Introduction
The first 3 years of life constitute a uniquely sensitive period for survival, physical growth, neurodevelopment, behavioral regulation, and the formation of lifelong health trajectories.[
1,
2] During infancy and toddlerhood, health problems that appear subtle in their early stages may rapidly influence nutritional status, psychomotor development, sensory function, and social-emotional wellbeing if they are not identified and addressed in time. For this reason, routine child healthcare services are regarded as a core component of pediatric prevention. These services usually include age-specific well-child visits, growth assessment, feeding and sleep guidance, developmental surveillance, immunization, anticipatory guidance for injury prevention, and early referral when developmental or health risks are suspected.[
3] When delivered consistently, routine preventive care not only benefits individual children and families but also supports broader public health goals by promoting equitable early-life health opportunities.
Despite the importance of routine preventive child healthcare, service utilization remains uneven. Even within settings where services are widely available, families differ in how regularly they attend scheduled visits, how promptly they seek vaccinations, how well they follow developmental surveillance recommendations, and how effectively they use health information to make decisions. Prior research has shown that healthcare utilization in early childhood is shaped by multiple interrelated factors, including parental education, household socioeconomic conditions, place of residence, perceived child health needs, trust in healthcare providers, accessibility of services, and the organization of healthcare delivery.[
3–
15] In many contemporary settings, however, one further layer has become increasingly relevant: the ability of caregivers to function within a digital health environment.
The digitalization of health information and service delivery has transformed the ecology of child healthcare. Parents and other primary caregivers now routinely use smartphones, social media platforms, online discussion groups, search engines, mobile applications, and hospital-based appointment systems to obtain advice on feeding, vaccination, fever management, developmental milestones, and everyday parenting concerns.[
2–
6,
8,
9] In addition, many healthcare institutions have incorporated digital interfaces into routine maternal and child health services, such as online registration, appointment booking, electronic reminders, health education modules, teleconsultation functions, and web-based access to educational materials or records.[
8,
9,
11] These developments have lowered certain access barriers and expanded opportunities for health communication, but they have also increased the demands placed on caregivers. To benefit from digitalized services, caregivers must not only possess access to devices and internet connectivity but also have the competence to search for information efficiently, distinguish reliable from misleading content, complete practical online tasks, and protect personal and family data in digital environments.
This broader capacity is commonly discussed under the umbrella of digital health literacy or eHealth literacy.[
16–
21] Earlier conceptualizations focused on a person’s ability to seek, find, understand, and appraise health information from electronic sources and to apply that knowledge to health-related decision-making.[
16,
17] Over time, the concept has expanded beyond information searching to encompass interactive and service-related abilities, including the use of digital platforms, communication with providers, privacy management, and the navigation of increasingly complex digital health ecosystems.[
18,
19,
22–
25] A growing body of literature has linked digital health literacy to health knowledge, preventive behaviors, healthcare engagement, and health outcomes in different populations.[
20,
21,
26–
31] However, most commonly used instruments were developed for general adult populations, students, or patients rather than for caregivers responsible for the health management of young children.
The caregiver context warrants specific attention. Caring for a child aged 0–3 years differs substantially from managing one’s own health in adulthood. Caregivers must interpret age-specific developmental norms, monitor rapidly changing needs, seek help for concerns that may be ambiguous, and often act as gatekeepers to preventive services on behalf of a child who cannot communicate or make independent health decisions. In this setting, digital competence is not merely an individual resource; it is a family-level capability that can shape whether information is sought early, how recommendations are interpreted, and whether timely action is taken. For example, a caregiver with strong information appraisal skills may be better equipped to verify vaccination schedules, recognize credible developmental guidance, and avoid misinformation. A caregiver with strong practical digital skills may find it easier to book visits, follow reminders, locate service sites, upload records, or communicate with service providers. By contrast, limited digital competence may increase dependence on fragmented online sources, expose families to misinformation, and intensify barriers to service navigation.
Recent studies have begun to address caregivers’ use of digital resources in pediatric contexts. Research has described how parents use mobile applications and web-based resources to seek advice on acute childhood illness, health promotion, and chronic care management.[
3,
5,
6,
8,
9] Some studies have also examined the relationship between parental eHealth literacy and parenting practices, child development, and digital engagement in pediatric health management.[
2,
4–
7] These findings suggest that digitally mediated health behavior among caregivers may have downstream implications for child wellbeing. Nevertheless, the current evidence base remains fragmented. First, many studies rely on general eHealth literacy tools that may not reflect the specific competencies required in child-rearing and routine preventive care. Second, existing studies often focus on digital resource use itself rather than on concrete utilization outcomes in formal health services. Third, there is still limited evidence from Chinese mainland concerning how caregiver digital health literacy relates to routine child healthcare participation.
The Chinese context is particularly relevant for this line of inquiry. Over the past decade, China has experienced rapid expansion of digital infrastructure, mobile internet penetration, and digital public services, including the health sector. Maternal and child healthcare institutions and community health service centers increasingly use digital registration systems, appointment booking platforms, reminder messages, and online health education channels. At the same time, China faces marked social and geographic heterogeneity in digital access, educational attainment, and service resources. Urban and rural families may differ in internet access quality, digital confidence, and familiarity with institutional digital tools. Even where smartphone ownership is nearly universal, the quality of digital engagement may vary substantially. As a result, the benefits of digitalized child health services may not be distributed evenly across caregivers.
Understanding digital health literacy as a potential enabling or constraining factor in routine child healthcare utilization therefore has both theoretical and practical significance. From a theoretical perspective, it helps extend digital health literacy research into family-centered and preventive pediatric care. From a public health perspective, it may clarify one pathway through which social inequalities are translated into differences in service engagement. If higher digital competence is associated with more consistent participation in preventive child health services, then digital literacy should be considered not only an educational issue but also a modifiable access-related factor relevant to child health equity. Conversely, if certain domains of digital competence are more strongly associated with service use than others, this may inform the design of targeted health education programs and service interfaces.
To address the measurement gap in caregiver populations, Jia and colleagues developed the Parental Digital Competence Scale (PDCS), a validated instrument specifically designed for parents or primary caregivers in childcare contexts.[
32] The PDCS includes three domains: Digital Methods Application, Digital Security, and Information Retrieval and Evaluation. These domains capture a broader competence profile than conventional information-centered eHealth literacy measures. Digital Methods Application reflects the practical use of digital functions and tools in caregiving contexts. Digital Security addresses privacy awareness and safe handling of digital information. Information Retrieval and Evaluation measures the ability to search for child health information and critically assess its trustworthiness and usefulness. The original validation study reported excellent content validity and internal consistency, making the PDCS particularly suitable for research on digital competence in families with young children.[
32]
Against this background, the present multicenter cross-sectional study investigated the association between caregivers’ digital health literacy and routine child healthcare utilization among children aged 0–3 years in three provinces of China. Specifically, we examined overall PDCS scores, domain-specific competence profiles, and their relationship with key indicators of preventive child health service use, including systematic child health management participation, timely vaccination, well-child visit adherence, and growth and developmental surveillance completion. We also explored whether caregiver sociodemographic characteristics and digital-use patterns were associated with digital competence levels. The study was guided by two main questions: first, whether higher caregiver digital health literacy is associated with better utilization of routine child healthcare services; and second, which dimensions of digital competence are most closely associated with service engagement. Given the increasing reliance on digital pathways in healthcare delivery, findings from this study may help inform caregiver support strategies and digital inclusion efforts in maternal and child health systems.
Methods
Study design
This multicenter cross-sectional study examined the association between caregivers’ digital health literacy and utilization of routine child healthcare services among children aged 0–3 years in China. A cross-sectional design was selected to allow simultaneous assessment of caregiver digital competence, family characteristics, and current preventive service utilization across multiple institutions during a defined study period. Because the design was observational and cross-sectional, all estimates were interpreted as associations rather than causal effects. The study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement for cross-sectional studies.
Study setting
The survey was conducted from March 2025 to March 2026 in Shandong, Henan, and Hubei provinces of China. Within each province, two to three tertiary maternal and child healthcare hospitals and community health service centers were selected to represent both higher-level maternal and child health institutions and community-based preventive service settings. All participating sites routinely provided child health management, well-child visits, vaccination-related services, growth monitoring, developmental surveillance, nutrition guidance, and health education.
These settings were considered appropriate because routine child healthcare in China is increasingly delivered through service pathways that combine face-to-face contact with digital elements, including mobile registration, online appointment systems, reminder messages, institutional health education platforms, and access to digital health information. Selecting institutions across different provinces and service levels improved diversity in socioeconomic context, healthcare organization, and caregiver digital-use environments.
Participants
The target population comprised primary caregivers of children aged 0–36 months who attended the selected institutions during the study period. A primary caregiver was defined as the adult family member mainly responsible for the child’s daily care and healthcare-related decisions. In most cases, this was the mother, although fathers, grandparents, and other family caregivers were eligible if they met the inclusion criteria.
Participants were eligible if they were aged 18 years or older, were the primary caregiver of a child aged 0–36 months, attended a participating institution for preventive child healthcare, child health management, or immunization-related services, were able to complete the questionnaire independently or with minimal clarification, and provided written informed consent. Caregivers were excluded if they had severe cognitive or communication barriers that prevented valid questionnaire completion, had previously participated in the survey, or submitted questionnaires that lacked key exposure or outcome information.
Sampling strategy and sample size
A stratified cluster sampling strategy was used. The three provinces constituted regional strata. Within each province, participating institutions were selected to include both tertiary maternal and child healthcare hospitals and community health service centers. Within each institution, eligible caregivers were recruited in clusters during child health management and immunization service sessions. Recruitment was conducted on different weekdays and service periods to reduce temporal selection bias and improve coverage of the regular clinic population.
The sample size was estimated using the single-population proportion formula, n = Z2p (1-p)/d2. Because the study included several dichotomous utilization indicators and precise prior estimates were unavailable for all outcomes, the conservative assumption of P = 0.50 was used. With a 95% confidence level and an absolute precision of 0.04, the minimum sample size was approximately 600. After accounting for the cluster-based sampling design with a design effect of 2.0–2.5, the required sample size was approximately 1,200-1,500. To allow for invalid questionnaires, at least 1,400 caregivers were approached. The final analytic sample included 1,326 valid questionnaires, yielding an effective response rate of 94.7%. The clustered sampling structure was accounted for in the regression analyses as described below.
Study instruments
Three instruments were used: a general information questionnaire, the PDCS, and a structured questionnaire on routine child healthcare utilization.
General information questionnaire
The general information questionnaire collected caregiver, child, household, and digital-use characteristics considered relevant to digital competence and preventive service engagement. Caregiver variables included age, sex, educational level, relationship to the child, employment status, place of residence, and household income. Child-related variables included age, sex, health insurance status, and general health condition. Digital-use variables included duration of digital device use, smartphone ownership, and frequency of online child health information seeking. These variables were measured because previous research and the conceptual model suggest that socioeconomic position, digital access, and child-related need may be associated with both caregiver digital competence and routine child healthcare utilization.
Parental Digital Competence Scale
Digital health literacy was assessed using the PDCS, a childcare-specific instrument developed and validated for parents or primary caregivers. The PDCS contains 23 items across three domains: Digital Methods Application (14 items), Digital Security (5 items), and Information Retrieval and Evaluation (4 items). Each item is scored on a 5-point Likert scale, with higher scores indicating stronger digital competence; the total score ranges from 23 to 115. The total summed score was used to represent overall digital competence. Because the three domains differ in length, mean item scores were used for domain-level description and domain-specific modelling.
Digital Methods Application reflects practical use of digital tools and functions in caregiving contexts. Digital Security reflects privacy awareness and data-safety practices in digital environments. Information Retrieval and Evaluation reflects the ability to search for child health information and judge its credibility, relevance, and usefulness. The scale was used in its validated 14/5/4 structure without modification. Internal consistency reliability of the PDCS in the present sample was assessed using Cronbach’s alpha for the total scale and for each of the three domains. Values of 0.70 or above were considered acceptable for group-level research. The observed alpha coefficients are reported in the Results section.
The PDCS was selected rather than a general eHealth literacy instrument such as the eHealth Literacy Scale because the present study focused on caregivers’ digital competence in childcare and routine child healthcare contexts. The eHEALS mainly assesses the ability to seek, understand, and appraise electronic health information in general populations, whereas the PDCS captures a broader parental competence profile, including practical Digital Methods Application, Digital Security, and Information Retrieval and Evaluation in caregiving-related situations. Therefore, the PDCS was considered more conceptually aligned with the caregiver role and the digitalized child healthcare service context examined in this study.
Routine child healthcare utilization questionnaire
Routine child healthcare utilization was assessed using a structured questionnaire based on core preventive child health service indicators for children aged 0–3 years. Four indicators were included: participation in systematic child health management, timely vaccination, well-child visit adherence rate, and completion of growth and developmental surveillance. These indicators were selected because they represent related but distinct aspects of preventive care engagement, including organized participation, timeliness, continuity, and developmental monitoring.
Systematic child health management referred to caregiver-reported participation in scheduled child health management activities appropriate for the child’s age. Timely vaccination referred to completion of age-appropriate immunization without substantial delay according to caregiver report and, when accessible during the survey, vaccination booklets, child health records, or electronic service records. Well-child visit adherence reflected the proportion of recommended visits completed. Completion of growth and developmental surveillance reflected whether scheduled surveillance activities had been completed during the relevant period.
Because record verification was not available for all participants, utilization outcomes were treated as caregiver-reported or partially record-assisted indicators rather than fully record-verified measures. This measurement approach was considered feasible for a multicenter field survey but was recognized as a potential source of recall bias and social desirability bias.
Pilot testing and quality control
Before the formal survey, a pilot study involving 50 caregivers was conducted to evaluate wording, response burden, field feasibility, and the logistical flow of questionnaire administration. Minor wording adjustments were made after the pilot to improve clarity and reduce ambiguity. A short-term retest in a subsample showed an intraclass correlation coefficient of at least 0.85, indicating acceptable stability of survey administration.
Investigators from all sites received standardized training on eligibility assessment, informed consent, questionnaire administration, on-site checking, and confidentiality procedures. Questionnaires were reviewed immediately after completion to reduce missing data and obvious inconsistencies. Data entry was performed using a double-entry process, and discrepancies were resolved against the original forms. This combination of pilot testing, uniform investigator training, on-site review, and double data entry was used to strengthen data quality across sites.
Missing data
The extent and pattern of missing data were examined before statistical analysis. Questionnaires were excluded from the analytic sample if they lacked key eligibility information, had missing data on major utilization outcomes, or had substantial missingness on the PDCS. Substantial PDCS missingness was defined as more than 20% missing responses on the total scale or missingness sufficient to prevent calculation of a domain score.
For retained questionnaires, item-level missingness was handled according to the amount and location of missing data. If fewer than 10% of items were missing within a PDCS domain, missing items were replaced using the participant-specific mean of completed items within the same domain. If 10% or more of items were missing within a domain, the relevant domain score and total score were treated as missing. Complete-case analysis was used for multivariable regression models. The number of excluded questionnaires and the extent of missingness were summarized in the Results section.
Statistical analysis
To examine whether the association between caregivers’ digital competence and routine child healthcare utilization was consistent across demographic groups, subgroup analyses were conducted by place of residence and educational attainment. Stratified models were fitted for urban and rural caregivers, and for caregivers with college-level education or above versus those with lower educational attainment. The same covariates as in the main models were included where model stability permitted, except that the stratification variable was not included in the corresponding stratified model. Interaction terms between total PDCS score and residence or educational attainment were additionally examined to assess potential effect modification.
Data were analyzed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables were summarized as means and standard deviations or medians and interquartile ranges, depending on distributional characteristics. Categorical variables were summarized as frequencies and percentages. Group differences in PDCS scores across participant characteristics were examined using independent-samples t tests or one-way analysis of variance when assumptions were met, and non-parametric tests were used when appropriate. Chi-square tests were used for categorical comparisons.
Internal consistency reliability of the PDCS was evaluated using Cronbach’s alpha for the total scale and each domain. Bivariate correlations between PDCS scores and utilization indicators were examined as preliminary analyses using Pearson or Spearman correlation coefficients, as appropriate. However, adjusted multilevel models were treated as the primary analyses because they accounted for covariates and the clustered sampling design.
Because participants were recruited through a stratified cluster sampling design, clustering by study site was explicitly evaluated. Site-level intraclass correlation coefficients (ICCs) were estimated for the major binary utilization outcomes using null multilevel logistic models with a random intercept for study site. Multilevel logistic regression models with a random intercept for study site were then fitted to estimate the adjusted association between PDCS scores and binary utilization outcomes. Province was considered in model specification to reflect the regional strata and to reduce confounding by regional service context. As a sensitivity analysis, models with cluster-robust standard errors or study-site fixed effects were examined to assess whether the direction and statistical significance of the findings were robust to different approaches for handling clustering.
The main binary outcomes included participation in systematic child health management, timely vaccination, and completion of growth and developmental surveillance. For well-child visit adherence rate, which was measured as a continuous percentage, linear mixed-effects models were used. If adherence was dichotomized for sensitivity analysis, the same multilevel logistic modelling approach was applied.
PDCS total score was first entered as the key independent variable. Domain-specific models were then fitted with Digital Methods Application, Digital Security, and Information Retrieval and Evaluation entered as domain-level mean item scores. Total-score and domain-specific models were estimated separately to avoid redundancy between the overall scale score and its component domains. Adjusted odds ratios (AORs), 95% confidence intervals (CIs), and P values were reported for logistic models. To improve interpretability, PDCS associations were reported both per one-point increase and per one standard deviation increase in PDCS score when appropriate.
Covariates were selected a priori based on the conceptual assumption that caregiver socioeconomic position, digital access and use, and child-related need may be associated with both caregiver digital competence and routine child healthcare utilization. Adjusted models included caregiver age, sex, educational level, relationship to the child, employment status, place of residence, household income, frequency of online child health information seeking, child age, child sex, health insurance status, and general child health condition, subject to model stability and missingness. Smartphone ownership was described but was not included as a primary covariate if variability was insufficient. Multicollinearity among covariates was assessed using variance inflation factors, and model diagnostics were examined where applicable.
All statistical tests were two-sided, and a P value of less than 0.05 was considered statistically significant. Given the cross-sectional design, the results were interpreted as associations only. No causal or temporal interpretation was made from the regression estimates.
Ethical considerations
The study was approved by the Ethics Committee of Shandong Vocational and Technical University of International Studies (Approval No. 2025012301). This study was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent before enrollment. Participation was voluntary, and respondents were informed that refusal or withdrawal would not affect the services received by them or their children. Questionnaire data were anonymized before analysis and were handled confidentially throughout the study.
Results
Participant characteristics
A total of 1,400 caregivers were approached during the study period. After exclusion of 74 incomplete or otherwise invalid questionnaires, 1,326 valid responses were included in the final analytic sample, yielding an effective response rate of 94.7%. The mean caregiver age was 30.6 ± 4.8 years. Most respondents were female (92.3%), 68.5% had college-level education or above, and 61.2% resided in urban areas. The mean child age was 16.8 ± 10.3 months, and 51.7% of the children were male. Most children had health insurance coverage (94.2%). As shown in Table 1, smartphone ownership was nearly universal among participating caregivers (98.7%), and 76.4% reported seeking child health information online at least weekly.
Digital competence scores and internal consistency reliability
The mean total PDCS score was 86.74 ± 12.36, with observed scores ranging from 27 to 115. At the domain level, the mean item score was 3.78 ± 0.56 for Digital Methods Application, 3.69 ± 0.61 for Digital Security, and 3.52 ± 0.68 for Information Retrieval and Evaluation. Internal consistency reliability was acceptable to good in the present sample. Cronbach’s alpha was 0.89 for the total PDCS, 0.85 for Digital Methods Application, 0.78 for Digital Security, and 0.81 for Information Retrieval and Evaluation. McDonald’s omega for the total scale was 0.90 (Table 2).
Routine child healthcare utilization
Routine child healthcare utilization indicators are summarized in Table 3. Participation in systematic child health management was reported for 959 children (72.3%). Timely vaccination was reported for 1,041 children (78.5%). The mean well-child visit adherence rate was 79.6% ± 18.3%. Completion of growth and developmental surveillance as scheduled was reported for 983 children (74.1%).
Intraclass correlation coefficients
Because the study used a stratified cluster sampling design, ICCs were estimated for the main utilization outcomes at the study-site level before fitting cluster-adjusted models. The ICCs were 0.032 for systematic child health management participation, 0.028 for timely vaccination, 0.041 for growth and developmental surveillance completion, and 0.035 for well-child visit adherence rate. These values indicated low to low-moderate clustering by study site (Table 4).
Cluster-adjusted associations between total PDCS score and routine child healthcare utilization
Multicollinearity diagnostics indicated no severe multicollinearity among the variables included in the final adjusted models. The variance inflation factor values ranged from 1.03 to 1.86, all below the commonly used threshold of 5.0.
Table 5 presents the adjusted associations between total PDCS score and routine child healthcare utilization after accounting for the cluster sampling design. In multilevel logistic regression models, each one-point increase in total PDCS score was associated with higher odds of systematic child health management participation (AOR = 1.025, 95% CI: 1.008–1.043, P = 0.005). Expressed per one standard deviation increase in PDCS score, the corresponding AOR was 1.36 (95% CI: 1.10–1.68).
Total PDCS score was also associated with timely vaccination. Each one-point increase in PDCS score was associated with higher odds of timely vaccination (AOR = 1.021, 95% CI: 1.003–1.039, P = 0.023), corresponding to an AOR of 1.28 (95% CI: 1.03–1.59) per one standard deviation increase. The association with growth and developmental surveillance completion was positive but did not reach conventional statistical significance after accounting for clustering (AOR = 1.018, 95% CI: 0.999–1.037, P = 0.061). In the linear mixed model for well-child visit adherence rate, each one-point increase in total PDCS score was associated with a 0.32 percentage-point higher adherence rate (β = 0.32, 95% CI: 0.08–0.56, P = 0.009).
Domain-specific associations with systematic child health management participation
Domain-specific multilevel logistic regression models were fitted to examine whether different components of the PDCS showed differential associations with systematic child health management participation after accounting for the clustered sampling design. The three PDCS domain scores were entered as mean item scores, and the models adjusted for caregiver and child covariates as specified in the main analysis.
As shown in Table 6, Information Retrieval and Evaluation was significantly associated with systematic child health management participation (AOR = 1.142, 95% CI: 1.048–1.245, P = 0.002). Digital Methods Application was also significantly associated with this outcome (AOR = 1.075, 95% CI: 1.013–1.141, P = 0.018). By contrast, Digital Security was not significantly associated with systematic child health management participation in the adjusted domain model (AOR = 1.028, 95% CI: 0.956–1.105, P = 0.465).
Sensitivity analyses
Sensitivity analyses were conducted to examine the robustness of the primary findings to alternative approaches for handling clustering and missing data. Models using cluster-robust standard errors produced results that were substantively similar to those of the primary multilevel models. Complete-case analyses also yielded findings consistent with the main analyses. In addition, models including province fixed effects generated estimates comparable to those from the primary cluster-adjusted models.
Across all sensitivity analyses, the direction of the association between total PDCS score and the main routine child healthcare utilization outcomes remained consistent. These analyses supported the robustness of the principal findings, while retaining the same interpretation that the association with growth and developmental surveillance completion was positive but did not reach conventional statistical significance after adjustment for clustering.
Exploratory subgroup analyses for systematic child health management participation
Exploratory subgroup analyses were conducted for systematic child health management participation, the primary binary utilization outcome, to examine whether the association between total PDCS score and service utilization was consistent across residence and educational attainment groups. In stratified analyses, the association between total PDCS score and systematic child health management participation remained positive in both urban and rural caregivers, although statistical significance was observed only among urban caregivers. Specifically, each one-point increase in total PDCS score was associated with higher odds of systematic child health management participation among urban caregivers (AOR = 1.028, 95% CI: 1.009–1.048, P = 0.004), whereas the corresponding association among rural caregivers was positive but not statistically significant (AOR = 1.019, 95% CI: 0.993–1.045, P = 0.152). The interaction between PDCS score and residence was not statistically significant (P for interaction = 0.583).
A similar pattern was observed in analyses stratified by educational attainment. Among caregivers with college-level education or above, higher total PDCS score was significantly associated with systematic child health management participation (AOR = 1.026, 95% CI: 1.006–1.047, P = 0.011). Among caregivers with lower educational attainment, the association was also positive but did not reach statistical significance (AOR = 1.021, 95% CI: 0.988–1.055, P = 0.218). The interaction between PDCS score and educational attainment was not statistically significant (P for interaction = 0.742). These exploratory findings suggest that the direction of association was generally consistent across key demographic subgroups, although estimates in rural caregivers and those with lower educational attainment should be interpreted cautiously because of reduced subgroup sample size and wider CIs (Table 7).
Discussion
Principal findings
This multicenter cross-sectional study examined the association between caregivers’ digital health literacy and routine child healthcare utilization among children aged 0–3 years in three provinces of China. After accounting for the stratified cluster sampling design, higher total scores on the PDCS remained positively associated with several indicators of routine child healthcare utilization. Specifically, a higher total PDCS score was associated with greater odds of participation in systematic child health management and timely vaccination, and with higher well-child visit adherence. The association with growth and developmental surveillance completion was in the expected positive direction but did not reach conventional statistical significance after adjustment for clustering. These findings suggest that caregiver digital competence is related to several, but not all, dimensions of preventive child healthcare engagement.
The revised analyses also clarified the magnitude and robustness of the associations. The site-level ICCs were low to low-moderate, indicating that clustering by recruitment site was present but limited. Nevertheless, the use of multilevel models was appropriate given the sampling design. Compared with the original unclustered models, the cluster-adjusted estimates were slightly attenuated but remained directionally consistent. When expressed per one standard deviation increase in the total PDCS score, the AORs were more interpretable than the per-point estimates: a one-SD higher PDCS score was associated with 36% higher odds of systematic child health management participation and 28% higher odds of timely vaccination. These effect sizes are modest but potentially meaningful at the population level, especially given the large number of families involved in routine early childhood preventive care.
Domain-specific analyses further indicated that the association was not evenly distributed across all aspects of digital competence. Information Retrieval and Evaluation and Digital Methods Application were significantly associated with systematic child health management participation, whereas Digital Security was not. This pattern supports the view that the digital skills most closely connected with routine service navigation are those that enable caregivers to find reliable child health information, judge its relevance, and complete practical digital tasks in healthcare settings. At the same time, the non-significant result for Digital Security suggests that not every dimension of digital competence has the same relationship with preventive service utilization.
Interpreting digital health literacy as an access-related capability
The findings contribute to an expanding literature that conceptualizes digital health literacy as a determinant of healthcare access and engagement rather than merely as a general technological skill.[
20,
21,
26–
31] In contemporary maternal and child health systems, caregivers often encounter preventive services through a hybrid pathway that combines in-person visits with digital reminders, online registration systems, appointment platforms, electronic health education materials, social media information, and provider communication channels. In this context, caregiver digital competence may shape how families identify recommended services, understand the timing of visits, respond to reminders, and translate digital information into service attendance.
This interpretation is particularly relevant for the first three years of life, when preventive care is structured around repeated contacts and age-specific monitoring. Unlike acute care, routine child healthcare requires caregivers to engage with services even when the child appears healthy. Families must understand why scheduled visits, vaccination timeliness, growth monitoring, developmental surveillance, and anticipatory guidance are important before visible problems arise. Caregivers with stronger digital competence may be better positioned to locate official schedules, distinguish credible advice from informal or misleading content, use digital appointment systems, and follow service instructions. These skills may reduce friction in service navigation and make it easier for caregivers to maintain continuity of preventive care.
However, the observed associations should not be interpreted as evidence that digital competence directly causes higher service utilization. The cross-sectional design precludes conclusions about temporal ordering. The relationship may be bidirectional. Caregivers who already attend routine child healthcare services more consistently may have more opportunities to interact with digital registration systems, receive official reminders, access institutional health education materials, and communicate with providers through online channels. These repeated interactions may strengthen their digital competence over time. Conversely, caregivers with higher digital competence may find it easier to use service platforms and follow preventive recommendations. The association observed in this study may therefore reflect a feedback loop between digital competence and healthcare engagement rather than a one-way effect.
Domain-specific interpretation of the PDCS findings
The strongest domain-specific association was observed for Information Retrieval and Evaluation. This finding is plausible because caregivers of infants and toddlers often face a high volume of online information about feeding, sleep, vaccination, fever management, developmental milestones, injury prevention, and common childhood symptoms. The quality of such information varies widely across search engines, social media, parent discussion groups, commercial platforms, and institutional sources. Caregivers who are better able to search efficiently, compare sources, evaluate credibility, and judge whether information is applicable to their child may be more likely to understand and follow preventive service recommendations.
Information appraisal may be especially important because routine child healthcare is anticipatory and schedule-dependent. Vaccination and developmental surveillance require caregivers to act according to age-based recommendations rather than immediate symptoms. If caregivers encounter misinformation about vaccination safety, misinterpret developmental advice, or rely primarily on informal sources, they may delay or miss recommended services. In contrast, stronger retrieval and evaluation skills may help caregivers identify authoritative guidance from maternal and child health institutions, public health agencies, or healthcare professionals. This may explain why this domain showed a clear association with systematic child health management participation.
Digital Methods Application was also significantly associated with systematic child health management participation. This domain reflects the practical ability to use digital tools and functions in caregiving contexts. Such skills are increasingly relevant because many child health services now require or encourage caregivers to use online appointment booking, mobile registration, electronic reminders, digital health education modules, vaccination scheduling tools, or hospital-based service platforms. A caregiver may recognize the importance of preventive care but still experience difficulty if appointment systems are difficult to use, instructions are unclear, or digital processes are fragmented across platforms. Stronger practical digital skills may therefore reduce administrative and logistical barriers to service use.
In contrast, Digital Security was not significantly associated with utilization in the adjusted domain model. This result should not be interpreted as evidence that Digital Security is unimportant. Rather, it may indicate that security-related awareness is less directly connected to the specific utilization outcomes examined in this study. Participation in systematic child health management or timely vaccination may depend more immediately on information appraisal and service navigation than on privacy management. In addition, the Digital Security domain contains fewer items than Digital Methods Application, which may limit variability and statistical power. It is also possible that Digital Security becomes more relevant in contexts involving sensitive electronic records, teleconsultation, app-based monitoring, or data sharing across institutions. Future studies should examine whether the role of Digital Security changes as pediatric digital health systems become more integrated and interactive.
Effect size and practical significance
Although the per-point AORs for the total PDCS score were small, this should be interpreted in relation to the scale range and observed variability. A one-point increase on a 23-item scale represents a relatively small change in overall digital competence. When expressed per one standard deviation increase, the associations were more interpretable: higher PDCS scores were associated with greater odds of systematic child health management participation and timely vaccination. These findings suggest that caregiver digital competence may contribute to routine service engagement, although it is unlikely to determine utilization on its own.
Routine child healthcare utilization is shaped by multiple factors, including caregiver education, household income, residence, provider communication, service accessibility, institutional organization, cultural beliefs, and family support. Therefore, digital competence should be understood as one access-related capability within a broader service environment. Even caregivers with strong digital skills may face barriers such as distance to facilities, long waiting times, fragmented information systems, or limited trust in providers, whereas caregivers with lower digital competence may still maintain good service use if they receive strong offline support.
The non-significant association with growth and developmental surveillance may reflect the domain-specific nature of this service. Compared with vaccination or scheduled well-child visits, growth and developmental surveillance may depend less on caregivers’ independent digital navigation and more on provider assessment, local service capacity, standardized screening procedures, and follow-up arrangements. Some surveillance activities may also be embedded within routine visits and initiated by healthcare providers rather than actively requested by caregivers through digital platforms. In addition, caregivers may differ in their understanding of what constitutes developmental surveillance, which could introduce measurement variability. These characteristics may explain why digital health literacy showed a weaker and statistically non-significant association with this outcome.
Measurement of service utilization and possible bias
Several utilization outcomes in this study were based primarily on caregiver report, although available records were used to assist confirmation when accessible. This measurement approach is feasible in multicenter field surveys but introduces potential bias. Recall bias may occur if caregivers cannot accurately remember the timing or completion of well-child visits, vaccination, or surveillance activities. Social desirability bias may also lead caregivers to overreport recommended behaviors, particularly when questionnaires are administered in healthcare settings where such behaviors are expected and valued.
Differential reporting bias may also have occurred. Caregivers with higher digital health literacy may be more familiar with recommended child healthcare practices, such as timely vaccination, scheduled well-child visits, and participation in systematic child health management. Because they may know what constitutes the expected or “correct” health behavior, they may be more likely to report service utilization in a socially desirable direction, especially when the questionnaire is administered in healthcare settings. If such overreporting was stronger among caregivers with higher PDCS scores, the observed associations between digital health literacy and service utilization may have been overestimated. Therefore, the findings should be interpreted as associations based on caregiver-reported or partially record-assisted outcomes rather than as definitive evidence of actual service completion.
The direction of this bias is important. If overreporting of service utilization occurred similarly across levels of digital competence, the associations may have been diluted. However, if caregivers with higher digital competence were more likely to know the expected answers or to present themselves as compliant with health recommendations, the observed associations may have been overestimated. The magnitude of such bias cannot be determined from the present data. Future studies should link caregiver survey responses with immunization registries, child health management records, electronic appointment systems, or institutional service databases to improve outcome measurement. Partial or full record verification would also allow researchers to distinguish between actual service completion, caregiver recall, and perceived adherence.
Self-report may be particularly problematic for outcomes that require precise timing, such as timely vaccination, and for outcomes that may be less familiar to caregivers, such as developmental surveillance. Some caregivers may interpret developmental surveillance broadly as any growth or health check, while others may report only formal developmental screening. Such variation in interpretation could contribute to measurement error. Clearer operational definitions and record-based confirmation would strengthen future research on the relationship between digital competence and routine child healthcare utilization.
Confounding, selection bias, and generalizability
The adjusted models included key sociodemographic and child-related covariates, but residual confounding remains possible. Factors such as caregiver beliefs about preventive care, perceived child vulnerability, trust in healthcare providers, previous healthcare experiences, provider communication quality, family support, distance to facilities, waiting time, service cost, and local health-system organization were not fully measured. These variables may influence both digital competence and service utilization. For example, caregivers who trust healthcare providers may be more likely to use official digital platforms and more likely to attend routine visits. Similarly, families living near well-organized health facilities may have both greater exposure to digital service tools and better utilization of preventive care. Because such factors were not comprehensively captured, the observed associations should be interpreted as adjusted associations rather than causal effects.
Selection bias is another important consideration. Participants were recruited from caregivers attending maternal and child healthcare hospitals and community health service centers. This sampling frame may overrepresent families already connected to formal health services and underrepresent those who are disengaged, geographically remote, socioeconomically disadvantaged, migrant, or less confident in using healthcare systems. As a result, the observed levels of service utilization and digital competence may be higher than those in the broader population of caregivers of children aged 0–3 years. The high proportion of caregivers with college-level education and the near-universal smartphone ownership further indicate that the sample was relatively advantaged in educational and digital access terms.
Generalizability should therefore be considered with caution. In rural or remote populations, lower socioeconomic groups, families with unstable internet access, caregivers with limited formal education, and non-clinic-attending caregivers, the relationship between digital competence and service utilization may differ. Barriers may be stronger, and digitalization may either facilitate access or widen inequity depending on how services are designed. The present findings are most applicable to caregivers who have at least some contact with routine child healthcare institutions and who live in settings where digital tools are already embedded in service delivery. Studies using community-based sampling or household surveys are needed to better understand families who do not regularly attend healthcare facilities.
Implications for maternal and child health practice
Despite these limitations, the findings have several practical implications. First, caregiver-oriented digital health literacy may be considered as part of routine maternal and child health education. Existing health education often focuses on child feeding, vaccination, disease prevention, and developmental milestones. These topics remain essential, but caregivers may also need support in identifying trustworthy digital information, interpreting online advice, using official service platforms, and responding appropriately to digital reminders. Brief digital navigation guidance could be incorporated into routine well-child visits, vaccination appointments, or community health education sessions.
Second, health institutions should not assume that smartphone ownership is equivalent to effective digital access. In this sample, smartphone ownership was almost universal, yet PDCS scores varied and were associated with service utilization. This pattern highlights the difference between device access and functional digital capability. Maternal and child health providers should consider whether caregivers can actually use digital registration systems, appointment platforms, reminder messages, electronic records, and online health education materials. Simple step-by-step instructions, plain-language interface design, consistent official communication channels, and staff support for digital tasks may reduce avoidable barriers.
Third, the domain-specific findings suggest that information appraisal and practical service navigation may be especially useful targets for intervention. Health institutions could curate authoritative child health resources, provide guidance on how to identify reliable online information, explain how to compare online content with professional advice, and clarify when digital information should prompt formal consultation. They could also simplify digital service procedures and maintain alternative non-digital access channels for caregivers who are less comfortable with online systems. Such measures are important to ensure that digital transformation supports equity rather than replacing accessible offline care.
Fourth, healthcare professionals may play a mediating role between digital information environments and formal preventive care. Caregivers frequently move between institutional recommendations and informal online content. During routine visits, providers could ask caregivers where they obtain child health information, whether they have encountered confusing or contradictory advice, and whether they have difficulty using digital service functions. These brief conversations may help identify families who need additional support and may strengthen the connection between digital information use and appropriate preventive service engagement.
Strengths and limitations
This study has several strengths. The multicenter design across three provinces increased contextual diversity and reduced the likelihood that the findings reflected only one institution or local service model. The inclusion of both tertiary maternal and child healthcare hospitals and community health service centers captured different levels of routine child healthcare delivery. The use of the PDCS, a caregiver-specific measure of digital competence, improved conceptual alignment between the exposure and the childcare context. The revised analyses also strengthened the methodological rigor of the study by reporting study-sample reliability, estimating ICCs, and accounting for clustering in the regression models.
The limitations should also be emphasized. First, the cross-sectional design prevents causal inference and does not establish whether digital competence precedes service utilization. Second, several utilization indicators were self-reported or only partially record-assisted, which may introduce recall and social desirability bias. Third, participants were recruited from healthcare facilities, which may limit representativeness and may have selected caregivers who were already more engaged with formal services. Fourth, the sample had relatively high educational attainment and near-universal smartphone ownership, limiting generalizability to underserved groups. Fifth, residual confounding remains possible because variables such as health beliefs, trust in providers, service accessibility, waiting time, provider communication, and local health-system characteristics were not fully measured. Sixth, the domain-specific model reported here focused on systematic child health management participation; additional outcome-specific domain analyses would be useful in future work.
Future research
Future research should extend this study in several directions. Longitudinal studies are needed to clarify the temporal sequence between caregiver digital competence and routine child healthcare utilization. Such studies could examine whether baseline digital competence predicts subsequent service use, whether repeated service use improves digital competence, or whether both processes operate simultaneously through a feedback loop. Intervention studies could test whether digital literacy education, guided platform use, official information-resource curation, or combined online-offline navigation support improves preventive service continuity.
Future studies should also prioritize more objective measurement of service utilization. Linking caregiver survey data with immunization records, electronic child health records, appointment systems, or regional child health management databases would reduce reliance on self-report and permit more precise assessment of timeliness, continuity, and completion. Qualitative and mixed-methods studies may further clarify how caregivers interpret online child health information, how they decide which sources to trust, what difficulties they encounter in digital service systems, and how provider communication affects their digital engagement.
Finally, research should include populations that may be underrepresented in facility-based surveys, including rural caregivers, migrant families, grandparents, low-income households, caregivers with limited education, and families who do not regularly attend routine child healthcare services. These groups may face different barriers and may benefit from different forms of support. Understanding digital competence in more diverse caregiving contexts will be essential for designing equitable digital maternal and child health services.
Summary of interpretation
In summary, the present study found that caregiver digital health literacy was positively associated with several indicators of routine child healthcare utilization among children aged 0–3 years. After accounting for clustering, higher total PDCS scores remained significantly associated with systematic child health management participation, timely vaccination, and well-child visit adherence, while the association with growth and developmental surveillance completion was positive but not statistically significant. Domain-specific findings suggest that Information Retrieval and Evaluation and Digital Methods Application are more closely related to service engagement than Digital Security. These findings support the relevance of caregiver digital competence in digitalized child health service environments, but they should be interpreted as associations rather than causal effects. Strengthening caregiver digital capability and designing inclusive digital service systems may be important components of efforts to support equitable and continuous routine child healthcare.
Conclusion
This multicenter cross-sectional study examined the association between caregivers’ digital health literacy and routine child healthcare utilization among children aged 0–3 years in three provinces of China. After accounting for the stratified cluster sampling design, higher total PDCS scores remained positively associated with several indicators of routine child healthcare utilization, particularly participation in systematic child health management, timely vaccination, and well-child visit adherence. The association with completion of growth and developmental surveillance was in the expected positive direction but did not reach conventional statistical significance after adjustment for clustering.
Domain-specific analyses further suggested that not all dimensions of digital competence were equally related to service utilization. Information Retrieval and Evaluation and Digital Methods Application showed significant positive associations with systematic child health management participation, whereas Digital Security was not significantly associated. These findings indicate that caregivers’ ability to identify, evaluate, and apply digital child health information, as well as their practical ability to use digital service tools, may be more directly relevant to routine preventive care engagement than general awareness of digital safety.
The findings should be interpreted as associations rather than causal effects. Because of the cross-sectional design, it is possible that caregivers with stronger digital competence are better able to navigate child health services, but it is also possible that more frequent contact with routine healthcare services improves caregivers’ familiarity with digital platforms, official health information, and appointment systems. In addition, some utilization outcomes were based on caregiver report and may be affected by recall bias or social desirability bias.
Despite these limitations, the study highlights digital health literacy as a potentially important access-related factor in early childhood preventive healthcare. In increasingly digitalized maternal and child health systems, improving caregiver-oriented information appraisal skills, simplifying digital service pathways, and maintaining alternative support channels for caregivers with lower digital confidence may help promote more inclusive service navigation. Future longitudinal and intervention studies are needed to clarify the temporal relationship between caregiver digital competence and child healthcare utilization, and to determine whether targeted digital literacy support can contribute to more equitable and continuous preventive care for young children.
The Author(s) 2026. This article is published by Higher Education Press at journal.hep.com.cn.
PDF
(380KB)
