This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
The literature has noted the need to use more advanced methods and models to evaluate physicians’ outcomes in the shared health care model that electronic health (eHealth) proposes.
The goal of our study was to design and evaluate a predictive multidimensional model of the outcomes of eHealth usage by European physicians.
We used 2012-2013 survey data from a sample of 9196 European physicians (general practitioners). We proposed and tested two composite indicators of eHealth usage outcomes (internal practices and practices with patients) through 2-stage structural equation modeling. Logistic regression (odds ratios, ORs) to model the predictors of eHealth usage outcomes indicators were also calculated.
European general practitioners who were female (internal practices OR 1.15, 95% CI 1.10-1.20; practices with patients OR 1.19, 95% CI 1.14-1.24) and younger—aged <35 years (internal practices OR 1.14, 95% CI 1.02-1.26; practices with patients OR 1.32, 95% CI 1.13-1.54) and aged 36-45 years (internal practices OR 1.16, 95% CI 1.06-1.28; practices with patients OR 1.21, 95% CI 1.10-1.33)—had a greater propensity toward favorable eHealth usage outcomes in internal practices and practices with patients. European general practitioners who positively valued information and communication technology (ICT) impact on their personal working processes (internal practices OR 5.30, 95% CI 4.73-5.93; practices with patients OR 4.83, 95% CI 4.32-5.40), teamwork processes (internal practices OR 4.19, 95% CI 3.78-4.65; practices with patients OR 3.38, 95% CI 3.05-3.74), and the doctor-patient relationship (internal practices OR 3.97, 95% CI 3.60-4.37; practices with patients OR 6.02, 95% CI 5.43-6.67) had a high propensity toward favorable effects of eHealth usage on internal practices and practices with patients. More favorable eHealth outcomes were also observed for self-employed European general practitioners (internal practices OR 1.33, 95% CI 1.22-1.45; practices with patients OR 1.10, 95% CI 1.03-1.28). Finally, general practitioners who reported that the number of patients treated in the last 2 years had remained constant (internal practices OR 1.08, 95% CI 1.01-1.17) or increased (practices with patients OR 1.12, 95% CI 1.03-1.22) had a higher propensity toward favorable eHealth usage outcomes.
We provide new evidence of predictors (sociodemographic issues, attitudes toward ICT impacts, and working conditions) that explain favorable eHealth usage outcomes. The results highlight the need to develop more specific policies for eHealth usage to address different realities.
In recent years, advances in health information technologies (HITs), electronic health records (EHRs), personal health records (PHRs), electronic health (eHealth) and telehealth applications and devices, and health information exchange (HIE) systems have profoundly transformed professionals’ health care practice, thus, contributing to the efficiency of their activities by reducing errors, improving early diagnosis of diseases, and offering better management of chronic diseases [
The application of eHealth services to primary care is of particular interest because it is at this level of care where patients come into regular contact with health care systems. Primary care also provides the highest volume of health services [
While there is considerable evidence in the literature about the predictors of some particular uses of eHealth, attention has recently been drawn to the need to use more advanced methods and models to evaluate the shared health care model that eHealth proposes [
Earlier studies have already attempted to model and predict eHealth usage among patients in Europe [
Data for this study were drawn from the Benchmarking Deployment of eHealth among General Practitioners Phase 2 (GPII) research project [
The study used survey data from a sample of 9196 European general practitioners, that is to say, physicians working in outpatient establishments in specialties such as general practice, family medicine, internal medicine, or general medicine. The sampling universe comprised 465,718 European physicians with an overall margin of error of ±1.03 in the case of maximum indetermination p=q=50%, for a confidence level of 95.5% (
The questionnaire used in the survey contained 38 questions grouped into 3 dimensions (
The survey was answered by European general practitioners in mixed Web-based, phone (Web-CATI), and face-to-face interviews lasting for half an hour each and in a native language of each country. A study presentation paragraph was written to inform potential respondents about the confidentiality of any data provided and the academic aim of the research. European general practitioners voluntarily answered the questionnaire and did not receive any payment in cash or kind. While the questionnaire was being implemented, an expert was on hand at all times (via email) to resolve any queries that the respondents had. The respondent general practitioners were selected by means of probability sampling applied to each country universe. The net response rate was 35.5%. The fieldwork period ran from October 25, 2012 to March 6, 2013. The GPII research project followed the Checklist for Reporting Results of Internet E-Surveys criteria [
From an empirical perspective, explanatory factors determining eHealth usage outcomes raise two particular difficulties. First, the approach to the concept requires a multidimensional basis that is not usually captured in a single variable. In fact, the most common approaches found in the literature perform partial analyses of its various dimensions. This type of analysis has the disadvantage of not taking a full snapshot of the explanatory factors, which gives rise to the second difficulty: statistical modeling. In other words, eHealth usage outcomes can be interpreted as a latent, nonobservable concept, which, therefore, calls for statistical techniques that allow variables of this type, which are not directly measurable, to be used [
In the empirical literature, structural equation modeling (SEM) with latent variables has been used to overcome this problem. A general SEM is a formal mathematical model. It is a set of linear equations that encompasses various types of models such as regression analysis models, simultaneous equation systems, factor analysis, and path analysis. The main advantage of this method of analysis is the incorporation of different types of variables into the SEM. Directly observable and measurable variables, and theoretical or latent variables representing concepts that are not directly observed can, therefore, be incorporated. When the variable to be explained (dependent) is latent, it must be continuous, whereas dependent observed variables can be continuous, censored, binary, ordered, categorical (ordinals), or combinations of any of these variable types [
This method of analysis allows us to define eHealth usage outcomes as a latent variable, thus enabling us to calculate the specific explanatory effect of the variables that it comprises. Hence, besides building an overall explanatory model of the determinants of eHealth usage, it is also possible to identify which of its explanatory dimensions are more important. In addition, SEM enables the relationships between the different observable variables included in the model (indirect effects) to be estimated. In this context, and in order to capture the factors that explain eHealth usage outcomes in a large sample of European general practitioners, we proposed and tested a two-stage SEM with latent variables and measurement errors for 2012-2013.
We applied the 2-stage empirical methodology as follows: in the first stage, we tested the relationships among 101 indicators and the 9 dimensions describing eHealth usage and eHealth usage outcomes by means of SEM and additive indicators (in those dimensions with primary data of a dichotomous nature), and in the second stage, we tested the relationships among the indicators constructed for those 9 dimensions (based on the coefficients and aggregations from the first stage). This methodology involved the design and statistical testing of 5 empirical SEM models (4 models for the first stage and 1 model for the second stage) and also 5 additive indicators in the first stage.
The 9 model dimensions and variables are as follows:
Model of eHealth usage and eHealth usage outcomes. H: hypothesis; ICT: information and communication technology; (+): positive prediction; (−) negative prediction.
Additionally, we performed logistic regression to model the predictors of 2 eHealth usage outcomes indicators using independent variables corresponding to sociodemographic circumstances, attitudes, and working conditions. For each independent variable, we calculated odds ratios (ORs) and their 95% CI. We used IBM SPSS Amos v24 (IBM Corp) for all calculations.
In the ICTUS dimension, the standardized coefficient variability was 0.58. The variables with the highest explanatory power in this dimension were related to describing the medical organizations in contact with general practitioners (0.640) as well as the existence of problems of compatibility in electronically exchanging patient data (0.481). In contrast, fewer explanatory variables were related to computer usage in general practice (0.064). In the BARRIERS dimension, the standardized coefficient variability was much lower (0.15) between the explanatory variables related to the lack of time or additional workload (0.681), lack of sufficient training (0.673) or lack of sufficient ICT skills (0.663), and increased patient expectations (0.528).
Regarding the 2 dimensions of general practitioners’ eHealth usage outcomes, the standardized coefficients obtained showed a different variability. In the case of eHealth usage outcomes related to IP, the standardized coefficient variability was 0.37. The variables with the highest explanatory power were related to the fact that eHealth was easy to use (0.801), that general practitioners found it easy to get it to do what they wanted (0.769), and that it was flexible to use or interact with (0.744). The variables with the least explanatory power were related to people who influence general practitioners’ general behavior (0.442) or clinical behavior (0.435) in respect of eHealth usage. In the case of eHealth usage outcomes in PP, the standardized coefficient variability was much lower (0.15). The variables with the highest explanatory power were related to improvement in the efficiency of the whole health care system (0.797), in the quality of treatment (0.784), and in the quality of diagnosis decisions (0.783). In contrast, the variables with the least explanatory power were related to the general practitioners’ perceptions of eHealth usage increasing patient access to health care (0.647) or reducing pharmaceutical expenditure (0.649).
For the remaining 5 dimensions relating to eHealth usage, namely PHR, THEALTH, HIEs, EHR_DSS, and EHR_DAT, and as a result of the dichotomous nature of the base data, we constructed 5 additive indicators. For each of the 5 dimensions, we created a joint indicator that adds together their internal uses (indicators). Thus, the PHR indicator takes a value between 0 and 6, the THEALTH indicator between 0 and 4, the HIE indicator between 0 and 15, the EHR_DSS indicator between 0 and 6, and the EHR_DAT indicator between 0 and 19 (
The direct standardized coefficients obtained validated the 21 formulated hypotheses and the signs of their relationships. For example, as we had hypothesized, the barriers to ICTUS determined a negative effect on ICT, HIE, and EHR_DSS usage. Regarding the relationships between different dimensions of eHealth usage, it is notable that ICTUS explained HIE (0.571) with considerable intensity that HIE had considerable explanatory power over PHR (0.398) and EHR_DSS (0.349) and that an important link was found between EHR_DSS and EHR_DAT (0.365). Concerning the indicators relating to eHealth usage outcomes, the indicator linked to IP was explained by ICT (0.132), EHR_DAT (0.215), EHR_DSS (0.050), PHR (0.112) and THEALTH (0.101) usage. In contrast, the outcomes indicator linked to PP was explained by THEALTH usage (0.027) and, primarily, by the IP outcomes indicator (0.607).
Descriptive statistics of eHealth usage and eHealth usage outcome dimensions, 2012-2013.
# | Dimension | Mean (SD) | Minimum | Maximum | Skewness | Kurtosis |
1 | Information and communication technology usage | 1.54 (1.002) | 0.00 | 4.54 | 0.085 | 0.738 |
2 | Barriers to eHealth usage | 9.99 (3.014) | 0.00 | 15.34 | −0.855 | 1.177 |
3 | Personal Health Records | 0.82 (1.320) | 0.00 | 6.00 | 1.872 | 3.360 |
4 | Telehealth | 0.62 (0.865) | 0.00 | 4.00 | 1.450 | 1.747 |
5 | Health Information Exchanges | 5.26 (3.962) | 0.00 | 15.00 | 0.551 | −0.569 |
6 | Electronic Health Records_Decision Support Systems | 2.38 (2.127) | 0.00 | 6.00 | 0.371 | −1.233 |
7 | Electronic Health Records_Data | 13.80 (6.015) | 0.00 | 19.00 | −1.428 | 0.628 |
8 | eHealth usage outcomes in internal practices | 8.86 (2.719) | 0.00 | 13.12 | −0.785 | 0.681 |
9 | eHealth usage outcomes in practices with patients | 11.50 (3.712) | 0.00 | 16.88 | −0.785 | 0.741 |
Explanatory factors of European general practitioners’ eHealth usage and eHealth usage outcomes, 2012-2013. Goodness-of-fit indices: normed fit index: 0.966; relative fit index: 0.914; incremental fit index: 0.967; Tucker-Lewis index: 0.915; comparative fit index: 0.967; root mean square error of approximation: 0.072.
Hypothesis (H) | Explained dimension | Explanatory dimension | Standardized coefficienta | SE | ||
H1 | ICTUSb | BARRIERSc | −0.118 | <.001 | 0.003 | <.001 |
H2 | HIEd | ICTUS | 0.571 | <.001 | 0.034 | <.001 |
H3 | HIE | BARRIERS | −0.035 | <.001 | 0.011 | <.001 |
H4 | PHRe | ICTUS | 0.062 | <.001 | 0.015 | <.001 |
H5 | PHR | HIEs | 0.398 | <.001 | 0.004 | <.001 |
H6 | THEALTHf | ICTUS | 0.031 | <.001 | 0.010 | .003 |
H7 | THEALTH | HIEs | 0.248 | <.001 | 0.003 | <.001 |
H8 | THEALTH | PHR | 0.081 | <.001 | 0.007 | <.001 |
H9 | EHR_DSSg | ICTUS | 0.191 | <.001 | 0.023 | <.001 |
H10 | EHR_DSS | BARRIERS | −0.012 | <.001 | 0.006 | .003 |
H11 | EHR_DSS | HIEs | 0.349 | <.001 | 0.006 | <.001 |
H12 | EHR_DATh | ICTUS | 0.190 | <.001 | 0.062 | <.001 |
H13 | EHR_DAT | EHR_DSS | 0.365 | <.001 | 0.027 | <.001 |
H14 | EHR_DAT | HIEs | 0.174 | <.001 | 0.016 | <.001 |
H15 | OUTINTPRAi | ICTUS | 0.132 | <.001 | 0.030 | <.001 |
H16 | OUTINTPRA | EHR_DAT | 0.215 | <.001 | 0.005 | <.001 |
H17 | OUTINTPRA | EHR_DSS | 0.050 | <.001 | 0.015 | <.001 |
H18 | OUTINTPRA | PHR | 0.112 | <.001 | 0.021 | <.001 |
H19 | OUTINTPRA | THEALTH | 0.101 | <.001 | 0.031 | <.001 |
H20 | OUTPRAPATj | THEALTH | 0.027 | .002 | 0.036 | .003 |
H21 | OUTPRAPAT | OUTINTPRA | 0.607 | <.001 | 0.011 | <.001 |
aRegression analysis: structural equation modeling (SEM); estimated coefficients: direct effects.
bICTUS: information and communication technology usage.
cBARRIERS: barriers to information and communication technology usage.
dHIE: Health Information Exchange.
ePHR: Personal Health Records.
fTHEALTH: telehealth.
gEHR_DSS: Electronic Health Records_Decision Support Systems.
hEHR_DAT: Electronic Health Records_Data.
iOUTINTPRA: eHealth usage outcomes in internal practices.
jOUTPRAPAT: eHealth usage outcomes in practices with patients.
To capture the main predictors of European general practitioners’ eHealth usage outcomes, we performed logistic regression using independent variables for general practitioners’ sociodemographic circumstances, attitudes toward ICT impact, and working conditions. The first step in this analysis was to recode the two composite indicators of eHealth usage outcomes. The mean value the composite indicator of OUTINTPRA was 8.86 (SD 2.72; minimum to maximum range 0.0-13.12). The mean value of the composite indicator of OUTPRAPAT was 11.5 (SD 3.71; minimum to maximum range 0.0-16.88). We, therefore, constructed a dichotomous eHealth usage outcomes indicator based on the mean of the composite indicators obtained. The two dichotomous eHealth usage outcomes indicators took the value 1 when the composite indicators of eHealth usage outcomes were equal to or greater than the mean and the value 0 when less than the mean. The mean value of the dichotomous composite indicator of OUTINTPRA was 0.547 (SD 0.498; minimum to maximum range 0-1). The mean value of the dichotomous indicator of OUTPRAPAT was 0.558 (SD 0.497; minimum to maximum range 0-1). For 54.75% (5035/9196) of European general practitioners, the eHealth usage outcomes in IP were more favorable (greater than the mean). For 55.82% (5133/9196) of European general practitioners, the eHealth usage outcomes in PP were more favorable (greater than the mean).
Concerning attitudes toward perceived ICT impact on IP, the results obtained showed considerable predictive power. European general practitioners who positively valued ICT impact on their personal working processes (OR 5.30, 95% CI 4.73-5.93), teamwork processes (OR 4.19, 95% CI 3.78-4.65), and the doctor-patient relationship (OR 3.97, 95% CI 3.60-4.37) had a high propensity toward favorable effects of OUTINTPRA. In contrast, European general practitioners who negatively valued the effect of ICT impact or considered that it had no effect on their personal working processes (no effect: OR 0.26, 95% CI 0.23-0.29; negatively: OR 0.12, 95% CI 0.09-0.15), teamwork processes (no effect: OR 0.31, 95% CI 0.28-0.35; negatively: OR 0.17, 95% CI 0.14-0.22), or the doctor-patient relationship (no effect: OR 0.51, 95% CI 0.47-0.56; negatively: OR 0.27, 95% CI 0.24-0.31) had a lower propensity toward favorable effects of eHealth usage. A clear link was also observed between a greater personal use of Web 2.0 (social media, blogs, etc) and more favorable OUTINTPRA. European general practitioners who had used 2.0 applications in their personal lives often (OR 1.77, 95% CI 1.60-1.97) or sometimes (OR 1.16, 95% CI 1.06-1.28) also showed a higher propensity toward favorable OUTINTPRA than those who rarely (OR 0.99, 95% CI 0.89-1.09) or never (OR 0.59, 95% CI 0.55-0.65) used 2.0 applications.
Finally, the study also found predictive power between some working conditions and favorable OUTINTPRA. Self-employed European general practitioners (OR 1.33, 95% CI 1.22-1.45) also had a greater predisposition toward favorable eHealth usage outcomes in IP than salaried European general practitioners (OR 0.82, 95% CI 0.75-0.89). For its part, the dynamic of the number of patients treated also had predictive power. In the case of favorable OUTINTPRA, the results obtained had an inverted U shape. European general practitioners who reported that the number of patients treated had remained constant in the last 2 years (OR 1.08, 95% CI 1.01-1.17) had a greater predisposition toward favorable eHealth usage outcomes in their IP. In contrast, general practitioners who reported that the number of patients treated had increased (1.03, 95% CI 0.95-1.12) or had fallen in recent years (OR 0.81, 95% CI 0.72-0.91) had less predictive power.
Concerning attitudes toward perceived ICT impact on IP, the results obtained showed considerable predictive power. European general practitioners who positively valued ICT impact on their personal working processes (OR 4.83, 95% CI 4.32-5.40), teamwork processes (OR 3.38, 95% CI 3.05-3.74), and the doctor-patient relationship (OR 6.02, 95% CI 5.43-6.67) had a high propensity toward favorable effects of eHealth usage on PP. In contrast, European general practitioners who negatively valued the effect of ICT impact or considered that it had no effect on their personal working processes (no effect: OR 0.26, 95% CI 0.23-0.30; negatively: OR 0.17, 95% CI 0.13-0.21), teamwork processes (no effect: OR 0.38, 95% CI 0.34-0.42); negatively: OR 0.20, 95% CI 0.16-0.26), and the doctor-patient relationship (no effect: OR 0.38, 95% CI 0.35-0.42; negatively: OR 0.26, 95% CI 0.22-0.29) had a lower propensity toward favorable effects of eHealth usage. A clear link was also observed between a greater personal use of Web 2.0 (social media, blogs, etc) and more favorable OUTPRAPAT. European general practitioners who had used 2.0 applications in their personal lives often (OR 1.94, 95% CI 1.74-2.15) or sometimes (OR 1.19, 95% CI 1.08-1.31) also showed a higher propensity toward favorable OUTPRAPAT than those who rarely (OR 1.09, 95% CI 0.98-1.20) or never (OR 0.51, 95% CI 0.47-0.56) used 2.0 applications.
Logistic regression (odds ratio, OR) models of the dichotomous composite indicator of eHealth usage outcomes in internal practices by sociodemographic circumstances, attitudes toward information and communication technology (ICT) impact, and working conditions, 2012-2013.
Sociodemographic factors | OR | 95% CI | |
Male | 0.89 | 0.85-0.92 | |
Female | 1.15 | 1.10-1.20 | |
<35 | 1.14 | 1.02-1.26 | |
36-45 | 1.16 | 1.06-1.28 | |
46-55 | 0.95 | 0.87-1.03 | |
>56 | 0.90 | 0.82-0.98 | |
Large city (more than 100,000 inhabitants) | 1.02 | 0.93-1.11 | |
Small or midsized city (between 20,000 and 100,000 inhabitants) | 0.98 | 0.89-1.07 | |
Rural town (fewer than 20,000 inhabitants) | 1.00 | 0.97-1.03 | |
Positive | 5.30 | 4.73-5.93 | |
No change | 0.26 | 0.23-0.29 | |
Negative | 0.12 | 0.09-0.15 | |
Positive | 4.19 | 3.78-4.65 | |
No change | 0.31 | 0.28-0.35 | |
Negative | 0.17 | 0.14-0.22 | |
Positive | 3.97 | 3.60-4.37 | |
No change | 0.51 | 0.47-0.56 | |
Negative | 0.27 | 0.24-0.31 | |
Often | 1.77 | 1.60-1.97 | |
Sometimes | 1.16 | 1.06-1.28 | |
Rarely | 0.99 | 0.89-1.09 | |
Never | 0.59 | 0.55-0.65 | |
Salaried | 0.82 | 0.75-0.89 | |
Self-employed | 1.33 | 1.22-1.45 | |
Increased | 1.03 | 0.95-1.12 | |
Remained constant | 1.08 | 1.01-1.17 | |
Decreased | 0.81 | 0.72-0.91 |
Logistic regression (odds ratio, OR) models of the dichotomous composite indicator of eHealth usage outcomes in practices with patients by sociodemographic circumstances, attitudes toward information and communication technology (ICT), and working conditions, 2012-2013.
Sociodemographic factors | OR | 95% CI | |
Male | 0.86 | 0.83-0.89 | |
Female | 1.19 | 1.14-1.24 | |
<35 | 1.32 | 1.13-1.54 | |
36-45 | 1.21 | 1.10-1.33 | |
46-55 | 0.84 | 0.77-0.92 | |
>56 | 0.93 | 0.85-1.01 | |
Large city (>100,000 inhabitants) | 0.97 | 0.89-1.05 | |
Small or midsized city (20,000-100,000 inhabitants) | 1.00 | 0.91-1.09 | |
Rural town (<20,000 inhabitants) | 1.04 | 0.95-1.13 | |
Positive | 4.83 | 4.32-5.40 | |
No change | 0.26 | 0.23-0.30 | |
Negative | 0.17 | 0.13-0.21 | |
Positive | 3.38 | 3.05-3.74 | |
No change | 0.38 | 0.34-0.42 | |
Negative | 0.20 | 0.16-0.26 | |
Positive | 6.02 | 5.43-6.67 | |
No change | 0.38 | 0.35-0.42 | |
Negative | 0.26 | 0.22-0.29 | |
Often | 1.94 | 1.74-2.15 | |
Sometimes | 1.19 | 1.08-1.31 | |
Rarely | 1.09 | 0.98-1.20 | |
Never | 0.51 | 0.47-0.56 | |
Salaried | 1.02 | 0.93-1.11 | |
Self-employed | 1.10 | 1.03-1.28 | |
Increased | 1.12 | 1.03-1.22 | |
Remained constant | 0.93 | 0.85-1.01 | |
Decreased | 0.91 | 0.81-1.03 |
Finally, the study also found predictive power between some working conditions and favorable OUTPRAPAT. Self-employed European general practitioners (OR 1.10, 95% CI 1.03-1.28) also had a greater predisposition toward favorable OUTPRAPAT than salaried European general practitioners (OR 1.02, 95% CI 0.93-1.11). For its part, the dynamic of the number of patients treated also had predictive power. In the case of favorable OUTPRAPAT, the results obtained had a growing evolution. European general practitioners who reported that the number of patients treated had remained constant in the last 2 years (OR 1.12, 95% CI 1.03-1.22) had a greater predisposition toward favorable OUTPRAPAT. In contrast, general practitioners who reported that the number of patients treated had remained constant (OR 0.93, 95% CI 0.85-1.01) or had fallen (OR 0.91, 95% CI 0.81-1.03) did not have predictive power.
The goal of our study was to design and evaluate a predictive multidimensional model of general practitioners’ eHealth usage outcomes in IP and in PP, comprising 9 dimensions and 101 indicators. To that end, we used a broad population sample of 9196 European general practitioners. The results obtained are very useful for two reasons. First, obtaining new evidence centered solely on general practitioners allowed us to focus the analysis better, particularly with regard to the eHealth usage dimensions (ICTUS, BARRIERS, PHRs, HIEs, THEALTH, and EHRs) that determine favorable eHealth usage outcomes. Second, the predictors we obtained (sociodemographic circumstances, attitudes toward ICT impact, and working conditions) provided evidence that complements studies that have taken partial approaches.
In our study, we constructed 2 composite indicators using a 2-stage SEM methodology, and the results obtained are consistent with this evidence: they showed that in 2012-2013: (1) for 54.75% (5035/9196) of European general practitioners, the eHealth usage outcomes in IP were more favorable (greater than the mean) and (2) for 55.82% (5133/9196) of European general practitioners, the eHealth usage outcomes in PP were more favorable (greater than the mean).
The outcomes indicator linked to IP were explained by ICTUS, EHR_DAT, EHR_DSS, PHRs, and THEALTH. In contrast, the outcomes indicator linked to PP were explained by THEALTH usage and, primarily, by the IP outcomes indicator. General practitioners’ eHealth usage outcomes had a 2-fold interrelated dynamic. Firstly, the set of eHealth uses had explanatory power over IP outcomes. And secondly, the IP outcomes, together with THEALTH usage, determined eHealth usage outcomes in PP. In the explanation of general practitioners’ eHealth usage, it, therefore, seems that a certain “experience effect” occurs. eHealth usage takes place initially in IP and is then transferred to PP. In addition, the results of this study revealed the key role that telemedicine would play in the sense that, through training programs, upskilling, and learning, it would enable eHealth usage to be transferred from IP to PP [
Our study confirms that the perception of eHealth being easy to use, of general practitioners finding it easy to get it to do want they want, and of it being flexible to use or interact with are explanatory factors that have a bigger effect on eHealth usage in IP. The statistical significance of these determining factors refers back to the importance of perceived usefulness and ease of use when the use of a technology needs to be explained. General practitioners’ surgeries are characterized by the high number of patients cared for. Patients have various health problems, some of which are clinically complex. As a result, general practitioners have to handle several aspects at the same time, which may lead to doubts in their daily clinical practice [
Regarding eHealth usage outcomes in PP, the variables with the highest explanatory power were related to improvement in the efficiency of the whole health care system, in the quality of treatment (0.784) and in the quality of diagnosis decisions. In this respect, eHealth is perceived as a technology that serves to reduce costs and increase the quality of health care provision [
The variables with less explanatory power were related to general practitioners’ perceptions of whether or not eHealth usage increases patient access to health care or reduces pharmaceutical expenditure. Again, the variables with less explanatory power were related to the belief that electronic systems would disrupt health care provision[
General practitioner’s eHealth usage in IP and in PP has been shown to have significant relationships with the individual characteristics of general practitioners, such as gender and age. It has been described in the literature that demographic factors such as age, education, gender, nationality, and clinical experience can have an influence on health care professionals’ predisposition toward eHealth systems [
Concerning attitudes toward the perceived ICT impact on internal and external practices, our analysis revealed that European general practitioners who positively valued ICT impact on their personal working processes, teamwork processes, and the doctor-patient relationship (IP OR 3.97, 95% CI 3.60-4.37) had a high propensity toward favorable OUTINTPRA. A clear link was also observed between a greater personal use of Web 2.0 (social media, blogs, etc) and more favorable OUTINTPRA. Various studies have described organizational and individual barriers to the implementation of eHealth services. Recognizing and understanding what the barriers and facilitators are is ideal for devising strategies and interventions to improve the effective eHealth usage and to address the barriers to implementation [
Regarding occupational status, our study has also revealed that self-employed European general practitioners had a greater predisposition toward favorable OUTINTPRA and OUTPRAPAT than salaried European general practitioners. Specifically, self-employed general practitioners attached importance to the potential of eHealth to reduce costs [
Last, from the perspective of demand pressures on health care practices, European general practitioners who reported that the number of patients treated had remained constant in the last 2 years had a greater predisposition toward favorable eHealth usage outcomes in their IP. In contrast, general practitioners who reported that the number of patients treated had increased or had fallen had less predictive power. This would, therefore, confirm the available evidence, which associates the implementation of digital technologies in the health care field with specific organizational circumstances, in particular with workflow pressures that are not too heavy [
In contrast—and this is the only result that clearly distinguishes between eHealth usage for IP and that for PP—demand pressures would have predictive power over positive eHealth outcomes. European general practitioners who reported that the number of patients treated had remained constant in the last 2 years had a greater predisposition toward favorable OUTPRAPAT [
Our study has several limitations. First, there was a time lag between the years the data were obtained (2012-2013) and the year we wrote the paper. However, we felt that the availability of a population database of 9196 European general practitioners deserved an analysis despite the time lag. In future research, and as they become available, we will use newer data and introduce dynamic comparisons. Second, the study provides information only from the perspective of physicians. In the future, we intend to address the issue of eHealth usage by health professionals and health users. By doing so, we will be able to improve our multidimensional approach and obtain mixing results and conclusions for all actors involved in eHealth usage and eHealth usage outcomes. Third, the empirical methodology could also be improved by looking at the intensity of eHealth usage (not simply usage or mean usage) and at a higher number of predictors.
The results obtained highlight the need for more in-depth research to be conducted into the link between eHealth usage, eHealth outcomes and predictors, and the different health care systems in Europe. By doing so, it will be possible to increase the resolution of our results and to establish whether the intensity of eHealth usage and eHealth outcomes varies depending on the health care systems or the extent to which health care systems determine the prediction of eHealth usage or eHealth outcomes. Similarly, strategic and public policy actions resulting from the research could be adapted more precisely to each health care system. Finally, the study results could be supplemented by the construction of a composite indicator of eHealth usage by health care professionals and health care users. The design, validation, and prediction of composite indicators of eHealth usage and eHealth outcomes that take into consideration the perspectives of both users (ie, patients) and professionals in the different European health care systems would provide us with a very comprehensive view of the issue and would allow us to round off our multidimensional approach. We shall focus our efforts on all of these approaches in the near future.
Again, appropriate innovations are needed to promote eHealth usage. European, national and regional authorities should take the results of these studies into account to develop suitable policies for greater integration of HIT among European health care professionals. This setting poses significant challenges for the formulation of public policies and strategies by states where decisions about eHealth should not be overlooked.
Statistical information based on General Practitioners II (GPII) survey.
Benchmarking Deployment of eHealth among General Practitioners II (GPII) Questionnaire.
Information and communication technology (ICT) usage by European general practitioners descriptive statistics 2012-2013.
Barriers to eHealth usage by European general practitioners descriptive statistics 2012-2013.
Personal Health Records (PHR) usage by European general practitioners descriptive statistics 2012-2013.
Telehealth usage by European general practitioners descriptive statistics 2012-2013.
Health Information Exchanges (HIE) usage by European general practitioners descriptive statistics 2012-2013.
Electronic Health Records_ Decisions Support System (EHR_DSS) usage by European general practitioners descriptive statistics 2012-2013.
Electronic Health Records_ Data (EHR_DAT) usage by European general practitioners descriptive statistics 2012-2013.
Outcomes of internal practices in eHealth usage by European general practitioners descriptive statistics 2012-2013.
Outcomes of practices with patients in eHealth usage by European general practitioners descriptive statistics 2012-2013.
Explanatory factors of some dimensions of European general practitioners’ eHealth usage and eHealth usage outcomes in 2012-2013.
eHealth usage by European general practitioners dimensions frequency statistics* 2012-2013.
barriers to eHealth usage
comparative fit index
electronic health
electronic health record
electronic health records_decision support system
electronic health records_data
Benchmarking Deployment of eHealth among General Practitioners Phase 2
health information exchange
health information technology
information and communication technology
information and communication technology usage
incremental fit index
internal practices
normed fit index
odds ratio
eHealth usage outcomes in internal practices
eHealth usage outcomes in practices with patients
personal health record
practices with patients
relative fit index
root mean square error of approximation
structural equation modeling
telehealth
Tucker-Lewis index
This work used data from the Benchmarking Deployment of eHealth among GPII research project, carried out by the European Commission Directorate General for Communication Networks, Content & Technology. Open access microdata are available from the European Commission Digital Economy & Society Webpage.
JTS participated in formulating the research question, study design, literature review, data analysis and statistical modeling, interpretation of the findings, and drafting the manuscript. ÁDC participated in data analysis and statistical modeling. ISR participated in the design and data analysis. FSR contributed to formulating the research question, study design, literature review, interpretation of results, and drafting the manuscript; he is the guarantor of the paper.
None declared.