Disabling the off switch
March 16, 2010 -- Recent reports in The Wall Street Journal and The Boston Globe discuss efforts to address potentially lethal hospital errors, with each piece using as its main example a case in which nurses did not spot a problem until it was too late and a patient died. It is very helpful for the public to hear that nursing (just like medicine) is so important to patient care that such problems can mean the difference between life and death. And we commend the journalists responsible for both of these pieces, which provide serious, thoughtful discussion of some important issues, including systemic factors beyond the nurses' control. But neither piece consults nurses to the extent it should, considering that the problems addressed are primarily nursing ones, and the effect is to undervalue nursing expertise and possibly to suggest that nurses report to physicians in providing the relevant care. Liz Kowalczyk's February 21 Globe report describes events surrounding the tragic death of a Massachusetts General Hospital (MGH) patient whose heart monitor alarm had been left off. The reporter includes a few helpful quotes from MGH's chief nurse, but none from national nurse experts or direct care nurses who deal with such monitors constantly, instead relying on physician safety experts and engineers involved in improving the safety of such technologies. And Laura Landro's piece in today's Journal discusses efforts to treat the health workers involved in errors fairly, focusing on the well-known case of Wisconsin nurse Julie Thao, who mistakenly gave a pregnant patient a fatal dose of a painkiller and actually faced criminal charges, but who has since worked to improve hospital safety. The story includes a little indirect commentary from Thao, but all the expert quotes are from physicians and other non-nurse safety experts, rather than the nurses who know more directly how and why such medication errors can occur. Perhaps as a result of inadequate input from nurses, neither piece mentions the extent to which nurse-related errors are due to inadequate staffing or other factors in the practice environment, which nursing scholars have shown remains a threat to U.S. patient safety. Nor does either piece discuss nurses' relatively low level of power, which discourages nurses from speaking up about problems, an issue on which Johns Hopkins physician Peter Pronovost has rightly focused in his efforts to improve safety.
The February 21 Globe article is "MGH death spurs review of patient monitors: Heart alarm was off; device issues spotlight a growing national problem." It reports that the patient died "after the alarm on a heart monitor was inadvertently left off, delaying the response of nurses and doctors to the patient's medical crisis." MGH "administrators" reportedly began investigating and quickly disabled every heart monitor alarm off switch in the hospital. The piece relies mainly on MGH representatives and outside "patient safety officials," who say
the tragedy at Mass. General shines a spotlight on a national problem with heart sensors and other ubiquitous patient monitoring devices. Numerous deaths have been reported because alarms malfunctioned or were turned off, ignored, or unheard.
From MGH, the report quotes "Dr. Gregg Meyer, senior vice president for quality and patient safety," as well as "Jeannette Ives Erickson (right), the hospital's chief nurse." Actually, Erickson is also MGH's "senior vice president for patient care." That's one of the amorphous titles that can mask the central role nurses play in hospitals because there is no explicit reference to nursing, so we would never suggest a reporter use such a title instead of "chief nurse." But including both titles here might have suggested that Erickson is on an equal footing with Meyer, and underlined how central nursing is to "patient care." And since the report devotes eight words to Meyer's title, it could have given Erickson nine.
Meyer (right) and Erickson are jointly quoted as saying that the "crisis" alarm on the MGH monitor was off, "but they are not sure why." They also reportedly noted that this alarm would normally "blare in the patient's room and at a central nursing station."
The piece does rely on Erickson for a few key details about what happened in this case. She suggests, in an indirect quote, that one possibility is that "someone turned off the switch during a previous patient crisis because they believed it would pause the alarm, not turn it off for good." She also reportedly said that MGH does not think someone turned off the alarm because it was "too noisy or annoying," as has apparently occurred elsewhere in the nation. And Erickson gets an entire long paragraph to describe the measures MGH took in the wake of the error:
Erickson said that during the hospital's investigation, administrators discovered that "alarms cannot always be heard . . . when a unit is very busy." So on Feb. 12, she put in place a plan to assign a nurse to every central nursing station to "ensure a timely response to alarms." She said this is a temporary solution until the hospital turns up the volume on all the alarms, installs new speakers, and assesses whether these changes solve the problem. The hospital also rolled out an educational program for nurses on working with the [General Electric] monitoring systems [involved in the error].
This is generally good, because it presents Erickson as a decisive executive who has knowledge of the monitoring systems and who seems to have played an important role in initiating and implementing a plan to avoid such problems in the future. It also suggests that nurses are primarily responsible for monitoring patients (and if you think the public already knows that, try watching an episode of popular hospital dramas like Fox's House or ABC's Grey's Anatomy.) And the note about the "educational program for nurses" also suggests that it is nurses who work with these systems, though we suppose readers might assume that physicians do too, but don't need the training. Of course, it's commendable that the reporter consulted nurses at all, as many would not have bothered, and we wonder to what extent this is due to the fact that MGH is one of the few hospitals to have a publicist who is dedicated to the hospital's nursing staff.
Unfortunately, some of the numerous quotes from Dr. Meyer probably do support the idea that physicians are more involved in these areas than they really are, and that physicians generally have expertise in matters that in reality are primarily about nursing. Meyer notes that patient confidentiality laws prevent the hospital from discussing details of the patient's death, but adds:
The death of this patient was tragic, and our hearts go out to this family and all the caregivers who were affected by this. Our priority is to find out what happened, why did it happen, and what can we do to make sure it never happens again.
Meyer also reportedly says that the hospital wants to "fix the underlying systemic issues with monitoring patients, which is why they disabled the alarms' off switches." Meyer notes that the hospital is not interested in assigning blame to individuals, because that would be "unfair and counterproductive," perhaps discouraging open reporting and discussion of the issues. The piece notes that the MGH president "praised staff for reporting the incident to hospital safety officials." This is mostly OK, because it's mainly at a general "patient safety" level.
However, Meyer is also quoted for details about the specific patient monitoring systems involved, which may wrongly suggest to many readers that physicians have significant involvement in such monitoring generally, and perhaps that physicians supervise nurses in their performance of that monitoring. It is Meyer who is first cited for the information that a GE monitor was involved in this error. In addition, Meyer is indirectly quoted for the following:
Screens in the room and at the nursing station display the patient's heart rhythm, but it's likely that in this case nurses were not in the patient's room because they were busy with other patients, Meyer said. The screens at the nursing station depict information for many patients, so one patient's abnormal heart rhythm might not jump out in the absence of an alarm, he said. A nurse discovered the patient was in crisis when she went to make a routine check, Meyer said. He would not say how long the patient was in trouble before the problem was discovered.
Of course physicians are broadly familiar with the existence of all of this monitoring, and presumably Meyer would know more than most as a result of his patient safety responsibilities and his involvement in this case. But why not consult Erickson, or a nationally recognized nurse expert, or any number of direct care nurses who actually take the lead in doing the monitoring? Wouldn't they be best to explain how these complex systems work and how such errors could occur? The text itself makes clear these are primarily nursing issues; where are the nurses?
The piece also relies heavily on several outside patient safety experts, none of whom appears to be a nurse. Once again, the issue is not that these people are not qualified to speak, since they certainly seem to be, but that the report fails to consult the countless nurse experts who would also be well-qualified to discuss broader issues implicated by hospital systems. No one is more involved than nurses are in operating such systems. Of course, nurses are not as well-represented as they should be in high-level policy-making and care systems analysis, even in areas in which nurses are more involved than anyone else, because of their generally underempowered status. That's why it's no surprise that Pronovost is seen as the leader of a movement toward checklists and other safety measures that are so central to nursing practice. So in fairness, this reporter doubtless encountered many non-nurse experts in researching her story.
In any case, the Globe report quotes two experts from the ECRI Institute, "a nonprofit research and consulting organization based in Pennsylvania that specializes in medical devices." Jim Keller (right), a vice president there, notes that alarms on patient monitoring devices are one of the top health technology hazards his group sees. He explains that workers can sometimes be confused by differences in brands of monitors; some don't allow alarms to be turned off entirely, only paused. The piece relies even more heavily on Kathryn Pelczarski, "director of ECRI's applied solutions group." She notes that a search of the U.S. Food and Drug Administration's database revealed 237 reports of "alarm-related deaths between 2002 and 2004," and that 12 percent of the "medical device" problems hospitals reported to ECRI between 2002 and 2006 involved alarms. Pelczarski says that "alarm fatigue" is "one of the most common problems she sees, where nurses and doctors are besieged with so many alarms that they lose their urgency." The piece explains that modern health technology means critical patients may be hooked up to many different machines with alarms, including heart monitors, ventilators, medication infusion pumps, and pulse oximeters to measure oxygen levels. Pelczarski notes:
There may be so many alarms going off it sort of becomes the background noise. We have seen situations where all the nurses are responsible for all alarms within that unit and there is the assumption that someone else will get that alarm. I frequently see alarms turned down to the point of being inaudible.
Obviously Pelczarski knows what she's talking about.
But wouldn't the nurses who actually do this work also have valuable information? Might some nurses have also noted that it can be hard to respond to all of the system warnings and other events in a nursing shift if nurses are understaffed? Also, what was the staffing like when the MGH incident occurred?
A nurse might also have noted that some alarms interfere with nurses' ability to hear other alarms. For example, on ventilators, some high airway pressure alarms are as loud and of the same tone as alarms indicating that the patient has been disconnected from the ventilator. So every time the patient coughs, an alarm sounds that might signal either of those events--one serious, one not so much. This can cause alarm fatigue, and soon, nurses might begin to ignore all the alarms. Why not make the disconnect alarm louder or more distinct?
The report quotes other outside experts. One is "Dr. Lucian Leape (right), a specialist on medical safety at the Harvard School of Public Health," who asks why manufacturers "would ever make a machine that allows hospital staff to turn off a critical alarm."
But if nurses had been consulted, they might have noted that there is actually a good reason to disable the alarms in certain situations--so that the alarms mean something when they do ring. For example, preventing the disabling of the alarms on ventilators while nurses suction a patient, or the alarms on heart monitors while nurses change the sticky pads on the chest, would overwhelm the unit with needless alarms. In these situations, nurses might develop severe alarm fatigue within a couple hours. In general, if nurses are there with the patient during an event and have the situation under control, the alarm should be temporarily disabled so that no one else's time or attention is taken away needlessly. Nurses are the best people to explain those issues to the media, as the failure of this article to explore the issues at all confirms.
The piece does consult GE Healthcare, but not surprisingly, its spokeswoman Jennifer Francis (right) is not exactly chatty, refusing to comment on whether other hospitals have had similar cardiac monitor problems, and offering standard corporate statements that the company is aware of the "tragic incident" and is now "partnering with MGH to investigate the incident. GE Healthcare is committed to the highest level of patient care and safety in the use of its products." Anyway, could it be that the manufacturers know that these alarms sometimes malfunction, going off constantly regardless of the patient's status, so they created a way to stop the meaningless noise until patients are squared away and ready to be left on their own--with alarms on? Could the nurse have mistakenly pushed the permanent alarm off button instead of the temporary alarm off button in this case? Are there any other reasons to turn the alarms off? Hey--how about if we ask the people who actually use the products all day every day? Naah.
The piece also consults George Mills, a senior engineer at the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), which accredits hospitals. Mills explains that early in the past decade JCAHO found what the reporter describes as "hospital staffers" who were "muzzling [alarms] with gauze and tape and otherwise blunting the noise." Mills says that JCAHO made an educational effort in 2005 to get workers to stop turning the alarms off and manufacturers to make it harder to do so, but that there has been "a resurgence of the problem in the last year or so." But why would the "staffers" go to such lengths to muzzle and turn the alarms off? Should we ask one?
On the whole, this piece pays more attention to nursing and relies more on nursing input than some others might have. But we have to wonder if the failure to really consult the nurses who use the technology in question every day stems from the familiar journalistic assumption that nurses are not real health experts, so it's necessary to consult non-nurses, even when it should be clear that the matter under discussion is more about nursing than anything else.
The Wall Street Journal article, part of the paper's "Informed Patient" feature, is headlined: "New Focus on Averting Errors: Hospital Culture." The piece's real subject is recent ideas about how to handle care givers who make errors, specifically the extent to which they should be punished and what steps might be taken to actually help them. The report focuses on a new study in JCAHO's Journal on Quality and Patient Safety, which examines Julie Thao's case closely.
The piece begins by reciting familiar statistics about the high number of errors made by "doctors, nurses and other medical caregivers," including preventable infections and medication errors, which together cause many thousands of deaths each year. But the piece says that hospitals are now "taking what might seem like a surprising approach" to the problem by not just trying to improve safety directly, but also "coming up with procedures for handling--and even consoling--staffers who make inadvertent mistakes." First it briefly cites two examples. The National Quality Forum, a "government-advisory body that sets voluntary safety standards for hospitals," has put forth a "Care of the Caregiver standard," under which hospitals treat the "traumatized staffers" involved in errors and involve them in the investigation of the error, if the error was not "reckless." And the article says that engineer David Marx's "Just Culture" model
stresses finding a middle ground between a blame-free culture, which attributes all errors to system failure and says no individual is held accountable, and overly punitive culture, where individuals are blamed for all mistakes.
But the main focus of the story is the new study, published in the April issue of the JCAHO journal and based on work led by researchers at the Institute for Safe Medication Practices. The piece says the study analyzes what went wrong in Thao's situation and "shows how assigning blame for errors can be a murky exercise." The article briefly explains the 2006 incident, in which Thao mistakenly gave a pregnant patient named Jasmine Gant an "epidural painkiller" by IV, thinking it was penicillin. The patient died, though her baby survived. Thao was reportedly fired by the hospital, and she was prosecuted for criminal negligence, a felony charge. She later pled guilty only to two misdemeanors, but she still had her nursing license suspended and was effectively barred from working for hospitals "for several years."
Then the piece devotes an admirable amount of space to what the new study found about Thao's case. Among the reported findings:
The study … concludes that while Ms. Thao consciously bypassed multiple safety procedures, there were also a host of system flaws that allowed and even encouraged her to do so, contributing to the fatal error. Researchers found that Ms. Thao failed to put an identification bracelet on her patient or use the hospital's bar-coding system, designed to match the right medication to the right patient. But the bar-coding system had glitches, and nurses hadn't been adequately trained on it, so they often bypassed it. Both medications--which looked alike--were brought into the patient's room before orders were given, a violation of policy. Fatigue increased Ms. Thao's likelihood of making a mistake, the study found. Ms Thao had worked two consecutive eight-hour shifts the day before and then slept in the hospital before coming on duty again the next morning, but there were no rules at the hospital to prevent her from being overworked.
These are all excellent points and represent a far more holistic analysis of the situation than was evident in many accounts in 2006. This analysis echoes concerns expressed by nursing advocates at the time Thao was charged; these advocates urged everyone to consider all of the factors that nurses confront in these situations. Of course, even if the error could be laid entirely at Thao's door, it's hard to imagine how that would justify a criminal prosecution.
The Wall Street Journal article says that editorials by patient safety experts that accompany the new study question how Thao was treated. The article focuses on an editorial by Charles Denham, "co-chairman of a National Quality Forum safe-practices committee," and Dr. Leape, the Harvard "health-policy professor" quoted in the Globe story above. These two are "harshly critical" of the way Thao was fired and left to face criminal charges without resources. Denham says that other nurses might have made the same mistake given conditions at the hospital, and that "Julie should be held accountable for her behavior, but she didn't receive support from her organization or treatment that was just." Denham isn't just talking; the piece reports that he hired Thao for two years as a "patient-safety fellow" on his research projects, and that she still does contract research for him. The piece also notes:
Ms. Thao, who was briefly hospitalized for depression after the event, says she considered taking her own life. She says her patient-safety work has helped her to cope with her despair over her errors.
The piece rightly gives the Wisconsin hospital, St. Mary's, a chance to respond. Noting that the hospital paid a $1.9 million malpractice settlement to Ms. Gant's family, the piece quotes hospital president Frank Byrne, who contributed his own commentary with the JCAHO study, describing safety measures the hospital implemented after Gant's death, including limiting how many hours nurses could work. Byrne says he knows that punishment alone does not improve safety, and that the hospital did not try to "shirk acknowledgement of [its] system issues," but that "we have to separate unavoidable error from reckless behavior and unjustifiable risk." He also claims to have been supportive of Thao. In this report's words, Byrne said that when he learned about the pending criminal charges, he "did everything he could to stop it and appeared at court proceedings to lend moral support." It's not clear if that moral support could be readily converted into the funds Thao needed to feed her family or pay her legal bills.
The piece weaves in comment from a couple other experts. Noting that safety advocates and "nursing groups" have questioned the use of criminal charges for errors, it quotes Sue Sheridan of Consumers Advancing Patient Safety. She calls charges "extreme" when used in "a system set up to fail," but also says there must be "accountability" when there's a tragic loss. David Marx, the engineer who developed the Just Culture Model, echoes these concerns, describing hospitals' struggle to "hold practitioners accountable for key safety behaviors," and to "address risky behaviors before they lead to the death of a patient." Marx's company, Outcome Engineering, consults with hospitals, states, and nursing boards. The company provides training, as it did for St. Mary's after the Thao case, on practices like hand-washing and doing safety checks when giving drugs.
Once again, this is all helpful information, and the description of the Thao case is generally very fair, giving readers plenty of reason to question how she was treated by the state of Wisconsin and by the hospital, and pointing out some of the deeper systemic issues that appeared to have played roles in Gant's tragic death. But even more than in the Globe article, nurses are missing from this account as sources of expert information. A couple indirect quotes from Thao about her personal experience won't do it. Where are the quotes from direct care nurses who actually provide the care under discussion here? Where are the "nursing groups" that the report itself says protested the use of criminal charges in the Thao case? If you talk extensively about nurses' conduct and how it might be changed, but never talk to the nurses themselves, what does that imply? At a minimum, it suggests that someone else is really in charge of all of this, and of course we all know who that would be: physicians. And frankly, ignoring nurses' views here also suggests that nurses have as much in common with the health equipment they use as they do with the "safety experts" who create and debate health policy; perhaps we should just disable the nurses' off switches. Let's ask some physicians and engineers if that would work!
A nursing expert might also have pointed out that it's not so simple as placing an emphasis on safety practices or better care systems, as important as those steps are. There are deeper issues this report does not raise. It may well be that fatigue from excessive hours was a factor in the Thao case. But it's not so easy as a hospital just choosing to cap the hours a nurse can work, when the hospital could easily reverse itself any time. And the effectiveness of a cap depends on what the cap is; is there any research about the effects of fatigue on nursing practice? (Yes.) Also, has anyone called for legislation to limit forced overtime? (Yes.) Is our society ready to allocate the resources needed for good care, or is it reluctant to do so because it undervalues nursing? Nurse short-staffing remains a critical problem around the nation, and this is likely to be a major factor in potentially deadly errors. But since nursing advocates and nursing scholars like Linda Aiken, Peter Buerhaus, and Penn's Ann Rogers are the ones who have focused on the safety and cost implications of nurses' practice environment, you'd probably have to actually consult them to hear about it.
In short, you can't be an "informed patient"--particularly when the subject is nursing errors--if you don't even talk to nurses and listen to what they have to say. We suppose the media's ongoing tendency to ignore nurses is "just culture," but it's also an unsafe health practice.
See the article "MGH death spurs review of patient monitors: Heart alarm was off; device issues spotlight a growing national problem" by Liz Kowalczyk, published on February 21, 2010 in the Boston Globe. Write to Liz Kowalczyk at firstname.lastname@example.org.