Use of symptom checkers to increase care efficiency - Competent patients and better decisions?

Symptom checkers offer high potential for addressing a wide range of challenges in the healthcare system [1]. On the one hand, the applications can contribute to increasing efficiency in doctor-patient interaction. On the other hand, they offer well-founded instructions for self-management of symptoms of illness from home and thus also provide a starting point for further patient navigation to the care levels. There is a steady increase in doctor's time per patient in both inpatient and outpatient care. The work pressure for health workers is also constantly growing. On the patient side, this is noticeable, for example, in the form of long waiting times for (outpatient) doctor's appointments and only short doctor-patient consultation times [2]. According to a meta-analysis of outpatient primary care, the average duration of a doctor's consultation in Germany is 7.6 minutes [3]. With a consultation time of around seven minutes, it is obvious that this must be used as efficiently as possible. After all, it is important to accommodate the presentation of the symptoms, the examination and diagnosis as well as the discussion of treatment options within this narrow time window.

>> The density of doctors per inhabitant in Germany is higher than the OECD average, with an average of 9.9 medical consultations per year for each insured person (see Fig. 1) [4, 5]. Despite this relatively high density of doctors, which would suggest that patients with their health complaints can receive a quick medical consultation, many people are increasingly visiting hospital emergency rooms for a wide variety of complaints.

Overcrowding in emergency rooms
Emergency departments are, by definition, centers where the priority is to provide very rapid care. This goes hand in hand with high work pressure and stress levels for those working there. At the same time, emergency rooms in Germany are freely accessible to all insured people. This means there is a very high potential for incorrect use. The number of patients in emergency rooms is constantly increasing, with the proportion of patients with low treatment urgency (so-called overcrowding) growing in particular [6]. Overcrowding emergency rooms poses high risks for patient safety because medical staff are often overloaded and valuable time for the life-saving treatment of emergencies is misallocated to less urgent treatment cases. The reasons why patients with low treatment urgency go to an emergency room include, in addition to time flexibility - compared to making appointments with private doctors - the hope for high-quality or better care as well as uncertainty and fears regarding their own state of health or symptoms [7]. The subjective urgency of treatment is assessed as high, especially by patients with low health literacy. Emergency patients in particular have lower health literacy compared to the German population, as a current German study shows [8]. These results indicate that health literacy is an important factor influencing the allocation of medical services.
Alliance for Health Literacy: less
Cyberchondria, more patient empowerment
Health literacy – also known as health literacy – is the ability to find, understand, interpret and apply relevant health information [9-11]. The continuous increase in unfiltered health information available on the Internet now makes it essential to have digital health literacy. Around 72% of Germans use the Internet to research health information, with the search for disease symptoms and causes being the most frequently cited motivation by respondents [12].
On the other hand, a large part of the German population (around 76%) has low digital health literacy. That is why the free search for health information on the Internet, on the one hand, entails risks for patients and for the community of those insured by statutory health insurance, for example through unnecessary anxiety for those affected (so-called cyberchondria2) [13], and on the other hand, it often leads to an overuse of health resources.
A clear trend is emerging: a large proportion of Germans initially search for health information on the Internet on their own. This search is a step in self-management of your own symptoms of illness. It is precisely this self-management – including the independent search and synthesis of information before seeking health care – that has already been addressed by the legislature. Through the adoption of the Digital Care Act in 2019, which allows for the self-determined, health-oriented use of digital applications, such as: B. symptom checkers, through which the insured are intended to promote the strengthening and promotion of the population's (digital) health literacy [14]. In this act of promoting the self-determination of patients - i.e. patient empowerment - health insurance companies and public institutions therefore offer offers and measures to strengthen health literacy. This begins with the training of basic health skills during school and continues all the way to the patient's bedside, where informed therapy and treatment decisions can then be made together and on an equal footing - ideally in a process of shared decision making. This legal approach to strengthening digital care and self-management can also be illustrated using the Kaiser pyramid (see Fig. 2) [15].
If self-care in the area of self-management of patients is inadequate, then there is a need for techniques that further support self-management until seeking medical advice and treatment becomes essential (see Fig. 2).
Health literacy “to go” and less pressure from a doctor?
Digital applications such as: B. Symptom Checker. These applications help classify symptoms and navigate the healthcare system (more efficiently) by systematically recording information, filtering it according to relevance and supporting assessments of the urgency of symptom clarification [16-20]. This also corresponds to the desire of every second German to use more digital technologies in order to be able to independently and medically assess their own state of health and complaints [21].
Symptom checkers have two main functions: supporting self-diagnosis and better assessing the urgency of treatment (triaging). The process of using symptom checkers based on an algorithm-based questionnaire (see Fig. 3) is similar to that of a medical anamnesis: location of the complaint, intensity of the complaint, symptom severity.
Intensity and their time of occurrence, possible triggers and accompanying symptoms as well as individual risk factors and limitations are recorded [22]. The questions may vary depending on the symptoms reported [23].
As a result, the symptom checker then presents a ranking list with possible diagnoses and an associated triage recommendation. Most symptom checker algorithms are based on the current state of evidence-based medicine and have learning AI systems (so-called artificial intelligence), which means they gain in precision with each new diagnosis [24, 25].
In recent years, a variety of symptom checkers have been developed and embedded in healthcare, such as: B. the now best-known symptom checker from Ada Health, the multifunctional symptom checker app from the Swiss company MiSANTO or the symptom checker from Infermedica, which is integrated on the hospital website in cooperation with Sana Kliniken AG [24-26 ], to name just a few examples. The focus of the various symptom checkers is on the idea: first check and inform digitally, then act and, if necessary, treat.
Symptom checker: the evidence
The evidence base for symptom checkers is still heterogeneous, as studies and reviews cannot show clear trends regarding effectiveness and patient safety, measured by diagnostic accuracy and triage accuracy [e.g. 19, 27, 28-32]. Positive results were achieved in the area of evaluating the usability of symptom checkers and in the area of interaction with medical staff using symptom checkers [29, 32-34]. The users felt a generally high level of satisfaction with the simplicity and flexibility in using symptom checkers [35].
Primary studies and a review showed that the diagnostic accuracy of most symptom checkers was low, i.e. i.e., the correct diagnosis was only listed among the first three suggested diagnoses in around 19 to 38% of the cases examined [19, 27, 29]. The triaging accuracy of the symptom checkers, i.e. the accuracy of an urgent recommendation for seeking medical help, was in a moderate to high percentage range (49 to 97%) [for example 27, 30, 32, 36, 37]. However, it turns out that symptom checkers have tended to become more risk-averse: According to a review, up to 40% of emergencies went undetected, which can have fatal consequences at the patient level and at the health system level [30]. However, there are large differences between symptom checkers and disease indications in terms of diagnostic accuracy and triage accuracy. The main reason given for this was the very different and therefore not easily comparable study designs [27, 32, 38]. While some studies compare the recommendation of the symptom checker with the recommendation of a healthcare layperson, other studies use medical staff as a comparison group.
Conclusion and outlook: a door opener into the digital supply landscape
Basically, it shows that symptom checkers can help find the way out of the paternalistic atmosphere.
health system towards a more self-determined health
to pave the security system.
The main reason for using symptom checkers is existing uncertainty about your health status. This uncertainty can apparently be reduced using symptom checkers. In addition to diagnosis, symptom checkers derive evidence-based recommendations for action, provide evidence-based health information and cover a wide range of languages. Symptom checkers therefore have an inclusive effect, even for culturally or educationally disadvantaged population groups (e.g. people with a migrant background) [39]. If patients engage with this health information - especially before the medical consultation - health literacy can be strengthened, which is also beneficial for the patient's role during the medical consultation. For example, a patient's need for information can be formulated more easily, which empowers the patient.
In general, it can be said that the study designs of previous studies on symptom checkers are only comparable to a limited extent, which is why a comparison of the individual studies is hardly possible. Nevertheless, it can be noted that the diagnostic accuracy of symptom checkers is generally still at a rather low level, while most symptom checkers perform better in terms of indications of the urgency of medical treatment (triage accuracy).
In any case, the goal cannot be to replace medical consultations with symptom checkers. Against this background, the diagnostic accuracy of symptom checkers may be considered less important. A possible solution to circumvent the current lack of clarity in diagnoses from symptom checkers would be to first transfer patients to a doctor chat mode. This option currently appears to be a better choice as it simultaneously helps mitigate risk and avoid misallocation of emergency capacity. Triage accuracy, on the other hand, is of primary importance - on the one hand to avoid fatal consequences due to non-treatment, on the other hand patients can be navigated more effectively through different levels of care, which would optimize the allocation of resources. Symptom checkers therefore offer relevant potential for reducing avoidable or incorrect use of doctor's time. In this way, medical staff can be freed up to treat urgent medical cases (e.g. especially in emergency rooms) with the intensity and capacity they require.
Nevertheless, there is a fundamental risk of overtriaging and the associated increased seeking of medical advice - which can also happen anyway through finding and unreflectively accepting misinformation from the Internet. The impact of symptom checkers on resource consumption in healthcare depends heavily on the behavior (e.g. level of anxiety as a characteristic) and the individual health literacy of patients. According to a review from Austria, the use of medical assistance through symptom checkers is decreasing, but at the same time telephone contact with doctors based on symptom checker recommendations has increased [32, 40]. However, the increased use of telemedical services is less significant because trained triage specialists are also used for this purpose and a telemedical consultation takes less time and can therefore relieve the burden on the care system.
In the future, symptom checkers can be integrated into care at many points, because these applications could help to better assess the urgency of the symptoms and the need to seek medical advice, especially outside of practice opening hours [41]. They join a list of digital solutions such as: B. the electronic patient file or teleconsultation. These applications can sustainably transform traditional care paths and thus counteract increasing costs [42]. For example, connecting symptom checkers to the electronic patient file through data donation from users can open up opportunities for quicker anamnesis and support of the diagnostic process. For example, For example, signs of rare diseases can be registered earlier and potentially diagnosed more quickly. All symptom data entered by patients could also be viewed by medical staff in the symptom checker database and thus provide information about recurring, early symptoms for rare diseases. In addition, the potential of the symptom checker database could also be optimally exploited in a public health context. For example, the pseudonymized data from the symptom checkers could support the RKI's early warning systems for infectious diseases [43]. Symptoms of infectious diseases and their local occurrence, such as Covid-19, can be recognized much earlier thanks to the real-time data diversity of the symptom checker databases than data that occurs through the traditional reporting process of the doctor's office to the RKI.
Symptom checker offers would also be conceivable, in which patients are connected to medical staff directly via chat or telephone during the process or upon request (e.g. in the sense of first-second-third-line support). The transfer to preventive and telemedicine measures or the initiation of a laboratory examination would also be possible starting points. Ultimately, it is possible to make the health data usable for research purposes after patient information and with informed consent.
Against this background, the potential of symptom checkers is far from exhausted, although important further development steps and associated studies appear to be necessary with regard to diagnostic accuracy, comparability and range of services. Symptom checkers offer a very good alternative to erratically Googling medical information and are therefore essential for avoiding cyberchondria and promoting patient empowerment. Symptom checkers should therefore continue to play an increasingly important role in the medical care landscape in the future. <<