The Erring Report

Additional Statement, Submitted by Seymour M. Gluck, MD, FACP

"To Err is Human" is an appropriate title for the National Academy of Sciences' recent report on medical errors and while the title clearly acknowledges the human tendency to make mistakes, the report makes a few mistakes of its own, especially when it downplays the responsibility of healthcare personnel for medical errors by shifting the blame to the "systems" under which they happen to work. (Berwick DM. Testimony before the Health, Education, Labor and Pensions Committee; U.S.Senate 2/16/2000.).

Shifting from personal responsibility to "system" responsibility is not unlike absolving people from murder, drug abuse, and marital infidelity because our economic, social, and penal systems are less than perfect. Of course hospital systems need improvement but suggesting that "perhaps 95-98%" of medical errors result from "system" failures while declaring that human failings "like carelessness or incompetence in individual doctors. nurses, and other workers" are so infrequent and inconsequential that they only account for 2 to 5% of medical mistakes distorts reality. Then to conclude that "if we simply fired every health care worker who was involved in errors, and submitted a new person, our future error rates would not change at all" is simply incredulous, and no data are offered to support such a conclusion.

One might as well suggest that if we took every murderer, drug dealer, and repeat sex offender off the streets crime rates would stay the same. The statement also implies that dismissing incompetent personnel would have no effect on the error rate and that hospitals can never upgrade their staffs because replacements will be just as incompetent. Even if this were true, some thought should be given to the positive message a justified dismissal sends to the remaining staff and its likely effect in motivating better performance.

The number of medical errors is grossly understated. The report estimates that medical errors cause between 44,000 and 98,000 hospital deaths each year. The range itself suggests that the data (such as it may be) is extremely soft. Every intern knows that "minor" errors are often "conveniently" overlooked and unrecorded. Even in situations where more serious mishaps occur more "acceptable" explanations are often found.

Whether the errors are those of omission or commission and whether they are noted or not, they don't cease to exist simply because they're ignored. The IOM figures appear to be based on studies suggesting that "adverse events occur in approximately 3-4% of patients" (Brennan et al. NEJM 1991;324:370-6.) although a study by Andrews et.al (Lancet1997;349:309-13) provides a figure of 17.7%. Even if we accept the 3-4% figure, an estimate of 44,000-98,000 deaths from medical errors occurring in U.S. hospitals annually (even if true), pales when viewed against a backdrop of "adverse events" from medical errors that might result during millions of patient encounters in thousands of emergency rooms, nursing homes, mental hospitals, outpatient clinics, ambulatory care centers, clinical laboratories, medical offices, and home-care settings every day.

Better hospital funding would avoid many of the frustrations that set the stage for medical errors such as overwork, under staffing, and fatigue. The applications of new technologies lessen the opportunities for error while improved operational systems assure better monitoring and liberate highly trained personnel from needlessly time-consuming secretarial and administrative tasks.

As the report points out, medication errors have been substantially reduced by patient profiling and electronic prescribing. Enormous data banks instantly retrieve vital clinical information and provide important algorithms for diagnosis and treatment. The application of advanced radiographic techniques, ultrasound, laser technology, and robotics continue to lessen opportunities for human error even though machines can also make mistakes (and sometimes on a very grand scale).

Nevertheless, in emphasizing the very positive and promising features of applied technology the report neglects to consider any downside. The more we succeed in substituting mechanisms and machines for the personal attention of "hands on" care, the faster the art of medicine will become cold, mechanistic, and dehumanized. Lack of time and personal attention represents one of the most significant complaints of patients as well as doctors even today. Our goal should not be to lessen the "human touch", but to enhance it with improved standards of care and increased attention to safety and efficiency in meeting patient needs. We need to seek the proper balance between people and technology. There is no "free lunch" and no quick technological "fix" for every problem.

Considering the realities of the present and the promises of the future, we should realize that it takes competent people to make a good system work and that in any extended chain of command it need only take one weak link to upset the entire operation through carelessness, indifference, incompetence or worse. No system can be stronger than its weakest link and as the links usually stretch from patients primary care are physicians to specialists, hospitalists, physician extenders, assistants, nurse practitioners, ward nurses, technicians, therapists, aides, and supporting personnel, weak links are more likely to pop up.

Each link is part of the final common pathway through which most "systems" travel. Sometimes because of their numbers they are difficult to identify, although close monitoring, continuing surveillance, ongoing audits, positive peer pressure, and responsible patient feedback usually succeed. They should be reinforced when feasible and removed when judged beyond repair to prevent some unlucky patient from becoming another medical error.

While there is no doubt that the application of new technology and systems engineering hold great promise in lessening opportunities for medical errors, the ability to adequately fund many of the changes proposed in the report raises many difficult questions not the least of which is where the money is going to come from. Until better systems can be developed, existing systems will have to be maintained more effectively.

One of the biggest stumbling block in coping with medical errors is the striking divergence between good theory and bad practice. Hospital service manuals are usually filled with good theory that often serves more as window dressing to show "paper compliance" with required procedures. The day to day realities suggest that acceptable protocols are much easier to find in service manuals and reference texts than at the patient's bedside.

Sometimes the distinction between simple errors and outright negligence is hard to make. For example, debilitated patients may be deprived of essential nourishment if they are served food that is cold, tasteless, or difficult to swallow. Assisted feedings and medications may result in choking or aspiration if improperly offered to reclining patients. Failure to check the position of stomach tubes prior to liquefied feedings can have the same effect. Lapses in handwashing and strict adherence to aseptic protocol during dressing changes, wound care; tracheal suction, placement of central lines, and dangling urinary catheters expose patients to serious infections. What often appear to be minor errors of omission often become highly magnified in their effects. Localized pressure areas left unattended, often because they are not sought out, can rapidly develop into life-threatening decubitus ulcers. Falls from bed because of improperly positioned side rails are a common cause of fractures and concussions frequently requiring major surgery.

Inasmuch as venipuncture in its various applications is a prime source of life-threatening bloodstream infections and represents the most common invasive procedure in medical practice, it serves as an important paradigm for examining the role of healthcare personnel in complying with aseptic guidelines. The most basic aseptic principles for venipuncture preparatory to the routine placement of peripheral I.V. lines include the following:

1. Providers must wash their hands for approximately 10 seconds before donning disposable gloves.
2. An aseptic field should be established by cleansing and washing the targeted area for at least 30 seconds.
3. After the area is prepared, it must not be touched to repalpate the vein. Should this become necessary the entire aseptic prep should be repeated.
4. Sterile venipuncture equipment must be used.
5. A sterile dressing is applied at the completion of the procedure.

In March 1992 the Occupational Safety and Health Administration (OSHA) issued new standards and final regulations for a program of universal precautions to protect health care workers from exposure to pathogenic microorganisms in blood and body fluids. The program was reinforced by the threat of sizable fines for non-compliance. Although the Centers for Disease Control (CDC) had published its own guidelines for patients recommending a basic 30 second aseptic scrub for venipuncture ten years earlier and an updated version in 1995 (Pearson ML Guideline for Prevention of Intravascular-Device- Related Infections. Infection Control & Hospital Epidemiology. 1996;17:438-473) there were no provisions for enforcement and no attempt to establish specific standards for routine peripheral I.V.'s. As things stand today, hands are frequently not washed prior to starting I.V.'s,, 3-5 second dabs with an antiseptic sponge generally replace the 30 second scrub, and the prepared site is usually repalpated (and potentially re-contaminated) immediately prior to the needle insertion. Although such breaches of aseptic protocol are universally prohibited they are very much in vogue.

Every venipuncture opens an immediate, direct line of communication between a patient's bloodstream and the outside world where antibiotic-resistant bacteria, emerging viruses, opportunistic organisms, and hospital-based pathogens abound. Microorganisms of every variety are easily transported from patient to patient on the hands and clothing of personnel, tourniquets, stethoscopes and blood pressure cuffs. It is little wonder that 20-40% of hospital-acquired bloodstream infections cannot be traced to their original source.

When aseptic protocols are judiciously observed, the chances of infection are substantially reduced (Hughes JM. SENIC Project.Chemotherapy 1988;34:553-61). Unfortunately entrenched bad habits, the constraints of time, and the absence of enforceable standards have fostered innumerable short cuts for many procedures, generally more reflective of individual attitudes than recommended guidelines.

In the present era of consumerism an informed and involved public may prove to be the most reliable force for change. Notwithstanding the latest technology, patients must still rely on erring humans for essential care and they should not be forced to accept pot luck for services that may literally determine whether they live or die. Information regarding basic hospital procedures should be widely publicized and conspicuously displayed at the patient's bedside. Perhaps that will give patients and their families the courage to speak out, and maybe it will also remind healthcare providers to do the right thing.

Current as of September 2000

Internet Citation:

The Erring Report. Additional Statement, by Seymour M. Gluck. National Summit on Medical Errors and Patient Safety Research. September 2000. http://www.quic.gov/summit/agluck.htm

Return to Additional Statement Directory
National Summit on Medical Errors and Patient Safety
QuIC Home Page
Department of Health and Human Services