Cases & Commentaries
A "Weak" Response
Approach to Improving SafetySafety TargetResource TypeSetting of CareClinical AreaTarget AudienceError Types
A primary care physician on call for his group received a call at 9:00PM from a 68-year-old man. He said, "They started me on a new pill for my blood pressure and now I feel really weak." The physician asked how long ago the new medication was started. "Three days," the patient replied.
The patient could not recall the name of the drug, but he found the bottle of tablets, whose label read hydrochlorothiazide 25 mg. He stated that he had been taking one pill per day as instructed. The patient reported also taking lisinopril 20 mg daily for more than a year.
The physician, attributing the symptoms to the new medication, instructed the patient to stop the hydrochlorothiazide. He told the patient to use his home blood pressure cuff after the call and to come into the clinic right away if systolic pressure went above 180 mmHg. Otherwise, the patient was told to make an appointment to see his regular doctor to get a different medication for his blood pressure.
Three days later, the patient was hospitalized with sudden onset of right arm and leg weakness, as well as difficulty speaking. He was found to be in atrial fibrillation with a ventricular response of 120 beats per minute.
On reviewing the patient's symptoms and confirming with the patient's regular physician that the atrial fibrillation was new, the admitting physician judged that the patient became weak due to the new atrial fibrillation and rapid ventricular response rather than to potassium depletion, hyponatremia, or other effects of the hydrochlorothiazide. The findings on neuroimaging were strongly suggestive of an embolic stroke. The patient was begun on warfarin for atrial fibrillation and received rehabilitation while in hospital, but still had weakness and some word-finding difficulties 6 weeks later.
In this disturbing case, a patient suffered irreversible effects from an embolic stroke because a physician committed a series of errors during a telephone call.
First, the history was inadequate. The telephone's inherent challenges—no visual cues and no physical examination—make a thorough history essential. In this case, once the patient ascribed his weakness to his new blood pressure medication, the physician accepted that, cut short his history, and prematurely moved on to the plan.
When taking a history on the telephone, one should begin with an open-ended question. This generally provides much more information than does an immediate barrage of questions. In other words, one should pause for a moment after a patient says why he or she is calling, or prompt a reticent patient with a question like: "What do you mean by 'weak'?"
Another challenge of the telephone—especially in a patient with a perplexing symptom like weakness—is the lack of vital signs. Yet in some cases, vital signs are available to us, if we look for them. In this case, the physician failed to take the bait: although the patient mentioned a home blood pressure monitor (these are often equipped to measure pulse rate as well), the physician asked the patient to use it after the call. If he had asked the patient to check his readings during the call, perhaps the rapidity of the pulse rate would have illuminated the potential urgency and seriousness of the situation. Even in the absence of technology, many patients know how, or can be quickly taught, to take their pulse.
Second, the plan was unsound. After advising the patient to stop the hydrochlorothiazide and to follow blood pressure measurements at home, the physician provided a parameter only for supranormal readings. In this case, the patient's weakness was just as likely related to hypotension. In addition, the physician did not provide symptom-based parameters for a call back—ie, what the patient should do if the weakness persisted or increased, or what other symptoms should prompt a call. Nor was he specific about when to schedule an office visit if the blood pressure did not surpass the given parameter. There might have been a better outcome had the physician asked the patient to come in the next day if he still was weak, or called the patient back a few hours later to see how he felt.
The actual prevalence of such telephone errors, which can range from communication mishaps to serious errors resulting in increased morbidity, is largely unknown. Few articles in the patient safety literature focus specifically on telephone issues.(1,2) Communication mishaps can cause confusion about changes in medications, misunderstandings about plans, patient dissatisfaction, and at times (as in this case) serious morbidity. It is not surprising that such errors occur; only 6% of residency programs teach telephone medicine.(3) As the telephone management of some medical problems, such as depression, becomes more prevalent, it is vital that future studies identify the most common and serious types of errors so that evidence-based approaches to preventing them can be developed.(4)
What kinds of systems changes might improve telephone outcomes? Telephone symptom algorithms specifically designed for physicians should be made widely available to residents and attending physicians.(5-7) Developing guidelines based on the available literature could improve outcomes and reduce errors; for example, forwarding all clinical calls to physicians or trained nurses rather than allowing patients or answering service personnel to make triage decisions, and repeating telephoned laboratory results from pathologists to physicians.(1,2) Telephone medicine should be taught more broadly, especially in residency training.(8) In the outpatient setting, regular telephone case reviews could help trainees learn sound approaches to telephone issues as well as tips for avoiding common errors; in the inpatient setting, conferences on telephone communication could include key topics such as calling patients or family members with bad news. Physicians in practice might consider implementing a telephone case review as a regular part of staff conferences.
Many telephone mistakes occur as a result of inadequate data available to the covering physician. Easier availability of computerized records would likely cut down on the error rate. The increasing use of telemedicine has the potential to improve the quality of telephone medicine in an additive way. For example, a patient recently called me about a growth on his forehead. As I tried to get a visual picture from his description, he mentioned that he had a digital camera. Moments later, I had a high-quality image of what looked like a squamous cell carcinoma and set up a next-day visit for him with a dermatologist.
Overall, the purpose of telephone medicine is not to diagnose but to triage. The question should be whether the patient needs emergency evaluation, and if not, to determine how the patient might be treated at home or whether he or she should be seen in the office within a few days. It is important to remember that many patients call primarily for reassurance.(9) Calling a patient back after a few hours to check on the progression of a symptom can be reassuring to both patient and doctor. For the patient, it can be a sign of the physician's concern. For the physician, it can be a way to check the correctness of the initial triage decision and an opportunity, if necessary, to revise the plan.
- Remember that your primary role is to triage, rather than diagnose, via effective and detailed history-taking (identical to that in an office-based visit).
- Try to elicit hidden concerns. Patients who call in the middle of the night with seemingly minor complaints often have some unstated but more serious concerns. In fact, the proportion of serious issues increases the later it gets.(10)
- Construct a clear plan that the patient understands and can repeat back to you. Always provide specific symptom-based parameters for a call back and always offer an office visit. Remember to document the call and plan, and pass the documentation and plan on to the continuity physician.
- Although you may feel alone in making a decision on the phone, don't hesitate to rely on telephone symptom flowcharts, MEDLINE or the Internet, general medical textbooks, and colleagues for support.
- Calling the patient back in a few hours can be reassuring for both patient and physician.
- Increased access to computerized medical records and telemedicine will assist in the provision of high quality and safe telephone medicine in the future.
Anna B. Reisman, MD Assistant Professor of Medicine Yale University School of Medicine
1. Hildebrandt DE, Westfall JM, Smith PC. After-hours telephone triage affects patient safety. J Fam Pract. 2003;52:222-7.[ go to PubMed ]
2. Barenfanger J, Sautter RL, Lang DL, Collins SM, Hacek DM, Peterson LR. Improving patient safety by repeating (read-back) telephone reports of critical information. Am J Clin Pathol. 2004;121:801-3.[ go to PubMed ]
3. Flannery MT, Moses GA, Cykert S, et al. Telephone management training in internal medicine residencies: a national survey of program directors. Acad Med. 1995;70:1138-41.[ go to PubMed ]
4. Simon GE, Ludman EJ, Tutty S, Operskalski B, Von Korff M. Telephone psychotherapy and telephone care management for primary care patients starting antidepressant treatment: a randomized controlled trial. JAMA. 2004;292:935-42.[ go to PubMed ]
5. Reisman AB, Stevens DL. Telephone medicine: a guide for the practicing physician. Philadelphia, PA: American College of Physicians; 2002.
6. Katz HP. Telephone medicine: triage and training for primary care. 2nd ed. Philadelphia, PA: F.A. Davis Company; 2001
7. Thompson DA. Adult telephone protocols: office version. Elk Grove Village, IL: American Academy of Pediatrics; 2004
8. Hannis MD, Hazard RL, Rothschild M, Elnicki DM, Keyserling TC, DeVellis RF. Physician attitudes regarding telephone medicine. J Gen Intern Med. 1996;11:678-83.[ go to PubMed ]
9. Curtis P, Talbot A. The after-hours call in family practice. J Fam Pract. 1979;9:901-9.[ go to PubMed ]
10. Peters RM. After-hours telephone calls to general and subspecialty internists: an observational study. J Gen Intern Med. 1994;9:554-7.[ go to PubMed ]