Digital clinic users in Finnish public primary care: their characteristics, reasons for visits, and equity considerations – a register study

Abstract

Introduction

Chat-based digital clinics have become an increasingly important gateway to primary healthcare in many countries. This exploratory study describes the socio-demographic characteristics of digital clinic users, examines patterns of digital clinic use among potentially vulnerable population groups, and compares the medical reasons for chat-based digital clinic contacts with those of traditional primary care.

Material and methods

We conducted an observational register-based study using nationwide data covering digital and traditional contacts to public and private primary care clinics in Finland from January 2022 to February 2025. Using a rich linked dataset, we examined several potentially vulnerable socioeconomic subgroups, including individuals aged 80 and above, those with multimorbidity, the unemployed, and members of the lowest income quintile. The study population comprised 1,599,588 individuals residing in eight wellbeing services counties in Finland.

Results

Public digital clinic users appeared to be younger, more often women, and more often live in urban areas, and have a lower prevalence of chronic illnesses compared with users of traditional public primary care. Potentially vulnerable patient groups also used digital services, although they used traditional services more. Medical reasons for digital clinic contacts were often relatively simple, but individuals with chronic illnesses also used digital services to address their care needs.

Discussion

We observed differences in the characteristics of users of digital and traditional primary healthcare services. We generated hypotheses for future research on equity in digital healthcare access. Further research is needed to evaluate whether improving access to digital services for vulnerable groups could support health equity.

KEY POINTS

• Differences were observed between users of digital and traditional primary healthcare services.

• Potentially vulnerable patient groups also used digital services, although traditional services appeared to remain more commonly used among these groups.

• A higher number of digital clinic contacts was observed among individuals with greater previous healthcare use

• Medical reasons for digital clinic contacts were often relatively simple, although individuals with chronic illnesses also used digital services to address their care needs.

• The findings highlight the need for further research on how digital clinics can be developed to support accessibility and equity across population groups.

• Future research should examine how digital tools may support continuity of care in primary healthcare.

Introduction

The use of synchronous and asynchronous digital messaging services has expanded rapidly in health care. As a result, chat-based digital clinics have become a fast and convenient gateway to primary care in many health systems. At the same time, the WHO and several national authorities have set ambitious goals for healthcare digitalisation [1]. Based on these objectives, digital health services are expected to improve access to high-quality care across all population groups [2,3]. Moreover, digital messaging services are expected to streamline access to healthcare in rural areas [4].

Digital services may provide a less costly alternative to traditional healthcare models [2]. However, improved access and lower barriers to seeking care may also generate additional demand and potentially increase the costs of health service provision [4]. Digital consultations are generally shorter and appear to produce similar in-consultation outcomes, measured by rates of diagnosis, prescriptions, specialist referrals, and patient satisfaction [3]. Digital consultations have been found suitable, especially in managing skin conditions, uncomplicated infections, mental health issues, and follow-up visits [5–10]. Yet, a key difference between traditional primary care consultations and digital services is that contacts in traditional services are more often associated with a higher prevalence of chronic conditions [3].

To date, digital health services in Finland have been predominantly developed and utilised within the private healthcare sector. Thus, expanding digital services to the public primary care sector, which serves multiple vulnerable patient groups, raises several questions about the equity of digital health care services and their ability to meet the needs of patient groups with complex or chronic conditions. According to several studies, the observed characteristics of digital clinic users differ from those of traditional public healthcare users. The users of digital clinics are generally younger, healthier, more often women, better educated, and more likely to live in urban areas than the users of face-to-face consultations [4,11–16]. At the same time, digital healthcare may not be accessible for all patient groups [17–19]. Lower socioeconomic status and higher morbidity have been associated as barriers to digital healthcare utilisation [14,20–23]. Additionally, insufficient digital or language skills, a lack of strong e-identification or suitable devices, and inadequate broadband service provision may hinder the use of telehealth [24,25]. Moreover, individuals with lower socio-economic status also have greater unmet healthcare needs [26].

The association between patients’ healthcare needs and digital clinic use has been examined in a few studies using large administrative datasets. In contrast to prior expectations, a recent Swedish study found that users of digital clinics appear to have greater healthcare needs than individuals who did not use the service and tend to seek care more frequently [15]. Simultaneously, in another Swedish study, the frequent attenders in digital primary care were found to be a heterogeneous group with fewer chronic conditions, while most contacts were due to temporary, less serious ailments [24]. However, evidence on the effectiveness of digital health services in managing chronic diseases and facilitating access for vulnerable patient groups remains limited.

Equity is a core dimension of healthcare quality assessment [27]. In digital health, equity frameworks highlight the importance of digital determinants of health – such as access to devices, digital literacy, and broadband infrastructure – and emphasise inclusive design, structured implementation, and continuous monitoring to avoid reinforcing health disparities [25,28–32]. Without systematic attention to equity, digital health tools may primarily benefit younger and healthier patients while offering less benefit to those most reliant on public primary care.

Equity considerations are also central to health policy decision-making. However, limited evidence on how the benefits of digital clinics are distributed across population groups restricts policymakers’ ability to make evidence-informed decisions and may slow both the adoption of digital clinics and the development of equitable digital solutions.

At the same time, digitally inclusive health system strategies may support the adoption of technology and the development of digital skills, particularly among recent adopters. Previous evidence suggests that exposure to digital services can lower adoption barriers and increase engagement with digital tools across domains, in line with digital inclusion frameworks [33–35].

This exploratory study describes the socio-demographic characteristics of digital clinic users, examines patterns of digital clinics use among potentially vulnerable population groups, and observes the medical reasons for chat-based digital clinic contacts in Finnish digital public primary care (PPC) clinics. Our focus is on potentially vulnerable socioeconomic subgroups, including individuals aged 80 and above, those with multimorbidity, non-native speakers, the unemployed, and members of the lowest income decile. We examine whether these groups use digital clinics and whether their reasons for contact appear to differ from those observed in the broader digital clinic user population.

Materials and methods

The Finnish health care system

The Finnish healthcare system is fragmented into multiple funding sources for first-contact care. These include the public healthcare system, partial reimbursements for visits to private clinics through the national health insurance system, and occupational healthcare involving employers [36]. The continuity of care in Finland has declined over time and is now lower than in other Nordic countries, with a COC-score of 0.28 in 2024 [37].

Public healthcare has been organised by 21 wellbeing services counties and the capital, Helsinki, since 2023. The boundaries of wellbeing services counties were primarily determined by existing administrative regions. Public health and social services centres offer primary healthcare services along with related social services, guidance, and counselling. When individuals require public primary care, they contact their local health centre. Traditionally, the initial contact is made by phone, where a nurse assesses the need for treatment. Following the assessment, the nurse either provides treatment independently or refers the patient to the appropriate professional at the most suitable time.

Access to first aid and emergency treatment must be provided immediately, and non-emergency medical care is available at health centres. Legally mandated time limits have been established to ensure guaranteed access to treatment [38]. The expansion of digital clinics in public healthcare is partly motivated by the need to comply with these time limits [39].

Alternatively, many employed individuals can choose to contact occupational health care, which offers faster access, less gatekeeping, and no copayments. The scope of occupational health services varies across employers. Employers are legally required to provide statutory preventive occupational health care, whereas curative services are optional. Statutory services include work-related health examinations, workplace assessments, support for work ability, and the evaluation of work-related conditions. The provision of curative care ranges from basic primary care consultations to specialist services, and some employers additionally offer insurance-based healthcare coverage. Additionally, people can also contact private clinics – general practitioners (GPs) or specialists – with or without voluntary private insurance, which involves much higher out-of-pocket costs but provides quicker access and less gatekeeping compared to public health centres.

Digital clinics in public primary care

The main function of the digital clinic is to enable patients to contact primary care through a chat-based service. In the chat, patients are initially connected with a nurse who can handle their concerns independently or, if needed, consult or refer them to the appropriate professional. In suitable cases, patients may be directed to an online physician within the same chat. Traditional contact methods, such as telephone or in-person visits, remain available.

To contact the digital clinic, the wellbeing services county’s website guides patients on the typical matters that can be managed through the digital clinic. However, it does not explicitly exclude any specific topics regarding when to contact. During the initial digital contact, most platforms use a standardised symptom checker, and after completing it, the nurse will assess the patient’s issue. The symptom checker is not mandatory; however, when available, it is often used to support nurses by providing standardised questions for common clinical problems and generating a clear summary of the patient’s complaint. If the patient’s problem cannot be addressed by the nurse or an online physician in the digital clinic, the patient will be scheduled for an in-person (F2F) appointment at their health centre or centralised urgent reception. Patient copayments in the digital clinics are the same as, or slightly lower than, those in traditional PPC.

The opening hours of digital clinics vary across regions, as do aspects of service provision like the platforms used, the level of centralised management, and whether services are fully or partly outsourced. From the patient’s perspective, however, the initial contact through the chat tool remains largely consistent and comparable across the clinics.

The data

We linked Finnish nationwide administrative registers containing individual-level data on contacts with primary health care (public, private, and occupational), morbidities based on special reimbursement rights for medicines, and socioeconomic and sociodemographic characteristics of the total population (not just patients), through unique person IDs (Figure 1). A contact includes all events that result in a note being added to the electronic health record system and may include indirect forms of interaction (e.g. prescription renewal and profession-to-professional consultations related to the patient) beyond the direct interaction with the patient. All contacts with digital clinics were included in the study. Guardians may access services on behalf of children under 12 years of age, in which case the contact is recorded in the child’s personal data. Adolescents aged 12–17 years may use the services independently if they have strong electronic identification credentials. The data span from January 2022 to February 2025.

Figure 1.

Retrieving and combining the data from the national registers.

The register containing primary care contacts is administered by the Finnish Institute for Health and Welfare [40]. Public healthcare providers are legally required to transfer data from their patient records to this national registry. For private providers, data transfer is voluntary; however, the largest private providers have been transferring data since 2021.

From the data, we extracted medical contacts, including care needs assessments, in-person visits, telemedicine contacts, and professional-to-professional interactions with nurses and physicians in public primary care, private outpatient care, and occupational healthcare. Within public primary care, we separate between digital clinics (mainly chat-based) and traditional health centres (mainly calls and in-person visits). In private outpatient care and occupational healthcare, we separate between in-person visits and telemedicine contacts (chat, telephone, and video) due to data limitations. Healthcare utilisation was defined based on contact dates: individuals received a value of 1 if they had any eligible contact on a given day. All diagnosis codes according to the International Statistical Classification of Diseases and Related Health Problems (ICD-10) were extracted for the study population from public primary care contacts (nurses and physicians), without restricting the type of contact.

The total population at the end of 2022, along with their socioeconomic and sociodemographic characteristics, is derived from the Statistics Finland FOLK and INFRA modules [41]. Based on these data, we also calculated family equivalised disposable income and straight-line distances to the nearest public health centre. We identified potentially vulnerable groups as follows: individuals aged 80 or older, those with a common chronic disease, insulin-treated diabetes and age over 40, multimorbidity, those in the lowest income quintile, individuals who experienced at least one period of unemployment during the year, and those who are not native Finnish or Swedish speakers. Frequent attenders were defined as patients who contacted their health centre ten or more times in the previous year [39].

Chronic conditions were identified based on special reimbursement rights for medicines at the end of 2023, as recorded by the Social Insurance Institution of Finland [42]. Patients with reimbursement rights were recognised for the three most common conditions: cardiovascular disease, diabetes, asthma, or COPD. Multimorbidity was defined as having reimbursement rights for at least two of the three most common groups. We also categorised insulin-treated diabetics over 40 years old, considering them more likely to have comorbidity.

Study population

To identify our study population, we first defined eligible target municipalities. We included municipalities where 1) the digital clinic was launched in 2023 or later, 2) there was at least a nine-month follow-up after the clinic’s launch (i.e. the launch occurred before June 1st, 2024), 3) there was no temporary pilot or closure of the digital clinic, and 4) the digital clinic service had a separate unit identifier (OID) allowing us to distinguish it from the traditional health centre.

In essence, the target areas include municipalities in eight different wellbeing services counties: Pirkanmaa, Kanta-Häme, East Uusimaa, South Ostrobothnia, North Karelia, parts of North Ostrobothnia, South Savo, and Central Finland. The study population was defined as the population that resided in these municipalities at the end of 2022. Overall, the study population covered 1,599,588 individuals. The observation period spanned 12 months before and 9 to 12 months after the launch of the digital clinic (up to a maximum of 365 days). Follow-up time varied between municipalities because all areas with available data were included to maximize sample size and power in the analysis, even when a full 12-month follow-up period was not available.

Statistical analyses

This study presents numerous means and differences in means. Our research is exploratory, not pre-registered, and aimed at generating hypotheses. Therefore, we do not do statistical testing. Instead, we provide relative (%) differences in means and standardised mean differences (SMDs) between users of the digital clinic and traditional face-to-face consultations, along with the means for the total population (Table 1). The relative difference in means was calculated using the following formula:

$$\mathit{\text{Relative}}\mathit{\text{difference}}\hspace{0.17em}\left(\mathit{\text{digiclinic}},\mathit{\text{traditional}}\right)=\frac{\mathit{\text{difference}}\mathit{\text{in}}\mathit{\text{means}}}{\mathit{\text{reference}}\mathit{\text{group}}\mathit{\text{mean}}}=\frac{x_{\mathit{\text{digi}}}-x_{\mathit{\text{traditional}}}}{x_{\mathit{\text{traditional}}}},$$

where $x_{\mathit{\text{digi}}}$ is the mean value among digital clinic users (Table 1, Column 1) and $x_{\mathit{\text{traditional}}}$ is the mean value among traditional PPC users (Table 1, Column 2).

Table 1.

The means, relative differences in means (difference %), and standardised mean differences (SMD) for the users of public digital clinics versus the users of traditional public primary care (PPC) clinics, including prior health care use in traditional PPC, private health care (HC) and occupational (occup.) HC, their sociodemographic characteristics, and the shares of potentially vulnerable people in different user groups. Total population refers to all residents in the region who have access to both traditional and digital healthcare services. Individuals were divided into two groups based on their use of public primary care after the launch of the public digital clinic: 1) digital clinic users who had used the public digital clinic at least once during the follow-up period after the launch, and 2) traditional public primary care (PPC) users who had at least one contact with traditional public primary care (including telephone contacts) but no contacts with the public digital clinic during the same period. Prior healthcare use was measured for 12 months before the digital clinic’s launch. The follow-up period ranged from 9 to 12 months, depending on the launch date of the digital clinic.

	Digital clinic users	Traditional PPC users	Total population
Individuals	162.298	694.963	1,599,588	Digital clinic users vs traditional PPC users
Digital clinic contacts	324.189	0	324.189
Traditional PPC contacts	725.094	3,484,215	4,209,309
	Mean	Mean	Mean	Difference (%)	SMD
A. Prior healthcare use (number of days with contact)
Traditional PPC contacts	4.33	4.68	2.91	−7.5	−0.05
Private HC: telemedicine contacts	0.16	0.09	0.11	+75.9	+0.11
Private HC: in-person visits	0.47	0.38	0.40	+24.7	+0.08
Occup. HC: telemedicine contacts	0.72	0.39	0.56	+81.9	+0.17
Occup. HC: in-person visits	0.69	0.41	0.59	+70.0	+0.17
B. Sociodemographic covarities
Age (in years)	38.67	49.58	43.56	−22.0	−0.45
Is female	64.6%	53.1%	50.3%	+21.7	+0.24
Language: Finnish	95.2%	94.0¤	93.2%	+1.2	+0.05
Language: Swedish	1.9%	1.9%	2.0%	+1.0	<0.01
Relationship or widowed	36.6%	45.9%	40.5%	−20.3	−0.19
Living in a city	70.4%	54.8%	62.4%	+28.4	+0.33
Distance to nearest clinic (km)	3.11	3.50	3.39	−11.1	−0.11
Tertiary education	30.2%	22.0%	26.3%	+37.3	+0.19
Pensioner	21.3%	42.5%	27.4%	−50.0	−0.47
In labour market	49.6%	34.2%	46.7%	+44.7	+0.31
Income (thousands of euros)	29.18	27.80	29.96	+5.0	+0.06
C. Potentially vulnerable groups
Insulin treated diabetes, aged 40+	6.2%	12.4%	7.4%	−49.8	−0.21
Common chronic disease	24.8%	36.8%	25.4%	−32.5	−0.26
Multimorbidity	4.4%	8.7%	5.2%	−49.4	−0.18
1st income quintile	20.5%	23.9%	20.9%	−14.2	−0.08
Unemployment spell	13.4%	9.5%	9.5%	+40.8	+0.12
Language: other	3.0%	4.1%	4.8%	−27.9	−0.06
Aged 80 or more	2.8%	9.8%	6.2%	−71.3	−0.29

The SMDs help compare differences in means between groups, adjusting for the underlying variability (standard deviation) in the data. Commonly, SMD thresholds of 0.2, 0.5, and 0.8 indicate small, medium, and large effect sizes [43]. However, these thresholds are often context-specific and do not necessarily imply clinical or social significance.

For the ICD-10 diagnoses, both primary and secondary diagnoses were included. The shares were computed separately for public digital clinics and traditional PPC clinics. Of the total population, the data was also stratified by age, gender, and wellbeing services county. Three-year age groups were used, and individuals aged 80 or older were combined into a single group. South Ostrobothnia and North Karelia were combined to ensure non-empty cells in each stratum. Each stratum included both digital and traditional contacts across both diagnosis categories. The prevalence of diagnoses was calculated separately within each stratum. The means were then averaged over all strata, with each stratum weighted equally. The follow-up period varied depending on when the digital clinic was launched.

Ethics and replication

This study relies on Finnish national administrative registry data without any contact with research subjects. Therefore, no patient consent or ethical review was required under Finnish legislation and research norms [44]. All individual-level data were pseudonymised and accessed solely through a secure remote system provided by Statistics Finland, in accordance with the European General Data Protection Regulation (GDPR).

The research data used in this study are governed by the Finnish Act on the Secondary Use of Health and Social Data (552/2019), which regulates the use of health and social care data for research purposes. Under this legal framework, access to pseudonymised administrative data is granted by the relevant data permit authority, subject to legal, data protection, and information security safeguards. In line with Finnish research integrity guidelines, a separate research ethics committee approval is not mandated for scientific studies that involve no direct contact with individuals, no interventions, and no special risks to participants (Finnish National Board on Research Integrity, 2019). The present study relies exclusively on previously collected register data and meets these criteria.

The individual-level data are sensitive and cannot be made public. However, interested researchers can apply for a data permit through Findata and Statistics Finland [45,46]. The replication codes, written in R, are available here: LINK WILL BE ADDED AFTER PEER-REVIEW. Artificial intelligence (AI) was utilised to ensure orthographic accuracy and to improve the fluency of the text.

Results

The characteristics of digital clinic users, traditional PPC users, and the total population are presented in Table 1. In the study population, 162,298 individuals (10% of the total population) contacted the digital clinic at least once during the study period. These individuals are referred to as digital clinic users. The digital clinic users had a total of 324,189 contacts with the clinic. During the same period, 694,963 individuals (43% of the total population) had no contact with the digital clinic but had at least one contact with traditional PPC. These patients are referred to as traditional PPC users.

Prior health care use

Considering prior healthcare use, users of the digital clinic had slightly fewer contacts in public primary care compared to traditional PPC users (−8%; SMD = −0.05). However, they had more previous contacts in other primary care sectors (private and occupational), with relative differences ranging from +25% to +82% (SMDs: +0.08 to +0.17), depending on the sector and contact type (telemedicine/in-person).

We further examined earlier primary care utilisation and current public digital clinic utilisation in Figure 2. Panel A in Figure 2 shows that individuals with a history of high healthcare use had a greater number of public digital contacts compared to those with previously low service use. Individuals who had used more private and occupational healthcare previously had more contacts with digital clinics. On average, individuals with no prior private or occupational healthcare contacts had 0.32 public digital contacts per person, whereas those with 10 or more prior contacts had 0.71 per person.

Figure 2.

The utilisation of public digital clinics by prior healthcare utilisation. Panel A presents the number of digital clinic contacts per capita, and Panel B shows the share of digital clinic contacts among all public primary care contacts in each group. In both panels, the data is grouped by the number of health care contacts in the previous year. Individuals were grouped according to their healthcare utilisation, measured as the number of contacts during the 12 months preceding the launch of the public digital clinic (x-axis). In Panel A, the y-axis shows the average number of public primary care (PPC) digital clinic contacts per capita, whereas in Panel B, the y-axis shows the proportion of public digital clinic contacts among all PPC contacts.

However, individuals with more prior contacts in public primary care appeared to have a smaller proportion of digital clinic use among all PPC contacts after the introduction of the digital clinic (Figure 2, Panel B). In contrast, individuals with previous use of private and occupational healthcare appeared to have a higher proportion of digital clinic use within overall PPC contacts. Among individuals with no prior contacts in traditional PPC, digital clinic contacts accounted for 12% of all PPC contacts, whereas among those with ten or more prior public contacts, the corresponding proportion was 4.2%. Of the total 324,189 digital clinic contacts (Table 1), 59,839 contacts (Figure 2) were made by individuals with ten or more prior public primary care contacts during the previous year, corresponding to 18% of all digital clinic contacts.

Socio-demographic characteristics of digital clinic users

Digital clinic users differed from traditional PPC users in several sociodemographic characteristics (Table 1, Panel B). On average, digital clinic users were younger (mean age 39 vs. 50 years) and more often women (65% vs. 53%). Most were native Finnish or Swedish speakers (97.1%). Compared to traditional PPC users, digital clinic users were less often in a relationship or widowed (37% vs. 46%), more frequently had a tertiary education (30% vs. 22%), and more often lived in urban areas (70% vs. 55%). They also lived slightly closer to the nearest public health centre (mean 3.1 km vs. 3.5 km). Regarding socioeconomic status, digital clinic users were less likely to be pensioners (21% vs. 43%), more often participated in the workforce (50% vs. 34%), and had slightly higher income levels.

In absolute value, the largest SMDs (>0.20) between digital clinic users and traditional PPC users were observed for being a pensioner (−0.47), age (−0.45), living in a city (+0.33), being in the labour market (+0.31), and being female (+0.24). The labour market and pension statuses are naturally strongly correlated with age.

Potentially vulnerable groups in digital clinics

When examining potentially vulnerable groups (Table 1, Panel C), digital clinic users generally appeared healthier than traditional PPC users. The proportion of insulin-dependent diabetics aged 40 or older was lower among digital clinic users compared to traditional PPC users (6% vs. 12%), as was the prevalence of any common chronic disease (25% vs. 37%) and multimorbidity (4% vs. 9%). These differences are likely partly explained by differences in age profiles. The share of individuals in the lowest income decile was smaller in the digital clinic group (21% vs. 24%), although unemployment spells were slightly more common (13% vs. 10%). The share of non-native speakers was lower than their proportion in the total population among both traditional PPC users (4%) and digital clinic users (3%). The proportion of people aged 80 or older was lower among digital clinic users compared to traditional PPC users (3% vs. 10%). However, this group still includes 4544 individuals aged 80 or older who used the digital service.

In absolute terms, the largest SMDs (> 0.20) between digital clinic users and traditional PPC users were observed for being aged 80 or more (−0.29), having a common chronic condition (−0.26), and being an insulin-dependent diabetic aged 40 or more (−0.21).

Figure 3 illustrates public digital clinic utilisation in potentially vulnerable groups in more detail. In two groups, the number of public digital clinic contacts per person was much higher than in the general population. These included the unemployed (0.34 vs. 0.20) and the frequent users (0.45 vs. 0.20; those with 10 or more PPC contacts before the launch of the digital clinic). At the same time, in some vulnerable groups, the number of digital contacts per person was much lower than in the general population, particularly among non-native speakers (0.12 vs. 0.20) or those aged 80 or more (0.09 vs. 0.20).

Figure 3.

The utilisation of public digital clinics by potentially vulnerable patient groups. The left y-axis (yellow columns) shows the number of public digital clinic contacts per person, while the right y-axis (purple line) displays the proportion of digital clinic contacts out of all PPC contacts compared to the total population (green column).

When examining the proportion of digital use of all PPC use, unemployed individuals had a higher share of digital contacts compared to the overall population (10% vs. 7%), and non-native speakers had an almost equal share to the total population. In contrast, in other vulnerable groups, the share of digital use was significantly lower than in the general population. For instance, the shares were 1.6% for those aged 80 or older and 2.4% for multimorbid individuals.

Reasons for visits in digital clinics

The 10 most common diagnoses accounted for 12,852 diagnoses. In digital clinics, 9.7% of all diagnoses were prescription renewals, followed by unspecified counselling (6.6%) (Figure 4). In comparison, these accounted for only 1.5% and 1.4% of diagnoses in traditional PPC. Other frequent digital clinic diagnoses included conjunctivitis (5.3%), urinary tract infections (4.6%; UTIs), and hypertension (3.6%) (Figure 4, Panel A). In contrast, conjunctivitis and UTIs were relatively uncommon in traditional PPC (0.7% and 0.9%).

Figure 4.

All individuals: the 10 most common diagnoses (ICD-10) in public primary care digital and traditional PPC clinics, and their adjusted shares of all diagnoses.

In traditional PPC, encompassing 742,097 diagnoses, the most common condition was upper respiratory tract infection (3.4%), a rate comparable to digital clinics (3.1%). Hypertension (3.2%) and pain-related diagnoses affecting the limb, abdomen, or back (2.4%, 2.4%, and 1.8%) were also frequent (Figure 4, Panel B). While pain-related diagnoses were less common in digital clinics, the proportions of hypertension and upper respiratory tract infections were rather similar across both modalities. Overall, the diagnostic distribution in traditional PPC appeared broader, with the leading diagnosis representing only 3.4% of all cases.

For the most frequent users, the distribution of the 10 most common diagnoses in digital clinics was rather similar to that of the overall population (Appendix A). We also examined a pooled group of unemployed individuals, those in the lowest income quintile, and non-native speakers, and found that their diagnosis patterns closely resembled those of the total population (Appendix B). In contrast, individuals with common chronic diseases showed a somewhat different profile, with diagnoses related to chronic conditions ranking among the ten most common diagnoses, setting them apart from the total population (Figure 5).

Figure 5.

The figure reports for the 10 most common ICD-10 diagnoses in public primary care (PPC), as well as their shares relative to all observed diagnoses. The figure is restricted to individuals with a common chronic disease, defined based on special reimbursement rights for medicines. Both primary and secondary diagnoses are included. We compute the shares separately for public digital clinics and traditional PPC clinics. This figure does not account for differences in observed background characteristics.

Discussion

Main results

This study provides a descriptive overview of digital clinic users in Finnish public healthcare and identifies patterns in the use of digital and traditional primary care services. Differences were observed between digital clinic users and traditional primary care users across several sociodemographic characteristics. Individuals with greater prior healthcare use also appeared to have a higher number of digital clinic contacts.

Individuals in potentially vulnerable groups seem to use digital services, although their share of digital clinic utilisation appears lower than in the overall population. Additionally, many digital contacts appeared to concern single or relatively simple issues. Individuals with chronic conditions also used digital services in contacts related to chronic illness management. As this study was descriptive and exploratory in nature, the findings should be interpreted primarily as observations that may help generate hypotheses for future research on digital healthcare utilisation and equity.

Considering socio-demographic characteristics, digital clinic users were generally younger, more often women, and from more urban areas compared to traditional PPC users, which aligns with previous literature [2,6,12–15]. Individuals with more frequent prior healthcare contacts also appeared to use public digital clinics more often, consistent with findings from a recent Swedish study [15]. In this dataset, higher use of public digital clinics was observed among individuals with previous contacts in private and occupational healthcare. Swedish studies have reported similar patterns in primary care digital clinic use, although the services examined there have mainly been privately provided [12,15].

This study also describes patterns in previous healthcare use and the proportion of digital service use. Individuals with more frequent prior public healthcare use had a lower share of digital clinic use within their overall PPC contacts, whereas those with more frequent prior use of private and occupational healthcare had a higher share of digital clinic use. One possible explanation is that individuals using private and occupational healthcare may already be more familiar with digital contacts, as digital platforms were introduced in these sectors earlier than in public healthcare. However, this study cannot determine whether public digital clinic contacts replace contacts in other healthcare sectors or represent additional use alongside existing services.

Previous literature has suggested that exposure to digital services may lower barriers to technology adoption and increase engagement with digital tools across domains [34,35]. This may partly explain why individuals already accustomed to digital services also appear to use digital healthcare services more often. Overall, the findings indicate that individuals with more frequent healthcare contacts also engage with digital services, although further research is needed to understand how these patterns relate to continuity of care. Our findings indicate that vulnerable groups do use digital services. At the same time, these groups generally have a higher overall need for services, especially in traditional PPC, resulting in a smaller relative share of digital contacts among their total PPC use. These findings are somewhat in contrast with the previous literature, which suggests that vulnerable groups use digital services much less than the average population [13,14,19].

We observed that people aged 80 and older had fewer digital contacts per person than the general population. This finding aligns with the previous literature [18,23]. Moreover, the use of digital clinics by non-native speakers was low, but the share of digital contacts out of all PPC contacts was in this group nearly as high as in the total population. These findings may support future research on whether digital tools and translation systems could improve accessibility for non-native speakers. Interestingly, the proportion of unemployed individuals was higher among digital clinic users. However, this might reflect the fact that young adults tend to use digital clinics frequently, have, on average, strong digital skills, but are overrepresented among job seekers.

Overall, individuals in vulnerable groups may have higher overall service needs through traditional channels. Continued reliance on traditional services may also have implications for continuity of care, although continuity was not examined in this study. After all, the first-phase digital clinics examined in the study were not primarily designed to promote continuity of care. Instead, they prioritised rapid response times, with professionals matched with clients sequentially from a queue, regardless of individual characteristics.

The most common medical reasons for digital clinic contacts appeared to involve relatively simple issues, which aligns with previous literature on telemedicine visits [3,5,6]. The digital clinic contacts were more often addressed by patients needing prescription renewals, counselling contacts without a specific diagnosis, and women with urinary tract infections. In contrast, traditional PPC diagnoses seemed to be spread across a broader spectrum, with the most common diagnosis accounting for only 3% of all diagnoses in traditional PPC, primarily focusing on symptoms that require examination, such as various pain complaints. Overall, the diagnoses treated at digital clinics appeared broadly consistent with previous descriptions of telehealth use, mainly through telephone contacts, even before the establishment of digital clinics [3,6]. The feature of sending photos via chat could have increased the digital treatment of conditions like conjunctivitis and skin problems, which is reflected in the most common diagnoses at digital clinics.

Nevertheless, we observed that individuals with chronic illnesses do engage with digital services to some extent in managing their conditions. Their distribution of diagnoses treated in digital clinics differs from that of the general population, with more chronic illnesses appearing among the top 10 diagnoses, and a larger share of prescription renewals than in the overall population. Some of these prescription renewal diagnoses could be associated with chronic illnesses if the diagnoses were more specific. To our knowledge, this finding has not yet been reported in previous studies. Based on ICD-10 codes, we do not know the specific issues, such as those related to diabetes, that are treated at digital clinics. Chronic conditions have traditionally been managed primarily through face-to-face consultations [5], but follow-up visits for these conditions have also been found suitable for digital contacts [5,6].

Strengths and limitations

The strength of this study lies in a rich and comprehensive dataset, which includes primary care records, socioeconomic characteristics, and a large study population. We were able to obtain 9–12 months of follow-up data from 8 different wellbeing service counties, covering over 1.5 million residents in the area, and collected over 300,000 digital clinic contacts. Another strength of this study is that we were able to examine multiple digital clinics across various regions. We also assessed service use in both the public and private sectors, as well as in occupational healthcare. This broad coverage increases the generalisability of our findings and allows comparisons across different healthcare sectors.

This study was based on data obtained from electronic health record (EHR) systems in the participating regions. Data entry practices differ between EHR systems, which may result in variation in data quality and completeness and may affect the analyses. Although public healthcare providers are legally required to submit patient record data to the national registry, the level of detail recorded in the registry varies. This variation appears to be related to differences in the EHR systems used across Wellbeing Services counties.

The diagnoses presented only cover part of the use of digital clinics, which are mainly nurse-led. Nurses do not record ICD-10 diagnoses, so our data reflects contacts with physicians. There is also regional and temporal variation in diagnosis coverage. The limited time professionals spend on documentation affects the thoroughness of records and diagnoses. Regarding chronic conditions and multimorbidity, only a subset of patients could be identified in this dataset, as not everyone with the diagnosis has the special reimbursement right. Thus, the criteria for special reimbursement for medicines are not identical to the clinical criteria for initiating treatment; not all patients with the condition are found in the special reimbursement register. Nonetheless, this approach was the most reliable means available for capturing part of this patient group.

In addition, diagnoses of chronic illnesses have been retrospectively submitted to the national database from some of the wellbeing services counties, at least during 2025, in an erroneous manner. This has resulted in diagnoses of chronic diseases appearing as contact diagnoses for patients already with a chronic diagnosis. There may be faulty diagnosis data from some wellbeing services counties in our study, but we note that the issue would also apply to traditional PPC.

The low continuity of care in the Finnish healthcare system may have influenced the use of centralised digital platforms in Finland, and the digital clinic user profiles might vary in countries with better continuity of care in primary healthcare.

Implications for practice

Digital clinic users were more often younger and healthier individuals with less complex health issues, whereas users in traditional PPC were older and less healthy. This may have implications for increased workload in traditional PPC. In addition, higher use of digital clinics was observed among individuals with high prior use of private healthcare and occupational health services. However, this study cannot determine whether service use is shifting from other modalities to digital clinics or whether digital clinic use is replacing previous care, indicating a need for further research on substitution patterns between different care modalities.

This study examines the initial phase of implementing public digital clinics. During the first year following their launch, these clinics reached only a small segment of the overall population in Finland. At this stage, it remains unclear how the coverage of digital clinics will evolve. Furthermore, the role of digital contacts in comprehensive patient care is not yet fully understood, nor is the extent to which observed patterns are related to the system itself or to its fragmentation. When designing digital services, it is important to consider the needs of potentially vulnerable groups carefully. Ensuring continuity of care is critical, as patients who interact with multiple services are also more likely to engage with digital platforms, which may increase the risk of fragmented care among those who need coordinated support the most.

Hypotheses for future research

Based on these findings, several hypotheses can be proposed for future research. It is possible that centralised digital clinics, on the other hand, provide effective ways to deal with simpler issues for healthier patients, but on the other hand, increase fragmentation of care, which may harm patients with multiple conditions and needs. Good continuity of care may decrease or modify the need to contact centralised digital platforms, particularly among patients with high healthcare utilisation. Digital platforms could support continuity of care and reduce the risk of fragmented care when integrated into existing care pathways, for example, by enabling interactions with familiar healthcare professionals. Future studies are needed to test these hypotheses and to examine how different models of digital service delivery influence continuity and care integration.

Conclusion

To date, digital clinics in Finland appear to be well adopted by patients, but they still account for a small share of all healthcare contacts. This study suggests that patterns of digital service use differ from traditional service provision, which may have implications for equity. As the introduction of digital services is still at an early stage, continued monitoring is needed to understand how these patterns develop and to ensure access to care across population groups.

From an equity perspective, it is encouraging that users of digital clinics include potentially vulnerable groups, as well as individuals with chronic conditions. However, the centralised way of organising the service may pose a risk of fragmentation. This may be particularly relevant for patients with higher healthcare utilisation and those who may have greater needs, for continuity of care.

Digital tools could be leveraged to support continuity of care. It is important to ensure that digital service options are accessible, safe, and fair, so that no group is excluded from seamless access to care. Additionally, there will remain individuals who are unable to use digital channels. Re-assessing the role of digital clinics within PPC services is important to ensure that their benefits are not limited to younger, healthier individuals.

AppendicesAppendix A. The 10 most common ICD-10 diagnoses among frequent users, compared between digital clinics and traditional public primary care, and their shares of all diagnoses

Appendix B. The 10 most common ICD-10 diagnoses compared between digital clinics and traditional public primary care among combined vulnerable groups – the unemployed, bottom income quintile, and non-native speakers – and their share of all diagnoses

Funding Statement

This work was supported by the Ministry of Social Affairs and Health, Finland, through the SoteDataLab project, and by the Strategic Research Council at the Research Council of Finland (GAINS, decision number 372591).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

This study relies on Finnish national administrative registry data without any contact with research subjects. Therefore, no patient consent or ethical review was required under Finnish legislation and research norms [37]. All individual-level data were pseudonymised and accessed solely through a secure remote system provided by Statistics Finland, in accordance with the European General Data Protection Regulation (GDPR). The individual-level data are sensitive and must not be made public. However, interested researchers can apply for a data permit through Findata and Statistics Finland [38,39]. The replication codes, written in R, are available here: LINK WILL BE ADDED AFTER PEER-REVIEW.