No other medical consultation is as broad in terms of topics to be explored (ie, medical and clinical, psychological, and biosocial) [6,10,23] and limited in terms of time as the primary care consultation: mean consultation time is 5 minutes or less [24] or around 9 minutes, with general practitioners feeling that a consultation of less than 10 minutes for primary care was inadequate [25]. A DMIA system could improve primary care consultations in several ways, based on the benefits related to the medical consultation, broadly conceptualized. Textbox 1 and the following paragraphs summarize potential benefits and disadvantages that DMIA systems may bring forth for medical consultations in general, based on both theoretical work and on studies evaluating the capacities of these systems. In the textbox, some articles are perspective or commentary articles, theorizing about the capacities of DMIA systems [1-3,19,26], and others are studies evaluating systems and presenting results [21,27-29].

First, DMIA systems have the capacity to acquire a more comprehensive set of information than that attainable during a face-to-face consultation. This could provide greater insights into potential risk factors and, possibly, suggest a more accurate differential diagnosis, prior to seeing the patient [2,3,30]. Secondly, it allows for better use of face-to-face consultation time, as previous existing conditions and information about presenting complaints have already been identified before the patient enters the physician’s office [2,3]. This allows for a more streamlined consultation, where time that would have been spent on history taking is shifted to discussing management strategies, building rapport, and focusing on the relationship with the patient.

Third, it improves data quality [2,20,21], which in turn improves the diagnosis, as the quality of the diagnostic process is significantly affected by the accuracy of the information made available to the primary care provider. Fourth, DMIA systems include more up-to-date information, including that available from tests and other health care providers, which helps maintain continuity of information across different health care providers [30].

Additionally, once a diagnosis has been made by the primary care provider and subsequent consultations have been scheduled, DMIA systems can help flag specific aspects that need attention related to the ongoing management or treatment of the condition, and they can collect subsequent patient concerns prior to the follow-up consultation. As a result, the primary care provider in follow-up consultations can focus on guiding the patient regarding self-management aspects and provide further education that might be needed [3]. Alternatively, the information in a DMIA system may flag the need for specialist visits or management from other team members (ie, allied health professionals) in advance of seeing the patient, leading to earlier referrals or timely, urgent treatment, if needed [30].

However, there are also reported disadvantages of DMIA systems, which include possible technical issues, related to problems with programming or design, that may lead to missing or erroneous information [21]. Potential problems could occur as a result of a patient interacting with a computer, such as perceived impersonality of the health care provider–patient interaction or missing body language cues [1,19,27]. This last issue could be overcome by new facial and/or body movement recognition technologies, such as affective computing, where automated analysis of facial expressions can provide accurate depression diagnosis, for example [31]. Also, it has been described that some primary care providers attempt to confirm all the answers, duplicating efforts [3]. Finally, DMIA systems, as with other technologies, require technical supervision and maintenance [1,19,28], which may result in additional resource spending.

We described the role that a DMIA system could potentially play in managing consultations on diabetes and depression within the primary care context. The anticipated benefits of incorporating this technology into the primary care consultations for these chronic conditions should mirror those described in the literature: namely, a more comprehensive set of patient information items, better use of face-to-face consultation time, better-quality and more up-to-date data, and more frequent interactions with the patient, among others.

All the information gathered outside of the face-to-face consultation means that the actual time in consultation is better spent. Conversation can be streamlined to address the needs of the patient, such as building rapport, providing education, or responding to concerns, and to improve communication and patient understanding, which ultimately leads to increased patient satisfaction [41]. Relatedly, better-quality data prior to the consultation, as well as support for additional data coming from other levels of care or lab tests, allow for early detection of a chronic condition, earlier and enhanced treatment at the primary care level, better monitoring, and early management of possible complications, all of which translates into better care. In addition, as the DMIA communicates with the patient before and in between consultations, it improves the continuity of care and it gives the patient the feeling of being better cared for, as communication with the primary care provider via the DMIA system occurs more frequently.

Additionally, DMIA systems could improve the treatment and management of chronic conditions, as seen with diabetes and depression, without too much additional effort once the system has been set up. As mentioned above, chronic condition management usually follows a pre-established pattern. Therefore, it would not be difficult to develop a rules-based system to check case finding and diagnosis and appropriate management strategies. Then, once developed, the system could be deployed and repeated throughout, using the same branching logic and platform. Moreover, artificial intelligence could be leveraged and introduced into a DMIA system, which would potentially improve questioning algorithms and language, among other improvements.

For such a system to work well, there are several considerations that need to be taken into account. As presented earlier, there are some disadvantages that need to be bypassed or considered when implementing DMIA systems. Some patients may not be able to read the materials in digital form or may not be digitally literate; about 10% of the population chooses not to complete their histories on computers [3], although this figure may vary depending on how technologically acquainted people from different populations are. General practitioners sometimes try to confirm all the answers to the questions, thereby duplicating the work of the computer, which obviously impacts efficiency [3], although as technology and program design improve, there should be more confidence from these physicians in the results obtained by these programs. Additionally, there may be patient data entry and potentially added work for the primary care provider, as more issues might be highlighted and in need of attention. Moreover, bringing a DMIA system into everyday clinical practice requires interdisciplinary collaboration with a team of specialists and primary care physicians, patient advocates, computer scientists, big data analytics, and experience design [2].

Another consideration relates to the language to be used. The ideal DMIA system should provide a human-like interaction. For example, for a consultation ahead of the follow-up, it should phrase the questions in such a way that the patient feels as if they were continuing a previous conversation. Technologies today allow for human-like conversations in the form of a chatbot—also known as a machine conversation system, virtual agent, dialogue system, conversational user interface, and chatterbot—where a computer program interacts with users by using natural language precisely enough to simulate human conversation [42,43]. As such, the system may ask “Have you been able to adequately manage the [chronic condition] we discovered last week? Is there something you’d need extra help with?”

Another essential consideration relates to the factors that influence the adoption of eHealth innovations in health care in general, and primary care in particular. On the one hand, there are factors related to the adaptability, flexibility, and cost of the technology to be implemented [28]. DMIA systems need to be sufficiently flexible and affordable so that they easily enter and adapt to different primary care contexts and into already-existing working practices and systems. On the other hand, the environment in which the technology will be implemented needs to be ready for such a transformation and must include supportive policies, incentives, and leadership, as well as the appropriate resources for implementation [28,29]. In addition, a primary care provider’s knowledge and experience with digital technologies needs to be taken into account, as well as their willingness to incorporate a new technology into their everyday practice [28,29]. Only the proper alignment of these factors will enable a technology such as a DMIA system to be successfully implemented and, ultimately, improve care.

Finally, by transforming the primary care consultation in the way we describe here, a DMIA system would standardize the consultation and follow-up of chronic conditions and improve clinical care. We foresee that, given the described benefits, it should become the norm and a regular practice accompanying the primary care consultation if implemented in the appropriate way, by addressing the challenges and factors mentioned above. Moreover, it can become part of a suite of digital health technologies, amplifying its impact. For instance, by directly connecting a DMIA system to an electronic medical record (EMR) system and/or to a clinical decision support system (CDSS), it can have synergistic effects that may transform the way health care is provided in the future, such as providing rapid and remote access to care, improved triage, and more accurate diagnosis, just to name a few benefits. The answers provided to a DMIA system, which are automatically stored in a digital format, can populate whichever data fields are needed for an automated CDSS, providing support across a wide variety of clinical fields and issues [2,3,20]. In addition, clinical data can be standardized across all patients in contact with a health care system and that data can be included in a database to feed and support clinical and population health research [2].

A DMIA system could enhance the primary care consultation and facilitate the management of diabetes and depression, and possibly other chronic conditions, which would hopefully make an impact in primary care. Earlier case finding and a more accurate diagnosis, due to more and better-quality data, paired with improved monitoring of disease progress, should improve the quality of care and keep the management of chronic conditions at the primary care level. A somewhat simple, easily scalable technology could go a long way to improve the health of the millions of people affected with chronic conditions, depending on the context and how successful its implementation is, especially if working in conjunction with already-established health technologies, such as EMRs and CDSSs.