Trust Before Tool
Your portal adoption problem is not a technology problem.
Healthcare organizations have spent 10 years treating portal adoption as an access issue. The premise is simple: patients don't use the digital tools because they don't have them, can't use them, or haven't been shown how. The solution, therefore, is more access, better onboarding, cleaner interfaces and better outreach.
A study published this month in Health Communication suggests the logic is wrong. Not partially wrong. Wrong at its foundation.
When researchers at UC analyzed nationally representative data from 3,475 adults, they found that patients' perception of the quality of their communication with their provider was the only factor predicting whether they would share their personal health information with a healthcare professional. It was not predictive of whether they would look at their test results in a patient portal. It was not predictive of whether they would read their clinical notes online.
Those behaviors actually depend on something else.
The portal is not the barrier. It's the relationship the patient has with the organization behind it.
What the study found
Lisa Jihyun Hwang and Rachel McKenzie drew on data from the Health Information National Trends Survey (HINTS 7), a nationally representative survey of U.S. adults conducted by the National Cancer Institute in 2024. The analytic sample of 3,475 respondents had all seen a provider in the past 12 months and had used an online medical record or patient portal.
They looked at three quite distinct digital health engagement behaviors: viewing test results via an online medical record, reading clinical notes via an online medical record and sharing health data from a personal device with a clinician. On the surface, they're all pretty much the same. In reality, they're very different.
Patient-centered communication was measured using a 7-item scale that assessed the extent to which the provider involved the patient in decision-making, responded to the patient's emotions, explained health information in an accessible way, and allowed the patient to ask questions. Institutional trust was measured as a single item (how much the patient trusts the system as a whole).
The results were very clear. Perceived patient-centered communication was a significant predictor of health data sharing (beta = .077, p = .003). It was not a significant predictor of either viewing test results or reading clinical notes. Institutional trust was a significant predictor of viewing test results (beta = .114, p = .026) but not of reading clinical notes or health data sharing.
The mediation results further clarified the picture. Institutional trust partially mediated the relationship between patient-centered communication and the viewing of test results (IE = .037, p = .028, 95% CI [.006, .111]). That same indirect pathway was not present for reading clinical notes or sharing health data.
Patient-centered communication only reaches the portal of behavior after it's passed through the filter of institutional trust.
Why these behaviors split
The mechanism they propose is definitely worth considering. To view the test results, you need access to the clinical data that the institution produces and maintains. To read the clinical notes, you need to read the documentation that the clinician has produced. The institutional infrastructure mediates both of those behaviors. The patient is interacting with the system, not the physician.
Sharing your personal health data with a provider is a different proposition altogether. It's an act of voluntarily transmitting information that you control. That behavior is more directly a function of the patient's trust in the person they are sharing data with, and that trust, in turn, is a product of the quality of the relational encounter.
The point is that digital health engagement behaviors are not equivalent. Each one is dependent on a different set of factors. Engagement with portal data is predicated on confidence in the system to hold that data. Sharing of personal health data is predicated on confidence in the person requesting it. Reading of clinical notes was not found to be significantly influenced by either of the two in this study. The authors suggest this is due to the involvement of both relational and institutional factors, with neither clearly dominant.
This is the bit that most digital health strategies completely fail to grasp. They treat portal use as a single behavior with a single lever. Access is the lever. The evidence suggests the levers are, in fact, multiple, contextual and relational in ways that access initiatives don't account for.
You can’t train your way into institutional trust. You earn it through the quality of the interactions over time.
What this means for healthcare organizations
The standard digital adoption playbook is: get the portal live, send the activation message, produce a how-to video, measure your activation rates and report on it quarterly. The underlying assumption is that if the tool is there and the patient knows about it, adoption will follow.
The Hwang and McKenzie work suggests that this playbook is targeting the wrong variable. Activation rate is a downstream metric. What's upstream of portal use (i.e., the act of looking at your own health records in a system maintained by the institution) is the patient's level of trust in that institution. Communication quality is what builds that trust. The portal is the destination. The relationship is the route.
This has very real implications for how organizations sequence their digital engagement work.
The organizations most likely to see real adoption of the portal over time aren't the ones with the most intuitive portal interface. They're the ones where patients leave their appointments feeling heard, informed, and treated as participants in their own care. The communication quality score in the HCAHPS is not just a patient experience score. That score predicts digital health behavior.
For community health centers in particular, this is worth noting. FQHC populations tend to have lower levels of baseline institutional trust in the healthcare system for historical, structural and other reasons. Digital adoption programs that fail to treat trust as a prerequisite are not failing because of bad technology. They are failing because the relational work required to enable digital adoption hasn't been undertaken.
The limits of the evidence
Hwang and McKenzie are very careful about what the data can't tell us. It's a cross-sectional study, so we can't establish any causal direction. Patients who are already portal users may well have a stronger sense of trust in the institution as a result. The mediation result should be interpreted as a statistically significant association rather than as a causal mechanism.
Institutional trust was measured with a single survey item. Patient-centered communication was assessed via self-report over the past 12 months (subject to recall bias). The trust variable was dichotomized for analytical purposes, which of course loses some of the variation that might otherwise have been present.
These are real limitations. They don't in any way detract from the direction of the finding. They serve as a warning against over-engineering a response to a single study. The evidence supports a new question rather than a definitive answer.
The right question is no longer how we increase portal activation rates. The right question is: what organizational conditions make patients willing to engage with the institution's health data?
The decision this changes
For healthcare leadership, the practical decision is about sequencing. If institutional trust mediates portal engagement, then communication quality isn't some soft metric on the side of digital strategy. It's part of digital strategy.
Performance conversations with care teams need to incorporate both communication-quality data and portal-activation data. Patient experience improvement is not a separate workstream to technology adoption -- it is a prerequisite for it. Poor digital engagement metrics are, at least in part, diagnostic of the quality of the relationship in care.
The first question to ask really is not how to improve activation rates. It's whether the patients currently trust the organization enough to engage with what the organization has about them.
A portal that a patient will not open is not a technology gap. It’s a trust gap in disguise.
Hwang, L. J., & McKenzie, R. (2026). Examining the impact of perceived patient-centered communication and institutional trust on digital health engagement behaviors. Health Communication. https://doi.org/10.1080/10410236.2026.2748945
Data: Health Information National Trends Survey (HINTS 7), National Cancer Institute, 2024. n = 3,475.
Some ideas are worth discussing in the context of your organization.
