Chapter 21. The Role of Patients

With increasing public attention to patient safety have come calls for more participation by patients and their advocates in the search for solutions. Some of these calls have focused on individual healthcare organizations, such as efforts to include patients on hospital safety committees.1 Most, however, have involved enlisting patients in efforts to improve their own individual safety—often framed as a version of the question: “What can patients do to protect themselves?” This chapter will explore some of the opportunities and challenges surrounding patient engagement in their own safety, including errors caused by patients themselves. The issues surrounding disclosure of errors to patients are covered in Chapter 18.

A previously healthy 10-month-old girl was taken to a pediatrician’s office by her monolingual Spanish-speaking parents, worried about their daughter’s generalized weakness. The infant was diagnosed with iron-deficiency anemia. At the time of the clinic visit, no Spanish-speaking staff or interpreters were available. One of the nurses spoke broken Spanish and in general terms was able to explain that the girl had “low blood” and needed to take a medication. The parents nodded in understanding. The pediatrician wrote the following prescription in English:

Fer-Gen-Sol iron, 15 mg per 0.6 mL, 1.2 mL daily (3.5 mg/kg)

The parents took the prescription to the pharmacy. The local pharmacy did not have a Spanish-speaking pharmacist on staff, nor did they obtain an interpreter. The pharmacist attempted to demonstrate proper dosing and administration using the medication dropper and the parents nodded their understanding. The prescription label on the bottle was written in English.

The parents administered the medication at home and, within 15 minutes, the baby vomited twice and appeared ill. They took her to the nearest emergency department, where the serum iron level one hour after ingestion was found to be 365 mcg/dL, twice the upper therapeutic limit. She was admitted to the hospital for intravenous hydration and observation. On questioning, the parents stated that they had given their child a tablespoon of the medication, a 12.5-fold overdose. Luckily, the baby rapidly improved and was discharged the next day.2

Any discussion of patient engagement needs to start from square one—do patients understand their care and the benefits and risks of various diagnostic and therapeutic strategies (i.e., informed consent)? If patients cannot understand the basics of their clinical care, it seems unlikely that they can advocate for themselves when it comes to safety.

Unfortunately, many patients are in no position to understand even the basics of their care, let alone serve as bulwarks against errors. First of all, nearly 60 million Americans (20% of the population) speak a primary language other than English at home, and 25 million have limited English proficiency.3 Both of these populations nearly doubled in the United States between 1980 and 2009 (Figure 21-1). Few hospitals have adequate translation services4—translation frequently takes place on an ad hoc basis, often by untrained clerical personnel or even family members.5 In one case, an 11-year-old sibling interpreter made 58 interpretation errors, 84% of which had potential clinical consequences.6

Language barriers increase the chances of adverse events.7 One UCSF study found that non-English-speaking patients had a 30% higher odds of 30-day readmission to the hospital than did English speakers.8 Other studies have documented poorer adherence to treatment and follow-up, decreased satisfaction with care, decreased comprehension of diagnosis and treatment plans, and more adverse drug events.9–12 The use of professional interpreters is associated with fewer communication errors, better patient comprehension, higher patient satisfaction, and improved clinical outcomes.13

Although state and federal legislation requires that adequate interpreting services be made available to limited English proficiency patients, the costs and scarcity of trained interpreters have made it difficult to ensure universal access to these services.14 The increasingly widespread use of new communication technologies, including telephonic (Figure 21-2) and video interpreter services and even Web-based voice recognition and translation services, represents a hopeful trend.15,16

Even when patients do speak English well, many are at risk because of low “health literacy,” which is defined as “the degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions.”17 It includes the skills that patients need to communicate with providers, read medical information, make decisions about treatments, carry out care regimens, and decide when and how to seek help.

In 2003, approximately 80 million U.S. adults (36%) had limited health literacy18 (Figure 21-3); the problems are greater in patients with lower income, less education, and lower English language proficiency. Low health literacy is associated with poor communication between patients and clinicians and worse health outcomes.18–20 Medication errors are a particular problem: studies have shown that patients with poor health literacy frequently misunderstand common dosing regimens (“Take one tablet by mouth twice daily”) and auxiliary warnings (e.g., Do not chew or crush; For external use only).21

A number of strategies have been employed to mitigate the effects of low health literacy on safety. Many of them have focused on identifying patients with low literacy,22,23 and then providing them with simplified health materials (such as brochures and medication labels), Web sites, and interactive videos.24 Newer interventions focus on training providers to interact with low health literacy patients in appropriate ways. For example, the use of the “teach back” (patients are asked to restate to the provider their understanding of their condition or treatment plan) can help ensure that patients truly comprehend their situation (Figure 21-4).25,26 An alternative strategy is the “Ask Me 3,” which prompts patients to ask their providers three questions: What is my main problem? What do I need to do? Why is it important for me to do this?27 Patients who appear not to understand their care plan receive additional help.

In addition to these targeted interventions at the point of care, it will be important to embed education about health literacy into the training of physicians, nurses, and pharmacists. For example, a number of the newer competencies for residents involve issues of doctor–patient communication and improving physicians’ sensitivity to patients’ varying communication styles and needs. Because so many of the problems surrounding health literacy revolve around medications,18,20,24,28 the involvement of trained pharmacists can be crucial.29

Even when low health literacy or language differences are not at issue, patients may contribute to errors in their care. The causes are usually similar to those leading to provider-related errors: generally competent humans interacting with a staggeringly complex system, sometime engaging in unsafe acts or violating safety rules. One group of New Zealand–based researchers divided patient errors into two groups: action errors (errors of patient behavior such as failing to attend an appointment or taking a medication with alcohol despite being advised not to do so) and mental errors (errors that involve patients’ thought processes, such as “assuming I must OK because I’m feeling good”).30 One study conducted within a large health maintenance organization found that nearly one in four adverse drug events (Chapter 4) was caused by patient error, such as the failure to take the medication at the proper dose or follow medication use instructions.31

The solutions to patient-generated errors are likely similar to those that address errors caused by clinicians: better communication and teamwork, more robust information systems, identification of at-risk individuals (such as those on multiple medications, and those with low health literacy and language differences) to trigger tailored interventions, and learning from past mistakes and experiences. The recognition that some errors are caused by patients is not an occasion to assign blame, since in the vast majority of circumstances the errors simply prove that the patients are humans who, as we know, err. The task for clinicians and healthcare systems will be to better understand these human frailties and design systems and approaches that make them less likely to cause harm.

Obviously, patients who do not speak English (or whatever the dominant language is) or who have poor health literacy are in no position to advocate for their own safety, and are limited in their ability to participate meaningfully in clinical decision making and informed consent. This is a problem that needs to be tackled on its own merits.

But what about the patient who is competent and engaged, or who brings a family member to the clinical encounter able and willing to play an advocacy role? I have listed a number of questions that I would ask of my doctor or healthcare system in an effort to keep myself or a family member safe in Appendix IX. Responding to the public demand for answers to the “What can I do to stay safe?” question, several advocacy groups and professional organizations have launched campaigns— such as the Joint Commission’s “Speak Up” campaign) (Table 21-1)32—to address this issue.

However, as any healthcare provider who has been ill or has had a sick family member can attest, there are limits to what a layperson can do to protect against medical mistakes. These limitations were illustrated most vividly by Dr. Don Berwick, the founder of the Institute for Healthcare Improvement and former director of the U.S. Centers for Medicare & Medicaid Services. Berwick’s poignant tale of his wife’s series of hospitalizations (at several Harvard hospitals) for an obscure neurologic disease demonstrates the limitations of depending on patients or their families—even those with highly informed and engaged family members—to prevent medical errors:

The errors were not rare; they were the norm. During one admission, the neurologist told us in the morning, “By no means should you be getting anticholinergic agents [a medication that can cause neurological and muscle changes],” and a medication with profound anticholinergic side effects was given that afternoon. The attending neurologist in another admission told us by phone that a crucial and potentially toxic drug should be started immediately. He said, “Time is of the essence.” That was on Thursday morning at 10:00 am. The first dose was given 60 hours later—Saturday night at 10:00 pm. Nothing I could do, nothing I did, nothing I could think of made any difference. It nearly drove me mad. Colace [a stool softener] was discontinued by a physician’s order on Day 1, and was nonetheless brought by the nurse every single evening throughout a 14-day admission. Ann was supposed to receive five intravenous doses of a very toxic chemotherapy agent, but dose #3 was labeled as “dose #2.” For half a day, no record could be found that dose #2 had ever been given, even though I had watched it drip in myself. I tell you from my personal observation, no day passed—not one—without a medication error.33

Before addressing the issue of patient engagement as a safety strategy, let’s begin with core principles. The ethical virtues of an informed and participatory patient or family are unassailable, and the importance of patient engagement has been emphasized by groups ranging from the Joint Commission to the Institute of Medicine, often under the broad umbrella of patient-centered care. And, shifting to a patient safety perspective, an extra set of eyes primed to notice hazards and empowered to point them out probably will prevent medical errors from time to time. Asking for proof that patient/family engagement works might seem like seeking proof of the value of parachutes, the subject of a BMJ satirical shiv aimed at the heart of evidence-based medicine zealots several years ago.34

Parachutes notwithstanding, I have several reservations about the premise that patients and families can and should be central players in efforts to prevent medical mistakes.

The first is that it might not work, or might even hurt. In the early days of “sign your site,” before the process was standardized (Chapter 5), patients sometimes tried to be helpful by marking their own limbs before they dozed off from their anesthetic. Unfortunately, some patients marked the limb to be operated on, while others marked the limb to be avoided. As you might imagine, this kind of anarchy introduces new error possibilities.

Second, even if a well-informed, alert patient or loved one can help prevent some errors, too many patients don’t have the resources to act as advocates for their own safety. Particularly in the hospital, many patients lack involved family members, are confused, anxious, or sedated, or—as we’ve seen—have language differences or problems with health literacy. All of this means that relying on a strategy of patient engagement introduces an element of uncertainty. As British safety engineer Melinda Lyons observed:

… in safety engineering terms, … patients are unlikely to provide a consistent and reliable contribution to the safety of the process of their own care. In a domain with a safety problem that is moving towards advocating the “systems approach,” it seems nonsensical to also advocate a solution with apparently decreased reliability.35

Third, some patients and families, focused on doing what they can to prevent errors, will cross the thin line that separates being empowered and appropriately skeptical from being bellicose and confrontational. The latter attitude may lead providers to adopt a defensive stance or even generate outright avoidance (I’ve seen it happen on the wards). Such a situation is not likely to promote safety.

Finally, we need to consider the sobering observation that patients and families often feel guilty about medical errors, which means that an emphasis on patients protecting themselves might amplify this guilt.36 As a family member of a sickle cell patient who became critically ill after a medication error recalled:

The feeling was impotence, because you can’t stay with a patient 24 hours a day. That’s why you rely on hospitals—you rely on nurses. You feel like you failed your family in terms of ‘I should have been there.’ That’s a guilt that everyone shares.36

Even with these caveats, certain patient engagement strategies may be helpful. One study showed that patients were willing to help monitor their providers’ compliance with hand hygiene, were able to do so accurately, and did not feel that this would compromise the relationship with their caregivers.37 (I should quickly add that another survey of Pennsylvania adults found that they were quite reluctant to challenge their providers.38) Researchers also found that patients often observe (and can report) errors that are missed through other ascertainment methods (Chapter 14).39

Turning our attention to the policy sphere, the involvement of patients and their families has been a major force for change, as has been vividly illustrated by those patients and families that have channeled their sorrow over a devastating medical error into helping the system improve. For example, Sorrel King’s grief over her daughter’s death at Johns Hopkins Hospital was a major impetus for Hopkins’ seminal contributions to the safety field.40–42 And patient engagement at the organizational level, such as by having patients serve on key committees, seems to help keep systems focused and honest.1

As I’ve reflected over the years on my own ambivalence regarding the role of patients in protecting themselves from mistakes, I’ve discovered that it flows partly from my lingering doubts about whether this strategy really works. But I’ve also come to realize that it stems from a more fundamental rejection of the premise itself: why should it fall to patients and family members to ensure their own safety?43 When I board an airplane, I know that there is nothing I can do to keep myself safe. This allows me to relax (mostly), because my feelings of impotence are outweighed by my trust that the airlines and their regulators have done all they can to ensure my safety.

Patients encountering the healthcare system lack this trust, which makes relaxation and passivity seem maladaptive, sometimes even near-suicidal. That patients and families would want to do whatever they can to improve their odds of emerging unscathed is completely understandable. I’m just not sure how well this strategy works or how to apply it most effectively. Clearly, this should be the subject of ongoing research.

In the meantime, while we should support efforts by patients and families to participate in their own safety when feasible, our primary focus should be on making such hypervigilance unnecessary. In my judgment, this problem is ours to fix, not theirs.

• Many patients have limited language proficiency and/or health literacy, which increases the risk that they will be victims of medical errors.
• Patients can cause errors in their own care; the solutions are likely to parallel those used for other kinds of errors.
• Patients and families can help participate in their own safety, but (a) there are limitations to this strategy’s effectiveness and (b) the responsibility to provide safe care should primarily be borne by providers, healthcare organizations, and policymakers, not patients.