74 S. Van McCrary / Neonatology and the specter of the law 0 30 60 90 High LD (top 10%) = 3 or above (average item score in range of "often," "very often," or "always") Texas --- New York 0 1 2 3 4 5 Fig. 1. Distribution of legal defensiveness scores in Texas and New York physician surveys. thus consistent with results from the US study showing that legal defensiveness and knowledge of medical law are inversely related. In addition to measuring the influence of law, we also studied the effects of physicians’ conceptions of medical futility. We asked physicians to assign a percentage of likelihood of success below which they would consider a treatment futile. Responses varied from zero to 60 percent, with most responses being in the one to 10 percent range [31]. One set of responses emerged as especially significant – 20 percent of responding physicians reported that their threshold of futility was zero percent. That is to say, these physicians would not view a treatment as futile if a one-in-one-million chance of success exists – in essence, a denial of futility. As we stated earlier, “in scientific terms, this is an extraordinary statement. It begs the question how could anyone know in advance when the probability of treatment success is actually zero? Perhaps more to the point, how would a doctor, who held to that standard of futility, act?” [31]. We analyzed the data and found that six physician responses and characteristics were significantly associated with this denial of futility: 1) male; a religious participant; feel not constrained by medical law; equate physician failure with inaction; are uncertain about efficacy vs. benefit; and, always remain emotionally detached from dying patients. It was especially interesting to us that this group of physicians reported not being constrained by law at all, but apparently were affected by other perceptions and attributes. These findings raise intriguing questions about how such physicians actually practice medicine, especially how they interact with patients and families in discussions about end-of-life decisions and the degree to which patient self-determination is respected. We next asked ourselves how extreme legal defensiveness might affect physicians’ perceptions of futility and how that might, in turn, affect interactions with patients and their families. Analyzing the same data, we found that a majority of physicians indicated that the probability of success defining futile treatment should generally be lower for patients with potential to benefit more from life-sustaining interventions (e.g., patients who are sentient), and higher for patients with less potential to benefit (e.g., patients in a permanent vegetative state) [30]. Stated another way, most physicians perceive longer odds worth pursuing for greater potential gain – a position that seems logically consistent with most patients’ rational self interest. However, physicians with attitudes of extreme legal defensiveness did not fit this pattern. Instead, they tended to define futility in a manner that would maximize physicians’ discretion to oppose patient preferences for abatement of end-of-life treatment. These findings sug- S. Van McCrary / Neonatology and the specter of the law 75 0 10 20 Knowledge of law: Low High Low High (n=232) (n=69) (n=122) (n=54) -------------------------------- ------------------------------- Texas New York Fig. 2. Physicians’ legal defensiveness by knowledge of medical law regarding end-of-life treatment in two states. Percent with high legal defensiveness. gest that physicians who are extremely concerned about legal implications take an adversarial position in their consideration of medical futility issues – an attitude that anticipates conflict with critically ill patients and their surrogates. Although not definitive, this analysis suggests that some physicians may inappropriately use their prerogative over medical futility as a means to guard their professional autonomy against perceived threats, and may thus limit patients’ self-determination under the cloak of legal fears. Another phase of our research replicated major portions of the previous study in a different state. Because the first study had been carried out with physicians in Texas, we wanted to ascertain which of our findings would pertain if the same questions were asked in a state with markedly different legal constraints – New York. The New York study surveyed physicians from two tertiary care medical centers in different geographic regions of the state and yielded 180 responses [33]. In general, the New York physicians performed comparably to their Texas counterparts on the quiz of medical law. Similarly, on the legal defensiveness scale the overall distribution of scores was quite comparable (see Fig. 1 for a comparison of data from Texas and New York). Like in Texas, general reported legal defensiveness was lower among New York physicians with more than 20 years experience in practice. However, when we assessed the interaction between legal defensiveness and knowledge of law, the results were strikingly different. In aggregate, the New York data show an inverse relationship to those from the Texas study. That is, in Texas performing well on the law quiz was associated with lower legal defensiveness, while in New York, higher scores on the law quiz were associated with higher legal defensiveness (see Fig. 2) [33] In order to understand this difference, it is necessary to describe two important