Costs don’t just mean financial — EMRs and patient lives

Part One: “Separating myth and evidence about electronic medical records,” here.

The public has heard little about the systematic reviews of the evidence on electronic medical records, which have found no significant benefit in reducing medical errors or improving quality of patient care, safety or health outcomes. The conclusions from those reviews, cautioning that health information technology is being implemented without sound evidence, has also had little media coverage.

The public has heard even less about the increasing numbers of investigators who have been questioning the claim that health information technologies save lives and who have even found that HIT can increase mortality. The media is largely silent about studies finding that EMRs can introduce entirely new types of medical errors and “many unintended and negative consequences.”

The paper that began the maelstrom against EMRs as not necessarily best for patients or patient care was an analysis published in a 2005 issue of Pediatrics. It was a study by physicians from the departments of Critical Care Medicine and Pediatrics at the University of Pittsburgh School of Medicine and Children’s Hospital of Pittsburgh in Pennsylvania.

Lead author Dr. Yong Y. Han, M.D., and colleagues said that the 1999 landmark report, To Err is Human** by the Institute of Medicine’s Committee on Quality of Health Care in America, has sparked safety initiatives at many hospitals focused on emerging health information technologies. As a result of that IOM report, extensive media coverage and congressional hearings have called for HIT as a potential way to reduce human error and ensure that care follows federal guidelines, and stakeholders have made its implementation a benchmark for care standards.

In response to the IOM report, the Children’s Hospital of Pittsburgh implemented hospital-wide a commercially available computerized physician ordering system in October 2002. All personnel had been trained and experts were on hand to support the transition. While HIT can reduce adverse drug events, they wrote, growing concerns are being raised about whether HIT actually improves patient outcomes. Some investigators have said that HIT can, in fact, facilitate risks for errors through “system integration failure” and “human-machine interface flaws,” they said. So, they examined the mortality data for all of the children transferred to their regional medical center over an 18-month period, both before and after HIT was implemented. They focused on these children because they require immediate processing of admission and stabilization orders, the most life-saving benefit suggested by electronic medical systems.

Overall, 3.86% of the children died during the study period. However, mortality rate was 2.8% during the year prior to HIT implementation and increased to 6.57% during the first two quarters of the year following. Adjusting for patient acuity, demographics and clinical categories to isolate the factors associated with increased mortality, they found that shock was associated with a more than 6-fold increase in mortality, followed by nearly a four-fold increased risk (3.71 OR) associated with HIT. This unexpected finding led their efforts to identify the reasons.

They found that the entire usual chain of events in admitting a critically ill child transferred to their center was altered by HIT. Typically, when we’re expecting a critically ill infant in the neonatal intensive care unit, we ready everything possible so that the instant the baby arrives at the ICU, we can resuscitate and swiftly stabilize the baby, and it usually requires the coordinated effort of multiple professionals and a rapid succession of time-critical interventions. These doctors found that HIT disrupted the ability to optimally stabilize and quickly respond to the needs of their critically ill patients:

Before implementation of CPOE [HIT], after radio contact with the transport team, the ICU fellow was allowed to order critical medications/drips, which then were prepared by the bedside ICU nurse in anticipation of patient arrival. When needed, the ICU fellow could also make arrangements for the patient to receive an emergent diagnostic imaging study before coming into the ICU. A full set of admission orders could be written and ready before patient arrival. After CPOE implementation, order entry was not allowed until after the patient had physically arrived to the hospital and been fully registered into the system, leading to potential delays in new therapies and diagnostic testing (this policy later was rectified). The physical process of entering stabilization orders often required an average of ten “clicks” on the computer mouse per order, which translated to about 1 to 2 minutes per single order as compared with a few seconds previously needed to place the same order by written form. Because the vast majority of computer terminals were linked to the hospital computer system via wireless signal, communication bandwidth was often exceeded during peak operational periods, which created additional delays between each click on the computer mouse. Sometimes the computer screen seemed “frozen.”

This initial time burden seemed to change the organization of bedside care. Before CPOE implementation, physicians and nurses converged at the patient’s bedside to stabilize the patient. After CPOE implementation, while 1 physician continued to direct medical management, a second physician was often needed solely to enter orders into the computer during the first 15 minutes to 1 hour if a patient arrived in extremis. Downstream from order entry, bedside nurses were no longer allowed to grab critical medications from a satellite medication dispenser located in the ICU because as part of CPOE implementation, all medications, including vasoactive agents and antibiotics, became centrally located within the pharmacy department. The priority to fill a medication order was assigned by the pharmacy department’s algorithm. Furthermore, because pharmacy could not process medication orders until they had been activated, ICU nurses also spent significant amounts of time at a separate computer terminal and away from the bedside. When the pharmacist accessed the patient CPOE to process an order, the physician and the nurse were “locked out,” further delaying additional order entry.

Communication and coordination of care between care providers were also adversely affected, they found:

Before CPOE implementation, the physician expressed an intended order either through direct oral communication or by writing it at the patient’s bedside (often reinforced with direct oral communication), with the latter giving the nurse a visual cue that a new order had been placed. The nurse had the opportunity to provide immediate feedback, which sometimes resulted in a necessary revision of that order. In addition, these face-to-face interactions often fostered discussions that were relevant to patient care and management. After CPOE implementation, because order entry and activation occurred through a computer interface, often separated by several bed spaces or separate ICU pods, the opportunities for such face-to-face physician–nurse communication were diminished.

Dr. Han and colleagues said that their findings of increased mortality coinciding with HIT implementation, and the problems they experienced, have also been reported by others. These findings, however, don’t support the studies reporting reductions in potential adverse events and improved resource utilization. In trying to understand these seeming contradictions, they said, “we are reminded by Berger and Kichak that ‘although the literature suggests [HIT] systems have the potential to improve patient outcomes through decreased adverse drug events, actual improvements in medical outcomes have not been documented.’” In other words, they said, potential benefits to date have been inferred.

They said that had they just looked at the reduction in medication errors, they might have been tempted to infer improved patient outcomes, too, but their independent investigation found mortality had actually increased. “Our unexpected finding might suggest that surrogate outcome measurements such as ‘medication error rate’ or ‘adverse drug events’ alone may not be sufficient to determine [HIT] efficacy.”

The inherent problems with HIT are that real life doesn’t proceed linearly, they said, and the complexities of patient care cannot be foreseen, let alone be predicted. The fundamental problems with HIT remained despite their best efforts to overcome them. While higher death rates were associated with HIT, they cautioned that their study design precludes statements of cause and effect. Their study did, however, identify clear problems imposed by HIT in the delivery of clinical care and suggests that some patients and settings may not benefit and may even experience negative consequences.

The low adoption of electronic medical records among healthcare providers and hospitals was explored in a series in Chest Journal by Dr. William F. Bria, II, M.D., FCCP, assistant professor of Internal Medicine and Co-Director of the Critical Care Medicine Unit and Medical Director of Clinical Information Systems at the University of Michigan. There have been numerous examples of failed clinical information systems over the past 30 years, he wrote, that relate in part to the poor understanding and oversimplification of the workflow in clinical care and the interactions necessary among care givers. Systems that might work well for some members of a team and for some providers, have been very poor at being safe and effective for others, and there are often competing agendas behind the implementation of HIT.

The most recent paper examining the unintended consequences of various HIT systems, such as EMRs and CPOE, on patient care was a sociotechnical analysis by authors with the federal Agency for Healthcare Research and Quality at Rockville, Maryland. It was published last fall in the Journal of the American Medical Informatics Association. “Unfortunately, there have been disturbingly mixed reports on HIT’s implementation and outcomes,” professor Michael I. Harrison, Ph.D., and colleagues wrote. “A growing body of research and user reports reveals many unanticipated and undesired consequences of implementation.” These “unintended consequences often undermine patient safety practices and occasionally harm patients,” the said.

Unfortunately, many healthcare professionals and even some IT specialists, are unfamiliar with this literature or its real life implications, they said. The same can be said for the general public and uninformed politicians who are promoting or mandating their adoption. “[T]hose who assume that computerization routinely enhances reliability may underestimate the safety contributions of clinical judgment, unmediated (direct) communication, and teamwork,” they added.

For their analysis, they drew upon key features of interactive sociotechnical analyses. First and foremost was the importance of examining actual uses of HIT, along with the impact of technical and physical work settings, users’ negotiation and interpretation of HIT, and its integration into clinical settings. HIT can’t be fully determined by the technology alone, nor can its success be viewed as depending primarily on training and technical support, they said.

“HIT implementation can alter or disrupt oral communication among clinicians, even when talk is faster, more clinically accurate, and safer than transmitting information through HIT,” they found. Not surprisingly, research has shown that continued use of paper after HIT is widespread among five institutions studied, with staff even printing out orders and manually reentering them into their department’s computers. “Poor interfaces between HIT and the physical settings in which it is deployed may also lead to HIT uses and workarounds which harm safety, quality, or efficiency,” they found.

As users try to make HIT work for them, it often leads to different uses and practices from those intended by HIT’s designers, they noted. For example, when nurses under heavy work loads encounter cumbersome software requiring multiple screens for medication administration, or cannot gain immediate access to a computer during a hectic shift or sit quietly without continual interruptions, they will often delay charting until the end of their shifts, which can create inaccurate recordings, reduce effectiveness and safety of communication between caregivers and eliminate the safety checks purportedly designed into the system.

Among the other unintended consequences the AHRQ authors identified were:

More/new work for clinicians — doctors and care providers having to spend more time on documentation and justification

Negative changes in communication patterns — decline of vital interaction among care providers and support service, and elimination of informal interactions and redundant checks that help catch errors

Workflow barriers — undermining of informal gatekeeping, for example by a clerk who catches whether patients really needed daily x-rays

New types of errors — busy clinicians enter data in miscellaneous sections, rather than scroll through screens for the ‘optimal’ locations, which can impede its use by others; fragmentation of information over multiple screens can lead busy care providers to miss key information; extensive reporting requirements lead caregivers to cut and paste whole reports, rather than extracting pertinent facts; inflexibility of documentation requirements lead to failures to record clinically meaningful information and cause difficulties in managing patient transfers; caregivers not following the system’s pre-authorization requirements for urgent care, and doctors and nurses working around cumbersome procedures; alert fatigue as doctors ignore the continual warnings and reminders; urgent orders are missed; loss of feedback from nurses that orders have been received and carried out; charting on wrong person

Compromised physician autonomy — remote monitoring by organizations undermines doctors’ autonomy; IT departments, ‘quality assurance departments’ and administrations and bureaucrats assume power over patient care by requiring doctors to comply with HIT clinical directives

Overdependence on technology — delivery of care becomes dependent on HIT and system failures wreak havoc when paper backup systems are eliminated; caregivers can become dependent on HIT to make clinical decisions and catch errors, resulting in reduced clinical judgment and trouble remembering dosages and medication contraindications; creates an illusion of communication among caregivers and a belief that information and orders that have been entered have been seen and acted on; secondary features added or reconfigured to make programs more usable for individual facilities quickly make systems unmanageable

As they concluded, “in the long run, some of HIT’s unintended consequences may play a role in helping HIT designers and healthcare professionals learn to deploy HIT more effectively.” However, “practitioners and medicine cannot tolerate developments that harm patients; HIT implementation, therefore, must proceed with caution.”

© 2008 Sandy Szwarc

** To Err is Human

The public misperceptions of the 1999 report on medical errors are that 44,000 to 98,000 people in U.S. hospitals are killed by medical errors every year and that these deaths can be prevented, largely through greater governmental oversight and enforcement of quality measures through improved health technology. The lay press and even Oprah have used the IOM report and these death figures to heighten concerns about the quality and safety of medical care in the U.S. and of dangers in our current system, said professor Kayhan Parsi, JD, Ph.D., assistant professor of bioethics and health policy at Neiswanger Institute for Bioethics and Health Policy of the Stritch School of Medicine Loyola University Chicago.

As he explained in the Virtual Mentor, the IOM report’s actual findings, however, have been widely misrepresented, fueling unsupported concerns among the public about the state of medical care and calls for government action. While the report itself itemizes the relevant literature, he said:

Unfortunately, not everyone who cites the report has read the entire document, and it is frequently misunderstood as a "study" that "demonstrated" the incidence of preventable deaths attributable to medical errors. Instead of being a study, the IOM Report is actually a policy document that discusses the scope of medical errors and makes recommendations to improve patient safety.

Only two studies cited in the report had even addressed medical errors and patient mortality, and both had limitations that must be understood in order to report and apply this information responsibly and credibly, he said. It is important to realize the types of “studies” these were, the definitions they used, the way the data was collected, the population examined and how the results were analyzed.

Both studies had used outdated data, the first used data from New York gathered in 1984 and the second used data from 1992 in Utah and Colorado. And both reports found similar rates of adverse events, but used different figures to estimate how many deaths were attributed to adverse events.

The New York study, known as the Harvard Medical Practice Study, reviewed 30,121 randomly selected charts for adverse events and extrapolated from these to estimate that the number of adverse events statewide was 3.7%, of which 1% was attributed to negligence. From this, he explained, they took the number of hospital discharges statewide and estimated that 13,451 patients died “at least in part as a result of adverse events,” based on the assumption that 13.6% of all adverse events led to death. The Utah and Colorado study reviewed the charts of about 2.5% of random discharges and found adverse events had occurred during 2.9% of hospital stays and this study estimated overall deaths by assuming 6.6% of adverse events led to death.

The IOM report then used these studies “and extrapolated this to the total number of U.S. hospital admissions in 1997” to estimate the total number of people who die each year due to medical errors, said professor Parsi. But while these figures are frequently cited in both medical and lay literature, he said, they have been denounced. For example, researchers from the Regenstrief Institute at Indiana University said the deaths were exaggerations, writing in a 2000 issue of the Journal of the American Medical Association:

Both were observational studies and were not designed to describe causal relationships. The Harvard study authors included caveats, such as "lead [sic] to death" and "died at least in part as a result of adverse event." The authors of the Colorado-Utah study reported a proportion of patients who died in the adverse reaction group, but said nothing about the cause of these deaths. The IOM did not mention any of these limitations in its report.

In other words, associations have been turned into causations. Dr. Troyen Brennan, M.D., one of the investigators in the New York study had made the point even more emphatically when he wrote in the April 13, 2000 issue of the New England Journal of Medicine that using the term “error” creates an impression that is not warranted by the science. The definition of error in science and engineering, as well as popular discourse as per “the Merriam-Webster Thesaurus, [means] blooper, blunder, boner, bungle, goof, lapse, miscue, misstep, mistake, and slip-up,” he wrote.

In both studies, the chart reviewers classified each event as whether they believed them preventable or not preventable. But “preventability is difficult to determine because it is often influenced by decisions about expenditures,” he said. For example, many drug reactions to antibiotics could be prevented if every patient underwent allergy testing before being given an antibiotic, but in reality the expense of such testing isn’t cost effective, nor is it practical in emergency situations, so they can’t accurately be classified as preventable. And decisions in both studies about whether an event was preventable “do not necessarily reflect the views of the average physician and certainly do not mean that all preventable adverse events were blunders,” he wrote:

For instance, surgeons know that postoperative hemorrhage occurs in a certain number of cases, but… even with the best surgical technique and proper precautions, however, a hemorrhage can occur. We classified most postoperative hemorrhages resulting in the transfer of patients back to the operating room after simple procedures (such as hysterectomy or appendectomy) as preventable, even though in most cases there was no apparent blunder or slip-up by the surgeon. The IOM report refers to these cases as medical errors, which to some observers may seem inappropriate.

Perhaps more to the point, neither study cited by the IOM as the source of data on the incidence of injuries due to medical care involved judgments by the physicians reviewing medical records about whether the injuries were caused by errors. Indeed, there is no evidence that such judgments can be made reliably.

These concerns aren’t just hairsplitting over definitions, he said, because the IOM report and the media give the impressions that doctors and hospitals are doing very little about injuries resulting from medical error, even though the data actually support a different conclusion. Extrapolating from our studies to calculate the number of deaths nationwide due to substandard care in 1992, he wrote, the number drops to about one-fourth that cited in the IOM report. In other words, based on more clinically realistic definitions of medical errors, the IOM report had overestimated medical errors nearly four times. And “the evidence suggests that safety has improved, not deteriorated” in the fifteen years since that report, he said, as surgical procedures, safety measures and technological advances have reduced mortalities.

“It is inaccurate to suggest that safety has been overlooked” by the medical profession, he wrote, nor have iatrogenic injuries been ignored. No one can disagree with wanting to work to make hospitals safer, but “most injuries from medical care are not due to mistakes,” which is the popular public perception. This is why federally-mandated public disclosures of medical adverse events and the increasing hostile responses to preserving patient confidentiality could lead to heightened distrust of medicine, spawn lawsuits and stifle healthcare professionals in coming forward to report and address problems, and hurt more patients in the end. Replying to letters to the editor, Dr. Brennan added:

I do believe that the use of the term "errors," although helpful when used to refer to a corpus of work from the engineering literature, does have perjorative [sic] overtones that the press has amplified. Berwick and I have written about the corrosive effect of heavy-handed regulation on attempts to improve quality. I have feared that the conflation of the meaning of "error" with the meaning of "blunder," combined with the IOM's call for mandatory, public reporting, would engender just that sort of regulation.

Susan Dentzer, health care correspondent for The Jim Lehrer Newshour, is one journalist who has criticized the media for its misleading scaremongering about the IOM report, as well as inaccurately equating errors with acts of medical malpractice. However, wrote professor Parzi, “the IOM Report itself contributed to this number craze with the following assertion in its executive summary: ‘More people die in a given year as a result of medical errors than from motor vehicle accidents (43,458), breast cancer (42,297), or AIDS (16,516).’ This type of comparison with stark numbers obviously makes good copy for most print journalists.”

However, journalists highlighting the misuse of statistics are published in obscure venues that don’t reach a mass audience, he lamented. Or, refused even publication at all, as this author has found.

Of additional note, while calls for HIT are foremost among solutions being proposed by stakeholders and public officials, the IOM report actually revealed medical errors that most of which would never be addressed by IT solutions.

This was illustrated in an AHRQ report looking at physical design of hospitals, for example. Patient falls are common in hospitals and can result in serious injuries and prolonged hospital stays; yet they can be reduced by changes in the physical design of hospitals, such as handrails, door openings and bathrooms. Hospital-acquired infections can be reduced with single-bed rooms, improved air filtration systems, and providing multiple locations for staff to wash their hands. Medication errors have been shown to be related to poor lighting, frequent interruptions and distractions while filling prescriptions which can be reduced with quiet, well-lit, facilities for pharmacists to fill prescriptions. Staff turnover and staffing shortages are serious problems that can lead to errors, with numerous measures possible to address burnout, physical fatigue and injuries and increase retention.

Addressing medication errors isn’t simply about writing prescriptions that meet a computerized protocol, as Boston authors wrote in the British Medical Journal. “Simple measures of known effectiveness, such as unit dosing, marking the correct side before surgery on paired organs, and 24 hour availability of pharmacists and emergency physicians, are often ignored,” they wrote. “[S]afe performance cannot be expected from workers who are sleep deprived, who work double or triple shifts, or whose job designs involve multiple competing urgent priorities.”

Dr. Donald H. Lambert, Ph.D., M.D., suggested multiple sources of medical errors in anesthesia, such as drug labels that look similar and how color coded labels or different shape vials for similar medications could reduce errors; and the wasteful and time-consuming practice of daily diluting dangerous vasoactive drugs where dilution errors have caused deaths that could be reduced simply by using prepackaged syringes.

The point being, that HIT won’t solve many of the most common medical errors, nor has it been demonstrated to improve health outcomes. In fact, the body of evidence suggests it can even cost lives. The evidence gives very different reflections to today’s announcement of plans for a new Office of Health Reform that will be responsible for the new Administration's “very aggressive initiatives around things like health IT and prevention.”

We'll examine financial costs and legal issues that have been raised next.

Y. Y. Han (2005). Unexpected Increased Mortality After Implementation of a Commercially Sold Computerized Physician Order Entry System PEDIATRICS, 116 (6), 1506-1512 DOI: 10.1542/peds.2005-1287