When a police officer threatened to arrest me last month while I was supervising a busy Philadelphia emergency department, I sputtered about the consequences of leaving the ER rudderless at 4 a.m. He wasn't impressed. After all, I could have simply complied with the officer's request: He'd asked me to draw a blood sample from my patient and turn it over to him for alcohol testing. The patient was a young woman brought to the ER after she crashed her car into a telephone pole at low speed. From the cop's point of view, the central point was that she was under arrest for suspected drunken driving. But from mine, what mattered was that she was unhurt. A physical and neurological exam showed that she did not need any tests: no blood work, no X-rays.
Drawing the patient's blood thus would serve only legal ends (which, given her slurred speech and tearful wails, seemed likely to benefit the prosecution). To step outside of my role as her caretaker seemed to violate fundamental tenets of our doctor-patient relationship. At the same time, I am an emergency physician trained in the field of public health and I know that alcohol is responsible for an enormous number of the tragedies that bring patients to the ER. I embrace efforts to reduce drunken driving, including helping to put the drivers behind bars. But the police wanted me to provide a tube of the patient's blood that I had no medical reason to obtain. A perfect storm of competing interests—those of bioethics, the law, and public health—had come crashing into my ER.
To be clear, emergency physicians routinely encounter such requests. Most of us have hospital or departmental policies that help guide us through the local laws that speak to these issues: Here's mine at the University of Pennsylvania. Previously when the police asked for blood samples, I took one of two approaches, both of which seemed to satisfy the cops, the hospital, the patient, and me. If the patient was injured enough to require a blood draw, I sent the samples to my hospital lab. The police could subpoena the results of a test for blood-alcohol levels, I told them. If the patient was uninjured and I saw no need for blood testing, I discharged the patient into police custody where, in Philadelphia, a nurse at the local jail draws the blood. Perhaps, on this particular night, the police were concerned that delaying the test would increase the chance of a result that was falsely negative, or maybe the jail nurse was out sick. In any case, my usual course of action did not fly.
In the end, the patient spared me arrest and defused the conflict: She decided that she wanted us to perform the test in the ER, so we did. Since then, I have learned that by refusing to draw the patient's blood, even for nonmedical purposes, I was breaking Pennsylvania law. The Pennsylvania Vehicle Code states that drivers may legally refuse blood-alcohol-level testing (at the risk of automatic suspension of their license). But physicians, nurses, technicians, and their hospitals may neither refuse to perform a requested test nor refuse to provide the results to the police except in "unusual circumstances." Other states differ. Some, in fact, prohibit physicians from being compelled to obtain blood samples for testing. Pennsylvania is unique in specifically prohibiting health-care professionals from "administratively refusing" to perform blood tests (though many states have laws requiring mandatory blood-testing and reporting of the results for drivers in fatal crashes). The Supreme Court confirmed the constitutionality of laws that permit police to obtain blood for testing, with the help of a doctor, in the 1966 case Schmerber v. California. The suspected drunken driver refused testing in Schmerber. But the court's ruling turned on the limits of the patient's right against self-incrimination. So, the high court has never considered whether medical ethics or the doctor-patient relationship should preclude a physician's involvement in testing that a patient doesn't need.
State laws that seek to expand blood-alcohol testing are driven by the desire to save lives. Drunken driving kills approximately 17,000 people a year in the United States and injures many more. Thirty-nine percent of all motor-vehicle fatalities are alcohol-related. In many areas of society we infringe on civil liberties to protect public health—think of mandatory vaccinations for public-school students, the fluoridation of municipal water supplies, and the reporting of epileptics to driver-registration bureaus. Ethicists have justified medical encroachments on personal liberties if they are reasonable, effective, as least invasive as possible, and when there are no other less invasive alternatives.
But there is no evidence to suggest that requiring doctors and hospitals to obtain blood-alcohol levels saves lives or even helps get drunken drivers off the road. Police officers with Breathalyzers or station-house phlebotomists are just as likely to obtain results that can be used in court without engendering a conflict with the mission to provide medical care. Two of the central tenets of bioethics are the principles of patient autonomy and non-maleficence—the charge to physicians to do no harm. In this case, autonomy speaks to the patient's ability to refuse medical testing. Even if she agrees to a test, the threat that she'll be punished for refusing undercuts the autonomous nature of the consent. And a doctor who forces a patient to submit to an unwanted invasive test would seem to violate the oath to "do no harm."
However worthy the goal of reducing injuries from drunken driving is, it is wrong to force a physician to flout central aspects of the doctor-patient relationship when the benefit to the patient is nil from her perspective and the benefit to society unclear, given the other available routes for testing. My judgment echoes that of the American College of Emergency Physicians, which supports law-enforcement alcohol testing of drivers but opposes rules that require doctors to obtain and report the blood tests. In the end, the bar for infringing on civil liberties should be set higher when it's a doctor or nurse who is in the position of undermining the rights of patients in the name of protecting the public. I am not sure this is a cause for which doctors would willingly to go to jail. But they shouldn't have to.