“There are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don’t know we don’t know.”1

“First, do no harm.” First, know where harm might be done. The greater risk to patients comes not from gaps in knowledge but from not knowing that we do not know. The problem is that, both individually and institutionally, we tend to disregard gaps—and, worse, may overlook the inevitability that they are there. In part this relates to our innate blindness to that which lies beyond our perception. It is fuelled by our discomfort in lacking answers, and by uncritical over-reliance on “objectivity”, at the expense, sometimes, of humanity. It is a by-product of specialisation and of professional approaches designed to regulate and reduce harm.

Safety requires as much stringency in recognising the implications of “unknown unknowns” as in managing knowledge and identified risk. We are, however, through increasing specialisation, an emphasis on diagnostic labels, guidelines, protocols and targets, and over-reliance on investigations at the expense of clinical skills, producing an impression of certainty which actively creates and perpetuates gaps. Iatrogenic harm comes in many guises. Protection of patients requires open-minded practice, lateral thinking and the ability to see the whole picture, not merely regulation. This means ensuring adequate professional freedom and trust.

CREATING GAPS: UNSUPPORTED SPECIALISATION

Specialisation matters. However, by eroding non-specialists’ experience it can reduce knowledge as fast as it extends it. It compromises the care of those whose ailments straddle or fall between specialties or within several.2

The professional psychological/physical divide is a particularly problematic example, the price of which is all the greater for being difficult to measure or attribute, and therefore readily disregarded. The truth that psychological issues and behaviour are integral to all illness indicates the necessity for every doctor to feel competent in considering their relevance. Specialised psychology services otherwise become an ironically blunt tool.

History demonstrates how assumption—frequently of “psychological” attribution—tends to fill gaps, creating a comfortable illusion of certainty. If non-specialists have little idea what they mean by this, “psychological” may seem somewhere near to “not real”, with toxic undertones of blame. The reference point tethering medical practice has shifted hazardously away from the testimony of patients. The perceived arrogance of paternalism may have gone, but in its place is a new and arguably more pernicious arrogance, that of over-inflated faith in the perceived omniscience brought by clinical investigation, such that negative results may be presumed to override the described experience of patients, far surpassing paternalism in its potential for harm. Clinical investigations, however, detect only what they are designed to detect. Science extends perception but can also blinker. The cost of its benefits relates in part to devaluation of basic clinical skills, at the expense of the subtlety of individual variation.

The divide between neurology and psychiatry presents intriguing dilemmas, encompassing as it does the perplexing conceptual wrangling of the distinction between “in the brain”, “of the mind”, “organic”, “non-organic”, and so on. Similarly beleaguered are lower areas of the brain, where, yet more problematically, autonomic regulation, influenced by both psychological and physical stress, touches every specialty but is owned by none. The cost to patients may be considerable if their various symptoms are each considered in isolation—or, worse, if they find themselves deemed hypochondriacal.

A CASE OF THE TAIL WAGGING THE DOG? DIAGNOSTIC LABELS, PROTOCOLS AND TARGETS

Definition, measurement, targets, audit and regulation increase safety, accountability and uniformity of standards within their field of vision. However, imposed structures also create blind spots, particularly when applied in time-pressed services. We are developing a system which, far from filling gaps, risks setting them in stone, creating an illusion of certainty, and blinkering us to what lies between. One model of practice does not necessarily fit all. For example, socially influenced areas of medicine better fit a “jigsaw” model than a linear “cause and effect” one. Cause, intervention and outcomes are frequently all multifaceted. Simplistic definable targets may detrimentally deflect attention from the more significant greater picture.

Diagnostic labels have meanwhile arguably acquired a status beyond their worth—no longer mere means of communication, but rather determinants of treatment options, funding and staff allocation, as though they were the absolute truth. This is fine with, for example, malignancies which are either present or not, and where management may follow directly from diagnosis. In less black and white areas, however, labels may seem more like straitjackets into which patients have to be squeezed, diverting attention from individuality and encouraging assumption. Guidelines are not necessarily applicable to everyone’s unique needs.3 For example, children who have experienced disturbed early attachment frequently show attributes resembling those of attention deficit hyperactivity disorder (ADHD), but not necessarily typically so in manifestation, implication, optimal management or priorities.4 Uncritical pursuit of diagnosis-based guidelines may deny the child sufficiently timely help: such symptoms can cost children a home, particularly if protective mutual attachment has yet to develop.5 The pragmatic “what would help?” may be the safer question than “what is the diagnosis?” Circumspection is warranted about the benefits and disadvantages of diagnostic labels for individual children.

NO-EVIDENCE BASED PRACTICE

“The practice of medicine is the art of drawing conclusions from incomplete evidence.”6

Evidence based practice alone is unlikely to be the final word in professional ethos, any more than has been so of its forerunners. Its value is clear—its ubiquitous adequacy less so.7 It is all well and good when there is evidence to be had—not so if there is not, or cannot be. It is fine if evidence is a valid representation of the whole. However, when allowed to determine practice it encourages the unmeasurable to be forced into measurable moulds, creating an illusion of objectivity.

The reality is that we also need to be equipped to practice “no-evidence based medicine”. Patients need care whether or not their problem has an evidence base. Most hazardous is a situation where evidence, far from being a tool for refining practice becomes hijacked as a means of rationing it—no evidence, no funding. This leaves beleaguered those whose difficulties fall in gaps in knowledge, between specialties, or in areas unresearched, unresearchable or beyond the commercial interest of research sponsors. It selectively disadvantages those for whom outcomes are difficult to quantify, long distant or within others’ budgets.

Anecdote and lateral thinking are necessary forerunners of evidence: attention to detail and time to reflect need to be protected and wisdom and clinical experience valued. Regulation without these is risky. Safe practice should be served by evidence, but not entirely determined by it.

FIRST, CONSIDER WHERE HARM MIGHT BE DONE

To “first, do no harm” presupposes that our concept of iatrogenic harm is watertight—a questionable assumption. In a peculiar irony of safety conscious, litigation-driven practice, risk tends to be identified only where it is sought, with a selective emphasis on the readily measurable and attributable. What about the cost of the ill-judged spoken word—often more damaging and less reversible than side-effects of medication? Or the consequences of attitude and intimation, of behaviour, of unspoken but subtly conveyed disbelief, of clumsy management of family relationships? What about the cost of inaction, or of disregard of the bigger picture of psychological and social context? What about the side-effects of professional systems, of uncritical specialisation, of stifling of initiative, of narrowing of professional vision, of over-control? Or the consequences of allowing commercial interest to affect research priorities? We need to stand back far enough to see where harm might be done if we are to avoid creating a mere illusion of safety.

LOOKING INTO THE GAPS

What are the safeguards? First, awareness of the inevitability of blind spots and of their risk. This means keeping an eye always on alternative perspectives against which to test and through which to illuminate what we think we know. It means learning from history, and remembering the distortion of seeing the narrow focus without its context, whether current or historical.

Individually, it means humility—the uncomfortable acknowledgement that time will reveal our shortcomings. Reversal of the devaluation of clinical skills of detailed listening and examination should be seen as a necessity, not a luxury. The patient’s testimony should be the presumed yardstick, any slide towards criticism prompting a search for flaws in our own understanding, and for the influence on our judgement of our own emotional reactions to children, their families, and our own role.8

Professional freedom needs to be protected, with wisdom, experience and judgement valued as necessary safeguards alongside regulation, bearing in mind that “It is the interpretation of knowledge, the engagement with new knowledge, the acknowledgement of uncertainty about knowledge, the sharing of knowledge, not the holding of knowledge, that are the characteristics of modern medicine”.9 Over-control should be resisted because it compromises initiative and encourages gaps to be overlooked.

The gains of specialisation must not be lost in gaps caused by non-specialists losing confidence or even ability. Generalism needs to be valued, not merely tolerated. All doctors need time and opportunity to see links with other specialties, with “professional development” pulled by recognition of the importance of doing so, not pushed by the requirement of fulfilling quotas.

The mind–body dichotomy is a bigger issue. As Holmes observes, “Psychological mindedness is always hard won”.8 It needs to be won though, if the value of specialist psychology services is not to be eroded by the gap surrounding them. First, we need to resist the impoverishment of medicine which comes from the illusion that mind and body can ever be separated, acknowledging the unmeasured iatrogenic harm of the ill-judged spoken word, of accompanying professional behaviour, and of perceived stigma of things psychological. Next, those on both sides of the divide need to recognise its risks and their responsibilities to reduce these. Mental health specialists can minimise mystique, and encourage paediatricians’ confidence that common sense, wisdom and life experience—no doubt the mainstay of our predecessors’ “psychological mindedness”—remain useful foundations today. Those referring to specialised psychological services should be clear of their purpose and value, and feel competent to know what question they are asking, as when requesting an x ray. “Would additional psychological support help his recovery?” may, for example, feel very different from the innuendo-laden “Is this psychological?” The harm of allusion, intended or perceived, is compounded by the unmeasured side-effect of distancing families from services which they might in fact welcome—few could overlook the psychological implications of illness; many comfortably acknowledge the relevance of stress to causation, and of psychological wellbeing to optimal recovery, yet sense stigma if services are poorly explained or offered apologetically. Paediatricians need to be equipped to work to the edge of the specialist mental health services they feed, and vice versa. Greater overlap during specialist training and continuing education would help, with priority to joint case discussions. Liaison mental health services need to be available to families and paediatricians as a routine, not an exception.10

We need to position the wing-mirrors carefully and remember to use them, not merely tighten up the Highway Code. We should resist the assumption that one model of medical care fits all, and be as stringent in evaluating the system itself as the practice it serves.