We used multiple strategies to identify eligible studies. The goal was to capture a broad range of study designs and methodologies to capture patient experiences across diverse clinical and technological contexts. A research librarian (KK) developed and executed the initial comprehensive searches across Web of Science (Clarivate Analytics), Embase (Elsevier interface), PubMed (including records not yet indexed in Medline), CINAHL (EBSCOhost), and Library, Information Science & Technology Abstracts (EBSCOhost).

The search strategy was structured around two primary conceptual domains: (1) laboratory test results and related constructs (eg, laboratory values, blood tests, and patient-facing access through portals or electronic health records [EHRs]), and (2) patient interpretation and understanding, including health literacy and related constructs such as comprehension, numeracy, and readability. Search strategies were adapted across databases to account for differences in indexing systems, controlled vocabularies, and platform-specific syntax. Controlled vocabulary terms – including MeSH (Medical Subject Headings) in PubMed, Emtree in Embase, and CINAHL Headings—were supplemented with keyword-based strategies to ensure comprehensive retrieval by including potentially nonindexed records. We refined the search strategies iteratively over this study’s period by expanding keyword sets and incorporating additional controlled vocabulary terms.

Searches were limited to English-language, peer-reviewed empirical studies involving human participants and published between January 2013 and April 2026. Where database functionality permitted, we applied filters to exclude nonempirical publication types such as editorials, commentaries, and letters.

We also hand-searched key journals (eg, JAMIA, BMC Medical Informatics and Decision Making, Clinical Chemistry and Laboratory Medicine, JMIR, and JMIR Medical Informatics) and examined the bibliographies of related systematic reviews [18,19,23,24]. Two reviewers (ZH and MLAL) identified key papers from the initial search, manually reviewed their reference lists, and conducted citation tracking using Google Scholar.

The initial database searches were conducted in September 2022 and updated in July 2023, June 2024, and October 2025. We reran the refined search strategy across all databases and supplementary sources in April 2026 to ensure that the review was both comprehensive and current. Complete database-specific search strategies, including exact Boolean syntax, field tags, applied limits, and search dates, are provided in Multimedia Appendix 1 [8-14,25-56] in accordance with PRISMA-S reporting standards [22].

Two authors, ZH and MLAL, developed the inclusion and exclusion criteria to identify papers eligible for full-text review. We included studies if they examined any of the following: (1) patient or caregiver perceptions or experiences accessing laboratory test results through patient portals; (2) factors associated with patients’ access to, interpretation of, or understanding of laboratory test results presented through portals or EHRs; (3) presentation styles, display formats, visualization approaches, or interface designs for viewing laboratory test results online; (4) strategies or interventions designed to improve patient access to, comprehension of, or decision-making based on online laboratory test results; (5) patient use of laboratory-test-related functions within patient portals or similar systems.

Studies were excluded if they met any of the following conditions: (1) focused solely on notification or communication of laboratory test results; (2) examined general patient portal use without a specific focus on laboratory result viewing; (3) focused primarily on the perspectives of clinicians, students, or other health professionals; (4) primarily dealt with imaging or radiology reports, or with diagnostic test results (eg, COVID-19 polymerase chain reaction tests, genomic tests, or pharmacogenomic tests); (5) did not report primary empirical data (eg, editorials, commentaries, narrative reviews, and protocols).

To guide data collection, MLAL developed and piloted a structured table to extract study characteristics and outcome domains relevant to the review objectives, including comprehension, risk interpretation, emotional responses, and activation-related behaviors. Further, 2 reviewers (LA and OA) independently extracted data using this table. Discrepancies were resolved through comparison and discussion.

Following data extraction, we conducted a recursive thematic synthesis. As the included studies varied substantially in design, populations, and outcome measures, quantitative synthesis through meta-analysis was not appropriate. Moreover, key constructs such as comprehension, interpretation, emotional responses, and activation-related behaviors were operationalized in diverse and noncomparable ways, precluding calculation of a common effect size. Furthermore, no subset of studies was sufficiently comparable in terms of outcomes, measures, and study design to support a meaningful subgroup meta-analysis. The narrative synthesis was conducted and reported in accordance with guidance for synthesis without meta-analysis [63] (Checklist 2).

We developed initial codes through close reading of extracted study findings, then 2 reviewers (LA and OA) independently applied and refined these codes. Coding proceeded iteratively, with codes merged, split, or refined as patterns emerged. Discrepancies were documented and resolved through structured comparison of coded content followed by discussion. When needed, 2 additional reviewers (MLAL and ZH) reviewed disputed cases to reach consensus.

We used thematic analysis to identify recurring patterns and areas of divergence across studies. These inductively derived patterns were subsequently organized within a 3-stage continuum reflecting patients’ experiences with online laboratory test results—access, interpretation, and activation—to reflect the sequential yet interconnected nature of patients’ experiences with online laboratory test results.

Measurement strategies varied substantially across studies. Comprehension was assessed both objectively and subjectively. Overall, about one-third of studies (n=12) used objective comprehension or behavior metrics, while the remainder relied primarily on self-report. Objective measures included value-interpretation tasks assessing accuracy in identifying whether results were within or outside reference or goal ranges [25-29]. Some studies asked participants to self-rate their understanding and confidence in interpreting laboratory results [9,13,25,30] or to evaluate comprehension of medical terms and reference ranges [31]. Further, 1 study measured verbatim memory for exact numerical values and gist memory for the qualitative meaning or risk implications of results to examine how older adults understand and respond to portal-based test results [32].

Risk interpretation captures the cognitive evaluation of the seriousness or health implications of a laboratory result, whereas perceived urgency reflects the perceived need for timely action in response to that information. Both were not measured consistently in the same way, and several studies assessed them jointly rather than as separate outcomes. Risk interpretation was typically measured through risk-appraisal tasks asking participants to judge the clinical seriousness of an abnormal laboratory value—whether it required follow-up or no action based on perceived severity [33]. Additionally, 1 study used broader risk-perception items adapted from Garcia-Retamero and Cokely [64], focusing on perceived likelihood of complications [32]. Perceived urgency focused on how quickly participants believed they should respond after viewing results. It was typically assessed using Likert-type items evaluating urgency of taking action or contacting a clinician or by asking participants about intended timing of response, such as whether they would contact a clinician immediately, wait, or take no action) [14,28,34].

Emotional responses to viewing laboratory results online were typically elicited through surveys or semistructured interviews. Only a few studies used validated affective scales [30,35]. Mák et al [30] assessed anxiety using the validated Global Anxiety–Visual Analog Scale [65], which measures emotions from “no anxiety” to “worst imaginable anxiety” on a 100 mm line.

Patient activation was assessed in several ways, mostly through self-report [35-37]. Struikman et al [38] used the Patient Health Engagement, a 5-item scale, measuring patients’ psychological readiness and involvement in managing their health across 4 progressive phases—capturing emotional, cognitive, and behavioral dimensions of health engagement. Further, 2 studies [9,10] used the 10-item motivation and confidence to act subscale of the eHealth Impact Questionnaire [66] to determine the extent to which eHealth technology influences patients’ confidence in taking health-related actions. Giardina et al [35] used the Patient Activation Measure [67]—a 10-item validated scale used to assess participants’ engagement and capacity for health self-management.

Retrospective studies captured objective measures of portal engagement (eg, frequency of portal logins and laboratory test viewing), capturing both initial access and downstream follow-up behaviors [8,39,40]. User log analyses also captured behaviors such as patient-initiated provider contacts, follow-up appointments, and no-shows [11,12,37,39].

Design-focused studies captured outcomes related to usability, user satisfaction, and visual engagement with laboratory result interfaces. Usability was commonly assessed using the system usability scale [29,41,68]. The net promoter score [69], a single-item measure of patients’ likelihood to recommend the portal experience, was used in 1 study to measure satisfaction [42]. A few studies used the 13-item information and presentation subscale of the eHealth Impact Questionnaire to evaluate users’ perceptions of the clarity, usefulness, and trustworthiness of health information [9,10]. Furthermore, 1 study used eye tracking to measure fixation count, average fixation duration, and dwell time to capture how participants processed laboratory result displays [33].

Access refers to the structural and contextual conditions under which laboratory test results are made available through patient portals.

A central task in interpretation involves determining whether a test value falls outside the reference range and understanding what that deviation signifies. The absence of explanatory context was the most consistently reported barrier to understanding. In a mixed-methods study, Giardina et al [35] reported that although many interviewees said they understood the information, several struggled to interpret specific values. One participant, for example, was unsure whether the term “positive” indicated a favorable or unfavorable result. Lustria et al [25] assessed comprehension of hypothetical laboratory panels in a survey of 276 US adults. While most participants scored well on objective tests, they rated their own understanding of the values as low and preferred physician explanation—particularly when values fell outside the normal range.

Across multiple studies, participants reported difficulty determining whether values were normal or abnormal when results appeared in dense tables without plain-language explanations or clear visual indicators [13,29,42,49,51,52]. In an experiment with 1817 US adults, Zikmund-Fisher et al [28] found that only 51% correctly identified a moderately elevated hemoglobin A_1c (HbA_1c) value (8.4%) as outside the reference range when presented in a standard table format. Many participants also struggled to rate glycemic control accurately or determine whether they would contact their physician. Monkman et al [51] asked experienced portal users to review a mock laboratory result on an existing portal display; even among these active users, many failed to recognize abnormal results when values were not clearly flagged or contextualized. In iterative design workshops, Nystrom et al [29] observed confusion around HDL (high-density lipoprotein) cholesterol values, with many participants assuming that higher HDL levels were unfavorable—reflecting a broader expectation that higher numbers are undesirable, even though higher HDL is typically associated with lower cardiovascular risk.

Experimental studies further demonstrated that the format in which laboratory results are displayed can substantially influence the accuracy of interpretation. In a randomized study with older adults, Morrow et al [32] found that verbally and video-enhanced displays improved participants’ ability to interpret the meaning of results compared to graphical formats, although recall of exact numerical values did not improve. Extending this to trend displays, Monkman et al [53] found that 18 of 24 participants expressed confusion about at least 1 graph feature, and only 2 said they would seek urgent follow-up for an out-of-range value.

Not all visualization styles improved comprehension. Displays incorporating multiple reference ranges, dense formatting, or layered graphical elements often increase cognitive burden and reduce interpretive accuracy. Scherer et al [26] found that comprehension was highest when visualized test results emphasized a single, clearly defined goal range. Simpler visuals that directly stated what a result meant and whether action was needed were rated more favorably, whereas layered displays with multiple reference zones and graphical indicators reduced comprehension. In user-centered evaluations of mock portal reports, participants described lengthy layouts and technical language as overwhelming, particularly when abnormal values were not clearly emphasized [51]. Similarly, Joseph et al [54] found that supplementing trend graphs with educational content was sometimes perceived as confusing or misaligned with the test result itself.

Risk interpretation and perceived urgency represent related but distinct dimensions of interpretation. As many studies assessed these constructs using overlapping measures—asking participants to judge either the seriousness of a result or how quickly action was needed—the findings are reported together here.

Display formats influenced how patients judged the seriousness of results and the perceived need for immediate follow-up. Visual formats generally reduced overestimation of urgency for near-normal results compared with standard tables. Zikmund-Fisher et al [14] found that participants viewing standard tabular displays were more likely to judge mildly abnormal alanine transaminase and creatinine values as urgent than those viewing visual formats. Visual formats improved differentiation between mildly and clearly abnormal values but did not substantially alter perceptions of extreme results. Scherer et al [26] found that showing only goal ranges—rather than both goal and standard reference ranges—was associated with more accurate judgments of urgency. Zikmund-Fisher et al [34] found that “harm-anchor” displays reduced perceived urgency for near-normal results and decreased intentions to contact a clinician or seek hospital care for those values, without affecting responses to more extreme abnormalities.

Visualizations were not uniformly effective in reducing misclassification of risk severity. Fraccaro et al [33] tested 3 result formats (baseline, contextualized horizontal bars, or grouped) with kidney transplant recipients across low-, medium-, and high-risk scenarios. Participants had the most difficulty identifying medium-risk scenarios and frequently selected follow-up actions that did not align with recommended care. Although eye-tracking data showed longer viewing time for contextualized formats, none of the 3 displays significantly improved interpretation accuracy [33]. Morrow et al [32] similarly found that participants in the graphically enhanced condition rated lower-risk results as more serious than those in the verbally enhanced, video-enhanced, and standard conditions, with a similar pattern observed for borderline-risk results. In a think-aloud study of a portal-style potassium graph, Monkman et al [53] reported that only 2 of 24 participants said they would seek urgent follow-up for an out-of-range value, while most preferred to look up information, follow up later, or were not in a hurry to contact a provider.

Risk appraisal and comprehension varied based on different characteristics. Zikmund-Fisher et al [28] found that participants with higher numeracy were more likely to correctly identify significantly abnormal HbA_1c values and reported stronger intentions to contact their physician. Higher health literacy was associated with a lower likelihood of follow-up for mildly elevated values. Zikmund-Fisher et al [14] reported that participants with higher health and graphical literacy were better able to distinguish between mildly and highly abnormal results, particularly when results were presented using visual displays. Zhang et al [13] reported that more participants with higher health literacy and technology proficiency said the portal helped them understand their results and feel supported. Lustria et al [25] found that older adults and frequent portal users were significantly more likely to score higher on objective comprehension tests, although many self-reported difficulty understanding their results and preferred to have doctors explain them.

The action of interpreting test results presented online can evoke a range of negative and positive emotional reactions. The evidence reported here includes both studies of patient reactions to actual test results and studies using hypothetical or simulated scenarios.

Negative emotions such as anxiety, worry, and confusion were commonly reported when results were abnormal, unexpected, or presented without a clear explanation. Concern and distress were strongest when terminology was unfamiliar or when patients were uncertain whether a clinician had reviewed the result [13,31,35,36,39]. In Giardina et al [35], several participants reported anxiety or confusion even with normal results—particularly when no interpretation or guidance was included. Similarly, in a survey comparing patient portal users with nonusers, Mák et al [30] found that users who were unclear about the need for follow-up were more likely to report higher anxiety.

Participants in the study by Robinson et al [31] described physician comments accompanying laboratory results as clarifying and reassuring, particularly when comments confirmed that the result had been reviewed and did not require further action. Although uncertainty frequently elicited anxiety, reassurance was also reported when results were clearly contextualized. In 1 study, among participants with chronic conditions, those who accessed results online were significantly less likely to report high anxiety [30].

In experimental and user-centered studies using hypothetical or simulated test results, presentation styles—including language and terminology, visual cues, and other enhancements—also evoked different emotions. Zhang et al [41] found that emotionally charged phrasing such as “not optimal” was perceived as alarming. Participants preferred neutral, explanatory language placed directly alongside test results. Scherer et al [26] examined emotional responses to hypothetical HbA_1c results presented in 3 formats: as a table, a 2-color number line, or a block-style number line. They found that block-style displays using color-coded diagnostic categories (eg, red for “high” or “abnormal”) were significantly rated as more discouraging than simpler number lines or tables. Similarly, Solomon et al [27] found that red flags and warning icons elicited mixed reactions—some found them helpful; others described them as anxiety-inducing.

To make portal-released laboratory results easier to understand, more trustworthy, and more reassuring, patients frequently emphasized the need for plain-language explanations to help interpret unfamiliar or abnormal results. They suggested providing summaries that clearly state whether a value was within a healthy range and what it meant. One participant in the study by Pillemer et al [39] said it would have helped to have “a note that said normal/abnormal or something like that.” In Monkman et al [42], users similarly asked for short, condition-specific explanations that clarified the purpose of each test and its relevance to their treatment.

Patients also expressed interest in receiving tailored and contextualized explanations, personalized resource links, and tailored prompts [41]. In 1 study, patients wanted explanations tailored based on users’ literacy levels and options to receive either basic summaries or advanced content. Similarly, Monkman et al [42] found that embedded explanations tailored to the patient’s medical context could help clarify the relevance and meaning of specific laboratory values. Monkman et al [52] found that many participants wanted to combine color-coded summaries with trend graphs and tabular details and preferred being able to switch between visual and tabular formats, indicating a desire to tailor not only explanations but also the format of laboratory result displays to their needs. In Zhang et al [13], some participants proposed integrating artificial intelligence to provide contextualized explanations based on individual health data.

To increase confidence in laboratory results, participants in 1 study recommended cues indicating provider oversight—such as checkmarks signaling that their doctor had reviewed their results [50] or some kind of indication from their doctor about the need for follow-up [31]. Some participants said that labels such as “your doctor is pleased,” and simple text overlays such as “you are here” or “your doctor wants you here” could provide reassurance and guide interpretation, especially when paired with graphical displays [27,31]. Lastly, patients often expressed the need for vetted, high-quality health information alongside laboratory results to help them interpret their results [41,54].

Patients commonly described reflecting on their health, monitoring conditions, and making small behavior changes in response to viewing laboratory results online [29,31,49,50]. Hulter et al [50] reported that online access helped Dutch patients prepare for appointments, identify errors, and engage more actively in health decisions. Monkman et al [49] found that participants routinely scanned for abnormal values and tracked trends over time, describing laboratory result access as a tool for self-care and personal accountability. In Nystrom et al [29], 64% of patients said viewing their results motivated them to change their diet or exercise, and over a third intended to follow up with their provider. Portal users interviewed by Robinson et al [31] also reported that online access improved their ability and motivation to monitor their health more closely, adhere to their medications, and make lifestyle changes.

Across studies, patients frequently reported searching for additional information, preparing for medical visits, or contacting providers after receiving results online [13,35,39,55]. Christensen and Sue [55] found that patients engaged in a range of follow-up information-seeking behaviors after viewing their results, including talking with others (21%), searching for information in the portal (20%), creating graphs (19%), and looking up information online (18%). These actions were more common when providers had not clearly communicated how results would be shared. Pillemer et al [39] found that patients often used early access to look up related terms, track chronic conditions, or prepare questions—sometimes reaching out even when results were not clinically significant.

Several experiments using hypothetical or simulated scenarios found that laboratory result presentation formats influenced patients’ intentions to act or engage in their care [14,32,34,38]. In the studies by Zikmund-Fisher [14,34], patients shown mildly or near-normal abnormal results in visual formats were more likely to report they would delay follow-up or take no immediate action compared to those who viewed tabular displays. Morrow et al [32] observed that graphically enhanced presentations were associated with stronger intentions to adhere to prescribed medications, particularly for lower-risk results, although visual format had less consistent effects on intentions to change diet or exercise.

Struikman et al [38] assessed how different laboratory result display formats influenced patients’ emotional, cognitive, and motivational readiness to engage in their own care using a shortened version of the validated Patient Health Engagement scale. Engagement scores declined significantly after viewing abnormal results without context. When visuals and explanations were included, engagement remained stable for normal and partially abnormal results and declined only slightly for abnormal ones.

Talboom-Kamp et al [10] found a strong positive correlation between perceived portal usability—particularly clarity, ease of understanding, and trust—and self-efficacy. Similarly, participants who rated the portal’s information and presentation more favorably reported higher motivation and confidence to act on laboratory results (r=0.77, P<.001).

Patients often advocated for features that would help them take informed next steps. Many expressed the need for automated alerts when new results became available. In Monkman et al [42], participants recommended adding follow-up support features, such as automated notifications when results were available and links to additional resources or options to message providers. Others stressed that these alerts should do more than announce availability; participants in Giardina et al [35] suggested pairing notifications with short explanations and recommendations about what to do next.

Some patients recommended tools to support follow-up visits with their providers. Participants in Zhang et al [41] recommended adding an “ask questions” button that would help clarify test results and help them prepare personalized questions for follow-up visits. This feature would allow them to save questions directly within the portal and generate question lists for clinical visits. This, they felt, would help structure conversations and remind them about key concerns they need to discuss. In another study, patients also suggested portals should include brief follow-up instructions, lifestyle suggestions, and access to trusted health sources [13].

While portal-based delivery of laboratory results has expanded transparency, it also amplifies the interpretive work patients must do to determine what their numbers mean for their health.

To make sense of a laboratory result, patients must judge whether deviations from standard ranges are meaningful, and that judgment depends heavily on how the deviation is signaled. In portal displays, values appear alongside reference ranges, color indicators, flags, and graphical elements that denote whether a test value falls outside expected limits. Experimental studies show that these formats shape perceived significance, but not always proportionately: some clarify trends and contextualize variation, whereas others amplify minor fluctuations or highlight borderline values in ways that overstate their clinical importance [14,26,34].

For example, a value only slightly outside a reference range may be interpreted as urgent when displayed in red, while more clinically meaningful patterns may be obscured by dense or layered interfaces [26,27]. Displays that are overly complex can increase cognitive load and reduce confidence in interpretation. By contrast, overly simplified cues can shift perceptions of seriousness and urgency in directions that do not always align with clinical meaning—especially when patients must interpret results on their own [32,33].

Moreover, these preferences and effects are not uniform. Some users prefer detailed, layered information, while others find it overwhelming. What enhances clarity for one group can increase confusion or anxiety for another [13,42,49,52].

Across qualitative and survey studies, patients report that they do not simply want to know whether a value falls within a reference range; they want to understand what the results mean relative to their health condition, treatment, and broader health trajectory [13,25,49]. In response, designers of patient-facing systems are often urged to “contextualize” and “tailor” laboratory results, but these terms are frequently used interchangeably. Although both are intended to help patients make sense of their results, contextualization and tailoring are implemented in different ways and serve related, but distinct functions within laboratory result communication.

Contextualization refers to providing information that clarifies the clinical meaning of a result, for example, by explaining reference ranges, using categorical risk labels, or adding brief explanatory text [26,28]. In contrast, tailoring involves designing messages for a specific individual based on assessed characteristics related to the outcome of interest (eg, beliefs, health literacy, numeracy, or motivation), with the goal of increasing perceived personal relevance and influencing how information is interpreted and acted upon [70]. When used appropriately, tailoring has been shown to promote deeper processing and increase how seriously information is taken [71,72]. In patient-facing laboratory result systems, this is most often operationalized through computer tailoring (ie, algorithm-based or data-driven messaging), where individual data are used to generate tailored information or guidance via mechanisms such as personalization, feedback, and content adaptation [73,74].

These distinctions, however, are rarely made explicit. Many recommendations invoke “tailoring” without specifying what and how much individual details are beneficial, which elements should be adapted, or how tailoring should operate in clinically ambiguous situations. As systems, including artificial intelligence–driven tools, become capable of generating highly individualized explanations, there is a tendency to assume that greater specificity—whether through contextualizing information or tailored messaging—will improve understanding. Yet additional detail can introduce noise, elevate perceived importance, or shift interpretive weight without resolving uncertainty. When these strategies are implemented primarily through layering additional data or explanations, they risk introducing complexity and cognitive load without improving interpretive clarity or reliably supporting more meaningful engagement. The challenge, then, is not simply to provide more contextualization or more intensive tailoring, but to calibrate how contextual information and tailored messages are introduced so that they sharpen clinical meaning without overwhelming patients or unduly inflating the perceived importance of test results.

Laboratory results are often viewed in moments of uncertainty, and patients commonly feel worried, confused, or overwhelmed by raw numbers and jargon [13,31,35,39]. These emotions are heightened in the absence of contextual framing [8,30]. Interface elements intended to clarify meaning, such as color coding, flags, summary labels, and risk terms (“abnormal,” “suboptimal,” or “high risk”), influence whether a deviation feels urgent or routine; bold alerts and red blocks may heighten alarm for marginal findings, while muted or cluttered layouts can obscure patterns that warrant concern [14,26,27,34,41].

Despite extensive documentation of these emotional responses, the literature offers limited guidance on how systems should deliberately calibrate them. Human-computer interaction research provides a useful lens for this challenge. Emotional design involves deliberately structuring visual layout, language, and interaction cues to shape how users feel in moments of uncertainty [75,76]. Potential strategies include neutral summary statements that precede numerical details, restrained and well-explained use of color and flags, visible indicators that clinicians will review results within a specified timeframe, and concise descriptions of typical next steps. Such features may help assure patients that they are not alone in making sense of their results. Without this kind of emotional calibration, increased transparency may lead to confusion or distress rather than informed understanding.