Background

The diagnostic process is a complex, iterative, and collaborative activity with a goal to reduce diagnostic uncertainty and develop a more precise and complete understanding of a patient’s health problem [1]. Measuring and improving this complex process and its outcomes has been recognized as a priority only recently, so that diagnostic quality, in comparison, for example, with treatment, has been long overlooked [2]. Diagnostic excellence, an emerging construct, builds upon six aims of high-quality care—effectiveness, efficiency, timeliness, patient-centeredness, safety, and equity—as a departure point for this construct’s further exploration [3].

Traditional efforts to improve healthcare quality most often rely on measurements based on data supplied by clinical teams and administrative data, but patient reporting is increasingly collected as a valuable source of information [4]. Health-related patient-reported measures elicit either patient-reported outcomes (PROs) or patient-reported experiences (PREs). PROs are specific phenomena or underlying constructs of individual’s health status in defined populations, while PREs focus on patient care experiences encompassing whether something that should have occurred in a healthcare setting, has happened, or how often it has happened [5,6,7]. The scope of PROs and PREs relevant to diagnostic excellence is unknown, yet that knowledge could help refine, measure, and achieve such excellence [5]. Identifying diagnostically relevant PROs and PREs could foster research that increases the accuracy, validity, and timeliness of measurements of diagnostic care and provide researchers with tools to inform interventions that improve care and patients’ quality of life [6]. PROs and PREs have potential to both reflect and enrich the specification of the construct of diagnostic excellence as patient reporting brings the values, knowledge, context, actions, and power of patients and their care partners to the foreground [7, 8].

Given the need to understand patient-reported aspects in the emerging construct of diagnostic excellence, this study sought to conduct an exploratory analysis that would identify the needs for and inform the development of patient-reported measures for diagnostic excellence. Specifically, we aimed to: (1) describe the scope of diagnostically relevant PRO and PRE domains; and (2) review the state of measurement of these domains.

Methods

Our exploratory analysis was an iterative and multi-pronged exercise where the findings of a scoping review of the literature were subsequently enhanced with targeted reviews in response to expert input and application of a published set of patient journey snapshots to validate diagnostic relevance of the domains. The exploratory analysis also included interactive activities at three expert convening sessions (see Figure). We maintained a prospective focus on diagnostic excellence, where PRO and PRE assessments could guide the diagnostic process moving forward, as opposed to after a diagnosis is established. We also were focused on effectiveness and efficiency of diagnostic care on overall patients’ well-being and health-related quality of life; thus, any domain described as specific only to a symptom or condition was excluded.

Fig. 1
figure 1

Elements of the study’s explanatory analysis

Full size image

Scoping review

The research team conducted a scoping review of a bibliographic database using multiple electronic search strategies designed to encompass patient outcomes, experiences, reflections, perceptions, or perspectives on any aspects of patient journeys that include or could relate to the diagnostic process and its outcomes. We followed the PRISMA-Scry (Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews) checklist acknowledging the limitations of what this systematic approach to map evidence can offer and the need for expert inputs and iterative refinements [9, 10]. Due to a large amount of literature in this area (nearly 10,000 articles meeting inclusion criteria each year) and the level of resources available, we used only one database (EMBASE™ that covers all MEDLINE and additional articles [11]), select years of publication (past 1 to 5 years depending on a particular subtopic), and additional inclusion specifications (systematic reviews; explicitly diagnostic and screening contexts; caregiver and family focus; quality improvement or applicability contexts; descriptions of frameworks or stakeholder engagement). (See Supplemental Information for search terms used and search yields). The research team used similar searches for grey literature sources of national and international organizations concerned with patient-reported measures (see Supplemental Information). Subsequently, we used a backward snowballing approach (reviewing references in the earlier identified sources) and forward snowballing approach (reviewing publications that cited already identified sources) to find additional publications, including those prior to the year 2015 in cases where the earlier literature presented how the identified domain had been captured in a measure.

The research team shared and discussed literature findings with an internal expert group (8 researchers and 2 patient advocates in total) and at the expert convening (session #1). When additional PROs and PREs domains were emerging during these consultations, the research team conducted targeted searches to identify literature to supplement experts’ input. For example, the domain of accuracy of information was identified as initially missing, and the domain of feeling of being reassured was supplemented with examinations of false reassurance.

In the identified sources, the research team sought and extracted excerpts relevant to the contexts of diagnostic processes, procedures, and outcomes. Those individual excerpts formed a narrative bibliography. If the identified study did not already directly report measurement tools, we searched for other published examples of measures that assessed the identified domains. Our unit of analysis was an identified domain rather than a study. The research team did not intend to find all examples of measurement and was guided by saturation necessary for illustrative purposes to inform further deliberations.

Diagnostic journey snapshots

To validate domains’ diagnostic relevance, the research team drew from the 2020 National Quality Forum report’s twelve snapshots of patient journeys containing diagnostic errors [12]. We then charted the identified domains onto these snapshots, which also facilitated clarification or identification of missing domains (see Supplemental Information for two examples).

Expert convening

A total of 24 international experts participated in a series of five 2-hour virtual sessions in June 2021 (see Supplemental Information for expert bios). Participants were identified from author lists and other sources as having expertise related to either development of patient-reported measurement, analysis of diagnostic journeys, patient advocacy, patient safety interventions, communication, and equity and patient-centeredness of care. The experts further critiqued revised domains (session #2) and made final decisions on all identified domains (session #3); all sessions were facilitated by the research team that included a human-centered designer.

Results

Overview of diagnostically relevant PRO and PRE domains

The narrative bibliography amassed 41 diagnostically relevant and condition-agnostic PRO and PRE domains. While these domains were mostly described as patient-reported, we present these experiences and outcomes as those that can also be reported by care partners (e.g., family, friends, patient advocates, and others closely engaged in a patient’s care). Ten of these domains were classified as PRO, 28 as PRE, and 3 as mixed PRO/PRE. Based on identified examples of measurement practice, 27 domains are assessed immediately post-encounter with patients and care partners reflecting on a specific encounter(s) and 14 domains are assessed subsequently and cross-sectionally to reflect the entirety of the patient’s diagnostic journey and diagnostic care. Among these domains, 19 were captured in existing instruments (Table 1) and 20 in qualitative studies by eliciting free-text responses on questions corresponding to the domain concept (Table 2). Finally, one PRO domain and one PRE domain were conceptualized during this study, but the research team did not find examples of these domains being assessed (Table 3).

Table 1 Diagnostically relevant patient-reported outcome and experience domains captured by existing instruments
Full size table
Table 2 Diagnostically relevant patient-reported outcome and experience domains described qualitatively
Full size table
Table 3 Diagnostically relevant patient-reported outcome and experience domains that are conceptualized but not assessed
Full size table

Domains captured in existing instruments

Respect is a well-established PRE domain, wherein patients and care partners assess their interactions with the care team during the diagnostic encounter. This domain is, thus, captured directly and immediately post-encounter, for example, by asking how often the patient felt treated with respect and courtesy by a specific provider or providers overall [13]. Questions assessing respect are part of existing instruments such as the Consumer Assessment of Healthcare Providers and Systems (CAHPS®) family of patient experience surveys [14].

Satisfaction is another well-established PRE domain that is tailored to reflect directly on a diagnostic encounter rather than the overall diagnostic journey. For instance, questions could ask patients or care partners to rate their visit on a scale from 0 (worst possible) to 10 (best possible) or the likelihood that they would recommend a provider to family and friends using a scale from “Definitely No” to “Definitely Yes” [15]. Our experts noted that collecting patient satisfaction during various stages of the diagnostic journey can mitigate concerns about ratings affected by the nature of received diagnoses (e.g., lower satisfaction when receiving “bad news”).

Listening and taking concerns seriously, duration of communication, and emotional support at communication are three PRE domains assessing specific qualities of provider communication during the diagnostic encounter. These three domains are captured by directly asking patients to assess frequency [15] or by Likert-type scales [16] evaluating these aspects of communication.

Overall communication experience is an umbrella PRE domain assessed immediately post-encounter; instead of inquiring into all specific communication-focused domains, an instrument might ask questions regarding communication as a whole [17]. Our experts noted that using such an approach to assessing communication is not desirable.

The feeling of being reassured is a PRE domain that encompasses both appropriate and false reassurance. Appropriate reassurance is assessed immediately post-encounter (e.g., via The Reassurance Questionnaire [18] or a question in Vaillancourt and colleagues’ instrument [19]). False reassurance is conceptualized in qualitative studies [20] but is not yet incorporated, for example, as an option describing patient feelings or suspicions among possible responses in reassurance assessments. It was noted that the timing for assessing a false reassurance versus timing to assess the presence of appropriate reassurance after the visit might preclude incorporating these two aspects in the same instruments.

Understanding of a health concern is a PRE domain captured immediately and directly post-encounter. As an example, questions ask the level of agreement with the statement “I have a clear picture or understanding of my health concern” [19]. The PRE domain perception of a care plan for a patient to follow is assessed similarly, for instance, as the level of agreement with a statement “I know what to do if the issue got worse or came back” [19]. However, it was noted that identified examples of this domain’s assessment do not examine patients’ attitudes towards the care plan or its alignment with their preferences, for example, preference for tradeoffs related to diagnostic procedures (e.g., invasiveness level vs. testing accuracy).

Emotional distress is a PRO domain that asks patients to rate their level of distress on their diagnostic journey, specifically attributing that distress to the diagnostic encounter. It may be captured via the Distress Thermometer [21] or other screening tools [22], including assessing the change in distress between several consecutive encounters. This domain is related to patients experiencing diagnostic uncertainty, described below, as the distress might be one of the outcomes of such uncertainty. However, distress might not be attributed exclusively to uncertainty and can be attributed to the diagnostic process as a whole.

The change in symptoms and related perception of the symptom change trajectory are two PRO domains captured immediately post-encounter [19]. Aligned with our experts’ suggestion, some assessments [23] specify that change in symptoms might not be just toward improvement or worsening but can have an intermittent pattern. Assessment of the perception of the symptom change trajectory is done, for instance, by asking patients to align themselves with statements that describe their understanding and accommodating to their health status trajectory in the past 24 h [19]. Alternatively, it might be inquired if the symptom change trajectory is perceived as expected or unexpected versus the diagnostic explanation received at the encounter. As noted by our experts, the assessment of change in symptoms may be potentially done via instrumental activities of daily living (IADLs) or other patient-reported measures of functioning [24].

The three domains inaccurate diagnosis, untimely diagnosis, and failure to communicate diagnosis are directly informed by the definition of the diagnostic error [25]. The first two domains capture both the outcome (diagnostic process resulted in a misdiagnosis or delay) and the experience (the patient has experienced delay or misdiagnosis). Failure to communicate diagnosis is classified as a PRE domain, given definitional debates about whether the communication gap stands alone as an error outcome. The literature reports on assessing these 3 domains subsequently by asking patients to reflect on the entirety of their diagnostic care (e.g., during the past 12 months). Two questions from the Patient Reported Experiences and Outcomes of Safety in Primary Care (PREOS-PC) [26] ask patients the number of instances they experienced problems related to their diagnosis and its communication. A qualitative study [27], directly elicited patient reports of inaccurate diagnosis, untimely diagnosis, and failure to communicate diagnosis via free-text responses.

Three PRO domains related to harms, specifically type of harms, severity of harms, and harms impact on recovery and/or permanence, were often assessed together (e.g., type of harm described via severity or severity of harm described via impact on patients) [26, 28]. These domains were captured cross-sectionally and asked patients about harms perceived as resulting from healthcare interactions, without attempts to attribute harms to particular providers, visits, or institutions. Our experts noted that harms should encompass not only harms due to diagnostic errors but harms resulting from diagnostic procedures, including overdiagnosis harms, which to date has not been incorporated into assessments.

Patient disengagement with the health systems, either as complete disengagement, not seeking care, or when patients are disengaged partly, for example, via changing healthcare providers, is a PRO that is not captured directly in the literature. Two instruments on trust in one’s healthcare provider [29] or physician [30] assess the domain indirectly, including the intention (frequency of ideation) to change providers. Exemplars of assessment are done in cross-sectional manner where a patient reports their attitudes toward a healthcare provider overall or even by reflecting on the entire care team involved.

Domains described qualitatively

At the time of the analysis, we have not identified existing instruments capturing the remaining domains in diagnostic contexts. Table 2 provides excerpts of qualitative analyses or narrative descriptions of 20 such domains. Fourteen PRE domains reflect aspects of patient-provider communication assessed immediately post-encounter and are included in the overall communication experience domain described above. These domains are: empathy and caring; modality of communication (e.g., in person, written, or telehealth); mode of communication (e.g., presented visually or verbally); congruence with patient preference for autonomy in decision-making; including family in communication; sufficient communication; clear communication; contextualized communication (e.g., patient referred to additional information source); interpersonal adaptation, including cultural awareness and adaptation for health literacy level in communication; safe, supportive, and comfortable environment for communication; patient preference for who delivers diagnosis; mitigating language barriers; providers’ checking of patient understanding; and accuracy of information.

The overlaps between the content of some domain concepts are most evident in this group. For example, empathy and caring relate to the domains of respect and listening and taking concerns seriously [31]; congruence with patient preferences for autonomy in decision-making overlaps with inclusion of family in communication [32], but family roles have additional aspects beyond decision-making. Questions in existing instruments do assess PRE of providers’ ability to explain things in a way that patients could understand [15]. However, those questions do not differentiate clear communication from contextualized communication or from interpersonal adaptation in the diagnostic context and merge those domains when assessing. Other exemplars merge mode and modality of communication [33]. There are conceptual overlaps between, for example, sufficient communication (that implies sufficient understanding [34]) and understanding of a health concern or understanding of the diagnostic process. Similarly, the domain of clear communication [34] encompasses the resulting clarity of understanding of all medical information. Each of these examples and others can be argued as conceptually separate constructs in the diagnostic context, such as providers’ checking of patient understanding assesses whether this process occurred as part of the encounter [35] rather than the understanding itself. Accuracy of information has elements of patient-reported breakdowns [36], as patients report on accuracy of their medical records, as well as experiences being stigmatized and labeled by information that is inaccurate, outdated, or uses biased or judgmental language.

Aside from communication-focused domains, we identified six additional domains described qualitatively. Care partner or family involvement captures perceptions of how well healthcare systems facilitate care partner participation in diagnostic processes ranging from care partners feeling shut-out to equal partnership [32]. As noted by our experts, this domain includes family members and broadly defined care partners, including patient advocates and patient representatives, which should be elaborated when assessed.

Understanding of diagnostic process is a PRE domain for post-encounter assessment of patient’s understanding of the process [10] and, as noted by our experts, aligns with meaningful shared decision making in diagnosis and diagnostic co-production. Thus, the immediately post-encounter assessment of this domain was suggested to support and inform those efforts embracing a distributed paradigm for diagnostic work, invoking a shared mental model and collective intelligence [37].

Experience of diagnostic uncertainty is a PRE domain of emotions and feelings that patients might assess immediately post-encounter [31]. As note by our experts, experience of diagnostic uncertainty is inextricably connected with the patient’s own tolerance of uncertainty, and those can be assessed in tandem.

Disconnected diagnostic encounters across settings and providers is a PRE domain that allows patients and care partners to reflect on the coordination of care (or lack thereof) throughout the diagnostic process. One qualitative study formulated this diagnostic journey as a “pathway until you were diagnosed with cancer” and inquired if a dedicated healthcare professional had oversight and responsibility for that pathway [38]. As noted by our experts and literature, this domain might include and assess the connectedness and coordination of diagnostic process with treatment [39, 40].

Future attitudes and behaviors toward health systems and providers is a PRO domain that captures how the diagnostic journey affected patient confidence, faith, and trust in healthcare for future interactions. In contrast with disengagement with health systems, this domain focuses on the future, and can be illustrated with a patient’s quote: “I just don’t like doctors anymore” [27].

Potential harms and their potential severity is a PRO domain that assesses how a patient perceives harm that could have occurred but was prevented, including the magnitude of its potential impact on them. It was suggested that this domain might be captured similarly to the three PRO domains of harm while the possibility of harm can be stratified, for example, from “Possibly Yes” to “Definitely No” [41].

Domains that are conceptualized but not assessed

Awareness of pending diagnosis was conceptualized as a potential PRE domain. While the domains of understanding of the diagnostic process and perception of a care plan to follow might include patient awareness of specific tests that are needed, experts suggested this domain might stand on its own. As patients and care partners traverse from one setting, institution, and provider to another, health systems often must rely on patients to coordinate different diagnostic procedures and ensure continuity. In those cases, patients and care partners should be aware of such roles and specifically that the results of a diagnostic procedure are pending, and that awareness can be assessed and confirmed immediately post-encounter.

Self-advocacy affirmation was conceptualized by experts as a potential PRO domain to capture cross-sectionally patient confidence that they can advocate for their own health throughout their diagnostic journey. It was conceptualized that patients might become less involved in their health (diminished self-advocacy) or more involved, or that advocating might become so domineering that it harms the patient’s well-being. This domain construct focuses on self-advocacy itself, making it separate from future attitudes and behaviors toward health systems and providers, and any disengagement from the health systems.

Discussion

We identified 41 diagnostically relevant PRO and PRE domains, though this list is not intended to be exhaustive. Nor is it intended to be conceptually crisp at this stage of bringing together three streams of inquiry: an emerging construct of diagnostic excellence, diagnosis as a process and outcome that contributes to health-related quality of life, and international interest in patient-reported measurement as a tool for achieving diagnostic excellence. Our exploration at the domain level provides a foundational approach to advancing patient-reported measures for the emerging and important broad construct of diagnostic excellence.

A large number of domains presents a challenge for measurement and prioritization. Only 19 of the domains have been captured using existing instruments. More than half of the domains would need to progress from conceptualization and description in qualitative studies to more rigorous measurement. Assessing the 27 domains where patients reflect on a specific diagnostic encounter immediately and directly after that encounter would allow more immediate opportunities to act upon that reporting. The remaining 14 domains require patients to reflect on their experiences or outcomes over the entire diagnostic journey, cross-sectionally. Thus, these two groups of domains introduce different timing and setting junctures for assessment. Importantly, our findings demonstrate both substantive differences and overlaps among the domains that future measure developers would face.

In response to the challenges of such a large number of domains and relative recency of diagnostic excellence as a broadly defined construct [3], some of our experts suggested that some of the identified domains should be treated as definitely constituting patient-reported diagnostic excellence, while others as describing contributors to diagnostic excellence. However, most experts embraced the multi-domain nature of patient-reported diagnostic excellence, emphasizing, for instance, that communication-focused domains are not only contributors or risk factors to diagnostic excellence but equally constitute diagnostic excellence [44, 45]. Thus, the identified domains were seen as evolving building blocks for the emerging construct of diagnostic excellence that is being formed inductively versus deductively, especially from those emphasizing the patient perspective.

Some of our experts questioned the benefits of considering diagnostic concerns separately from other problems patients face, such as treatment concerns. Given the iterative nature of diagnosis even with empiric treatment phases, teasing out problems specific to the diagnostic process itself can be challenging. On the other hand, diagnosis has its uniqueness and emphasis; it is the foundation of the relationship a person will have with the health system [1, 12]. Additionally, the diagnostic process is often overlooked [42], and it is important to distinguish between diagnostic excellence and treatment excellence, especially given a framing of diagnostic excellence as condition-agnostic and prospectively focused.

This study also suggests care partner reporting options for identified domains following similar developments in broader patient reporting [43, 44]. Family members and other partners can help fill out patient reports in certain circumstances, such as when the patient is experiencing distress during encounters or when patients are dealing with homelessness or some types of trauma or institutional betrayal [45].

Future directions and implications

This study underscores the need to develop definitions for identified domains and create and validate item banks for domains yet unmeasured. Another possible future direction is to review domain classification as either PRO or PRE, given potential impact on measurement approaches (e.g., distress measured as an outcome of the diagnostic process versus as an experience at varying time points during the diagnostic process). Assessment of the domains and amassing data will allow establishing how domains relate to each other under the overarching construct of diagnostic excellence. This iteration between construct definition development and attempts to capture a construct is common. Recognizing the iterative nature of doing so at the domain level for an emerging broad construct will also lead to better understanding of how to prioritize domains for further assessment, and how to utilize them to explicate the overall meaning of diagnostic excellence. For example, efforts might be made to map the identified domains into the six aims used for the initial definition of diagnostic excellence [3]. All these directions will also highlight the important role of patient-reporting for improving all aspects of patients’ diagnostic experiences and outcomes and in assessing effectiveness and efficiency of diagnostic care on patients’ well-being.

Limitations

Our methods did not allow us to systematically identify all relevant domains or comprehensively supplement domains with many examples. Instead, we aimed to survey and assemble domains and associated measurement opportunities and challenges to facilitate future research. We might have missed existing instruments for domains, thus misclassifying them, for example, as captured only via qualitative studies. Our scope precludes providing answers to measure users on instruments to choose. However, given the relative novelty of diagnostic excellence, this exploratory analysis is expected to inform and guide measure developers on where to start. Additionally, this work did not include reports from patients on, for example, perceptions of contributors to diagnostic errors that are neither patient outcomes nor patient experiences. These patient-reported breakdowns [36] might inform diagnostic excellence but, arguably, not fully constitute the construct of patient-reported diagnostic excellence. As this construct and field matures, updates to this exploratory analysis will be needed to further systematize patient-centered assessment possibilities and realities.

Conclusions

As the construct of diagnostic excellence matures, at least 41 patient-reported domains can be identified and contribute to inductive development of this construct. As the existing domains are both substantively different and overlapping, further research is needed to continue to iterate on domain definitions and development of instruments to assess those domains. Capturing these domains via patients’ and care partners’ reporting allows maintaining diagnostic-sharp focus on care aspects that are condition-agnostic, prospective, and impact patients’ well-being.