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 use of patient portals for presenting health-related patient data, such as blood test results, is becoming increasingly important in health practices. Patient portals have the potential to enhance patient health engagement, but content might be misinterpreted.
This study aimed to discover whether the way of presenting blood test outcomes in an electronic patient portal is associated with patient health engagement and whether this varies across different blood test outcomes.
A 2x3 between-subjects experiment was conducted among members of the Nivel Dutch Health Care Consumer Panel. All participants read a scenario in which they were asked to imagine themselves receiving blood test results. These results differed in terms of the presented blood values (ie, normal vs partially abnormal vs all abnormal) as well as in terms of whether the results were accompanied with explanatory text and visualization. Patient health engagement was measured both before (T0) and after (T1) participants were exposed to their fictive blood test results.
A total 487 of 900 invited members responded (response rate 54%), of whom 50.3% (245/487) were female. The average age of the participants was 52.82 years (SD 15.41 years). Patient health engagement saw either a significant decrease or a nonsignificant difference in the experimental groups after viewing the blood test results. The mean difference was smaller in the groups that received blood test results with additional text and visualization (meanT0 5.33, SE 0.08; meanT1 5.14, SE 0.09; mean difference 0.19, SE 0.08,
Adding text and visualization features can attenuate the decrease in patient health engagement in participants who receive outcomes of a blood test via a patient portal, particularly when blood test results are (partly) normal. This suggests that explanatory text and visualization can be reassuring. Future research is warranted to determine whether these results can be generalized to a patient population who receive their actual blood test results.
Patient health engagement (PHE) can play an important role in personal health and can improve adherence and satisfaction with received care [
Currently, electronic patient portals are increasingly offered by health professionals to communicate blood test results to the patient. Even though these portals are not optimally used yet [
One of the main problems is that patient portals often rely on numerals to purvey information. This is a concern for people with low numeracy skills (ie, people who lack the ability to use and draw meaning from numbers). People with low numeracy skills have shown more difficulties with identifying out-of-range test results [
There are good reasons to believe that textual explanations with explanatory visual aids can benefit patients [
The basis of our study was the comparison of two portals. Both portals communicated blood test results accompanied by reference values (ie, the range that expresses the normal values for that test). These types of portals are most often used in Dutch clinical practice, which was the setting of our study. The first portal was a fictive basic portal only providing the patient’s blood values with the corresponding reference categories. The second portal was based on a more sophisticated portal as developed by Saltro, one of the largest diagnostic centers in the Netherlands. The Saltro portal adds two main features compared to the first portal: (1) textual understandable information explaining the test and its outcomes and the action the patient can take afterwards and (2) visual support by using traffic light colors to indicate whether the outcome is within the normal range. The content of this portal was cocreated with health care professionals, communication experts, and patients. The text was written to be understandable for the majority of people. The level of health literacy of the results information has been estimated at communication level 1B on the scales of the Common European Framework of Reference for Languages. Furthermore, the content is frequently evaluated by patients and adapted according to their recommendations.
A 2x3 between-subjects experimental design was employed to test the effects of the blood test results outcome and the addition of explanatory text and visualization on PHE. Participants were so-called analog or simulated patients (ie, people who imagine themselves being in a hypothetical health care situation) presented with a hypothetical case (see
We tested the two types of portals (ie, with vs without explanatory text and visualization), within which we distinguished three possible outcomes of the blood tests: all values within range (normal), partially deviating values (partially abnormal), and all deviating values (all abnormal).
You have been tired for a couple of months now, and it does not get any better. No matter how much you sleep, you remain tired. You even fell asleep at work once. You do not feel tense or stressed. It does bother you that the tiredness does not go away. You decide to visit your general practitioner (GP) to describe your symptoms to better understand/get a grip on your situation. The GP asks if you have any thoughts on the cause of the tiredness. You have no clue. Your private life is fine, you have never been this tired before, your diet is healthy, and you do not smoke or use medication. You have become worried about the situation; what is going on? The GP suggests to first run some blood tests before making any decisions. The GP explains that she wants to measure 3 types of blood values: tests for an underactive thyroid (thyroid stimulating hormone), anemia (hemoglobin), and a deficit of vitamin D. In the morning, your blood gets drawn, and you are told that your results are available at your convenience on the website of your GP through a patient portal the same day.
The 6 groups of the 2x3 design.
Portal type | All normal values | Partially abnormal values | All abnormal values |
No features | “Green basic” | “Partial orange basic” | “Orange basic’” |
Added featuresa | “Green Saltro” | “Partial orange Saltro” | “Orange Saltro” |
aExplanatory text with added visuals (see
Basic portal (top) and Saltro portal (bottom), both displaying the same partially abnormal results. All text is in Dutch and is shown in a similar fashion for hemoglobin and thyroid stimulating hormone.
The Saltro portal adds explanatory text that gives an explanation about which function the substance has in the body and what exactly is being measured, while showing a bar that presents the patient’s blood value in an orange (abnormal) or green (normal) range. A marker shows where the patient’s blood test result falls within the range (see
We approached a sample of 900 participants of the Dutch Health Care Consumer Panel to complete an online questionnaire. This panel aims to measure opinions on and knowledge of health care as well as the expectations of and experiences with health care among a cross-section of the Dutch population [
Data were analyzed anonymously and processed according to the privacy policy of the Dutch Healthcare Consumer Panel, which complies with the General Data Protection Regulation. According to Dutch legislation, there is no legal requirement to obtain informed consent nor approval by a medical ethics committee for conducting research through the panel [
We used the original 9-item version of the PHE scale as the starting point for our measurement [
For the randomization check, we examined whether participant characteristics were equally divided across experimental conditions using
For the main analysis, a mixed analysis of variance was conducted with “text/visualization added” and “outcome of blood test results” as the between-subjects factors and “pretest measure of patient health engagement” versus “posttest measure of patient health engagement” as the within-subjects factor or repeated measure. Main effects of the between-subjects factors (ie, “text/visualization added” and “outcome of blood test results”) and within-subjects factor (ie, “pretest measure of patient health engagement” versus “posttest measure of patient health engagement”) were calculated. Furthermore, the interaction effect between the between-subjects factors (ie, “text/visualization added” × “outcome of blood test results”), two-way interaction effects between the between-subjects and within-subjects factors (ie, “text/visualization added” × “pretest vs posttest” and “outcome of blood test results” × “pretest vs posttest”), and three-way interaction effect between factors (ie, “text/visualization added” × “type of blood test results” × “pretest vs posttest”) were assessed. Simple effects analyses were conducted in case of significant interactions between variables.
Of the 519 participants that started the survey, 487 had sufficiently complete data (ie, ≥60% of the PHE measure completed).
Sample characteristics (N=487).
Characteristics | Values | ||
Age (years), mean (SD; range) | 52.82 (15.41; 24-90) | ||
|
|
||
|
Male | 242 (49.7) | |
|
Female | 245 (50.3) | |
|
|
||
|
Low | 37 (7.7) | |
|
Middle | 219 (45.8) | |
|
High | 222 (46.4) | |
Health statusb, mean (SD; range) | 2.65 (0.87; 1-5) |
aValues for education do not add up to 487 due to missing data.
bSelf-reported health status (how would you rate your own health) ranges from 1 = excellent to 5 = poor.
We found that the outcome of the blood test results, regardless of the portal design, impacted PHE after exposure to the blood test results (
Furthermore, adding text and visualization that explained the blood test results impacted PHE after exposure to blood test results (
A significant interaction effect between the outcome of the blood test results and the addition of explanatory text and visualization revealed a nuanced insight into how PHE develops after being exposed to blood test results. It showed that, for all outcomes of blood test results, a lack of explanatory text and visualization decreased PHE after being exposed to the results (
Descriptive statistics for patient health engagement (PHE) before (T0) and after (T1) exposure to the patient portal across the experimental conditions (N=487).
Experimental conditions | n | PHE at T0, mean (SE) | PHE at T1, mean (SE) | ||
|
|
|
|
|
|
|
Without text | 83 | 5.02 (.14) | 4.73 (.16) | .03 |
|
With text | 82 | 5.51 (.14) | 5.46 (.16) | .71 |
|
|
|
|
|
|
|
Without text | 87 | 5.16 (.15) | 4.31 (.16) | <.001 |
|
With text | 74 | 5.16 (.14) | 5.19 (.16) | .81 |
|
|
|
|
|
|
|
Without text | 78 | 5.39 (.14) | 4.61 (.09) | <.001 |
|
With text | 83 | 5.31 (.15) | 4.73 (.17) | <.001 |
The aim of this study was to discover whether the way in which blood test outcomes are presented in a patient portal is associated with PHE and whether this varies across different blood test outcomes. Adding textual and visual explanations to blood test results minimizes the decline in PHE when receiving blood test results in an electronic patient portal. When presenting blood test results through an existing patient portal, the group that received explanatory text and visualization in addition to their results experienced less of a decline in PHE than the group without these features. This was particularly true for patients who received normal and partially abnormal results (ie, combination of normal and abnormal results). For patients who received abnormal results for all three blood tests, health engagement significantly decreased independent of whether explanatory text and visualization were added. It can be concluded that adding text and visualization to a patient portal can attenuate PHE and therefore involve patients more in their health management, but only when blood test results are normal or partially abnormal.
The focus of our study was to gain insight into how providing explanatory text and visualization, when presenting blood test results via a patient portal, influences PHE. While previous research has shown that patients are generally satisfied with the use of a patient portal to check their blood test results [
One recurring concern with patient portals is the fear of the misinterpretation of results [
In this study, we did not distinguish between the added value of explanatory text only and the added value of visualization only. Hence, we do not know which of these two features was the “active ingredient” or whether the combination of the features caused the effectiveness of the patient portal. In two previous studies, information provision in a patient portal was evaluated in isolation. In the first study, visual support using a color scheme to differentiate between normal results and divergent results was positively evaluated by elderly patients [
In addition, our study measured PHE using a questionnaire. Previous studies have used different ways to measure (patient) engagement. To illustrate, Phelps et al [
A strength of our study is that it systematically tested the impact of patient portal design on PHE in a highly controlled experimental setting. Although patient portal designs have been introduced to patients, these have not been studied yet in terms of how their design can impact PHE. In addition, our measurement of PHE gave a broad depiction of PHE through patients’ feelings, thoughts, and behaviors. Therefore, our study design gives this study a good foundation for its conclusions and implications.
There were also some limitations to our study. The first one is the potential bias in our sample. Although the sample was randomly selected from the Nivel Consumer Panel, which represented a relatively diverse group of patients, the participants of this panel are people who agreed to fill out questionnaires about health and health care. This means that there is good reason to believe that a majority of the people have, at least to some extent, affinity with their personal health (and health care). Furthermore, it can be assumed that the average panel user has a higher health literacy than normal. Health literacy entails the capability of obtaining, processing, and understanding information about health and health services [
The necessity for participants to empathize with a fictional case was another limitation. Actual patients, for whom results might indeed have an impact on their life, might have reacted differently to the results of the blood test as they might be expected to be more emotionally involved. Based on the meta-analysis by Van Vliet et al [
Another possible limitation is the specific design of our portal. Our results could not cover all portal designs, so the scope of the study may be limited and not generalizable to other patient portals. For instance, the colors of our “complementary” portal design could have influenced the feelings of the participants. Research has shown that the color red can evoke feelings with patients [
Our study has provided insight into how portal design can benefit PHE. Our results suggest that institutions that use a patient portal for blood test results might want to consider adding complementary information and visual support. Further testing of patient portal features is recommended, but institutions would do well to start adding visualization and textual support to their portals as this can benefit PHE. Adding text and visualization that explain the blood values and the implications of an abnormal result might support their patients in their health management. It might lead to patients who are more involved in their diagnosis and treatment, which could lead to, for instance, patients being more comfortable voicing concerns or asking questions. As mentioned, there is a possibility that our sample consisted of participants with an above-average health literacy. Therefore, future research could focus on patients with below-average health literacy who might need a different portal design. Yet, more research among actual patients is needed to test the portal with patients who are personally involved and therefore more concerned about the outcomes of blood test results. For example, in case a fictional study is chosen, this fictional case could include a disease with a higher emotional impact and familiarity, such as cancer. For such a disease, a more intense reaction could occur while viewing blood test results, such as relief or anxiety, which in turn could have a stronger effect on PHE. However, ethical considerations should be taken into account, as participants often do not know that they are being confronted with such an emotional case. A warning up front for the participant might be considered if the need for a more emotional investment is deemed necessary. Lastly, we used an adaptive version of the validated, original PHE scale. By deleting 4 items, it is possible that it did not cover the full domain of PHE (ie, did not measure the domains of feelings, thoughts, and behaviors as fully as the original scale).
Patient portals have been designed to improve patient involvement. When blood test results are communicated to patients, it can negatively affect their PHE and consequently their involvement in their health management. However, when these outcomes are supported by explanatory text and visualization to help interpret the outcomes, the decrease in PHE can be attenuated, especially when test results are partially normal. As receiving test results can cause feelings of uncertainty in patients, which can lead to lower PHE, our results suggest that explanatory text with visualization can cause feelings of relief in patients. Future research should focus on repeating the experiment with actual patients who receive their own blood test results to test whether the results hold in more ecologically valid settings.
general practitioner
patient health engagement
BS, NB, JvW, AB, and LvD do not have conflicts of interest to declare. AG, ETK, and SvD are employees of Saltro, the organization that developed the portal that was tested in this experiment.