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 JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Chronic disease patients often face multiple challenges from difficult comorbidities. Smartphone health technology can be used to help them manage their conditions only if they accept and use the technology.
The aim of this study was to develop and test a theoretical model to predict and explain the factors influencing patients’ acceptance of smartphone health technology for chronic disease management.
Multiple theories and factors that may influence patients’ acceptance of smartphone health technology have been reviewed. A hybrid theoretical model was built based on the technology acceptance model, dual-factor model, health belief model, and the factors identified from interviews that might influence patients’ acceptance of smartphone health technology for chronic disease management. Data were collected from patient questionnaire surveys and computer log records about 157 hypertensive patients’ actual use of a smartphone health app. The partial least square method was used to test the theoretical model.
The model accounted for .412 of the variance in patients’ intention to adopt the smartphone health technology. Intention to use accounted for .111 of the variance in actual use and had a significant weak relationship with the latter. Perceived ease of use was affected by patients’ smartphone usage experience, relationship with doctor, and self-efficacy. Although without a significant effect on intention to use, perceived ease of use had a significant positive influence on perceived usefulness. Relationship with doctor and perceived health threat had significant positive effects on perceived usefulness, countering the negative influence of resistance to change. Perceived usefulness, perceived health threat, and resistance to change significantly predicted patients’ intentions to use the technology. Age and gender had no significant influence on patients’ acceptance of smartphone technology. The study also confirmed the positive relationship between intention to use and actual use of smartphone health apps for chronic disease management.
This study developed a theoretical model to predict patients’ acceptance of smartphone health technology for chronic disease management. Although resistance to change is a significant barrier to technology acceptance, careful management of doctor-patient relationship, and raising patients’ awareness of the negative effect of chronic disease can negate the effect of resistance and encourage acceptance and use of smartphone health technology to support chronic disease management for patients in the community.
Due to its large impact on patients’ health status and health care expenditure, there is a growing interest worldwide in developing programs to support consumers to self-manage chronic diseases [
Despite its potential benefits, mobile health (mHealth) technologies have encountered various challenges in patient acceptance [
Although there have been studies on consumer acceptance of health technology [
We conducted preliminary interviews with 10 patients who were frequent users of the smartphone health app Blood Pressure Assistant to understand why they used it. We identified 3 factors influencing their usage behavior: the need for hypertension control, compliance with their health care providers’ advice, and the reluctance to use it. This preliminary knowledge was taken into account in our conceptualization of the research model. The other constructs of the model were drawn from the relevant theories such as technology acceptance model (TAM) [
Among its wide adoption in all fields of technology acceptance studies, TAM [
Realizing the limitation of TAM in not considering the social factors that very much likely would influence a person’s perceptions about the technology, Venkatesh extended TAM to TAM2, which includes social influence (SI) as a key determinant of perceived usefulness and use intention [
Moreover, users’ prior technology usage experience can shape their belief in the new technology [
Cenfetelli developed a dual-factor model of information technology usage to compensate the limitation of TAM being solely focused on users’ positive (enabling) perceptions but ignoring the negative (inhibiting) ones [
Resistance to change (RTC) refers to people’s attempt to maintain their previous behaviors and habits in the face of change required. A study into physicians’ resistance toward health information technology finds that resistance to change is the inhibitor that has significant, direct influence on both behavioral intention and perceived usefulness [
In essence, adoption of smartphone health technology is a patient’s behavior to promote, protect, or maintain their own health [
The perceived benefits of action in HBM are embodied in perceived usefulness in our new model. Barriers to action are modeled as resistance to change. Self-efficacy is the extent of patients’ beliefs in their ability to complete various tasks and reach the goal of controlling hypertensive condition. In the social cognitive theory (SCT), self-efficacy refers to users’ confidence in their ability to use a technology, and has been modeled as a determinant of perceived ease of use [
The positive effects of doctor-patient interaction for chronic disease management have long been established [
A systematic review of studies on patient acceptance of consumer-centered health information technologies reveals that the most studied demographic variables on technology acceptance include sex, gender, and education [
The hypothesized theoretical research model.
Thus, we tested the moderating effects of these 3 variables on intention to use and propose the following hypothesis:
In this study, we define actual use as the ratio of a patient’s actual use of the app to that prescribed in their management plan for a certain period of time. The predictive power of TAM is undermined if actual use is not included in the model [
On the basis of the above reasoning, we propose that 4 social factors—resistance to change, social influence, perceived health threat, and relationship with doctor—and a technical factor (ie, perceived ease of use), and a personal factor (ie, smartphone usage experience) affect patients’ perceived usefulness of smartphone health technology. Three factors, relationship with doctor, usage experience, and self-efficacy, affect patients’ perceived ease of use of the technology. Perceived usefulness, perceived ease of use, perceived health threat, and resistance to change affect patient’s intention to use. Three demographic variables, gender, age, and education, mediate the effect of the above variables on intention to use. Ultimately, intention to use affects patients’ actual use of smartphone health technology (
The smartphone health app Blood Pressure Assistant was developed by the Biomedical Informatics Laboratory in Zhejiang University, People's Republic of China. It was designed to enable communication and collaboration between the outpatients and their health care providers in hypertension management. It included a smartphone app for the patients to use, named Blood Pressure Assistant (
The functions of the smartphone app for patients included a reminder for blood pressure measurement, medication, and exercise; the form to enter and submit blood pressure measurement records; and receiving physician feedback and access to the health information published through the app.
Screenshots of the smartphone-based Blood Pressure Assistant application.
The functions of the physician portal included continuous monitoring of patient health data, data visualization and reminding of abnormal situations, assessing patient health conditions based on the collected data, classification of patients according to their health conditions, and management of regular follow-up.
To improve population health, a chronic disease management program had been piloted to develop a model for chronic disease prevention and control in Ningxia Province in China. As the primary health care system was still in the emerging stage of development in the province, the program was run by the 2000-bed General Hospital of Ningxia Medical University, the only tertiary hospital in the province. The initial focus of the program was hypertension management. Therefore, the study population was the hypertensive outpatients in the Department of Cardiovascular Medicine in the hospital.
The health care providers who participated in the hypertension management program included a cardiovascular medicine specialist and a certified health manager. The program started once an outpatient was recruited and the hypertension management plan was developed for the person. A patient was requested to submit the blood pressure data via smartphone regularly according to the care plan. The system would assess whether the data were normal. An alarm would be flagged to the health care providers once any abnormal data were recorded. The health manager would then phone the patient to discuss the person’s abnormal health condition, reevaluate, and adjust the self-management plan. The patients could also read the information about chronic disease management published on the smartphone app. The system went live in November 2015.
The health care providers recruited the outpatients into the program. Patients who met the following inclusion criteria were recruited: (1) aged 18 years or over; (2) no other serious complications except hypertension; (3) had a smartphone and sufficient network connectivity at home; (4) able to read and write in Chinese; and (5) resided in Yinchuan city so as to be contactable. After being recruited by the specialist, the health manager provided face-to-face training to the patients. The content of the training included knowledge about hypertension self-management, and the method to download “Blood Pressure Assistant” and use it, either from Apple Store if the person used an iPhone or from the specific website if the person used an Android phone [
Questionnaire survey was conducted between June and September 2016, 1 month after a patient entered the program. It was conducted either through the telephone survey or the electronic questionnaire survey.
We started conducting telephone interviews to collect questionnaires. A researcher made a phone call to an eligible patient. After informing the person about the survey and seeking the respondent’s oral consent, the researcher read and sought the person’s answer to each question, and then entered the answer into the electronic questionnaire survey form. After collecting 23 responses, we found this method to be resource-intensive and not efficient. Therefore, we piloted the method of using the mobile phone app to conduct the electronic questionnaire survey. In this method, a patient could fill in the electronic questionnaire survey form that automatically displayed on the smartphone health app interface 1 month after the person was recruited into the program. The information presented included the survey purpose, its voluntary nature, and insurance about anonymity of results in any related research publications. A patient could tick the check box to give consent. Implicit consent was assumed if a patient sent the completed questionnaire survey form back without ticking the check box to express consent.
After collecting 23 copies of electronic questionnaire responses, a
The researchers extracted the questionnaire responses from the database for data analysis. In addition, data about each respondent’s actual use of the app were obtained from the system log in the database. The person’s number of interactions with the smartphone health app was tracked over a 7-day period, including 3 days before and 3 days after the day of response to the questionnaire. At the time of the survey, the system log only tracked the number of times a patient submitted the blood pressure (systolic and diastolic pressure) measurement data. Therefore, the patient’s actual use of the smartphone health app was calculated as the ratio of the number of times of submitting blood pressure measurement to the recommended number of times of submission in 7 days in the management plan.
A total of 24 questionnaire items were used to measure the 11 constructs in the theoretical model. These items were drawn from the previous validated instruments. A 5-point Likert scale was used for measurement, ranging from 1, strongly disagree, to 5, strongly agree (
The questionnaire was piloted on 5 patients to test the content validity. All of the measurement items except 1, “I am able to use Blood Pressure Assistant without much effort,” were easy for the patients to understand. We modified the “effort” into “time and energy” to improve readability. The patients’ demographic information was also collected, including age, gender, and education.
The research model was tested by the partial least squares (PLS) path modeling, a well-established statistical method to model the relationship between variables in social sciences, econometrics, marketing, and strategic management [
The constructs, measurement items, and source references of the measurement items.
Construct | Item code | Measurement items | Source referencea |
Demographics | — | Age, gender, and education | — |
Perceived usefulness (PU) | PU1 | Logging or sending blood pressure values would make me cope with hypertension better | [ |
PU2 | Knowing that a doctor checks my blood pressure data gives me confidence in hypertension management | ||
PU3 | Overall, Blood Pressure Assistant is useful | ||
Perceived ease of use (PEOU) | PEOU1 | Learning how to use the mobile app would be easy for me | [ |
PEOU2 | I would find Blood Pressure Assistant easy to use | ||
PEOU3 | Blood Pressure Assistant is not cumbersome to use | ||
Social influence (SI) | SI1 | People who are important to me think that I should use Blood Pressure Assistant | [ |
SI2 | People who are important to me use Blood Pressure Assistant | ||
Usage experience (UE) | UE1 | I use smartphone to search health information on the Web | [ |
UE2 | I use mobile apps to help with managing health issues | ||
Resistance to change (RTC) | RTC1 | I do not want the mobile app to change the way I deal with hypertension | [ |
RTC2 | I do not want the mobile health app to change the way I interact with other people | ||
Perceived health threat (PHT) | PHT1 | I am aware of my hypertension condition | Drafted by authors |
PHT2 | I am very concerned about hypertension | ||
PHT3 | I would take effort to manage hypertension | ||
Self-efficacy (SE) | SE1 | I am able to use Blood Pressure Assistant without much time and energy | [ |
SE2 | I get the best value from using Blood Pressure Assistant | ||
Relationship with doctor (RWD) | RWD1 | Doctors are my most trusted source of health information | [ |
RWD2 | When I have a health concern, my first step is to contact a doctor | ||
Intention to use (ITU) | ITU1 | Given the opportunity, I would like to use Blood Pressure Assistant | [ |
ITU2 | I would consider to continuously use Blood Pressure Assistant | ||
Actual use (AU) | AU | Ratio of the actual number of measurements to the physician’s recommended number of measurements in care plan | — |
aThe symbol — denotes that the item has no source reference.
It is most suitable for models with relatively small samples in comparison with the covariance-based structural equation modeling technique [
The Ethics Committee of the study hospital claimed that since this study did not involve patient data, there was no need for an ethics audit.
There were 279 patients who used the system for more than 1 month. One hundred and fifty-two (54.5% (152/279) of them completed the questionnaire survey: 30 through telephone and 127 via electronic questionnaire. Giving 18 scale items to be tested, according to the minimum sample requirement of 5:1 subject-to-parameter, 90 questionnaire responses were sufficient for the PLS modeling [
Composite reliability (CR) and indicator reliability were used to assess the reliability of reflective constructs. All the constructs had adequate CR (ranged from 0.822 to 0.935) and indicator reliability (ranged from 0.710 to 0.976), both exceeding the recommended value of 0.70 [
The average variance extracted (AVE) of the construct was higher than the threshold of 0.50, confirming the convergent validity. AVE of each latent construct was higher than the construct’s highest squared correlation with any other latent construct (
The model was assessed by checking the significance of path coefficients (β) among the independent variables and the latent variables. The demographic variable education was excluded from modeling because of large number of missing values. The variables age and gender were found to not have any significant influence on intention to use. The results of the PLS modeling are shown in
Demographics of the participating patients.
Characteristics | n (%) | |
Male | 106 (69.7) | |
Female | 46 (30.3) | |
<30 | 5 (3.2) | |
<40 | 15 (9.9) | |
40-49 | 55 (36.2) | |
50-59 | 57 (37.5) | |
>60 | 20 (13.2) | |
<Middle school | 9 (5.9) | |
Middle school | 12 (7.9) | |
Vocational and technical education | 18 (11.8) | |
High school | 25 (16.4) | |
Three-year college | 34 (22.4) | |
University | 38 (25) | |
Missing information | 16 (10.6) | |
iPhone operating system users | 20 (13.2) | |
Android users | 132 (86.8) |
Descriptive statistics of the variables and the reliability coefficients.
Construct | Items | Mean (SD) | Standardized loading | Composite reliability |
Usage Experience (UE) | UE1 | 3.23 (1.56) | .945 | .8948 |
UE2 | 3.07 (1.77) | .835 | ||
Relationship with doctor (RWD) | RWD1 | 4.59 (0.74) | .870 | .8223 |
RWD2 | 4.40 (0.76) | .801 | ||
Perceived health threat (PHT) | PHT1 | 4.13 (0.87) | .762 | .8775 |
PHT2 | 3.29 (2.05) | .863 | ||
PHT3 | 4.35 (0.63) | .890 | ||
Perceived ease of use (PEOU) | PEOU1 | 4.58 (0.79) | .908 | .8702 |
PEOU2 | 4.26 (1.13) | .866 | ||
PEOU3 | 4.49 (0.84) | .710 | ||
Perceived usefulness (PU) | PU1 | 4.17 (1.19) | .942 | .9413 |
PU2 | 4.68 (0.55) | .944 | ||
Resistance to change (RTC) | RTC1 | 1.87 (1.25) | .921 | .8802 |
RTC2 | 1.66 (1.09) | .852 | ||
Self-efficacy (SE) | SE1 | 4.33 (1.02) | .889 | .9035 |
SE2 | 4.47 (0.62) | .926 | ||
Social influence (SI) | SI1 | 2.42 (1.98) | .944 | .9150 |
SI2 | 1.76 (1.93) | .891 | ||
Intention to use (ITU) | ITU1 | 4.53 (0.94) | .955 | .9350 |
ITU2 | 4.64 (0.56) | .976 | ||
Actual use (AU) | AU1 | 0.84 (0.13) | 1 | 1 |
A heat map showing correlations and discriminant validity. The diagonal elements denote the square root of average variance extracted, and all other elements are correlations between the constructs.
The validated theoretical model. *
With a loading factor of 0.616, perceived usefulness had a major, significant positive effect on intention to use (β=.616,
This study proposed a hybrid smartphone health TAM for chronic disease management. The model was developed based on an extensive review of the related models and theories, including TAM [
First, as hypothesized, the antecedent variables—including resistance to change, perceived health threat, relationship with doctor, usage experience, and self-efficacy—influenced the patients’ acceptance of the smartphone health technology for hypertension management, along with the traditional TAM constructs, perceived usefulness and perceived ease of use. As these factors are considered by HBM to influence patients’ engagement in health promotion behavior, therefore, our finding supports the applicability of HBM in explaining patients’ behavior in using smartphone health technology for chronic disease management.
Moreover, 0.323 of variance in the perceived usefulness was explained by 3 variables: perceived health threat, relationship with doctor, and resistance to change. First, there were cascading effects starting from perceived health threat, to perceived usefulness, and to behavioral intention. The effect of perceived health threat was also found by Kim et al [
Second, a major contribution of this study is to validate the significant influence of 2 antecedent factors, relationship with doctor and perceived health threat of hypertension, on the 3 intermittent factors for intention to use: the significant positive influences on both perceived usefulness and perceived ease of use, and strong negative influence on resistance to change. This demonstrated the vital role the health care providers play in any intervention that requires patients to self-manage their chronic diseases. In this study, the patients held highly positive evaluation of their relationships with doctors. This was suggested by their agreement with the statements that “doctors are my most trusted source of health information,” which scored 4.59 out of 5, and “When I have a health problem, my first step is to contact a doctor,” which scored 4.40 out of 5. These positive feelings were likely to be derived from the full attention and excellent service they received from the health care providers. They received 1-hour personal training from the health manager on hypertension management about how to download and use the smartphone app. If any abnormal blood pressure recording was reported, the health care providers would call the patients to discuss and adjust the hypertension management plan. These positive interactions built up rapport and patients’ trust with the health care providers. The trust could enhance the patient’s interest in using the smartphone health technology to communicate with the health care providers. Therefore, patients valued the usefulness of the technology.
This high evaluation of the relationship with doctor also led to the highly positive evaluation of the intermediate factors, perceived ease of use, and intention to use. It also strongly impeded resistance to change, with values for both items “I don’t want the mobile app to change the way I deal with hypertension” and “I don’t want the mobile health app to change the way I interact with other people” laid at the very low level between strongly disagree and disagree.
Third, the study confirmed that resistance to change indeed had a biasing effect on patients’ perception of usefulness of the smartphone technology. Its negative direct effect on behavioral intention was in accordance with that found in the middle-aged Chinese people’s acceptance of mobile health services, but not in the older people aged 60 years and above [
In accordance with the previous literature [
Perceived usefulness and perceived ease of use are 2 significant predictors of intention to use in the previous literature [
However, different from the previous studies’ findings [
Contrary to the suggestion from the previous literature [
Finally, 1 step further from the previous consumer health technology acceptance studies [
This study is, understandably, limited by its empirical scope of the study population, their social, economic, and geographic location; the smartphone health app to be used; and the type of chronic disease they suffered from. The results may vary from place to place [
The measurement of constructs can be further developed. For example, relationship with doctor may include multiple aspects, in addition to the 2 items measured in this study. Previous studies found that privacy concern is an important factor influencing patients’ acceptance of information technology [
Another limitation was the means by which the study participants were recruited. As only the patients who already used the smartphone health app Blood Pressure Assistant were recruited into the study, they were the innovative group of patient population who were likely to have a higher level of social economic status to afford to have smartphone than others; therefore, although the finding was valid for this population group, it may not be generalizable to the entire patient population even in our study site.
Only a moderate level of variation in use (0) was explained. The study captured actual use from only 1 dimension, patients’ submission of blood pressure recording. It did not capture use of other functions, such as accessing health educational information. Future research can identify other constructs influencing patients’ smartphone health technology use. The internal validity of the study was also confined by nonrespondents. Therefore, the future study can fine-tune the measurement of use. It also needs to evaluate the relationship between actual use and outcomes.
The study developed a theoretical model about patients’ acceptance of smartphone health technology for chronic disease management. It found that patients’ perceived usefulness of smartphone health technology was positively influenced by their perceived health threat, relationship with doctor, and perceived ease of use, but negatively influenced by resistance to change. Good patient-doctor relationships can alleviate patient resistance to change. Usage experience and self-efficacy positively influenced patients’ perceived ease of use. Intention to use was influenced by the enablers of perceived usefulness and perceived health threat, and the inhibitor of resistance to change. Intention to use had a significant, weak relationship with actual use.
Although the rapid growth of smartphone technology has opened new opportunities for chronic disease management, the opportunity can only be captured by the patients who accept and use the technology. The findings suggest that 3 antecedent factors, relationship with doctor, perceived health threat, and resistance to change, are important for patients’ acceptance and use of smartphone health technology. Therefore, for the successful introduction of smartphone health technology innovation for chronic disease management, efforts need to be focused on improving patient-doctor relationship and providing continuous patient education to raise awareness of the disease’s threat to health. These strategies will be effective in overcoming potential resistance to change and encouraging acceptance and use of the new technology.
actual use
average variance extracted
composite reliability
health belief model
intention to use
mobile health
perceived ease of use
perceived health threat
partial least square
perceived usefulness
resistance to change
relationship with doctor
social cognitive theory
self-efficacy
social influence
technology acceptance model
usage experience
This study was supported by the National Key Research and Development Program of China (No. 2016YFC0901703), the Science and Technology Program for Public Wellbeing of Zhejiang province, China (No. 2014H01010), and the Zhejiang Provincial Public Welfare Technology Application and Research Project of China (No. 2017C33064).
None declared.