What we lose by being overly scientific about healthcare