab Manager
Ouellette, who currently is President and CEO of Management Decision Systems, Inc., in Holden, Massachusetts, came to Maryland General Hospital as a contractor with Chi Solutions, Inc. He shared the interim laboratory manager’s role at the MGH lab and then shortly thereafter took over as project manager. His company’s quality management systems, MAST (Management Accountability, Service & Staffing Tracker), and the ORF Tracker risk management system, were installed at the MGH laboratory during this time as well.
In this role, Ouellette had input in shaping the laboratory’s response to the public and the media, while taking the corrective actions needed to find the patients (and their physicians) who had received the erroneous results. These doc- tors and patients were then notified of the inaccurate results, along with the need to perform a retest.
Massive Effort For Retest
This required a massive effort. The retest strategy and plan at Maryland General Hospitals was developed by a 15-member multidisciplinary team. Included were pathologists, physicians, laboratory consultants, executives and members of the community.
THE DARK REPORT asked Ouellette to share his experience and make recommendations about how a lab should react when errors are made, such as in the case of the erroneous Vitamin 25(OH) D lab test results that Quest Diagnostics Incorporated has admitted reporting to physicians and their patients during 2007 and 2008.
“When the public learns that a laboratory allowed systemic errors and reported inaccurate results to patients, that laboratory will find itself in the center of a hurricane of criticism,” observed Ouellette.
“This is the moment when a good crisis management plan and a coordinated team effort makes all the difference in whether a laboratory can recover from the unfavorable publicity—and whether the laboratory will be able to re-establish trust with its referring physicians and patients,” he added. “Every lab should have a contingency plan which addresses two dimensions of the problem. The most important dimension is mandated by CLIA and relevant regulatory processes, informed by ethics and the appropriate clinical issues which come into play. The second dimension is the public reputation of the laboratory.
“After any mistake, there are processes mandated by CLIA to which labs must adhere,” continued Ouellette. “The goal of these quality systems is to establish standard operating protocols and quality control programs so that, as the laboratory conducts testing, the results are deemed accurate and reflective of each patient’s condition.
“The quality process takes into account the pre-analytical and the analytic testing processes,” he noted. “One goal of these protocols is to immediately identify and correct errors. That means corrections should be made before any results are reported to physicians and their patients.
Inaccurate Test Result
“Errors happen in every laboratory,” Ouellette continued. “An inaccurate result could go out for any number of reasons. If either the laboratory or the physician discovers the inaccurate result, the protocol is for the lab to immediately retest, assuming they still have the same specimen. If the result of the retest is different from what was originally reported, then a corrected report process takes place.
“The initial result remains on the patient’s record,” he said. “It is highlighted manually or electronically to: 1) let people know that it was the wrong result; and, 2) guide them to the corrected report. The lab notifies the physician of that error. The laboratory also generates an incident report and takes corrective action. Next, an assessment takes place to determine what caused the incorrect patient test result.
“Equally important in this process is to implement corrections to procedures and processes that, when followed, verify that the error will not take place in the future,” advised Ouellette. “Ultimately, the laboratory must conduct a rigorous root-cause analysis to determine what caused the error and how to prevent it from happening in the future. Some laboratories will have already performed a FMEA (Failure Mode and Effects Analysis), a proactive tool and quality method that enables the identification and prevention of testing process errors before they occur. The goal of this effort is to avoid adverse events that could potentially cause harm to patients, employees or others in the pre-analytic, analytic, and post-analytic phases of the testing process.
Root Cause Analysis
“The root cause analysis will help find the breakdown in the lab’s existing testing processes, as well as breakdowns in the lab’s quality assurance and quality improvement processes,” he said. “By the way, all lab accreditation agencies expect that, in these situations, the laboratory will do a root cause analysis and implement the required corrections to ensure that this source of inaccurate test results does not recur.
“What was different about the systemic problems at Maryland General Hospital—and what we are learning about failures in Quest Diagnostics’ Vitamin D testing program—is that errors at these labs reached what I will call an almost ‘catastrophic’ level,” Ouellette related. “It wasn’t a random event that happened, after which the lab then isolated the cause and fixed the problem.
Thousands Of Patients
“At the MGH lab, thousands of patients over several years were involved,” continued Ouellette. “Quest Diagnostics is telling the press that its systemic problems with Vitamin D testing occurred during an 18-month period. It has acknowledged that about 7% of Vitamin D results were inaccurate during that period. So it is likely that a very large number of patients are involved in its retesting program.
“In each example above, the laboratory must answer the question: “Do we know if the error caused harm to a patient?” commented Ouellette. “Generally, determining if there was any harm to a patient comes down to a judgment call on the part of the medical director who is ultimately responsible for laboratory operations. It also involves an intensive investigation of the affected patient’s medical record, their diagnosis, and any other testing that may have been performed at the time.
“I can share with you the process we used to address these systemic problems at the MGH laboratory,” offered Ouellette. “It was a time when the lab and the entire hospital were undergoing intense scrutiny by Maryland health authorities, CMS officials, as well as by inspectors from The Joint Commission (then JCAHO), and CAP. We were meticulous in our assessment of the issues. MGH followed the ethical high ground in dealing with affected patients and physicians, as well as with the media and health officials.
“If the laboratory has good reasons to believe that no patient was harmed, then the laboratory can start to make corrections that will prevent this type of error from happening again,” he stated. “However, if there was patient harm—or if the lab is not sure if any patient was harmed—then the laboratory must institute the patient retesting protocol.
A Judgment Call
“In resolving the known lab errors at Maryland General Hospital, the judgment of the new medical director was to follow the patient retesting protocol,” Ouellette explained.
“However, this is not a simple black/white decision, particularly when inaccurate results were reported by the laboratory to large numbers of patients over a period of several years,” he added. “The medical director makes the decision to institute a patient retest protocol, which then triggers several questions that the lab must answer.
“Some labs might ask, for example, ‘Is there any wiggle room when making this assessment?’ Yes! Absolutely there is wiggle room,” stated Ouellette. “At the same time, the laboratory must keep in mind that what keeps labs on the straight and narrow is intense regulatory scrutiny and their continued ability to provide patient care in their community.
“Inspectors from the state health agency, the federal Centers for Medicare & Medicaid Services (CMS), or the College of American Pathology could show up at any time to inspect the laboratory,” he explained. “Those inspectors could say, ‘Wait. You didn’t do the error assessment appropriately,’ or ‘We disagree with your assessment.’
Billing For Medicare Testing
“If CMS determines there’s a problem, it can affect a laboratory’s ability to bill for testing provided to Medicare/Medicaid beneficiaries and be paid for services,” Ouellette added. “The pathologist’s license can be jeopardized or revoked as well.
“If a mistake occurs in a hospital laboratory, it can cause the parent hospital to come under regulatory and public scrutiny that would affect the hospital’s ability to attract patients,” he stated. “For all these reasons, most laboratory professionals will play it by the book when a mistake is made that might affect patient care, taking careful steps to properly correct that lab test error.
“In our case, because the lab test errors had occurred during a two-year period, we knew we would have to collect new specimens from many patients before conducting the retest—but first we would have to locate these patients to notify them of the inaccurate lab test result and arrange to collect a new specimen,” recalled Ouellette.
“Another question that we had to answer for each individual patient was how the retest might have different clinical significance for the individual patient,” continued Ouellette, “After all, the health status of patients changes from day to day. Patients may be undergoing therapy. The laboratory must collect all this information to make an informed decision. It is responsible for reporting inaccurate results to the physician and the patient. What impact could those inaccurate results have had on that patient’s course of treatment?
“Inaccurate Vitamin D results reported by Quest Diagnostics illustrate this challenge,” said Ouellette. “Assume that a physician closely monitoring his/her patient had three vitamin D tests done over 18 months. The lab determines that the first Vitamin D test was inaccurate and the second two test results were accurate. What is the proper, ethical course of action for the laboratory in this example?
“Does that lab let the first inaccurate Vitamin D result go, because it has provided accurate results on the two most recent tests?,” he asked. “My perspective is that these questions are clinical, meaning the medical director should answer them.
“Even though the medical director must make the first assessment, my recommendation is that, when the issue is questionable, the laboratory has the responsibility and ethical obligation to notify the attending physician, advise him of the lab test error and its possible consequences to the patient’s health. That gives the attending physician the relevant information he/she needs to provide proper care to the patient.
“The next question is: What happens if the physician or the patient doesn’t cooperate in your lab’s effort to do a retest?” he asked. “In that case, you conduct due diligence and document your efforts to reach the physician and the patient. The laboratory did what was reasonably expected by a community’s standard of care. Now it becomes a legal question and your lab has documentation of its actions.”
Ouellette next turned to another issue that is triggered when a laboratory has reported inaccurate test results. “As the patient retest protocol proceeds, the laboratory must next address its obligations to the payers,” he stated. “The laboratory cannot bill for a retest that is being performed because the original test was erroneous.
“The lab must eat the costs of retesting,” said Ouellette. “Moreover, the laboratory must assume all of the costs associated with the retest program. At MGH, the cost of the testing process was actually the least costly part of the effort. The logistical nightmare of locating and informing physicians and patients was the most challenging and incurred the greatest cost.
“It’s important to remember that mistakes happen,” noted Ouellette. “When mistakes occur, it’s best to have an effective chain of command in the laboratory and open communication throughout the organization. More importantly, the laboratory needs to foster a working environment where staff understands that the potential for errors exists—and when errors are discovered, staff can do the right thing to: 1) report the error; 2) correct the cause of the error; and, 3) notify the referring physician and the patient about the error and arrange for retesting.”
“Staff in your laboratory want to do a good job,” noted Ouellette. “They want to provide accurate, high quality test results to every patient. At MGH, we worked hard to foster communication and a nurturing work environment. A constructive atmosphere may be the single most important key to identifying errors and making effective corrective actions.”
In laboratory medicine today, there are not many lab professionals who have had to clean up after a major laboratory testing failure. Ouellette’s experience at MGH is unique. The unfolding story about systemic failures in Vitamin D testing at Quest Diagnostics is a reminder to all pathologists and lab managers that the best crisis or contingency plan is to always produce reliable, accurate lab test results.
But when things go wrong in a laboratory, it is important to be prepared. That means having a plan to contact those physicians and patients who may have received inaccurate results, as well as a plan for handling what are likely to be hostile news reporters. The high-profile news coverage of the problems at Maryland General Hospital’s lab in 2004 and now at Quest Diagnostics during 2007-2008 demonstrate how quickly a laboratory’s fortunes can change.
Whistleblowers Raise Flag at Maryland General’s Lab
NEWS ABOUT CATASTROPHIC LAB TEST ERRORS in the laboratory at Maryland General Hospital broke on March 10, 2004. State health officials found, beginning June 2002, that HIV and HCV testing done at the hospital’s laboratory had produced unreliable results over 14 months. The Baltimore Sun was first to report this story.
Within days of this disclosure, public health officials estimated that at least 4,500 individuals tested for HIV and HCV had been given potentially inaccurate results during the 14-month period of flawed testing.
“I’m really quite disturbed. They [laboratory personnel] apparently knew there was a problem,” stated Baltimore Health Commissioner Peter C. Beilenson at the time of the news. Beilenson and Secretary of the Maryland Department of Health Nelson J. Sabatini both stated that two inspections of the laboratory by state officials in January 2004 had uncovered other potential problems in how the laboratory was operated.
Two days later, on March 12, 2004, came another startling disclosure. A medical technologist formerly employed by the hospital laboratory had sent a letter to state health officials in December describing serious safety and accuracy problems in the MGH laboratory. Now it was learned that this med tech was infected with both HIV and HCV, which she attributed to exposure while operating the HIV/HCV testing instrument in the lab.
Public news that hundreds of patients may have received inaccurate results from their HIV and HCV testing created a public relations disaster for the hospital.
Hostile Press and Poor Morale Just Two Challenges at Maryland General Hospital Lab
WHEN RICHARD OUELLETTE, FACHE, MT(ASCP)H, stepped in as an interim project manager at Maryland General Hospital (MGH) in 2004, he faced one of the most difficult challenges a lab manager can face. He was charged with coordinating a team to identify and correct potentially thousands of lab tests errors while rebuilding the laboratory’s processes and protocols from the ground up.
“At MGH, most errors were produced from a faulty instrument,” Ouellette recalled. “A procedure was modified incorrectly with- out being documented properly and that procedure went on for about two years. Over that time, about 4,000 to 5,000 patient tests went out with incorrect results. There were other procedural issues with this lab that needed to be corrected. It was a big challenge to correct flawed procedures so these inaccurate results would not happen again.
As Many As 5,000 Patients
“We reviewed all the patient records at the MGH laboratory and determined that 8,000 samples—representing between 4,000 and 5,000 patients—were likely to be involved in the problems with HIV and hepatitis testing,” he explained. “During our detailed examination of each patient’s medical records, we wanted to determine if any patient was harmed because of inaccurate lab test results. According to CLIA, that’s the key issue in any case involving a lab error: was any patient harmed? An important secondary question is: how were patients affected?
“When the public learned that thousands of patients may have received inaccurate results, this created a hostile press environment that contributed to a morale problem in the MGH laboratory,” recalled Ouellette. “The Baltimore Sun was very critical of the hospital, but the hospital had made a decision not to fight the issues in public or in the press.
“Instead, the hospital issued progress updates and explained what it was doing,” Ouellette said. “Otherwise, it didn’t comment. That’s quite different from other situations in which I have worked, where hospitals hired public relations consultants to respond to criticisms from the press and the public.
“Another challenge in restoring this laboratory into a reliable, quality testing organization was the need to work with different accrediting and licensing agencies,” he said. “During my time at MGH, the laboratory was undergoing inspections by federal and state agencies, as well as the College of American Pathologists and the Joint Commission. The multiplicity of inspecting agencies made the implementation of our laboratory improvement plan much more difficult. Our team needed to balance our limited resources to both respond to these inspections while continuing to improve the laboratory’s performance.
“During our patient retest program, MGH used outside help to locate patients who had been tested within the previous two years,” said Ouellette. “Obviously patients move in and out of the city. Some patients are transients. Other patients are elderly and some may have died since the testing was performed. These are some of the obstacles involved when notifying patients of inaccurate results and arranging to perform a retest.
“The program to notify physicians and patients, then perform the retesting, required about two years,” he concluded. “On the first round of contacts, we reached about 70% of the patients. Toward the end of the process, it became progressively harder to move the total number of patients contacted and retested from 95% to 98%.”