History taking cannot follow strict rules and has to be adapted to each individual case and person. It is often wiser to take advantage of opportunities that may arise during conversation with the person relating the history than to try to maintain a chronologic and logically constructed questioning.
As previously indicated, history should be scaled in detail in proportion to the seriousness of the illness; therefore, it should be meticulously detailed for epilepsy. In some cases, apparently casual details are of tremendous importance.
Although the basic aims of history taking are fundamentally the same, there may be some variation depending on the position of the person taking the history, and the emphasis may not be the same for primary physicians and those in secondary or tertiary care. More importance may variably be given to clearing previous diagnostic confusion, to assessing overall life situation, or to reviewing treatment. History taking is also the first act of the doctor–patient relationship, with long-lasting consequences for the subsequent development of this relationship.
With children, the history is usually obtained mainly from parents or guardians, and such third-party questioning poses special problems. The child’s account, if any can be given, should be particularly facilitated. Older children and adolescents can contribute information on subjective phenomena that are unobtainable from any other source; such information may prove crucial for the diagnosis and be important to the young patients in coming to terms with their illness. One of the difficulties of third-party questioning is the increasing risk for biased and overrehearsed accounts, with the description of the attacks conforming more and more to a preconceived, once-and-forever established idea of what they are like, without the possibility of rectification on the basis of phenomena personally felt by the patient.
Another risk is to give credence to only one parent’s account, usually the mother’s, and ignore the other’s. Listening to both parents, however, might result in contradictory statements. A decision of which account is more credible must be made, which may be in part arbitrary. For these reasons, a fresh history and not a simple repetition of previously given accounts should be obtained at each new consultation. Accepting previous accounts uncritically is a major source of diagnostic errors. Jeavons
10 emphasized the trap represented by a previous
diagnosis of known epilepsy often based on a history that may not have been properly obtained. Such a diagnosis may nonetheless remain accepted for years, thus delaying the true diagnosis. Doctors’ and nurses’ accounts are potentially of great value. In some cases, however, they can also be misleading because professionals often have a tendency to describe attacks as they should theoretically be rather than as they are. Whoever the witnesses, critical scrutiny of their accounts is essential. A distinction should be drawn between first-hand versus second-hand witnesses and between witnesses who have been able to observe the whole of a seizure and those who saw only a part of it. An assessment of the degree of reliability of the witnesses has to be done. It should not be based only on how articulate they are; attempts at checking the veracity of their accounts (e.g., by gathering details that indicate whether they were really in a position to observe some reported phenomena) are important.
Obtaining a description of the paroxysmal events is the most important part of history taking, and it requires a high degree of attention and considerable time. The history, however, is not limited to this description. Such events occur in individual persons who react in their own way and in a given setting. Thus, there are two approaches to history taking—obtaining a history of the complaint and obtaining a history of the person.
History of the Complaint
It is often very useful and always recommended to ask for a description of specific attacks. This should especially include the last attack actually witnessed because it is better remembered. The physician should not forget that there may be two or several types of seizures. In such cases, efforts should be made to obtain a description of the most recent seizures of every type.
A description of the first seizure is also of great interest because it not infrequently differs from subsequent attacks. For example, it may have been a febrile convulsion followed by complex partial seizures. In addition, the impact it had on the patient’s life and the way the patient reacted to the first epileptic event can tell much with regard to later adjustment to the illness.
A description of the worst attack is also to be sought. It provides indications about the capacities for adjustment under maximal stress and about the scale of the clinical problem and the risk to life. The circumstances of occurrence, duration, and possible presence of postictal phenomena, such as Todd’s hemiplegia or aphasia, can give valuable clues to lateralization and localization.
The setting in which the attacks occurred may be of considerable significance. Epileptic attacks or nonepileptic paroxysmal events that simulate seizures are often nonrandom phenomena, and the circumstances of occurrence can be important for differential diagnosis and planning of investigations. Points of particular interest include the temporal relation of events to the sleep–waking cycle; for example, was the patient awake or asleep, and, if asleep, did the attack occur at onset of sleep, before or shortly after awakening, or in the middle of the night? In the case of daytime attacks, the following should be determined: Whether the attacks usually occur at a particular time (morning or evening) or apparently at random; the type of activity in which the patient is engaged at the time of seizures (resting, exercising, in bed, at school, playing, in a bath, eating or fasting, standing or reclining, using a computer, watching television or playing a video game, bored, emotionally disturbed, or engaged in a pleasant occupation); and the patient’s general state of health (concomitant or recent febrile or other systemic disease). In the case of a child taking antiepileptic drugs on a long-term basis, the temporal relationship of fits to drug ingestion, the possible failure to take one or several doses, and events such as vomiting or other digestive disturbances that may have interfered with absorption of ingested drugs should all be noted.
A prodrome is a long-term indication of a forthcoming attack. Prodromes (e.g., changes in behavior, such as irritability, sleepiness, feelings of hunger, sweating, hypothermia, distant feeling, and the like) are mostly absent or of secondary importance. Behavioral changes are more often seen before migraine attacks than before epileptic seizures; hunger and hypothermia can suggest hypoglycemia.
The search for a
stimulus regularly associated with the occurrence of attacks is of extreme diagnostic interest and consequently should be systematic and thorough. Identification of the regular association of attacks with an immediate stimulus is often of greater value than even a description of the attacks. The significance of the fit may also be better pointed to by the type of stimulus than by the clinical phenomenology of the paroxysm. The
regular precipitation of attacks by such stimuli as the sight of blood, blood letting, or injections, prolonged standing in hot or confined places, holding breath, or pain from minor trauma, such as bumping one’s head, is practically diagnostic of “anoxic seizures,”
15 even when the induced attack is difficult or virtually impossible to differentiate from a convulsive generalized epileptic seizure, as in convulsive syncope.
7,11 Conversely, such precipitating factors as flashing lights or startle strongly favor a diagnosis of epilepsy after such rare conditions as hyperekplexia have been excluded.
3 Certain maneuvers—for example, Valsalva’s maneuver—can induce attacks that are very difficult to distinguish from seizures but are easily recognizable when the initial phenomena are well described.
2
The
analysis of the aura or first symptom of the seizure is so important that detailed questioning is required; the patient and witnesses should be asked to concentrate on that moment. A wide range of symptoms, from vague sensations (especially abdominal and visceral) to highly elaborate motor, sensory, and psychosensorial phenomena, occurs in epileptic seizures,
9,17 but auras can occasionally precede attacks of other types, such as migraine (especially visual aura) or syncope.
4,16 In addition, no description of auras is available with young children. The occurrence of an aura can sometimes be deduced from a child’s observed behavior, such as screaming, appearing terrified, and running to his or her mother, or suddenly stopping any ongoing activity and assuming an appearance of preoccupation and concentration. Although such events are often fairly vague, their regular presence indicates that a seizure is probably of localized origin or that any more apparent first symptom is not in fact initial and can be misleading if interpreted without knowledge of previous phenomena.
Reconstitution of the
ictal sequence itself is difficult because most seizures are brief events with rapidly changing and multiple symptoms. So many things take place in so short a time that even the best witnesses are unlikely to be able to describe all
aspects of the whole sequence. For that reason, monitoring with video in a hospital or, more easily, at home is extremely helpful. It is not, however, a substitute for history taking because it often does not allow a full view of the patient and provides no information on essential clues such as awareness, responsiveness, memory of events, and any cognitive or language involvement. Inquiry about ictal incontinence and tongue biting is traditional, but the presence of these symptoms is neither necessary nor specific for a diagnosis of epilepsy. Although lateral tongue biting is unusual in disorders other than epilepsy, it has been reported occasionally in syncope.
11 The subjective estimation of the
duration of a paroxysmal attack is very often exaggerated. The person taking the history should attempt to obtain temporal clues, such as the time elapsed before taking the child to the hospital or calling a neighbor for help or phoning the doctor. It is imperative to separate the duration of the active seizure itself from that of the immediate postictal coma, confusion, or sleep and therefore to ask specific questions in this regard. Very often, the duration initially reported includes the whole of the ictal and immediate postictal phases.
The manifestations of the postictal phase and their duration are of interest because they can provide useful lateralizing or localizing clues and indications about the severity of the ictal phase.
Asking witnesses to complement descriptions by miming seizures is often more informative than a purely oral account. Often, miming reveals features that would have easily gone unrecognized, such as asymmetry of posture. The doctor’s miming of features of seizures can be of considerable help to witnesses in distinguishing between jerks and vibratory tremor and understanding what an oro-alimentary automatism or different types of head turning look like. Certain types of seizures, such as infantile spasms or other minor seizures, are difficult to recognize, and parents often recognize them only when their features are demonstrated.
Finally, the rate of repetition of the attacks should be carefully recorded, as well as the regularity or irregularity of repetition and the occurrence of clusters. This should be done for each type of seizure.
