Preserving Patient Stories: Bioethical and Legal Implications Related to the Shift from Traditional to Digital Anamnesis

It is since the beginning of the so-called ‘digital revolution’ in the 1950s that technological tools have been developed to simplify and optimise traditional, time-consuming, and laborious anamnestic collection for many physicians. In recent years, more and more sophisticated ‘automated’ anamnestic collection systems have been developed, to the extent that they can actually enter daily clinical practice. This article not only provides a historical overview of the evolution of such tools, but also explores the ethical and medico-legal implications of the transition from traditional to digital anamnesis, including the protection of data confidentiality, the preservation of the communicative effectiveness of the doctor–patient dialogue and the safety of care in patients with poor digital and health literacy.


Introduction
In the context of medical practice, the anamnesis (also called medical history) represents an absolutely fundamental and indispensable moment, as it is the tool that allows the caregivers to form a complete and exhaustive picture of a patient's health condition.In the absence of a correct and accurate anamnesis, it is basically impossible to implement an appropriate planning of the diagnostic and therapeutic pathway.In fact, it is well known that defects in the collection of the anamnesis are often at the origin of adverse events [1][2][3].It is estimated that the collection of a correct and complete anamnesis, carried out using appropriate communication techniques, enables a correct diagnosis to be made in 76% of cases [4].
Traditional anamnesis collection is carried out through an interview between the physician and the patient, but over the past 70 years, with the advent of the digital age, more and more innovative modes of anamnestic collection, based on technological tools, are emerging.So-called "digital medical history" or "digital anamnesis" harnesses technology to collect, store, and analyse patient data, including by examining the documentation represented by electronic medical records and the outputs of wearable devices.
The digital anamnesis looms as a potential tool to ensure greater completeness of anamnestic collection and a reduced workload for healthcare professionals, while at the same time weakening the doctor-patient interrelationship, since the patient can fill out the anamnestic questionnaire in total autonomy, even from home, using digital tools.It is evident that the transition from traditional to digital anamnesis raises important medicolegal and bioethical issues.This article aims to explore these implications by comparing and contrasting digital anamnesis with its classical counterpart, in order to shed light on the challenges and opportunities presented by this digital evolution of healthcare.

The Traditional Anamnesis: Rationale and Critical Aspects
The term anamnesis derives from the ancient Greek "anamnesis" (a calling to mind) and consists of the collection of information about the patient's current and past health status.In antiquity, the first work on anamnesis that has come down to our times was by Rufus Ephesius (150 BC-80 BC).This work was published and translated from Greek into French by Daremberg and Ruelle in the 19th century, and the term anamnesis only came into use in the mid-19th century in German-speaking countries and the Netherlands [5].Traditionally, the anamnesis takes place orally, and is developed through a series of questions asked by the physician to the patient (or to family members, friends, or witnesses of a particular event in the case of special situations, such as in the case of mentally unreceptive individuals, or victims of an event they do not remember, or unconscious patients).
In order to understand the centrality of medical history collection in the diagnostictherapeutic process, it is appropriate to quote a phrase conventionally attributed to William Osler (1849-1919), a famed Canadian physician and one of the fathers of modern medicine: "Listen to the patient.He is telling you the diagnosis" [6].Generally, the anamnesis is divided into five parts:

•
Family history: The current and past state of health of surviving relatives is investigated, as well as the causes of death of deceased relatives; • Personal history: Its aim is to find out the patient's general characteristics and lifestyle habits, namely weight and height, dietary regime, alcohol and drug consumption, smoking habits, sporting activity, sexual habits, drugs and supplements taken, any allergies or intolerances to drugs, foods or environmental substances, regularity of bowel and urinary function, current and previous employment, childbirth patterns, psychomotor development in childhood, compulsory military service (for males), any pregnancies (for females); • Past history: The patient's entire medical history is investigated, from birth to the time of the anamnesis.In particular, all illnesses, surgeries, and traumas suffered by the patient must be investigated; • History of presenting symptoms: The reason why the patient came to the attention of the doctor is investigated specifically and in a high degree of detail.It is important at this stage to gather information about any similar incidents in the past and to have the patient explain how he or she managed the onset of symptoms before seeking medical attention, and with what results; • Current status: The symptoms and signs presented by the patient at the time of the anamnestic investigation are thoroughly investigated.
The collection of the medical history is as fundamental as it is difficult to do optimally.This is primarily because it is a time-consuming activity.There is no codified time cut-off indicating per se whether and how adequate the history collection is, varying greatly from case to case.The anamnesis of a young, healthy subject without pathologies and previous traumas or interventions may be absolutely complete even if collected in a few minutes.On the contrary, the history of an elderly person, perhaps with difficulty in expressing himself/herself and slow speech, and with a rich past medical history may take more than 30 min.For example, in a 2017 paper, a Turkish group conducted a study to determine the minimum appropriate duration for the examination of patients in the pulmonology setting.The investigation revealed that, of all examination phases, the collection of the medical history took the longest time, with an average duration of 5.0 ± 3.6 min (minimum: 0.25, maximum: 36.6 min) [7].
The questions that should be asked are: does the doctor have sufficient time to conduct adequate anamnestic investigations?And above all, does he/she have the appropriate communication skills to be able to carry out a high-quality anamnestic collection?
As early as 1972, Slack and Slack [8] wrote about traditional anamnesis: "Dialogue between doctor and patient is a time-honored process revered by the medical profession.During conversation with his patient the doctor can establish rapport, evaluate his patient's ability to engage in productive discussion, observe his patient's nonverbal behavior and collect historical information of clinical relevance . . .doctors as interviewers are busy, expensive and sometimes hard to find.It seems reasonable, therefore, to look for substitutes that will serve at least some of the purposes of medical interviewing in widespread and inexpensive ways" [9].
What emerges from the analysis of Slack and Slack, true pioneers in interactive history taking, is that doctors tend to be bad interviewers, because they have neither the time nor the communication skills to conduct an optimal interview.
The communication aspect is also paramount.The possession of good communication skills on the part of the physician has been shown to predict a higher level of patient satisfaction, resulting in a higher likelihood of adherence to the treatment programme [10,11].The moment of the collection of the medical history represents the first real contact between doctor and patient, i.e., a moment in which the two actors in the therapeutic relationship interface by establishing a connection, including an emotional one, between them.Underlying a quality anamnestic collection is a whole series of precise communication rules, which are far from obvious, and for which preparation and a considerable share of individual sensitivity are required.These rules are effectively illustrated by Priscilla Peart [12]: speaking clearly, ensuring that the patient can hear without the need to shout, unless they have a significant hearing impairment, using open-ended questions initially to collect comprehensive information, then transitioning to closed questions to refine understanding of specific details or events, interrupting the patient tactfully when the conversation veers off-topic, gently steering them back to the relevant information with phrases like "Can I bring you back to" or "Can you tell me more about", utilizing moments of silence effectively, giving the patient space to think and respond without feeling rushed, employing closed-loop communication techniques, explaining concepts or instructions and then asking the patient to repeat them back in their own words to confirm understanding, summarising the information discussed to ensure that the correct events or timeline are understood, and seeking clarification if there are any discrepancies or ambiguities.
Objectively, it is difficult to assume that all doctors possess the communication skills to conduct the medical history interview in accordance with the rules outlined, primarily because in the vast majority of cases they have not received the necessary training.Not to mention the aspect of time, which is of paramount importance: conducting a communicatively effective anamnesis requires time, time that is not available in departments with often hectic activities.
Another limitation of the traditional anamnesis is the fact that it takes place in one go, unlike the digital anamnesis, which involves a "pre-anamnesis" taken using digital instruments (in the waiting room or even at home) followed by the actual anamnesis.The fact that the traditional anamnesis takes place in one go actually makes it more difficult for the patient to recall his or her past medical history effectively.In fact, when it comes to recalling events that took place many years earlier, the patient often needs time to put ideas in order.It is therefore far from uncommon (especially in the case of elderly patients and patients with a very extensive past medical history) for patients to provide incorrect, incomplete, or partial information during the anamnesis.
Another aspect that should not be underestimated is that of the language barrier [13]: very often the taking of a medical history is greatly hindered (sometimes made completely impossible) by the fact that the patient does not speak the language spoken in the country where he or she is being treated.This problem is usually remedied in the traditional anamnesis by having recourse to translators (who may simply be the patient's family members or duly appointed staff on duty at the hospital) or, in the worst case, by speaking in "universal" languages such as Spanish, the second-most spoken language in the world, or English, ranking third in worldwide prevalence.The language barrier is, however, a fundamental problem of medical history taking, which can lead to miscommunication underlying medical errors [14].
A further weakness of the traditional anamnesis is that the doctor often forgets to ask certain questions, omitting to report some important anamnestic features.Only in some cases, in fact, is the anamnesis collected through the use of pre-filled forms (which act as a "checklist").In other cases, the doctor simply writes on a blank sheet of paper or at most the titles of the sections to be completed, without specifying the individual aspects to be investigated.
Indeed, it can be difficult for the doctor to remember all the aspects that need to be investigated.Suffice it to say that in order to avoid forgetting, mnemonic techniques have been developed to enable the physician not to forget to investigate important aspects of the anamnesis.The best-known example is the acronym "SOCRATES" for the study of pain characteristics: site, onset, character, radiation, associations, time course, exacerbating/relieving factors, and severity [15].Other examples concern acronyms that help to remember the individual phases of the history, such as AMPLE (allergies, medications, past medical history, last meal or other intake, and events leading to presentation) and SAMPLE (like AMPLE but with the addition of the S, which stands for signs and symptoms).
Listed below are the aspects that, in the authors' clinical experience, are most commonly overlooked by the physician taking the traditional history: Complete pharmacological history: Including the use of over-the-counter drugs, food supplements, herbal medicines, and the patient's adherence to the prescribed regimen.

The Digital Anamnesis: History, Potential and Critical Issues
A digital anamnesis is a procedure through which the patient answers questions about his or her personal medical history without interacting directly with a doctor, but rather with a technological tool (computer, tablet, applications on mobile phones) through traditional software or chatbots and virtual assistants based on artificial intelligence [16].
The first digital anamnesis systems date back to 1949, when a study was conducted at Cornell University that processed paper questionnaires via computer [17].The Cornell Medical Index, with 195 questions for patients, revealed that this system collected 95 percent more information than the clinical one and was thorough and satisfactory for patients.The use of the computerised questionnaire made it possible to obtain a wide range of data on the patient's medical history without taking up the physician's time, facilitating the interview and ensuring that no significant symptoms were overlooked.Of course, at this early stage, the computers were technologically backward, so they were large, located in special rooms, operated with punch cards, and controlled by an information technology specialist.It was not until the 1960s that there was a significant technological advancement in the interface between patient and computer: Slack and others introduced video terminals connected to minicomputers to interview patients about allergies, no longer using punch cards [18].Since the 1970s, a large number of studies have tested the advantages and disadvantages of computer-generated medical histories.
Table 1 summarises the characteristics of the most relevant digital anamnesis studies between 1968 and 2001, taken from Bachman's work [9] (modified and supplemented).Today, there are mainly 18 digital instruments [78] used to collect the medical history, the main features of which are summarised in Table 2, listed in chronological order of development.Six of them (Instant medical history, HELP System, AIDA, ParentLink, CIDI-Auto, and MEDoctor) were created between the 1980s and early 2000s, while the remaining 15 were developed more recently, in the last 20 years.Introduces a concept of a self-anamnesis implemented as a mobile application for patients -Uses a conversational user interface to simulate the conversation between patient and therapist, improving the interaction and accuracy of the information collected -Equipped with 63 questions posed successively to the user, using a rule-based approach and Artificial Intelligence Markup Language (AIML) to manage the chatbot conversation -Offers advantages over traditional digital questionnaires, such as encouragement to complete all questions and the possibility for the user to ask for clarification -Integrated into the care process, allowing collected data to be stored in a structured manner on an eHealth platform, making it accessible to therapists or physicians for a better information base for initial consultation and subsequent decision-making

Tool Key Characteristics
App for digital medical history taking in urgent care practices [110] -Designed to collect patients' medical history based on symptoms in out-of-hours urgent care settings -Allows patients to select their acute symptoms and guides them through a series of related questions, such as specific symptoms, relevant pre-existing conditions, and previous treatments or medications -Questions and answers are formulated in plain language, with a simple design and intuitive navigation, making the app easy for patients to use without prior instruction.After completion, it generates a structured medical history summary that can be transferred to the patient's electronic medical record (EMR) The digitised anamnesis by means of self-compilation by the patient (thus, in fact, a 'self-anamnesis', which does not replace but complements the classical anamnesis) offers undoubted advantages for both healthcare professionals and patients.
First of all, patients come to the traditional anamnestic interview better prepared, having already gone through their clinical history and thus having in their memory all relevant information to provide to the doctor.As explained above, in fact, digital anamnestic collection systems do not replace but complement traditional anamnesis, acting in fact as a pre-anamnesis.This facilitates the patient, who has time and a way of recalling his or her medical history, but also the physician, who, having the pre-anamnesis data at his or her disposal, can conduct a more targeted, precise, and detailed anamnesis.
Secondly, the automation of the anamnesis process significantly reduces the administrative burden on healthcare personnel [111], both in terms of time and bureaucracy.This has a significant effect not only on the collection of the anamnesis itself, but also on the entire activity of the physician, who will perform his work with less stress and less risk of burnout [112].
Another aspect of primary importance is that a digital collection of the medical history enables a structured, and thus more complete, recording of the information provided by the patient [113].It should also be noted that the increasing development of AI (artificial intelligence) systems could bring the efficiency of digital anamnesis to an even higher level in the coming years.AI algorithms can analyse patient responses, identify warning signs, and prioritise critical information for healthcare professionals, thus facilitating efficient decision-making [114].
A final positive aspect to mention is the ease of collecting information that the patient finds embarrassing or ashamed of (e.g., risky sexual behaviour).Interfacing with a digital, non-human instrument undoubtedly facilitates the communication of such aspects pertaining to the private sphere, there being on the other hand no possibility of a moral judgement regarding the patient's conduct.However, the collection of anamnesis through digital tools also has its downsides.
First, there is the fundamental issue of the so-called "digital literacy" [115,116].Not all patients are inclined to use digital tools, and not all patients are able to interact effectively with them.It has to be said that the more technology advances, the more effective and engaging the interfaces become, so it is certainly easier today to get patients to adhere to them than it was with the first systems developed in the 1960s.
A second potentially negative aspect of relevance is that whereas in traditional anamnesis the patient is assisted and helped by the physician in reconstructing his or her medical history, in digital anamnesis he or she is in fact alone.This means that there is no guarantee that the patient adequately understands the questions he or she is asked, and there is a risk that he or she may provide inaccurate answers.
Another major negative element is privacy and security [114].If patient data are collected via digital systems, they are inevitably exposed to possible intrusion by hackers or unauthorised parties.This imposes the need for healthcare facilities to equip themselves with protection systems (antivirus software) that are often very expensive [117].
A further critical aspect related to digital anamnesis is the depersonalisation of the therapeutic relationship.The traditional anamnesis is normally lengthy and time-consuming, but during such an interview an emotional connection is established between the physician and the patient that forms the basis of the therapeutic alliance.The anamnesis collected following the compilation of the pre-anamnesis through a digital tool is certainly quicker and more effective than the traditional anamnesis, but it has the disadvantage of not being long and in-depth enough to allow the establishment of a solid emotional and communicative bond between doctor and patient.

Bioethical and Medico-Legal Profiles
A first aspect of medico-legal relevance is undoubtedly that of confidentiality.In fact, this aspect is also of bioethical importance, if we consider how sensitive certain types of information can be, and how much the doctor-patient relationship can be affected by a breach of trust in maintaining professional secrecy.In the United States there are specific regulations that protect confidentiality, such as the Health Insurance Portability and Accountability Act (HIPAA), in force since 1996, which enforces the protection and confidentiality of patient data and provides for articulated privacy and security rules (including, for example, encryption and audit logs on access).In Europe, however, there is the EU Regulation 679/2016 (GDPR, General Data Protection Regulation), the privacy regulation that became fully applicable in all EU member states on 25 May 2018.The differences between HIPAA and the GDPR are essentially fourfold [118]: the GDPR provides for greater restrictions on the processing of health data in the absence of patient consent; HIPAA does not provide for the so-called 'right to be forgotten', which is covered by the GDPR; HIPAA provides for a much simpler process than the GDPR with regard to the handling of data breaches; the economic penalties are much harsher in Europe (maximum of EUR 20 million or, for companies, 4% of total annual worldwide turnover versus fines of 100,000-250,000 per breach in the US, with an annual cap of USD 1.5 million).What is crucial to do when using tools such as those that make digital anamnesis possible is to invest significantly in IT security systems and staff training.This means implementing robust security measures such as firewalls, antivirus software, and intrusion detection systems, as well as ensuring that mobile devices used to transmit confidential information are properly encrypted.
A second aspect of purely bioethical interest is that relating to the characteristics of the doctor-patient relationship, the physiognomy of which is evidently modified by the use of digital tools for the collection of the anamnesis, which represents the first act in the relationship between the two protagonists of the therapeutic alliance.Although originally developed in the field of psychotherapy [119], the therapeutic alliance has gradually asserted itself, together with the overcoming of the paternalistic model, as the ideal model of empathic relationship between doctor and patient, and since the collection of the clinical history represents the first moment of contact between these two figures, it is legitimate to wonder whether depersonalising it might not run the risk of jeopardising the establishment of an effective relationship of mutual trust between the healer and the treated.This is indeed a real problem, considering that a physician's knowledge of the data emerging from the outcome of the digital 'pre-anamnesis' will undoubtedly lead them to shorten the interview time, and above all to conduct a 'surgical', targeted, selective anamnestic examination.This accelerated mode of collecting the medical history undoubtedly entails a major saving in terms of time and effort, as well as obvious advantages in organisational terms (more time and energy to devote to the care of other patients), but it puts the aspect of empathic communication in the background, if only for mere temporal reasons (establishing an empathic relationship takes time).It is therefore necessary to ask oneself whether, from an overall perspective, one can consider oneself willing to accept a reduction in the probability of establishing an effective therapeutic alliance in order to benefit from time and energy savings and enjoy an optimisation of the organisation of healthcare.It should also be considered how the flattening of the doctor-patient relationship can be particularly important in certain socio-cultural fabrics, such as rural communities, where doctors are often not only health professionals but also integral members of the local community.Here, the doctor-patient relationship transcends the purely clinical realm, intertwining with community dynamics and cultural expectations.The erosion of personalised interaction facilitated by digital tools may not only hinder the establishment of trust and rapport, but also weaken the community fabric in which the physician plays a central role.The loss of this interpersonal bond could be perceived as a decrease in community cohesion and could lead to a sense of alienation among patients who rely not only on medical expertise, but also on the emotional support and solidarity provided by their local healer.It is therefore imperative to recognise that the implications of communication flattening vary in different healthcare contexts and underline the need for nuanced and context-sensitive considerations when implementing digital solutions in medical practice.
A third and final aspect that is important to highlight as it has important implications on both the bioethical and medico-legal fronts is that of safety of care.Digital medical history collection can only be used effectively and safely if it can guarantee the acquisition of real and reliable data.In contrast to a clinical interview conducted by a physician, a computerised anamnestic collection poses problems relating on the one hand to the patient's digital literacy (his familiarity with the use of technological tools) and on the other hand to his health literacy (his degree of knowledge of medical terminology and medicine in general).The first problem is probably relatively easy to solve, since, as shown by the above-mentioned literature, the development of increasingly simple, attractive, and engaging interfaces makes it possible in the vast majority of cases to overcome the obstacles related to the unfamiliarity of users (especially older users) with technology.The second problem, on the contrary, appears in our opinion to be more complex to solve.How can one be sure that the patient adequately understands the meaning of the questions posed to him by a digital system with which he cannot interface as directly as he would with a human?How can the digital system confirm that the patient's answer is not the result of a blatant misinterpretation?How can a computerised system adapt the type of language and the way medical information is explained to the degree of culture and preparation of the individual patient?These are questions that currently remain with a question mark.Indeed, the paradoxical scenario cannot be ruled out whereby a physician faced with a patient with a limited technological proficiency background does not trust the information resulting from the digital pre-anamnesis at all and prefers to take a traditional anamnesis from scratch, with the absurd result that instead of saving time, time would be lost.
Again, with regard to safety of care, it should not be overlooked that there is a risk that a systematic use of digitised medical history collection systems could lead doctors to over-rely on digital technologies, which could be accompanied by a decrease in clinical skills.If physicians rely too much on results generated by algorithms and chatbots, they might neglect to carefully examine patients and consider other important factors in the diagnosis and treatment of diseases that are not already considered by digital systems.In other words, a kind of 'laxity' of doctors could occur, which could lead to a decrease in the quality of care provided and an increase in malpractice cases.
The issue of safety of care is intimately connected with the issue of medical liability.Should malpractice occur due to errors in anamnestic collection, who would be held liable: the doctor, the software developer, the software programmer, or the company marketing the device?This is a thorny issue that has been widely debated for several years.The European Parliament and the Council are in the process of finally adopting the proposal for a regulation presented by the European Commission on 21 April 2021, on a legal framework for artificial intelligence ('Artificial Intelligence Act').Both the product liability directive and the artificial intelligence non-contractual liability directive are under discussion.Not all digital anamnestic collection systems are based on artificial intelligence, but the liability profiles applicable to such systems may in many ways appear overlapping, since they are non-human entities interacting with patients and actively involved in the diagnosis and treatment process.
The issue of liability from digitised collection of informed consent would actually be much easier to handle than liability from malfunctioning of devices physically operating on the patient (such as robots used in surgery).This is because the activity provided by digital anamnestic collection systems is purely intellectual, which evidently implies that it is much more easily controlled and governed by human doctors.It is quite clear that a human surgeon has very little (if any) control when faced, for example, with a mechanical arm of a robot that autonomously performs (on the basis of AI algorithms) an incorrect manoeuvre causing injury to an anatomical structure.It is equally evident how an omitted or erroneous anamnestic information represents, on the contrary, an error that is much easier to remedy, also considering the fact that digital anamnesis, as widely observed, represents only a complement to the traditional anamnesis, not replacing it.It therefore seems legitimate to argue that the greater the possibility on the part of the physician to control and govern the work of the non-human system, the greater the liability profiles for the physician.In the case of the digital anamnesis, therefore, it is plausible to assume that the heaviest burden of responsibility would be on the physician, who in any case has the burden of verifying that the information assumed by the digital collection system is correct and true, and above all that no important information has been overlooked.However, it should be pointed out that if the physician were to go through the traditional anamnesis again to verify the reliability of every single piece of information collected by the digital system, the digital anamnesis would lose its meaning, leading to no advantage in terms of saving time and work for the physician (who would basically be forced to take a traditional anamnesis from the beginning in order to avoid any liability charges).This is an issue that is difficult to resolve, since the fact that the final responsibility for the adequacy of the anamnestic collection falls on the physician makes the latter obviously interested in ensuring that the preliminary anamnestic collection made by the digital system is correct and accurate, and this would precisely lead to the nullification of the advantage of time and effort that the use of the non-human system would bring, since the physician would take a new anamnesis from scratch (with the risk, moreover, of collecting less reliable information than that collected by the digital system in the first instance, since the patient would be called upon to repeat his or her entire medical history from the beginning, and would therefore be more fatigued and less attentive, and thus more prone to making mistakes).

Conclusions
In conclusion, in the authors' opinion, it is completely unreasonable to halt technological progress by hindering the development and dissemination of technological tools that can concretely facilitate the work of healthcare professionals, saving them time and energy.What needs to be done, rather, is to attempt to redefine the established structures of the doctor-patient relationship in such a way as to make it adhere to the new clinical reality.
As has emerged from the outcome of the literature review, the most concrete issues related to the implementation of digital anamnesis systems are those of communication flattening and of safety of care in relation to the treatment of digitally illiterate patients.These challenges have to be addressed proactively, e.g., by fostering the development of new communication skills on the part of physicians and by developing digital systems that are able to compensate more and more effectively for the gaps (digital or medical knowledge) of patients.

Table 1 .
The main digital anamnesis tools developed from the 1970s to the 2000s.

Table 2 .
The main digital anamnesis tools used in clinical practice today.