Individual Patient Factors Associated with the Use of Physical or Digital Primary Care in Sweden

Background

Primary care has traditionally been a physical visit at a primary health care center. However, telemedicine using telephone contacts, photography, and video visits has been used as a complement to physical care for several decades.^1–3 In recent years, telemedicine using digital communication technologies (digital care) has become more widely used in primary care.^4–6 This accelerated during the COVID-19 pandemic.^7–9

The adoption of digital care is concentrated to specific groups in the population; patients that use digital care are younger, female to a larger extent, and have higher socioeconomic status than patients using physical health care,^4,10,11 although this may have changed during the COVID-19 pandemic.^12,13 Factors such as travel time and geographical distance, cost, previous use of digital care, and quality of care seem to influence patients’ choice of seeking physical or digital care.^11,13–16

The differences in adoption of digital care has raised the question of how individual preferences affect patients’ choice of seeking digital care over the traditional physical alternative. Previous studies focusing on patient preferences have only included patients that have used one modality or the other.^15,16 There is a lack of studies comparing patients seeking physical or digital care with regard to what patients consider important when making their choice.

To better understand the drivers that influence the choice of seeking physical or digital care, this study aimed to compare self-rated health, internet habits, and what patients deem important when choosing health care between users of physical and digital primary health care, and to determine if these characteristics were associated with seeking either physical or digital care. To our knowledge, this is the first study comparing these factors between physical and digital health care in a primary care setting.

Methods

In this cross-sectional study, we collected questionnaire data from primary care patients in Stockholm, Sweden, during October 2020 to May 2021. The study was approved by the Swedish Ethical Review Authority (Dnr: 2020-00860 and amendments 2020-06506 and 2021-04602). The guideline for Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) was followed.

SETTING

The Swedish primary health care system is publicly funded, and patients pay only a small fee for each visit (0–250 Swedish kronas [SEK], equivalent to about 0–24 United States dollars [USD]), up to an annual total of 1,150 SEK (110 USD), above which all visits are free for the patient. Patients are listed with one physical primary health care center in their region of residence, but are free to visit other primary health care providers.

During recent years, providers that offer direct-to-consumer patient-initiated telemedicine outside the listing system have entered the Swedish primary care system. In 2020, they constituted 11% of all primary care visits in Sweden.¹⁷ These providers offer both synchronous communication (such as video visits and text messaging where both participants are connected through chat system simultaneously) and asynchronous communication (such as text messaging where participants may have a continuous conversation, but need not be connected to the messaging system simlutaneiously). Different providers promote different types of communication. They also have access to patient records through a national health information exchange infrastructure, through which the majority of Swedish electronic health records can be accessed.¹⁸

Participants were recruited at six physical primary health care centers (physical care) and four providers of digital primary health care (digital care). The physical primary health care centers covered a large portion of Region Stockholm’s geographical area with widespread socioeconomic status. For digital care, three of the largest providers of digital care in Sweden participated in the study. In addition, Region Stockholm’s own system for digital care participated.

RECRUITMENT

Patients were consecutively recruited in connection to a visit with the participating health care providers. Participants responded to the study questionnaire either on paper, or through a secure online survey system.

Inclusion criteria were: (1) ≥18 years of age and (2) having visited a participating primary care provider during the study period either as patient, as guardian of a minor, or as a carer of someone else. Exclusion criteria were being unable to give consent. Patients whose consultations were due to COVID-19 were also excluded to minimize the impact of the ongoing pandemic on the results.

QUESTIONNAIRE

The questionnaire contained questions on sociodemopgrahics, self-rated health, internet habits (Supplmentary Table S1). Questions about self-rated health and internet habits were derived from public surveys.^19,20

We also asked what participants deemed important when seeking health care for similar reasons as their current visit. Participants were asked to choose up to 5 of the 16 predefined response alternatives, including “none” and “other” (Table 1). The response alternatives were based on factors that have been shown to affect the choice of using physical or digital care in previous studies, such as accessibility, waiting time, competence, cost, and geographical proximity.^11,14–16 In addition, relational continuity (i.e., seeing the same health care professional every time) and informational continuity (i.e., health care professionals having access to previous patient records) were included.²¹ We also added alternatives that were based on the research team’s clinical experience from working with digital care, such as avoiding leaving home, avoiding waiting areas, and patients’ own access to digital health care records.

More than 5 response alternatives were chosen by 426 respondents (16%). The majority of those (58%) had chosen one or two extra alternatives. Excluding these, respondents had limited effects on our results and therefore, all observations were included in the analysis.

To evaluate the questionnaire we performed cognitive interviews. Seven adults commented on the questionnaire, its structure, the questions, and their wording, as well as the written instructions while completing the questionnaire. The evaluation only led to minor clarifications.

ANALYSIS POPULATION

In total, 2,182 patients who had visited a physical care provider (physical users) and 2,206 patients who had used digital care (digital users) provided informed consent and responded to the questionnaire. For physical users, the response rate was 42%, based on the number of questionnaires that were distributed. The total number of patients asked to participate was not recorded, nor reasons for nonparticipation. Among digital users, 41% of those who clicked on the link to the invitation message to participate in the study responded to the questionnaire. The number of patients who viewed the inviation without clicking the link was not recorded.

In this study, we limited our comparison to respondents living in Region Stockholm who had met with a physician during their visit. We also limited the study to patients who had scheduled the appointment themselves; patients who had been scheduled for a visit by their care provider were not included. Furthermore, we excluded respondents who were at the health care visit as a guardian of a minor or as a carer of another adult. In total, 1,669 respondents were excluded (Supplementary Table S2).

Participants were categorized into four age groups based on quartiles: 18 to <38, 38 to <51, 51 to <64, and ≥64 years. Education was categorized into ≤12 or >12 years. For self-rated health, we combined the two response alternatives, “bad” and “very bad” self-rated health into one category as <1% rated their health as very bad. Participants were categorized as either working, retired, or other (e.g., studying, unemployed). Furthermore, respondents were categorized as having access to an internet device if they had reported access to at least one of the following: smartphone with applications, computer, or tablet. They were categorized as daily internet users if they reported using internet daily on at least one of these devices.

STATISTICAL ANALYSIS

Descriptive statistics were reported as n (%) for all categorical variables. Differences in distribution between physical and digital care were compared using χ²-tests. We used multivariable logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs) of seeking digital care. An OR >1 indicates that the characteristic was associated with higher odds of seeking of digital care, whereas an OR <1 indicates lower odds of seeking digital care, that is, the characteristic is associated with the use of physical care.

We fitted three different models. Model A included sociodemographic characteristics, self-rated health, and internet habits. Model B included what respondents deem important when seeking health care, where each response alternative was treated as an independent dichotomous variable as we considered it possible to have contradictory preferences. Model C included the variables from model A and B. Models were stratified by age groups for sensitivity analysis.

The p-Values <0.05 were considered statistically significant. Data were analyzed using Stata 15.1 (Stata Corporation. College Station, TX, USA).

Results

Our analysis population consisted of 2,716 participants (1,414 physical users and 1,302 digital users). The mean age was 51 years and 67% were female (Table 2). The majority (67%) worked and 60% had >12 years of education. Overall, 70% of physical and 74% of digital users rated their health as good or very good. Less than 8% reported having a bad or very bad health. Almost all respondents reported access to one or more internet devices (99%) and daily internet use was high, which is 98%.

IMPORTANT WHEN SEEKING HEALTH CARE

Competence of the health care professional was rated as important when seeking health care most often by both physical and digital users, but it was rated as important by a larger proportion of physical than digital users (90% vs. 78%, p < 0.001). More than 50% of both physical and digital users also rated convenient contact with the care provider and the attitude and demeanor of health care professionals as important when seeking health care. Digital users rated having digital access to care, avoiding leaving home, after-hour appointments, and avoiding waiting areas as important more often than physical users. Conversely, physical users rated relational and informational continuity, attitude, and demeanor of health care staff, and proximity as important more often than digital users. The cost of the appointment and access to electronic health records were rated as important by <10% of both physical and digital users. Results are presented in Table 3.

REGRESSION ANALYSIS

In model A, having access to internet and using internet daily were associated with seeking digital care (OR 5.39, 95% CI 1.17–24.77, and OR 3.21, 95% CI 1.30–7.90, respectively) (Table 4). Very good self-rated health, compared with average self-rated health, was also associated with seeking digital care (OR 1.31, 95% CI 1.01–1.69).

In model B, considering it important to receive care without leaving home was the factor associated with the highest OR of seeking digital care (OR 31.95, 95% CI 13.79–74.01). To avoid waiting areas (OR 4.13, 95% CI 2.61–6.53) and having access to care outside office hours (OR 4.05, 95% CI 2.99–5.50) were also associated with increased ORs of seeking digital care. Conversely, patients who reported relational or informational continuity as important more often sought physical care (OR 0.23, 95% CI 0.18–0.30, and OR 0.70, 95% CI 0.55–0.88, respectively). Competence of health care staff, which was the response alternative rated as important most often by both physical and digital users, was associated with seeking physical care (OR 0.54, 95% CI 0.39–0.74).

When combining model A and B in model C, only university education (OR 1.52, 95% CI 1.20–1.92) and daily internet use (OR 5.64, 95% CI 1.58–20.07) remained significant from model A. All factors from model B remained significant in model C. Despite the increased number of covariates in the model, the direction and magnitude of the coefficients remained stable. In sensitivity analysis in which the models were stratified by age group, the results were overall robust (data not shown).

Discussion

We found that patients seeking digital care considered themselves to be healthier and reported having access to and using the internet more often than patients visiting physical primary care. Important factors when seeking health care for both physical and digital care users were, the competence of the health care staff, the attitude and deamenor of the health care staff, and to easily and quickly get in touch with the health care provider. What patients consider important was associated with what type of care they chose to seek. Valuing relational continuity (i.e., seeing the same health care professional every time) and informational continuity (i.e., professionals’ knowledge about patients’ medical history) was associated with seeking physical care. Considering it important to have access to care without leaving home, avoid waiting areas and have access to care outside office hours was associated with seeking digital care. Using internet daily was also associated with higher odds of seeking digital care, compared with not using internet daily.

Our study supports results from previous studies that have reported that factors such as age, education, and internet access are associated with the likelihood of seeking either physical or digital health care.^4,11,22 While we assessed self-reported daily internet usage, previous studies have used information about residential internet access.¹¹ However, sociodemographic factors had little explanatory value for choice of physical or digital care when taking into account what patients deem important. This may indicate that factors such as age and education are not drivers for the choice of physical or digital care by themselves, but rather covariate with what patients perceive to be important when seeking health care.

While previous studies have indicated that cost may be important when choosing between physical or digital care,^11,14,15 this was not the case in our study; <10% of both physical and digital users rated cost as important. This may be due to the fact that health care in Sweden is publicly funded independent if the appointment is physical or digital, public or private. Previous studies have also identified long distance and travel time as factors associated with choosing digital care or other types of telemedicine over physical appointments.^11,22 However, geographical proximity and minimizing travel time were only rated as important by 21% of respondents in our study.

The results of our study may help policy makers in two ways. First, it offers concrete examples of how both physical and digital care can improve to better meet patients’ preferences. For example, physical users value both relational and informational continuity from their provider, but both of these can also be achieved using digital care. Sweden has a national health information exchange infrastructure enabling all care providers to access the majority of all Swedish electronic health records through one unified view,¹⁸ enabeling informational continuity for the patient. Similarly, physical care centers could also be able to offer patients’ care outside office hours. Physical care without leaving home, that is, home visits, already exist for selected groups, such as elderly and those with different types of handicaps or terminally ill patients, where leaving home is impossible.²³ This solution could be made available to other groups of patients as well.

Second, our results indicate that what patients deem important when seeking health care has an impact on what type of care they seek. This impact seems to be larger than that of sociodemographic factors. As such, our results complement previous studies and inform policy makers of the importance of considering patient opinion rather than assuming that this can be predicted through sociodemographics. As use of digital care may be a way of reducing health care costs, it is important it is adapted to patient needs and preferences.²⁴

STRENGTHS AND LIMITATIONS

A strength of our study is the large number of respondents from both physical and digital primary care. Previous studies, apart from register-based studies, have primarily been interview studies with <20 participants,^15,25,26 or have only included users from either physical or digital care.^16,27–29 We recruited both physical and digital users simultaneously to increase comparability. Furthermore, we recruited participants from multiple health care providers, covering different geographical and sociodemographic areas of Region Stockholm.

A limitiation of our study is that we do not know how many patients that were invited to participate from the providers of digital care since different methods of screening patients for participation were used. Furthermore, only the number of distributed questionnaires was recorded for physical providers, and we do not know the actual number of patients that were asked to participate.

Another limitation of the study is that we do not know the reason for the patients’ care visit. Previous studies have shown that comorbities and reasons for visit differ between physical and digital care.^12,22,30 Nonetheless, the purpose of this study was to investigate what made patients choose physical or digital care. Studies also show that patients feel competent to choose the type of care they need, based on the health problem they seek care for.³¹ There may be many reasons to prefer either type of care, such as having old, sick, or immunocompromised relatives at home. Although we could not include all possible reasons in the response alternatives, we also gave the respondents the possibility to give free text answers.

During spring 2021, vaccines for COVID-19 were introduced in Sweden. Despite this, the national recommendations to avoid contact with other people remained in place until July 2021. As such, we doubt that the introduction of vaccines had any major impact on visiting patterns to physical or digital primary care during the study period.

We chose to only include visits to physicians done by patients living in Region Stockholm that had booked their visit themselves to ensure a fair comparison between physical and digital care. This may, however, limit the generalizability of our results as patients from Region Stockholm may differ from that of other parts of Sweden in that they have had a faster uptake of digital care.^4,17

Conclusion

Considering it important to receive care without leaving home, and access to care outside of office hours were associated with choosing digital care. Valuing seeing the same health care professional every time and health care professionals’ knowledge of the patient’s medical history were associated with choosing physical care. What patients considered important when seeking health care was associated with the form of care they chose also when adjusted for self-rated health and sociodemographic characteristics. This should be considered when planning health care to optimize resource allocation.

Authors’ Contributions

D.S. was responsible for data acquisition, analysis, and interpretation of the data, and drafted the article. S.E.B. participated in data acquisition, analysis, and interpretation of the data, and was a major contributor in writing the article. L.S., A.D., and I.M. participated in data acquisition and analysis. I.A.W., B.C.B., N.F., H.H., M.T., and M.H. contributed to the conception of the study and interpretation of data. Y.T.L. was responsible for the conception and design of the study, participated in the acquisition, analysis, and interpretation of the data, and was a major contributor in writing the article. All authors contributed to the critical review and revision of the article. All authors read and approved the final version of the article.

Data Availability Statement

The datasets from this study are not publicly available due to ethical restrictions, but are available from the corresponding author upon reasonable request.

Disclosure Statement

I.A.W. and N.F. have previously worked at two of the participating providers of digital care. No other disclosures were reported.

Funding Information

This work was supported by the Swedish Research Council [2021-006426]; Region Stockholm [2018-1076]; and AFA Försäkring [190210].

Supplementary Material

Supplementary Table S1

Supplementary Table S2